 which gives rise to multiple small impressions. And that small impression doesn't inflate the cell and stretch the cell wall. So there is certain tension in that wall, which is different material, and the cells usually add about, in the case of the syringes here, which is a sphere that I'm going to talk about most of the time. This is about three hours here. We have two big pressures, but a wider one of the pressure. So deep in for something that is of a wider length here, roughly 2, 3.1, and the cell walls that I was mentioning have a more semi-permeable membrane right there that separates the inside and the outside. So that's semi-permeable membrane, semi-permeable membrane. Sorry, I think it's not on still. Let's get that there. Now wait. Now it's on, it's a trigger in the water. One, two, one, two. One, two. OK. Yeah, OK. Now we're done with this. OK. So the semi-permeable membrane consists of two lipid bilers. Let me not get tangled up. So two lipid bilers. One is on the outside, and the other one is on the inside. And in between is this stiffer material called cell wall. So the two lipid bilers with the cell wall compose the membrane. That is semi-permeable to water. So water can freely go in and out of the cell. So because you have two lipid bilers, there is this intermediate smaller compartment. So you have the inside, which is usually referred to as cytoplasm. But I will be telling you inside, big inside compartment. Then there is the outside. And then there is this in between compartment. And actually very little is known about this in between compartment, because it's roughly 30 to 50 nanometers, sometimes even estimates at 20 nanometers at width. So this is below the fraction limit of an optical microscope. It's less even than some of the super resolution microscopes. So if you want to do something that is live and fast, it's very hard to measure what is going on in this compartment. So there is still some argument within the community of whether that intermediate compartment is the same as the outside in terms of pressure and pH, or is it same as the big compartment inside. So if the membrane is semi-permeable to water and you suddenly change the external concentration, first thing that should happen is really just physics. The water should move according to the chemical potential of water. The fact that there is a concentration difference ordinarily, it is really due to life. It's due to the bacterium putting in work to maintain that steady state of concentration difference. So when the external concentration state of it increases, what happens first? And some of you have seen this yourself during the hands-on session. The water comes out of the cell, so that the inside of the cell and the outside equilibrate, and the cell will shrink. So the other way around, and we have also seen this at the hands-on session, or for some of you who will still come, when the external concentration decreases, so now in order to equilibrate the cell and the outside, the water is moving in, and it's inflating the cell to deconcentrate the inside, and the cell risks bursting. The aim of this inflating of the cell is the same, to equilibrate inside and the outside. So both of these scenarios, one is you're now a shrunk cell, and you can shrink up to 50%. I'll show you some videos. And if you're about to burst, both of these scenarios are not beneficial for the cell. And it does have mechanisms to cope. So in the case of an increase in external concentration, so remember now you're shrunk, and you want to come back to your normal volume. So what the cell does is uses these channels that sit in this semi-permeable membrane, and they will pump either ions, potassium ion, or organic osmolides. These organic osmolides are just molecules that you can accumulate at high concentration, and they will not disturb the charge balance across the membrane too much, and they will protect some of the proteins. So when you bring these osmolides and potassium inside of the cell, the water follows, and then the volume expands. If the volume expands enough to still keep the concentration difference equal inside and outside, you're still not pressurizing. But at some point, you'll be expanding the volume up to the point where you start stretching that cell wall, which is the stiffer material. And that's where you start pressurizing, because you're no longer expanding enough to equilibrate the inside and the outside. Interestingly, other than these pumps, the cell can also make extra solids inside. And the goal is exactly the same. You increase the concentration of solids inside the water will follow, you will repressurize. So this increase in external concentration really refers to in the literature as hypothermotic shock. So the other one, when you decrease the external concentration, and now the cell is about to burst, it has expanded, what is thought to happen is that these mechanosensitive channels, these are these iris-like structures, open up and let the solids out. And I will come back to that. I will show you how from some of the volume traces, we can understand how does this mechanism work? How do they actually recover? But before, I'll tell you about how we do, how we look at these individual volume cell area traces to be able to learn about how they shrink and also how they recover. And this is what we have been discussing in the hands-on session. The experiment is as follows, and it's exactly like you have been doing in the hands-on session as well. So we have some bacteria that are attached onto the cover slip surface. They are here in green and are as exact same strain that we have been using in the hands-on session. The reason they're green is because they're expressing a fluorescent protein that is freely diffusing in that inner, bigger compartment. So because it's freely diffusing and it's small, so this is really fast, there's quite a bit of it, the whole inner compartment is essentially labeled. So green area is that inner compartment. We attach them to the cover slip surface and then we also attach a plastic bead onto that cover slip surface. This is not something that we have been doing in the hands-on session. So that plastic bead is there to allow us to stabilize the slide in X, Y and Z so that when we administer in these shocks and when we're imaging for a longer period of time, nothing moves. That makes our analysis a lot easier in terms of X and Y, but in terms of Z is essential because changes in Z can influence our area. We're looking at just one area. Okay, so now that plastic bead we put into an optical trap, which Pietro told you about yesterday, except this optical trap is heavily attenuated. So it's not really trapping. There's just not enough photon momentum to trap, so it's heavily attenuated. And what we use it for is something called back focal plane interferometry. It just allows us to detect the position of the bead. So if the bead moves in X, Y and Z, we can correct it by moving the stage. And because the bead is attached to the cover slip surface, if the bead moves, the slide has moved, so we're effectively correcting for it. And how this then looks like when we subject the cells to shock, you see them here, they're fluorescent in green and it's a real time, cranking, here's an increase in the concentration and then they're slower, the recovery slower. This is now the active part where they're pumping stuff in and generally recovering. Okay, so on microscopes, although the slide preparation is as we have been doing it in the hands-on session, on microscopes are a bit more expensive. We have two right now, have a Bowser and Mario. They are custom-built, so we do machine the parts. So compared to a commercial microscope of the same price, we will be about half a price. We're still at around 50 to 80,000 pounds. So it's not a tabletop experiment in this case. But it is cheaper than if you're buying commercial and in case you're wondering, yes, we do have a character Luigi, it's a smaller microscope, smaller nickel microscope that is complimenting Bowser and Mario. But we do most of the experiment in Bowser and Mario. Okay, so how do we analyze? So that part, we have covered in the hands-on session. So you have a trace of a cell and we choose ones that we call flat cells. So that means that they're uniformly attached to the covers of surface, so they're not tilted up. And that ensures that when we subject them to these shocks, they're not coming down. If they come down, it will look like fluorescent intensity has increased, which could be misinterpreted as an increase in area. And it's not, it's just the cell coming down. So we really choose the flat ones, we subtract the background, and then we go across the cell and find the pixel intensities. And then we set a threshold value and that threshold value is based on the point spread function. And we count the pixels above that threshold value as belonging to the cell. Apart from the inner compartment, we can also dye the outer membrane. And that outer membrane is attached to the stiffer material. The outer lipid bilayer is attached to the stiffer material, the cell wall. So if we do that, then we have this inner compartment and the outside of the intermediate compartment. So if we subtract the two, we have information about the intermediate, that small compartment between the outside and the inside of the cell. Okay, so we do it slightly differently there. We use a level set method. But we can extract the information about the cell area from these images and we convert it to volume usually. If the cell is not changing size too much, we will say it's a sphere of cylinder and that's how we convert it into volume. If it's shrinking quite a bit, we'll say it's rotationally symmetric and that's how we convert it into volume. We could do with the area as well. So to be honest with you, is usually the reviewers who asked this would be converted to volume and then we just do it. But the area will give you the same information. Okay, so how do the traces look like? This is now a video of a hyposmotic shock. The cell will expand a little bit, about five to 10%. And if you watch, you might be able to see it. There it is. I definitely saw it, but I have seen hundreds of these cells. So you haven't seen it. Don't worry about it because here's a trace. This is now volume in micron cubed and this is stym in minutes. The frame rate is one hertz, one second. And you can see the initial volume, the fast expansion, and then the slower recovery and then the cell go above their initial volume, they continue to grow. They're in a rich media where they have enough nutrients, sugars and amino acids so they can grow. Show us again. The video? Yes. Okay, if you haven't seen it, let me show you this one because this one is the shock increase in the concentration. You'll definitely see this one. There it is. Right, so they shrink. And in the volume trace, again volume in micron cubed against stym, you see the initial volume shrinking and then that slower recovery. This here is physics as is this. Then the recovery, slower recovery is then the cell trying to get itself back into steady state and the concentration different exists. Okay, so the first thing that we've done with these traces before we went into these active mechanisms of recovery is to learn something about how the shrinking occurs because if it's just physics, we should be able to fully understand this. And in the literature, there was some controversy about how this shrinking occurs. So remember we have three compartments. We have the big inside compartment, the intermediate compartment and the outside. And what you would normally see in literature in some older literature, this is now 1963, is an EM image of a shrunk cell. So the outer membrane in the cell wall is this here. This is going to be read in our images. The inner membrane and the inside compartment is there. So this is the cytoplasmic, so-called cytoplasmic compartment. And you can see that it has peeled off. So it shrunk, but only the inner compartment and the wall have stayed intact. And this was labeled plasmolysis. And in some experiments, they would see it and maybe if we waited longer, maybe we would see it or maybe it happens on a slightly different time scale. But what we thought like all of that has been seen in the literature should fit. If you think about where the solutes that you used to increase the concentration can go. So if a solute cannot pass this membrane, then the outside and both of these compartments are going to be different. So the outside is going to be much higher than both of these compartments. And then the whole thing should shrink. If the solute that you used to increase the concentration can pass into this intermediate compartment, then the intermediate compartment and the outside are the same high concentration. So only the inner one is at a different one. So only that one should shrink. So if you have a solute that can slowly go into that compartment, initially not, then initially both should shrink. And as it slowly goes in, the two should separate. So we found the solutes that do exactly that. This big one does not pass. This salt passes straight into this intermediate compartment. And this sucrose, so E. coli does not eat sucrose for some bizarre reason, unlike us. It likes glucose, it does not like sucrose. So you can add its sucrose, it will not eat it. It will just serve as a solute. And the sucrose goes slowly in. So if I show you the videos, you'll see this one here. Oops, let's go up. So this one here, the two shrink together first. You see it here. And then you see the separation as the sucrose slowly goes in. This one here is the big molecule that can't pass in either compartment. And you see that both, even the stiffer one shrink and they stay shrunk. Versus this one here, sodium chloride, that goes into the intermediate one, you see that it goes straight into that in between compartment. And that's why the cell wall does not shrink. It's only the inside that shrinks. So this is purely passive. And it does confirm that the water can pass quickly through that semi-permeable membrane. And what happens is really just due to the chemical potential of water. Okay, so now I'll go back to one of the mechanisms of recovery. And I'll go back to that mechanism of recovery where the cell has expanded. So we have decreased the external concentration. Water has rushed in. The cell is expanding. And the estimated pressure in the cell from the literature is now around 300 atmospheres. So this is really pressurized. And it risks for the cells to burst. And you can see sometimes if you do these large shocks you just see the cell. It can rupture, it can break, or it can gently leak, or it can just literally burst. You just lose the fluorescence all of a sudden. The whole thing pops. So in order to stop it, mechanosensitive channels are involved. And structures of these channels are known. Here's one of them. So here on the left, you're seeing the side view and this is the top view. It's an RS-like structure. For some of them, there's seven in total in E. coli. Open structure is also known. So here now you see the closed structure on the left and the open structure on the right. The top is the top view and the bottom is the side view. So you see that when the open structure is about three nanometers, two and a half nanometers wide. So this is a big hole. This can let proteins out. And I'll show you a little video of how this looks like. So there are going to be some of the parts of this protein labeled in the video. Ignore that, that's not important. Just for this talk. I'm sure the authors think it's very important. But for this talk is not important. So just watch how this happens. Here's the video. So this is now from the side. And when there's tension in the membrane, this is how the channel opens. So you can see it as well from the top. You see that RS-like structure and it's opening because the membrane is now on the tension. So usually this mechanical sensitive channels have been studied in vitro. So people can get them into lipid vesicles. So there is no cell wall. There is no two lipid bilayer. It's just one lipid bilayer. And they can reconstitute them and then they can suck it into a pipette, literally. They get from that vesicle, they suck in a little bit of lipid bilayer and then they have a channel in there. And then what you do this is a classic experiment in vitro for mechanical sensitive channels. You apply pressure. And as you apply pressure, you're watching the current across that membrane. So when a channel opens, you get current across. You have ions on both sides. So pressure, pressure, pressure. You get to the threshold value and you open one mechanical sensitive channel. Then you open another one. Then you open a few and then a bigger one. And you should be able to close them straight away. In vitro, they close straight away when you decrease that tension. So what is important for the rest of the story is that that opening time is around 60 milliseconds. So they really see them opening fast and closing. You increase the tension, you decrease the tension, open, close. There is seven of them in E. coli. These mechanisms of channel of large and small conductance are to be the main ones that are involved in recovery. These three small ones have only, in 2010 they have been found and they contribute very little. The number of these channels are on the order of hundreds. But in different media, they change the number of different channels. One of the questions that I'll come back to if I have time is why seven? If you can control the number, then why would you just not do it? And can you do with just one? Can you recover from all the cases? Why would you have seven? Okay, so the question that... Yeah. It is some given concentration of salt. Yeah, yeah, yeah. So just so that it can detect the current across. But they're not interested in which particular ions or anything like that. It's just a signature of a channel is open stuff is coming through. They're supposed to, because so there is some controversy in the field whether they're specific or not. But I mean, it's a three nanometer hole. They can't be specific things. And the 300 atmospheres is just water and solutes will come out. Whether it's purely, whether on a side of that channel, you can have some kind of interaction that will slow it down and make it a little bit more preferable for certain solutes or certain ions, maybe. But until you prove that, if you tell me there's a three nanometer hole, I can't assume that this is specific, right? You need to tell me something else for, and most of the field will agree that they're non-specific in the case of E. coli. That's not necessarily the case in all of the life cells. Okay, so our question was, so we know about that they open and close in vitro. So what happens in a live cell? How does this opening and closing actually cover it? So just to remind you, that cell has two lipid bilayers and has this cell wall. And this is a bit more realistic, still a cartoon representation. So inner lipid bilayer, outer lipid bilayer, here's this stiffer cell wall. And these mechanosensitive channel are on the inside. So they open, and now when there's a lot of pressure, they are going to get pushed against this cell wall. So we were wondering, how does this actually lead to recovery? And to illustrate, I'll show you a trace of one cell. So this is now that same volume trace that I showed you at the beginning, except here is now normalized. So everything starts at one, because we only care about relative changes. So everything starts at one, and then this is time axis in minutes. This is a very large shock. And we have chosen a cell that accentuates the phases. So it starts at one, and then it expands. That's quite fast. And then there is a slower recovery that can go below the initial value. And then they recover back and continue to grow. So the difference between recovering back and growth is arbitrary, we said it at one. Like if it's bigger, a lot bigger than one, it has to be growing. But where exactly the growth starts, it could have started a little bit before. Growth means insertion of new material and making a new material. It could have started a little bit before they fully recovered to one. But we don't know, so we said, okay, above one you have to have extra material. Okay, so just to convince you that this is not just one trace, that this is consistent among different cells. I'm showing you now 50 to 100 cells that are averaged in each one of these traces. The black is the average line and yellow is the standard deviation. So the magnitude of the shock goes from low to high, top to bottom. This here is that normalized cell volume against time. Time is now five minutes only. So what you can see that they start at one and they expand and the expansion does increase with the magnitude of the shock. You can also see that the recovery bed is slower. So it takes time to recover and it takes several minutes. And they also can overshoot. And that overshoot going below the initial volume is slightly bigger with the higher you shock them, the higher you initially expand them. So the first question is, is this anything really to do with mechanosensitive channels in vivo? So what we've done, we have removed, we deleted the two main of those seven channels and we call it a double mutant. And if those two main channels are not present, you can see that they still all expand, but the recovery is a lot less and they definitely don't overshoot. There was one other difference. If I expand the scale, I'm gonna show you the exact same image, but I'll expand the scale to go longer in time. Here's the original one with all the channels. You can see that after some time, all of them continue to grow versus when we remove those two main channels, not all continue to grow. But it's higher shocks. You see that even after 35 minutes, these cells are not continuing to grow. We don't know why. Some of them here are also bursting and breaking. Okay, so just to put some numbers in because this tells us something about that semi-permeable membrane. So here now is the maximum volume, normalized maximum volume. And this is the shock magnitude. We're going from lower to higher. The one type is blue and the double mutant, the one without the two main channels is red. So you can see that the expansion does go higher with the magnitude. The more you shock them, obviously they go higher, but it's saturated. So at some point you can't expand the cell anymore. But the saturation is around 15, even 20% for the double mutant, which if you calculate how much lipid membrane, how much you can expand the lipid membrane if it's under tension before you start expanding, you can't expand that much. Cell wall, the stiffer material you can, according to the estimates in the literature, but the cell membrane you can't, it would burst. So it does suggest, and there is some other evidence in the literature, that there is some extra membrane there. So that it does not need to be under pressure so that there is some extra membrane there. In terms of the time it takes to increase, so this is, according to Rheinstein, this is that fastest slope of expansion against the shock magnitude. And you see that this is second. We say that it's the same for both of the strains, although maybe something happens there, but we don't think we know enough to go into that. If we look at the overall time it takes to reach the maximum volume, that now is a bit slower, but still on the order of seconds. And I'll come back to that. We think we can understand why is that a little bit slower. Okay, some other numbers is just that minimum volume. Remember, they increase, they slowly go down, and then they can even overshoot. So we went for the minimum volume. What's the minimum volume? And this is now plotted at normalized minimum volume on Y and on X axis you have still the shock magnitude. Same thing, red is the double mutant, blue is the wildtap, you can see that the wildtap goes down more and the more you shock them. And the double mutant doesn't really go down, right? They recover very little, which is what you can see here. And that time, this is now the time that it takes to reach this minimum volume against the shock magnitude. You really see that these are minutes. For higher shocks it's five to 10 minutes, so that recovery is very slow. Okay, so we asked if you open a channel on a 60 millisecond time scale, why does it take you five minutes to recover? Why is that so slow? Can we understand that? So the first thing that went to say, okay, let's just check that the water can still freely move in and out of the membrane, that the biology has not done something we're not expecting after these shocks. So what we said, okay, because when we're down-shocking them, we need to grow them at a higher osmolarity and then bring them down by exchanging the liquid like we were doing in the hands-on session, right? But because we're growing them in these higher osmolarities, maybe the membrane is such that the water no longer passes as freely. Or if we expand them, we grow them at higher osmolarity, we now shock them, they expand, so maybe that expanded membrane can no longer pass the water as fast as we expected. So we wanted to just check that before we go further into understanding why it takes so long to recover. And the way we've done that, we have grown these cells at higher osmolarity now and we're gonna increase the osmolarity even further. And then we expect the cell to shrink, right? And we wanna see whether that happens as fast as we're expecting it. So here's now the normalized trace again. This is a 10 second time scale, time. Again, the average of both 30 cells and the purple is the standard deviation. So you see them at initial volume, we increase further and they shrink fast, less than a second. So water still goes fast in and out. Here, we have them at a higher osmolarity, we down shock first, they expand and then we wait a little bit and increase the concentration and they go fast down. So the water is moving fast in and out. So that's not the reason why the recovery takes several minutes, up to 10 minutes. Okay, so then we went and said we need to think about this a little bit and we need to work out what is the chemical potential of water in the solids to be able to understand this slower recovery. So we start by saying we know the osmotic pressure. So the osmotic pressure is equal to the concentration difference between inside and the outside. And because solutions are non-ideal, we have some fudge factor. Otherwise it would be really just equal to that concentration difference. So at very high concentrations, it's not exactly, so there is a factor that corrects for that. But this is, this part is not important. The other thing that we will take into account is the elasticity of the cell wall and we define it as true stress over true strain. So this is d sigma over L, where L is the thickness of the cell wall and dr over r, where r is the radius of the cell. And sigma is a surface tension and we assume this to be a hollow cylinder with a tin wall and we write it as pressure times the radius. So this is now the Laplace pressure. Okay, so the water will be moving across this membrane, according to the difference between these two, osmotic pressure and the mechanical pressure that has built up in the cell wall. So we can write that. We can write change in volume of the cell in time is going to be proportional to this difference. This here is just a molar volume of water and this constant here does have the time. If you're wondering when did the time go, this is the effective conductivity. So it has one over time in there. Right, okay, so we know this one here. This is the osmotic pressure. It's proportional to the concentration difference. We need to work out the mechanical pressure in the cell wall. So just to remind you, we have defined the elasticity as true stress over true strain. We have here the expression for the surface tension. We also have an experimental result which says that the elasticity of that cell wall is proportional to the pressure. So exhibit stress stiffening. The more it's inflated, the stiffer it gets. So the experimental result actually has this gamma coefficient there, but their coefficient is 1.22. So for our model, we said one. It will not change what the model tells us. Okay, so now we have these two expressions. We have this one and this one and we'll just solve it. You see there will be DP over P and then there will be R squared, R and R squared. So you'll get the expression for the pressure. So this is now the mechanical pressure. You have an exponent. This V to one third is just R. That's just the radius. Delta C zero is the initial osmotic pressure, the initial difference between the two. Okay, so we can write our first equation. This is now how the volume of the water will be changing. This is the osmotic pressure and this is the mechanical pressure in the cell wall. And the water will be moving according to the difference between these two. The only thing that I still want to add is this A. A says that when a mechanical sensitive channel opens, more of that can happen, right? Because now you have a hole in the membrane. So A is linked directly with a mechanical sensitive channel. So the other equation that I need is the solutes. So when you open that hole in the membrane, now the solutes can pass in and out as well. So they will pass, there will be diffusion. There will be a concentration difference. So this is just a diffusion component, but also they will be pushed with this mechanical pressure out as well. So we have two parts in this equation. Here is the mechanical pressure that it's pushing them out and time the concentration that you find inside. And this part here is the diffusion. So there's the concentration difference, thickness of the membrane. And all of this is to do with diffusion of the solutes, number of mechanisms in the channels and area of the mechanisms in the channels. Here is that constant A in both cases because the channels are open, right? So now I have these two equations. One is for the solute flow and the other one is for the water flow, the changes in the cell volume. And I can solve it and get the volume, water influx, water efflux, and solute efflux to help me understand why are they recovering slower? Okay, so for the volume, this is now in thermal leader, so it's the real volume. Water influx, water efflux, and solute efflux, this is all coming from the model. So this is just solving those two equations. We don't have access to this experimentally, we have access only to this volume, which is why this is really helpful to help us think about it. Okay, so there are four lines. This is now time in seconds, it's quite fast. So the first one is the hyposmotic shock. So at the point of hyposmotic shock, the water influx is large, water goes in, right? There is still no solute efflux or water efflux, mechanosensitive channels have not opened yet. So now as the water is going in, at some point tension is high enough and the mechanosensitive channels open. So that is the channel open at the second light. At that point, you'll see that little bit of a king. The chemical potential is still driving the water in. So a little bit more water goes in, right? You open a hole so even more can go in. Yet at the same time, the solutes can now start coming out. Both the solutes and the water will be pushed out by the mechanical pressure in the wall that is now large. So each time you push that one solute out, you have less incentive, less chemical potential driving the water in. So you have a competition between the two. At some point that equation, this one here goes to equals zero and that's where the volume of the cell reaches maximum and now it's going to start recovering because less water is going in and more solutes are coming out. And this takes time because originally when you open the mechanosensitive channel, the chemical potential is still driving the water in. So you need to push with a mechanical pressure one solute out and then water comes out as well. And that happens slowly, it takes time. Also, depending on what happens with your small pressure, you could lose a small pressure while there is still mechanical tension in the wall, mechanical pressure in the wall. So then you need to squeeze the volume a little bit more so that you balance it out with your small pressure, which is why the volume can go below the initial value, which is why you overshoot. Right, okay. So this helps us understand what's going on and we think this is a passive control mechanism. You just open a channel and then you have a competition between the two, which takes a little bit of time. So it's not an instant recovery, but it's just driven with the chemical potential of water and chemical potential of solutes. And because it's a passive control mechanism, you can overshoot. You can go below your original volume, but that's okay because that's not as dangerous as you bursting, right? You wanna prevent bursting because you're dead for sure. Like if you overshoot a little bit, then you can use those other mechanisms to bring the volume back up, right? You will not die. So we fitted the model to a trace. This is now volume in femtolidors against time. A trace of one type, a characteristic one. We just chose one. And you see the expansion and the recovery in black. And then the model, the fit is in blue and the orange are the confidence intervals. We can do that for the double mutant as well. So this is again volume in time and double mutant, but look at the number of parameters, right? So we have seven parameters. Three are from the literature. This is the initial, osmotic pressure initial. This one we can measure initial volume of the cell. Initial elasticity and the thickness we can get from the literature. We're still left with four, right? So of course we eventually fitted these traces. But these two, A and alpha, are to do with mechanosensitive channels. So the wild type value is in blue versus the double mutant is in red. So you can see that the wild type is a bit bigger compared to the double mutant, which is what we were expecting because the wild type does have those mechanosensitive channels, right? But from the model, we can also learn something about what we expect when these parameters change. So what I'm showing you here now is volume on y-axis and times in seconds on x-axis. And I'm going to choose one of the six parameters and vary that parameter from low to high. So green is low value of that parameter, red is high value of that parameter. All of the other parameters stay fixed. So I'm only changing that one parameter in the model, this is nothing to do with experiment, which you can probably tell because it looks too pretty. And what I'm plotting is the volume of the cell. What happens with the volume of the cell if I change that particular parameter? So if we start with A and alpha, which is to do with mechanosensitive channels, what it tells you, if you have less mechanosensitive channels there, you'll expand more, which makes sense because if the mechanosensitive channel is there, it'll open at a given tension and start recovering that volume. So if you don't have them there, you're going to expand more and you're risking bursting more. There's only a small change in how much you will overshoot for these two parameters. The biggest change in how much you will overshoot is to do with initial osmotic pressure, which makes sense because that overshoot is, most of it is to do with how much pressure loss you encounter during this time that the mechanosensitive channel is open. So if the initial concentration is higher, the osmotic pressure is higher, the cell is pressurized more, then it's more likely to overshoot, overshoot more. So the less it's pressurized, the less they're less overshoot. This threshold value is when the mechanosensitive channel is open, so if they open at a higher threshold value, obviously you will expand, you will expand more and also they will close at that higher value so you will overshoot, overshoot less. This one, these two I will not touch upon because it's a little bit intricate to think about it. This one is the water conductivity coefficient and this one here is E0L, the initial elasticity of the cell wall. Okay, so I have told you that we were able to fit the volume of the cell in black with our model in blue. Here's the volume in time and this works well but I have not shown you the whole trace. So this is not a whole trace, this is the bit that I've shown you here. If I show you the whole trace, you'll see that the volume will go under but then it will recover. There's this part where it recovers. The model doesn't capture that. Model is really passive, right? So we thought, okay, this here could be something to do with active recovery then. And just to remind you, we said when the cell shrinks, it loses just more pressure, it goes below, below the initial volume, it can use some of the pumps to bring the solutes back in and to reinflate that volume. So okay, if this is the case, then when there is nothing available to pump in, we should not see this recovery. And this is what we have done here. So we have this normalized cell volume again and this is now time in 15 minutes. The yellow and black average trace is what I've shown you before. This is average of 50 to 100 cells. Black is the average trace and orange is the standard deviation. So what I'm showing you now is this red curve which is again average of around 30 cells and the standard deviation is a cyan. This is now the same shock magnitude. It goes from lower to higher, but it's in a buffer where there is nothing available for the cell to pump it back in. So if there is nothing available to pump it in, you will not see the recovery from that minimum volume. None of them go back up. So we thought okay, but we can now add just one. We'll add into that buffer, we'll add potassium and if it has potassium, it should be able to pump potassium in and recover the volume and this is this trace now. So again, normalized volume against time. So the cells start at some volume, expand, overshoot, but if there is potassium, you'll see this little kink. This is the same thing, but just one cell. This here is average of about 30 cells. So you see that they do, they do this is the active recovery bit. So if we just put active recovery into the model, we get another parameter. So it's not too surprising and then we can fit the trace that recovers as well. Okay, so at this point I will conclude this part to say that upon a hyposmotic shock, you see a fast initial expansion as expected because this is driven first by the chemical potential of water rushing in to equilibrate the inside and the outside. But the control mechanism of opening these mechanosensitive channels is still passive. So now you'll have a competition between water wanting to go in, be driven in and the solids coming out and each time the solid comes out, less water wants to go in and that takes time. This is why the recovery is so slow and also because it's a passive control mechanism, it can overshoot, it can go below the initial volume, but that's okay because it doesn't have mechanisms to cope with that. It can recover from that scenario. Okay, so the question that we are asking next from here is some of them, one that we'll probably address last is why does it have seven channels in total, right? So it has seven different ones and I already shown you that when you delete the two main ones, the recovery severely changed, right? So they contribute already a lot and each of these seven can have a different number. So if you can control the number, why would you need seven? Maybe just the number of one would do the trick. So that's one question. There does not necessarily need to be a very concrete answer to that. It could be that they were there initially and then you acquired a bigger one and it doesn't cost you, it's not detrimental to have others present so you don't waste it. Evolution will not get rid of it because it doesn't bother you. So it could be as simple as that. You just have them there because at some point you had them, now you no longer need them, but they're not bothering you and even you will not get rid of them. That's potentially the answer, but potentially there's something more intricate, like so you need the smaller one because if you have only bigger ones in certain situations, you'll just lose too many solids and it will not be beneficial. So that's one of the questions, but the simpler ones are, what is the cause of that small recovery? So if you remember, even when we deleted only the two channels, the two main ones, although the trace has been severely changed, less recovery, there was still a little bit of recovery. So the first question is, is that due to the rest of the channels or is it maybe just the bilayer and the membrane, maybe even just the membrane on its own will be able to at that pressure leak some of the solids. And then a different one is, how does this different number of channels of the same type influence the recovery? I can which vary the number of just one type and achieve recovery in all of situations or do we need to have other ones in certain scenarios? So these are the first two questions that we went to ask next. And this was a fun bit that took about half a year where a student was making all, so you can calculate how many permutations you have if you have seven channels. So we only chose few and the ultimate goal was to get down to seven. And although we're lab of mostly physicists, we did go and make these mutations, these seven mutations and our approaches. So in E. coli doing genetic mutations is relatively straightforward. So you have certain two kids and people will tell you, microbiologists will tell you what you need to do. However, they don't fully understand it. There is not an equation that tells you do this, this and this in that scenario and therefore you get your genetic mutation. The reason for this is they don't really understand what are the factors you need to get the expression of that gene, how does it, the machine that makes the protein, how does it best work for that particular gene? So they don't really fully understand that so you don't have an equation that will predict that. So our approaches start with a common know-how and if it doesn't work within two, three months, give up. Because otherwise you can spend a lot of time doing trying all sorts of things and they're not, you don't fully understand it. So there is no end. You could still not be successful after years of doing this. So we try, we give it a certain period of time depending on really how much we care, never longer than a half a year. And then if it doesn't work, we say, okay, this is now something that needs a patience of a microbiologist, I guess. Okay, so we made this deletion seven, took a half a year. Not too bad, it wasn't too painful. The student was crying. She was just frustrated at the point but she wasn't crying. Okay, so this is now that mutant, no mechanosensitive channels at all. This is the volume on time on x-axis. The same type of stuff that I was showing you before. The shock goes from lower to higher. She has only three conditions now. And now you can see that in that mutant it expands more with higher shock but it stays pretty much flat. So even that small recovery is due to the presence of mechanosensitive channels, those other five ones. So they still do something little. Okay, so now we have like a blank state where without mechanosensitive channels in this strain that has none of them, there is no volume recovery, they burst faster. So now we can put a circle, a bit of DNA that allows us to control how many of these channels we have. And think that control fuzzy, right? Like we're talking 50 plus minus 20, 100 plus minus 30, we're not talking 21 plus minus one. So it's a rough control of the expression. And we do have from the model, we have predictions of what should happen with the volume. This is again that normalized volume in time. This is the wild type trace, the black one. So shock, one shock expansion, a little bit of overshoot. Here is the mutant with no channels all the way up, no recovery. And then she's just playing around with a number of channels. And you can see that the shape, it's really not important that you see exactly what each color is. Just to see that the shape changes quite a bit when we'll play around with different number of these channels. So we can then hope that in our volume traces, we will be able to distinguish whether really when we express, when we produce a given number of mechanosensitive channels, it really fits with our theory of what we expect. Okay, so I will stop here to leave time for questions. I'll just show you how the people in the lab look like. Here's the one that you know, sitting in the back. So yeah, and I'll stop here for questions.