 Ja, het mooie deel in de wereld en deze mooie conferentie, ik geniet er veel van aan het doen. Ik ben een biophysicist, ik werk meestal op soft-matterfysieken en ik probeer er wat op te doen op biophysieken. En ik ga vandaag meestal over wat werk ik in het gebouw heb gevolgen met tijdens een sabbatical leave in Keltaque, terwijl ik met oud-vrienden van UCLA een collaboratie heb. Dus het officieel titel van Baitalk is een dynamisch of in vivo bacteriopage ejectie, genome ejectie. En de populaire titel is hoe je in de kruid gaat. Want ik zal je laten zien dat in vivo ejectie van een genome, van een vijf genome in een live bacteriopage. Er is een short term, dus er is een short range behaviour, het is gedreven door de pressie in de vege, maar er is een lange tijd behaviour ook. En dat behaviour heeft te doen met de groepen van de cytoplasm. En die groepen is het meestal kruidendes, zoals we in een minuut laten zien. Dus dat is waar de populaire titel komt om hoe je in de kruid gaat. Alright, dus dit is de kruid. Dit is de kruid ik ben werkenen met. Missie Rue, ze was een groot student wanneer ik werkte in Keltaque. En Rob Phillips, ik denk dat hij veel van jullie wel is, wrote a beautiful book about physical biology. Deze zijn twee oud-vrienden. Hij ging terug, meer dan 20 jaar geleden, wanneer ik een postdoc in UCLA was. En hij was ook aan de virus werken, dus dat was echt leuk. En een inspirerende collaboratie. Maar om te beginnen, feestjes, waarschijnlijk zijn de meeste van jullie ergens, dus ik zal jullie een introductie geven. Het is ook mijn probleem om je te houden, na de lunch. Ik hoop dat ik je wil houden, om je te houden met iets wat je al weet. Maar ik wil jullie een introductie geven. Het is altijd goed, hè? Hier is een feestje, een proteinecapsitie die genetische materiaal heeft, het genome, het ligt op de bovenkant of het stem van een bacteriën, waarin het bindt op typische proteinen of sugars op deze stemmen. En dan ejecteert het genome. Dus hier is een scherpke foto, hier is een reale foto. En de lifecycle is eigenlijk, je hebt deze feestjes die lopen op het bacteriën, ze injecten hun DNA, het bacteriën, gebruiken de reproductie, de machinery van het bacteriën. Nieuwe protein, feestproteines worden genererd, nieuwe feestjes, en je krijgt nieuwe feestjes. Op een gegeven moment, de bacteriën ligt of explode en de cycle is klaar. Nu, Dennis zal zeggen, in de volgende seizoen, zal de DNA weer in deze feestjes zijn. En nu gaan we over hoe het gaat uit, oké? Dus, waarschijnlijk, dit is een historieke ding, het is een hele bepaalde experiment. Ik denk dat, wanneer ik het niet bewijs was, tot een paar jaar geleden, de Hershey Chase experiment, dat was een van de experimenten die geholpen hebben, dat in effect DNA een genetische materiaal is, een genetische materiaal. En het is een heel smarte experiment. Een lange tijd geleden, in 1952, about a year before Watson & Crick wrote, they published their paper on the alpha helix structure of DNA, but still at that time there was still some debate whether DNA constitutes really the genetic material of organisms. And the argument, the counter argument of that, was in fact that DNA is so simple. It couldn't code for these complications that organisms display. Not even bacteria, but particularly higher organisms. But this is typically a proof, I think a very nice and elegant proof. So they radiactively labeled the sulfur groups in the proteins of these phages. And they also radiactively labeled the phosphorid bits in the DNA. Then they could show that the DNA is in fact in the bacterium and the proteins outside the bacterium. Dat was a really additional confirmation that DNA is really the heritair material and not protein. I'll get back to the Hershey chase experiment in a minute. I will talk about the easy part. How DNA gets out. Initially it's pretty easy for DNA to get out, but because it's highly pressurized. It's wrapped inside the phage head. We have, for example, a delta DNA. The contour length of 50,000 base pairs, 50 kilo base pairs of DNA is about 17 micron. Then these phage heads are about 30 nanometers. So it's like if you have a cable and you spin it over the golden gate bridge, you try to wrap it into a van. And it's not the very thin cable. So there's a lot of pressure involved in stacking of having DNA packed into these phage heads by 56 atmosphere. So this is a robotically lysed phage. And here you see all the DNA that's supposed to be in the phage head. So it's enormous disparity of length scales in that sense. So this is what you can call a single molecule Hershey chase experiment. So this is stuff we'll be mainly talking about. The question we addressed or what Rob and his collaborators already addressed is what the mechanism is of phage DNA translocation. So it is proof of Hershey chase indeed that DNA is the carrier of genetic information. Now the question is how does it work? What is the mechanism? Can we quantitatively describe that? What are the time scales, for example? And I think the experiment that they did is really elegant and smart. So what they did is they labeled phage DNA with so-called cytox orange and that's dye that cannot penetrate the bacterial membrane. So the inside of the bacterium only gets fluorescent if DNA penetrates the membrane by other means, right? It's not a spontaneous process. So at the beginning of phage infection there will be all the fluorescence will be in the phage of the bacterium. En on a later stage the bacterium will continuously become more fluorescent and the bacteria and the phage head will become less fluorescent. So this is a grammatical picture of how that's supposed to go. Yeah, so... And then a priori. So for some people this is probably a familiar picture, right? So a priori you could think about how this affects influence the dynamics of phage-action over the speed of how DNA is translocated from a phage to a bacterium. So of course it's a stored energy in the... So it's like the very closely packed DNA that constitutes a huge driving force for ejection and I think it's safe to say that in vivo ejection is the main driving force, right? It's almost all there is. En then of course there are other things that are the properties of the solvent or the cytoplasm like the osmotic pressure, other macromolecules, salt, you know, divalent ions, you name it. All these things are important. And there's also something... Potentially there's also something else and that's so called regiting proteins. So the idea is that at the first stage of DNA ejection the process is relatively fast even by the internal pressure. But then after a while when the driving force becomes comparable to the turgor pressure in a bacterium it gets slower. And the idea is that there are so called regiting proteins that very strongly bind to the DNA that comes out of the phage so that it doesn't get back. So it rectifies in fact the motion. I'm not going to defend it in detail because this is a spoiler but we're going to show that we'll at least let the phage but they could be there. This is one of the aspects that could be important in the phage ejection or in DNA ejection by phages. So this is a typical and I have to learn to show it. So this is a typical movie if it worked, yes it does. So it's a little fast but you see that in the beginning in the beginning all the fluorescence in the phage had so this is a phage that sits onto a bacterium here's the bacterium so all the fluorescence in the phage had, there's no fluorescence here and then after a while in order of minutes there's a redistribution of fluorescence and you can tell here that it's on the order of minutes so let's say in 5 minutes most of the DNA is out there's also a lot of noise but that's a different story so schematically it looks like this so at first all the fluorescence in the phage had and then gradually there's a redistribution of fluorescence that coincides with the redistribution of the DNA and these are some typical stills now in vitro the situation is much faster and I'm going to show you that as well because here you see it really goes over let me first stop what the idea of the what the what I'm actually showing you so the setup is that you have phages adsorbed onto a microscope cover slips and then there's a flow of gold particles that stick to DNA so they're being visualized the idea of the flow is that it kind of stretches the DNA that comes out of these phages and this lamb B initiates let's say the uncorking of the phage so it binds to the it's like a receptor that binds to the phage tail and that induces ejection of the DNA so that's the setup and here are some typical movies this is a typical movie you see it goes very fast I'm inviting you to look over here because here you see how it triggers and how it here goes and then so it's like seconds right in vivo it's extremely fast so there's a huge difference or huge well there's at least a magnitude difference in timescale of the ejection process and there's more I will show you later there's some more pictures over here just because I like them so here are some stills of the process you can see how it goes eventually in time so this is seconds so it's relatively as I will show you later this is fast compared to I'll show you here it's fast compared to in vivo experiments and also there are differences because depending on the the kind of ions in the in the evitro system it goes faster or slower so it's very sensitive to the counter ions and a lot of other things alright so this is I think this is a really very beautiful and elegant way to get grip on what the driving force really is and it's using osmotic pressure I'm a physical chemist by trading so I appreciate the smartness of this setup so what these authors did I think many of them are here right Alex is here and Bill is here Avi je was also involved in that in the early stage so the idea is to have a so they call it the osmolite that's a relatively long polymer that cannot penetrate the phage tail so it exerts effectively an osmotic pressure and by playing with the osmotic pressure of the osmolites these people were able to to stop ejection at some point so it stops at the point where the pressures are equal and one of these pressures is being set by the concentration of the osmolites and by that you could where the ejection stops that's a measure in fact for the pressure inside these phage heads I think that's a beautiful experiment now back to where I really want to talk about back to in vivo it's more complicated this is an experiment of two mutants of lepta phage with E. coli a long chain length the blue ones, short chain length and the red ones and this is the ejection velocity so kilobase per per minute as a function of the DNA that's inside the capsid and a priori the idea was that the DNA inside the capsid determines the driving force and that principle determines the ejection speed the curves have the same shape but they're different curves because they're different DNA now if you look carefully so the first thing to think about is what the relevant quantity or the relevant parameter is to study the relevant description and if you look at these curves you can tell that if you translate that red one about 10 kilobase per that's about the difference between these two mutants they coincide now if you do that typically the difference between these two mutants that means that the starting point also coincide and starting point here means here's DNA in the capsid so here's no DNA in the bacterium so the idea is if you translate it first you shift the curves and then you flip it you will get universal for two curves behaviour if you take the parameter DNA ejected right so here's how it looks like so you shift it and flip it so then you plot the ejected DNA rather than the DNA in the capsid and then you see that on a relatively long time these curves coincide at the beginning there's a difference and that makes sense because that larger genome of course exerts a larger osmotic pressure so it's faster but in the longer run it was unexpected so that means that since in fact the relevant parameter at least experimentally is the properties of the ejected DNA as well as the properties of the bacterial cytoplasm that set the time scale for ejection here in these relatively long times alright so and if you look at that carefully there are two regimes right so there's one regime we call that pressure driven I'll get back to that later why I call it that so that's different between these two mutants and there's another regime we call that late stage and I'll tell you where it comes from in a minute so people are know a bit about glassy dynamics they know where short stage en late stage come from but I'll tell you that in a minute so you see that it becomes similar you can also plot it differently you can just plot the the number of ejected base pairs is a function of time take the logarithm for both and then you clearly see two regimes there's one short time regime and it has a slope of one which means that it is it's a driven process and velocity more or less not exactly but more or less and then here there's another slope and that slope turns out to be 0.2 so one of the things we try to do is to find out to come up with a mechanism that explains the value of that slope on the long times alright so let's start with the early ejections well many talk about the late stage but let's we know about the early ejection stage the idea is that at the early ejection stage when the DNA is closely packed into the phage heads and they have a typical separation between the DNA strands of ds of s then the idea is that at these short distances is mainly hydration force that determines the interaction between the particles so here and that's exponential that's exponential there's also electrostatics that's also pretty important but at this point that contribution is small compared to the bending energy bending energy is also large on an absolute scale but it's small compared to the hydration force at least close as these very close packings then what you do so this is typically the force per unit area so if you integrate that over the typical distance to the typical separation you will get the energy that is stored into the system and then if you take a model at least a geometrical model for how DNA is packed and people all agree that it's hexagonally packed you can get the force just by differentiation and it turns out that the force is an exponential force in the length of DNA and it's not directly easy to see increasing function of the DNA length cause it's minus 1 over the square root of L DNA but it's an increasing function of the DNA length it's weakly increasing compared to a real exponent but it's weakly increasing and then of course there's also a friction so there's also a friction contribution and that's typically a friction coefficient times the velocity and we could calculate the velocity just by by force balance and that the friction factor that's also turns out that's also be exponential in the length of DNA but with a slightly different coefficient than the driving force there for completeness there should also be a stokes contribution cause you could think of the blob of DNA being ejected that also has a friction effect but at this point that we ignore that it turns out to be a pretty small effect I'll get back to that later cause in lake stage it becomes very important alright, so you get the injection speed but you do compare to so this is all known this all worked by Bill and A-legs and a lot of other people and the only thing that we did is we add or we subtract the turcopressor of the bacterium and then you get a pretty good agreement for the short stage, for the early stage injection it's not totally perfect if you choose the quantities that are in the theory for reasonable values it more or less qualitatively describes what's going on so this is the ejection DNA and this is the speed and it goes to a maximum that's because of the competition between so the friction factor decays faster than the driving force so that's the reason why you have a maximum here so that looks reasonable like let's say acceptable in terms of the simpleness of the model but there's a big problem and the big problem is that at some point when almost not even, well a bit more than half of the DNA is being ejected there's no driving force so purely mechanically the process should stop at that point but still it goes on and that has to do with the two so that's consistent with these two length scales they show that first the fast regime and the slower regime back to the question that I'm addressing now so we understand more or less the fast process now let's get to the late stage process and what I'm assuming is that it has to do with the dynamics or with the diffusion dynamics of a blob of DNA the ejected blob of DNA of cytoplasm we've already seen experimentally that at some point the relevant parameter experimentally the relevant parameter is the amount of ejected DNA inside the bacterium so the only quantity that may go into this description is the properties of the DNA as well as the properties of the cytoplasm nothing else so that's what we assume so we need to know what approximately the hydrodynamic radius is because we want to calculate at some point whether it is diffusion coefficient or friction coefficient we know it's not simple self diffusion a priori because the pretty large blobs of DNA at some point I will show you that in a minute and we know that the cell cytoplasm is classy it's a crowded environment and we start simple by just assuming that there are no ratcheting proteins so this is the idea so there's a blob of DNAs already out and what we are pursuing is that the dynamic properties or the diffusion of these blobs determines the ejection speed so what we need to know the ejected DNA in a number of base pairs we call it n the function of t then we need to know what the hydrodynamic radius of that blob is which depends on n, it depends on t and then we need to calculate what the diffusion coefficient is and it depends on n, it depends on t and then you need to close to the story well first of all about the size this is more or less a sort of proof but it's more like an argument by reductio at absurdum I think that's the most beautiful argument in mathematics personally so it's not stretched because we know that the size of a bacterium is about 1 micron and we know that the contour length of DNA is 17 micron no deal also it's not just purely a random log or it's not purely a self-avoiding chain self-avoiding chain the hydrodynamic radius is given by the so this is the persistent length times the number of base pairs to the power gamma gamma is about 3 over 5 that will give you for that typical DNA about 1.5 micron no deal either so it should be more condensed than either of those two and what I'm just assuming simply is just like it's close packed it's close packed on the scale of the persistent length and then you get something reasonable that's about 300 nanometers so it could be that it's even closer than that but then you have to bend it a little bit but I don't think that's not necessarily problematic it could be a little bit less closed it's also fine I will show you later on at the end that the result is not very sensitive to gamma so that's good because this is kind of a gas and you could have that smaller the denser state just by the so-called nucleic associating proteins we already heard about that rapid proteins whatever so now about the cytoplasm so I think this was you can say that if you are a large maker molecule and you are in a cell cytoplasm what you may call diffusionally challenged because diffusion is not easy if you are in such a highly viscous crowded environment and this has been shown not too long ago by this group is Christine Jacob-Parker and Eric Newfrain and they showed that relatively large maker molecules say larger than 30 nanometer if you are a typical protein you are fine in the cell cytoplasm it is self diffusion within a couple of seconds it is relatively fast but if you are larger than 30 nanometers you have a problem you see this is the mean square displacement this is time it is not linear even over more than an hour it is sub diffusional here you see the same thing they effect show that if you switch off the metabolism completely so you can do that by some chemicals and you stop all the ATP hydrolysis you can just temporarily put the bacterium sleep so to speak then everything stops so at least for these large maker molecules it is a typical glass so there is no movement at all at that point so you may say that just summarizing anomalous diffusion anomalous diffusion means a slope smaller than 1 if you plot the mean square displacement it is a function of time when the cell metabolically active in a rest when it is inactive so this was something I wanted to share with you because I like that I understood that there are also students here there is something you want to hear so this is just in writing what all scientists know that is if you start studying something you usually find something else or it doesn't work usually the other thing that you find is often much more interesting than what you were really looking for these guys just put it in writing because they start their paper that study began with something else and now I am going to talk about what they actually discovered so I like that so I just wanted to share that with you typical solid fraction we have about 3 million proteins in E. coli their volume is typically about 100 cubic nanometers so your volume fraction is about 0.3 and then I don't count the hydration layers and all these kind of things if you count that you may be close to about a half and we know that for example studies of hard sphere hard colloids that behave as hard spheres at about the volume fraction of a half things really become slow and that is probably what is going on here as well so this is some other so we just checked some other relatively large molecules so this is the logarithm of the mean square displacement this is the logarithm of time so this is self diffusion so these are some typical so these are plasmids this is lab.dna here and there are some other proteins relatively large proteins here and for all these components these slopes are pretty close to about half so it's self diffusion with a coefficient of about half it's something that I don't understand it's something that we apparently see I will get back to that in a minute so the idea of that sub diffusion is that for example here if you have a so this is this typical cytoplasm here is a relatively large molecule so at short times at relatively short times before any collisions with their neighbors has occurred diffusion is itself diffusion even shorter times there is a ballistic regime but I'm not going to talk about that but at very short times it's diffusional because there is no it doesn't feel the neighbors yet then intermediate times these particles are temporarily caged by other particles and that's called so that gives rise to anomalous diffusion and it's typically transient behavior because in these systems if the system is fluid like after a while these motions randomize again and you will get diffusion again but now it's long time diffusion so this is typically an intermediate regime and there are also experiments because these are hard sphere colloids and you see that at relatively long time so this is the log of the mean square displacement this is the log of time so at short times it's linear at long times it's more or less linear and at intermediate times intermediate times it's a region of anomalous diffusion and you see here that it could be everything and the slope in between in between 1 and 0 and what we find in the cytoplasm it's typically a half and I don't know where that comes apparently these densities and these interactions between proteins in cytoplasm are such that it leads to an anomalous diffusion coefficient die is around a half I think that's the most reasonable explanation I can come up with at this point so it's still transient behavior but transient can be very long because we have a good long life and we are all transient alright back to phase DNA so we assume we have anomalous diffusion so that the mean square displacement of the blob of DNA is proportional to time to the power of a half where a half is to the power of alpha where alpha is equal to a half and then we assume that the diffusion coefficient still has a stokes eindzine form which is that it's kt over 6 times pi times the viscosity times the hydrodynamic radius which as we see as we saw depends on the number of ejected base pairs which depends on time and just put a delay factor in the larger viscosity effectively and it should have a unit of time to the power of 1 minus a half in order to have the mean square displacement being square nanometer so whatever it means it's something that is on the order 0.1 so that's about the difference in viscosity between cell cytoplasm and water so now couple the dynamics of the blob by the ejection of DNA but just assuming that this is the mean square displacement of the blob given by this and here and that's related to the amount of ejected DNA in the system so the blob dynamics determines how fast the DNA is being ejected at long time that's actually all I'm assuming here so then you combine everything you get an expression for the diffusion coefficient which depends on the hydrodynamic radius which depends on the number of ejected base pairs the number of ejected base pairs itself depends on the mean square displacement which depends on the diffusion coefficient and time the diffusion coefficient itself depends on n you solve for the number of ejected base pairs and you get something like some constant times time to the power alpha over 2 plus gamma and here you see as long as gamma is around 1 over 3 you always get the same coefficient so in effect that I was assuming gamma is 1 over 3 it's not very sensitive for the result as long as it's not 1 or 3 over 5 that will give you a difference then you just plug in the numbers we predict that the number of ejected base pairs function of time should be proportional to 3 to the power 0.21 and here that's what you see because here I plotted this you've seen it before but this is the fast regime so this is the short time regime so this is the long time regime and you see that the slope is indeed 0.2 and also the pre-factor because in this case I would think that even with ridiculous models you may get a reasonable coefficient but the pre-factor is also reasonable so in this case that's also important to check that of we just take a viscosity of between 10 times 200 times more than in a simple solvent you would get a reasonable value for A that we get here so you can also do it in terms of velocity it's just a matter of differentiation and it's a bit more noise but you will get to predict minus 3 over 7 and we find with an error minus 4 so that's all pretty okay so this is discussion well conclusion indeed so the late stage ejection seems to be quantitatively in agreement and just with a simple diffusion scenario and at this point we don't need to invoke the properties of ratcheting proteins it could be that the dormant stage bacteria also have a dormant stage so they can get into a kind of hibernation at some point and it could very well be that that stage it should work as a defense mechanism against phages because we know that diffusion stops when there's no metabolic activity so in terms of phages phages will only be able to to insert about half of its genome into bacteria and then the process stops so it may be I don't know it could be a mechanism er are also other things that I didn't talk about one of them is that ejections also often it's often observed that ejection goes on for a while and then it passes right and then it goes on again etc so the ejection process is kind of irregular in some cases and that could have to do with so-called dynamic hydrogenaties because we have already seen that at these length scales the molecules that are so large as these blobs of DNA that are half or totally ejected then we have glassy we know in glassy systems that they're dynamically heterogeneous and that's because they're pretty closely packed so they hardly move but when one particle turns out to be able to move it will just cause a cascade of other movements so it's a collective process and that could be the reason of why these pulses have been observed alright I think with that I'd like to stop and maybe there are questions