 Yeah, nothing on dormancy today. So in case you didn't like that topic Maybe I can try to change things up a little bit Yeah, so I think the thing that it's been interesting about this meeting so far I have to leave today. So if I don't get a chance to meet and chat with all of you, I apologize But you know, you kind of get a sense of how people approach different problems differently thinking about cellular biology and physiology metabolism communities, maybe how they function and So the way in which we kind of collectively approach those problems comes from like two ways, right? We we can we can take top-down approaches where We look at patterns and have hypotheses and kind of reject what might happen based on maybe the distribution of abundance of species And we can take computational approaches comparative approaches statistical approaches And we do that kind of work to that. I really like that and then there's sort of the other way in which Biologists tend to approach complexity in life, which is to try to break things, right? Like we find some gene of unknown function and we can Remove it put it back in and see if we can get things to work And you know through that process, maybe we can find some generalities about how a system works maybe together those two approaches we can kind of come up with some Model for her for about the complexity of life and some rules that we can use for you know prediction and understanding Such so so I see some of that going on here this week And I'm gonna decidedly today take a different approaches, which is more of his bottom-up approach, which our group doesn't always necessarily do But in this particular example was interesting to us so I'm gonna talk about the evolution of a minimal cell For people who don't aren't familiar with that concept It's the idea that there might be a form of life that only has the essential genes that are needed for basic functioning property of life being able to metabolize divide and Conduct care homeostasis of a single cell autonomous form of life So so one type of complexity that is important in cell biology is reflected in the genomes of organisms And there's different ways that we can look at the complexity of the genome But one is it's its size So taking the sum of all a's t's c's g's in a genome When we do that there's a lot of variation Across the tree of life spanning 10 orders of magnitude So some of the simplest Biological entities on the planet which have the most streamlined genomes are things like single stranded DNA viruses They may have a handful of genes maybe a thousand or two Total genes base pairs. So those are the simplest of course those organisms can't live on their own They're persisting through the infection and taking over the cellular machinery of their hosts But on the other end of spectrum, there are some organisms like the marbled lung fish that have upwards of 100 billion base pairs in their genome so so 10 to the power 11 And Tens of thousands of genes So there's a lot of variation in the essential core material and information that makes up an organism and what I'm showing on the on the left is a scaling relationship between Genome size and protein coding genes and it's generally positive But if you see in the upper right hand quadrant the orange color dots those are all for for you carry-outs And so this relationship becomes sublinear Due to the fact that protein coding genes are going down in those organisms with larger genomes due to the retention of things like introns And other types of what sometimes we will call junk DNA that might not actually be Contributing to the performance or function of an organism And so there's questions about how that information why it's retained if it has any function So the maintenance of genome structure, but also again the consequences of what what all those A's T's G's and C's mean for the performance The ability of an organism to be a generalist and live and tolerate different environmental conditions, for example so I'm going to talk about a synthetic bio biological approach to Investigating questions related to genome complexity and genome size in particular So this is a relatively new field depending how you define it that's been around for maybe a decade or two That draws on Principles and ideas from different fields So so drawing on fields of genetics molecular biology engineering physics computer science and Generally what the the goal of synthetic biology is is to to modify an organism to use its parts Modules circuits that can either be constructed or deconstructed to create an organism that can perform some function That's useful typically for society, right? So it could be can we create new drugs? And can we create them at a yield that might be economical? Can we can we create an organism that might aid in for example bio remediation of an oil spill? Or are there new crops that we could create that might be tolerant to environmental stressors that are associated with global climate change? So these are some of the questions that applied questions that synthetic biologists might ask But i'm going to argue that I think we can also use synthetic biology to address basic questions in biology and specifically in evolution So i'm going to talk about how people have used synthetic biology to Create a minimal cell to address basic questions in biology and cell biology and evolution And so this question has been around for quite some time actually probably since the early 80s And the question this came up yesterday in one of our summers the idea of the analogy of the hydrogen atom So when I've heard this story and the motivation behind this work people have said if we want to Understand a quantum theory of an atom, you know, you would not start with uranium You would start with something that is much similar. We have one proton one electron and we can Describe orderables and come up with You know equations that can describe this simple atom and then we can start to build up and understand the diversity of elements within the periodic table And so the the motivation for having a minimal cell was that this would be You know perhaps an equivalent of a hydrogen atom for cell biology What are the minimal the bare minimal elements components needed for self-sustaining autonomous form of life? And so they started to build a candidate list of species that could be Used for constructing a minimal cell back in the 1980s and there's one group of bacteria that Showed up on this list if they did identified as being a potentially good organism to start with were bacteria called micho plasmas So these are groups as a class of bacteria in the in the molecules None of these organisms have cell walls And they tend to be commonly associated With hosts often pathogenic So we find them in humans we find them in non-human primate populations And in ungulates and goats in particular Or they form Pathogenic relationships and infections and the upper respiratory tract and urinary tract of those of those organisms And so over Like other endosymbiotic and symbiotic bacteria these aren't endosymbiotics, but other symbiotic bacteria over millions of years What happens through this close association? Is that microorganisms provide some benefit to the host, but they're also Getting something from the host usually metabolites And so what this does is this kind of relaxed selection on certain genes within the genome of the microbe And that they're over time you can see the the loss of of genetic information the deterioration Of traits over millions of years of evolution Resulting in streamlining or genome reduction of many host associated microorganisms So this is common And it basically means in the end a lot of those organisms can't be free living anymore But many micho plasmas still have streamlined genomes, but do have the ability to live on their own So now we have some candidate organisms with with small genomes that perhaps we can even minimize further That's the idea So in the 2000s Early 2000s at the venture institute in san diego People started to develop a platform where they could manipulate the genomes of mycoplasma And the idea was that you could create a synthetic organism by coming up with a list of genes synthet in silico and Constructing synthetically chemical chromosomes, which could be transplanted into vectors like E. Coli or yeast assemble those genomes and then move those Genomes in their entirety into another species that lack the chromosome just this just the cell membrane of a vacant cell And so now what you've effectively done is you've you've booted up a new form of life from computer So when they did this for Some of the mycoplasma cells the quote from Craig renter was this is these are the only species on earth whose parents were a computer So in 2016 some of this was published and it was You know kind of touted as being a really a major breakthrough in cell biology and synthetic biology Yeah, and so so it was around this time That I I was at a Gordon conference in new england in 2018 and Clyde Hutchinson the lead author on this work had given a talk and It was a good talk and in the end we we understood what was essential for You know creating a synthetic organism that that they could only live off the essential genes and the way this is done Just just to back up a second. There's a process called global transposed on mutagenesis so they had a wild type organism that had approximately one million base pairs and 901 total genes and they used this global transposed on mutagenesis to knock out One at a time every single gene and then you would score whether or not that mutant Was viable or not Okay, so if if you did a mutation in one of those 901 genes Randomly and that gene was mutated and it couldn't form a colony then it was determined to be essential and you have to keep it If you delete another gene of the 901 and the colony still forms then you know that that's a non Essential mutation and you could toss that out So I was wondering how environment dependent this definition of minimal cell is And it sounds like they went for the environment of colony growth So does this mean everything else you have to do in colonies? So it's a very important and good question So when we first started doing this work, we were talking about the evolution of the minimal cell and we Quickly realized that this that that was not appropriate It's not the minimal cell. It's a minimal cell and it's dependent and conditional upon the the techniques that you use for selection The growth media the conditions the pH the temperature I thought about the the the selection process of thinking about colonies forming on plates But yet that would be another thing that would describe what this minimal cell is It's a really important point We could create other minimal cells I've been kind of interested in like whether or not this model could be useful for thinking about Origins of life because we know that life probably evolved from simpler means, but this is like A chemo organo heterotroph. It uses oxygen as a terminal electron acceptor and it lives in association with ungulates that probably evolved, you know a few hundred You know a hundred million years ago, right? So it's probably not a good model for that But I was you know last week I was traveling in Zurich and we were talking to people about well Why what could we do this for a chemo litho autotroph, right? That would have a small genome it might that be a better model for thinking about origins of life Through this process of simplification and genome minimization. Yeah, so that would be an example of another minimal cell that you could create This is the only one we have right now though Yeah So jade the way you described the evolution of these these micro plasma cells and Do you think you could only you know take an E. Coli cell and put it in a chemostat and just You know wait for 20 years and you arrived at a minimal cell For this for this sort of environment The the chemostat environment, you know, so we mentioned some of the work by rich lansky and crew that they've been evolving under semi batch conditions semi continuous culture for what 90,000 close to approaching A hundred thousand generations. I don't think there's been any evidence for massive genome reduction In those there is gene loss, but not at the you know the magnitude that we're seeing here So I don't think the time scale of phd students or or faculty even That that would be the way we would want to go about but but yeah at some point If you relax selection under those conditions that you're suggesting you should perhaps see some erosion of genes, but It might take thousands or millions of years to see that okay, so Yeah, so we have two cells now. We have this non minimal cell Which is in blue and then you can see on the inner side is the genome of the non minimal cell And there's a lot of gaps and the architecture of this genome Basically, that's where all the non essential genes were removed. Okay, and we're going to contrast The performance and evolution of those two cells There's obvious I want to point out there's a shared component too, right? All the essential genes between those two the blue and the red strings are shared It's just the the sort of a nestedness of we've removed all these non essential genes I was thinking what about genes that are not essential on their own But they're essential when coupled to other genes like there could be a gene that like If I remove gene a the the the cells are still viable if I remove gene b singularly the cells are still viable But if I if I remove I mean sorry if I remove Any of them the the cell are not viable, but if I have both of them the cell are viable Like if they look into these so this is some kind of epistasis Yeah, you know it was actually to admit that I was thinking about this the morning and I wanted to look this up So, you know was this done sequentially and in what order was it done? And would that determine what genes were essential or non essential? And I don't know if this was all done at like one time where you could have different Transposons that were mutated across and you just screened all cells or if it was literally done one gene at a time And I have to I'd have to go back and refresh my memory on that But but I do think about And sorry, maybe you mentioned that but do we know the function of all these 500 genes that are left? What what are the functions? Yeah, I mean there's uh, I I can't summarize all of them because it would be about You know roughly 500 genes But anything that I mean there are a lot of things that are important that are removed and that are Especially important for for for evolution cells like repair of mutations Any kind of accessory genes that aren't essential for life? And so there's a long list of them And I'm going to also say that there are some genes that they did define that are quasi essential Things that affected the performance and the ability to work with this organism in the lab That the choices were made to retain them even though technically you could get a cell To to to live and divide. It just wasn't really made it made it difficult to do experiments with and one of those genes I'm going to talk about is really important is involved in cell division and the evolution of cell size So so but there's only like maybe a dozen of those genes Okay, so so I want to uh, you know, what are the questions that we had um, so So I meant to go back to this gordon research conference where I'm talking to Clyde Hutchinson And it's just this marvel of being able to make this cell But as soon as the cell was made it was put in A glycerol or whatever cryopreservant and placed in minus 80 They didn't want to domesticate this organism. They wanted to keep it the way it was the way they constructed it I think this is the a bit of the mindset of an engineer, right? You want you've created this to the device a synthetic cell You want to maintain it in its pristine form? So I asked Clyde, well, what would happen if you let this cell go What would happen to it? And I don't think it was a question. Well, he told me it was not a question that had ever come to his mind So there's a lot of money a lot of effort huge teams that went into being able to make this minimal cell But the the follow-up question was one that particular for me As an ecologist and evolutionary biologist is how is this organism going to behave perform subsequently when It has no redundancies no degrees of freedom, right? And what's going to happen as soon as you start to grow this organism? Invariably, there will be mutations that will arise And it's going to hit only an essential gene So if those are deleterious, which most mutations are Then you would expect that perhaps this would compromise the the robustness of this cell and perhaps even leave populations to extinction You could imagine that would be the case So I was really interested in this idea and I talked to Clyde immediately on the spot and I was like can you send me these Cells expecting that maybe they wouldn't be so keen to do that. It's like sure so within One month they had we had got our USDA permits and we had JCVI send us these cells And we wanted to address two two fundamental questions one could be the first one is related to mutations That arise in these populations. There's two reasons to know this if you're a synthetic biologist You've gone to all this work and you've got this minimized cell and if mutation rates are really off the charts Then maybe this will kind of erode the function of the desired function and performance of your cell So that would be a bit bad for for practical purposes But it's also important if we want to understand the evolution of a minimized cell to know what's going on in terms of Inputs of mutations because this could be really important for for driving the evolutionary dynamics of this population So we wanted to characterize the rate and the spectrum of mutations that are Arising and whether or not that differs for an organism that's undergone extreme genome streamlining Question one question two is already kind of suggested is that because there's no degrees of freedom and there's no redundancies Then this could potentially constrain adaptation and from other systems outside of microbes We know that a lot of adaptive evolutionary changes occur in in non-essential and accessory Genes and we've eliminated all those So so we could hypothesize we did hypothesize that the the ability In the way in which the minimal cell adapts Would be different from that of the non-minimal cell from which it was synthetically derived Yeah, just out of curiosity um, did they look into the fact that This this minimal cell basically lacks a lot of genes. So does it have a growth advantage? Um, so that could be an alternative hypothesis. That's that's a that's a good question And that turn I don't want to give away the punchline just yet But maybe there are some advantages and it's hard to I don't have data to To speak specifically to that question, but you might imagine that there there could be certain things Well, the interpretation could be the alternate. This is good. So Maybe an organism is constrained through the process of genome streamlining But there's also we learned yesterday a lot of things that are going on in cell a lot of random proteins that are interacting And that perhaps with less material, uh, maybe this opens up some opportunities for adaptive evolution Due to the, you know, just just noise and complexity in this cell That that that was brought up in the review process of this work We needed an alternate explanation for the results I'm about to show you Yeah So we I'm going to talk about two experiments The first was, you know experimentally, maybe these aren't as exciting but this addresses the question about Inputs of mutations. We did what it's called a mutation accumulation Experiment and classically in evolution what you would do is you you breathe inbred lines You passage them over time through your bottlenecking under conditions that are supposed to relax selection And allowing for neutral And mildly deleterious but not lethal genes or mutations to accumulate through this bottlenecking process over time And what this allows you to do is to estimate the the spontaneous mutation rate Okay, so this has been done for a lot of other organisms It's also been done for microbes and the way you do it is you grow up a population of cells And you plate them out on a petri dish and you pick an individual one random individual colony from that plate You grow it up You plate it out again pick a random colony Grow it up and so it's going through repeated bottlenecks, but you're also maintaining Populations at very low sizes so that if there are any deleterious mutations that arise It's not going to be experiencing selection from its neighbors So we did that for 87 replicate lines or lineages for the the non-minimal cell and 57 Lineages for the non-minimal cell we passage those for about half a year And then at the end of the experiment we grow those cells up extract nucleic acids and we We we map the reads back onto the reference genome and we we Classify mutations And we know how many divisions and generations occurred during that time point so we can express the mutation rate In terms of mutations per nucleotide per generation So this figure doesn't at the surface look all that interesting But there's two important things or interesting things to me that that emerged from this One the we expected that mutation rates would be higher in the non-minimal cell because we've removed all of these non-essential genes that are important for Correcting mutations that would arise like mismatch repair And that's not the case the statistically the mutation rates here are comparable But the thing that is sort of noteworthy is that the mutation rates were exceptionally high These are the highest mutation rates are recorded for any free living organism ever So there's a lot of a lot of mutations coming into the population The non-minimal cell was the original ancestor that they had in the in their study that they deleted everything The so the blue dots are referred to the non-minimal cell and that was synthetically created But it contains all the genes before prior to minimization and the red are the ones that were also a synthetically constructed cell But it had all the non-essential genes removed So So you can't so Yeah, and so now we're just measuring the rate at which new mutations arise in this population over this experiment and The rates are high, but there's no no effective genome streamlining The the mutation rates are the same they're comparable And so by mutation rate Do you also take into account like synonymous mutations stuff like that because this you would expect not to have an effect on sort of the viability of the genes right so Yeah, so there's no we're kind of not really worrying about fitness of the the mutations We're trying to create an environment where we passage them where the those Cells that arise that have mutations that are not lethal will be able to be continued to be passage I'm saying that because I thought I understood from what you were saying that you would expected the minimal cell to have A lower mutation rate due to the fact that they only have essential genes But of course if you're looking at synonymous mutations, they shouldn't play a role right Because you wouldn't change the function of your gene Since the synonymous mutation are not changing the amino acid that end up in the protein So you're saying silent mutations. Yeah, these are total number of mutations. Yeah So maybe it's maybe you would observe a different effect. Yeah, we do. Yeah, okay Yeah I think I understand Just to be sure that I understand so this is like To be like More precisely is the non-lethal mutation rate. So the mutation rate that gives rise to mutations that are non-lethal. That's right Independent of their fitness effect That's right And and we don't I don't know of any other really good ways to measure like the mutations that would arise that would Be lethal but but presumably they're happening and maybe they're happening at a higher rate in the minimal cell Hi, so here you just mentioned that both of these are synthetic cells Yeah, so was any comparison done with the non-synthetic version of the non-minimal cell We didn't do that You could imagine that there's a lot of comparisons you could make There are other micro plasma cells that could be used a comparison But we we thought that for this experiment We just took the synthetic non-minimal cell and the synthetic non-minimal cell So is there a possibility that the mutation rate is similar because these are both synthetic? Yeah, I mean it could be I mean the content of the non-minimal cell is as I understand identical to the non-synthetic But there could be perhaps something that happens in the process of synthesis But it's not surprising that the the mutation rates are high. That's pretty typical for Organisms that are undergoing massive natural genome So so it wasn't surprising to us that the mutation rates were high And in fact the reason why we probably don't see an effect of the non-minimal cell is that the baseline Mutation rate for this organism is just really elevated because of its small effect of population size Sorry Jay. I realized that I didn't understand How the ancestor cell in what way is it synthetic? I'm sure you said it Yeah, so what what you do is it's the same process. It's just one cell has It's been in silico the construction of the genome contained all of the essential and non-essential genes of the wild type and the non-minimal cell has Was synthetically created without any of the non-essential genes. Yes. Thank you. Okay all right so Jay for A lot of population genetic work. I didn't ask this question. I was in your lab and doing this but For a lot of population genetic work. It's not so much the per base mutation rate that matters or even like the genome Scaled rate. It's that per base rate times genome size Times the size of the population and there's the input of mutations per generation. Yeah, and so I mean You do evolution experiments. You just say that the size of the population is effectively the bottlenecks population Yeah, but so once you if you were to take these quantities multiply them by genome sizes and multiply them by the typical population size For a given environment, how would those effect how would those compound parameters compare for the two bugs So I think population sizes are comparable In a subsequent experiment. So we we think they both get to about 10 to the seven cells per mil The the thing that maybe When you ask this question that I queued in on was is this these are reported on per per base pair So the genome of the non-minimal cell is twice the size So maybe if I understand your question And I'm not sure if this is expressed here is that the There might be a factor of two difference. Right. Yeah Yeah, so Yeah, I don't think that would be I don't know if that would fundamentally change We're looking at an arithmetic axis. So maybe it would change things A bit. No, that's that's good because then it means that for evolution experiments They're roughly in the same parameter regime for that Important variable. Yeah Okay, um So this is this is great to get some questions. This was like I was like is this really interesting I think it's interesting. It's just like we have to it's a prerequisite We have to figure out like what's going on if you know mutation inputs are different So the other thing that we we we did is we We want to know about the spectrum of mutation and so for people who aren't familiar with that nomenclature We're just looking at the composition or the differences and the types Of mutations that are entering these populations And so this is where we can get categories of insertions deletions and single nucleotide mutations And the proportion of those mutations arising in the non minimal and the non in the minimal genomes are the same But we see that most of the mutations entering the populations are single nucleotide mutations about 90 So we wanted to look at those in a little bit more detail And so what we have now are three panels that are characterizing the classes of different types of bias that can arise In types of single nucleotide mutations so in the middle would be just like there's no no effect There's changes, but there are neutral effects. There can be mutations that give rise to gc bias We don't see many of those most of the mutations that we see result in at bias This is pretty common among bacteria and archaea in general So that wasn't surprising but in this panel over here Where we have c's and g's going to t's and a's we see that there's a 30 fold discrepancy in that type of transition mutation So we wanted to try to see if we could come up with some explanation for that discrepancy There's a mutation. There's an essential gene called ung which is dna glycosylase So when cytosine becomes deaminated, so there's a loss of an amine group This can be lead to a replacement of that with a uracil and then in subsequent Rounds of dna replication you can get adenines that will fall into its place So so that that gene and what it does and how it corrects for these types of base Paraxition can is consistent with what we know about at least one type of DNA repair gene that was removed from the minimal cell But the conclusion from all this is Effectively two one is that the mutation rates are really high We're estimating that every single site in that genome is getting mutated four times per day So that's a pretty big mutational burden or load potentially And we do see some bias in spectrum. There's a lot of more single nucleotide mutations that are going to give rise to at bias but By and large we didn't see like fundamentally that the genome and the mutational input should be like dramatically altered Through the process of removing 50 percent of an organism's genome So so at this point we're now thinking about moving away We've characterized what's what's coming into a population do the effects of genome minimization The second thing that we wanted to now explore is is to what degree Is this minimal cell able to undergo adaptive evolution? And that's different than what we just talked about with mutation accumulation When we're trying to estimate like inputs of spontaneous mutations So so now we're going to move into the direction of trying to understand how this organism after undergoing genome streamlining Is it constrained in its ability to evolve and adapt? That's the question Are there genes that stay That are not mutated or genes that are less that mutates Less than other genes. So there is a distribution of rates of you. Can you find ideas out? The question is about the first bullet point Each size so it's average That's the average. Yeah, so we know how many mutations what the mutation rate is We know what the size is and we assume that mutations are random. We're occurring at random So this is the average is the estimate for the average number of time each gene should get a mutation But I guess you can check whether this average is a good characterization Or is there genes that are less mutated than others? That's something you can check because perhaps there are genes that if you Mutate this gene the cell actually die Yeah, we talked a little bit about that. I think that's a hard thing to know to estimate what Are lethal mutations, but I have an experiment where we're going to do an adaptive evolution experiment Where we sequence and we can look at where what genes get mutated so But this is just to be on average. We know what the total rate is. We know what the size is How many times on average assuming mutations are random should each gene get Hit per day and it's it's high Never ever where no degrees of freedom. These are all essential genes. So what the question is what are the consequences of that? So I described this two days ago It's a it's a similar And it's a common approach of using experimental evolution where instead of maintaining populations at low sizes We allow them to reach Large population sizes around 10 to the 7 per mil and then every day We just passage one percent of those cells into a new And into a new vessel And so there under there is some bottlenecking that's going on But but the population sizes are large and we know that large population sizes should favor Deterministic changes and adaptation of these strains as they're experiencing mutations in the population So we started off in the beginning with the ancestor. We created four replicate populations. Perhaps we could do more And then we just passage them over time and about every 30 Generations or so I think we would take a population and put it in a tube and freeze it so that we could resurrect it later We could characterize its its fitness and other phenotypic properties and then At the end of the experiment we had extracted DNA and again using pool population sequencing We could look at the mutations that arose under these conditions that should favor adaptive evolution So that's the experiment Is that is everyone okay with those that that type of study approach? So I think this might be I don't know maybe the one of the main Results from the from the from the paper that we worked on this is work that I did with Roy Moser Reicher Who's a phd student in my lab at the time? And so what we did is we wanted to measure the relative fitness of these strains And I'm just showing you what the relative fitness was at the beginning of the experiment and what it was at the end of the experiment We have other higher resolution ways of measuring fitness And and that's sort of relevant because you might ask well what's going on in between those 2000 generations So about three 300 days we passage these cells Are those relationships of fitness change over time? Are they linear? Do they take on some power law relationship which has been described in the long-term evolution experiments of rich lendski, etc And you'll have to trust me. I can show you later that the the rates of Change and adaptation over time are linear So this is just a simple way of looking at what happened by looking at the beginning and the at the end of the experiment And there's a couple things I want to emphasize number one There it is So we do these competition experiments where we have an independent Reference reference Strain, which is the non-minimal cell and it has a red fluorescent protein fused into its genome So we can put that marker strain in combination with other strains. We start off at 50 50 densities We let those incubations go For a day and we measure the the frequencies of those cells after 24 hours And from that we can estimate the the fitness of those organisms obtained from different conditions So you'll you'll note that the relative fitness of the non-minimal That's what we started from its relative fitness is one, right? So everything is kind of referenced to that that starting point and if we look in this column here at the ancestor What we can infer from this is that genome minimization Resulted in a very sick cell So there's a 50 reduction in the relative fitness due to reducing the genome of that organism down to its bare essentials And that's consistent with just being able to do work with these cells in lab It was not easy like they're just not happy when we got them shipped from jcvi It was really difficult just to kind of tweak the right conditions in our laboratory to get them to kind of to grow You can't measure these things with optical density So you have to come up with different ways to quantify cells. We use flow cytometry here But there's a big cost to genome reduction and then what we see Observation two Is that all of those fitness costs are regained in 2000 generations? So here i'm comparing that symbol to this symbol, right? There's no difference quantitatively 100 fitness regained in 2000 generations. So This is not consistent with our expectation or prediction that That the minimal cells ability to evolve would be hampered Seems to be gotten better very quickly And now If you will those lines that connect those points those vectors those aren't statistical lines Those are just for i but we can see and I have other data to back this up as the rates of adaptation are the same statistically If anything in this this figure they're actually the minimal cell evolves a little bit faster than a non-minimal cell So that's sort of interesting um and Maybe this isn't a good hypothesis, but what we were thinking at the time is like well, maybe they're adapting Via selection on the shared genes, which are only essential That would be different from what most people in evolution think the adaptation happens due to mutations and in non-essential or accessory genes But the given the the common pattern of those slopes Let us the hypothesis that maybe despite genome minimization these two cells are evolving via the same mechanisms So we we we sequenced the genomes of these organisms at the end of the experiment and the results suggest that that's not the case So there were about 14 in one group and 18 Genes in another that we call putatively under positive selection Those genes were ones where the mutations went to 100 frequency in the population And we also used a statistical approach where we This is similar to some work that that will has done. I don't know if you'll talk about it later today where you can Take into account gene size and you can create a random distribution of mutations by throwing them on the genome and creating an expectation of How many how many mutations you would expect to see in a gene under a random null model? And then you can compare that to the observed number of hits on each of the genes that you saw And from that we see that there are a lot of over enrichment of hits on genes under certain conditions We we we infer that that's a signature of positive selection So that gives us 14 versus 18 candidate genes It may be important in the adaptation of these cells over the 2000 generations problem To looking at that pattern of shared adaptive changes in those rates what we see is the genes are Completely different and we're only comparing essential genes here so they're Divergent mechanisms of adaptation that are arising some of them fall into similar categories like dna replication and transcription in both groups We thought that maybe the minimal cell where we would see mutations and transport Etc. The membrane we don't see that we actually did not see that in the in the minimal cell We see mutations in atp even have those and lipid metabolism So just you know divergent ways in which these organisms are undergoing adaptation based on sequence data It's very interesting. Um, so can you say or guess if the mutations on and the minimal cell are loss of function? or just adaptation of the say Function level Um I haven't thought about this. So so let's work through it. So if it were a loss of function in an essential gene That would be problematic, right? Well, I'm thinking again that this is a different environment than the one that you use to define a minimal cell Same media. Uh, I mean, yeah, I mean, maybe there's some subtle differences And so what you're saying is that we defined essential genes In san diego at jcvi under one condition that maybe in the environment that we're working in there are genes that are No longer that could be actually non-essential Yeah, it sounds like for example, they were using cultures like on on uh ego plates to define the minimal cell Okay. Yeah. Yeah. I mean, that's reasonable. I have we don't we don't have any information on that but we assume that Yeah, I did not consider that issue that that that the changing the conditions or the labs might change to what degree A gene could be defined as what is essential or non essential So what you're saying is that there could be genes that are mutating that are actually Uh, we call them essential, but they're actually not essential Yeah, I wonder for example if you can like say something about the stability of these proteins taking the mutated sequence And then see if they still fold for example Yeah Yeah, we haven't done that but um, and I haven't thought about that question before so thank you You know just along these lines. So you you didn't see any nonsense mutations You know premature stop codons this sort of thing Um, I think maybe we did and I would need to go back into the the tables In the supplement to to to know for sure. Yeah Because that would indicate that the genes That would be a loss of function probably, you know, okay Yeah, I'll have to think about it some more. Thank you You declare just to get stared back at you. We'll wait this out and see where we can go. Um the the point is that uh We we thought well, maybe there would be some shared mutations involved in this common Quantitatively common adaptive change and it's not so they're evolving divergently And it doesn't appear right now that the minimal cell is hampered in terms of its ability to adapt It's just doing it via different means At the same rate So there was one mutation so I'm gonna this will be the last part of this story. I think Um I mentioned that there was these quasi essential genes and one of those is f t z f t z And this is a tubulin homologue That orchestrates the process of cell division And uh when this protein these little Monomers polymerized it forms this dynamical z ring at the center of the cell And that z ring recruits other proteins and other types of bacteria would be recruiting peptidoglycan We don't have peptidoglycan and mycoplasma Um, and so it creates this ring which is involved in regulating And the division of the cell from from one cell into two cells And so when that gene Nonessential gene was removed the morphology and division of the cell got a little bit weird So jcbi said we're gonna we're gonna keep it even though it's technically a nonessential gene Um f t z has also been observed to have effects when mutated in other populations of bacteria and archaea as well So there's some good sense that this is a something that's affecting cell size It was retained and so we thought it might be good for us to kind of it was it was hit in every single replicate population across all of our All of our minimal cells and non minimal cells So I convinced roye that that we should um, I know Europe people are really into microscopy for some cultural difference in the united states It seems to be harder to get people to look under microscopes at least in the field of ecology and evolution We're like wanting to sequence and sequence. It's like roye. Can we just look at these under the microscope? Please And so and roye's always responsive, but I have to use Really strong logic to get him to do things. It's like we have to do this because if we do this then it will Okay, okay, we'll look at the cells underneath the microscope Maybe i'm exaggerating a little bit, but not too much And so we we did that and so these are uh, the the sizes the diameter of these cells Measured through scanning electron microscopy. We're also used epifluorescent microscopy microscopy and So the non minimal cell this had been reported before in the science paper in 2016 The genome minimization leads to a reduction in cell size. It's a pretty small. We're trying 300 nanometers, right? And and then we can see that there's a five fold increase in cell size over 2000 generations And statistically did we did not see an increase in cell size in the minimal cell? So there's a couple things so so now we've I've told you up to this point that we see no constraints Uh On evolution due to minimization, but now we're seeing a big phenotypic change and something that's really important for resource acquisition Whole organism metabolism Efficiency and how everything scales with body size, right? And so we see that adaptation for the non minimal cell It goes up and it's accompanied by an increase in cell size But in the minimal cell we see increases in growth rate and adaptation And it's not accompanied by an increase in cell size. So that's sort of a difference that that that struck us as being Sort of interesting and potentially important So what do you do in these because we have a synthetic organism? We could get the mutations in ftsc We told jcvi about them So can you take those mutations that we observed in the experimental evolution trial? And put them back into a clean background of mycoplasma mycoidesynthetic cells both the non minimal and non minimal cell And we'll see two things. Can we Recapitulate the effects on cell size and also test whether or not these genes and the mutations that we saw that were Putatively adaptive are actually adaptive by doing competition experiments like we did earlier So i'll show those results Next on the left panel is cell size. So we put the wild type cells On the non minimal and minimal We can see this is the same sort of data that we showed in the previous slide, but we put this ftsc mutation in there We see that cell size increases. We can recapture 60 of the change in cell size with this one mutation And if anything putting that mutation Back into the non minimal cell it does not increase size if anything actually decreases cell size But in both cases that mutation went put into a clean background results in higher fitness So this means that they're adaptive and this means that we can recapture the changes in cell size with one Essential gene that's involved in cell division regulation of cell division Yeah, um, I didn't so these are non evolved. These are the the cells that we started with at the beginning Yes, they hadn't under none of these cells had undergone evolution the the adaptive evolution We're just gonna step back. We're gonna work with the non-evolved the ancestors and then we have mutants So i'm just calling them something different here. That's The wild type is the stuff that you started with that's on yeah And then the one on the right hand column is the the cell that had the observed ftsz mutation from the Um from the evolution experiment Uh crispered back into the wild type. So we've we've controlled for everything. There's just one mutation That's accounting for these differences. No, sorry. I missed that. Uh, so on the right hand side, uh on the You know the right side of the picture of each picture. Yeah, you have The cells that you evolved through these 2000 generations No, no, you're just looking at a single mutant It's a the mutation that we observed from that experiment We we got that sequence and we went back to the original stocks and we said synthesize this mutation take out the ancestral ftsc And we're going to put in the observed mutation that we saw from the experimental evolution back in Without any of the other baggage and other mutations that arose during so only a single mutation single mutation is being compared between left and right That's right Sorry if that wasn't clear I've had this question out a few times but now I'm willing to ask it Is evolution just maybe a better synthetic biologist than we are? Yeah, I think that's um Well, I have some thoughts on that topic. We can talk about it now. Um I'm not an engineer and I'm not a synthetic biologist But I imagine that sometimes people we think that we can create a certain organism that can do a certain thing The way it should be and I would say that maybe we can leverage natural selection to In combination with synthetic approaches to create organisms that can do things. I mean, uh, you know, somebody a couple years I just got the noble prize for using directed evolution to to kind of improve the You know the performance of enzymes, right? So, um, yeah Yeah, I think that that that would be one thing that you can infer that Yeah, maybe in the in the field of synthetic biology We should be thinking about not trying to preserve as a snapshot in time and organism and keep it free from the forces of evolution Because that's inevitable that this is going to happen And I was saying in some cases at least for certain functions Whether or not fitness or cell size is important to you Then maybe those are things that we can use in concert with synthetic approaches um So here fitness do you think is mostly due to the fact that the Well, the non-minimal grows faster than the Than the minimal Um, so so which one grows faster the non-minimal? Yeah, they Yes, that's right. That's that's just it. Okay. Um So you're talking about the panel on the right. No, no, no, I'm actually thinking about the fitness fitness. I'll say That you showed before um No, because it was because in in that case you grow them for one day, right? So the only important thing is how fast you grow in in this That's right. Yeah maximum growth rate Yeah is basically what we're using for fitness But but the head-to-head competition experiments here do capture something a little bit different But we've also done a lot of experiments where we just measure kinetics and the the growth rate maximum growth rate of these strains those two approaches match up Is that is that was it? Yeah, I was just thinking What exactly these fitness I'll say reflect and yeah, yeah, these are also these head-to-head competition assays where we're competing experimental strains against the m cherry labeled. I think I said red for Russian purpose. It's m cherry labeled reference strain So measure its ability to increase in frequency relative to a labeled strain Can I ask you another question? Yeah so you said that these Minimal cell is very picky in terms of Where they grow so So you have to be basically calibrate the media And then and then evolve it in this media Do you think so let's say that you find another media where it can grow. Yeah Yeah Yeah, I mean we can talk. Yeah, I think there's there's a there's a sub question there that you're getting at The media issue is like one if you want to talk about media These are hard to grow and the media is really expensive. It's got like fetal calf serum And we're supplying it with everything that it needs to grow Because it doesn't have the genes that are needed for that. I think that's important for the types of questions There's now defined media just last year This group from jcvi published a paper where they did a bunch of other things but like nested within that They've actually defined they've developed a defined media Which I think could be really important for flux balance models and like maybe simulating mutations and what we would expect to see So that's that's Interesting and if we continue to work in this direction, that would be something we want to include But I think you're asking another question too, which might be like, well, what if you tweak the media a little bit? Or the environmental conditions, would we see these same results and that's something I've been Thinking about a bit and I was inspired actually by the the talk we had yesterday Eric's talk was thinking about, you know, what happens when you Think about noise and populations and how noise is propagated Under an organism that's undergone a massive, you know, genome minimization. I think Eric said that there's only You know three transcription factors in these Minimal cells and so I've been thinking a lot about how that might allow organisms to contend with And evolve in different environments. So it was a really nice conversation we had yesterday Thanks Yeah, so so I have a very basic question about this plot. So the the Mutation that the new engineering the ancestral train is a mutation that occurs in the non-minimal or in the minimal cell That's a good question too. So what what mutation was it and I wanted to go back and look this up too There are different mutations that arise in the natural selection that hit fdse But there's some subtle differences in what those mutations are You know, none of the mutations are going to be exact. I don't know which Like consensus sequence we took it from it would have been either from the non-minimal cell or the minimal cell But we didn't we only pick one And put that back into one of the cells. I have to go back and check on that. I've just been thinking about that um, okay, so um I just want to wrap up. I think we're at the top of the hour now. Anyways, uh, you know, some people don't like these these examples of thinking about fitness landscapes, but You know 100 years ago We started to to see these these ideas being presented in evolutionary biology where you have genotypes That are mapped out on this fitness landscape that are being shaped not only by the genome but also by the environment and I think sort of Can imagine that this non-minimal cell over 2,000 generations is marketing up some kind of fitness landscape to an increase in its its fitness And then we did a genome. I didn't do the genome JCVI did the genome immunization the reconstruction of the synthetic cell lacking half of its genome And we can see that it's really sick and so it's going into this trough on the fitness landscape and the question was Can it get out? And if so, what is the path that it takes up that mountain, right? And so it seems like we don't know what the fitness landscape actually looks like It could be really rugged and there's multiple peaks that are being climbed and the two there's two high peaks Here i'm depicting that there's just one one peak But the the cell seems to not be struggling with getting back up to that fitness Yeah, so so this is a generalized model of like some some ideas about how Cells might evolve after undergoing genome minimization, which is really quite common We see it in a lot of endosymbiotes of You know things like aphids but also in the global ocean Some of you may be familiar with the sar 11 palagibacter ubeak strain One of it's the most abundant microbe on the planet has an extremely small genome So I think some of these findings are not only important for environments of host associated microbes that we know undergo genome Reduction but also in free living environments where organisms are important for carrying out important functions at global scales We've already addressed this issue about What what does evolution mean for synthetic biology and that instead of thinking about this as a hindrance or something that's Moving us away from a desired product. It might actually increase the productivity and yield of of of synthetic organisms Um, and I mentioned a little bit about like the idea of models of origins of life And why I don't think this is really probably the organism to be asking those questions But I think this approach could be used with other organisms To gain insight about, you know, what what what a progenitor cell or primitive life form may have looked like Um, you know three and a half four billion years ago And I think we also have this ability to think about general questions in in biology related to the evolution of cell size Why in some cases Adaptation is associated with an increase in cell size in other cases. It's not And we're thinking about also like building up the next step would be we've now Investigated what happens with genome reduction. Can we Recomplexify an organism and I've been really thinking a lot about genome duplication And how that might be some a question where we can look at neo functionalization and diversification of microbes in different environments due to Relax selection on an additional copy of a gene while there's preservation and Stabilizing selection on another copy of gene and we have the ability in this in this This this platform of synthetic biology to actually construct a gene Where there are two copies and we can concatenate those genomes and perhaps use things like barcoding to test some of those hypotheses So there there are some financial Things that i'm thinking about but you know taking these types of experiments and asking questions about diversification and duplication as as Organisms genomes become more complex. So the flip side of of simplification Anyway, um, yeah, so just importantly. I want to point out that you know all this work was led To a large degree by rory moiser-reischer who was a phd student lab when will was in the group as well Along with some people at the venture institute arizona state and usgs Thanks I think we have time for a couple of questions Yeah, yeah, thanks. Yeah, that was a very interesting talk and um I'm actually a bit confused trying to reconcile Your last conclusion. I mean some of your conclusions including the last one of that waddington landscape picture Okay, yeah with a slide that you had shown earlier So in this picture you said you seem to suggest that Both the blue and the red are starting at the same point one the blue is going up the red is going down slightly down Right, whereas that earlier picture that you had I thought Where you have plotted the relative fitness I thought that um They both started at a different point. They do so You know the red one started at a fitness of 0.5 and the blue one started at a fitness of point At a fitness of one that's right by definition. That's right But both increased it increased their fitness by the same amount. That's right, right? so How does how do these two pictures square with each other? I tried to capture that that's what exactly what I was trying to reflect in my cartoon and drawing But maybe maybe it wasn't effective so that they both get to the same point at the end in terms of their Fitness or approximately But there's initial the cost that you're describing the the point five Value of relative fitness versus 1.0 was reflected by that red arrow where it's moved the genome reduction the dash red arrow Brings you down in the fitness landscape. I see that was what I was trying to convey I was yeah, I Yeah, I didn't know I've mixed feelings about whether or not this is a good metaphor for what's going on in our experiment, but No, okay now that you've clarified Okay, I'll try to make that clearer. Thanks. Yeah. Thanks, Sanjay Maybe a little bit long same lines Not about itself, but if you now evolve the minimal cell after your 2000 generations And the non-minimal cells at the ancestral string that both have a fitness of one Yeah, do you expect that linearity to continue? I don't know. So once they're sort of equal, you know, what happens? So, um, you know the the Classical example to look to is rich Lensky's work with E. Coli where he evolved You know 12 populations over time and The the fitness gains saturate over time and have been fit to power law distribution Some people debate whether or not that's the right model for those But the point is is that in the first 2000 generations, which is the time scale that we have here You see the most amount of adaptive evolution and after that it becomes more incremental Over time, but still goes up. It's just a lot smaller So the question is like what will we expect with this organism a synthetic organism where 50% of its genome? Will it keep on increasing linear? I mean we had to put a stop to this so rory could eventually graduate But yeah, like, um, you know, I could imagine With the right person we could continue to ask these types of questions. I don't know They're linear though. There's no there's no effect that that within 2000 generations is any sign of leveling off That's really fun. Thanks. Um, we had this, um Jay and I had this fun talk on the way back from lunch yesterday talking about, um, de novo gene birth and We were so sorry the genomes sizes just constant throughout all of this the only thing that's going on is Just nucleotide changes. Yeah, we didn't see there are some There are some deletions in the non minimal cell about nine most of them are so not in the minimal cell But in the non minimal cell and one of those is sort of big I don't remember how big but a lot of them are like eight of them are small and one of them is big This also seems like a good place to look for, you know, do you know of a gene birth? So I wonder What do you have to do weight or you have to mix in some other cells? You have to feed in some other DNA I'm wondering how you can get some some genome growth. Well my My approach, which is artificial would be to just create it from scratch. We have this chassis this platform that we could use to add back genes And instead of waiting to see them or not having the control to know what genes so again also in the lte I think we talked about this yesterday on the way back from lunch. They did the What is the The citrate utilization example where there's an inversion There's a duplication you get this dosage effect that allows for the initiation or potentiation And then there's the refinement steps that happen later So rich had to wait. I think it was 65,000 generations for a citrate utilization gene to undergo that process of duplication and inversion I don't have that much time left to do that so Yeah, so one thing that's kind of interesting the synthetic biology is you can just like all rules are off Right, like we doesn't have to necessarily mimic something that we've observed in nature Uh, let's let's try to build something up and maybe we can test general principles, even though it's not consistent with whatever what we see Yeah, maybe one more comment. I'm also getting really excited about the duplication idea um, there are people who Um have tried this um, especially when they try to say take some bacterium and try to um Give it the ability to degrade linin lininus is very difficult to degrade compound that's found in uh, I think wood and and all these shrubs and um So they do experimental evolution for that and they also need some sort of acceleration of it and they Are looking for gene duplication. So what they do is they create um Tundam repeats so they take the same gene and they sort of glue it together in a specific way And the fact that it's a tandem repeat then makes it much more likely that um, you can increase the dosage of that gene And in this evolution. So this this I think ellen nightliffe, for example, that does this um You see that it's like a cassette. So you start with a tandem repeat. That's two genes. It goes up to like eight Something happens and then it goes down again So I have been wondering like if you were to do this synthetic genome, um, um duplication You could just take two pieces of the two copies of the genome and concatenate them Uh, or you could do what you're suggesting is put copy one copy copy a one copy The second copy in tandem together around and I think I have to think more about what Yeah, so maybe that would A promote dosage effects. Yeah, so I think what happens is they create an artificial structure of two genes that are fused in a way that makes um Additional copies of this tandem much more likely at DNA replication Yeah Cool. Yeah, I think I mean it makes me a little nervous because like if you're gonna build a cell and spend the money to do that You better have thought through all these things and I could imagine there'd be a lot of things that would be off my radar When building an organism, right? Oh, I didn't think about that. Yeah Uh There's origin to replication and yeah, and then there's also limits to how much material you can really just pack into these cells So they're really small We're thinking a lot about we've been talking about ribosome content in cells and growth rate and efficiency and And the rate at which chromosomes are replicating relative to the division of the cells. So I think there's a lot of Things to consider Um Do you think that if you throw a plasmid around they will be able to Take genes out of the plasmid or do you need do you need something that they don't have? Um, do you think that these cells would would take up a plasmid? I don't know If you throw some DNA in the environment and uh to acquire something Through horizontal gene transfer. Do you think they will be able to do that or Or If there is some specific machinery that they don't have so I understand that this organism doesn't undergo homologous recombination Um, but I don't know that and and there's no cell walls, but I don't know what how that Effects its ability to take up exogenous DNA through transformation And I don't know I don't have no knowledge about whether or not my go plasmas have or can obtain plasmids I have no clue Yeah, so I hadn't thought about that Asking out of I don't know anything about my go plasmas Okay Okay, I think we can thank jay again for the three fantastic lectures