 Our first speaker this afternoon is Dr. John Holland, Professor Emeritus of the University of California at San Diego, where he was in the Department of Biology and the Institute of Molecular Genetics. Dr. Holland is no stranger to Minnesota. After receiving his Ph.D. in microbiology from UCLA, Dr. Holland spent three years as both a postdoctoral fellow and assistant professor in the Department of Bacteriology and Immunology at the University of Minnesota. It was here that he began his viral research, studying the relationship between poliovirus and the cells that the virus could infect. Dr. Holland left Minnesota for a faculty position at the University of Washington, where he continued his viral research. He has assured me that it had nothing to do with the winters here in Minnesota. In 1964, Dr. Holland became the first chair of the Department of Molecular Biology and Biochemistry at the brand new University of California at Irvine campus. And in 1968, he moved to University of California, San Diego, where he retired from in 1996. Over the course of his career, Dr. Holland and his colleagues have investigated viruses from the perspectives of their biology, their biochemistry, molecular biology, immunology, and ultimately their evolution. They demonstrated the role of cell surface receptors in the susceptibility of human and other animal cells to infection by the intact viruses, as well as the susceptibility of some receptor-less cells to infection by either naked DNA or RNA. For more than 20 years, Dr. Holland's work has focused on the evolution of the RNA viruses, whose genes accumulate mutations at rates more than a million-fold faster than the DNA-based genes of the host cells, such as our own. These studies have had important implications for the treatment of viral diseases such as HIV infection, viral epidemics, and the emergence of the so-called new and often deadly viruses in the world today. Dr. Holland is a prodigious author and the recipient of many honors, including the Eli Lilly Award for Research in Microbiology. It's a great pleasure for me to welcome him this afternoon as he speaks on virus evolution implications for diseases. Thank you very much, Colleen, and thank you all for inviting me. It's a great pleasure to be here, and as Colleen indicated, it's also a homecoming. It was 41 years ago last summer that I came to Minnesota to learn virology. And I was fortunate enough to study in one of the world's greatest centers of virology at that time. It was a new field, and there weren't very many people in the field. Jerry Siverton, who died tragically young at age 51, was my mentor here. I learned virology very well from him and from Lee McLaren, another very fine scientist. So this is a nice homecoming, and thank you again. I'd also like to thank the excellent student musicians and their director for the entertainment that we're getting. It is an often that I get to give a talk and be entertained at the same time. Thank you. What I'll be telling you about today is evolution. And I might as well warn you, those who don't like evolution, I doubt that there are many of you. But evolution to me is not a theory. Evolution is a fact. It's something that we study every day. Evolution occurs with viruses as with other living things. And the difference is that with the RNA viruses, and I'll be telling you mainly about the RNA viruses, with these RNA viruses, the error rate of replication is so high that they mutate and have the ability to evolve. Now, they do not necessarily evolve, but they have the ability to evolve and they mutate at rates approximately one million times the rates of higher organisms. So we host to the viruses are evolving a million times slower than the RNA viruses. Now, what that means is that one year of study in our laboratory of virus evolution is equivalent to a million years of study of human or animal evolution. This gives us a tremendous opportunity to speed up evolution, to look at naturally speeded up evolution and see how it works. So my talk isn't so much about viruses as it is about evolution generally. And you might question whether viruses are good models being the simplest living things. Let me assure you they are good models. Let me also apologize because some of the things I'm doing, I refuse to speak down to students. I never have done so when I speak. I'm going to give you some difficult concepts. I believe that I hope, I pray that I'll be able to present it to you in an understandable manner, but I may fail to do so at times. So bear with me. I want to show you the insights we have into evolution. So I'm going to go in to sequence space and some other difficult concepts. Nothing that you can't deal with, but it requires that you listen carefully and think hard, okay? And that's something I'm not too good at myself. So let me start with the first slide to give you an idea of the kind of material that we work with. Can you put that first slide on please? I'm having trouble activating. Okay. Now you see here what we call a plaque assay. The viruses that we work with, these RNA viruses, have very high mutation rates as I've already mentioned. And we are able to isolate them and to study them overnight. That's one of the keys to being able to do the kind of work I'm going to tell you about. Overnight, we can count the number of infectious virus particles we have, by a technique that Bill Yawklick told you about this morning, that Renato Del Beco worked out many years ago, decades ago. This is called a plaque assay. If you look at the blue background on these cells, what you have are human cells in this case attached to the glass, one layer thick. A one layer thick layer of human cells is called a monolayer. And approximately two to three times 10 to the six, two to three million cells are stuck to that glass bottle. Now what we do is attach virus to them. And when we do that, we get plaques. We use a semi-solid medium to feed the cells while the virus is growing overnight. So in about 17 to 18 hours, we can come back and see how much virus we have. That's called a plaque assay. Now you'll notice there are two sizes of plaques. The plaque arose from a single virus particle hitting one cell. That one cell releases about 10,000, again, 10,000 infectious virus particles. And many more thousands, probably about 20 to 30,000 dead particles because of the high mutation rate. There's a price to pay for high mutation rate and high adaptability. And as I tell you about this rapid evolution, you must keep in mind, the virus is paying a terrible price, a very high non-infectious particle ratio. You can see another thing about this. The single virus particle has grown out to the side and involved hundreds of cells. Then when we stay in the living cells, you can see where the dead cells are where the virus grew. The first cell infected released 10,000 virus particles and they spread to adjacent cells to form this plaque. Again, a single infectious particle produced this plaque. Now if we reach in with a pipette and pull out the virus that's in that plaque, we find nearly a million particles usually depending on plaque side. We can now use this to produce mass cultures of virus. We put it into bottles like that, excuse me. We'll put it into bottles like that or larger bottles. And we can then produce an amazing amount of virus. Let me give you a feeling for that. We get 10 to the 10 infectious virus particles per milliliter of fluid. In a large bottle, we get hundreds of billions to nearly a trillion virus particles depending upon the size of the bottle. So one virus plaque in a period of two days or less will give us countless virus particles so we can study their evolution. But I do want to emphasize that that one virus plaque came from one single particle and we call that cloning. Now you know about molecular cloning where you put genes into viruses and other vectors to transfer the gene. This is not that. When I say cloning in this talk, I'll be talking about picking the progeny. That is the offspring of a single virus particle and we do it by doing what you see here, making a plaque and then picking the plaque. Now the next thing I want to talk about very briefly, I want to emphasize that there are two major kinds of viruses. There are the RNA viruses and the DNA viruses. The vast majority of viruses are RNA viruses. But the DNA viruses which are smaller in number are very successful parasites. Now they have a different life strategy. The life strategy of the DNA viruses is to find optimal sequences in their DNA genome. The genome is the collection of genes in a virus particle. They find an optimal sequence for successful strategy of survival and reproduction. The strategy of the RNA virus is because they have a higher mutation rate in general is to overwhelm hosts with a swarm of mutants and hope that some of those mutants will be very successful. And that's how they not only infect us, but that's how they evolve. They can undergo antigenic changes as occurs with influenza virus and HIV and many other viruses. They can undergo changes which render them resistant to drugs and so on. So the strategy is different. The DNA viruses, let me give you an example of a DNA virus infection. We can take varicella zoster which causes chickenpox. Early in life you become infected with varicella zoster and you get a generalized infection throughout the body. The immune system drives that down. And then the virus hides in certain neurons in the body. Later on, often in old age, 80 years later, if an older person loses a spouse and undergoes severe stress due to psychological stress say of losing a spouse, they will get shingles. Shingles is a reactivation of that childhood infection. And shingles then can lead to new outbreaks in grandchildren or anyone else coming in contact with this shingles patient who is anyone else who is nonimmune and they get chickenpox again. Now you see that strategy as one of persisting for a long period of time without causing disease. And it causes disease only as a mechanism for being able to spread to additional susceptibles. Now the RNA viruses tend not to do that so much. They tend to be explosive to cause overwhelming quick infection and either get driven out by the immune system or in a very few cases persisting. They're not nearly as good in general as the DNA virus are for persisting in the body for long periods. Hepatitis C is an exception to that. So there is a real difference and I'll be telling you only about these rapidly evolving RNA viruses. Why do they evolve so rapidly? The reason they evolve so rapidly is both quantum mechanical and due to thermodynamics. And it has to do with the fact that no machine is perfect, that the replicases are polymerases which copy one strand of the RNA into an opposite strand and then back to the first strand. These are imperfect machines that make mistakes. They make mistakes at about one in 10,000 bases inserted into this genetic code, into this long string which represents a genetic code. When they make those mistakes, we have a mutation. One in 10,000 is the number we've derived. I'm not gonna bore you with any data on mutation rates and mutation frequencies. You'll have to take my word for it. I'll show the slide from another laboratory and show you how this works in a minute. Again, I've already implied this, but I wanna point it out again because it's critical to my talk. RNA virus rates of mutation and evolution can exceed those of their human and animal hosts by millions fold. And therefore, evolution runs a million times faster. That's what allows us to do the studies I'll be showing you. Now this is not from my laboratory. I decided to choose an example from somebody else's laboratory. This is Roland Rookert's laboratory at the University of Wisconsin. This is a common cold virus, a rhino virus. And they took this rhino virus and they developed drugs they in collaboration with drug companies. These are wind drugs, they're called. And when they add the drug in increasing concentrations, the virus yield goes down. I'm not gonna bother with the numbers here, it's not important. What do you see as you add more and more drug coming across here on the abscissa, as you add more and more drug, the yield of virus goes down on that monolayer of cells. Until finally you reach a plateau. Now that plateau drops off later due to killing of cells. But that plateau is a very definite plateau and it occurs at exactly one in 10,000 of those virus particles being a mutant which is resistant to drug. Now this happens with HIV, it happens with a wide variety of other RNA viruses. The remarkable thing is that mutations of almost any kind that we tested and others who have carefully tested, they all show up at about that rate, one in 10,000. But this virus only has approximately 11,000 nucleotides, 11,000 coding signal bases. So just about every virus that's produced is a mutant. Now when we first proposed this, when we first saw this, people said, I should mention that we collaborated with a very good person in Spain and his laboratory, Dr. Esteban Domingo, a Spanish investigator in Madrid. And we've been collaborating for over 12 years. Both he and I independently observed his laboratory in mind that the mutation rate was this high that almost every particle produced is a mutant. This has enormous consequences. We didn't believe it at first and people didn't believe us at first. They said, you're crazy, Holland, you're absolutely crazy, this can't be right. Well, as it turned out, they were right and we're right. We are crazy, but the viruses do mutate at this rate. Okay? So, you probably think I'm kidding. So, I am crazy. I've busted my body up on motorcycles and nearly killed myself in airplanes. I'm crazy, but that's beside the point. The point here is that these viruses mutate at an enormous rate and give us an enormous opportunity to study evolution. Now, the next thing I wanna do is move into what happens. If you take a single virus particle, and we could call that a wild type, and genetics, wild type has a definitive meaning. It means that something is what's found in nature and so we call it the wild type. Then we get mutants and we say these are non-wild type mutants. But this really doesn't apply very well to RNA viruses because they're all different. So, there is no real wild type. This is a difficult concept to get across. If you take a so-called wild type quote unquote and you pick a plaque from that so you have a plaque of wild type, as soon as you grow that plaque up, you have a mutant swarm, a swarm of mutants related to each other, usually by one or two mutations or in some cases no mutations related to the original plaque. But you cannot produce a single thing. If we were to try to plot in sequence space, and I'll get back to that later, but if we were to plot in three dimensions, what happens? If you take any kind of a living thing and clone it, take a single particle and clone it, you're supposed to get something identical or nearly identical. Now, in fact, you don't do that even with DNA organisms, but I don't have time to go into that. The genomic mutation rate, as John Drake showed, is very close to one. But when you clone these viruses, what you get is a quasi-species swarm, and this quasi-species swarm can be represented instead of a single point dot representing that sequence in sequence space, in three-dimensional sequence space, what you get is a swarm like this, a swarm of mutant. I'd like you to keep this in mind that every time we pick a clone and grow it up, we don't get the same thing that we've picked. We get something that's a swarm of mutants related to that clone. This has enormous implications for virus disease and for the nature of disease and for the nature of epidemics, and I'll try to give you a little insight into that. It's very difficult. I hope I won't fail. Now, the advantage of this swarm is the following. If the original virus clone is indicated in blue there, and you put this clone into a selective environment, which is very slightly different than your original environment, some of those mutants, which are often a periphery, out here in this particular case, will give rise to a new mutant swarm. And I can tell you right now that I'm a Darwinian. I believe that Charles Darwin was a genius who was absolutely right, and we see every day in the laboratory that he was right. We see positive selection, negative selection, and a lot of neutral drift and neutral mutation, but in general, the most important kinds of mutants are those very rare mutants which confer an advantage. And the way this happens is on the periphery. If you just produce something in the middle of this, it'll be very much like the other mutant swarm, but if you change the environment and try to select something new, it comes from the periphery. So in a quasi-species mutant swarm, and we call these quasi-species because they're not a single thing, it's the periphery that gives you the ability to jump ahead in evolution. This is a very difficult and important concept to grasp right now to understand the rest of my talk. It isn't the fact that you have a lot of mutants, it's that one of those mutants is likely to be useful if the environment should change. Now, the other thing I wanna emphasize, I'm not gonna give you the background for this experiment. This was two years of work, and this was done by Patrick O'Hara and Stu Nickel, and Patricia Fultz, and a number of other really fine people in my lab, Frank Horadiski. And what we did was select for mutants that would escape from the original virus replicase, the enzyme that replicates the genome, and we sequenced the end of the genome without going into detail. As you continue to select for mutants at the end of the genome, you can see as you move down in time with more and more passages in cell culture where you're doing this selection, you get swarms of mutants. But note what happens. The replicase attaches to this portion of the genome in order to replicate it, and it's only in that portion where you've carried out selection that there is any change. When I tell you to produce a mutant swarm, the first thing you think is that the whole swarm is selected, and that's a mistake. What actually happens is we selected for change in that limited area, and that is the only area where the whole swarm changed. This is very counterintuitive. If you put strong selection on, you will get changes there. You'd expect a lot of junk changes elsewhere, but out for hundreds of nucleotides, there were no changes at all. Darwin was right. Selection of the fittest does not necessarily bring along a lot of junk. With those high mutation rates, you'd think it would come along, but we've seen this repeated. I want to give you one disease example. I don't have a lot of time to go into the disease examples here. I'll try to do a little bit of it because I know people are interested in virus disease. This is Nancy Beck and her colleagues, and what they showed very briefly, I have to be very brief here because I'll run out of time, what they showed is a very simple thing. There are coxsacuviruses, which cause human heart disease called myocarditis. They cause, that means simply that they infect the heart muscle, the myocardium, and cause inflammation and sometimes serious damage to the heart and even death. Now this occurs very, at a very low probability when you're infected with this particular coxsacu virus called coxsacu B3. Coxsacu is named after a ton in New York. Big name, small ton. I, coxsacu B3 occasionally then causes human heart disease and they set up a model system in mice. They injected lots of mice with the wild type, so-called wild type mutant swarm, and nothing happened. Occasionally when they injected a lot of mice, some of the mice got serious heart disease. Now here's the critical point about quasi-species. When they take out the virus from the heart, deliberately select the virus from the heart and put it into other mice, the mice regularly come down with myocarditis. They found six base changes in the entire coding sequence of the virus and only one of them, only one base and one amino acid change was critical for this disease. Now in order to get the original virus in the mice, they had to put in selenium deficiency. Selenium as you know is an antioxidant vitamin-like supplement and mice and humans, it turns out humans in south of China have selenium deficiency. They have a lot of problem with coxsacky myocarditis. So they got the original change in the mutant swarm, but now when they take out that mutant swarm with only one critical change plus regularly five other changes in the bases, it regularly causes myocarditis. Once again, we go back to the mutant swarm. It's not there. I'll have to go backwards just a minute. We have to go back to this mutant swarm. What they did again is what I've shown you earlier. They selected for mutants, not so far out on the periphery, only six base changes out. Then they got a new mutant swarm, very slightly different than the other mutant swarm, and this was a deadly virus causing myocardium infection in the mice. Is that clear? This is a very strange kind of result. Every time they did this, I think five or six or seven times, every time they did it, the same base changes occurred. What you select for in evolution is what you get. If you don't select for it, you'll have all these changes around there and they can represent junk until you put strong selection on them. Now I want to talk about evolutionary fitness. When you think about fitness in general, very often young people in particular, not old decrepit people like myself, but young people will think about fitness as being physical fitness, but I'm going to talk now about evolutionary fitness and evolutionary fitness is reproductive fitness. The important thing isn't how fit you are physically yourself, but how many progeny, how many offspring you produce because evolution doesn't care about the individual survival, they care about the lineage, what survives into the future. So when I talk about fitness now on, I'll be talking about reproductive or evolutionary fitness. And evolutionary fitness is defined and measured as the relative reproductive capacity in a very defined environment. As I told you earlier, change the environment, you change the selected forces and you change the fitness. This is a theoretical curve. We've decided that the best way to present this information is as fitness vectors. Now very briefly, the way we plot this, we put viruses, we have a type of virus which we consider wild type and it represents an internal standard in a mixture. We passage the viruses in competition with genetically marked viruses and if they stay at the same ratio that we start at, they'll come straight across, okay? These aren't vectors where the length of the arrow means anything. We've made them so that coming straight across is neutral mutant that has the same fitness and competes equally with the wild type. If it goes up, the fitness is a high fitness and the higher it goes, the higher the fitness. This is a thousand times higher fitness here and this is only 1.2 times. Now in just 12 passages in cell culture, you can get something like 1.2 fitness which is obviously overgrowing the other competing mutant or you can have a thousand fold which is a very high fitness arrow. The same with decrease in fitness. You get down to one one thousandth of the fitness or 0.9 of the fitness. So when I show you these arrows, up is increase in fitness, down is low fitness, straight across to the right is equal fitness or neutrality. First thing we did is passage the virus in the same old cell we've been passaging in it and if we pass it a long time in BHK, the fitness will go up as I'll show you later but in this particular experiment, it stayed the same. Now we passed it in a new kind of cell line 10 times. This is passage 10 and only 10 passages over 10 days. We got an enormous fitness increase of hundreds fold on a dog kidney cell line. In a human cell line, a human tumour cell line, the same thing happened, heedless cells and in mouse macrophages again. What you select for is what you get. We selected for the ability to grow in different cells and you can see the fitness went up enormously. What does that mean again? That means they out reproduced the competing virus after they were trained to grow on a certain cell. Now we took a very low fitness virus. We have clones that are low fitness and clones that are high fitness and we passaged it 60 times on BHK cells. That's a hamster kidney cell line and after only 10 times, we didn't see much change but in this particular case, it was a virus that was not as well adapted to the cell line and we got after 61 passages an enormous fitness so that after 61 passages, this is P61, we have an enormous fitness increase from the low fitness clone we started for. Evolution and attestive. There's no question what's happening here. There's Darwinian selection to increase fitness on this cell line. Now what happens if we change the environment? This was that high fitness clone we selected with 61 passages. What if we put it in a new environment? It's not nice to do but we injected this into mice and intracerebrally in the mouse brain it has no fitness at all. It's extremely low fitness. Very high fitness in the cells and culture, essentially terrible fitness in the mouse. What happens if we pass it in the mouse? Same clone, only four passages in the mouse brain, enormous fitness in the mouse brain but now the fitness in the BHK cell line has gone down. What you select for is exactly what you get. These quasi-species swarms are trouble waiting to happen you have all these mutants in a big swarm, change the environment and one of those or some of those or many of those will be fit in a new environment and your whole quasi-species swarm will shift in sequence space. I'm gonna go through this very quickly. In 1964, a man named Herman Muller proposed that where you have asexual organisms and our viruses don't carry out sexual recombination where they're asexual and they have a very high mutation rate then when you go through genetic bottlenecks such as one individual or several individuals the fitness of the organism should ratchet down. This is a famous proposal by Muller but people tried to study it in our organisms where everything goes slow. But Lin Xia, who's at the University of Maryland and is moving over to the University of California at San Diego in a few months. Lin Xia in a bacterial virus showed that Muller's ratchet actually operates and we jumped on his discovery. What I'll be showing you here is just jumping on Lin's work and showing that he's right in an animal virus. Doing the same thing he did and all we did was reduce the virus at each passage to a single particle then pick a plaque, dilute it out so it form more plaques, pick it again, do that 20 times in a row and what happened is high fitness went down to low fitness after 20 passages and all of the clowns and neutral virus generally went down a couple of them went up. This is a stochastic or probabilistic process because mutation is probabilistic. You don't know where mutations will occur. You don't know which mutations you will get is probabilistic. And so Muller's ratchet doesn't operate. I'll show you in a second here. If you actually follow at each passage what happens in this particular experiment in Muller's ratchet we actually gain fitness through passage six, 10, went back to neutrality at 13, passage 14 genetic bottleneck. You're going through one particle at each passage. It dropped in fitness by passage, what is that, 16? It was down very low, 18 it was even lower. So it's not like a regular ratchet that ratchets down as Muller proposed but being probabilistic it tends to go up and down with the overall effect being deleterious. Now this is not surprising. Whenever you have a mutation the highest probability is that you will either get a neutral mutant or you will get a deleterious mutant or with quite a high probability you'll get death of the virus or death of the organism. This is because you can't control a mutation where you have your mutational hits. If it hits a lethal site which destroys a functional protein the virus is dead. So we don't even study the dead viruses but we can see the sickly ones with low fitness. Now if this happens and viruses are often transmitted when you cough and cough and influence a virus into the air studies at the U.S. Army germ warfare research unit back in the 60s showed that very often there's only one particle on the amount of air you'll breathe in a crowd of barracks with other troops that are spreading the virus. So very often bottlenecks happen. A single particle will be inhaled and cause flu or adenovirus infection whatever it was they were studying and that will tend to cause virus fitness to go down. So what will bring virus fitness up? As was implied in Muller's original suggestion high population passages will do that. We took some extremely unfit virus very very low fitness virus and passage it instead of plaque to plaque where we're going down to a single virus particle we transferred roughly a million virus particles from cell culture to cell culture to cell culture bottle to bottle to bottle and every time in a regular manner they didn't always go to the same endpoint but the virus population fitness gained. Now this alone tells you that evolution is working according to Darwinian positive selection. Why does this happen? The bottleneck you have a probability high probability of getting a sickly virus. When you raise the population size at each passage you pick the best of the best of the best of the best ad nauseam until you end up with a very high fitness virus. So viruses which tend to spread to you in large numbers will tend to be more damaging and tend to take a better foothold against the immune system and so on than if you get a bottleneck. So next time you catch influenza and cold weather breathe in one virus particle. You'll probably get no disease you'll probably get a mild infection. What you don't want to do is come up to somebody who is coughing with the flu. They cough into their hand, they shake your hand. They give you a hand full of virus. There can be tens of thousands or hundreds of thousands of virus particles in there which you immediately rub into your eye and it gets down into the respiratory tract. So a genetic bottleneck as opposed to a population passage gives quite a different outcome. I'm not going to go through all this data. There's not much time. Let me tell you about one rather interesting thing. David Clark and Elizabeth Dwart did this work. There's Dan Dalen a long time ago, 20 years ago suggested a red queen hypothesis. The red queen hypothesis is based on an Alice in Wonderland sentence in which the red queen said to Alice, around here it takes all the running you can do just to stay in one place. And what this hypothesis states is that either a host and its parasite are in a constant genetic race changing all the time trying to keep ahead of the other or the two populations feeding off the same resources are racing to use those resources. We decided to test this hypothesis because again we have this million fold increased evolution. And it's true. We took virus particles that had, we had previously measured had identical fitness and we just let them compete for a long time and either one or the other would drop out of the race. They kept running and running and finally one went up in fitness and the other disappeared or the other went up in fitness and the other disappeared, then the others go down. Now when we tested the fitness of the winners, naturally the fitness of the winners was up as you might expect. They were mutating and getting better and better to compete with the losers. But they're rather unexpected thing, not in retrospect but unexpected at the time. Right before the losers died out we got some of them and tested their fitness and it had gone up too. Around here it takes all the running you can do just to stay in the same place even if you're gonna lose, okay? This is another example. We had matched five passage population. We looked at the size that's required for genetic bottleneck and without going into detail it depends on the fitness you start with. It's between five and 50 clones, five and 50 virus particles. They were, you see, they were all competing equally and suddenly one of them jumped up in fitness but high fitness tends to be unstable when they drop back down. And one of the strange things that I always wondered about when I was a young viral is just 175 years ago is why is it that only RNA viruses are able to grow both in insects, insect vectors, and in plants or in insects and in humans. Many viruses are so-called arthropod-borne viruses and we wanted to see what happens to fitness when we take one of these vesiculostomatitis virus and pass it for a long time in sand fly cells and adapt it to the sand fly cells without letting it see human or animal cells. And what happens is rather remarkable. After 10 months in cell culture at low temperature this sand fly adapted virus, these are logarithmic scales here, it had gone up tens of thousands of fold in fitness on sand fly cells but when we tested it in BHK hamster cells or directly in the mouse brain it had no fitness at all. It didn't make mice sick, it should have and it didn't really grow well in cell culture at all. The difference between its depth, the difference between its fitness on the sand fly cells which are vectors, arthropod vectors and in mouse brain or cells in culture, the difference was two million fold. Enormous difference in the ability to grow after only 10 months of single type adaptation at low temperature. A lot of this is due to temperature because we have a high body temperature. Now what happens if we take that virus, it's control now that has very low fitness and pass it one time on BHK cells or seven times on BHK cells. It regains almost all of its fitness. This is what quasi-species swarms are all about, rapid adaptation. One passage overnight, 18 hours and we get back virus that grows quite well in the animal cell. When we put that into mouse brain, it killed mice again. So it regained virulence as well as regaining the ability to grow in culture and in mouse brain. Quick adaptation due to swarms. Now here comes the hard part. I apologize in advance, let me do my best. A long time ago, back in the 30s, it's all right. Came up with a mathematical paradigm for evolution. And it basically reduced to a very simple but not absolutely correct visual demonstration. When things are adapting, he viewed them as climbing fitness peaks. In other words, when they're gaining fitness, they're climbing an adaptive peak out of an adaptive valley. And so we'll visualize that and we'll talk about this very quickly, much too quickly because I don't have that much time. Move ahead here and look first of all as an example of what he's saying in two dimensions. Two dimensions is nowhere near correct. If we view adaptive peaks, that when the virus has moved through sequence space, that is mutated so that its sequences will fit it on this peak here, it has enormous fitness. When it's down on this peak, it has very low fitness or rather low fitness. When it's down in this valley or plateau, it's essentially dead. And so the viruses tend to get on to fitness peaks or any living thing according to Sewell Wright's theory gets on a fitness peak and it wants to stay there. It has a reluctance to get off even a low peak and find an even better fitness on a different peak by mutation. So it becomes improbable to move from peak to peak at times depending on how high the fitness is and how broad the quasi-species is, how big is this cloud or swarm of mutants? I don't know if you can visualize that. This is two dimensions is not right. Now let's move to the hard stuff. This is Manfred Eigen. Manfred Eigen had come up before we started our work, he had come up with a paradigm, mathematical paradigm, which fits our work beautifully. Manfred Eigen is in my book, a real genius like Charles Darwin and he has insights into evolution that most of us aren't capable of. And what he came up with, he and Christophe Bebriker and several other of their colleagues, came up with sequence space in which V is the length of the genome in basis. In other words, in the case of the virus we're studying it's 11,000 bases long. It's a V dimensional hypercube with all the possible permutations and combinations of sequences with genome length V. I want to emphasize V is not a constant. Viruses can change their genome length so this is almost infinite. Now to show you how quickly sequence space builds up, if you have coding units, two coding units just in a binary sequence like you have in a computer, these are the options, four options. You go to three it gets greater, you go to four it gets greater and when you get to six you can see it's becoming pretty enormous. The important thing here though is as big as this is you can get from one sequence to another by V jumps. Now how do you know how to make the right jumps? If you want to get from up here to down there or from over here to over here or from here to here, how do you do it? You have to be guided because sequence space is V dimensional. You have four to the 11,000 possibilities. That's easy to say but it is beyond comprehension, beyond all comprehension that kind of space. The number of possibilities that have to be tested to find all viable viruses, to find all viruses that might break out as new virus outbreaks in the future are beyond the size of the universe, a thousand universes, a trillion universes. You can't begin to comprehend how much four to the 11,000 is. There are approximately 10 to the 80 atoms in the universe and we're talking about four to the 11,000. Now, sequence space is vast beyond imagination but only minuscule fractions can be explored in the spacetime of our universe. The pathways that are explored by evolution are probabilistic, stochastic, and totally unpredictable. People are always asking me when is it likely that the next new virus will emerge or what virus will emerge? Totally unpredictable because of the vastness of sequence space. We don't know what pathways will be explored but in our terms and human lifetime terms they are almost infinite, nothing's infinite but the idea of four to the 11,000 staggers even the imagination. We can't even think of four dimensional space time well let alone four to the 11,000. Now that brings up another thing which I've already implied, the subject of selection is the whole swarm not the individual variants. The reason is again that it's the periphery of the swarm which gives you your big jumps toward where you wanna go, up a fitness peak. Once again we're back to this, you see that you can jump from a swarm that's over here to a swarm that's over here in sequence space simply by changing the environment. Now the other thing is robustness in sequence space and that means the following. If we look at this area down here where we have fitness peaks we have a number of fitness peaks close together and all of them are pretty high fitness. Now if we have virus on this marked peak right here it can move around very easily to other peaks over here as you can see relatively easily but how can it get over to this high fitness peak? It has to go through non-adaptive valleys to get over there. Now this is not real space we're looking here at three dimensions and what you have to understand is this is 10,000, 11,000, four to the 11,000 hopeless number of dimension. What guides it along? One of the things we found is that the most feeble and unfit virus clones always get back to high fitness. They can always climb a high fitness peak if we give them enough opportunity to select. We have deliberately tried to get the sickest virus clone mutants we can find. They always can move up to become robust viruses. So much of the sequence space that is actually occupied by viruses is probably high fitness space which we call robust. Kaufman refers to this as robust which means you can sustain a lot of mutations without getting away from pretty decent fitness. Most of sequence space happens to be dead. It has to be dead because almost all of it would be junk information. This just shows viruses climbing in fitness space and we actually plotted the peaks around we took a clone and then we took sub clones of that clone to see what their fitness was and we did this 98 times. You can see this is what the peaks actually look like. That's actual data, not theoretical. All right, most sequence space is dead. Most viable space is non-adaptive. That means it's not really gonna work in nature. However, the viable adaptive regions are countless beyond imagination, exploration or explanation. I can't explain to you how vast these areas must be which are viable and adaptive. The reason again has to do with large numbers. It's easy to say 10 to the 80. It's not easy to imagine that being all the atoms in the universe. It's easy to say four to the 11,000. It's hopeless to try to imagine how much adaptive space there must be. So these swarms of viruses are moving through sequence space unpredictably, climbing peaks in robust space and moving we know not where. This, looking at our data, this is the kind of thing that we think it's really like in these robust areas. We think there are a lot of pretty much neutral. We think the neutrality is defined by robustness of the space. That means there are lots of close peaks together. And we can get viruses up these very high fitness peaks but they tend to drop right back because every mutation they throw when they're very high fitness moves them back down into the less adaptive peak. All right, finally, with regard to disease. If we view a quasi species swarm as being somewhere in fitness space, now fitness space because it's multi-dimensional isn't even anything like this. It's not a symmetrical cloud. As the virus grows it will throw a symmetrical cloud of mutants, most of which are dead or non-adaptive they can't live in the real world. But as they become selected, you narrow that down and you move it into very adaptive areas of sequence space and the virus has to crawl through these. If we view a quasi species swarm as being the size of that ruby laser beam, very tiny, it will crawl through sequence space in unpredictable manners. I show one possible route here to get to very adaptive space. But there are countless, absolutely unimaginably countless other routes that it could take. This is why evolution is not predictable because of this vastness and unpredictability of pathway. Now if this happens to be an influenza virus which is able to grow in humans after recombination and mutation, then when it gets in there it will be fairly robust because this whole area is filled with pretty good adaptive peak. But what about a new virus? What about something that's never existed before? How likely are they to arise? Very, very unlikely but they could be anywhere in this labyrinth and space, we don't know. Whether we ever hit them or not is probabilistic. How about one that's fairly removed? Let's say this is very robust sequence space but out here if we got a virus that happened to have the right sequences to occupy that part of space, we would have a lethal epidemic. Look at the distance of non-adaptive or dead space between these two. You can't mutate and crawl out there. You might be able to go over here and produce a really big mutant swarm that'll jump over there. But most likely you have to carry out sex, recombination to bring together pieces of DNA from different places and reassemble them in order to make such a big sequence jump. So one of the questions in the brochure that was handed out here is why the mystery, why do so many viruses die out? Why does a virus like the avian virus which broke out in Hong Kong last Christmas, it got started in humans, why didn't it go human to human? The reason is there are two damn, excuse me, two damn many ways that it has to change to get into the right area of sequence space to be able to adapt well to humans and spread by respiratory spread. Let me give you an example. Dr. Crosby will be talking about this in some detail. In 1889, there was an influenza virus epidemic and we now know that one of the key proteins called the hemagglutinin or H protein was what we call the H2 type. In 1900, there was a new epidemic, the H3 type. 1918, that was the H1 type which caused this horrible Spanish flu that Dr. Crosby will be telling you about. Then that stayed around until 1957 and it was supplanted by the Asian flu which was H2 coming back again. Then the Hong Kong flu came back in 1968. That was H3 coming back again. Now that's what's circulating today along with H1 which reappeared in 1977 after a hiatus of many years. These things come out of nowhere. They're in the bird population, a wild bird population all the time. We know they're closely related to the past strains. They do come back into the human population but we don't know when, we don't know through what pathways. The reason that Hong Kong flu, the one that's last Hong Kong flu last winter, reason it didn't get going in humans, it takes too many mutations or maybe reassortment of genes to pick up a new gene. Influenza virus A has eight different genes that can be reassorted. So we can't predict what will happen. We only know the following. New viruses, new and parenthesis, all new viruses come from old viruses. There are no new viruses. They have to have parts of old viruses, parts of cellular genes very often and RNA viruses can pick up cellular genes. But new viruses and new epidemics are not predictable in advance but they are absolutely inevitable. They will occur. Their severity, their nature, the sequence of the virus. We can never predict this. People question me on that but they don't understand sequence space when they question it. It is too vast for the greatest minds in the world to even comprehend what it's like and let alone predict the pathways through it. There are a lot of human factors I'm not gonna talk about. One of them obviously is human density, jet travel, rapid transportation, commerce and so on. I'm gonna end my talk here. Just emphasize once again that evolution does occur. We observe it every day on the test tube. It's beautiful, it's marvelous and it's complex. The reason it's complex is because of the complexities of the alternate path. Thank you very much and thank you for inviting me. Good to meet you. I'm John Graph, my book for making a sculpture. It's a big skier, I love skiers. Hey, parents, I'm glad you're here. That's nice. Enjoy it. It's an interesting character. You've written several books. Yeah. Four or five, that's cool, I guess. Four or five in the area of genetics and theology and genetics and ethics, but... We're ready to begin our conversation, resume our conversation. Thanks a lot. Now, focusing on Dr. Holland's comments on the evolution of viruses. Do other speakers have some questions to ask Dr. Holland that were raised in his presentation? Dr. Peters. John, you didn't expand on the last slide, but I wondered if you thought that had anything to relate to the growth of mega cities, the way we raised chickens and cows and some of those things. As you're well aware, CJ, the growth of great cities of Megalopolis is a 100-year-old phenomenon, and it was made possible by the control of infectious disease. And the infectious diseases that were controlled pretty much not by medicine, but by public health measures, modern sewage measures, and so on. As we get more and more crowded, it gets harder and harder to maintain this. You go through Europe, I spent a sabbatical in Switzerland. I got sick a number of times, and it was due to sewage lines, ancient sewage lines crossing over water lines. You're well aware of this kind of thing. As civilization gets more and more packed and more complex, the probability for diseases which are virus-caused and therefore not easily treatable or prevented goes up. Density of the human population increases the probability of quasi-species swarms being transmitted as you're well aware. You know the answer to this better than I do. You explain it so much better, and you have so much better basis. You just gave my lecture, but okay. You describe the competition between the virus and its host, and even the loser gains in evolutionary fitness. No, let me emphasize, those were two viruses competing with each other. Oh, two viruses competing with each other. That theory actually has to do with hosts and parasites, too, so I wanna clarify that, okay? Yeah, well my question has to do with a general outlook. Is competition the only way to look at it? What about those people that like to speak of co-evolution in terms of mutuality and symbiosis? Do these things make sense to you when you see that evolution is just continuing and the competition continues? Absolutely, one of my former students is a professor at University of California, Irvine, where I was the first department chair and the first professor. I've been around too long, it's time to leave, check out. But in any case, he has a theory about DNA viruses that not only do they co-evolve with their hosts, but they literally help to drive the evolution of their hosts because they're capable not only of integrating onto the chromosome and making regulatory proteins like Bill told you about today, but these regulatory proteins can upset the usual pathways and regulatory schemes. And one of the beautiful things, even about a DNA virus which mutates hundreds of times less rapidly is that you take them out of the host if they were on a chromosome and now they'll tolerate much more mutation. The advantage is small size is you can have a much higher mutation rate. That's why RNA viruses can do this. If you did that in the human genome, you'd have a polymorph, okay? You can't possibly have mutation rates like that, but take out a human gene on a virus, let it be subject while the virus is growing somewhere to rapid mutation even with a DNA virus and now you go back in with a new human gene. Or you use the human gene as Bill told you today to regulate the immune system. So co-evolution is absolutely a fact that was implied today. Can you say anything about the role of evolution in a species in which the virus is evolving in a pathogenic way? You showed the acute infections in each other versus a reservoir species where the virus may be relatively dormant and is there any interplay between those two? Let me, that's a difficult pathogenesis pathology, virulence, these are multi-genic, very complex. Let me tell you a story. I live in a small town and I've retired and moved to New Mexico. I live in Taas and last winter when I was an influenza virus epidemic, fairly young man died and they were burying him as the hearse went up the hill, the back of it opened up on this icy winter day and the coffin fell out and skidded down the road. It skidded all the way into town, crashed over a curb and went into a drugstore. As it did, the lid of the coffin opened up and the corpse sat up and as he skidded by the pharmacy counter, the pharmacist leaned over and said, can I help you, sir? And the corpse said, yes. Can you recommend something to stop this damn coffin? Ha ha ha ha. Ha ha ha ha. Let's talk about this damn coffin in two, okay? One can see right away that the symptoms of coughing are very useful to a respiratory virus. Once again, I don't need to tell any of you. So that it's quite conceivable that there is some selection for respiratory spread. Now, many respiratory viruses, I happen to believe, I don't know about CG or you, I haven't to believe we get more flu off our hands, shaking hands and so on than we do by respiratory spread, but it certainly does happen. And so there could be selection for that kind of virulence. There is clearly selection for rabies virulence. Rabies moves centripetally, it tends to move up to the central nervous system, and then it goes out centrifugially, goes out to the other parts of the body at the same time the animal, the carnivore, is stimulated to bite. It's very clearly, it's clear that this virus has evolved a very complex mechanism affecting sexual behavior, aggressive behavior, biting behavior, or somnolent dumb-lap rabies, which helps to spread the virus. And you can see the populations of carnivores in Europe go up and down with the virulence of this virus and with the spread of this virus. So it's very complex. The answer is some viruses clearly gain something from pathogenesis, from pathology, from virulence. And my feeling is that most viruses, again, I'd like to hear CJ, most viruses, it's accidental. They have to grow in cells. They often kill cells. They just rub pathways. And so most of this is accidental. I don't know, what do you think, CJ? Yeah, I would agree. It's just particularly in viruses that are maintained in a single host that they have to have some properties to accelerate, to maintain their transmissibility. I think with zoonotic viruses, it's a little different. They're maintained outside humans and they just pop over occasionally and cause trouble. And that's just an accident. A question from the audience. Knowing that viruses are so highly mutagenic, are we shooting ourselves in the foot by developing antiviral medications, particularly with implications with the mutations that are occurring in HIV? I think there's no choice. You have to try to lower the virus population when you can. Of course, it's a hazard that, well, let me back up. I think so many variations of the virus already exist that probably the bulk of those that are fit best to replicate and cause disease are already there. And there is a danger of therapy and having something emerge that resists the drugs. And it might give unexpected other effects, but there really is no choice. The greater good is to lower the amount of virus or to spread more of the epidemic and the person dies. But I can't say it's not without risk. Do you want to play further with that? Question, another question from the audience. We generally think of evolution as occurring in living things. Do you think viruses are alive, not alive? Or is the question totally irrelevant? Let me give you my view of the origin of viruses and the meaning of viruses. First of all, I think most scientists now have come grudgingly to the point of view that the earliest forms of life on Earth, the really meaningful information-containing forms were ornate. Hugh Robertson and others, and I believe it's correct, Don Gannam and others, now become convinced that viroids were among the earliest forms of life on Earth. These are little long, these are small, long pieces of RNA, very tiny. They fold back on themselves and they cause pathology occasionally in plants. These things have been proposed to be among the earliest forms, and I think it's a very good proposal. And I believe that other viruses probably arose from these, including ultimately DNA, which became our genetic material. DNA has the advantage of being stable. RNA has the advantage of being able to rise early in a primitive soup. When you have mutation rates like this, the origin of life does not become mysterious, it becomes probable, okay? What was Jostek has shown that when he randomly generates RNA sequences, these RNA sequences, one in 20 trillion, and that sounds like a lot, but it's not very much when you consider these small molecules, one in 20 trillion takes on a particular type of enzymatic RNA enzymatic ribosome activity. So I happen to believe that life is probable. We don't need to invoke any guy, a hypothesis or anything else, and that evolution is absolutely determined by climbing up sequence space. There are so many pathways that it's almost inevitable. And by the way, I want to say something. Every time I give this talk, people bombard me with evolution doesn't fit with a Bible and so on. To me, it's inconceivable that somebody should reject God's world because they believe in God, okay? I think that evolution is the way the world works, that's the way God made it, and that's the way we have to accept it. Now, back to the question. Viruses, modern viruses, Bill already told you, modern viruses have picked up all kinds of cellular genes so they can use these cellular genes to non-regulate immune responses, to control cytokines, which you work on yourself and so on. So the viruses are very clever, moving in and out of cells, picking up genes and moving them elsewhere, helping to drive our evolution, driving their own evolution, but when they pick up our genes, not only do they use them to cripple our immune response, they can mutate them and we can get them back in a new form, which might be a new gene, okay? So tiny elements like this, the transposable elements, the retro-transpose on the viruses, they're very useful in evolution and there's no question that they contribute to evolution. That was shown back, the early Nobel Prize winners for their RNA-phage work and their DNA-phage work. There's no question about the interaction of hosts. Dr. Galklik, with your years of studying RNA viruses, how does your perspective of the world of RNA go? I agree with what John said. I mean, I think it's very likely that the first organisms were in an RNA world. Furthermore, I think the origin of viruses is very ancient and probably coincides with the formation of the first cells. Viruses have been around for a long, long time. Well, what John has discussed so very beautifully today is virus change. I somehow, I don't really see that, to me, virus evolution is like the dinosaur whose footprints you see, it's in a certain direction, which indicates that he knew where he was going, we wanted to get there. Now, for evolution, the same thing to me, it's more direct there, you know? You talked about change, like the influenza virus, the formation of the species and all that sort of thing. But in a well-adapted host, it doesn't get anywhere because the virus has got nowhere to go, unless, as I said before, you give the protein, you give a new protein a chance to perform better than the parents did. Do you like to call me? Yeah, I agree with that, Bill. One thing I didn't have time to tell you, and I can't do it without AIDS very well, but let me try to get it across. Manford Eigen's ideas on this are in precise accord with what you said. You said it looks like it's directed or it looks like it's moving somewhere. Now, as you move out in sequence space and get a longer and longer genome and get more and more possibilities, every time you add a single nucleotide or a single coding unit to an already fairly long molecule, you increase sequence space dramatically. And as Manford points out, and I can't really do this well, but let me try to do it. Let's say you have a mountain peak and water drops down off that mountain peak. This is Manford's idea, and it fills a cup. If that cup is two-dimensional, you only have one way to get out of there, you have to overflow here. Now, what if we had another cup in the second dimension? Now you've got four ways to get out of there. Now, what if we had four to the 10,000 dimension? Now you have all of these peaks with an easy reach from any point of evolutionary selection. So it actually has the appearance of being directed, and in fact, in a sense, it's directed along the peaks to move in the robust areas of sequence space. Part of your response is related to just the basic math question. Some of the members of the audience are curious as to how you arrived, just the basic four to the 11,000th power. They're just curious as to how you got there. That's just simple probability. Okay. I know how you did that, but some of the other people are. If you, let's say you have four bases, and it's two bases long. You have four to the two possibilities, okay, eight. Now if it's three bases long, what do you have? Four to three, or? 16. You just go on and on like this, and it just gets out of control in a really big hurry. We do this in Principles of Biology, but it wasn't one of my Principles of Biology students that asked that question. A question from the audience again. Do the path, part of this question, I should let you know the background. One of the topics of Nobel conference several years ago was chaos, the theory of chaos. So do the pathways that viruses form in sequence space form chaotic patterns as studied long enough? As a matter of fact, we've published that. I'm not good enough at math and physics and nonlinear dynamics to do that myself, but some physicists at UCSD and the Nonlinear Dynamics who took our data, the raw data, and analyzed it, and in fact, you see chaotic interactions. We put two populations in competition. This is published in a physics journal a couple of years ago, and this is Werthe, is the senior author for those of you who might wanna look it up. And basically what we showed is that if you put the same things in competition in the Red Queen, I told you about the Red Queen, one will take over and the other dies, but that doesn't always happen. I didn't have time to show that. You get a nonlinear dynamics with some of these systems and they start to go through a chaotic oscillation. And our physicists have analyzed that and it's pretty hard to know what the hell's going on, but it's a chaotic oscillation. Well, I think it's time for the bell to ring for changing classes, so we will take a 45-minute recess and reconvene at 3.30. Excuse me, I wasn't trying to be update, but I just wanted to know if there was...