 I would like to continue where I finished yesterday. And I told you that one of the processes which I would like to speak about today is this spreading process over here, which I sort of explained yesterday, where the blaster down, which is sitting on top of this very big yoxa, starts to spread around the yoxa. This spreading process is mediated by two different cellular rearrangements. One is that an epithelial cell layer on the surface of the blaster down, which is called the enveloping cell layer, undergoes active spreading. So it's not increasing its volume, but it spreads. It becomes very thin and large. And the other process, which I second half of the talk will talk about, is how deep cells are rearranging and how this rearrangement of deep cells is orchestrated with the spreading of the epithelial cells in this process of tissue spreading at the onset of zebrafish gas production. OK, so the first part of my talk will be about the force-generating mechanisms by which epithelial cell layer spreading is being triggered during this process of epithelial here. And I just put in the schematic diagram. There have been different models how epithelial cell layer spreading is being triggered. And in the schematic here, what you can see is a fish embryo. Again, at the onset of gastrulation, the EVL, which is the scrimmage epithelial cell layer on the surface of the embryo, has made it halfway over the york cell here. And what had been noted a number of years ago is that there's significant accumulation of actinamizing right at the margin where these EVL cells are making contact to the underlying york cell. And this accumulation of actinamizing appeared to be in a ring-like fashion, in a sort of a cable-like fashion. And what had been sort of speculated by a number of studies, including our own lab here, is that this ring might be a constricting ring once it moves over the equator of the york cell. It just needs to constrict. And then it covers through the curvature of the york cell. And it can pull on the margin of the EVL and can pull the EVL down that. So there was sort of an attractive hypothesis which we could experimentally address. And what we wanted to is to see if that is the only mechanism by which, if it is a mechanism at all, and if it's the only mechanism by which the EVL makes it over the york cell. One sort of limitation of this model of having a constricting actinizing ring is that force generation and pulling on the margin of the EVL can only happen once the ring has moved over the equator. Because if it would constrict before the equator, it would just go in the opposite direction. So we had to assume that the EVL spreads by an unknown mechanism up to the point where it sort of crosses the equator. And then this actinizing ring would get into action constricting and pulling on the margin of the EVL and pull it on that. Okay. So there was sort of the starting condition. And the first thing we did is we wanted to visualize actin and myosin in real time. In the embryos, we generated two transgenic lines. One transgenic line which visualizes myosin. And the other transgenic line which visualizes actin. And consistent with previous observations on the accumulation of actin and myosin, this process what we find is that indeed there's an accumulation of myosin and an accumulation of actin in this region of the York cell over here, which is overlapping with this assumed actinizing ring which might pull on the margin of the EVL, of course. So these observations are consistent with previous observations, that there's indeed actin and myosin accumulating at the position where the EVL makes contact to the York cell and that it could in principle act as an actinizing constricting ring which pulls the EVL downwards. Okay. Then a PhD student in the lab, Martin Bairn, he together with Stefan Guild actually addressed it this time. He thought that we're gonna use a UV laser cutter to see how forces are being generated within this actinizing ring, to see if it is exclusively acting as a constricting actinizing cable. Okay. So the experiment which Martin did here is, you're just looking over at this little window here looking at actinizing accumulation in this ring, that question here. He did cuts in different orientations. What he assumed is if he is doing a cut in a particular orientation to the margin of the EVL, he will reveal a lot of recoil and this recoil will be proportional to the tension along this circumference within this actinizing band. And he did also control cuts over here which he's cutting parallel to the margin of the EVL and he was expecting if it would be a simple contractile ring, he would reveal much less tension along this orientation than along this rotation here. So he's monitoring recoil and we are deducing from the recoil than the tension. Okay, this is what is happening. If you cut perpendicular to the margin of the EVL, there's indeed a significant recoil here in this orientation indicating that there's tension around the circumference which again would be consistent with the idea that this is a constricting actinizing ring. But what was more surprising is when we did these control cuts over here that there was actually a significant amount of tension also perpendicular to the along the width of this actinizing band here. So that indicated that perhaps it is not exclusively a simple constricting actinizing ring but there might be something more about it which might generate forces. Ah, because the time scale of this experiment is in seconds, I mean, and the ring moves in the time scale of minutes to hours, so the total movement from its formation up to the point where it closes is approximately seven hours, six to seven hours. And here we're looking in the time scale of a few seconds. So it is actually, I mean, if you look down here, there's a little bit of movement but you know, there's something very hard to... Okay, this is a quantification of the recoil velocity. And at different stages of EVL or PIPL, so at different stages of the ring position, basically it gives you an estimate where this actinizing ring is at different stages of PIPL. In this case, the actinizing ring is approximately at the equator. Here it has crossed the equator and then it is halfway down and it nearly reaches the vaginal pole down here. And if you look now at the tension, these are the perpendicular cuts where you reduce the differential tension and these are the horizontal cuts where you reveal tension along the width of the actinizing band. If you just look at this stage here before the ring has crossed the equator, it was quite surprising to see that circumferential tension and the tension along the width of the actinizing band were actually not far apart from each other, indicating that this is not just a constricting ring. Interestingly, if you go through different stages of castellation, the tension along the width of the actinizing band remains relatively constant. It doesn't change too much while the circumferential tension is going up. Okay. So what we wanted to know is where this tension is coming from and what it really means for the process for forced generation and pulling on the margin of the gap. So the next thing we did is we looked at the dynamic distribution of actinizing during the formation of the ring and during the ring constriction. And what we do is again, we're looking at this little window over here at a stage where the ring has crossed the equator and we're looking here at myosin-2 dynamic redistribution, f-actin, and then I merge between myosin-2 and f-actin. And in this movie, what you will see is that there's actually flow and you see myosin flowing in the opposite direction to the movement of the EVL down, that's right. So you have a redrawed flow of myosin-2. Yes, in parallel to myosin-2 flow, you have a flow of the f-actin and you can very nicely see that these things are flowing in the approximately the same velocity in the opposite direction to the movement of the EVL down, that's right. So if a redrawed flow of myosin-2 and f-actin in the region where this actinizing ring becomes apparent, right? And I think this is just a quantification of these flows. Positive velocities are velocities which are going down towards the vegetable pole. So the EVL march moves downwards. This would be the connection between EVL and the York cell and this is actinizing within the York cell which are moving in the opposite direction to the march and of the EVL. So you have a redrawed flow of actinizing. So we are teamed up here again. Stefan was already part of the experimental analysis but he took Guillaume's, he's a theorist on board and together with them we developed two different potential scenarios but which forces could be generated by this actinizing band including these redrawed flows now of actinizing. So the first sort of motor activity we assumed is the one which has been assumed before already in several studies and that is essentially what we called a cable constriction motor which means that this actinizing band predominantly constricts around the circumference and once it has moved over the equator it generates a pulling force on the march and it puts it down there. The other motor activity at this point is still a speculation is a flow friction motor but we assumed here is that there's constriction, there's contraction of the actinizing band not only around the circumference but also along the width of the actinizing band. Then assuming that this actinizing band would be coupled here to these junctions which connect the EDL to the underlying York cell and it would be more free to constrict on the other end then you would get a net flow, a retrogate flow of actinizing. You contract the network, you couple it on one end and you only loosely couple it on the other end then you get a flow which is back that's oriented. But assuming that this flow would not be completely free otherwise it would collapse but it would be resisted by some friction then you would generate a net force which is pulling in the opposite action and pulls the march and downwards. So we call it a flow friction motor where you have the contraction of the actinizing band along its width which in case it is resisted by friction to any adjacent structure within the York cell, this could be York granules, micro-tubules, anything you can imagine which is in there it would cause pulling force downwards. Okay, at this point was obviously a pure speculation and we were looking for experiments how we can distinguish between these two motor activities. And the experiment which we came up which was originally actually suggested by Peter Schwilde and then we sort of took it on is the idea that we take the embryo and we put it in an agro-scapillary and deform the embryo from its original spherical geometry into something which looks like a cylinder. When we have a cylinder then we know that this cable constriction motor activity would be rendered inactive because there's no curvature of the embryo and it cannot pull by a cable constriction activity anymore. Okay, so the experiment is this, you take the embryo, you pull it in and then you're asking would the EVL in such a configuration in such a geometry still move downwards? Is there any sort of pulling activity which might be due to a motor activity which is not the cable constriction motor? Let's put it in a very sort of, you should say. That's what the embryo looks like if you're putting it in there, fine. You can take them out and they develop into normal embryos. Excuse me, this is now the embryo in a capillary, a time magnification. You can see the EVL up here, the york down here. And you can see a bit of actin accumulating here in this margin which is not very obvious yet. Now, the question is what would happen? And the original assumption was that it would not move but to our surprise, what was happening is actually that the EVL moved downwards approximately at the same velocity as it could move downwards in an unconstrained embryo here. And you see nice accumulation of actinimizing and you even have a regular flow of actinimizing. Indicating that the cable constriction motor activity which had originally been assumed to be the critical motor activity to pull on the margin of the EVL is actually dispensable and you can even generate forces which are sufficient to pull the EVL downwards in cases where this motor activity is not, is rather inactive. If that is the flow friction motor activity is still an open passion but we assume since we see these flows that the flow friction motor might actually be doing the job at this case. That's what the embryos look like at the end of constellation. And again, you can take them out and sort of find. Now what we concluded on this and I meant very quickly to it because this is the old stuff that she published a few years ago is that the EVL spreading here is driven by two motor, presumably two motor activities. One motor activity is the cable constriction motor activity which you can which is to a degree dispensable. It doesn't mean that it doesn't work in a normal geometry but if you render this inactive there's another motor activity which might likely be a flow friction motor activity which pulls on the margin of the flow friction motor activity. Okay. So I'm not saying that the cable constriction motor activity is not active, it will contribute but its contribution can be substituted by another motor activity once it's being rendered inactive. Okay. Now what I would like to do then is tell you a bit more about the EVL, how it spreads over the oxel and particularly what I would like to speak about is cell divisions within the EVL and how possibly cell divisions within the EVL contribute to spreading of the EVL over the oxel. Okay. So what I show you here is the EVL again, you would assume that there's active inamizing, accumulating in this marginal region already and the EVL starts to spread and what I have visualized in this movie, I, Pedro, who was doing these experiments here, are all cell divisions happening within the EVL. And you know that there are number of EVL cells undergoing divisions and they are dividing at pretty much random places within the EVL. There's a slide-by initially to the end of the pole but eventually cells dividing at every place. Cell divisions becoming more rare during the course of the PIVLE and most of the cell divisions are happening before the EVL has actually reached the equator. Once it crosses the equator, the frequency of divisions are going up. But what we noted by these divisions is not only their frequency at certain stages of the EVL at PIVLE and their spatial distribution within the EVL but they also noted that these divisions seem to be preferentially oriented along the animal-vegetable axis. And just stop the movie now and it's perhaps not the most representative picture but you can see these two daughter cells being aligned along the animal-vegetable axis. These two have a slight tendency, this perhaps less than these two not but overall if you look at the orientation of cell divisions there is a preferential orientation along the animal-vegetable axis of the embryo. So the daughter cells are being placed along the animal-vegetable axis which is the movement axis of the EVL darkness. So we asked two questions based on these observations. First of all, how is EVL cell division orientation being controlled within the EVL? And the second one which is probably the most important one is what is the function of oriented cell divisions in EVL tissues during a PIVLE if there's any function at all, right? I should say one thing about these divisions within the EVL, this is not growth. It doesn't, you know, the volume of the total volume of the EVL is not changing during the process of the EVL and PIVLE movements. So once the EVL is being specified and there are no cells joining the EVL anymore, any division of an EVL cell is just subdividing the volume in two parts, right? There's no volume increase, no growth involved. So the only thing we are seeing here is basically a redistribution of the original volume of the dividing cells in a specific spatial configuration, okay? So no growth. Okay, so the first question is how is EVL cell division orientation controlled? And what we have hypothesized here is based on studies which have been pioneered by Matthew Piel in the city of Korea in France, Paris. And what Matthew did is he did an experiment where he took a single cell and then he stretched the cell along a main axis of tension and then he asked how would the methotic spindle be oriented in response to tension applied to these cells? So he did it in culture cells and single culture cells and what he found is that the main axis of tension determines the orientation of the methotic spindle. Now, the most simple explanation would be in such cases that the methotic spindle that's known for more than 100 years tries to find the longest axis of these cells, right? And once you are applying stress on a cell, then it will elongate along the main axis of tension and then the methotic spindle falls into the main axis of tension. What he then did in this study which was published a few years ago is he originally grow the cell on pattern substrates so that they have a morphology which is actually slightly elongated and then he pulls them into a more round morphology by an ectopic tension axis. So the morphology where ectopic tension is being applied is actually not elongated along the main tension axis and then he was asking would the methotic spindle still find tension irrespective of the morphology of any sort of provincial elongation along the main tension axis and he still finds evidence that the methotic spindle goes along the main tension axis. Now what he attributes that to is that the cell undergoes a slight metotic grounding and metotic grounding leads to retraction fibers and the retraction fiber and acting accumulation close to the retraction fibers is usually a long-drain tension axis. So tension triggers active accumulation of the atomizing cortex along the main tension axis and these accumulation of active leads to a preferential anchoring of the methotic spindle along the main tension axis. So somehow tension modulates the atomizing cortex in a polarized manner and this leads to anchoring of the spindle along the main tension axis. So it's a mechanism which is independent of shape. That's what he proposed. OK. Based on his experiments, we thought anisotropic tissue tangent within the EVL might indeed be responsible for orienting the methotic spindle and leads to this preferential orientation of EVL subdivisions within the EVL. Now, the experiments we did here are very similar to the experiments I showed you before. Now, we're just looking at the EVL. We're not looking at the York cell down here. And we are doing two cuts and two orientations. One cut is parallel to the margin of the EVL, which would reveal tension along the animal vaginal axis of the EVL. And we do cuts which are perpendicular to the margin of the EVL, which should reveal circumferential tension within the EVL. We did these cuts at different stages of the EVL, of people and at different positions within the EVL. And in a very short, just summarize that the stage, I will show you dependency on stage. There's no dependency on the position other if, of course, if you go to the animal pole, you don't have any. The orientation doesn't count anymore. But it doesn't matter where you do these cuts here, here, or here, you would get approximately the same ratio of tension distribution. OK. Now, that's what he got is when he did these cuts. This is the quantification over here. Different stages of EVL, of people and movements. And the perpendicular cuts are the ones which are revealing circumferential tension. And the parallel cuts are the ones which are revealing tension along the animal vaginal axis. If you compare now the red to the green bars here, you can see that from 50% down to 80%, there is a tension on esotropy where the main axis of tension is indeed along the animal vaginal axis, revealed by the green cuts here. And the minor axis of tension is around the circumference. So once the EVL has moved over the equator, there is another topic tension building up within the EVL, which could be, in principle, consistent with the assumption that the orientation of the martyric spindle and the placement of the daughter cells depends on the main axis. It depends on tension on esotropy within the EVL. Excuse me. That's obviously a hypothesis which you could experimentally address functionally in the dependency. Now, the experiment which Pedro came up, which is a very simple experiment to see if tension within the EVL can indeed orient the martyric spindle is an experiment where he looks at a cell within the EVL at 60%, 70% epibly. And this is now an EVL cell, which is about to divide. And what he has marked now in white is the orientation of the martyric spindle, which is already visible in this binding cell. So the martyric spindle has found its position where it wants to be. And now what he's doing is he's inducing in a topic tension axis within the EVL by doing a little trick. He's using a laser and he's ablating a few cells up here and a few cells down here. Once he's ablating these cells, the cells are being extruded. You get wound healing. What you have is a lot of stretching between these two points of ablation which lead to tension building up between these two points of ablation. This is a very strong topic tension axis within the EVL. And he's positioning now these ablations in a way that the tension axis would be perpendicular to the original position of the martyric spindle. And he wants to see if he can reposition the martyric spindle now to this new main tension axis within the EVL he is inducing. To see if anisotopic tension can reorient the martyric spindle within the EVL, OK? So now that's the experiment. He's ablating a cell over here, a cell down here. Can very nicely see that the martyric spindle now finds the new tension axis which he has induced between the two points of ablation, OK? So that's direct proof that anisotopic tension influences the positioning of the martyric spindle. That's, I should say, a few more words about the mechanism because I don't have slides in here. But one thing you can already see is if you look at this cell, that it is elongating, right? And you could assume that one of the mechanisms and that something we are certainly supporting is that the main tension axis leads to a preferential elongation of EVL cells. And that this elongation leads them to a preferential positioning of the martyric spindle, right? So that's a likely mechanism. You're not excluding it. We also wanted to know if perhaps the accumulation or the activity of actin and myosin, and potentially the asymmetric accumulation of actin and myosin, consistent with the idea from Marty Pl, in response to tension distribution within the EVL, might have an influence on the martyric spindle positioning. So the experiment, which I unfortunately don't have the slides in here, that the inhibited myosin to activity partially, inhibited myosin to activity within the EVL, we were asking would, in such an experiment, the martyric spindle still find the main axis of tension, which we have induced the dependency of myosin to, right? And what we find is evidence that in cases where myosin to activity is partially inhibited, the martyric spindle would not find very easily the longest axis of these cells anymore. Indicating that myosin to activity is required. You're not saying it's asymmetric distribution of myosin to activity, but myosin to, presumably, actin-mising contraction, is required for correct positioning of the martyric spindle in response to tension. What we don't have evidence yet for is whether actin or myosin would now preferentially accumulate at these poles of the cells, and if that would lead to a professional anchoring of the martyric spindle. That's something we don't know. We have not found any evidence for that yet. The only thing which we found is that that's quite interesting. If you look at the apex of the surface of these cells and you look at myosin, myosin is present in myosin mini filaments, and the orientation of myosin mini filaments is along the main axis of tension. So if you look at these mini filaments, they align along the main axis of tension. If there is any influence on the positioning of the martyric spindle, we don't know. Yeah. I come to that. I have one experiment where we quantify that. Yeah. Because I mean, that would be an important experiment if you indeed to actually induce an anisotopic tension for this experiment. We assume, but is it actually in the range of what the embryo shows or not? I'll show you, I'll show you evidence for that. Okay, this is just a quantification of the experiments I showed you here that in a control experiment, the martyric spindle is always where it was before at 90 degrees to the atopic tension axis. And once you are inducing an atopic tension axis, you can be in the martyric spindle. Now, but in just a summary up to here, what we propose here is that there's another atopic tension building up within the EVL during its movement downwards, leading to a main tension axis along the animal-vegetable axis with anisotopic tension within the EVL leads to a stereotypical cell division orientation. Now the, did I, let me see if that's still in there. Yeah, it should be in there. Okay, so what do you, yeah, yeah, yeah. No, you don't. So what you might initially release tension because you get a little opening, but then what you have is a wound healing. You extrude the cells which you have ablated, right? So you reduce area here and you reduce area here and then you stretch the tissue in between, right? Initially, directly after the cut, you might release tension, but then cells are being extruded. And then you heal that you have a little wound healing ring and then you, you stretch the tissue. Yeah, so it's the cylinder, the embryos, and then spindle orientation. Yes. Yeah, we have done experiments, perhaps I can come back to that at the end of the talk because I show you, well, I'm not sure if I show you, but you know, perhaps I go through the functional evidence and then I come back to your question because I'm, that's a relevant one. Yeah, I mean, I can tell you what we have done along that, but before I have to tell you what we did on the function. Okay, so what's the function of tension oriented cell division and the media tissue spreading during your presentation? And what we assumed is that tension oriented cell divisions might reduce tension on a zotopee within the tissue and by reducing tension on a zotopee they might be facilitate tissue spreading. So the idea would be that what you have is you have cells keeping and then you're positioning the two daughter cells along the main axis of tension which would be equivalent to an elongation of the cell along the main axis of tension, right? Because the volume is not changing but you're positioning the two daughters as a specific spatial arrangement which would be equivalent to an elongation. So you preferentially elongate cells along the main axis of tension which should lead to a tension release along the main axis of tension and which should then reduce tension on a zotopee within the tissue and that potentially could facilitate EVL effectively spreading. That was the idea we had on a possible function media spreading. Now the experiment in that comes back to one question I forgot some of you. Do we actually have tension along this axis here? So what we did is we uplaid a few cells up here cells down here and then we're doing a cut. The reveal tension which is being built up between these two points. We do it in two different conditions. In one condition where there's no cell dividing in between and in another condition where there's one cell undergoing an oriented division, right? The question is would tension be reduced in response to having an oriented cell division between these two parts? Again, this is not growth, it's just an oriented cell division. The volume is not changing. So would the tension recorded in the absence of a cell division the same as the tension present in a cell division? It's a very simple experiment. You can do that and I mean you see recoil here and you see the recoil here, it doesn't tell you too much, but excuse me the recoil velocity. If you compare now the recoil velocity in cases where you have no division to cases where you have one cell which is doing before the experiment actually is done has undergoing a division, then you see the tension is being reduced in the presence of a division compared to cases where there's no division. Indicating that this assumed function of the oriented division reducing tension as an anisotropy is actually a realistic one, at least at the time scale of this experiment here. I should go over that and then the sort of functional experiment we did here is doing two functional experiments. We thought we can either inhibit cell divisions completely or we can randomize cell division orientation. In both of these cases we would assume that tension anisotropy is not being reduced as efficiently as in vital embryo and potentially EVL epibely movements are being impaired. I mean that was the assumption we had originally and these are the experiments that you can already see in cases where the inhibits cell divisions or in cases where the randomized cell division orientation by injecting an antibody against dynein which inhibits dynein function and dynein is required for coupling the methotic spindle to the optimizing cortex. In these cases we know in this case that the very efficiently inhibits cell divisions within the EVL and in this case we randomized cell division orientation in both of these cases the take home message is hardly anything happens. So the EVL moves down at the same velocity and at the same efficiency roughly as you would see in a vital embryo. So we thought that sort of the end of the project which just tells us that oriented cell divisions do happen within the EVL they are being induced by anisotropic tension within the EVL but apparently they are dispensable. You can get rid of them and the EVL is different. Okay, until the point where Pedro was looking at movies of the EVL in cases where he inhibited cell divisions or in cases where he randomized cell division orientation. He made a very interesting observation which he didn't anticipate before and just look I think between these two cells here. Now this is an embryo in which he has inhibited cell divisions. What he noted is that instead of having cell divisions what you get is cell fusions within the EVL. So you have two cells which are fusing within the plane of the EVL. That's something you would see in a vital embryo very rarely to not at all actually. So you get very infrequently you get fusion of cells. Now in cases when he's inhibiting cell divisions or when he's randomizing cell division orientation he gets a very large increase in the frequency of the cell fusion events within the EVL. Now the question is what he speculated at this stage is what these fusions might do. There might actually be a compensatory mechanism which is being activated in an epithelium in which oriented cell divisions are being impaired. So potentially what he was speculating is that you get an oriented cell division, oriented release of tension during the fusion process. The idea is if you look at these two cells which are about diffuses that once these cells are fusing they're elongating the common boundary between these two cells very rapidly and eventually this boundary is collapsing and the cells are fusing. So what you do is you get release of tension along this orientation here which is perpendicular to the orientation of fusion. So there was this idea and you want to see that is actually a plausible idea. So what he did is he looked at the dynamics of junction boundary extension between two cells which are about to fuse versus two cells randomly two cells which are not fusing, right? Just to see if the dynamics of junction shrinkage and extension are different between these two cells. And he finds that the junction growth rate is significantly higher in cases where cells are fusing versus cases where cells are fusing. He's done many cases and we compare that. So what he came up and what I still think is a plausible assumption is that you have two mechanisms within the MBA which could potentially reduce tension on esotropy and could facilitate EDR spreading. One mechanism would be that EDR cells are undergoing divisions and that the two daughter cells are being placed along the main axis of tension which is equivalent to an elongation of the mother cell here and that would release tension and would release tension on esotropy and facilitate spreading. And if this mechanism is defective you might activate a second mechanism which is a fusion of cells and in the fusion event again you have two cells which are fusing along this orientation and release tension along this orientation by ending up as a very simplistic diagram of course but it would intuitively tell you what the underlying model is. Okay, so just to summarize up to here what we are speculating is that they have optimized infant medial pudding forces which pull on the march and put it downwards. This leads to a buildup of anisotropic tissue tension within the EDR where the main tension axis is along the animal-vegetal axis and the minor axis along the circumference. In response to this anisotropic tension cells are preferentially elongating along the main axis of tension but there might also be an optimized independent process which is independent of cell shape which leads to preferential positioning of the daughter cells along the main axis of tension. So you have a stereotypical cell division orientation, yes? Say it again. So if the cells are fusing would they be white again? Would they be white again? Ah, would they be white again? That's a bit, well, I mean, in cases where we have suppressed cell division they would not divide again. In cases where we have randomized cell division orientation the fused cells, I think in most cases, well in all cases we have seen actually not dividing but the problem is that cell fusions that the frequency of divisions goes down over the cause of the pivley and they are only very, very infrequently dividing at the end of the constellation and once we have seen a fusing cells we already waste the constellation so it's hard to find actually cell divisions still happening within the EVL to test this. You know, if you have a bi-nucleated cell and the question would be would that be able to divide and we don't have a good answer for that. The cells are still functional. They are staying on the surface of the embryo and the good thing about the EVL is the embryo doesn't need the EVL at data stages. It's just shedding it off and gets rid of it. So it's not an indispensable structure for long in development. So you're asking if more than two cells are fusing together? Yeah, so that comes back to the experiment where we did the cylindrical embryos but we did is actually try to do cylindrical embryos to at least transiently build up tension by changing the surface to volume ratio in these embryos and then to see if you can actually induce fusions of cells. We have no way to see if we are changing the unisotropy of tension within the EVL because once we are putting them into little agarose capillaries, we can't really cut CV laser cuts so we cannot reveal tension within them. But what we observed is when we are taking these cells out again that we are inducing a massive amount of fusion of cells, even in wild type embryos in which cell visions are. So if we transiently presumably increase in plane tension within the EVL, we can trigger fusion of cells and also multiple cells together. So we have very large patches of EVL which contain multiple in such cases. So the downside of this experiment that we can actually not look at tension within the EVL so we assume that there's more tension. Okay, so importantly in this model is that there's an equilibrium between another topic tension instead of the cell division orientation which keep an equilibrium where there's a certain degree of tension on the zotopee, keeping the divisions oriented along the main axis of tension and once you are inactivate this mechanism you might activate another mechanism, a compensatory mechanism, which might be EVL cell fusion. There should say one more word about EVL cell fusion. You might have noted that the number of fusions is much lower than the number of divisions which already indicate that fusion cannot be the only compensatory mechanism because you would assume that at least the numbers would roughly match between these two events but they are not. So the fusions are much more infrequent than divisions indicating that fusion is only one potential compensatory mechanism. Okay. I think that's for the anti-division if you have more combination. Ah, okay. See. I think that's the answer. I mean, I think there's good evidence from particularly from Trussoffler that the tension, in-plane tension is not only determining the orientation of divisions but also the frequency of divisions. If you have very dense tissues and you release tension within the tissue then you activate I think hypersickling and you know, you get tasks out of the nucleus, you have tasks out of the nucleus and you don't get divisions. There's a, the activating of the cell division program is tanging dependent via certain biochemical signage pathways like the hypersickling pathway. In our case, we know that the hypersickling pathways actually only these cells and the EVL cells. We know that tasks and gap go in a tension dependent manner into the nucleus with mechanosensing and the frequency of cell divisions within the EVL is not tension dependent. The frequency seems to be a developmental program which runs in these cells and they have loads of divisions at early stages of EVL differentiation and very little divisions at later stages. No tension dependency in this case. There's modern tension dependency which is quite interesting. I didn't put these slides in here. As I indicated, YAP tasks, you probably have heard about it. These are transcription activators, co-activators. And what had been shown before is that tension can regulate the nuclear localization of these transcriptional activators, YAP and tasks. What we noted is that when tension builds up within the EVL, tasks, one of these two activators goes in a tension dependent manner into the nucleus and once it is in the nucleus, it activates the differentiation of these cells and the ability of these cells to undergo squamous spreading. In mutants for tasks, in mechanosingotic mutants for tasks, EVL specification and EVL epiblity is strongly affected indicating that there is a mechanosensitive process going on in these cells which in our case is not the radar cell divisions but cell type specification and the ability of cells to undergo spreading. Okay. So time-wise, how much time do I have? Probably not. So I can tell you a bit about the interaction between EVL spreading which would be on the outside here and what the deep cells are doing because what we did is now we exclusively looked at the surface of the embryo but obviously the surface is not working independently of what happens below, right? And the really interesting question was how would these two processes rearrangement of deep cells and spreading of surface cells be coordinated in space and time during the spreading process? Okay, so I just show you what we did is here we looked at a very early stage of EVL spreading and deep cell rearrangement, a process which is called zebrafish doming and you will see why it is called doming. Now that's the embryo and you will see that it appears as if the yolk moves into the blastoderm but in fact what is happening in the blastoderm starts to spread around the yolk, right? So this process, it's a very simple process and you might think you can understand that straight away but what you have is what this process involves is you extend the surface here and you have a lot of rearrangement of deep cells in this process of doming, right? So, okay, so doming in zebrafish again is a process of morphogenetic process which has been looked at for many years already and two different potentially force-generated mechanisms have been implicated in doming. One is the one which I indicated before is radial intercalation of deep cells and here I'm saying active radial intercalation of deep cells meaning that cells are actively moving along the radial axis undergoing intercalation along the radial axis thereby thinning the axis along its radial extent and expanding the tissue along its remaining two axis. Okay, so the driving force being cells actively moving along the radial axis undergoing radial intercalation and then leading to a deformation of the tissue thinning of the tissue along its radial axis and then that has been proposed with a number of publications that just put the most reason in here. There was another study which is now more than 10 years ago and it was not really noted much in the community is that they suggested that what you're seeing here is actually the result of the York cell actively pushing into the blaster dome. So they are speculating that the tissue which is spreading is passively deformed by a pushing force which comes from the York cell and pushes into the blaster dome. And we found those are these ideas quite attractive and that's something we can actually experimentally address. So again in this case we teamed up this Guillaume to first build a very simple mechanical model of how spreading of the tissue could be achieved. And again it's a model which is based on the balance of interfacial tangents. Where we assume that the blaster dome has a surface tangent, TB, the York cell has a surface tangent which is TY and then you have an interfacial tangent deep in the blaster dome and the York cell which we call the TBYI and the blaster dome to the York cell interface, okay? And now what we postulate is that the specific geometry of the embryo during the dorming process is given by the distribution of interfacial tangent at each stage during the dorming process. We're looking at steady state where the change would not, where the embryo would not undergo any shape change and we assume that the balance of these interfacial tangents would give you a specific geometry. Now in this phase diagram over here what we have plotted is the ratio of surface tangent of the blaster dome to surface tangent of the York cell on this axis here and the ratio of interfacial tangents in between the blaster dome and the York cell to the surface tangent of the York cell along this axis. If you look at that you can get embryos which look a bit like a doming embryo in this region down here, right? Everything up here is basically unrealistic because these embryos are not undergoing doming but in this case you achieve a geometry where embryos could in principle look like an embryo at the onset of doming. Now if you're in this region you have to assume that the blaster dome tangent is smaller than the tangent of the York cell and that this interfacial tangent here between the blaster dome and the York cell is much smaller than the surface tangent of the York cell, right? So we wanted to see if that's a prediction from this very simple interfacial tangent driven model and we wanted to see if you can experimentally achieve that. Now the experiment we did here is an experiment which Guillaume suggested and we thought it's gonna be very simple but it took us more than a year to do. It's taken an embryo between two plates and then deforming it along its animal-vaginal extent. Then looking at the deformation of the blaster dome at the upper plate and the deformation of the York cell at the lower plate as a function of force which is being applied onto the embryo, right? And the contact angle and the area of compression here the contact area and the contact area down here and the contact angle to tell you something about the surface tangents at the blaster dome of the surface tangent of the York cell. Here in the process of talking you can do that, right? But you have to incorporate here in this model and that's something we realized only over time is gravitational force that this the York cell down here already deforms without any compression. Okay, now that's the experiment we are putting the embryo together and then we're waiting and we're calling force till it goes into equilibrium. And once this is in equilibrium which takes quite a while then we can go and measure the contact area up here the contact area down here and the contact angle and to write then the respective surface tension. Now we did these experiments in quite a few embryos and incorporating then eventually the gravitational forces led us to results which indicate indeed the surface tangent of the blaster dome during the process of doming becomes smaller than the surface tangent of the York cell. This is now not what has been experimentally determined but this is the prediction from the geometry of the actual embryo on the distribution of the interfacial tangents, right? So this is a prediction based on morphology of embryos but if you're measuring you're getting to a value which is very close to one and smaller than one as well, okay? So our experimental observations match this point which would predict from the geometry in the framework of this theory here. So we get a quite close match between experimental observations and the derivative predictions. Okay, so what we wanted to know is then how is the blaster dome reducing its surface tension and is that the only mechanism of its doming? Triggered. So what we speculated there are two potential for generating process. One would be a reduction in blaster dome surface tension by EVL expansion. The EVL could actively expand by actively expanding it could reduce the surface tension of the blaster dome but we also wanted to see if the generation of radius stress within the blaster dome by active radius amplification could contribute to this process. We didn't want to exclude the original assumption that there's actually stress generation within the blaster dome which leads to spreading of the blaster dome. So this would be the mechanism which we sort of addressed in this very simplistic model of interfacial tangents and this would be an alternative mechanism which has been proposed in the previous studies on this topic here. Now what do we have to put in in this dynamic model of doming then? We have to put in the surface tension of the blaster dome and the oxen. That's something we can measure which I showed you in the previous experiment. We have to put in the discosities of the blaster dome and the oxen. That's a bit harder to measure and what we did is we essentially took to measure the discosities, we took pieces of the blaster dome, we took the blaster dome, we cut it into two pieces and then let these two pieces of the blaster dome fuse over time. But the relaxation and the fusion time we can deduce the viscosity of these tissues. And we did the same for the blaster dome as we did for the oxen to arrive at the discosities of these two tissues here which we can put into the model. Then what we ideally would like to know is the interfacial tension between the blaster dome and the oxen. This interfacial tension of the blaster dome to the oxen interface. That's something we cannot experimentally obtain and we have to fit this value to the morphology at the end there. And the radius stress within the blaster dome again is something we assume, but we have to fit because we cannot measure it currently directly. Okay, I'll just show you this one experiment here where we there we have two ways of measuring the viscosity of tissues. In this case, we take a blaster dome tissue, we compress it, then we relieve the upper plate and we look how quick the tissue regains its very good geometry relaxation. And here we are taking two tissues of blaster dome, we are fusing them at the middle and then we ask how quick they would round up again. Those cases, we would get a readout for viscosity. We also did one experiment to learn something very roughly about the interfacial tension between the blaster dome and the oxen. What we thought is that perhaps the amount of actinamizing localizing at this interface might give us some idea about at least the extent by which tension would change at this interface. And that's very, very indirect, but what we see is that there's very little change in the localization of actinamizing at the interface between the blaster dome and the oxen, suggesting that this tension, at least, that something we can sort of propose is not changing during the course of the typically. How big it is we cannot say, particularly we cannot say that it is much smaller than the interfacial tension of the oxen. You might remember from the phase diagram that's a prediction from the model. Okay, now that's what we put into this dynamic model and then we did first the simulation which is only based on the assumption that what is happening during oscillation during doming is that EVLs are actively expanding and reducing the surface tension of the blaster dome. That would be one mechanism. We're exclusively looking at this mechanism now and that's what you're getting in a simulation. You get an embryo which looks like an embryo which is undergoing doming, but really not. I mean, if you look now at this triple point here, and particularly at this region down here, that doesn't really look like an embryo which is undergoing doming. So that already suggests that the assumption that it's only active expansion of surface health might not hold true. I mean, at least based on these simulations now. So we are not putting in any radio stress generation within the blaster dome. Now we did also a simulation where we only assume that what is happening in the embryo is radius stress generation, but there's no active expansion of the EVL, right? I mean, whatever expands on the surface of the EVL is passive, but there's no active expansion of the EVL cells, right? I mean, the only thing which is happening is active stress generation. And to our surprise initially, when you look at these simulations, that looked much more realistic to what is happening during doming in comparison to the simulations I showed you before, right? So we sort of arrived from these, based on these two types of simulations that active stress, radius stress generation within the blaster dome seems to be a more plausible mechanism which triggers doming compared to this reduction of surface health. And so we were not entirely sure how we put our experimental observations now together with the theoretic prediction, but what we wanted to know is we wanted to experiment where we can actually interfere with these two different processes. One would be active radius stress generation within the blaster dome and the other one would be active spreading of the EVL cells. Now, experimentally, we can address these two things and this is an experiment that Toshi came up which is a tricky experiment to do. But he took an embryo at the onset of doming so no doming has happened and then he wants to take out most of the deep cells and replace the deep cells with buffer. I mean, you change viscosity, but we can discuss it later. But essentially what we do is we reduce the amount of deep cells and thereby you reduce the amount of radius stress generation which can happen within the blaster dome. And he was assuming when he has an embryo which is consisting only of a York cell and EVL and then a few cells in between that he would look at a scenario where radius stress generation is being very strongly inhibited in the embryo and he would not see any doming, right? Because if, you know, from the theoretical predictions we thought that radius stress generation is gonna be the main driving force. Now that's what is happening and that was really surprising. If you look at this embryo, doming seems to be completely fine, right? Look over here. It's all okay. So even in the absence of, you know, a large amount of cells between the EVL and the York cell the embryo can undergo doming. And that again indicates that radius stress generation or at least a large part of radius stress generation within the blaster dome is dispensable for the embryo undergoing doming. Okay, so that was quite surprising. So we wanted to see if he can address now to see, you know, what is the contribution of active surface cell expansion. And these experiments which, where we took advantage of a mutant embryo and this mutant embryo is called, it has a name, it's called Kokey. And what is happening in these mutant embryos is a mutant in a gene which is called IKK1. And IKK1 is a part of the NFCAPB signaling pathway. It's just a background information. What is happening in this embryo is that the surface cells, these EDR cells are not undergoing differentiation. And because they are not undergoing differentiation they cannot actively spread anymore. These cells are remaining relatively small and they are not undergoing, they are not becoming really screened epithelial cells. Right. So we can take advantage of these mutant tissues and now doing transplantation experiments to see if we have mutant cells which are unable to expand would we still get doming in embryos, right? I mean the first thing he did is he took a mutant embryo and asked, would this mutant embryo actually undergo doming? That had been noted before already that this embryo is not doming whatsoever, eventually it actually collapses and dies. So in the complete absence of EDR spanning, doming is very strongly being inhibited. Just a quantification where we're looking at the B by I surface, this surface area as a function of time during the doming process. There's hardly any change in these mutant embryos and there is an increase in wild type embryos. Now the real experiments he wanted to do is now doing transplantation experiment. That's, you know, it looks relatively simple but it's a complicated experiment. What he did is he did two types of experiments. He takes now mutant surface cells from a mutant embryo which are unable to expand and then he replaces a patch of wild type surface cells with a patch of mutant surface cells. Now he has an embryo which is a mosaic embryo which is largely, which is widely wild type but contains a little patch of mutant surface cells which cannot expand. And he's asking, would that be able to interfere with doming movements in an otherwise wild type embryo? So assume Brady's stress generation is fine, everything is fine, just these cells up here cannot expand. Okay, and then he can do the reverse experiment and he can put wild type cells which are expansion competent and put it on an otherwise mutant embryo and see if he can actually rescue doming in a mutant embryo. Okay, now the first experiment is one which I just indicated he has a mutant embryo which usually would not undergo doming. He puts these wild type surface cells onto the mutant embryo up here and then he's asking, would he be able to rescue doming? Now that's what happens, you get spanning up here and whoop, you get a little dome forming right below the transplanted cells. So once you're putting only a patch of surface cells which can actively expand, you can locally rescue the upward bulging of this interface here which is indicative of doming. Let's just a quantification to convince you in a pokey mutant, this is a mutant embryo without transplanted cells, there's hardly anything happening once you're putting these transplanted cells in to get a local rescue of doming. Okay. You can do the reverse experiment by putting our mutant cells onto a wild type embryo. Now this wild type embryo usually forms a perfect dome as wild type embryos do and you're putting a patch of mutant cells up here and these cells are unable to properly expand, actively expand, right? They're just staying small, remaining small cells. And if you look at these embryos, you can see that doming is partially suppressed right below the transplanted cells where the transplantation has taken place, indicating that there's an inhibitory influence of these surface cells on doming movements. Very clear evidence and again, if you look at the quantification, this effect is partially transient but there's at least at this stage is doing doming. There's a clear inhibition efficiency of undergoing doming below cells which are unable to expand. Now, if you do transplantation between mutant wild type embryos, there's always a discussion about what you're actually interfering with. You can take mutant cells and you're assuming that the main defect in mutant cells is that they are unable to expand. But you could argue mutant cells are not only unable to expand but they are not secreting and chemotractin and they might do many different things or might be unable to do many different things which a wild type cell will do. So can we do an experiment where we specifically interfere only with the ability of surface cells to expand irrespective of their specific genetic background or a mutation where we don't know what exactly is being affected in that. Now, we did three different experiments. I show you to address that. In the first experiment, we are taking wild type mutant wild type surface cells and we are overexpressing a form of the small row GTP as row A, which increases optimizing contractility in these cells and which interferes with the ability of surface cells to expand. These cells are otherwise wild type of their over-activating row A, over-activating leading to activation of my Z2 and they impair their ability that undergoes training. Okay, so while it's happening in such a case again, you get a local reduction of doming right below the patch of row A overexpressing cells. This effect again is convenient and you can see it now in here and eventually they're catching up but there's a clear recognizable and consistent effect on the ability of this interface to deform in response to the inability of surface cells to expand. We can do the opposite experiment and now doing a transplantation between two mutant tissues. We are taking mutant surface cells, we're expressing now on the mutant surface cells amyosine phosphatase. And by expressing amyosine phosphatase, we reduce optimizing contraction of the surface cells and we increase the ability to undergo spanning. Okay, now this is the experiment we are taking amyosine phosphatase overexpressing surface cells and putting it onto a mutant embryo and ask for the possibly rescue doming in this homotypic transplantation experiment. And again, you see that there's a dome falling right below the patch of transplanted cells here indicating that by only affecting my optimizing contraction in these cells, keeping the genotype constant, you can affect doming in these areas and again, the quantification of it. Now the last experiment was the most tricky one where we discussed we want to do an experiment where we are not interfering with any gene function in these cells and the only thing we want to do is changing the surface area locally. Now the experiment he did here is taking a very large patch of mutant surface cells and replacing a smaller patch of mutant surface cells with this large patch here. So he's just increasing locally and transparently the surface area at the transplantation side. The genotype is exactly the same and he's asking, will that rescue doming? And that's what is happening. He gets a local rescue of doming. Eventually the embryo collapses and dies or there's a recognizable effect and consistently deactivates it. And I think that really was clear evidence that what triggers doming is the ability of surface cells to undergo spreading and increase a local active increase in surface area. Now, there was one difficulty in these experiments which we noted at the end of the transplantation experiment in discussions at the YOMA about the potential contribution is that what we find in these experiments we are putting for example, mutant cells onto a vial type or we are putting vial type cells onto a mutant embryo that the effects on the deformation of this interface are rather local. We see a locally inhibition of doming and we see a local rescue of doming. That is something you would not necessarily assume because in these transplantation experiments what you do is you put cells which are either unable to expand or which are competent to expand onto the other genotype but what you would assume is you globally change surface tension on the blaster belt eventually because these cells are making contacts to the surrounding cells. They should globally change surface tension, not locally. If you change globally surface tension you should have a global effect on this interface but you get a very local effect. That was a problem because that would not be what we would predict from simulation. If you globally change surface tension you globally change this interface you don't get these very local effects. So we had to find an explanation how we are arriving at these very local effects and what we came up with using a couple of additions to this dynamic model is the only way we can get local effects is assuming two different things. One thing is that these transplanted cells globally change surface tension but they have to do one more thing. They have to locally, right below the patch of transplanted cells, they have to locally modulate active radiocell intercalations. Only in this scenario we can get local effects on the interface. So we have to assume that what these surface cells are doing when they're expanding is not only changing the surface tension of the blaster belt but they also locally change the ability of deep cells to undergo active radiointercalations and thereby generating active radio stress within the blaster belt. So that was a prediction on models and a higher prediction on models which comes also for the local inhibition of domain movements. They would only work if you're assuming that these surface cells would locally inhibit radiointercalation of deep cells. Okay. So what you wanted to know is how would you provide to you here only with a slightly hand-raving explanation how would expansion of surface cells trigger active intercalation of deep cells? That's something we, you know, and apparently the theory predicts it. What we found is, if you're comparing now whitehead cells to mutant cells in which EVL sparing is being defective and you're looking at the density of deep cells below the surface cells, what we found is that the density in whitehead cells is smaller than the density in mutant cells indicating that spreading of EVL cells reduces the density of deep cells below. And conversely, if you're looking at the motility of cells below the surface in whitehead embryos which are competent to expand to mutant embryos which cannot expand, then you're finding that motility is higher in whitehead embryos compared to mutant embryos indicating that spreading of surface cells, one, you know, potential mechanism which could be in place here is that spreading of surface cells not only reduces surface tension but reduces density of deep cells locally below the sparing cells and this local reduction in density leads to deep cells becoming more motile and therefore becoming competent to undergo active radiative calculations. Okay. So what we end up is now a model where we have these two effects, surface tension reduction and active radiative calculations which have to be both taken into account. If you're taking both into account in a model which combines both of these effects, then you're getting a very close and we're taking all our measure parameters in there for tissue viscosity and surface tension, then we're getting a very clear match between the experimental observed changes in geometry during the process of doming to the experimental prediction which is really striking by looking at features of the relative distribution of surfaces from the york cell and the blaster term changes in the height of the blaster term relative to the york cell and changes in the contact angle at the contact line over here. And for all these different geometrical features, get a very close match between our theoretical predictions and the experimental observations. That is not really over-fitting anything here because you're putting a lot of parameters in this constraint, the remaining parameters that you have to fit here. Okay. So what we're ending up is now a model where we have blaster term surface tension where we propose that blaster term surface tension reduction is one of the main driving process which trigger doming. That radiative contraction facilitates doming and acts, you know, it's also required for doming that EVL expansion triggers both blaster term surface tension reduction and radiative cell layer contraction. And that's it. Okay, thank you.