 So, ladies and gentlemen, dear colleagues, dear member of the FBM, many thanks to have answered to the invitation of the CIG and the FBM to honor Andre Stajak today for his inaugural lecture. He will speak today about revealing the secret of chromosome organization by numerical assimilation. But before I let him the stage, I will just introduce you in a few sentences and using a few images Andre and his work and his parkour or his scientific parkour. So, first, Andre is coming from a far, far away country, Ezekiel here. Ezekiel is from Poland and he studied, he was born in Poznan but then studied in Warsaw where he did his master, he was interested in microbiology and during his studies he's become more and more interested in electronic microscopy on EM. And so, basically throughout his career and interests in EM, he went to different places and actually during his PhD in Warsaw, he went for two stages in Professor Kole's lab at the ETH in Zurich to learn and basically carry on in his learning phase on electronic microscopy and this is why Professor Kole asked him to join his lab to perform a postdoc in Switzerland at the ETH Zurich and this is where basically, and I use this image this time because you will see that he's bound to Switzerland somewhere as, you know, pictured here by this ring and because he's still now in Lausanne, it's also, and I open a parenthesis, it's also that he's not alone here and Alicia is together working with him and I think it's a good sign that there is also this ring here, this note here, which is shown here. So basically ETH Zurich then, so through contacts and that I think is an important message for you younger students or younger scientists is that he went to courses, he had contacts and so he was invited to do a postdoc at ETH and then during his postdoc at ETH, he came here, invited by Jacques Deboucher to give a seminar and they saw that this was, so they have interesting points and interesting question to answer together and this is why Andrei basically decided and was invited to join the faculties here in Lausanne and he was in the Bio4 building and he moved into general broad building and joined CIG, what is it now, ten years ago roughly basically when the CIG was created. And so I will also show you that, I don't know, I mean do you have, do you have one ancestors war, it's a marine, it's a navy, no, boy scouts eventually, because there is a passion for notes here, okay and not only there is a passion of notes but I think what it's beautiful in science is that it can be also picturesque and throughout my presentation you have seen that not only Andrei is asking interesting question but he was able basically to answer those questions in providing also the message with beautiful images and so there is also an artistic touch that I think is important to see in science and in nature in general and so basically to untie the notes and to finish here, please welcome the king of the notes and I think he will disentangle those notes in his presentation. Congratulations Andrei, the floor is yours. Alexander thank you very much for this very kind introduction. I will talk not much about notes but then so I wanted, well first to go to the title of my presentation and so what I will be presenting to you is how we can study a little bit organization of chromosomes by simulating behavior of microscopic chains. Those chains which we will simulate they will behave like fluctuating chain, very microscopic chain fluctuating by thermal motion and that's what we can simulate. We can put some constraints on the chain. We can put a torsion that they will be supercoiled and now from those observations which we can simulate what we can conclude about the behavior of chromosomes. So that's let me go. So where we are interested in mainly in metaphase chromosomes, sorry, we all know very well metaphase chromosomes and now this type of image is something which everybody knows for as an image of a chromosome. This is also a little bit in graphics of Seizhe we have such images of chromosomes but in metaphase that stage where cells divide DNA is just packed very well for the division. It's not the time the DNA is doing its function. It's not the time when DNA is transcribed. It's not the time when the DNA is replicated. Really interesting things happening in the interface which means after division when chromosomes kind of swell, they become less organized and it was a time people were thinking that they just are kind of disorganized. Now we know that's not the case that chromosomes maintain very interesting features after they are decondensed from the metaphase state. So that's the image. So again we start from interface then when we go through division we see those chromosomes separating and then after the division then again they decondense, they feel the look close and that's the very interesting state which we would like to understand. Now there are many secrets of chromosome organization. I wanted to discuss two today with you. One is what is the structure and the mechanism of formation of tats. So what are the tats? Tats are regions within a chromosome where there is a kind of higher frequency of contact. This drawing here is a cartoon. It's only a cartoon. We really don't know exactly how the condensed chromosomes look like. There is a technique which is called 3C technique which allows to find out what are the regions which contact each other. And this information, so one can sequence the regions which contact each other and this information is used by many groups to reconstruct the chromosome structure. So this is a cartoon and now in this cartoon what is drawn there are blocks of chromatin fiber so that is a chromatin fiber, there would be nucleosomes. Then there are some genes here, what we see is some interactions between enhancers and promoters and there would be expression. So what is this organization? And this is now the main information is through this technique of 3C which is chromosome conformation capture. And so those are how the data from those 3C techniques are presented. So imagine that again it's only cartoon, we don't really know how it looks like but now I'm stretching it here, this fiber here and now this matrix here tells me what's the frequency of contacting, so that point here will show the frequency of contacting that region here with that region here. So that's the information we have, this we can get to quite high resolution what are the regions which contact each other. And so now out of those information this 3C information which is the contact matrix or contact map of chromosome one is we will try to a little bit to get back what is the structure of chromosome. Then the second question which I will and the second secret which I will try to discuss it is what is the mechanism by which enhancers find the target promoters. Enhancers are the distal regulatory elements. They can be as far as a megabase from the promoter. Now for the expression of developmentally regulated genes it is required that enhancers come in the contact with the target promoters. And then only the developmentally regulated genes are expressed. And so they are far away, they find each other and it's, and they find each other in very kind of complex structure and so this is in a way a problem how two needles can find each other in a high stack but full of other needles. So this is even more difficult at finding a needle in a high stack. So again here would be enhancer promoter contact, sorry. Now how we model our chromatin fibers. So if chromatin fiber would be like this with kind of, now we are mostly thinking that chromatin is not forming 30 nanometer fibers but 11 nanometer fiber or 10 nanometer fibers. So that if that would be individual nucleosomes we kind of average for this which means we don't take into account individual nucleosomes. We treat the chromatin fiber as an elastic filament, elastic filament which would be here I have for my lectures on DNA topology but we treat chromatin fiber as elastic filament and we know from biophysical experiments how difficult or how easy it's to bend it and what is the, if you turn it like this what is the resistance it provides. And so we put this so we don't see such details like very important details of course where are the linkers, where are nucleosomes, how it affects. This will be left for let's say for future years to go to higher kind of accuracy of simulations. Now we are doing something which is called coarse grain simulations where we average for the properties. So the fiber which we can which you could measure how it reacts to the formation we model as an elastic filament. So now about those organizations of chromosomes so in a way it started relatively recently in 2012 several papers at the same time observed and by using those high resolution free sea techniques which means they found out more than billions of contacts in one experiment. What are the regions which contact each other and out of this they got those maps. Again so if you think about that say here it is a chromosome 2 it's region and then you would see that here there is a region which shows increased frequency of contact. Then there is something like a barrier and then there are not much contacts between that because if I let's say take a point here this point has almost no contact and that would be contacts between those regions. So that was an observation and then the question which came right away is what is the structure and function of those topological associated domains which are called TATs in interface chromosomes and how TATs form and that's what we try to discuss today. For the function it is now most it's almost accepted now that the function of TATs is to divide chromosomes into regular regulatory regions and enhancers and promoters which normally interact with each other are always in the same in the same TAT. So there was this interest how TATs what is the structure of TATs and already in this first paper which proposed which observed those blocks of increased contacts the authors and the authors is Dixon et al. proposed such a model or a cartoon and where they were saying well one TAT would form a kind of a blob like this then there would be some separation and another blob like this and then if you think about this structure fluctuating then of course this region would contact frequently this one it will contact this one and so if you measure by this high C techniques what are the regions which contact each other then you would see that there is a lot of contacts here that there is a lot of contacts here and not much contacts here. So this cartoon is nice but it somehow does not tell us what keeps this chromatin to form a global here and that's very important and what is the boundary here and so that's so right away after publishing this paper in 2012 groups doing molecular dynamic simulations similar to what I will be presenting tried to reveal what is the structure of TAT and so a group of Mario Nicodemi proposed a following model they proposed that if we have a region of a chromosome that there in this there would be a specific that could be specific sequences in that region which would bind a specific protein and then a next region would have another type of sequences which would bind another another protein and then if we have those binders so this was strings and binders model if we would have binders which will so there will be region which binds green protein and then the region which binds this pink protein and if we put green and pink proteins then what will happen there will be this kind of separation and so that was the simulation and then one could also simulate how what how the contacts will come in in the in the system and again this was a nice model in a way it it was the first model which proposed a realistic mechanism which kind of explained why there could be this enhancements of contact before it was there was no mechanism there was just a proposal that they are separated region now the problem with with that model is that it requires many different specific binders for different TATs so here we have just two but if it's long and and in the chromosomes we have thousand of them then we would need a different for at least for those which are not too far away which can contact each other because otherwise if there would not be different those TATs would fuse with each other and so so that was a problem with with the model and then the model also does not explain that in fact TATs can fuse when one removes the border element so if let's say if if I have here the border between this region with which binds green proteins and the region with binds pink protein if I remove the border element nothing will happen but in reality in experiments then they they fuse and so so there is a there is this need of other models and a little bit we were thinking about those other models and somehow to have an elegant elegant model and so in that elegant model so let's say that's a chromatin fiber in elegant model every TAT should be undergoing the same type of interaction and and so what we were thinking and then every border should be of very similar similar character so in in this type of a model here so we put so that would be a TAT a region between now we put here an attachment point it's a little bit like a matrix attachment sides which and now what could be the this interaction which would lead to increased frequency of of contacts between those regions and then we thought well supercoiling could do it supercoiling because now when we have of course now we have more contacts between region and what in addition there are those contacts are increased here but now this distance here increased so there are less less contacts between so that's that's nice so now what would happen if if one would remove that border that's like how it was done in in experiments and in experiments when the border was removed then I think the two TATs were fusing so if if you would remove that border it would just form a longer longer supercoil and so that would be a fusing of of TAT so we thought well that that could could could work and but there was a little bit of problem supercoiling is very well established in bacterial cells in eukaryotic cells supercoiling is generally let's say in the general textbooks you would find that chromatin in eukaryotic cells is not supercoil okay but but at the same time you would also find in general textbooks that transcription induces supercoil and so and that's already the papers from late 80s by a group of of Jim Wang and so what white white transcription induces supercoiling because when we have ongoing transcription it turns out that it is not the RNA polymerase which circles around the DNA but it is the DNA which which kind of rotates during during this process okay so if you if you think about so in the kind of very early model RNA polymerase would be like a screw which rotates but now if the screw is stopped then there if the nut is sorry if the nut is stopped then only the screw could could rotate and so now again early papers on on this where in end of 80s it the subject of supercoiling of chromatin is a current active subject and interesting things which were observed is so when when there was this ongoing transcription there is always this accumulation of positive supercoil which which is before the RNA polymerase and accumulation of negative supercoils behind now in the cells positive supercoils are are blocking progression of RNA polymerase the they also block progression of DNA polymerases so cells evolved to remove positive supercoiling quickly so there is a quick removal of positive supercoiling but kind of slow removal of negative supercoiling so finally there is accumulation of negative supercoiling in transcribed region and so we wanted to to model this and this is our early modeling approach when we tried in 2014 so we wanted to model a supercoiling of two topological domains which will be here and well it could be a still a longer chain like this but now to keep the supercoiling here we were somehow we didn't have a good idea how to do it better at the time and so we were closing those loops with kind of accessory chains which were keeping for us this tension in the supercode morning because if I close it I just want to supercode if I close it keeps the tension and of course if it's open it will lose the tension so we were using this little bit like a trick like a trick here we were closing this and then so that was our we were closing it but we were able to introduce a given supercoiling into into the domain we could rotate several times before connecting connecting the ends so that's that's how how we were doing it and then so that's our starting configuration then it got supercoiled here we put certain level of supercoiling which produced more or less free super terms for but now we this kind of shows so that's a chain we are interested in but but we still have this other chain around but we are not putting this data into the statistics we just kind of ignore it it's a little bit like in kabuki theater in in japan there are actors which move the well kind of which are kind of gray but that's normally you should just don't look at the actors but look just at the at the kind of puppets they they they they move and so now but we went this and so okay I still wanted to say how it goes so we have those equilibrated configurations now they flew to a due to thermal motions and we score we look how many times that beat contents that beat how many times when this when this simulation is going on so the for every configuration we and then we could we could enter those data into into those type of much maps so let's say that point here would be a contact between that region and that region so when we play when we created the situation like this it was a weak supercoiling here okay so that we looked at this region here but you might still remember that there was another region here which is which was needed for us to keep but supercoiling and then another situation here and then we could calculate those contact maps and what was well visible that when we put certain level of supercoiling we could reproduce experimental data so the level of supercoiling in eukaryotic cells is low it's difficult to measure and so we had a little bit we were trying to see how much supercoils we should put here and here we put about eight supercoils per a loop of which corresponds to 400 kilobases and with that supercoiling we were able to reproduce experimental data so I'm still sorry for this so the experimental data are provided by the paper Noray et al in in nature where they were measuring how quickly the frequency of contact decreases when you increase the distance between points between a lot of for which you measure the interaction and this you can measure within et al this you can measure between et al and those the experimental data are the points here and continuous lines are our simulation so we were able to fit the data to just by changing supercoiling to fit the experimental experimental data and so now we were able to have a reasonable fit to to again our model is very simple one could do and that's the model we more or less arrived to model says well decondensed chromosome would be composed of those supercoiled loops and so every supercoil loop will be one one tap again so we published this in 2014 and our model is I would say quite frequently mentioned in the topical reviews well we have about 50 quotations for for the that paper but still the the concept of of the tats are supercoiled are not really well accepted because it's not well accepted that chromosomes are really supercoiled and so one needs more data and I think one also needs better simulations than which we had this having those accessory chains was was not really helping so now for the new data there was a paper published in 2014 a Mitsuguchi is the first author among the authors there was a group of Leonid Mirny and what it was a paper on organization of chromosomes in lower eukaryote in espombe and when what what was observed there that there was also formation of tats in in lowering now what the paper said in the text but it never developed this this idea they observed the perfect correlation between tats they have here and between a domains of divergent transcription in in this east so what are those domains of divergent transcription and then I would like to go to this so if you would have a gain it's a it's a my drawing here so you would have here a region of chromatin and then you would have a RNA polymerase going in that direction and then you would have RNA polymerase going in opposing directions mean they would read other different strands of of the DNA and so those are regions of divergent of divergent transcription then the same would be would be here again the transcription will go this way toward the border between between those tats and then again in this so if so here I in the drawing I just have a now RNA polymerase is bound to promoter if transcription would start what would happen they would produce this this supercode because now they what what happens now this is so if one polymerase goes this direction another goes in this direction then they introduce they turn in opposing direction and this introduces supercode so this is kind of a classical system where those early studies of transcription induced supercolding bacteria were were done regions with divergent or or convergent transcription and so so in espombe there was this perfect correlation between those tats and the regions of divergent transcription so we wanted to model it but model a bit better way than we modeled before and so what we so before what I what we modeled when we wanted a supercode DNA we had to use this accessory chain now we thought well can we do it better can we still have an open chain or and what would be if we just put a motor which turns in this direction and what would be if we put a motor which turns in the other direction and so we modeled now the motors which induce rotation so and now again so here would be just to to illustrate how how it goes so so now here it is closed just for the purpose of illustration so there will be one RNA polymerase turning in this direction and another RNA polymerase turning in that direction so that would be a body of this of this domain with divergent transcription and that that would be a border between them and in the border we put a a swivel a free swivel where the stress can be can be relaxed and now with with this setting so we could start our our motors and then that's that's what happens molecules get supercoiled our modern molecule and again rife is the measure of supercoiling we see it kind of grows and then saturates because our motors have a certain strength that like RNA polymerase can exert a torque of up to five picon Newton's time nanometer so we put our strength of two picon Newton's time nanometer and then that's that's what what what we okay so that's that's just for one domain but then we thought what will be if we combine several such domains so so each color here is a different domain and in each one we have those motors at the end because in each again in as from the in each domain there it's a domain with divergent transcription so so that so we have motors we have regions where the stress can be dissipated here and we have also the regions which where segments can pass through each other they are drawn here as a a transpine because top DNA is relaxed chromatin is relaxed by type one topoisomerases which can be kind of compared to free swivel and also by passages of duplex through duplex DNA which we could model in this way that we have a semi transparent zone through which our duplex our chromatin can pass through and so that's that's a image of the snapshot from the simulation so individual domains get supercode now we don't need to to have those accessory chains and so this is kind of we were more satisfied with with this with this simulation now what is the result of the simulation so now again when we I go back to this so again that's one snapshot but this evolves over time there are thermal fluctuations and again we can we can measure what are the con score all the contacts during the the simulation and then that's our simulation results you could have it at higher resolution or lower resolution comparing to to compare better with experimental data and so more or less with the strength of the torque which we apply we our data reproduce reasonable the experimental experimental data and the simulation is kind of a bit a bit nicer because it represents directly how the RNA top how RNA polymerase would introduce supercode into into the now in the original paper by Mitsuguchi at all the authors suggested that this those uh tats in esponda were due to the presence of of cohesin and then cohesin is a is a protein which is important for shaping chromosomes and now they made an experiment they devoided cells of of cohesin and then they didn't have those tats and so that's how they concluded that that's because of the present that cohesin makes this border and that's physical border between regions of chromatin so we tested that model by by simulations and again so we we placed cohesin rings more or less dimensions corresponding to well it's a question whether this ring should be straight or more more collapsed new papers showing that this ring might be more collapsed than this but I think it does it does not change the result of the simulation so when you put just rings but no motors just the rings by themselves cohesin rings there there was essentially no no tats formation if one changed completely the color scale from the normal linear to logarithmic one one could see regions that just in the place where tats was present there was a little bit less contact but there was where the cohesin ring was sitting but there was no tats formation so somehow we rather think that cohesin might be required for targeting top or two two borders of of tats and that might be why if we don't have cohesin we don't see this this nice organization so now a second part it will be shorter than the first one is enhancer promoter interaction and and tats so again developmentally regulated genes require this interaction with between enhancer and promoter and so so here let's say you have an enhancer and normally your target gene will be here so that's that's one tats so that's very important that this enhancer at at some point will contact the target gene and then there would be expression and but it's also very important that it will not contact another promoter which is just not far not much farther in in in next in next tad now think that this is in the chromosome it's not a kind of extended configuration they so the the physical distance between this region and the that gene here might be very similar between this region and the gene gene here so and so it's very important that there is no contact here now there were interesting experiments done where this border was removed and when the border was removed that enhancer contacted this promoter here and this led to pathological situation so that's very important that really that enhancer should only target here and not not here so now i mentioned before that frequency of contacts in the tad is increased but this increase is kind of two to three four contacts within a tad are two times more frequent than contacts between between the tad so this looks like not sufficient so now we wanted to understand the situation better and now we modeled again as supercoil loops where we placed enhancers and and promoters with a certain affinity i will so we put the affinity which is 10 kbt that's the strength of the of the interaction and now what happens in interesting situation so now when we increase the supercoiling so that's the super we see that that fraction of time enhancer contacts promoter increases with with supercoiling and that's somehow very important so that's the the first thing but now let me go to next situation and that that supercoiling increases intra tad preference of enhancer promoter interaction so we modeled following situation we have enhancer which can contact a promoter in the same loop or in the neighboring loop in the neighboring tad that's how we we modeled it and the distance here is the same kind of this genomic distance between enhancer promoter within the same loop and between is is the same now we that's how we modeled it and now we looked what happens when we increase increase supercoil so and now we look at and the intra domain contacts so that's enhancer enhancer with promoter one and interdomain contacts enhances with promoter two when we increase supercoiling intra domain contacts and the time enhancers and promoter stay together increases now the interdomain decreases and the ratio between the two now it starts from the value of about three and then it goes up to the value of about 30 when we do this so so now there was this problem a small difference in contacts between a kind of generic site get generic sites are would be normal beats in chroma but now enhancers and promoters they contact each other and now if there is a if they not only contact each other they have affinity to each other and if they have affinity to each other this enhances this effect there is like additive effect between supercoiling and affinity which makes that finally we have this a difference of the preference for contacts within the same the same that and between that comes to the now the preference comes to the value of 30 so value of 30 is already a safe value it is then it will really be much value of two or three was not sufficient really to to understand this this situation now why are enhancers not placed near the target promoters so this is so again enhancer and promoter can be far away they can be a mega basic part why nature has chosen this this solution why not to place enhancer just near near the promoter okay and so we tried to we tried to understand this little bit so now we have we put different pairs of enhancers those are distal enhancer promoter and proximal enhancer and promoter and if enhancer promoter are close close by they find each other quicker than if they are when they are far away okay and so that's without supercoiling where what we what we see here this is for large separation this is for small separation like 28 kilobases 16 kilobases so always when they are close together they they contact each other quicker they always win whatever the supercoiling is those which were close together they they win but an interesting thing is that when they were far away there was a big difference in the interaction with and without supercoiling so we conclude from from this that large genomic distance between enhancers and their target promoters permit control of the interaction by supercoiling and this gives additional regulatory possibilities and now in bacteria this regulation by supercoiling is well established and very interesting so let's say if you put bacteria from you change to you cut the amount of oxygen there is the supercoiling changes and special genes which are needed for for response to this are regulated by by supercoiling so here by putting far away enhancers and promoters there is this possibility of regulating this interaction by by supercoiling so now still like at question a little bit so we have this developmentally regulated gene and okay now I kind of suggested before that there was supercoiling helps to bring promoters together with enhancer this is needed for the for the transcription so now let's say we have this we start from in an inactive development regulated gene and wherever so we so there is no transcription if there is no transcription there is no supercoiling if there is no supercoiling there is no enhancer promoter interaction and there is no there is no transcription and so how we go out of of of this and that looked kind of difficult to go out of but since 2010 it is known that enhancers are transcribed and that the transcription is non-dependent on interactions with other with other enhancers so they they are regulated by binding of specific transcription factors so in this case what we have we have we have transcription of enhancers the RNA which is produced from enhancers called E RNA this produces supercoiling this makes enhancer promoter interaction and then finally there is an enhancer controlled transcription of developmentally regulated genes which we could again show in this in this scheme so we have inactive that E is enhancer promoters red light they cannot now there is there would be now activation of of that of that gene so there would be some production of some transcription factor which starts the now starts transcription from enhancer so E RNA is produced this introduces supercoiling now we it brings together enhancer promoter are brought together and now the promoter the mRNA of this developmentally regulated gene can be produced and of course this in addition if we would have the some top topoisomerases we could regulate it we could allow a generation of supercoils or not and so now the role of RNA is a puzzle still it is quickly degraded so so somehow but according to our model it's its role is to produce supercoiling and therefore it can be quickly degraded it would it would not not matter so that would be the end of scientific part very short cv okay so I studied biology with specialization microbiology like Alexander said in the University of Warsaw in 1972 1977 that's University of Warsaw again if you visit Warsaw I invite you to see it it's a green oasis in the center of a big city and that's and that's so that's a main library the biology building is little bit on on that side but let's say almost every day I was passing through it it's very lively place you see here it's nine o'clock but I think it was like Easter Sunday when they took this picture so it's that's that's that's why there are no no people okay so biology study then I went to do my phd in the Institute of Biophysics and Biochemistry of Polish Academy of Science so Polish Academy of Science that's a big building here it's a main building of Polish Academy of Science Institute of Biophysics and Biochemistry is maybe not as nice as this do you recognize this statue here in the front it's a Copernicus of course in Polish Academy of Science that would be a Copernicus and I think somehow I was very much influenced by different Copernican view of of so that's now this is an administrative building which is also very important because to go to Switzerland I needed the passport and to have the passport I needed to go to that building and then afterwards I went to Zurich I was at Institute of Cell Biology at Etihad Holmkeberg I spent eight years in the group of telecollar and that's the building that building here Giovanni is your building that's the where so we were at the same time and now afterwards I was at biology building in the group of of Jacques de Beauche and now I'm here at this I would say it's the nicest place from all I've seen and again every time I walk here I just kind of admire this this view it is I feel very much privileged to there and I I'm not so kind of surprised that this is so frequent in so many good paintings like like this one of of Hodler okay so now scientific highlights little bit I I just selected of course the best of of so kind of saying about so 1982 my first nature publication still I would like to say out of 11 and it was accompanied already then by news news and views but those were still a different time biology has grown so much that paper file it was in a way not even full two pages two pages without any supplementary materials and something entire paper then interesting thing 1996 that was already in in in Lausanne I had this rare event two papers in the same issue of nature two papers and it also had the accompanying news and and views and that was on as you said on on knots and then I there was and so there was electrophoretic mobility of DNA knots and then together with Giovanni we later followed up this this subject a bit and that was also geometry and physics of knots with some new concepts which are called ideal knots and then again another cover that was we had a paper in 97 on again on knots and this one even made it to the cover okay so what can be better than than this of course the better can be this promotion which I received and then shortly after promotion shortly after promotion in October 2016 I got this very interesting message from Stockholm that was and then what what they say they asked me to nominate someone for a Nobel Prize and so that's a great honor the snag is you cannot propose yourself okay so so anyhow that that's how I will be watching closely you to whom I should propose okay for for the for the next few few years and then I would like to thank you for first for your attention here and then I would like to thank Alitza my wife and also my very close collaborator we worked for for many many years together we made all this way from University of Warsaw to be I Zurich to to here and Alitza took last year early retirement and she's enjoying more of free time and now I also wanted to thank member current members of my my group Julian Dorier who is not present here today Fabrizio who yes oh very good okay okay so then Fabrizio which who is who is present here Duchan had some unexpected and but he said he's watching this from Bratislava today so so that's and Demos who joined recently was not working on on that project but it will be working on notes that's what then I wanted very much to thank Juli who is our secretary I think she helps so much in work of the of the group and that's that's very important I also wanted to thank special thanks to Nicole because there are so many administrative important things with grants with everything and without her help I would not got some some of my grants that was very essential I'm thanking also Nuria who is not not here today because it was great when to have chance to discuss with Nuria and what was important for me I had several questions and some small problems and also Nuria was very much helping this and again after Nuria left Alexander took care about those small problems and this and Alexander I also thank you very much for this and again it's it's great that there's always a chance to come there's almost like door open and to have such a chance then I would like to thank Giovanni Dietler from a PFL for many years of of collaboration and now we have a active collaboration within within so that's with the group of Vincent Dion and Gustavo and Oscar who are probably present here yes and so again we work together on modeling of chromosomes and and then of course I would like to thank all Seyge and SIP colleagues and I would like to thank you for coming here for your attention and I hope that you you will have time also to join us for the APRO which probably I would be also happy to answer some questions but if there are more questions we could continue or during the during the APRO thank you very much for you so question please wait for the microphone because this is streamed live and so they will be it's a listener need to to get a question from you so maybe I didn't get the point but the one of the last slides you said well if you have no transcription you have no supercalling you'll go away and answer and but there is there is the other problem you have transcription more supercalling more an answer and so it could run away in the other direction too so what about this so that's that problem is kind of regulated by by topoisomer raises in eukaryotic organisms which always have the chance to relax supercalling we don't have topoisomer raises in eukaryotic organisms which can produce supercalling but so the supercalls soon there is excessive supercalling of positive or negative the DNA structure or chromatin structure is changed so much that it is now the topoisomer raises have high affinity to those changes of of chromatin or DNA structure induced by excessive supercalling of positive or negative so there is a kind of a healthy level where they don't intervene and then when it comes about this they so this takes care of this excessive supercalling philly thank you for a beautiful lecture I think it's it makes me dream you know it's very complicated at first and luckily you had this route that was very useful but it makes you think of the complexity of this whole regulation so you have you have supercalling and you have to regulate that at the right time with enhancers with with with with all transacting things proteins and these tasks might be very must be very important because they are kind of translational units that may then interact and and and give signals to other how many genes do you have in one tad so so from the first discovery of tats it looked like there are several genes per tad now the now when there was this size was of about one megabase now somehow with this kind of resolution and then people talk about sub-tats sub-tats because so the and there now we are coming to about 50 000 100 000 and then this is now we are at the level one gene one tad supposedly it's tad now yeah it's a question from a tar yes exactly so thanks a lot for this very interesting lecture um i was wondering whether the model predicts certain periodicity of the frequency of the contact within a single tad because well the the coils impose certain periodicity of contact but that might be obliterated if the tightness of those coils changes as there's more and more torsion so i don't know whether yeah one predicts anything yes so i i think i thank you for for this very very good question so that was a little bit little bit also problem with my drawings i draw them drawings are very stationary if one would observe such a fluctuating molecule what would happen the the regions of context would change this this is a motion this motion is called slivering and or this is kind of conveyor type of motion which is also driven by a thermal motion so finally although if i draw it like this you would think that there was a preferential interaction no they would average out over over time and that's that's what i think would happen i have a question can you tell us a bit more what what are the anchor points or what is the nature what is different there okay so uh what the important thing for the anchor points it was we had it in this early early models we had the anchor points uh now because there was still a little bit of this uh uh let's say uh matrix attachment region i was influenced by this but in this our new simulations we finally now when we have the borders are just the regions where the topological stress or torsional stress is released so in these new simulations of the we don't have any more uh any more attachment points the board what is this to so to have this organization of tats it's sufficient to have a generators of of of supercalling anywhere within a tatt and the sinks where the supercalling can be released between them and that's our those are our current models kustavo um so so so certain tads that are gene rich do they exhibit sort of higher contact frequencies between the sequences of that tad compared to tads that are uh not gene rich well it will not be the answer to exactly to your question but but it is are the tats where the there is more intensive transcription more dense do they show more contacts yes that was shown that the higher the transcription the more contacts that was this correlation not necessarily with the number of genes because of those story with the sub tats and this which finally seemed to separate into into much smaller units and then sort of thinking about the time that it takes for supercoiling to affect the tad formation and structure um what are your thoughts about like in you know in one cell cycle is is it is the time where transcription occurs enough sort of to impact tad formation so well i think this when you generate super supercoil this again RNA polymerase generates about five supercoils per per second and now that would be that's a large now how quickly they will diffuse how how this this entire region will become supercoil this of course is a complex story we could model this in water in water it would be very quick it will be a it will be a fraction of of a of a second in a dense environment like it it might be a slow process so so that's so do if so imagine that it's in the honey then of course it will it will take take long time but i think even in the honey it would be uh you know the time of the order of minutes and then i think we are fine with with this time scales if that's of the order of minutes the last one so you hinted even though you didn't say this explicitly that this urna produced through a transcription of the enhancer is actually useless in itself and i was wondering because that there could be other places nonlessly enhancers themselves were transcription my function just to essentially induce this supercoiling a regular transcription of other um so what does it do you think that says something about mod encode project and this whole idea that people some people advocating that actually almost everything gets transcribed therefore all this is is important and all this RNA that you detect is actually important do you think that it sort of says maybe many of those RNAs actually a byproduct of a process that is really physical to introduce the coiling so maybe that it will be not very direct answer to your but those studies which a kind of established a transcription from a from enhancers they were kind of a very interesting they were two phenomena there was the first so so let's say with whatever with the estrogen you you start a transcription of several several regions the first what you see which is which is coming there is this wave of transcription of enhancer that that comes first and then comes the wave of transcription from the target genes so so this kind of tells you that that was not a not a noise that you see this way this wave is going down and then there is this another so that that is not the noise now the question of of the would noise do this maybe yes but it's much better to have a kind of a safe mechanism where you would not depend on that there would come at some point the transcription which will introduce supercon you need this super