Jack Szostak: Reconstructing the First Cells

and thanks for inviting me here to take part in this conference so I think yeah the origin of life is is one of those fundamental questions it really catches people's interest it's also it's kind of a different and difficult area of scientists to study because you know we can't ever really go back and find out what really happened so the best that we can really hope to do is come up with plausible pathways that provide reasonable explanations for how life might have got started on the early earth and so I think one of the best ways of doing that is basically by extending the approach of synthetic organic chemistry and just sort of building structures that behave like simple cells or at least what we with our limited imaginations think simple cells so the first cells might have looked like okay so before I get into that experimental approach though I do want to just set the stage by reminding you of what I think is really one of the major revolutions in biology in the last two or three decades which is the extent to which we see that our planet has been colonized by life and even the most unlikely environment so here you see the classic picture of a deep sea hydrothermal vent with abundant life living at high-temperature high-pressure there are these acidic systems Rio Tinto but also acid mine drainage sites going down to essentially pH of zero they're not sterile indeed there's abundant life here we see life in the rocks this layer of cyanobacteria under the rock crust but of course we also know that there's life kilometers down in the crust and porous rock so really every where you can imagine that life could adapt and spread it has so the point of all this though is that this tells us about the adaptability of life it doesn't really say anything about how life got started and the real issue there is what are the range of conditions that would have been compatible with the very beginnings of life that transition from the chemistry of the early Earth to the simplest and first biology and to me coming from a background in biology that key step is the emergence of Darwinian evolution so the question really is how do we get a chemical system to start exhibiting Darwinian evolutionary processes and thus behaving like biology okay now that question has also really been set into a broader stage through the just incredible recent advances in astronomy and so the detection of exoplanets by many different methods now but most recently in most amazingly sees the Kepler mission which has now a candidate list of almost of over 1200 new exoplanets the extrapolations from the statistics of the Kepler findings suggest that in our galaxy alone and pointer here did you did you steal the remote Craig oh wait a minute wait wait I I got a present I got a present I want it no no I want to try this okay right but you know I got this as a present and I want to see how it works ooh new blue laser fighter really cool okay so in our galaxy alone and extrapolations are that there could be on the order of 500 million earth-like planets orbiting sun-like stars and I think we just all would really like to know you know if there's life out there or are we alone you know did life only start here you know Craig mentioned this starship project which I hadn't heard of before but I think you know even if that's incredibly successful it's going to be a long time before we get direct observational evidence as to whether there's life on any of these other planets and so question is what can we do in the meantime to address the question which really comes down to whether this transition from chemistry to biology is something that's easy you know just a series of simple steps that we don't quite understand yet or whether there's some bottleneck something that makes that transition incredibly difficult and therefore rare and so I think the thing to do is just go in the lab and start doing simple experiments and trying to work at pathways and see if we can get clues as to how life started and and and hopefully some answers to this question now there's a related question which I also find really fascinating which I probably won't have time to talk about today but that is this question down here you know if there is life there is it gonna be just kind of boring lay the same as what we're all familiar with here you know life based on water with some kind of nucleic acid to mediate inheritance protein lake molecules to carry out most of the cellular functions or could there be really really different kinds of life different molecules carrying out these these functions maybe even life in different different solvents again something that I think is actually quite easily addressable with simple experiments long before we have direct observational evidence okay so let me get jump into the things that we're actually doing in the lab to try to answer these questions and the basic idea is to to try to build up simple molecular systems that start to show in a very basic way these properties of Darwinian evolution and biological behavior so I don't know if you call this synthetic life or artificial life but in any case we're not there yet so it doesn't really matter that much the main thing to think about here is that we're thinking that systems which are much much simpler orders of magnitude simpler than the simplest bacterium on earth today right we're trying to understand the spontaneous transition from chemical systems to the simplest possible cells so we don't want to have a system with hundreds of genes we're just trying to understand how you go from zero genes to one gene and and so that's something I'm going to come back at the end so the way we think about this is that a simple cell would have two fundamental components a cell membrane to basically make the distinction between inside and outside and to keep all the valuable goodies on the inside from just diffusing away into the environment and then some kind of genetic material that can mediate the inheritance of useful functions and so that also implies that whatever that genetic material is it it has to be able to do useful things so it could be an RNA like molecule which could not only mediate the inheritance of genetic information but also carry out structural or catalytic functions that would confer an advantage to to its cell ok so it turns out that actually assembling structures with these two kinds of components is extremely simple it happens more or less spontaneously there are various ways of catalyzing it making it more efficient but the tricky part the interesting part fun thing to work on and think of that is how you can make systems like this spontaneously grow and divide and so that means that both the membrane has to grow spontaneously and divide and the genetic information has to get replicated and distributed into daughter cells and all of this has to happen without any of this highly evolved complicated machinery that is basically responsible for running all existing living cells all of this has to happen from just chemistry and physics and so that's in a sense one of the fundamental questions is you know is is that possible and if so how could it work can we demonstrate it experimentally okay so before we go into the specific experiments and the basic idea is that if you had a system where a cell carry protocells simple cell can grow and divide this makes a cycle as you go through this replication cycle you'll explore more and more of sequence space eventually some sequence that does something useful should arise and as soon as an inheritable sequence emerges that confers an advantage switch to its host cell it should start to take over the population that change in the genetic structure of a population is essentially the definition of Darwinian evolution and that's what we would really really like to see in the lab ultimately okay so so what I'm going to do is in the first part of the talk focus on the membrane system and explain a very simple pathway that allows very primitive simple membranes to grow and divide in response to physical changes in the environment and then I'll say a little bit about our approach to the problem of replication of the genetic material and my colleague Jerry Joyce over there will talk about one approach tomorrow I'm going to talk about a more chemical approach and if either one works out in the end we should be able to combine these and and generate simple living systems okay so so what can we make these primitive cell membranes out of so we don't want to use the kinds of molecules that make up our membranes today okay modern cells use complicated components that make membranes that are good barriers and that means that cells because they make machinery to get molecules across the membrane have total control over everything that goes in and out of the cell in a primitive state that's not what you want at all you want primitive cells to be able to take up molecules from the environment and that means the molecules have to be able to get across the membrane so the solution to that problem is actually very simple and satisfying in a way if you make the membranes out of much simpler components such as fatty acids like you see here the membranes have very different properties which are more appropriate to what you'd like to see in a primitive kind of cell so these fatty acids are what we use as our typical convenient lab model systems this shorter chain saturated fatty acid is more realistic basis for what you might find on the early Earth in any of these cases if you just take the fatty acid dissolve it and water shake it out with some buffer they self assemble into beautiful bilayer membranes that close up into these wonderful vesicles here you see a large vesicle with smaller vesicles in between and inside it okay so those membranes as I said have very different properties from normal membranes and I think we're going to need the light stand for this and the several of the coming slides but I just wanted to show you an animation that was generated by Janet iwasa in my lab to illustrate some of the dynamic properties of these membranes this is a lot of motion there's thermal fluctuations if we look in cross-section there's a bilayer membrane but the okay yeah so there's a lot of dynamic behavior that you don't see in modern membranes based on phospholipids a small because it constantly flip-flopping from one side to the other they're entering and leaving the membrane and all these exchange processes play an essential role in the ability of these primitive membranes to grow and divide in in several different ways okay so so how can we look at this issue of replication in in these simple membrane systems so for many years when we first started studying this we were we're actually kind of terrified of the complexity they that you get when you when you do just shake these fatty acids up in water you you don't just get nice uniform single layered vesicles you get them real mess right they're tiny vesicles big vesicles compound vesicles multiple by layers and in in the walls here and studying a heterogeneous mix like this is is kind of problematic and so for for a long time we tried to simplify everything make everything controlled and and homogeneous and well-defined so that we could do simple experiments and understand what was happening and that led to some proof of principle but highly artificial ways of driving growth and division it wasn't until I got a really talented really brilliant graduate student in the lab that we that we actually came back to this kind of system and so to in qingxue came to the lab he said you know I think we could we could take mixtures like this and just sort out vesicles of roughly the same size and make a relatively homogeneous population that we could actually study and he proposed a simple way of doing doing this and I told him that would never work sounded impossible to me he came back two days later and showed me beautiful pictures like this and and so that I love having students like that who will just follow up on their ideas and show me I'm wrong so what you see here are vesicles that are about four microns in diameter they typically have multiple by layers and often smaller vesicles inside but you know they're relatively homogeneous so now we can do a simple experiment just look at these in the in the microscope add more food meaning just throw in extra fatty acids and watch how they grow so we don't have to use any of these indirect fluorescence or light scattering approaches we can just watch and of course what we expected was that well these are spherical vesicles they should probably just grow by just turning into bigger spheres or maybe they get a little elongated if if the surface screw faster than the volume what we did not expect was yeah I think we might need the lights down I don't know can you see the little tails coming out of these okay all right so from this angle it's I can see it very well so yeah this is how they grow they this is after five minutes every spherical vesicle has a little tail coming out of it and after half an hour they have all morphed into these filamentous vesicles and I want to show you a movie this is not an animation this is a real-time not real-time speeded up video micrograph and for this we really will need the lights down further because it's a little bit thing ok so again now we're looking at a single vesicle we add more food and yeah you should be able to see a faint filament fluctuating around here over time the parental vesicle gets smaller as its material is distributed into this growing filamentous structure and after about half an hour you see this long branched thin filament so that's how that's how growth actually happens when you do it in a very simple way and so so this phenomena was totally unexpected raised lots of interesting mechanistic questions that we've been looking at but which I don't really have time to go into today but in addition in addition to those questions it actually solved what was the hardest problem of getting the cycle to go which is division we had to go to great lengths to get spherical vesicles to divide into smaller daughter daughter vesicles these filamentous structures are so fragile that the slightest agitation or sheer mild shear forces and the solution will will trigger division and and so I just show you one example of that here so there's an interesting physical instability here you see the sort of bead on string morphology but eventually the filament will just snap in response to gentle perturbations and in the fluid and so that's a really easy way of making a cycle of division now you grow into filaments gentle agitation causes division so you can think about wave action on a shallow prebiotic pond for example and that cycle can be repeated now there are other ways the division could happen and so I just want to show you another movie that's a little bit more dramatic we don't really know if this is prebiotic lee plausible it's this is a photo chemically driven process we see this again this purling instability and then over time driven by the photo chemistry here the individual beads separate from each other and drift off so this is a way of dividing in to a large number of small dotter protocells okay so in both cases we have a cycle you add food you grow into a filamentous intermediate division into small structures these can again over time can grow larger and grow into filaments you can go through this cycle indefinitely okay so we think that's a pretty simple and robust process that takes more or less takes care of this problem of how the cell membrane could grow and divide in response purely to physical fluctuations in the environment so then the harder part is the question of the genetic material and how could that replicate okay so so just to remind you we really need some kind of genetic material doesn't necessarily have to be a nucleic acid but we're pretty conservative so we're sticking with nucleic acids for the time being we need it to code for heritable functions right and and so we need a way of replicating it so that as the the host membrane vesicle grows and divides these sequences can also replicate and be distributed to daughter cells so there are two general approaches to solving this problem of replication one is enzymatic catalysis and so you can think of for example an RNA molecule that's an RNA polymerase that could catalyze its own replication or at the replication of its own kind of sequence and and lead to an auto catalytic amplification so Jerry Joyce is going to talk about this approach tomorrow at least in part I hope and what I'm going to talk about is an alternative approach which if it were excited in the end might provide a simpler and more gradual transition from chemistry to biology which so this is a question of whether they're just truly chemical processes that might drive replication okay now no matter which approach you take there are two fundamental considerations okay the there are two thresholds that have to be overcome the rate of synthesis has to be greater than the rate of degradation okay so you can't go arbitrarily slow and the accuracy of that copying chemistry has to be good enough to get over the eigen error threshold we want to transmit useful information and so so the error rate has to be low enough to allow that to happen indefinitely okay so here's another one of Janet's animations just to emphasize what we're after in this chemical system so we want to have template strands of you know RNA or some related polymer floating around in solution with activated building blocks that just find their complementary bases and click into place and gradually build up a complementary strand and then we need some mechanism presumably thermal denaturation to pull those strands apart and allow this to happen again okay so how can you actually do this so this was a problem that was actually worked on for several decades by the late Leslie Orgel and many of his students and colleagues again including Jerry Joyce and as in the case of the membrane forming building blocks the building blocks that are used by modern biology the nucleoside triphosphates are not appropriate for a primitive cell and Leslie recognized that and studied in large part activated analogs such as this they're much more intrinsically chemical reactive so they can polymerize spontaneously on a reasonable timescale there's another consideration that I find very important and that is that these modern building blocks are highly charged extremely polar so they can't leak out of cells but they rely on extremely good catalysts in a primitive state you want something less polar such as molecules like this so that they can get across membranes spontaneously and again more activated so that you don't need such good catalyst so maybe you don't need any catalyst okay now it would be really nice and simple if we could just start with RNA and maybe we can but we don't really know how to do that with at our current level of knowledge and the replication of RNA has has a number of intrinsic difficulties one is well which hydroxyl are you going to use here to make the linkage to the next base all right and typically when you do these experiments you get a mixture and that just seems very problematic there's also effects of the basis so copying GS and C's tends to work pretty well copying a huge sequences is slower and can be actually quite difficult the accuracy of the copying is not really quite good enough and there's another problem with RNA itself which was even if you succeeded in copying the sequence and making a duplex those duplexes are so thermally stable that it can be extremely hard to get the strands apart so in view of all those problems which we don't really know how to approach at this point what we've decided to do is sort of step away from the strict considerations of what we'd like to start off with prebiotic ly and just look at some model systems and try and see if we can learn some some general principles that might either lead us to an artificial replication system or maybe give us some clues as to how to actually go back and replicate RNA in a chemical way and so this is just a panel of nucleic acids that we've been studying in the lab over the last few years if you just focus on the one in the middle this is essentially DNA except for one small change we've changed the what would be the three prime hydroxyl here to an amine and so the linkage is now phosphoramidite otherwise this is extremely similar chemically and conformational e this little change speeds up the polymerization chemistry by a lot and so that's why we're looking at this series these are in in terms of an order roughly of conformational constraint so this polymer over here was the open acyclic backbone is extremely flexible the two over here are much more rigid and DNA and this alternate linked phosphoramidite DNA are sort of intermediate and we're looking at these in order to try to get some insight as to whether the degree of flexibility or constraint is is something that is is important in solving these problems okay so I don't have time to tell you about all these different polymers I'm going to focus on this one over here where we've made the most progress and we decided to concentrate on this initially for a few different reasons so here is one of the four standard building blocks for this to prime amino system one thing that's nice about it is that this the rigid sugar structure holds the nucleophile here the amine away from the electrophile let's phosphorus so they can't reach each other they can't cyclize so this whole side pathway that takes away your activator Barnabas just goes away in the natural RNA system the normal bias is to make 2 5 linkages so we thought well let's just go with the flow and do what nature's trying to do anyway and make polymers like this and see if it works better also these kinds of duplexes are known to melt at lower temperatures so we thought it might solve the problem with strand separation or at least make it easier okay so it took us a while to make these things but once we did we could start to look at how they polymerize on different kinds of templates and so this is one of the first results so here the template is essentially this string of C's and what we're looking at is is a primer annealed to this template and we're going to add a bunch of G's on to this template so here's the activated monomer we add it to this primer template complex and we look at the synthesis of the elongated primer and so here's a time course of the reaction it looks pretty much like an enzymatic DNA polymerase reaction but there's no polymerase it's Jeff chemistry okay so after 12 to 24 hours you have pretty much full likes copied a template so so the good news is the chemistry looks great you know if we could copy arbitrary sequences this effectively we'd be done we'd be ready to start putting this kind of chemistry into into vesicles and evolving new functions so unfortunately it's not that simple this works really well with G's and C's as in the case of RNA when you try to copy a DH and use it basically doesn't work okay so then you have to start asking well what can you do to solve these kinds of problems and there are also fidelity issues so we can look at well we could change the nucleobases we could try to you know play around with the chemistry of those recognition modules we can explore all the different backbones and see if for example more constrained backbone my position things more accurately and lead to faster and more accurate chemistry we can try to do less chemistry by putting in preformed building blocks or we can give up and let Jerry solve the problem was a good catalyst okay so so the first thing we did is just go back to the few mood and start making analogs okay so here's the problematic base pair and I just want to go through this quickly just to show you that the kind of the course of the thinking and what happens as you go through this cycle of experiment and and and thinking so you look at the a you based parents obviously weaker than a GC base fare in part because it's only got two hydrogen bonds instead of three but also in part because the you nuclear base gives you the least energy from stacking interactions of any of the four standard bases okay so both of those problems are solvable in simple chemical ways they have been solved by other groups and the Switzer lab has looked at this base pair in the context of DNA so by changing a to diamino purine you now get 3 hydrogen bonds and by changing you to five per Pinal you you put in some extra stuff here which gives you more stacking interactions and this is just as good as a GC base pair okay so we made these in the two prime immuno system made the activated monomers and indeed copying D and Pro Pinal you works great okay so that problem seemed solved but is it no turns out that when you make this modification you're changing all of the electronic properties of the bases you're changing the pKa of this group it ionizes more easily the fidelity goes to hell so it's actually worse in terms of accuracy by a lot okay so now what do we do we have to go back to the drawing board the drawing board is really easy you know it's ChemDraw you can draw these things in five minutes and then two years later you get to do the experiment after you've made the molecules okay it wasn't quite this bad in this case we've started to look at and get very excited by this very simple modification just changing this oxygen atom to sulphur this has been studied by several groups in fact by Doug Turner's group ten or twenty years ago this is known to make a stronger base pair and in fact it's quite dramatic and so we see that when we make the activated monomer of 2000 the pKa problem should not be worse in fact if anything it should get better just sterically this is known to to weaken wobble base pairs and so we're thinking or hoping the the simple modification might solve both the rate and the accuracy problems it has the added virtue of maybe being prebiotic ly plausible in a sulfur rich environment so we'll see so this is kind of the the the way these chemical experiments evolve in the lab okay so it'll probably take us a while to work out a complete replication system but meanwhile we can use what we've learned to go back to the vesicles and ask questions about compatibility okay we actually put all these things together and so we can do these simple template copying experiments where we put the primer template not floating around freely in solution but inside these fatty acid vesicles okay so here the experiment is to add monomers to the outside so they have to diffuse spontaneously across the membrane no transport machinery chess spontaneous physical changes get to the inside and then do the chemical copying okay so the panel over here is what I showed you before this is the reaction in solution the panel over here is with an encapsulated primer template and you can see the time course is a little bit slower reflecting the added few hours it takes for these monomers to get across the membrane but at the end of the day you still have full-length template copying so that was extremely encouraging and it's motivated us to continue the search for a chemical replication system because now we know if we can do it it will be compatible with the membrane system and we can then start to think about such interesting questions as what might those first selectively advantageous functions have been what kinds of new functions could emerge spontaneously that would drive competition and allow the emergence of Darwinian evolution and so I just want to tell you one short story again the work of a really brilliant graduate student each ibudan in the lab and this came from actually a consideration of a larger evolutionary transition from the primitive fatty acid based membranes that I've been talking about up to more modern membranes based largely on phospholipids as well as sterols and all the other components so the primitive membranes as I've said are very dynamic they let polar molecules get across they have exchanged processes that allow for growth and division and modern membranes don't have any of those properties right there are much more static structures that rely on evolved machinery to carry out those functions so how do you get from the print of state to the modern state you can't make the jump all at once that would be suicidal because now a cell couldn't get any of its nutrients in before they before to devolve transport machinery or metabolism so it seems logically that you have to go in a stepwise process and so an early stage ought to be a membrane made mostly of fatty acids and perhaps doped with this very small amount of phospholipids which you can be you can imagine being synthesized in the cell for example by a ribozyme so okay that's fine but what we couldn't really understand is what selective advantage would this confer all right if a cell could make a little phospholipid okay we'll go into membrane it probably wouldn't change the properties very much but if it didn't confer an advantage it wouldn't take over the population and it wouldn't initiate initiate this this series of transformations so of course the answer came from doing an experiment and what each I did was just to make vesicles that were mostly fatty acids but doped with small amounts of phospholipids and then just look at their properties and the the the results were really extremely dramatic so these vesicles was a little phospholipid are perfectly stable on their own they just you know live there like normal they don't do anything strange but as soon as you mix them with vesicles that are just fatty acids these guys start to grow and the fatty acid vesicles start to shrink okay so this is a real result of these exchange processes I've been taking talking about and mechanistically this is because phospholipids allow membranes to hold on to neighboring fatty acids for a longer period of time the result when you look at this and the microscope is shown down here so here's one of these initial vesicles there was 10% phospholipid mixed with a large excess of pure fatty acid vesicles and over the course of two or three minutes you see this transformation into again this filamentous morphology so we could drive the same cycle of growth and division growth into a filament agitation causing division to give us a cycle and this would be driven by the ability could be driven by the ability to make phospholipids okay so making phospholipids from activated precursors is just a single step simple acyl transfer reaction lots of ribosomes have been evolved in the lab that do that kind of chemistry it shouldn't be an issue so if you grant that that could happen there's an interesting series of consequences so as you've seen experimentally phospholipids can drive the growth of fatty acid vesicles implying that there would be an incredibly strong selective advantage for any primitive cell that could start doing this of course that cell it's progeny would take over the population but now the entire population would be composed of cells that are making a little bit of phospholipid okay so then the only way to grow by eating your neighbors is to make more fossil of it than they do so that would lead to an evolutionary arms race favoring cells that make more and more and more phospholipids at some point the permeability of those membranes starts to decrease so it gets harder and harder to import stuff you need from the environment and that means that it would be favorable to either start making stuff inside yourself so that's metabolism or and maybe in addition start evolving membrane transport machinery that will help you get stuff in and out so we think that this whole series of consequences could follow from a very simple initial metabolic function okay so then just to sum up from all these simple experiments what have we what have we learned and I think there are two really general lessons so the first is that we've seen just an amazing number of completely unexpected but very simple physical phenomena for example the way these vesicles grow is a complete surprise the permeability of member two charged molecules was a surprise and there are many other examples that I could go into a second lesson is something I think we're being driven into reluctantly which is that we always start off trying to make simple constrained well-defined systems and they don't usually work very well and then as we learn how to relax the constraints appropriately and work with systems that are Messier but more natural things actually start to work a lot better that's certainly been the case in the membrane system and I think it's going to turn out to be the case with the problem of nucleic acid replication okay so let me just end with something that's a little bit more speculative this is a figure from a recent Scientific American article that I wrote with Alonso Ricardo who is then in the lab and and so this just gets to you know based on on the chemistry and the physics of the systems that we've been studying what are the implications for the appropriate environments you know could there be places on the early earth reasonable places where all of this could happen and what would a cell cycle really look like in the beginning so the basic idea is that you have a fundamentally cold environment either Arctic or Alpine but but in a volcanic landscape so there were localized sources of heat and that can set up convection cells where our cells would be cold most of the time but every now and then get entrained in a plume of hot water and so the idea is the template copying copying the genetic material goes well at low temperatures so that can happen slowly in the cold then you make a duplex inside a vesicle which can also grow under these conditions a brief exposure to very hot conditions will lead to strand separation also leads to an influx of raw materials as the permeability goes up and then division as a consequence of the gentle shear forces gives you now replicated daughter cells which can go through the cycle again so from what we know about these membranes this has to be a freshwater environment you can't have too much calcium or magnesium around or all the the acids precipitate we need a heat source for strand separation I don't think it can be something like the day/night cycle I think it almost has to be something more like geothermal heat that will give you transient exposures to high temperatures and so I think this is is it's clearly very speculative but it's the way that we can at least get some clues about potential early earth environments that would support the kind of chemical systems we've been looking at and so obviously we're hoping in the years to come to be able to demonstrate more and more of the cycle and eventually watch new functions such as this acyl transferase activity emerged spontaneously in the laboratory that would be the the real ultimate goal as far as I'm concerned okay so I've tried to mention the key contributors as I went along as certainly each ibudan and ting zoo played a major role in the vesicle work along with sri frenzy earlier on on the nucleic acid site a large number of talented students and postdocs an essential Craig Blaine Alonso Ricardo Sheng Wang Shang Jason Trump and Schuler shines their local NMR expert and many others and thank you for your attention you

Channel: MoleCluesTV

Views: 12,212

Rating: 4.7707005 out of 5

Keywords: Cell, Nobel Prize (Award), Molecular Frontiers, Jack Szostak, origin of Life

Id: jmMU0b20FSg

Length: 41min 20sec (2480 seconds)

Published: Wed Mar 26 2014