 All right, so today, Vince Fior from Rockfeller University will tell us about morphogenesis, mechanics and malignancy about development of cancer cells in skin. So, Vince, please take the lead. Fantastic. Yeah, pleasure to speak with you all. Please feel free to interrupt as we go if there are questions and I'll try to go maybe a bit more rapid because of the late start. But yeah, it's a pleasure to share the work and hear your feedback. So, you know, one of the things that really fascinates me is the diversity of forms that biological systems take and in particular, within the human body. If we look at our tissues at the micro scale at the cellular scale, we recognize that all these different tissues throughout our body have these different structures and these structures are intimately linked to their function. How these arise are really still questions that we want to understand. And if we kind of zoom in and think about things that may be a bit of a conceptual level, what we understand is that the tissues have this arrangement of different cell types in different spaces. So in the middle panel here, I'm showing just cartoons of on the top, the skin, and it's a appendage, the hair follicle, or on the bottom, the intestine, and its distinct structure, which includes a crypt and the villus, the villus is where nutrients are absorbed. Crypt is where stem cells reside. And they produce all of the the other cell types of the tissue. So we I just color code cells, you know, for, for example, but we can think that, okay, red cells different than an orange cell, which is different than a yellow cell. And these cells have different functions. And really, the question is how do these cells come to be in the right place at the right time? And how does that influence the the whole function of the tissue? These are big questions. But I think the way kind of I think about it, perhaps, in engineering terms, is there's some sort of a signal, and there's some output. And there's this black box that the cell interprets the signal, and it makes a choice about what what sort of behavior will it will undergo, whether it's motility, growth, or death. And this can change the organization of the tissue. And this probably goes back into an iterative feedback loop sort of manner to ultimately give rise to these complex architectures. And one of the important things to think about is that each cell state, so how in what say flavor of cell you exist in at any given time, probably colors the this input output relationship. So what do we know about this process, we know quite a bit about how tissue tissues change their shape. And a lot of this comes from work in a simple model organisms such as the Drosophila, which is a monolayered epithelium, showing here the the the ectodermal cells here that will give rise to these folds, these dorsal folds, which is one of the one of the say prototypical mechanisms of shape change that have been observed. And so how does this happen? There's I would say two major mechanisms that have been described quite quite abundantly in the literature, which are ideas of differential growth and differential tension. So in differential growth to two tissues or two structures that are growing at different rates will give rise to potentially deformations within a confined area. Or you can just have growth, like extended growth within an unconfined space that will give rise to to shape change. The other is differential tensions. And this has been widely described in the context of the actomycin contractility, where say, if we're looking at a monolayer, the apical side of a cell might contract and change its shape, preferentially to the basal side, and this can give rise to shape changes like these dorsal folds that are seen here. So these are really these are ubiquitous mechanisms throughout the animal kingdom. And they've really been largely studied in simple epithelia. But some questions that have been really instrumental to my work is what other mechanisms could there be? As we know that there is really a diversity of architectures that of tissues that are seen throughout more higher order complex mammalian animals. Lastly, I'll just say that this this architecture is important because one of the hallmarks and maybe maybe the the synphonone characteristic of cancer is a loss of normal tissue architecture. And so we see in the schematic here on on the screen that as as cancers initiate and then progress to a real problematic state, a normal architecture say of this epithelium, where you have cells at the bottom, which are different from cells at the top, that really gets destroyed and disorganized and understanding that process is we think very important to understanding cancer. So some of these overarching questions are what are biophysical drivers of these tumor architectural changes? Do these changes influence tumor progression? And if so, how? And are say similar developmental mechanisms that we're that we're learning about are they co-opted during tumor genesis? Or are there similar mechanisms between tumor genesis and normal tissue development? So my work in the past few years has focused on the skin as a model tissue. This is a very interesting tissue, not only for its relevance in terms of disease and homeostasis of humans, but also its inherent architectural features. What I'm going to tell you about today is how cancers arise in the skin. And one of the things that really interested me when I joined a lab here at Rockefeller that focused on skin is that normally you have this architecture that's what we call a complex epithelium. There's cells at the base of the tissue that give rise to all of the differentiated cell layers above it. When those cells, what we call basal cells, when they get a mutation that can confer cancer, they give rise to two distinct shapes in terms of the pre, what I'll call premalation lesion. So before they totally invade the tissue and become really problematic, they have two different shapes. And those are schematized here. So if a basal cell gets a mutation in the hedgehog pathway, it gives rise to what are called basal cell carcinomas. And what I realized was that they have these distinct features in terms of their shape, which we've termed budding. And I'll describe that more on the next slide. Whereas if a basal cell gets a mutation in the brass or map kinase pathway, they give rise to premalignant lesions that we term folding. They both originate from those basal stem cells, but they drive distinct tumor architectures. And it really wasn't clear why. Importantly, however, these two cancer causing mutations give rise to very different prognoses and outcomes in terms of tumor progression. So while BCCs or basal cell carcinomas are rarely really invasive and metastatic, squamous cell carcinomas are. And so this kind of set up a nice question for us to try to ask the question of how do these distinct architectures arise? And might they be important for tumor progression? So to study this, we set up a system where we could introduce these lesions within the skin of mice and follow the growth and morphology of these lesions over time. And so to do this, we I'll spare some of the details, but we inject lentivirus into the amniotic sacs of mice. This will infect the earliest layer of cells they see, which happens to be those basal cells. We can then, by using this method, introduce any variety of oncogenic or genetic changes to those cells in a mosaic fashion. So in this case, we are introducing these cancer causing mutations, which are in the background of the mouse, either atras or smoothened through a through a cremated excision of their floxed alleles. The power of the system is we can also introduce other types of genetic alterations like knockdown of a specific gene of interest or CDNAs to overexpress genes. And so this gives rise to these these distinct folds, which features, which I'll characterize by these these morphometric features. But let me let me show you first that this works. So if we introduce these mutations, we'll give rise to if we introduce mutation in the hedgehog pathway, which I'll abbreviate as SMO or SMO M2, which is a which the details aren't critical, but that will be activating mutation in the hedgehog pathway. We get these butted like features. And if we introduce a mutation in the RASMAP kinase pathway, we get these folded features. And so I'll characterize them by a few different morphometric parameters. But what we what we decided was a very easy non dimensional way to parameterize the system was this shape factor s, which is essentially the indentation depth of the of the growth from the basal plane of the tissue over the curvature radius of the distal tip of that tissue. And so these look these look in our system, if you let them develop, they look like the lesions that we see in humans. And so he and the the morphometric parameters are here. So so we see that there's differences in indentation depth where these butted smoke smoke lesions made a little bit further have a very have a much smaller curvature radius, and that gives them a higher shape factor compared to the RAS holds. So so that so I'll say that the take home here is that for this shape factor s, these butted lesions that we get from from from hedgehog pathway activation, or if a high shape factor, and these folded lesions that we get from RASMAP kinase pathway activation have this low shape factor. Okay, so this was the system that we we set up, and we wanted to try to start to understand, you know, how how are these different architectures being driven? One of the first things we looked at was proliferation as these are oncogenic mutations. So if we look at how quickly cell cycle once they've been activated with the mutation, we find is that not surprisingly, the smoothen the BCC's and the SEC's, which I'll use for for shorthand, but think BCC is the smoothen activating mutant and the SEC is this RAS activating you. In both cases, cell proliferation is activated. And interestingly, we see that the density of cells within these mutant clones is elevated as are some of their shape parameters like the aspect ratio of a cell is actually increased. So these cells are getting kind of packed in within these clones, their density is increasing and they're becoming more columnar. They also experienced this anisotropy actually around the periphery. But this was all kind of well expected within and has been demonstrated within the literature as a function of growth within a confined space. So basically we have cells, they're proliferating higher than their neighbors. And so they're experiencing these compressive forces that are changing their shape and density. Is this what's giving rise to differences between these architectures as we see it really in both cases? To do this, we derived a functional genetic experiment to try to test this, which is if we diminish the proliferation of cells within these mutant tissues, do we change the architecture of these lesions, i.e. do we make a bud into a full or vice versa? And the answer is no. So what we see is that actually when we decrease the proliferation rate of cells within these mosaic tissues, the main take home that we see, the effect we see is actually, of course, the growth area of the individual clones diminishes. And this is just by titrating the amount of the cell cycle inhibitor within the tissue as does the indentation depth or general deformation of these lesions within the tissue. So in a dose dependent manner, if you decrease proliferation, you decrease the growth of these lesions and you decrease their ability to deform the out of plane of the normal epithelial plane, but we don't see changes in, this is, I should say, this is the same irrespective of the mutation that's being activated. So this kind of told us... Vince, what about the shape factor? So indentation depth sort of clear that it decreases, but what about shape factor? Yeah, so okay, so good question. And the answer is that, well, the shape factor actually, it diminishes slightly for both cases, but it doesn't switch the architecture from a budding to folding. So the shape factor will diminish slightly as a function of decreasing indentation depth, but it's not the drastic, it's not what is changing or dictating budding or folding. Yeah, but that's a good question. So what about, so it seems like differential growth is certainly happening and we think that's what's, you know, the important thing for driving these deformations, right? You need the growth to actually get out of plane deformation, but not the difference in shape. What about differential tension? This would be another hypothesis. So to do this, we did what are pretty classic experiments. We do a laser ablation at the junctions, in this case either within a mutant clone at the periphery of a mutant clone and wild type cells or say just normal wild type, wild type neighbors. And what we find is that, and so this is quite a bit of data, but I'll walk through one of the most important parts. If you look at, excuse me, could you explain a little more detail what is the effect of laser ablation? How does it influence the tension? Yeah, great, great. So the idea here is that we'll bisect a junction, it's kind of schematized here with the red dotted line, and then we'll follow the vertices of two junctions and we'll see how they change in their position over time. The idea is that as there's more tension on a junction, their displacement over time increases at a faster, at a higher rate or velocity than a junction that has lower tension on it. So this is purely a way to measure whether there is differential tensions on junctions within these mutant clones. There are some caveats to this, which is basically, you know, how are the material properties, let's say, of the tissue slightly affected between these cases, which could also influence retraction velocity, but for a good, for first proxy, these very initial reaction velocities, you know, seem to be a pretty well accepted and pre-robust measure of tension on a junction. Essentially, you cut it and you see how fast it moves apart. If you look at mutant junction, so the junction between two mutant cells in both smoothened or wrasse, it's very low, and we think this is because as these cells are compacting together, basically their tensions, their lateral tensions are being diminished. This is compared to two wild-type cells that are distilled to a clone where they actually do have a retraction velocity. That is quite substantial. The interesting thing here was that if you look at the retraction velocity between a mutant wild-type cell and a wild-type wild-type cell within the smoothened mutant tissue, there we do see a significant difference in their retraction velocity between these two cell types. So this really does suggest to us that at the edge of these mutant clones, there is this differential tension that could perhaps be giving rise to that shape, whereas in the case of wrasse, we don't see a difference in the interfacial tension between a wild-type mutant cell and a wild-type wild-type cell. Wins, can we also conclude from this that mutant cells push on the wild-type cells in this picture? So basically there is a tension more in the non-mutant tissue, right? Yes, yes, this is correct. This is correct. So if you look at, and yeah, within the center of the clone, there's compressive forces. Within the periphery of the clone, there's extensile forces, if that makes sense. If you looked at a wild-type cell right next to the mutant clone there, you would also be experiencing these differences in compression and tension, this pushing of mutant non-wild-type cells. In these cases where we're looking at these wild-type cells, they're actually quite distilled to the clones. We wanted to have a condition where it's kind of the base case, if you will. So I should also say that there is this slight trend where in totally wild-type tissue it is a bit elevated, though not statistically from the in the mutant tissue, which again gets to that point. This was corroborated by, if you look, if you're staying for things like actin and myosin, which are these major force-generating components of the tissue. If we look at the border of these smoothened clones, you can see these differences in concentration of these molecules that we don't see in wild-type or in the RAS case, and this is quantified here. So indeed it seems like the smoothened lesions, they exhibit differential interfacial tensions that are driven by actin myosin. At this point and throughout the rest of my talk, I will be going kind of back and forth between simulation and experiment, which hopefully I think for the audience will be some of the really interesting points of the work. Because yeah, maybe I should have prefaced this. Obviously there's a lot going on within these tissues in vivo, so it's quite a complex system, and we're able to make these genetic manipulations, but still if we want to try to maybe gain from some first principles what we think is going on, we thought it would be very beneficial to have a simulation component to see how well we were with our hypotheses and our measurements compared to what we could actually derive from first principles. So at this point I'll introduce you to the vertex model, which hopefully most are familiar with, but this is essentially describing the tissue with cellular resolution as with really two terms at this point that describe the work of the tissue that we minimize, which is this cell incompressibility term and these active line tensions that are at the interface between these cells. So incompressibility, we have this constant that is quite high, cells don't like to be incompressible, so this is essentially saying cells aren't going to disappear into nothing, they're going to maintain their area in this case, and the line tension saying that this gamma term can describe some differential tension or force on these junctions between cells like we had measured actually in the real tissue. The cool thing about this model is that like the skin is these, we built this model in this multi-layered fashion such that there's these blue cells at the bottom which are the proliferative cells in a normal case. If we introduce a mutation then that actually is going to drive this differential proliferation between the mutant and basal cells, and it's only the basal cells that'll proliferate, but they can rearrange their junctions, they can move upwards in the tissue, and eventually within this this bounded box they'll give rise to tissue deformations. So hopefully any, if there's any questions please, please. Yes, I wonder, but this vertex model, the number of vertices is not fixed, is it right? So it's not all hexagonal. So you don't use all hexagonal cells, right? No, no, it's not all hexagonal, and as the simulation runs, cells grow, divide, and once they reach a certain area, we bisect them, and so then they give rise to two cells. Yeah, so not all perfect hexagonal, and also, yeah, the number, these junctions are constantly rearranging as more cells are added throughout the simulation. So if we run the simulation and we, now I'll show you for cases where we're driving different tensions between the mutant and wild-type cells, as we observed in the smooth-in case, what is the effect? What is the effect? No, here, let me, let me, let me see if, yeah, it runs, it runs. It runs, okay. So, so, and I'll run it a few times, but basically on the top, there's no difference in tension between the mutant and wild-type cells, and as we move downward, we are increasing this parameter, this gamma parameter to a case where at the bottom, it's actually quite super physiologic, but we did want to explore the parameter space, and the middle is what we, what we actually, we measure in the tissue. So if I run it again, essentially what we can see is, yeah, cells are dividing, rearranging, and in, in all cases, we're arriving to this, this really a butted shape that, that looks most like the smooth-in lesions. So this for us was quite interesting, which essentially told us that, well, in this case, where we're just having cells divide, and we're having this multi-layered model, whatever parameters, the system is always giving rise to these butted shapes. So, so while we have this differential tension in the smooth-in lesions, which is associated with budding, it's really, we're still not able to figure out what is different between the budded and folded architecture. So this was kind of one of the outcome. I mean, it didn't, we didn't, we didn't solve our problem of what was distinguishing the two. And the main features that we do see is that you, you tend to see these cells become more excluded from their neighbors as their tensions get higher. So this would end, and yeah. And this is, I will say, very different from what we know about in a simple monolayer epithelium. So in this case, we're running the simulation, increasing the tension, either on the base or the basal side or the apical side of the tissue. And here we do really get a diversity of shapes, like just periodic folds or these kind of more apical or basely-oriented folded structures. And this is known throughout the literature for a long time that really, all you need is these differences in tension at a junction to get a real pretty, pretty large diversity of shapes. So clearly there was something about the multilayer aspect of the tissue that was different in being able to generate a diversity of shapes. And this is just, this is just shown here. This is just a graph of these results where we're showing, sorry, I'm showing the result of shape factor from our in silico simulations for a variety of interfacial tension parameters. And you really, you see that indeed the shape factor is not changing dramatically. And it's not getting us to distinguish between these two mutants. The last thing we did was, we did indeed do a genetic manipulation of this, where we knocked down myosin within these smoothen lesions, which is the main isoform that is generating these actomyosin tensions. And while, again, we see this subtle decrease in shape factor, which actually we predict from in silico as well, it's really still, it's not drastically changing the shape. So we can see the differential tension is involved, but it's not dictating the difference between the buds and folds. So what is, we kind of then took an agnostic approach, we decided to isolate these cells between these smoothened or ras embryos and see what kind of molecules are differentially expressed. Maybe that gives us a clue. And this is where we were quite excited by the fact that we found that the most highly enriched terms, if you compared smoothened to ras mutants, were genes involved in the extracellular matrix. And in particular this specialized extracellular matrix that's the basement membrane. So just a very quick, a very quick background on this basement membrane is this this polymer network that underlies epithelial cells. It's this network of laminin and collagen molecules that really provides this structural separation between epithelia and mesonkine. And these networks can additionally be crosslinked to change their biophysical properties by a certain number of accessory molecules. And in the skin, this membrane connects to the epithelial cells through these specific adhesion components and their termed hemidesmosomes. The interesting thing was that the, you know, really we know a lot about how these proteins behave biophysically by purifying them from tissue, maybe self-assembling them in vitro and measuring their properties. But what is the, over the biophysical properties of these networks really in the in vivo tissue was largely still a mystery when we started this work. So that was what we set out to do next. First I'll describe the model, the modeling aspect of it, and then we'll go into the measurements. So then in the model we implemented, we basically implemented terms for the stiffness, so both the stretching and bending modulus of the basement membrane, as well as a timescale for remodeling or assembly, which I'll talk about more in general. But basically how quickly is this basal membrane adapting to the growth timescale of the cell? So how quickly does that, when a cell is growing, how quickly does that membrane grow with it? I hope that makes sense. But okay, so let me show some of the simulations. And so here's where we started to get excited, as we could see really getting a more diversity of shapes by including the material properties of this basement membrane. And here if we look at the full kind of spectrum, stiffness on the Y and assembly on the X axis, we see that by bearing these parameters we get a quite rich spectrum of shapes, where you really need a high bending modulus to get folded type architectures, a lower modulus is giving rise to buds, and really with high assembly rates, we get these budded features to dominate. Sorry, Wins, can you comment a little about the model? So tau is the characteristic time of, and in your model, and lambda is connected, you're connecting the area, right? Yes, correct. Exactly. Yeah, exactly. So with respect to the previous model that you discussed, the area term, the effect of the area of the cell is still included, but through this lambda times square root of thing, right? Yeah. Okay. And it's not clear, how is the bending included in a way? How do you, what is C? Is it the curvature? C is the, yes, correct. The curvature of the Yeah, the curvature of this local. Yeah, the local, the element, yes. Yeah. So B, what is B? So B is our bending modulus, which we show for thin film, really, bending is dominating over stretching, and so that is our characteristic bending modulus of that network. Okay. And then what of all these quantities, so you can measure bending modulus in tissues experimentally, like a curvature, you already explained that you have been measuring to determine the shape factor, correct? Yes, thank you. Yeah. So now I'll get to the measurement of modules, but yes, this is the good point. So it was, we did have to devise a way to actually measure the local bending modulus, or we use actually small indentation and Young's modulus and converts to, it really is a good proxy for bending modulus in this case, but essentially we use a atomic force microscopy to, once we separate the epidermis from the base of membrane, we can go and locally measure the mechanical property of this tissue, which just is schematized here, but it was a headache, but we were able to do it. We find is that the base of membrane is much stiffer than the underlying dermis or the stroma, so we think by first proxy, that's the most stiff mechanically important element of this tissue, and as the model predicted, if you look at the smoothen lesions, actually those have the lower modulus, whereas the brass folds have a higher modulus. If we look at these proximal regions versus the distal tips, actually they have a high modulus, even higher modulus, which if I have time I'll get into describing why we think that is, but essentially at least to a first approximation, the model correctly predicted that you have a lower bending modulus in the smoothing case versus in the brass case. So is this really important? Does this influence the shape, architecture of these tissues? Again, we use a functional approach to basically vary the modulus by knocking down components of the base of membrane. If we knock down a collagen trimer molecule, we decrease the modulus. If we knock down one of these proteoglycans, we can actually increase the modulus, and if we knock down a crosslinking enzyme, we can also decrease the modulus, although to a bit lesser extent. So we went and then measure the effects of these knockdowns on the tissue, and we I'll just kind of point your attention to the quantification here, but here we do see that indeed if we change the stiffness, we can affect the shape of these tissues in a way that is kind of consistent with the model, where if we decrease the stiffness, we increase the shape factor, and if we increase the stiffness, we can we decrease the shape factor. And this is actually the trend is similar for both types of lesions. And so it seems like this modulus is indeed important in driving the shape, but we're not all the way there in terms of describing what is different or what is really the distinguishing features of these two. But I think we're getting close. So the last thing I mentioned was this, which was that assembly of the basin membrane is important. So we went to we went to investigate this. Essentially, if we look at types of molecules that are incorporated, either in an actively actively assembling basin membrane versus a more mature basin membrane, these are enriched within these distal tips of smoothened buds versus folds. We derived an assay to actually measure this in an X plant. So X vivo, we can take these tissues, we can culture them with just a small amount of fluorescently labeled basement membrane molecules and see how quickly they get incorporated into the existing basin membrane. And what we find is that indeed a high basin membrane assembly is is characteristic of these smoothened buds. Lastly, if we knock down laminin to decrease basin membrane assembly, we actually diminish the shape factor of these of these tissues, which is again, what the model predicted. So so here, I'll kind of describe what what what we really in in measuring and varying both the stiffness and the assembly of the basin membrane. We found that again, we were able to quite describe the morphometry of these lesions in ways that really, I think, started to make a lot of sense. So if we look at these regions here with a very so here shape factor is in cut in a heat map scale here. So you really you get these high shape factor lesions where we have a a high assembly rate and moderate to low stiffness. The shape factor decreases as we decrease assembly rate of the basin membrane. And it also decreases as we increase the stiffness. And so this was and so this is really this is, you know, quite consistent with with the results of our experiments. Yeah, the things I just want to want that in five minutes, we should go to the questions sessions. Yeah. So what I'm going to do is I'm going to skip through a a part here, I'll get to the conclusions. And I think, yeah, we can do questions and discussion. And I think that would be the most most fruitful. And I'll just go to a summary slide, which will kind of, yeah, I think get us there. But essentially, what we did was there's one, there's one more aspect that is important here. So we did uncover that these tumors, what I'll call tumor specific basin membrane mechanics are important in dictating these tumor architectures, as is a tumor specific cell differentiation program, which I'll just get into the summary of that on the final slide. So I'll skip all that. And I'll tell you and I'll just tell you all the summary, which is that the architecture, we think really is dictated by interactions between expansive forces, which are driven by the proliferation of those basal cells. The assembly dynamics and stiffness of that basin membrane that I talked about. And also the material properties of the super basal cells. And this is due to their differential differentiation program and the expression of care and molecules. And just to describe it briefly, it is that as these cells, as they differentiate and move upwards, they actually change their material properties as well, which all, which influences the folding versus budding shapes of the tissue. These mechanisms, at least these last two, thinking about basin membrane and super basal cells, these are distinct from super simple epithelia, where it's really been shown that this differential actinized intention is was driving things. Although, what we find is that actinized intention does kind of further quantitatively or subtly also these architectures, which we could think maybe they're really critical and normal morphogenesis. And what I didn't have time to talk about, but we think is really important, which is that actually the tension, the the extensile strain that's experienced by this basal membrane is elevated in a case where we have lower assembly of basal membrane and stiff super basal cells, which is exactly what we see within these swamous cell carcinomas that are more prone to invasion and basal membrane breakdown. So we think this could be a new mechanism by which this basal membrane is actually disrupted during tumor genesis and then leads to worse prognosis. So just kind of getting back to the general kind of questions, you know, I think that really these networks, them being the the extracellular matrix, the the base membrane and these cytoskeleton, these in particular intermediate filaments are really key, key networks, key drivers of how this this what I'll call mechanotransduction kind of feedback loop arises to give these different tissue architectures. And I think we've kind of at least understood a bit about how different architectural diversity can arise due to the nature of a complex versus a simple epithelium. And that is really this combination of basal membrane and differentiation of those super basal cells that expands this architectural diversity. So with that, I'll take questions. And if we want to go back through any parts of the data itself, I'm happy to. But I'd just like to acknowledge Elaine Fuchs, who's my postdoctoral mentor here at Rockefeller contributors, in particular, Mette Prank, who was at Princeton during the time of this work in the lab of Stops Schwarzman, who were really the drivers of the simulations and the computational work, as well as other collaborators that were involved in the study. So with that, I'll be happy to take questions. Thank you. Thank you, Vince. All right, we, because of the late start, we'll take a few questions over the time that was allocated. And for the talk, so if anyone wants to ask a question, you're free to connect your audio microphone, to connect your microphone and ask. Otherwise, if you have issues with your microphone, you can also type in the chat. All right, just while the audience thinks of the question, I will ask a question of my own. So you mentioned that the end of the talk is that while the cancer tissue proliferates, and it may also some layers of this tissue may change properties. So they will have different, so it doesn't mean that they will have different like bending models and other mechanical properties that you need for your simulation model. Is it correct? And if it is correct, what is the mechanism that drives this change? Yeah, yeah, so I'll just show a bit of the data. But yeah, essentially, it's just that. One of the things that we noticed was if we look, so this is doing this RNA-seq, right, and looking at different genes and how they're expressed. One of the other things, so I showed you what was really highly expressed in the smoothing case, which was these extracellular matrix proteins. What's highly expressed in the RAS case are these, we call them their differentiation genes of the skin, but it's essentially these proteins that make up the barrier. And that barrier, the outside of our skin obviously needs to be robust mechanically to keep water in and keep microbes and such out. And so this is again one of the major differences between these two states. And what I'll, yeah, I guess to more, these are like, you know, these looking at different keratin molecules, which are known to be important for this making the barrier. But if we just think about the material properties, again, now we could go in and actually measure these with AFM, looking at the layers, and we see that it's really quite beautiful. You see this fast change in the material properties of those cells as you move upwards in the layer. And so, you know, basically what it's like your skin gets harder as it's further outside of your body so that, you know, you kind of have a nice barrier at the surface. And this also happens in the tumors. So to answer the question, there is this change in material properties. It seems to be due to just the normal process of barrier formation that the stem cells undergo. I see. Thank you. All right. Do we have questions? Yes, I have a question. Yes. Thank you for your presentation. I have some question. Why is it important to study the architecture of skin cancer? If it's to define the stages of all the skin cancer, is it to provide the medical treatment or what? Yeah. Good question. So, yes, I mean, at the end of the day, what we do when we uncover is kind of novel mechanisms of how cancer progresses. And whether it's, you know, skin cancer per se, or whether these are more generalizable principles to other forms of cancer, that's, you know, that's like the next set of questions and where we want to go. But we really, you can think about how these principles can inform mechanisms of progression and how it depends on the architecture of the normal tissue. So there's multiple, multi-layered complex epithelia within the body. We like to know what are these same mechanisms are at play within those tissues. But the bottom line is just that, yes, we still don't have mechanisms to stop even that initial process of barrier breakdown, the base of membrane, and stopping cells from spreading from that initial primary tumor site. So, yeah, the goal is to actually use these findings to inform, to inform treatment strategies, although I mean, we can talk about what they are. They're not, they're, they're not easy, but that's what we're thinking about. Yeah, well, I would add that to Boris' question that, first of all, architecture is very important also in the design of drug delivery systems. So one of the things is that when you treat the cancer, you need to ensure that drugs that need to, that fight the cancer will, will be delivered there and diffused through these tissues. So architecture of those cancer is important for designing cures and medications. That means when, when we find the architecture, we know exactly almost the kind of treatment we are going to apply to the skins. Well, there are several questions to that. Okay, first there are, there is a search for drugs and drug delivery. So for that, you need to know mechanical properties of the tissues and architecture. But there is also a question, or why this, right, why, why this cancer form this architecture, these features in the skin. All right, so, so to answer your question, I mean, this, this is new. So what we do know is that when, when a pathologist would go for a cancer, look at skin cancer, let's say specifically, they would see, okay, it's got this budded histology or this folded histology, or at some maybe later stage of that, but really then they will classify them as two different cancers and they'll, they will treat them differently. You know, again, when we set out to start this, we didn't know why they looked different. But, you know, now that we understand that and what some of the, the kind of constraints are that drive these different shapes, I think, you know, nobody, we don't have the treatment that says, okay, we need more base of membrane assembly in these brass driven squamous cell carcinomas, which again, before we even identified the, the, the cancer causing gene, we, we, we treated them differently. But now that we know that, you know, that that would be the goal is to, yeah, we're working on some screens to try to increase the expression of these important molecules that help to halt the, the next stage. Thank you. You do have another question, Boris? Yes, I have the last one. Yeah, go, even if it's not. Can you, can you go back in your slide where you show the, this, the, in a figure half walk video starting by the initial point of skin cancer. Sorry. Can you go back? Yeah. Could you say again to, to, to, to where is it? In the figure, I think it's this slide video though. No, no, no, no. The beginning of the presentation. Yes. About slide number 10 means one. I'll just, I'll just keep clicking. You can say, no, no, it's, it's before, it's before. Before. Yeah. Go back. That's the second slide. Ah, no. Move, or why is it not here? Could you just formulate the question without the slide? The problem is in the figure, he, he used two, two rules. One is like reversible. And I would like to know if, if I move from the beginning of the, the skin cancer, that means the, the first stage and move from the, the second stage card, is it possible to go back to the, to the first stage? Because in the figure, I have two rules. I think you want this kind of schematic maybe? Uh, it, yes, but yeah. The, the rule, because here you used to, you have, for instance, male stage and moderate stage. Can we move from the moderate stage to the male stage? Yeah. I mean, yeah, I think in principle, those are, those are the ideas that, that, that we're trying to explore, which is, yeah, I mean, clearly you can. So clearly you can push, obviously one thing is to eliminate the tumor cells, but then to push, push those cells to say differentiate, differentiation therapy to promote them to kind of be removed from the tissue through normal means. Is, you know, a potential way to move from, say, a disease state back to a more normal tissue architecture. So I think if that's, if that's the question is, can we and how can we, that would be an example of, of an approach. But I think we're, you know, we're still a long way from actually being able to do that. But, but that would be a goal. Obviously, you know, if it's, if it's a tumor that you're able to see and operate on, then you would move it. But if it's not, and it's, you know, say, small or it's, you know, there are many, many of them, then you do have to find a way to, to get rid of the tumor genetic tissue while still maintaining the normal tissue. Yeah. Thank you. Thank you. Are the questions, do we have? Not at the time? Okay, just