[MUSIC] So this is the title of the paper that was received by the journal in May 43 and appeared in November issue of Genetics. And I leave you this little note, theory by MD, experiments by SEL. This is a sharp division of labor, which you find more often in journals that was completely absent for many number of years. The paper begins by discussing the phenomenon of secondary growth that occurs when a bacterial culture is first lised and then gives rise to a new turbidity, a new bacterial growth. So the question that the paper address is, what is in nature of these variants and how do they occur, the manner in which they originate? They are considered to be a new stable character, and it is a mechanism that they want to try to understand. There are two basic ways of thinking about it. The first way is that the phage induces the bacteria to change, and that the bacteria therefore, some bacteria become resistant. That is the induced action. The second hypothesis is the mutation hypothesis. The mutation occurs at anytime during the culture. And the mutation is independent of the presence of the virus. So these two are shown here, and this is directly taken from the paper. These hypotheses are discussed several times, three times in the paper. But this, I think, the most clear and concise way of discussing it. The mutation hypothesis, there is a probability, low but not none, for any bacterium to mutate during its lifetime, that the lifetime which sounds a little bit strange. The lifetime of a bacterium is a time between the bacterium's birth, through sition from the mother and the bacterium giving rise to two bacteria. So, that's what's mean by life. Every offspring of a resistant mutant is resistant unless reversion occurs. The term resistant means that the bacterium will not be killed if exposed to virus, how the interaction will occur left open. This resistant is the same as a resistant for the R form, the rough form of Avery. The second hypothesis is acquired hereditary immunity. This is basically what happens when a bacterium is infected by a lysogenic virus. There is a small finite probability for any bacterium to survive an attack by the virus. Survival confers immunity to the individual and to its offspring. The probability of survival in the first instance does not run in clones. If we find that a bacterium survives, we cannot from this information infer that close relatives of it other than descendants are likely to survive the attack. What does this mean? You take, A first bacterium. This first bacterium will give rise to two to four to eight. Now, this bacterium will be resistant. If this bacterium is resistant we can infer, or we can infer that this one is also going to be resistant. This one, this one, this one, this one because they're all descendent. But the relatives, other than descendants, are this guy and this guy. Can we infer anything about these guys? They are relatives, but they're not descendants. This is a brother or sister. We cannot infer anything about it just by isolating this particular individual. We know that its offspring will be resistant. We don't know about the parents and brothers and sisters. That's what it means. So this is the essential difference between the two hypotheses. In this case when a mutation occurs, it will give rise to a clone. This will occur whether there is a virus or there is no virus. This occur by accidents of replication of the DNA, errors in typing. And when you make an error in typing, you stop making the right protein and the right cells and the cell surface, which will attach the phage. In the mutation hypothesis, the mutation can occur at any step during the growth of the culture. During the acquired immunity hypothesis, the virus has to be there to trigger the phenomenon. So the virus has to be at this stage, has to come into the play at this stage. That's a major difference. So they predict experiments that will be done six and ten years later and will confirm the model. We'll discuss one of these experiments in a short while, okay. So they are saying right away that this study may encourage the resumption of quantitative work on other problems of bacterial variation, quantitative which is what was missing with Avery. They're perfectly aware of Avery's work because Delbruck knows and met with Avery's brother. And he went to see the work in New York. But it's not quantitative. So we'll skip the theory because this is best left to those who are keen on doing math and equations. And we'll jump straight ahead to the experiment. So in the experiments Luria states right away that he's not dealing with lysogeny. He was aware of lysogeny because of his stay in Paris between 38 and 41 and 40. So they're not lysogenic. They don't bind phage. They don't attach phage. They don't do anything with phage. They behave like phage is nonexistent, and the phage behave like bacteria is nonexistent. They will discuss, and this is clearly Luria's work and Luria's thinking. They discuss the physiology, remember Luria is an MD. They discuss the physiology. Cells may grow in rich medium, in poor synthetic medium. They didn't test at all temperatures, but they tested at least two different media. Cell may grow in big volumes or in small volumes, so they test a number of generations. So, they test various conditions before they're going to analyze their mutants. And the first experiment they do is something you would think is trivial. But it's very important for them because it's the proof that what they're measuring is something real. So, I will go to the document. Luria is using three cultures, independent cultures. Independent means that each of them start from a single bacterium, which by itself is sensitive to the phage, three cultures. In these three cultures, he takes ten times the same amount of liquid. We call it an aliquot. And he plates it to count how many phage resistant bacteria there are. You can think of it as five drops, if you want. And he plates these five drops on ten different plates with a virus, a lot of virus to kill all the sensitive bacteria. And the next day, he counts the number of colonies. Each colony is a clone derived from a single bacteria. And what he sees, basically, is whether the system is solid and reliable. And if you take the first culture experiment, 10A, 14, 15, 13, 21, 15, 14, etc. Average 17, variance 15, basically, everything behaves in a perfectly normal way. There is no bias in the experiment, same thing with the other culture. Now even with a culture that has a very few number of colonies 4, 2, 2, 1, 5, 2, 4, 2, 4, 7. Average 3.3 variance 3.8, everything behaves normally. You don't have one plate with 100, that would be odd. So, if you take a big culture and you sample this big culture as many times as you want, you will get more or less the same number of events, or of colonies on the next day. So far, so good, this is not worth a publication. This is just telling them that their method is acceptable and reliable. Now they take a series of cultures, not one, a series of cultures. And out of each of these cultures they take an aliquot, a small sample. For instance, if you take experiment 16, they have 20 cultures, 20 tubes. Each is an independent expansion from one bacteria. These cultures have a volume of 0.2 milliliter. If one serves you 0.2 milliliter of scotch, you won't be happy in a bar. They take about slightly less than half of that volume, and they put it on the plate. And what do they see? They see cultures with 0 like this one, this one, this one, and one culture with 100. 0 means that there is no mutation occurring in that culture. 107 means that there is at least one mutation. Of course, you don't know whether there was one mutation that gave rise to 2, 4, 8, 16, 64, 107, 128 bacteria, or whether you have 107 events, each independently occurring in a different cell. You have no idea. All you know is the total number of mutants. You don't know the number of mutations. But right away, you see that the number of events, no matter how you count them, is very variable, high, low, very low, nothing, nothing, nothing, very low, very high, 64, very high. And that's the same for all the different experiments. The only thing you may notice is that there are more mutants in the large cultures than in the small culture. The large culture being the cultures than in the small culture from experiment 16 and 17 and 18 and 19. Now, the real key experiment is experiment 23. Here Luria has used 87 tubes. 87, it's a pretty large number. 0.2 ml each, he let these cultures grow. And then he plated the entire culture on the virus, so he tested the entire culture. There's no bias. And 29 of the cultures do not contain a single phage-resistant bacterium. 17 contained 1, 4 contained 2, and 4 contained more than 200. This is not the distribution of resistant bacteria that you would expect from an induced immunity. This is typical of what you would expect from a mutational event. So from this and this series of experiments, it looks like The mutation are occurring at random and will give rise to large and small clones which is favoring imitation hypothesis. On top of that what Luria could calculate are mutation rates, which are of the order of 1 in 100 million per division, 1 in 100 million. Remember a fully grown tube contains a billion per ml. We're dealing with large numbers here. Now there is one problem that was The culture will grow with time as an exponential culture and reach a plateau. So far, so good. When you reach the plateau you stop dividing. If a cell becomes mutant at this level, at this moment in time, you agree that this red cell will not divide to give rise to a clone. So this red cell will give you one bacterium resistant in the culture. And there were several tubes with one bacterium. There were 17 tubes with 1 bacterium. If a mutation occurs, and I'm going to change the color and use the green, occurs here. The green will multiply two, four, six etc. Will give let's say, 128 bacteria. You agree with that and you agree that the red one we'll give one bacterium. That's what you expect. Now if you think about it in physical terms. The bacterium, the white type bacteria carries at it's surface receptor for phage. This is the phage receptor, this is the structure where the phage will bind. In the case of this phage, it's two protein attached together called TonA, TonB. Now, the cell in these conditions has about 500 copies of this protein at The surface. I only drew one but think of it as 500. Now, this cell is a white out cell, it has a chromosome. When you're going to mutate this cell by destroying the gene for the receptor. So you only touch the DNA. So the DNA now carries a deletion of this gene. Now, what is this cell? This cell is genetically resistant to the virus, but this cell still carries the receptor. It's phenotype is sensitive. Because you need to get rid of all these receptor molecules at the cell surface, before the bacteria becomes resistant. So you can do this either by division, Dilution Or by turnover. A degraded protein of the cell surface and you replace it when you by new protein since you have no genes. The gene is gone, you can not replace that protein. So either you Destroy and replace the proteins or you divide. You have no choice, it's either one. It can't be both, because this cell is. So the green cells, you have no problem to believe that the green cells have lost be the evolution. So the green clone, the 128 bacteria, you have no problem seeing. It's the one bacteria that pose a problem. How could those bacteria happen? Well, it turns out that one can calculate that if Luria was at that time what he was later, he was not somebody who would work at strange hours. He would leave the lab by 5, he would come back by at 8 or 8:30 in the morning. If he leaves the lab by 5, that means he's inoculated his culture in the afternoon after lunch. Basically, the cultures are saturated before it's been done. And the bacteria sits in the saturated face from midnight to 9 o'clock in the morning when you come to test him. So they have plenty of time to destroy and replace if they have the genome, not replace if they don't have the genome. This is something which is called which was certainly, is discussed by the authors but not in a quantitative way. And although they had the tool to identify this in a quantitative way, at the time they didn't think about it. So with this we've seen that they have analyze the distribution of resisting bacteria. They have studied experimentally the distribution, conform with the conclusion drawn from the hypothesis arrive by mutation and measure a mutation rate, a number. The probability for division to become resistant. That's what made this paper so impressive. The combination of the three methods and the prediction. And among the prediction is one which is that the notion that, if you were to analyze a large number of micro colonies, very small colonies, thousand bacteria per colony, you could identify clones and non clones. And this is basically what Newcomb managed to do. Okay, this is basically what Newcombe was set up to do in 1949. In 1949, Newcombe publishes his little letter in Nature. So this little Nature note, which is not even a page, had two tables. The first table describes the instances in which the Luria–Delbrück proposal has been tested experimentally. What I find interesting is that if you had asked me before reading this, I would have said that all the examples were from E coli. This is a prejudice of somebody who's worked most of his years on E coli genetics. In fact, at the time, there were quite a few E coli work, but there were also experiments done with Staph aureus, Clostridium, Haemophilus influenzae, and Eberthella. The data about lactose fermentation and salmonella fermentation were considered to be less serious, because non-quantitative. So the experiment of Newcombe is extraordinary simple. He plates a certain number of cells of bacteria on a plate. In this case, 51,000 on many plates. And then he waits for three hours, four hours, five hours, six hours. And so, there's an increase in the number of bacteria. That's standard, nothing special. And then, one plate for each time-point is either treated with phage, and this time, he uses a spray like a makeup spray, he uses a spray of phage on the plate. And that's for one of the plate, and for the other plate, before he sprays he redistributes the bacteria on the surface. He mix everybody, he reshuffled the cards, and he sprayed the plate and he kept. And so, the three and four hour time-point are not very useful, five and six hours are useful. At five hours, unspread plates, maximum 28 colonies, spread plates 353 colonies. Six hours, 240, there's a typo, the two is missing, 12,000 colonies. This may sound very dry, so I've prepared a little cartoon to illustrate this. All the numbers have been rounded up to make it simpler. 50,000 colonies instead of 5.1 and 10 to the 4th. Five hours, about 12 generation, six hours about 15 generation. Each bacteria gives rise to a little clone. The little clone has 5,000 bacteria after five hours, and 40,000 bacteria after six hours. Roughly 15 generation and roughly 12. Think about it, 2, 4, 8, 16, 64, 128, 256, 512, 1024. 1,000, 10 generation, 1 million, 20 generation. It's pretty easy to calculate. You have here the number of bacteria, the number of clones, and the number of bacteria per clone. Of course, it's bigger here than here. The number of clone doesn't change. This plate is unspread, and this is a pool, and I made it this as a pool, it's 28. 28 colony clones out of the 50,000 clones contain at least one Ton resistant bacteria. Why do I say, at least? Because I don't know how many resistant bacteria are present at that spot. One, two, four, eight, etc. When I do the respreading, or when you come do the respreading, he counted all the bacteria that are resistant. And he found now 353, which means that each clone, on the average, contains 13 resistant bacteria. Of course he does not know the distribution, because he is not doing a tube for tube like Luria was doing. But, look at what happens at 6 hours. The number of clone will go up. 240 versus 28, but the number of bacteria goes up 52 fold. 52, the average of 52 Ton resistant bacteria per clone. And 12,000, it's basically saturated plate. So that tells you that the number of clones increases, and the size of the clone increases, which is exactly what you would predict for a mutation. This occurs all before you add T1. Remember, you spread or don't spread, and then add T1. You spread, don't spread, then add T1. So, the number of resistant bacteria, and the number of clones will increase, because in each of the one that were already present at five hours, are still there at six hours, except they're bigger. Each of them has more bacteria. Approximately ten times more bacteria. So, this was the experiment that finally illustrated, if you want, the concept of the fluctuation test, even though this is not a fluctuation experiment. The last nail in the coffin will come when Lederberg will show, in the 50s, that you can isolate a strep-resistant bacterium that has never seen streptomycin. So, that was the end of the induced adaptation. But, that came very much later.