Engineering Yeast: Synthetic Modularity at the Gene, Circuit, Pathway and Genome Level

Video in TIB AV-Portal: Engineering Yeast: Synthetic Modularity at the Gene, Circuit, Pathway and Genome Level

Formal Metadata

Engineering Yeast: Synthetic Modularity at the Gene, Circuit, Pathway and Genome Level
Title of Series
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
Synthetic biology seeks to understand and derive value from biology via its re-design and synthesis using engineering principles. After a decade of work to improve DNA assembly and the control of gene expression, synthetic biology can now tack le cell-scale problems. By applying modular assembly from a kit of parts we can design complex genetic circuits that reprogram how yeast grows or endow yeast cells with new metabolic pathways that produce valuable molecules such as antioxidants and antibiotics. Or is aim to convert yeast into a prototyping factory for new phenotypes, and this will be aided by a modular synthetic version of the S. cerevisiae genome that enables evolution of gene content on cue. A part of the global Sc2.0 project to assemble a human-designed yeast genome, our lab is working on assembling synthetic chromosome XI and has already begun exploiting the new possibilities that it offers. After 2 years, we’ve nearly completed our 665 kb chromosome and have also developed new lab and software tools that will enable the future of genome engineering and yeast synthetic biology.
Area Potenz <Homöopathie> Synthetic biology Operon Gene Genome Multiprotein complex DNS-Synthese Wine tasting descriptors Genome Anomalie <Medizin> Systemic therapy Synthetic oil Chromosome Functional group Materials science Tool steel Zunderbeständigkeit Sea level Peptide synthesis Baker's yeast Modul <Membranverfahren> Process (computing) Stuffing
Sedimentation Flocculation Chemical element Bohrium Cell growth Area Memory-Effekt Gene Block (periodic table) Molecule Tiermodell Protein Library (computing) Methylmalonyl-CoA mutase Periodate Hydrophobic effect Lactose Surface science Organische Chemie Sedimentation Mashing Pigment Synthetic biology Gene Genome Tanning Systemic therapy Branch (computer science) Vancomycin Tiermodell Optische Analyse Azo coupling Cell (biology) Thermoforming Modul <Membranverfahren> Stuffing Pipette Polyurethane Differential calculus Tetracycline Wursthülle Cyclin-dependent kinase Man page Synthetic oil Deterrence (legal) Doxycyclin Oppression Volumetric flow rate Cell (biology) Gene expression Tube (container) Controller (control theory) Mortality rate Tool steel Genregulation Sea level Modul <Membranverfahren> Process (computing) Setzen <Verfahrenstechnik> Potenz <Homöopathie> Glucose Galactose Repressor DNS-Synthese Ethylene-vinyl acetate Genotype Volumetric flow rate Combine harvester Functional group Lake Pelletizing Pan (magazine) Genregulation Storage tank Hexagonal crystal system Baker's yeast
Bohrium Cell growth Zellskelett Binding energy Protein domain Gene Set (abstract data type) Senescence Anomalie <Medizin> Elf Fire Protein Library (computing) Periodate Hydrophobic effect Nucleobase Ozone Gene Wine tasting descriptors Electronic cigarette Branch (computer science) Systemic therapy Stem cell Vancomycin Sample (material) RNA Cell (biology) Metallmatrix-Verbundwerkstoff Polymer Dielectric spectroscopy Biomolecular structure Modul <Membranverfahren> Controller (control theory) Separation process Activity (UML) Sunscreen Wursthülle DNS-Synthese Man page Ring strain Deterrence (legal) Cell division Tea Table wine Volumetric flow rate Enol Lactitol Cell (biology) Gene expression Tetraederstruktur Mortality rate Tool steel Active site Genregulation Sea level Area Potenz <Homöopathie> Glucose Galactose Repressor DNS-Synthese Singulettzustand Sequence analysis Genotype Gap junction Gesundheitsstörung Functional group Pan (magazine) Genregulation Base (chemistry) Baker's yeast Screening (medicine) Gene expression S-Adenosyl methionine
Schwarzbier Ozone DNS-Synthese Plasmid Assembly (demo party) Constitutive equation Acid DNS-Synthese Genotype Man page Synthetic oil Metabolic pathway Functional group Gene expression Insulin Tool steel Library (computing) Sea level Base (chemistry) Process (computing) Breed standard Nucleic acid
Solid phase extraction Structural steel Singulettzustand Plasmid Assembly (demo party) Gene Erdrutsch Man page Anomalie <Medizin> Systemic therapy Deterrence (legal) Collecting Metabolic pathway Functional group Abbruchreaktion Azo coupling Open reading frame Baker's yeast Modul <Membranverfahren>
Digital elevation model Kohlenhydratchemie Carotine Cell growth Plasmid Growth medium Gene Penicillin Organelle Chemistry Genome Benzylpenicillin Anomalie <Medizin> Molecule Phantom <Medizin> Chromosome Protein subunit Enzyme Saccharomyces cerevisiae Marker pen Protein Stratotyp Library (computing) Xylose Breed standard Xylose Peptide Octane rating Pigment Penicillin Gene Genome Chemical reaction Electronic cigarette Systemic therapy Rapid Yield (engineering) Acepromazine Ribosome Beta sheet Stereoselectivity Modul <Membranverfahren> Biomolecular structure Rekombinante DNS Peptide Electron donor DNS-Synthese Man page Cell (biology) Abbruchreaktion Pigment Gene expression Amino acid Open reading frame Explosion Mass spectrometry Pain Modul <Membranverfahren> Setzen <Verfahrenstechnik> Glasses DNS-Synthese Singulettzustand Assembly (demo party) Plasmid Chemistry Antibacterial Genotype Metabolic pathway Functional group Breed standard Peroxisome Baker's yeast Peptide synthesis
Setzen <Verfahrenstechnik> Glucose Sunscreen Bohrium Singulettzustand Genome Gene Mill (grinding) Chemical reaction Genome Man page Anomalie <Medizin> Alpha particle Systemic therapy Metabolic pathway Cell (biology) Ballistic trauma Gene expression Insulin Library (computing) Explosion Baker's yeast Modul <Membranverfahren>
Digital elevation model Rekombinante DNS Golgi apparatus Arginine DNS-Synthese Genome Man page Genome Synthetic oil Systemic therapy Chromosome Synthetic oil Chromosome Functional group Cigar Peptide synthesis Baker's yeast Modul <Membranverfahren> Lymphangioleiomyomatosis Basenpaarung S-Adenosyl methionine Modul <Membranverfahren>
Digital elevation model Enzyme inhibitor Setzen <Verfahrenstechnik> Bohrium Cell growth Telomere Growth medium Metabolism Binding energy Gene cluster Gene Penicillin Set (abstract data type) Recombinasen Calcium hydroxide Genome Estradiol Chromosome Shear strength Insulin Library (computing) Carotinoide Xylose Telomere Cave Synthetic biology Gene Sand Genome Gene cluster Growth medium Wine tasting descriptors Electronic cigarette Systemic therapy Vancomycin Collecting Yield (engineering) Optische Analyse Hydroxybuttersäure <gamma-> Survival skills Modul <Membranverfahren> Stereoselectivity Rekombinante DNS Biosynthesis Chemical plant Secondary metabolite Man page Synthetic oil Ring strain Pan (magazine) Cell (biology) Chemical plant Gene expression Mortality rate Active site Flux (metallurgy) Tool steel Kernproteine Modul <Membranverfahren> Inclusion (mineral) Process (computing) Setzen <Verfahrenstechnik> Beer Golgi apparatus DNS-Synthese Rearrangement reaction Assembly (demo party) Translokation Plasmid Chromosome Synthetic oil Gesundheitsstörung Metabolic pathway Functional group Baker's yeast Gene expression Nitrosamine
Cosmetics Flocculation Insulin shock therapy Cell growth Secretion Growth medium Metabolism Gene cluster Memory-Effekt Gene Penicillin Organelle Cytoplasma Chemistry Anomalie <Medizin> Chromosome Lipide Library (computing) Cell membrane Survival skills Surface science Patent Rock (geology) Telomere Genome Wine tasting descriptors Branch (computer science) Systemic therapy Blue cheese Azo coupling Thermoforming Stuffing Sense District Reaction mechanism Density Chemical plant Wursthülle Secondary metabolite Ring strain Stress (mechanics) Cell (biology) Tube (container) Gene expression Irrigation Sea level Modul <Membranverfahren> Area Setzen <Verfahrenstechnik> Potenz <Homöopathie> Phase (waves) Gap junction Volumetric flow rate Synthetic oil Sodium Metabolic pathway Initiation (chemistry) Silencer (DNA) Genregulation Particle-induced X-ray emission Peroxisome Baker's yeast
they okay thank you for this opportunity to speak here in this amazing venue I'm gonna talk about some of the work we do in my research group on yeast so we do engineering and all sorts of different scales on synthetic biology systems in yeast and probably about two-thirds of the work in my group I was going to build up to some of our stuff and eventually talk about work on the synthetic genome project with yeast but Jeff is coming after me and he's going to give a much more detailed talk on that so I'll probably concentrate a little more on some of the stuff that the circuit level and maybe at the individual gene level in terms of using modularity as a process and maybe if there's time I'll talk about some other work we've done that answers that sort of answers more some of the questions that were posed yesterday so that my group generally works on foundational tools a whole variety of different things with an aim to try and make engineering of biology more predictable and something a bit more akin to an engineering discipline and we started now moving to apply some of these tools in different areas diverse areas such as materials production and therapeutics production in particular we kind of see that within nature there is some level of modularity most genes are considered as modules you see operons those modules and these scale as you build up chromosomes built from these various modules and so we can also see comparisons and this is from a Ron Weiss review paper with how modularity allows sort of scaling of complexity in other disciplines and how you can also see that naturally in biology I'm not a firm believer that these modular rules exist and they're never going to be broken biology is very good at finding ways to break rules but I still think that as we do work in synthetic biology and construct at different levels being able to use modularity gives us a lot of power and enables us to really progress in the subject very quickly so I first
started getting involved in synthetic biology in Jim Collins's lab back then at Boston University and what we did was work with mod the parts rewiring them into different circuits and you can see here a couple of different circuits we published on one of them being a feed-forward loop Network and one of them being a bi-stable toggle switch which allowed us to look at how noise is influences in differentiation of cells both of the networks you'll see here contain pretty much the same parts are the same modules promoters that are regulated at different times by different regulators bacterial regulators in this case that we've moved over into yeast and placed in the yeast genome and by just simply changing the topology of the interactions of these regulators we could get different effects which is kind of the way we're just with a very limited number of parts we can get a lot of different diversity in the behavior of yeast and so just to give you run through one of the examples of these projects we wired up these two bacterial repressors against one another in a what is effectively a leaky bi-stable switch system and this became a way to program in predictably the timed message within the cells of gene expression turning on it after a certain period of time so we call these kind of like slow release genetic timers so just to talk through this network design here we have two regulated promoters that are producing to bacterial represses these repressors when expressed will repress the expression of that opposite repressor so that you have either one system Domon one side of the system dominating or the other but because this is a leaky system almost always the 10-hour side of the system would dominate unless you came in and gave a signal to the cells which was an Hydra tetracycline you could also give doxycycline which blocks tatar from binding on his repressor and will then instead see the system start producing more Lac I so you can see here in our normal system we start the cells they immediately revert to their favored scenario where Teti dominates Tata dominates the system we add an Hydra tetracycline for a period of time and then we can wash it off the cells that has flipped them into being a Lac I dominated system and then over a period of time due to the the leakiness of some of the promoters here you end up with the system slowly reverting back in this case over about two days worth continual yeast cell growth so--that's many generations of yeast so this is a message that they're passing on to their progeny not by changes in the DNA so it's a form of epigenetic memory passing down and so this system very much works like if you had a switch on the wall and you flipped it up and then it slowly went tick-tick-tick tick-tick-tick over a period of time and then went down and we could plug this in to our yeast and see GFP levels change has the population group we then realized that because the nature of how this time of reset was to do with how much leakiness there was an expression from the promoters and how how well they were controlled if we did point if we did general mutation of regions of the core of these yeast promoters we could get a whole library of promoters that led to different expression levels but also different leakage of expression and we could then use these different regulated promoters in our system and we can take data from this part library put it in together with the prototype data from the initial version of the network we made and we could use something we could use predictive modeling to kind of work out different reset times from different combinations of of modular parts we're putting together so the way we do that
is quite straightforward we build our network we do mutation within specific regions of the promoters to get our promoter library we characterized the promoter library for both that output how much they produce when unregulated and there and there are kind of their input which is how repressed they are when Ted aural Akai is overexpressed and this data can then be fed into a predictive model and this was done by Zhao Wagon Jim Collins's group and we can then look to see when we have different combinations from these part libraries how long the reset time if this these generic timers would be and this these blue lines are the prediction space for these things and the red lines were actually when we went and took a few of these combinations together and then measured the reset times this is what we got some of them nearing almost a week of exponential growth before the reset gets all the way back to normal this is a good example of how modeling can be used in synthetic biology projects especially circuit circuit projects like this we had three promoter nodes in our system here and we had 20 promoters in our each of the promoter libraries that could go in here so we could have produced eight thousand possible networks that actually could be done probably now by PhD students in in good groups with good tools but back in the late 2000s when I was doing this work that would have been a hell of a lot of effort by me and Jim's lab with the basic tools that were available so it's really nice then that by having a predictive model we can then use that to define a small number to build rather than have to build very large numbers these networks were only controlling GFP expression so we wanted to show that they could be used for something slightly more interesting and we managed to discuss with Kevin first reference group who was then at Harvard about an interesting gene flow one which when knocked out of yeast or or under expressed in yeast the yeast grow normally they you know they float around within the liquid perfectly fine but when you overexpress it in serra vc i used you really start to get tight binding of the cells to one another and this is because it's this protein which forms on the outer surface of the yeast cells and acts like velcro getting cells to start linking together we call this process flocculation and so when you get enough cells flocculating within a population they then have a gravity and so they then sediment down and you end up coming in in the morning and seeing this pellet at the bottom of your tube and actually if you go to the microscope you often see very very tightly packed yeast cells sometimes even looking like hexagonal when i wanted to get microscope images in the lab sometimes i would have to be fishing these pellets out and then smashing them and breaking them off with pipette tips to be able to get some that i could see under the microscope so this is a really cool natural glue that yeast has so we linked up our timer switches to this system and now the different timers with combinations of different promoters in the key nodes of the of the circuits can now give us sedimentation and pellet formation within our exponentially growing cultures of different predicted times and again going on one here almost a week afterwards so this was a nice way where we could tie in a really interesting phenotype of yeast with genetic timers that we built and we'd use predictive modeling and we get new kind of phenotype scene and control of that as well we've been kind of pushing that a little further in my research group inspired by Royal College of Art fashion designer who came to our group for a few months with a wacky idea of that it would be great in the future that instead of mass manufacturing of clothes where all the clothes have the same pattern instead you have microbes performing the patterning so that every one of these mass manufactured items of clothes has an individual pattern so she investigated how you could grow bacteria in patterns and we got bacteria to secrete a pigment molecule as this was as they were growing and we even tested this out on fabrics and showed that this was okayed in an autoclave and take out of the lab with this patterning but a goddess thinking about whether we could ever make patterns with the way yeast grows so he snore Mele grows obviously as a single-cell organism which then buds off and then another single-cell organism grows but in a colony what happens is they all start just growing on each other but yeast also has hidden away in its genome and has the genes that allow it to do kinds of multicellular growth most of these have been I've got now mutations within them that silenced this because the lab yeast we use we don't want to see that happening but you can resurrect these mutations and start getting things like filamentous growth pseudo Highfill growth out of yeast things that are seen in natural yeast and filamentous fungi so there is a poster I think outside here where someone else is looking into this is the person who's posted yeah I haven't had a chance to talk to you yet but it's super cool stuff so we will we then wired up these some of these genes my PhD student was very happy that one of the key genes involved is the gene PhD one so he gets his PhD straight away with the mutation there and then so now we have using the same promoters that we used in the previous network we can link up to control of flow eight PhD one and also another one cdk eight and so now we can have extra cellular control of filamentous growth so that these cells instead of growing and budding off from one another when we hit them with specific induces they now grow like this where they don't leave they grow more elongated with branches coming off so they're starting to look a little like a branch fractal but one that we don't have much control over so
just to show you how they grow this is kind of like filamentous growth where we've induced this and we managed to do this in haploid strains which was the first time people would have done that because it's normally only seen in diploid strains and you can see their branches come the filamentous growth coming off that occasionally branches coming off that as well and we can put the whole system now wire it up so it's linked to our genetic timer switches so over a period of hours and days you see filamentous growth initially and then as the timers start to release the system you'll now start seeing them start forming colonies as normal so they start filling up in the gaps between the branches my student George is finishing
off work on this we currently have unregulated branching which is which is great for making filaments but actually being able to control the pattern really requires us to be able to control the branching raid and particularly when and where how often you get a branch for me so what currently happens is a cell will divide off and make the next cell and this will make the filament so this is we call it in yeast the mother cell and then in daughter cell and what we're trying to achieve now is differential regulation between the mothers and the daughters so that the mother cells then enter a period after they've produced a daughter cell where over a certain amount of time or a certain amount of gene expression they don't produce another daughter cell so that prevents them making another branch but then after maybe several hours or days we could see that signal come back down again to allow it to then produce another daughter and that would then produce another branch into the pattern here so to get that we're working very hard on regulated promoters that are regulated so that they're only expressed in the mother cell and of the daughter cell and those can be tied up to senescence genes we have some initial work there that I'm not going to show here that that seems to show that that is working as well that kind of piggybacks on some of the work my group did earlier on a few years ago where we looked into designing promoters from the bottom up because what we need in that scenarios we need a promoter where we know we can have repressible external irrepressible expression that we also needed that the regulation turns on in mother cells but not daughter cells that's a complex conundrum for us because we have to take regulation that is found naturally in yeast in upstream activating sequences but we need a core promoter that can be repressed so we're starting to have to make hybrid promoters where we're bringing in different regions from different promoters putting them together so that our logic can be like we want it to be so we started doing work towards this end on a different set of promoters a few years ago and this was published way back in 2012 was one of the first papers from my group when we first got going five years ago and this was we we looked at could we kind of start with a very very boring promoter that worked as a promoter so that genes would be transcribed in yeast and then could we begin layering on our own regulation on that in a more predictable way as opposed to taking an issue initially well regulated promoters and trying to work with those where they have a lot of things written into their sequence which define when and where they're switched on so we did possibly the world's most boring bioinformatics screen which was we looked at thousands and thousands of gene expression data banks that were available online to look for mRNAs that almost never changed their gene expression in lots of different conditions because we wanted something that was just on and always on this gave us a sort of a top ten list here and one of them we settled on was the profile in one promoter pfy one this produces a cytoskeletal protein quite a small one and what we really liked about this promoter apart from being very very constitutive in that its expression levels don't change much is that it's also very short and very simple it simply has a poly T track which is bound nearby by rebbe reb Wan protein and then it seems to leave a core promoter region here which RNA polymerase can keep binding to and firing to just keep doing gene expression of whatever you put downstream of this gene and so we tested that in a variety of different conditions here and we always saw very close constitutive gene expression at different times in exponential growth of yeast and in different conditions as well and then we did this trick again that we'd done back in the previous work in 2008 where we mutated bases just in a very specific region in the core region of this promoter and we showed that by mutating all these bases you could select out a whole library of promoters that the slight changes in the base sequences there now lead to the promoters having different strengths but they're all very constitutive that also showed us as well as giving us a library to do lots of different levels of gene expression in yeast it also showed us that there were regions within that core promoter where you could change the base sequence you would still get output so that means we now have a kind of an area within the promoter where we can start adding in our modules so simple one we did first was to add in the recognition sites for the bacterial tet repressor so that this now switches from being a constitutive promoter to being a regulated promoter that is usually repressed in the presence of tet repressor and then as you add in increasing amounts of and hydrojet recycling you can switch this on so this allows us now to become a regulated promoter at the time back in 2012 the the nuchal in vogue thing or 2011 actually was tala factors and we showed that tala factors which are very modular DNA binding proteins where you can swap out domains of the protein so that they could target any pretty much any different DNA base sequence we could then get them to target the different sections of our promoters and they would be what we call orthogonal which is that one repressor would specifically bind one promoter and switch that off whereas the presence of another repressor would not interact with that promoter and so that you would have a very defined relationship so now we're able to scale this so that we've gone from one boring pfy one promoter to now having multiple regulated promoters with multiple regulators for each one DNA 2.0 offered to analyze our mutation sequence from this library they went through this sequence and they have an algorithm that looks to add all the base sequences within the mutation library and what the output of that promoter was and then tells you the relative importance of every base as well as they can from the very small data set we provide them and what was quite cool here is this led them to initially say to us ok we we think based on this sequence analysis we could design you a strong promoter from this relationship here and this is the one they designed pretty strong not the strongest but pretty impressive that they get to second best just with their first guess but this also gave us much more insight into the bases that are available within this region to mutate which ones we need to keep as a certain base if we want to maintain strong gene expression and which we can tolerate all sorts of mutations with and so then for designing all of our further regulation we found that there was a specific region where we could change a number of different bases as long as we kept this T and this a the same and we could generate all sorts of different targets within that region which could all then be individually targeted by different tell effectors but the gene expression that would come out from the promoters would be exactly the same in every case so you have a whole whole load of different regulated promoters but they behave almost exactly the same both when activated and when repressed and so we haven't yet published this work we're kind of wrapping it up at the moment but this is how the orthogonality scaled this was the six by six matrix Ben's taken this up now to a 10 by 10 matrix or you can see all the different telefax are based repressors matching and repress down only at the specific promoter pair that they're supposed to be with so this gives us a whole lot of different tools to do the rewiring here just to cover we've also done other
promoter libraries here this is constitutive promoter library this one's based on the very strong ADH one promoter with a whole range of mutations and this was done in a slightly different way because this is a much longer promoter region where we used highly new degenerate PCR to introduce about 10% errors into the promoters and this again leads to a different level of expression what was kind of cool about this was this was done in the same pot as doing DNA assembly to make the plasmid of the first time this was part of work and we've done lots of work in my group and with Jeff
Baldwin at Imperial College on different DNA assembly methods and this showed that in using a certain method you could in the process of constructing your plasmid you could say I would like this part within the plasmid to appear as a library of mutation in the same go as you're doing the whole experiment and the assembly so on the same day that you assemble your plasmids you can also be defining that one position within that network or within that pathway is mutated so DNA assembly is something we've like I said we've done a variety of different papers and reviews on and I just want to now move on to talking about some of the success we've been having since we moved to Golden Gate
based assembly we aligned ourselves with John DuBose lab at UC Berkeley who developed this fantastic moko use toolkit which allows you to take a whole load of different modular parts assemble them together first to make kind of a gene system so to an open reading frame a terminator and then you can take multiples of those and then bring them together to make pathways or to make signaling pathways or just multi-gene systems and the kit that you can go out and get now in a gene if you're a yeast group and you want to again using this contains a whole variety of different parts there's about a 96-well plate is what you get in our own lab and in John's lab the these collections have now ballooned to many hundreds of parts and actually John sent me this slide a
few a couple of months ago looking at when his lab switched to everyone on the yeast project using this specific framework of modular mokele obaid assembly the massive increase in the productivity of the lab in terms of how many plasmids they're they're all making and I certainly think it's been the same in our lab since we moved to that about a year and a half ago so we'd be mostly
using that for pathway engineering and we heard plenty about pathway engineering yesterday without having tons and tons of math specs like they do up at Manchester we've instead been looking at pathways it's very easy for us to see a good output for so Geoff's group have been helpful in sending us over some pigment pathways we've also been working on selectable pathways such as the production of the antibiotic penicillin and selectable growth pathway is a commonly used one in yeast researchers introducing the genes that allow yeast to grow on xylose a lignocellulosic sugar that is it doesn't normally grow on penicillin is probably the one of most interest and it's one that we've really enjoyed working on and we're just wrapping up the work now as a paper and so this is how penicillin is made firstly a gene condenses two amino acids and an unnatural amino acid into a molecule called AC V which is very quickly then modified to become I so penicillin in the peroxisome of the filamentous fungi penicillin this is then modified to make pengie which is then secreted out of the cells and this goes out into that as external media outside the penicillin cries odium cells so we decided to put these pathway libraries these guys all on a plasmid to add to yeast this first gene here the a CBS is a type of gene called a non-ribosomal peptide which is a assembly-line enzyme which is effectively one protein but it actually encodes multiple different enzyme subunits within it catalyzing lots of reactions as it goes along so you can imagine this as one sort of multi modular enzyme and as such it's very big it's almost bigger than any gene naturally found in the yeast genome so it's a bit of a pain to clone because 12 KB you know PCR is kind of off the table unless you can guarantee really good accuracies but we managed to clone this and have it under a regulated system into the genome of our cells and so that goes into one of the chromosomes the rest of our pathway goes on plasmids and we use these modular DNA assembly
kits to put in some different promoters different versions of the open reading frames that encode the enzymes adding on tags and different three prime UTR or Terminator sequences because that can also affect gene expression in yeast and when we put all these together combine them first into their individual gene plasmids and then assemble them into a final these pathway plasmid we can start getting ourselves producing penicillin so firstly this is the pure chemical standard and the mass spec trace shows that that big gene on the enzyme is working so we get the non-ribosomal peptide synthesis working in yeast and it produces the first part of the system which is a cv and then the rest of the pathway then allows us to get pengie production and what's really great about this actually is that the pengie is highest in the external media because as well s cerevisiae is secreting this out naturally so we find much more of the penicillin G in our growth media of our yeast than in the actual cells themselves so that makes life very easily and what was great about the moko yeast toolkit in this system was that the use toolkit included localization tags to send certain enzymes into compartments of one of those was the paroxysm targeting tags so for example it we when we got the kit from john's group we immediately ali immediately stuck fluorescent reporters with the sequence tags on the end and turned on the gene PEX 11 which causes overexpression of peroxisomes in the cells and we could see all of these expression now of these proteins these enzymes being sent into the peroxisomes and many more large peroxisomes in our cells and that really helped us boost the yields in our pathway and especially boost the yields we were getting in the extracellular media as well because so much more of the reactions are occurring now in big peroxisomes that are fusing and leaving leaving our yeast we've also slightly improved in the last few months this mock low CRISPR system moko system to include CRISPR so now we have a very fast protocol in the group for putting in a variety of different things into the cell at the same time and at the same time making changes on the genome and what happens here is this is so efficient now that you do not need any selectable markers on the things that you're integrating in because in yeast the CRISPR is cutting the genome until you fix the sequence there so that sequence doesn't exist and all the cells that can't repair that sequence are going to die and you then provide in donor sequence maybe the gene that you want to add in or the mutation you want to put in to the to the yeast chromosome and that will repair in very efficiently because yeast is very good at homologous recombination and so then you will end up with all your changes so we'll first go trying this whole protocol you had greater than 90 percent success rate with no selectable markers doing simultaneous two deletions and two insertions into the genome so this is now scaling up and so the sort of things that people in my group well is just a first year PhD student he's now doing
things like looking at signal signaling pathways looking at doing multiple knockouts of different components at the same time as putting in refactored components now with a whole variety of different promoters to change the gene expression of the components in a single empath way and so in in a one one shot
experiment here as an example he knocked out four genes on the genome and added two more genes back in but on the library of different promoters and was able to pick out colonies that then maintain the signaling system so they still respond to M F alpha this mating factor so that increased mating fact that leads to increased expression from the reporter gene not quite the same as well type it's a bit more leaky but if you see if you haven't added these genes back in the whole thing is constitutively on and so this is pretty impressive cuz this is six different changes going on in just two in a one week in a one-pot reaction in the cell so the last thing I was going to talk about on yeast was the module at least
yeast genome synthetic Keith but I'm gonna let Jeff give you the whole introduction here but I just want to point out that the reason why I call it the modular modular yeast genome and why I'm so interested in working on it is because it has this recombination system written into the chromosomes that I'll go into in a second our lab is involved
in building one of the chromosomes chromosome 11 with the kind of scary number of 666 killer base pairs inside and I'm happy to report that we've got all of our DNA gets synthesized for us by Gina and gets sent like this it's very boring it's just a FedEx package that costs about the same as a small car and then been in my group with the assistance of Maureen has been putting this together and 76% of the synthetic DNA is now integrated into our chromosome so we're nearly done and you can see our progress on a
resource web page along with the progress of some of the other chromosomes in the project but why I'm calling it a modular genome is because the design that Jeff and others came up with was to place a recombination site looks piece in landmarked throughout all of the chromosomes immediately downstream of all non-essential genes and an other significant landmarks this is a recombination site where if you provide a recombinases this site and you add estradiol which is induces it to go into the nucleus it will bind together two of these sites link them together in a loop and catalyze a rearrangement at the DNA leading to either translocations so that one part of one chromosomes shifts to go to another part of another chromosome inversions so that a region of the genome flips over so it's now pointing in the other direction and also deletions that's a very important one the genes could be lost and potentially as well insertions if you provide new DNA at the same time that's within this format and so this is from one of Jeff's papers this shows that the majority of the time if you have synthetic chromosomes in your cell and you induce this system most cells are going to die and so the reason for that is quite obvious because they're deleting or rearranging essential genes so that the new versions of the chromosomes that arise from this recombination are not viable so you get lethality but what's very cool is that many of these colonies here there are survivors right so these are cells that have maybe slightly rearranged their chromosomes or even dramatically rearrange their chromosomes but they're still viable so the modules of the individual genes have changed in both in terms of that content some of them might have disappeared some of them might be present in multiple copies instead and the arrangement of them have all shuffled around but the cells still grow and so I'm very excited in this because gene expression has been shown by others to be significantly affected by the location you are in in the genome and things like the secondary metabolite pathways like the penicillin pathway were working on and others and are often found to cluster in plants and fungi you would have seen that from an Osborne's talk earlier this week and here's an example of a metabolic pathway that's clustered through evolution in s service yeah the DAO cluster and so the modular genome of synthetic yeast allows us to to look at this in a way and try and understand how a genome could be laid out to do things like optimize biosynthesis and then eventually one day I would love the challenge of having a full set of parts and thinking from bottom up not copying what nature does how would you put these various modules together to make a viable genome from scratch to give you an example of some of the effects of genomic positioning with tools that were sent over from Geoff's group we put in a violation pathway in different positions around a partially synthetic chromosome putting it in the middle of the chromosome putting it up close to the telomeres spreading it out so there at either end of the chromosome close to the telomeres and the production of this pathway if it working perfectly well violation pathways should produce some very dark purple pigmented colonies and what you can see is certain versions of these constructs do that very well others the whole system doesn't work possibly because of local silencing close to the telomeres and other ones are a bit mixed and so we even occasionally saw colonies that were half white and half pink because some sort of silencing effect was going on within that colony where we're going with this kind of work is that we think that the Scramble plus the heterologous pathways is is a good interest so we could put our carotenoid pathway pan' pathways ILO's pathways in cells that have partially synthetic chromosome regions or even full synthetic chromosomes we can induce this scrambling system and what that does is that creates a whole diversity of different yeast strains that your metabolic pathway is operating with maybe more favorable you can even rather than construct your whole pathway you could provide your genes as just units that are ready to be scrambled in and we did that with the xylose pathway which is two to three genes best if there's three but it can survive with two and we were able to do this rambling and pick out a strain which actually started to grow on xylose media because it must have inserted these genes in the genome so where I think this is quite useful is in the metabolic engineering world linking that in with synthetic biology the traditional method is quite to get high yields is quite multi-step you would initially design and clone a version of your pathway put that in your cells test it and then you'll go back and start swapping in alternative genes different promoters of different strengths to try and optimize the expression and flux of the pathway you'll then probably take your best winners out of that take them out and now move them into lots of different strain backgrounds so you might do lots of different strains that people have worked with before that show good yields in industry or you might do put them into knockout collections because it may be that certain genes when knocked out will help improve your pathway and then finally probably the one that in industry really matters is optimizing the growth conditions for scale-up and so all these tools the modular DNA assembly with promoter libraries part libraries plus the synthetic yeast can really narrow this down so that this whole process can all be done so you can have one pot assembly of many thousands of versions of the pathway and you can have one pot strain evolution so that your your best pathways here can then be just scrambled so that the all of these different strains are created in the same pot and you can then look for those as long as you've got a good selection system and you can go straight up to optimizing conditions to get you a better yield and this is part of a project that I had fun I've got funding for at the moment where I tried very hard to get a really good acronym for my project so this is it automatic chromosome rearrangement for optimizing novel yeast metabolism I was very glad the funders saw that as a good acronym and a lot of these methods that were developing we're going to be teaching next summer at a synthetic genome engineering summer school that I encourage you to send your students or your cells ask your supervisor to go for it will be heavily discounted so it'll be very good value 5-day residential summer school up in Edinburgh looking at things like CRISPR mediated integration synthetic yeast chromosomes scrambling evolution and it's going to then lead on to a brilliant two-day conference on synthetic genomes and synthetic yeast which just will give a mention off and a final thing I would like to shout out is that there is still time if you have
researched that Patrick tie up in Edinburgh and myself are asking for people to send in European based research to a special issue of ACS synthetic biology this was originally with a deadline at December but they've extended that now so you can get it in by about mid January so let me know if you're interested in sending in any research so I'll just wrap up now because I guess I'm almost out of time and thank my group this is most of us
proving that the density of rock is significantly more than the density of humans so whatever we do we can't unbalance this rock up here on the top of a mountain in Wales and I particularly want to highlight the people whose work I spoke on then Ben who worked on the tell effector and the promoter library stuff and who's also now leading the work on the synthetic yeast chromosome George who's been working on the patent formation with the yeast colony growth from filamentous form and I'd also like to particularly mention Ali who did who's doing all the work on the penicillin pathway and the collaborators here and especially I'd like to thank Jeff a great collaborator as well and my funders so thank you we can see just humble quick exceptions your palm and then we'll take a five minute break before we have right so I was interested by what you said about the the position the genes that encode the secondly metabolic pathways they often cluster in the sock telomeric area and this is as you mentioned is known to correspond to silence silence regions and generally however it has been shown at least in the case of low-level irrigation that some stresses will wait partly tells us from the nuclear ample and will in conduct in reducing this silencing there is in over expected over expressing the softer American Telemedicine so that's in some sense could it be for the first of all could you be that the chemical stress or has it been shown that chemical stress well induce will induce this response by inducing the pathways of secondary metabolic makeup much like the ones you mentioned is it no and and second of all is it possible that or do your data speak to this phenomenon and we are shown that you put your pathway genes at different positions could it be interpreted could the results be interpreted in any way that says that they are induced there they are silenced and near the pillars and the door induced by okay so it's a really good question I had not heard that hypothesis before so there's I speak with an I was born about this quite a bit there are generally a couple of hypotheses why you have clustering of secondary metabolite pathways in things like the plant genomes and the filamentous fungal genomes and so one of them is deemed evolution that by having these genes in the sub telomeric regions where you're more likely to get crossovers between chromosomes it leads to quicker evolution of new secondary metabolites another one is co regulation it's obviously you don't want the secondary metabolites to be continually expressed because they're often very stressful for the cells so be nice to have them in an area where it's easy to Co regulate in silence the whole one I think your hypothesis that adding a stress that then turns that on is a is a very good idea as well that would I could also see that being a benefit the problem is I don't think anyone has studied that because these pathways are where they're naturally seen is in hard to grow hard hard to work with filamentous fungi and with plants as well so those kind of experiments very difficult so one of the main reasons we're doing this project this acronym project is to put in a selectable secondary metabolite that the cell might use to defend itself in a way which is what some of these secondary metabolites in the plant systems are so the production of an anti antibiotic so that it can then compete with bacteria that are in the growth media and so that we have a system then within yeast which is very tractable where we can then start to ask a whole load of these questions to see whether these things are you know good can we have the whole thing repressed in terms of its expression and then induce of stress and we will see more survival that's something we could test that's a really nice idea so just as a for I'd like to make it clear it's a fact that low level irrigation which softly and over expresses the subtalar marriages this suspect the hypothesis is that it could happen with other types of stresses and would it be possible to use the procreation for wanting to be mentioned I think so I think when we started that project we looked at a variety of different things and we just saw that the Sudha - growth was very kind of close to where we wanted to get to anyway which was these branch fractals so we stuck with that but yeah flocculation is a fantastically interesting phenotype as well did you try greens in with the flocculating strains that I showed at the beginning the flow one strains yeah they would form pretty horrible colonies unless a blue was repressing the the flow one gene and then but most of the experiments we did way back when we were doing that flow one over expression was was mostly focused on we would keep the colony not we would turn off flow one in colony growth and then switch it on specifically for the exponential growth phase so that we could then get those pictures of the different tubes with the palette forming over time Tom so I guess another question related to the growth based patterning I was there is a particular advantage to doing kind of growth based as opposed to kind of self patterning techniques it's actually printing cell type printing patterns that people are they have your pattern just print the cells of different types of cells and things like this is there well so one of the justifications for doing this project and actually it's kind of cool because the funding agency the labor hoop trust who this nice charity in the UK they they actually encourage you to submit projects that you know you can't have a good justification for like in terms of commercial they they say you know their remit is to fill the gaps between other Research Council's for the interesting cross-disciplinary things but we did we do think that in 2d growth a growing branched fractal is one of the highest surface areas and a lot of what you need for you know testing out things in metabolic engineering might require good surface areas so good secretion as opposed to if you grow the cells as a single 2d colony they're just going to grow on top of each other now if you were to print the cells in a 2d branch fractal we're still gonna want them to keep growing because we you know all of these pathways run best when the cells are growing pretty well right so they will just start growing and they'll start growing on top of each other filling in the gaps right and so there will be dead zones there will be areas in the middle where the existing cells are no longer exchanging nutrients because they're surrounded by other cells so they don't have access to the full surface area what you said quite quickly about the peroxisomes initiative directly at the Sun membrane so I think in east unconventional that huge record yes and memory so is it that you are meaning it as anyway yeah so yeah I did miss I said that wrong if I go back to the pathway actually here they they do they don't go at like the paroxysm doesn't fuse and the whole thing empties out it diffuses from the peroxisome out into the cell but the overexpression of the peroxisomes led to much higher levels that maybe because there's much more catalysis going on because we have bigger peroxisomes or it may be because now there's more peroxisomes closer to the membrane so it's coming out a lot more but the ratio between product that was secreted versus not secreted went up when we increased the peroxisome so that means that there is a mechanism to transport the penicillin from the cytoplasm - yeah it's this part of the machinery that you know that's something that occurs yeah question to you first of all do you think that this is a paroxysm merge with the membrane but to induce the formation of filamentous i'm not sure that those two things are related the these these don't form filaments in this project this is a bit different having said that the the big penicillin producer Pico sodium is a filamentous fungi and it grows this filament so that it can have good surface area well I have to go and look that one up I don't know if anyone here knows what the pix 11 is involved in lipids