 Okay, I'd like to thank the the organizers for inviting me to this really interesting conference and giving me the opportunity to Tell you a little bit about our work Although so I like to start with the central dogma introduced by Francis Crick in the 1970s. I think that shouldn't need too much of an introduction. So this was formulated to To illustrate the belief at the time that information flow in biological systems is unit directional from DNA Into RNA and into proteins Later it turned out that this is only partially true at least between RNA and DNA the information flow can be reversed By the actions of reverse transcriptases and in a limited Limited way by telomerase as well so The fact that these two polymers can exchange information at information can be written from DNA Into RNA and from RNA back into DNA I think naturally leads to the question why just to kind of could there be other polymers with which DNA or possibly RNA could exchange information So we ask ourselves the question if we could conceive of simple chemical alternatives to DNA and RNA And would these also be capable of heredity and evolution? And to do this you need a chemical framework and we call these kind of XNAs for xenonucleic acids They don't occur in nature and to sort of add these two extra arrows to the central Dogma diagram what you really need is an XNA polymerase to write Information from DNA into XNAs and once you're there you need an XNA reverse transcriptase to Write information back into DNA and if you can do that if you can close that that loop We have a replication cycle for for XNAs via a DNA intermediate a little bit like retro viral Replication and that opens the XNA sequence space for exploration and and in that sequence space You know, there will be functions and I'm going to tell you a little bit about ligands We can make XNA active after most you can make catalyst XNA enzymes and you can also build Simple nanostructures and devices which I won't tell you about and when we started this We didn't just want to do this for one We wanted to be able to do this at will for a number of backbones and really as far as many as we can do So step one we need an XNA polymerase to go from DNA to XNA and the XNAs we started with These two are going to talk mostly about HNA So this is HNA and CNA where the five-membered ribofurinose ring of DNA in RNA is replaced with six-membered congeners cyclohexanil in CNA and hydrohexatol in in HNA and the reason we chose these is really because they have Remarkable properties so these at the nucleoside level and also as polymers are non-toxic So they do not interfere With the cellular metabolism they're orthogonal and this is really can't be taken as a given many many Nucleoside and nucleotide analogs are potent detoxic the potent antiviral and anti-cancer drugs So these are orthogonal that's as polymers. They're also completely resistant to nucleus degradation So this is HNA you can see here DNA goes very quickly with DNA is one and even quicker with bowel 31 Which is sort of shredding microbial shredder enzyme while HNA is largely impervious same in serum DNA goes relatively quickly and really HNA stays there for as long as you like But importantly they they retain the ability to specifically hybridize to both DNA and RNA and that's clearly important for for information transfer But Orthogonality as I said, you know has its price. They're truly terrible polymerase substrates when we started You know, this is sort of how good we could get I'm just going to quickly Introduce what these gels tell you so this is a primer extension gel going to see a few of those So this is a nice semi qualitative way To measure polymerase activity you label a primer at the five prime end You let the polymerase extend it and resolve the products on a on a polyacrylamide gel So to each so each of these rungs of the ladder To the closest approximation denotes an incorporation step and you can see this is the best polymerase We could find this is a Replicative polymerase from an archaea bacteria called thermococcus gorgonarius and you can do about one two three four five six Maybe seven incorporation step and then it's curtains. You won't go any further. However long you wait So at the time this was really Not not good enough for The type of screening and selection systems that were available So we built our own and this is called compartmentalized self-tagging CST That's how we call it and it's really very simple method It's based on a positive feedback loop whereby polymerase extends its own encoding Sorry tax its own encoding gene or in this case the genetic element containing its own encoding gene by extension of a primer And this primer importantly hybridizes only in a meta stable way To the plasmid so it needs to be stabilized this interaction needs to be stabilized by this extension And the primer also has a five prime capture tag and to ensure Genotype phenotype linketry encapsulate these into the aqueous compartments of a water and oil emulsion So there so to talk you through it the red polymerase can use X in a triphosphates It can extend the primer so when we break the emulsion you can capture this plasmid By the capture tag on a magnetic bead the light blue polymerase cannot and This this gene will be lost from the gene pool and once we were able to get this to work You know we very quickly arrived at the polymerase that can do H&A So this is 6g12 you can see this our starting point now we can synthesize Easily 73 we can go a lot further But this is just a tRNA gene just to show that we can do an actual biological gene Okay, so we're here now. We need to go back. I mean this is just a band on a gel not not very interesting We really want to know we want to decode the information that's contained in that a band and for that We need to build a reverse transcriptase again. No natural polymerase is display any H&A our tea activity So we decided to build run from the ground up again based on TGO because we know it well and this time we used a bioinformatics technique called statistical coupling analysis So this was introduced by Loclas and Ranga Nathan and it's really based on the idea that That distant areas of a protein communicate with each other through interacting networks of amino acids and that these networks Reveal themselves in phylogeny and can be mapped out through statistical techniques Really the the quality of this analysis rises and falls with the quality of your Sequence alignment, so we build a hand-corrected database of polymerases for alignment this just shows the cluster analysis and then the the scar residues plotted onto to polymerase structure in orange together with the highly conserved residues in cyan and you can see it sort of makes sense kind of you get a continuous surface that Encloses the primer template duplex, but you what you can also see you get far too many hits kind of it's not clear where to start Really to build the desired activity that we're after So we try to route this Analysis to this residue here So this is a residue that was described by the biotech company strategy to convert RNA reverse transcriptase activity on a related polymerase from pyrococcus When we mutated it It gave us nearly RNA nor HNA reverse transcriptase activity, but We decided to look in the vicinity We thought maybe a solution can be found in the close vicinity of this residue of these 15 residues in the close vicinity four were also part of these scar networks and one of these When mutated actually, I think isoleucine to leucine sort of really subtle mutation gave us very very excellent Reverse transcriptase activity and that really closes the loop So now we can move the genetic information from DNA into HNA this entirely unnatural genetic polymer And we can read it out with an HNA reverse transcriptase Because we're going by a DNA intermediate these can be cloned and sequenced very easily and the information transfer proceeds with a I think very decent Fidelity of about one in ten to the minus three. This is roughly equivalent what you get in viral in in RNA viruses Okay, and once we figured out to do it with HNA we very quickly figured out to do it how it was CNA Anna and Fanna are based on Arabino furanose rings rather than ribo furanose rings TNA is an interesting case based on a tetrose rather than a pentose And some of you may have may have encountered the interlocked rings of LNA which are important in in biotechnology different Different polymerases are needed for different of these XNAs, but they all share mutations in the so-called Some domain of the polymerase we never find any mutations in the active site And exactly what these mutations do we're still in the process of figuring out, but that's a story for another day So we can do Information transfer backwards and forwards into XNAs. What about evolution now to look at evolution? We turn to these aphthalmers. These are biomolecular drugs based on single stranded nucleic acids I'm not going to introduce them again. I think Echirro did that very well yesterday Just to show you again kind of this is a structure of a thrombin aphthalmer bind to thrombin in orange just to show you that these can really fit into the Surfaces of a protein in really very much the same way as antibody cans and have the potential to bind the targets with equal affinity and specificity but they haven't really Made good progress into the clinic because when they discovered as DNA and RNA aphthalmers They need to be extensively modified like in this case with macogen to to survive it in the body So if we could start with HNA, which is on degradable, you know We might have an interesting starting point and wouldn't need to go through all of this and medicinal chemistry So this is our HNA aphthalmers selection protocol. This is completely standard We start with a DNA random sequence library. We synthesize HNA and we get rid of the DNA with DNA is one We Bind the library to a solid phase target. We wash away the non binders. We loot the binders We reass transcribe we PCR and we can start again and the first target we we went for is this RNA motif HIV one this is called tar the virus needs this to grow and many many DNA and RNA Aptimers have been described to it they seem to interact mostly why this kissing loop interaction You can see an RNA aptimer here and we saw this we thought we'd be really clever because we required a very small library We would only randomize a loop and Would get a kissing loop interaction in our HNA aptimers But as it turned out that was not to be so so this is a specificity assay For the different after most so this is a cartoon of the tar RNA motif These are variants of tar where the sequence regions shown in Shown in magenta kind of like have been scrambled while retaining secondary structure You can see the RNA aptimer will keep binding to all the HIV motifs That retain this loop interaction But not to the ones where the loops sequences are scrambled But the h&a aptimer require seems to require the full motif Including both this loop and the bulge and there's a second one in the library Which doesn't care about the loop at all what seems to instead target this bulge region and We measure the affinities so these have you know respectable affinities in the in the sort of mid to low nanomolar range But really making Nucleic acids bind other nucleic acids Not that Interesting so we try to make aptimers against the protein target Henic lysozyme is nice and cheap So we started with that again many examples in the literature for comparison and you can see you know We get decent binding with cross reactivity to the related human lysozyme and again You know not great, but respectable affinities and a very very intriguing G and T rich sequence motif and we're trying very hard to to crystallize the The complex of these aptimers with Henic lysozyme because we'd really like to know how these fold up Just to show you that these are the proper Liggins they don't just bind to protein coat on plastic. So this is a plasma cytoma cell line, which we which expresses Henic lysozyme as a membrane protein Compared to the wild type line you can see the aptimers very very nicely pick out the The Henic lysozyme on the cell surface in the context of all the other proteins specifically In a fax essay you can see the wild type shows no signal at all And I've already told you these are non degradable by nucleases We've thrown the whole New England biolabs catalog at them. They don't get degraded But they're also chemically quite stable. So this I don't if you can see this These are two biocore curves overlaid one drawn out one dashed and this is before and after a three-hour exposure to pH one at 40 degrees followed by pH nine trust me DNA does not survive this treatment There's nothing left while the two H and aptim as you can see neither can neither binding kinetics nor amplitude is is affected by this and And My colleague Jeffrey the Stefan has also been making aptimers against HIV reverse transcriptase These are picomolar binders. So they're really excellent. And this is in the fluoro Arabino Framework, so clearly I think there's a fairly general opportunity there to to generate binders in this framework So we have ligands, what about enzymes? So this shows our discovery strategy for enzymes We wanted to make initially X and A times that would cleave a certain RNA sequence So this starts with with with basically synthesizing our X and A's from an RNA primer and having them fold back onto their RNA substrate you can then separate active X and A's I'm simply by a gel shift assay and and go back to DNA by reverse transcription and We very quickly found some In the nucleus in the in again fluoro Arabino frame or you can see the enzyme here You can see the cleavage and the catalytic rate is You know, not stunning again, but but respectable and And again once we managed to do it for Fana, we managed to do it in a whole range of other X and A's with with, you know, slightly different properties But this is still catalysis where the substrate is is natural So in this case, you know, you could argue maybe, you know Maybe actually the catalytic function is at least partially encoded by the substrate substrate. This is the catalysis is common in Many enzymes. So can we make a completely synthetic catalyst and and this is this here So this is an X and a ligase X and A's I'm so both substrates are now also X and A's as well as the enzyme shown here And this legates two pieces of fluoro Arabino Nuclear gas it together and you can see here These are the proper controls only in the presence of both substrates and the enzyme Do you get a ligation and this now allows you to play games where we can ligate together the The Fana's I'm in the nucleus Sub-hybridized it to its RNA substrate and get cleavage. So it's a sort of two-step assembly Of where an X and A's I'm assembled another X and A's I'm so I'd like to summarize this part of the talk So I think what this what this shows really that at least in this limited way as far as we have explored it both heredity and evolution Two fundamental properties of life are not restricted to DNA and RNA But can be implemented in you know quite a range of of different polymers and Presumably evolution is an emergent property of all information carrying properties that can be That can be replicated And there's clear opportunities now for maybe Re-invading biology with these orthogonal backbones and trying to build orthogonal genetic systems within the cell There's also opportunities in biotechnology. This clearly expands the chemistry of nucleic acid polymers that can be replicated and As I've shown you we can make aptomers and enzymes and also nanotechnology objects And since this is a sort of shows a family picture of the various X and A's we have a few more added to this and you can also As of recently we can you know increase the molecular diversity further by introducing an on canonical backbone linkages Okay, so for the second part of my talk, I'd like to Get back to the central dogma and I think another question that is immediately obvious when you look at this is Is how did the system get started we need nucleic acids to make proteins and proteins to make nucleic acids? How do you? How do you build this up? How do you boot up life? Now over 40 years ago Francis Craig Leslie Orgel and Carl Those Proposed what seemed at the time a pretty far-fetched idea and their idea was that our biology was preceded by a primordial biology That lacked both DNA and proteins But relied on RNA as its main You know Molecule not just for for genetics, but also for metabolism and At a time as I said it's pretty far-fetched idea But in the meantime, I think a lot of compelling if if circumstantial evidence has accumulated that this there is really a lot Lot to go for this idea I mean the structure of the ribosome is maybe the smoking gun of this what is called the RNA world hypothesis commonly but one of one cornerstone of this hypothesis is missing and that is a That is a replic case to replicate those nascent RNA genomes of the RNA world and RNA replic case and So it hasn't been found in biology. What can we do now true to the dictum of my former teacher Albert Eschenmoser Great chemist from Switzerland the origin of life cannot be discovered. It can only be reinvented We decided to try to build such a replic case from scratch And our starting point is this amazing Ribosome discovered in David Bartels lab at the Whitehead Institute still sometimes blows my mind that a Molecule so simple can carry out such as sophisticated Molecular function. So this is an RNA polymerase ribosome. So this can read out a sequence on an RNA template and Synthesize its complement you can see its action here another primary extension gel And you can see it can go to about 14 incorporations. It's not very fast. It takes 24 hours to do that And and then it can't go any further. So it's it's pretty amazing But unfortunately we're some distance away from replicating Really itself or an RNA genome So as a minimal goal we set ourselves because the RNA the r18 what it's called polymerase ribosome is 200 nucleotides long So the minimal goal we set ourselves is that we should be able to Improve this to at least the level where it can synthesize RNAs as big as itself. So we're at 14 We need to go to 200 And I'm gonna tell you now how we get there Again, we let nature doing all the hard work. So there's another selection strategy that which we call CBT For compartmentalized be tagging now this looked complicated, but fear not all I will explain and it's not that Difficult at all. So there is it starts with one micron beats They contain a single a double-stranded DNA gene Encoding the ribosome as well as about 10 to the 4 copies of this little squiggle here This is an RNA hairpin and we will encapsulate these again Into the aqueous compartments of a water and oil emulsion to ensure Genotype phenotype linkage and then we carry out a couple transcription ligation reaction and what this does is it decorates The beats with about 10,000 copies of ribosome. So when we break the emulsion, we now have these population of clonal beats each containing a Multitude of ribosome, but each beat just containing one species of ribosome We can then add the primer template duplex Re-emulsify again encapsulate them again and now the ribosome can do its job synthesize RNA and really the rest of the workflow is all about converting this Primer extension signal into a fluorescent signal that we can read out in again site by flow cytometry Isolate the beats containing the best ribosomes and start a cycle again and we initially started with a Just complete random sequence library appended to the five prime end of the ribosome After just three rounds of selection we isolated this clone here C 19 and this is now just one hour extension reaction You can see the wild type can barely do anything while this isolated clone is significantly superior And it has this rather intriguing hairpin domain here. So we thought like that's where the function lies But when we analyzed it in detail it turns out Most of the improvement is encoded by this simple hexanucleotide sequence at the five prime end Which is also complimentary to the five prime end of the template So clearly what this does it it binds the enzyme to its template presumably Increasing it it's local the local concentration and thereby increasing polymerase activity Now what we found is that especially for longer since trying to synthesize longer RNAs Not only did this hairpin domain not do anything. It was actually inhibitory So we replaced it simple placeholder for a's and that really unleashed the polymerase activity of this ribosome. We can now synthesize RNAs up to 96 Nucleotides long at least on a on a sequence template that the that the ribosome likes particularly unfortunately, it's not a general RNA polymerase at this point But we certainly were able to do on a much shorter sequence to begin to synthesize rather than arbitrary RNA sequences We can see now synthesize RNAs that actually encode a function. So this is an Hammerhead RNA endonuclease ribosome. You can see the wild type kind of cannot make any full-length product While our best ribosome now can and when you put it together with temp with with its substrate strand here in red You can see you do get cleavage. So this is the ribosome catalyzed transcription Of another ribosome clearly a process that must have gone on During the the RNA world where you know the replicase would transcribe genes from the RNA genome now For the third part of the talk, I'd like to get back to This picture here of the RNA replicase now There's another thing that always bothered me about the RNA world hypothesis Now I don't know how many of you have worked with RNA It's perfectly okay in in the lab in a nice clean eppendorf tube, but it's a rather fragile molecule It falls up to bits very rapidly at high pH in the at high temperatures or in the concentrate in the presence of high concentrations of Metallines and certainly in combinations. They're off. We'll destroy it very quickly. So really the problem is Really RNA is a questionable if not downright downright perverse choice of primordial genetic material Because clearly the surface of the early earth wouldn't have been such a benign place as our little clean eppendorf tubes in the lab and This RNA polymerase ribosome is a case in point even in the lab under the best possible conditions It actually falls to pieces in about two days because it requires high concentrations of magnesium mines for activity so really either the RNA world hypothesis is wrong and some people think so and have Proposed pre RNA worlds based on more stable polymers or alternative scenarios or and we like that Approach better. We need to think of an environment where RNA makes more sense where RNA could be stable both stable and active As for stability, I think the solution is is relatively obvious What do you do in the lab if you want to preserve a molecule that is that is perishable? That's right. You put it in the freezer. That's where they stay for long But freezing enzymes is not a great way to preserve activity so this is a Proteinaceous RNA polymerase T7 RNA polymerase you can see nicely laddering at 37 degrees But when you freeze it, it's dead But the RNA polymerase ribosomes different So when you freeze it, although it's a little bit slower, it will eventually go as far as it does in and on their ambient conditions and what's more because it's now frozen and more stable There is something of the tortoise and the hare about it. Well, it's faster at ambient temperatures in the ice It's slower, but it just keeps going and going and going and if you adjust the conditions You can actually go considerably further in the ice phase than you can go on the standard conditions and The reason for that is that it would be wrong to think of ice as a homogeneous solid medium and it's in fact a biphasic medium. So what happens when you freeze a Aqueous solution containing ions and RNAs and primers and triphosphates the water freezes out first But all the solutes get excluded into an aqueous brine phase that surrounds the ice crystals So this can be seen in this scanning electron microscopy picture. This is a freeze fracture You can see the hexagonal ice crystals here and these are surrounded by these riches the so-called eutectic phase This is a brine phase which surrounds the ice crystals and stays liquid at sub-seru temperatures and that's where the ribosome Does its work? So the next question we asked ourselves can we you know, we can synthesize RNA and ice What about evolution and to do that? We simply replaced that the second emulsion step in our selection protocol by freezing so instead of encapsulating the The the clonal beads into the aqueous Droplets in an oil phase. We just enclosed it into the eutectic phase pockets in the ice phase and carry through our selection procedure and when we did that we ended up with this ribosome here You can see not only is it better than the wild type, but it's actually now more active and frozen than it is at ambient temperature So this is now perfectly adapted to the ice phase and When we combine this now with this hexa-nucleate type this quasi-shein-dal-garner sequence that we discovered earlier on we can now synthesize RNAs that are 206 nucleotides long. So this is the first ribosome To be able to synthesize RNAs longer than itself It's clearly an important milestone on the road to self-replication So I'd like to just summarize this part. So we found that I stabilizes the RNA ribosome structure and activity it enhances its activity actually while maintaining fidelity I haven't shown you that It and I haven't shown you that either. This is quite important in the context of the origin of life because of the tremendous Concentration effect that happens during the freezing This allows RNA replication from highly dilute starting concentration and it provides sort of a quasi-cellular Compartmentalization within the eutectic phase microstructure and supports evolution So clearly, you know ice is attractive and really we were inspired by previous work On on the properties of ice by the late Stanley Miller who's found that ice Catalyzes various aspects of nucleotide synthesis and also RNA oligomer synthesis from activated monomers So this is work by demer monar and Shostack as well as some RNA ligase activity that is Laurel Landweber's work and and really this provides the raw materials for RNA evolution in C2 inside the ice phase And how much time do I have? Okay, so for the last part of the talk I'd like to Tell you a little bit about some of our recent work On on other aspects of RNA assembly and replication. So if this is a sort of Cartoon of how we get from chemicals the primordial soup to Luca They're now some pretty good There's some pretty good evidence some pretty good studies that suggest how we can go from simple primordial chemicals to at least some of the activated nucleotides and also amino acids and lipids and Other people have found ways to assemble in an on-template way on templated way activated nucleotides into short oligomers and You know these may well be capable of non-enzymatic self-replication and it is something that among others Jack Shostack is exploring and Eventually there is a sort of idea that these will replicate and eventually progress to an RNA world Where at some point we will have to progress to enzymatic self-replication to get to look at the last common Ancestor but but really there's a little bit of a gap in between here between those short oligomers Which actually have to be quite short and enzymatic self-replication, which we think would probably require quite a complex beast of at least 200 nucleotides long so one question we asked ourselves, you know wouldn't it be possible to actually have an intermediate replication stage where Where the RNA transitions between? sort of pools of short oligomers and fully assembled Complex ribosomes this would also make the replication task easier because these short oligomers would be unstructured and therefore easier to replicate by the ribosome So we wondered can we basically can we assemble? the ribosome from its pieces and To do this is an initial experiment. We split it into four chunks about each about Between 50 and 60 well between 40 and 60 nucleotides long We use the hairpin ribosome This is a ligase or rather a right ligase slash endonuclease enzyme, which carries out transesterification reactions And again, we use the ice phase to drive the reaction and you can see you can you can you can you can get full-length assembly About 96 hours, but the yields are really not inspiring just 3% yield This is okay for some for some things, but really to set up a replication cycle that is that is really not sufficient So the question is why is the yield so poor? Now one hypothesis that we that we had is that maybe a large fraction of these ligase ribosomes Remain inactive in ice and why do we think that? Because misfolding is a pervasive issues in many functional RNAs This is because unlike with proteins the energy landscape of RNA folding is is rather flat So RNAs tend to get trapped in lots of local minimus and as this review says from 2008 nearly every RNA whose folding has been studied has been found to adopt misfolded conformations Now in biology. This is all solved by having a army of RNA chaperones and helicases which Refold and re-equilibrate the misfolded RNA structures and many of these are ATP dependent motors But we don't have this in the RNA world Or at least not at this point So what there's there to do so we need to find a way to refold RNAs maybe in a cyclical way to drive them towards their Most active kind of maybe most stable confirmation But the obvious the obvious regime to do that would be thermal cycling But we can't do that because high temperatures in the presence of magnetium will degrade our RNA So what can we do? and one thing we found that was rather Counter-intuitive that freezing also tends to refold RNA So ice can act as an RNA chaperone This just shows a hairpin Invasion assay you have a fluorophore and a quencher next to each other no signal in this structure But when this invading strand invades the structure you get Fluorescence turns on and you can see when you heat these Little hairpin structures at 37 degrees in about 60 hours the fluorescent turns on quite well but a single cycle of freezing will turn it on to a hundred percent and Maybe even more striking when you have an intramolecular Association so these are hairpins you have an invading oligo. There's essentially nothing happens at room temperature But when you freeze it the the The fluorescence turns on very quickly. So actually freezing acts as a as a RNA refolding engine So we wondered if we could now drive the assembly using freeze source cycling and we built Because it's a bit tedious by hand. We built a little Free source cycling machine in the lab. You can see this is the temperature profile cycling between minus 30 degrees and up to 37 degrees and holding for an extended time at minus 9 in the eutectic phase and You know you can immediately see this is much more striking We can now drive the yield up to 30% Thanks to the cycling and what's even more impressive the RNA polymerase ribosome activity turns on almost instantly in the very first cycles So that means we can now now we have an efficient process we can actually Fragment things into much smaller pieces. So this this is now the assembly pathway of the ribosome split into into four pieces, sorry seven pieces and Neither of this is 30 nuclear longer than 30 nucleotides, which is approximately what is within range from the prebiotic Oligomer assembly Processes and you can see again free source cycling can drive this to completion Now but all of these all of these RNA fragments were still pre-activated chemically to drive the ligation reaction because the Hairpin ligase ribosome requires a 2 prime 3 prime cyclic phosphate to to carry out the ligation But actually the hairpin ribosome also carries out the cleavage Generating such a 2 prime 3 prime cyclic phosphate So we wondered can we drive assembly even without activation chemistry starting from naked RNA Such a scheme would look like this. So we took this in green. We have the fragment we want to assemble We give it a little 3 prime extension Which is recognized by the hairpin ribosome for cleavage So it cleaves it here and generates the 2 prime 3 prime cyclic phosphate shown this red dot and then we go through a strand exchange reaction driven by the free source cycling which Positions the orange strand here, which is there to be ligated upstream of the 2 prime cyclic 2 prime 3 prime cyclic phosphate and now we carry on the ligation reaction and So this is the scheme now We now have we now have to proceed in three dimensions because fragments first need to be cleaved Before they can move down into the plane where the various assembly pathways can operate and just to show you in a gel How this works? So this is our first piece with a little tag it needs to be cleaved Now this is the cleaved form with the 2 prime cyclic phosphate that can be ligated And needs to be cleaved again So this is this one here that can be ligated cleaved again, and then we go full length. So This just shows that actually the assembly can work without any Preactivation chemistry from simple naked RNA Yeah with about 10% yield So I'd like to summarize so complex ribosomes can be assembled from pools of RNA oligomers less than 30 nucleotides long And they can assemble from these pools without Preactivation and free store and potentially other physical chemical cycles I mean, I think this is a area worth exploring can drive such assembly reaction by re-equilibrating a misfolded RNA Confirmers and they seem to function akin to an RNA chaperone by Effecting iterative refolding and re-equilibration of kinetically trapped RNA structures and complexes and Just to finish on I might like to make the case for ice as a medium that should be considered In prebiotic terms, you know if we think of our solar system as maybe a typical planetary system in the cosmos It's worth noting that ice is abundant Liquid water is not in fact, there's whole celestial bodies in the outer solar system that are built mainly from water ice and certainly freeze those cycles are still going on at least on one planet and With this I'd like to end and thank the really brilliant people who did all this work Peter Pinero is now a system professor at UCL London. He built the CST selection system Alex Taylor did the X&A after mirror and X&A's I'm selections. He's going to be a professor in Montreal and Chris cousins did the Scar analysis and yellow walk there a wonderful postdoc from Germany Did the built the CBT selection system Jamie did all the ice work and Hannah's did the free-store cycling. He's Now moving to Munich as a young group leader Max Planck. We had a wonderful collaboration with Peter Devine on the H&A and generally the X&A Field and it came in weeks on mapping out the X&A's I'ms and these but I'd like to thank these Various bodies for funding us and you for listening. Thank you