 one two one two yeah there we go so last week I told you that bacteria are motile so before I do anything else I will show you some of these different forms of motility so just to remind you motility is just a word used in biology predominantly now in physics as well it means self-propulsion and it's usually a directed motion towards chemicals so it's a way or towards a chemical that bacteria might or likes or dislikes but it doesn't have to be just chemical it can be towards temperatures or it can also be towards oxygen so there are different ways of doing this and we have been looking and I'll also be going to talking mostly about flagellated movement so that's either with several flagella or with one flagellum but you can also have actual filaments that help the movement you can have gliding and you can have twitching motility there's still some motilities with some bacteria they're difficult to grow or difficult to do genetics where we don't really know how they're how they're motile so to start off where I'm just going to entertain you and show you some of the movies so this one here is E. coli and the cell body and the cell filaments have been fluorescently labeled by Harberg at Harvard in his group what you will see is how they move so you have a filament that pushes the cell and then one of the filaments falls out and you can see that the cell turns a little bit so that's E. coli there's another flagellated bacterium this one is called rotobacter spheroid it has just one filament and this one will stop and start so you'll see the filament pushing the cell forward and then you see the guys this one will stop there we go and then after a while it keeps going these guys that I'm showing you next have two flagella bundles and either end and they look quite different still motile but quite quite different also you can have gliding this is mix of gliding mix of Agassanta is gliding on a surface you see that these bacteria are not swimming but they're still motile they're gliding one other bacterium this one is called pseudomonas aeruginosa twitches so it's going to extend a filament type of structure and it's going to pull itself forward so this video is quite short so have a look there it is and then you pull pulls it up so I'm going to go back to E. coli I showed you what it does in 3d but it 2d that swimming looks slightly different and it's called swarming so this is now E. coli again with several filaments but this looks quite different and it's called swarming there's quite a quite a bit of research going into the swarming mobility close to close to surface E. coli is not the only one that people are looking at the surface there is another one called Vibrio and this one here is again swarming close to the surface and and this here is one bacterium so what it does somehow when it when it senses that it's close to the surface it stops dividing and forms this gigantic one long bacterium that forms a spiral and you can actually study how it goes into the spiral and how it comes out of the spiral so this is just to give you an idea of different types of motility but I'm going to swing back to E. coli now this is the one that we have been looking at the hands-on session so this is an illustration of how the propulsion works so you have a cell body this is roughly 1 to 3 microns long and then you have about 6 and we say about 6 because if you look at different literature sources you'll find 3 you'll find 8 you'll find 6 and each bacterium has a slightly different number but let's go for 6 because it's nice and round half a dozen of these filaments spaced around the cell body randomly and they form a bundle at the back so this bundle then rotates and propels the cell forward this is what generates the the motion at the end of each of these filaments there is a motor structure called bacterial flagella motor and this is a very impressive motor structure it's about 15 nanometers in size and it spans the whole of the cell envelope so in the case of E. coli cell envelope consists of two lipid bilayers so they're called outer membrane and inner membrane and there's also a structure called cell wall and this structure is a stress stiffening type of material so it's a stiffer material and the more the cell gets pressurized the the stiffer it gets so the motor spans all across it's 15 nanometers in size and it does have a rotor and it does have a stator part of the motor so the rotor part it's right here and this will rotate pushed by stator units so stator units use something called protomotive force although it's not actually a force it's an electrochemical potential so the cell is charged inside outside inside is more negative and then when the ions protons are passing through these stator units they're giving energy to generate the movement to push the rotor so this rotor then it's coupled to a structure called hook which is around 50 nanometers and it's quite sharply controlled the size is quite sharply controlled it is really 50 nanometers and this one is a universal joint it's very flexible which is why allows it to bend all of those filaments that are in the middle of the cell and then they connect to the filament which can be three to ten microns long and this is a little bit of a stiffer stiffer structure so what is quite important is that a motor in a call I can rotate both ways it can go clockwise and counterclockwise most of the time it spends rotating counterclockwise so the cell swims forward because the filament is formed every soften every 10 to 20 seconds for about 0.1 second the motor goes the other way one or two and when that happens you can imagine you have these filaments rotating one way and a few motors start going the other way and they disrupt the bundle and the bundle force apart and then the cell is doing Brownian movement so during that period there is a good likelihood that when all the motors go back into the same direction the cell is going to be going in a different direction so it does that normally just does a Brownian Brownian walk so it goes for 10 to 20 seconds roughly straight and then 0.1 second of wobbling about and then continues 10 to 20 seconds when it wants to when it wants to navigate and it plows something called hematactic network and I'll take you through it but you can think about this network is just a series of feedback circuits then transmit the signal from the outside to the motor itself the motors the output of the network the input here is a chemical I am presenting is as a square because all you need to know about this chemical right now is that it either likes it or dislikes it so when that chemical is present in binds to the gray proteins these are the receptors and then these gray proteins send the signal down the network there's few of these proteins that you see colored and they really are just feedback they're just controlling one another the only one that I'd like you to remember is the green protein this one is called key why when it gets phosphorylated right at the end of the network output this green key why in a phosphorylated form sits down onto the motor and that binding changes the direction of the motor so when you have a signal from outside what will you do to direct the movement of the bacterium you will change the frequency with which the bundle is disrupted I the motor goes the other way right so if you like something you're swimming 10 to 20 seconds straight and you don't really want to have that wobble so that you go away from what you like and then you will suppress the wobble you will change the rotation of the motor less frequently if you dislike something you will increase the frequency of these wobbles so that you have a hard likelihood of turning and going the other way right however to efficiently move along the gradient of things that you like or away from a gradient of things that you don't like you need to be able to do one more thing and this is in chemo taxis called perfect adaptation it just means that you reset the network base back to its original steady state where by steady state I mean you swim for about 10 to 20 seconds and then you occasionally wobble and you swim 10 to 20 seconds when you encounter something you don't you like or dislike you change the frequency of these wobbles but very soon you need to go back and this is illustrated in this experiment so on the y-axis I'm showing you the bias the amount of time cell goes straight and on the x-axis is the time in seconds so around five seconds you see that the cell encounters a pulse of something that it likes what it does it starts swimming straight for longer so it suppresses these stumbles but very quickly within few seconds about five seconds it resets back to the same value so now when it encounters something else it can still sense it so that's one important characteristic of the information processing of this network the other one is that it's very sensitive so what I'm showing you here is the bias this time the bias how what is the free these are previous experiments so I have not chosen to switch the bias now so in this axis is the probability of tumbling rather than swimming straight so how many times does the motor switch and here is the concentration of the green guy key y in a phosphorylated form which is responsible for generating this is the protein responsible for generating the switches of the motor and what you can see that the response the bias changes very very sharply so hill function there with a few micromolar concentration difference so it responds really finely to micromolar concentrations it's a very sensitive sensitive network what I'll also tell you is that E. coli is pressurized so there is about a bike tire of pressure inside the cell and there's quite a bit of research of understanding what exactly does that pressure do how does it how does the cell uses this force against its surface area in order to grow and and and survive so inside again just to remind you it is separated from the outside by some semi permeable membrane in the case of E. coli these are two lipid bilayers one on the inside the other one on the outside and in the middle is this cell wall which is a stiffer material inside there is a higher concentration compared to the outside normally and this generates a smaller pressure so the cell is pressurized about a bike tire of pressure when the concentration outside changes say that for example it increases what happens and it needs to before I go on it needs to actively maintain that pressure obviously because it has a semi permeable membrane when the external concentration changes say that for example it increases first thing that happens is just physics so what wants to happen is that the inside and outside equilibrate and that is what happens so the cell shrinks and it can shrink up to 50% and then the inside concentration and outside concentration are the same the other way around when the external concentration decreases the water rushes in and the purpose is the same to equilibrate the inside and the outside and then the cell expands and risks potentially bursting so in both of these scenarios you lost 50% of your volume you shrunk or you're about to burst to expand it E. coli doesn't appreciate it very much so it does have mechanisms to cope so in the case of this shrinking so now it's shrunk and in order to recover pressure and recover volume what it starts doing it employs these pumps and these pumps pump stuff in and I'm really going to stay at stuff here there are ions there are some organic osmolytes but it really matters that there are molecules that go in as they go in the water goes in because it's equilibrating and at some point it starts pushing against the cell wall and at that point of volume no longer expands by the same amount of the stuff that comes in and you start repressurizing the cell the other way around now the cell has expanded because we decreased the external concentration and it's about to burst what it has are these mechanosensitive channels these are like rs-like structures that open up a hole for the solutes as well so now the membrane is semi-permeable not just the water but also to all the other solutes and these solutes are going to come out and then the water is going to come out and the cell cell volume recovers so what we the reason I'm telling you this is because these shape changes can happen when the equalized swimming as well but before I pose that question I'll show you how some of these shrinking behavior looks like so top one we have increased the external concentration by adding some sucrose in red we're marking the cell wall we're marking that stiffer material in green we're marking the inside of the cell just the inner compartment so because we have two lipid bilayers we effectively have two compartments the inside compartment and then this around ring compartment and the outside so what happens when you increase this concentration you see that the cell shrinks and you can even see detachment on on on the sides if we increase it with something really big like dextran that cannot pass through this outer membrane the whole thing shrinks and it shrinks quite a bit so you can see that the material the red the stiffer material is actually not that stiff when it's not pressurized it does exhibit stress stiffening so it gets stiffer when the cell is more pressurized when we increase it with sodium chloride you see that the peeling off happens straight away and then you lost about 40 percent of the water and the cell wall actually stays unchanged and this is real real time so it is really really fast this fast initial response is really fast it's just it's just equilibrium inside and the outside so what we wanted to ask is what happens when you do the two together so you're swimming but you're also experiencing these osmotic changes and this is something that will be happening in the gut and i'll motivate it a little bit more in a second but there is previous literature that suggests that something interesting does happen so if i show you this plate so what is happening in this plate is that you have a plug of stiffer agar with a high concentration of something on the top one that high concentration is of something that E. coli does not care about it it's in this case it's ribotal but it can be sucrose E. coli doesn't care about sucrose it's not like us it likes glucose but it doesn't like sucrose so at the bottom is a high concentration of something that it does like the rest of this plate is agar of a smaller concentration so E. coli can swim through it here in this heavier concentration higher concentration of agar plug it cannot so this plug creates a gradient of high osmolarity downwards and what you can see this black ring around it where E. coli does not reach so there is a negative form of taxes it swims away from this very high concentration even when this high concentration is of something that it likes because at smaller concentration of something that likes you'll see accumulation of of cells there is previous indication that these high osmolarities even for molecules that it does not care about in terms of their neither their attractants like neither food or not something toxic it will still exhibit a form of taxes there's a little bit more evidence that this signal could travel this high osmotic signal could travel down the network in this experiment so what i'm showing you here is this receptor molecule the molecule that normally sends this chemicals that E. coli either likes or dislikes and what in this experiment has been done is that there was a fluorescent protein put in the place where normally the chemical that E. coli likes or dislikes would sit and then what you look at you look at the amount of polarization and the difference between two polarization so have parallel and perpendicular the difference and then you then then you normalize it so when there is a lot of wobble this number is going to be higher when there is less wobble it's going to be is going to be smaller and what happens in the experiment is if you look at the red one here r on the y-axis is this number it's essentially the amount of wobble in this protein and on x-axis it is 9 so what you see that when you add that sodium chloride sucrose or dextran those that i showed you there is quite a bit of shrinking in the cell there is less wobble so the the when the membrane shrinks it could mimic the binding of some kind of chemical on top here you have just a fluorescent protein there is present in the cytoplasm it's not on the side of the cell and there is no difference there is no difference in the signal i should also say that these guys here in harvard bergs lab have looked at whether the signal could possibly travel down the network and they found that there is indication that it could but i'm not going to go into details of that just now because the experiment is slightly more complicated okay so there is an indication that a small exchange change in a small pressure of the cell could go and travel down the network but i also want to motivate you why would this why would this be interesting so normally the chemo-tacting network the sensitive chemicals things that buckler like or dislike is done in a very very limited environment so there is not much else it's just a buffered solution with micromolar concentration of things that bacteria like or dislike but this is an indication of what might happen in your gut so after a steak meal what they have measured here is the osmolarity in your stomach and your small intestine and when you can see that it goes to about 200 300 osmol this osmol is the amount of like for example something that is 150 millimolar concentration will be around 300 osmol so it's the measure of the the the amount of chemicals present and the osmolarity can be increased even with things that E. coli does not like to eat or it's not afraid of so this could be just very passive molecules yet the osmolarity is higher same thing happens after a bit more sugary meals so here's a donor milk meal and again on y-axis is the osmolarity and on x-axis is your intestine essentially so we're traveling for stomach to small intestine and here's slightly different you go from quite high osmolarity and then you drop it still stays relatively high as you travel down the intestine so these kind of shocks these kind of changes in osmolarity E. coli could encounter while it's swimming through our gut and potentially trying to trying to infect so we wanted to know what happens when you have the two together how does the network actually respond when its experiences increase in the osmolarity and here's how the experiment looks like so what we do we attach the cell onto the coverslip surface so the cell is now stationary we cut the filament and then onto the filament we attach a plastic bead a polystyrene bead index of refraction around 1.45 and that bead can be of different sizes in this particular experiment we use 0.5 micron bead but because that's a little bit small to see what I'm going to show you how the experiment looks like when we attach a 1 micron bead so this is a real life experiment you'll see when it when I play it you'll see bacteria on the surface I'll be going in and out of Z so you'll see some of them on the surface and then you'll see a bunch of beads and what we're doing there we're looking for a good spinner we're looking for something that rotates quite well and then we place it in an optical trap and I'll take you through Y in a second so here's how the experiment looks like so you see some of these rotation and you see the cells on the surface so now you're passing through the slide and it really is a single motor experiment you're looking for one individual motor that you will do the experiment on and the ones that are rotating quite a bit are probably the ones that we're not going to record we record those that you can hardly see your eyes get used to the little wobble that could be a good spinner because we want the filament to be as short as possible so that the drag on the motor comes mostly mostly from the bead so we want really really short filaments okay so then we put the once we found a good spinner we put it into a heavily attenuated optical trap and this is how the system looks like so this part here is just a bright field illumination that gets sent to the condenser and the objective and then down here we have a set of cameras we have an em ccd uh stemos we don't have an eyepiece we just have a what we call a crappy camera it's just a ccd standard ccd that gets onto a monitor and this part here is for fluorescence that we're not going to use in this experiment but what forms a trap is this laser here 855 nanomineral laser that is just expanded so it overfills the back aperture of the objective and is then sharply focused and forms an optical trap it's sent after the condenser so the back focal plane of the condenser is imaged on the position sensitive detector or a quadrant photodiode the reason for that is that in the back focal plane of the condenser you have a Fourier transform of the image so if you image that onto position sensitive detector you're sensing interference patterns of what's happening in the in the image plane this is how the microscope looks like real life so here's the condenser and an objective this part here is where the where the position sensitive detector sits and here is our laser and to show that we really just have one telescope i'm zooming in here here's the laser here are the two lenses that expand the beam and then send it onto the objective and right there is our position sensitive detector so the way you can think about it so once we put our bead into this optical trap that is heavily attenuated is not trapping it's just really sensing the rotation of the bead the experiment looks a little bit like this so the signal that we're going to be detecting here is our flagella motor here is a very short filament we're going to attach a bead and this particular cartoon the bead is a lot smaller than our rb is a lot bigger compared to the cell the bubbles are ions artist impression of ions so the the bead rotates and then you put that you image this the back focal plane the Fourier transform of this rotation and the laser onto the quadrant diode and you can think about it as sensing different amounts of light falling on the on the diode like so although that's not exactly correct it is an interference pattern it's not about shades and then you get a signal you get a speed of the motor one way and speed of the motor the other way right so here's a signal that you display real life you convert that into the speed and you get speed one way speed the other way and now we're interesting to know how the motor is going to respond in terms of changing the frequency of rotation which sets the direction of movement we need the amount of time spent rotating one way over the amount of time spent uh rotating the other way and that's what we call bias so it's the amount of time spent going the other way where the tumble occurs over the amount of the total amount of time and the experiment itself when nierco does the experiment this is what you see so here's the speed of one individual motor and here's time in minutes and this is a representative trace so you see that the speed before any kind of stimuli is around 80 hertz and you see that occasionally the motor goes the other way so about 10 seconds swim roughly straight and then occasionally for 0.1 second go the other way this blue line is where quite high in this case osmotic shock a step function comes in what you can see is that the motor stops spinning and then increases the speed and in this period there is no changing of motor rotation and then it ends up there's this lots of switching happening it ends up at a steady state these are minutes so it's for quite a long time where it just wobbles a lot more and it does not recover back to the initial level this is also in this histogram so this is now the histogram of bias the amount of time spent this way over the total amount of time and you can see that after the shock which is here the minutes are the same you see a lot more switching events and the map here the color map is just condensed histogram it's just 1d histogram because what i want to show you i want to show you quite a lot of these cells i come back to this in a second so these are now these 1d maps of biases for a lot of cells each line is one cell here is a shock of 100 millimolar then 200 millimolar and 400 they're getting higher and higher what you can see that before the shock which is a yellow line on x axis is your time at minutes that before the shock the bias is relatively low so it goes straight most of time then switches a little bit and then after the shock you see a period increasing in length with the magnitude of the shock where there is no switching of the direction and then there is a steady state increase in the amount of time that the cell switches so it does not go it doesn't adapt it does not go back what i want to show you is also this one here where we comparing this response of the cell without yes that's a little bit what it does thing here i'll call mark after to demonstrate that yeah yeah i'll discuss what it might so actually in this period of time it's 10 to 20 minutes where the volume does recover so we keep it in a buffer where we allow it to to recover recover the volume so here is a cell that i showed you this is the one type cell that just means it's it's we haven't genetically touched it and this one here is we deleted that green guy the one that sits on the motor just that protein so we call it delta key y without that protein and when you see without that protein here's the speed of the motor on y axis and the time on x axis in minutes again is that you have the speed response pretty much the same you're just lacking you're just lacking this switching events that does not necessarily mean that the network is involved in generating this behavior it just means that the key y protein binding is responsible for it and i'll touch upon that in a second okay so here are the speed changes so again putting the same number of cells this is one shock magnitude higher and higher on the y axis is the speed in hertz here's the color color scale and then on the x axis you have time again same time periods exact same cells so you can see that after the relatively high shocks there is a bit of an increase in the speed but also for very high shocks there's a drop in speed and then increase increase in the speed right so we have two things that we have not normally seen in a chemo attacking network this is that the bias increases and it does not go back to the same value so there is no perfect adaptation it's a steady state increase in bias and also the speed can change so there are two type of questions that we can ask number one is how is this actually happen on a molecular level within the cell so we left that aside for now and the other question is can this generate taxes can with this type of behavior can we explain what we see and that was if you remember that swimming away from a plug of high concentration and here i'm presenting it here so if this is a high osmolarity there will be less and less cells as you as you go out so to answer this question whether this type of behavior can generate taxes i'm splitting the the the way the response is generated in two possible ways so number one is that there is no information processing right so as soon as you step into the higher osmolarity environment your bias and your speed can change at that position and that could happen even without any kind of sensing because for example at higher osmolarity there is more stuff inside so the pressure inside is actually higher which could make binding to the motor and switching easier it could just change the binding constants and therefore as you step you actually straight away change the bias and change the motor speed so then the question here is can the change in the diffusion constant give you a tactic negative type of behavior so to answer that question we need to think about what is the diffusion constant on a microscopic level right so if you remember when you're going from a brownian walker into a continuous limit where your step size is getting smaller and the time period between your random steps is getting smaller as well that you find a solution only if the ratio of the two is a constant then you have this macroscopic diffusion diffusion constant so to answer the question of whether or not a spatially dependent diffusion constant will give you taxes you need to think a little bit about these microscopic steps and time periods in between so if your speed it could change in space but it does stay the ratio of the two stays the same then your equilibrium concentration will not change but if you either keeping the step constant or you're keeping the time constant or you're changing all of them you will get negative taxes also if you're changing all of them you can no longer as easily find the diffusion constant you need to be a little bit more careful here you can write you can write the diffusion equation but the diffusion constant will not be as easily defined that is if you are only changing your you're changing your bias and your speed in in space without any information processing the other scenario that the network is involved and there is information processing so now it goes the signal travels down down the network there you need to be a little bit more careful so now the most basic question is can you get a negative type of taxes in the scenario where your Brownian Walker now it really depends on where it came from right because it's going to be sensing a signal across a certain time period so it matters where I came from because I was sensing for a given period of time before I came to this spot so it's now it's now becoming an anomalous diffusion it's no longer a straightforward diffusion so in that case because of this dependency on where you came from you will get negative taxes if you exhibit no adaptation what happens with the speed with the speed changes on top of that that that's where it gets more more complicated so we're thinking that what we observe this increase in a steady state bias so no adaptation and a change of speed in both of these scenarios can give negative taxes exactly how the swimming behavior will look like and how will they now behave where there's high osmolarity gradient or a chemical that they like or this like at the same time as high osmolarity gradient which is most likely the gut scenario that's the next the the next thing to do so I will leave with the conclusions so unlike the hema taxes osma taxes exhibit an elevated steady state bias so there is no perfect adaptation you never adapt you stay at a higher bias a response can include the final change in motor speed and that really depends on how sharp the gradient is as you swimming through it so a transient response to a very sharp gradient can involve more complicated things like the drop of speed and then recovery of the speed with no switching events and this osmotic response can lead to negative type of taxes in an osmotic gradient like they have observed in a plate but obviously the it can be more complicated depending on exactly how the gradient how the gradient looks like and here's where we are of the people in the group these are the students and the postdocs so the one that has done the work that I've talked to you about you met this is Jericho and this is how they all look like and I will stop here to leave time for questions