 But this is a dream of every molecular biologist to get his gene of interest to find gene therapy. But at the end, it seems like, tell me what you think. You don't need all these basic research. All you needed to know, you don't need the rubs, wheat rub, which is my interest. All you need to know is the gene that is defective and try? Well, yeah. I wanted to demystify a few things because I come from the same perspective that you come. And all I wanted to say is that sometimes things that we thought were very important were not and others that we thought were not so important are actually very important. So the key to this is actually to have collaborators that are experts in the different areas to work together and complement and find the right path. So I think that's what I learned a lot with the ophthalmologist, even in terms of, I mean, the way he approached this and what, really, it was quite a learning experience and some things. I mean, perhaps I made too strong a point or I made it too simple. I mean, it's not, but also luck is really very important because all of a sudden, you know, you had the loss of function mutation, you had the eye that turns out to be a good system. And we could not predict that in the beginning. On the other hand, I think my medical training made me more keen to do this because I think in a lot of people, most of you are not medically trained. And so you are sometimes a bit afraid about venturing a little bit out of your comfort zone. And for me, that was more natural because I spoke the language and I could understand. And it was also my own interest because when this happened, and I was obviously as a cell biologist, very interested in the pathogenesis and then, you know, I did a lot of work on RAPGTPS and then RAPGT-7 turned out to be a very interesting story, completely independent of Choroidremia, but it was always, I mean, it was always there, this aim of the application and sometimes it's just belief or something or motivation that you want to pursue that path. And it's not easy, but you persist on it. Questions? Thank you. Do you want the microphone? No. No. OK. I have a question for Joel. At the risk of seeming overly aggressive, I wonder what is the motivation? I can understand the motivation for doing like one or two synthetic chromosomes, but for doing all 16 synthetic chromosomes, it just seems like it's a lot of resources and a lot of time, you know, when those resources could be used to do other things and the time could be used to solve diseases or something. So why, you know, you learned a lot from doing one or two chromosomes. Why, it seems like they're diminishing returns, you know, by the time you get to chromosome 16. So why is it important to do all 16, make synthetic copies of all 16 chromosomes in yeast? Can I answer one just for a second? He's doing disease on the side. OK, so you could shift your resources to disease. So that's a really good question. And that's why it took us, that's one of the reasons why it took us so long to get the project funded. Like we started working on it like 2006 and didn't really get funding for it until like our first funding was a Microsoft like seed grant where I convinced them that revision control systems that people use for software like we need similar things for genome editing. I think that there's some things that you, well, there's some things that we want to do that we can really only do by doing a large scale project like this. So for example, doing the stop code on recoding is useful functionality added to yeast. And you have to edit the yeast throughout the genome to do it. And you could say, well, you could do that using homologous recombination to do that one by one. Getting rid of all of the repeats to see the function of repeats in the yeast genome, like really have to get rid of them all. You could do it one by one. Putting in the locks piece in recombination sites. So what that lets us do, so I'm sorry, I missed Charlie, Charlie, I missed your talk. It was just like the travel was too hard and I was either on the plane or passed out when you were speaking. Well, it might be better, it might be better. So, but so it lets you do screens like synthetic lethality, but like letting the yeast do all sorts of copy number variations. So for the yeast project, there are a lot of things that are proving to be useful that we could really only do by doing it genome wide. So now in terms of return on investment for what we're doing versus other things, you know, the costs have come down so much. And like you could say the same thing about genome sequencing, like after we've sequenced one genome, why do we have to worry about sequencing another? But because the costs have come down so much, it opens up new possibilities. And one of the things that maybe has been driving genome synthesis costs to come down are projects like ours. So like in the end, I think it was worth it. And I'm happy for the six, eight, nine years I devoted to it with only getting like one or two papers out and like convincing my postdocs it's worth sticking on this project because we're gonna like build something great. So it's a huge question for people talking about mammalian and doing like, you know, what's the point in synthesizing mammalian genome? And there, you know, I'm a little harder pressed to say that like right now is the time to really start a mammalian genome project. But on the other hand, it's really taught us a lot about synthetic chromosomes. And I think where my group's research is going to go at least is not really synthetic genomes, but more functional synthetic chromosomes. Okay, yeah, I mean, I wasn't, I wasn't arguing with doing a couple of chromosomes. It was just like, what do you learn, you know, just further down line. Is your aim entirely basic research? You could imagine using this. Oh no, a lot of it's industrial. So for example, for strain engineering. So for strain engineering, they're like they're a lot like it's very useful to be able to engineer an industrial microbe in order to maximize production of some product. And right now, like a lot of that just sort of trial and error with just standard mutagenesis. And what we found in I guess some papers that I'm kind of in yeast and other like other work going on is it's incredibly powerful for a pathway engineering to take a pathway, put on a chromosome of its own, put in these LOXP sites that let the cell try different copy numbers, select the ones that do best and go from there. It like, you know, it seems like, oh, standard mutagenesis should work fine, but it's really doing like a lot better, a lot faster. Like the reason is that we get a lot of yeast cells in a test tube that are all reorganizing their genomes in different ways and it becomes very useful. Questions, no questions? Yes, Jerry? Yes, a question for Ben-Lehner. Maybe I missed something during your talk about transgenational inheritance, but do you know because some epigenetics marks can be erased or reprogrammed during after fertilization on a UNID why your specific Schwanngel is still expressed? No, so why this rare example does show epigenetic inheritance and the rest of the genome does not? No, I mean we think it's to do with this particular chromatin state that it's this heterochromatic repressed locus in the germline that is then activated every generation, which is a very unusual epigenetic thing, but why it happens, we don't know. Didn't you say it has all the other repeats were behaving the same way, the natural repeats in the genome? So some of them, the ones which repressed by the same histone modification seem to behave the same way to a lesser extent, so in the genome. So it's something to do with the chromatin, this particular chromatin state that is temperature sensitive. So you think that this is the end of, I mean you try to understand what is the mechanism maybe, but is this the beginning of trying to understand inheritance of epigenetic markers? Where are you going with it? Okay, maybe it's to... Well, we would obviously like to know if this is used anywhere, either in C. elegans connecting to proper higher level phenotypes, fitness, life history traits, or in other species, I mean it may not be used for anything in C. elegans or other species, so we're systematically testing all of that basically. The other thing we work on which I didn't talk about is just these one generation effects, so there are many more one generation effects where the physiology of the parents affects the traits in the next generation, and one of the big influences in C. elegans is the age of the mother, so the very young mothers are producing osprey which are very different throughout their lives. This is the dominant cause of phenotypic variation in the lab, actually, we think. This one is partially to do with the loading of a protein complex into the embryos, so it changes over time, and we don't understand the mechanism underlying this at all. This is not probably a chromatin-based anything, it's probably a protein-based effect. I want to say one more thing about the cost of the project. So, no, because when we started the project, DNA cost... That's right. No, because I was thinking more about it, seriously. When we started the project it was like a dollar a base for synthetic DNA and like 12 million bases, that's 12 million dollars just for the DNA, and everything else is like 50 million dollars costing, if everything had been done at the cost of the beginning, but now DNA costs are like down to ten cents to one cent depending on like at what stage it is. So then if you think about more like in the end this cell to redo it would cost between a hundred thousand dollars and a million dollars. Then it becomes more reasonable, and I think that's very... That's like there's a lot of exciting stuff you can do in that range, so I mean that's less than PI salary for a year. Now depending on the PI, I'm not saying not my salary, depending on the PI. But is it going to... Are you going to be able to do it faster the second time? Oh yeah, definitely. You do it faster because first of all you can order bigger pieces, so a lot of the entry stage work that we had to do you don't have to do. So it's... Yeah, you can order a bit, there are better methods for... How much faster? So I wouldn't want to... Like if you listen to twist they have the capacity to do it tomorrow. In the end like I don't know how much... I don't know how much faster. I mean honestly you might be able to do it very very quickly. Well like the part of it that has taken the longest time now is working out the bugs. Like you know you order a chromosome, you put it together, you find a fitness defect, working out what nucleotide caused the fitness defect can take months. Oh, I want to ask... So maybe I missed it. What is the phenotype of a mouse without rep one in the retina? It has the same phenotype? So I showed you the carrier, so the carrier mouse has retinal degeneration, yeah it's the same phenotype, yeah the same phenotype. Although we have not looked at the coroid in great detail, so we don't know if there's also degeneration of the coroid, but in terms of the RPE and total receptors it is pretty similar. You tried to recover this with the injection? Yeah we could, we could recover, yeah yeah yeah yeah. I just didn't, I didn't show you the functional analysis because it's quite slight and even I don't understand very well all the electro-retinograms and all that. There's slight changes that are sort of normalization of that imply better vision, that that was all done. And mice can also occur. Actually I have to say that an interesting comment, there is a disease called LCA, Leavers Congenital Amoroses, who's now has been out of Philadelphia and there's been a company called Spark Therapeutics that's now been approved by the FDA to license this. So they're there two or three years ahead of Nightstar for Chroedremia for this disease. And the interesting thing was that the preclinical work in that case was harder because they were first ones, but it was much simpler because it's a disease that involves no electrical activity in a much more dramatic disease with no electrical activity. So the rescue and the little blip will give you a proof of concept in terms of for Chroedremia that was much more difficult. So our preclinical work because of the lethality, the conditional knockouts, the subtlety of the phenotype, all that work was much more difficult. But now clinically Chroedremia is much more attractive than LCA because it's a much more severe dramatic disease. So the kids by age 15 they're almost blind so it's very hard to and then much rarer because LCA is heterogeneous and this is just a subpopulation of these. And so so it's interesting because everything that was bad, it's a complete switch preclinical and clinical in terms of these two diseases. Glaucoma is one of these complex diseases. There's many you don't find one. No, no, no, no, no, no, what could happen? In terms of expression level, not in terms of... Now our major work now is actually on macular degeneration, age-related macular degeneration because involves RPE changes that are actually quite similar to the ones we've seen in Chroedremia mice. So we're coming from a perspective of now pure pathogenesis because nothing is known and there it's just completely virgin that trafficking defects and it's macular degeneration is thought of an extracellular disease and we're trying to make the point that it is definitely an intracellular disease that then has an extracellular manifestation later on. I have a question for Ben actually and it's about... So you said that you talked about how the model of transgenerational option and inheritance you have is not a good system for transferring information and that it's not going to be representative of a Lamarckian inheritance model. I think that's downplaying a little bit that kind of idea because you can have a cute selection that happens within just one generation and that can definitely affect the fitness. So like even though it's not necessarily a long-term natural selection thing, like a change in the gene, I mean how do you feel about that? I don't know. So I think this could be a selectable trait if you so mean that they have this ability to transfer information but because it has this very short half-life relatively on an evolutionary scale it's not like environmental liquid information that then gets fixed so it's not like the giraffe stretching its neck and then all its descendants forever are going to have a long neck. But yes this can just as a heat shock response within a generation can be beneficial. It could be beneficial across generations. So in toy models it's easy to come up with conditions where this kind of thing could be beneficial as long as the generation time is shorter than the time scale that the environment is changing, then it becomes beneficial to transmit information about the environment because the environment now predicts the environment tomorrow. Well what you're saying also is that the more times you do that to the generate, like you know you said you do to five generations and you have sort of an increased effect on the HSP90 expression. So like actually if the environment, it can go on forever until the environment stabilizes right? Yeah but the environment is not stable. I mean this is like an adaptation. I mean our system is synthetic okay so we're not claiming it's beneficial to the organism in any way but you can you can cook up in a computer model situations in which something like this can be beneficial provided that the time scale over which the environment is changing is longer than the time scale of the generations then it becomes beneficial to transmit information across generations. In humans it doesn't make any sense because the you know the generation time is not is massively longer than the time scales of environmental change you know. It's not predictive what the temperature was when my mother was born versus. Do you think like a famine? People like that the epigenetic memory for stress conditions including maternal stress. Ah no biologically you can see these effects like a very acute starvation in mice and rats etc does have an effect on this generation but no one would claim that's adaptive or beneficial. He's asking about whether this is a... If it can be... Well some people may claim that this is adaptive or beneficial but I do not agree with this. Nadu would like me to ask you if you have any more questions. If not I think we may thank our speakers.