Session 5: Cell Motility
Transcript of Part 1: Protein Polymers, Crawling Cells and Comet Tails
00:07.3 Hello. 00:09.1 My name is Julie Theriot 00:10.2 and I'm a professor at Stanford University. 00:12.1 I'm delighted to be taking part in this iBio Seminar series 00:15.1 and in the first part of my presentation today 00:17.2 I would like to talk about protein polymers, 00:19.1 crawling cells, 00:20.3 and comet tails. 00:23.2 One of the things that I find absolutely fascinating 00:25.1 about biological systems 00:27.0 is how the enormous complexity 00:28.2 that we see in living cells and living organisms 00:30.2 can arise by the assembly 00:32.3 of relatively small and simple subunits or parts. 00:35.3 For example, the protein structure 00:38.1 that's shown here 00:40.1 is one of the most abundant proteins in eukaryotic cells, 00:42.2 called actin. 00:43.1 It also happens to be my favorite protein. 00:45.2 And it's a small, globular protein, 00:47.2 not particularly distinguished in terms of its overall shape. 00:50.0 It has the capacity to bind and hydrolyze ATP. 00:54.1 However, it also has the capacity to self-assemble, 00:57.1 to make filaments made up of many copies, 00:59.1 many identical copies, 01:00.2 of the same protein, 01:02.1 and by that self-assembly process 01:04.0 actin and other cytoskeletal proteins like it 01:06.1 are able to elaborate tremendously complicated structures 01:09.0 inside of living cells, 01:10.1 giving rise to shapes 01:12.1 like the beautiful Purkinje cell 01:13.2 that's shown here in this classic drawing 01:15.2 by Santiago RamÃƒÂ³n y Cajal. 01:17.2 So, to me this is one of the most profound problems 01:20.2 facing our understanding of cell biology, 01:22.2 is to try to figure out what are the rules 01:24.2 by which these tiny nanometer-scale proteins 01:27.3 can come together and organize themselves 01:29.2 in order to make structures that are many orders of magnitude larger 01:33.0 - 10^4 or 10^5 orders of magnitude larger. 01:36.2 And that question becomes even more profound, I think, 01:39.0 and even more puzzling, 01:40.1 when we realize that in the context of biological structures 01:42.2 everything is dynamic. 01:44.0 This is certainly true of the cytoskeletal elements 01:46.1 and I think that's really beautifully illustrated 01:47.2 by this classic movie of a neutrophil 01:50.1 chasing down bacteria. 01:51.3 This movie was made by David Rogers at Vanderbilt University 01:54.2 in the 1950s, 01:55.3 and what we're looking at here is a blood smear, 01:58.0 where the little round dark objects 02:00.2 are red blood cells, 02:02.0 the white blood cell is labeled there in the middle, 02:04.1 and you'll see right at the front end of the white blood cell 02:06.2 is a tiny little pair of bacteria. 02:09.1 Now, when the movie starts to play, 02:12.0 you can appreciate that the white blood cell 02:14.1 is very well organized 02:15.2 -- it has a definite front and a definite back -- 02:17.1 but it's also very active, 02:18.3 so it's chasing after that bacterium. 02:20.2 As the bacterium is bouncing around, 02:22.1 because of fluid motion in the background, 02:24.1 the neutrophil is able to change direction 02:27.0 and keep going after the little bacterium, 02:29.1 until finally it's going to catch up 02:31.2 and then engulf it, 02:32.2 which is its biological function. 02:34.2 Now, that happens to be a human neutrophil, 02:37.0 probably taken from David Rogers' own blood, 02:39.1 but if we look at other types of eukaryotic cells 02:42.2 we can find examples like this amoeba. 02:44.1 This is Acanthamoeba castellanii, 02:45.3 which lives in the soil, 02:47.1 and it also makes a living by crawling around 02:49.0 and eating bacteria. 02:50.1 And when we watch it move under the microscope, 02:52.1 it looks actually, remarkably similar 02:54.2 to the movement of the human neutrophil. 02:55.3 It would really take an expert to be able 02:57.2 to tell the difference between these two different cells. 02:59.3 And yet these two cell types 03:01.1 have been separated by about two billion years of evolution. 03:04.3 No, we appreciate today that not only 03:07.1 is the overall appearance of motility quite similar 03:09.1 between these two different cells, 03:10.2 but also the fundamental molecular machinery 03:12.3 that drives the motion 03:14.1 has also been conserved all of this time. 03:17.1 And one of the most important elements of that fundamental molecular machinery 03:20.2 is the actin that self-assembles into filaments, 03:23.2 as I just told you about. 03:26.2 If we look at the leading edge of a crawling cell, 03:28.3 and in this particular case 03:30.3 the image on the bottom here 03:32.1 is a fixed cytoskeleton from a cell called a keratocyte, 03:35.1 and we can label the actin filaments in that cell 03:37.2 and see their overall distribution. 03:38.2 In this particular case, 03:39.3 they are very much focused towards the front, 03:43.0 towards the leading edge. 03:44.1 And in this beautiful electron micrograph 03:46.0 by Tanya Svitkina 03:47.2 you can see at the ultrastructural level 03:50.1 exactly how abundant 03:52.2 and how densely intertwined and crosslinked 03:54.1 these actin filaments actually are. 03:56.0 The top shows a relatively low magnification 03:58.2 and then the bottom part here 04:00.1 shows individual boxes 04:02.2 magnified from the marked portions in the top 04:05.2 where you can see the actin filaments 04:07.2 are not only crosslinked to each other, 04:09.1 but also actually branching off one another 04:11.2 to make this large structure 04:13.0 that is able to push the cell forward. 04:16.0 Now, when we think about how 04:18.1 these tiny little nanometer-scale proteins 04:19.2 are able to assemble 04:22.1 tnto such an enormous structure that's able to perform 04:24.2 this vast amount of physical work, 04:26.0 of pushing a whole cell forward, 04:27.2 the most important place to start 04:29.1 is just with the filament itself. 04:30.2 As I mentioned, 04:32.1 actin is a globular protein 04:33.2 #NAME? 04:35.0 is shown here at the top -- 04:36.3 and this image gives you a sense 04:39.0 at the structural level 04:40.2 of how all of those monomers are able to come together 04:42.1 in order to make a filament. 04:44.0 Now, this textbook diagram over here 04:46.2 emphasizes another very, very important aspect 04:48.2 of actin behavior, 04:49.2 which is critical for understanding motility, 04:51.2 and that's the fact that it's able to bind and hydrolyze nucleotide. 04:54.3 And although this diagram is somewhat oversimplified, 04:57.0 the point is that 04:59.1 when actin is bound to ATP 05:01.1 it has a propensity to assemble into filaments. 05:03.3 Once it enters the filament, 05:05.1 the hydrolytic activity of actin is activated, 05:08.0 and so within the filament 05:10.0 the ATP is converted to ADP. 05:12.1 Once that has happened, 05:13.2 the ADP-containing filaments 05:15.1 are relatively less stable, 05:16.3 and those monomers are able to come off again. 05:19.1 Now, the consequence of this is, 05:20.3 as long as there's free ATP present, 05:22.1 actin will bind ATP, 05:24.0 assemble into a filament, 05:25.1 hydrolyze, 05:26.2 disassemble, 05:27.2 and keep this going in a treadmilling-type cycle 05:29.2 indefinitely. 05:30.3 And it's that assembly and disassembly 05:33.0 that actually drives a lot of the dynamic movements of cells 05:35.2 that we'll be talking about. 05:36.3 A few other important things to bear in mind about actin 05:39.1 when we're thinking about how it works inside of a living cell 05:41.2 is that in order to form a filament 05:44.1 it's nucleation that's the rate-limiting step. 05:46.0 So, there's lot of actin monomers 05:47.2 floating around all the time, 05:49.0 some of them are bound to other proteins 05:51.1 that modulate their propensity to make filaments, 05:53.2 but when three of them come together, 05:55.1 or when they're catalyzed to come together 05:57.1 by a specialized nucleating protein, 05:59.1 that's the initiating event 06:01.0 that will allow filament elongation to begin. 06:03.3 Because of the constant hydrolysis of ATP 06:06.0 in this process of assembly and disassembly, 06:08.2 equilibrium is never reached inside of a cell. 06:10.2 All of the actin filaments 06:12.0 are always constantly turning over. 06:13.3 This is obviously true in something like the neutrophil 06:15.3 that has to rearrange all its actin filaments 06:17.2 so fast in order to crawl, 06:19.0 but it's also true, 06:21.1 although to a somewhat lesser extent, 06:22.3 in cells like skeletal muscle or cardiac muscle, 06:25.1 where the actin filaments are relatively static. 06:28.0 Always, there's ATP hydrolysis 06:29.2 and always there's some kind of turnover 06:31.1 associated with the filament structures. 06:33.3 In the kinds of cells we're going to be focusing on, 06:36.1 so, relatively rapidly moving cells, 06:37.3 the average half-life of the filaments 06:39.1 is on the order of a few tens of seconds 06:40.2 or maybe a few minutes at most, 06:42.2 and they're organized in order to make 06:45.1 these large-scale structures that can span 06:46.2 throughout the entire cell and govern its behavior 06:48.3 by literally hundreds of different kinds of 06:51.2 actin binding molecules 06:53.0 that bind along the filament, 06:54.2 regulate their assembly and disassembly, 06:57.0 crosslink them, sever them, bundle them together... 06:59.2 anything you can imagine that a protein can do to these filamentous structures, 07:02.2 there's probably some protein out there that does it. 07:05.1 And one of the things that I particularly want to focus on 07:07.1 in our discussion today 07:08.2 is this last fact down here at the bottom here, 07:10.2 that polymerization of actin, 07:12.2 and also actually of tubulin into microtubules, 07:15.1 can actually generate physical force, 07:17.2 very similar to the way that a molecular motor 07:19.2 can generate physical force. 07:21.1 And in the context of cell motility, 07:23.0 like those movies that I just showed you, 07:24.2 it's actually the assembly of actin 07:26.1 at the leading edge 07:27.3 that is thought to be one of the major drivers 07:29.1 of that particular kind of motion. 07:32.3 So, let's take a step back for a minute 07:34.3 and think about how it is that 07:36.2 polymerization of a protein 07:38.1 can actually lead to physical force. 07:39.3 One way I find helpful to think about it 07:42.1 is to imagine the protein polymerization reaction 07:43.3 simply as a biochemical binding reaction, 07:45.3 where on the left-hand side 07:48.0 we have a small subunit that's four subunits long 07:51.1 and a free subunit. 07:53.1 Those are going to be the reactants. 07:55.3 When they bind to each other 07:57.1 with some on rate designated by k-on, 07:59.0 they'll make a filament that's one subunit longer 08:01.2 #NAME? 08:03.3 And because this is just a binding reaction, 08:05.1 it can go to equilibrium, 08:06.2 and the ratio between the k-on and the k-off 08:08.2 is going to give you some equilibrium constant, 08:10.2 which for cytoskeletal proteins we call 08:13.1 C-crit. 08:14.1 That stands for critical concentration. 08:16.1 Now, if we think about a filament like this 08:20.2 that is growing not at equilibrium, 08:22.2 but rather is growing 08:24.1 with some sort of non-equilibrium situation 08:26.1 in its environment, 08:28.0 we can end up with a situation like 08:30.1 what's illustrated at the top here. 08:31.2 If we have excess monomers present, 08:33.2 so that is, you know, 08:34.3 just one little filament and lots of monomers, 08:36.1 but more than would be present at equilibrium, 08:38.2 then those filaments will tend to polymerize. 08:41.0 Because we have excess monomers, 08:42.3 that will put us in a regime 08:44.1 where polymerization is favored thermodynamically 08:48.1 #NAME? 08:49.3 is gonna be negative overall. 08:52.2 Now, if we imagine taking one of those little filaments 08:54.2 and putting it between, 08:56.2 on the one side, a rigid barrier like a wall, 08:59.1 and on the other side 09:01.0 a smaller barrier that can move back and forth, 09:03.1 as long as the free energy 09:06.2 that is released by the polymerization reaction 09:09.1 is greater than or equal to 09:11.2 the amount of work 09:14.1 that it takes to move that little black barrier 09:16.2 through that distance delta, 09:17.3 then the overall polymerization reaction 09:19.1 will still be favored 09:20.2 and the barrier will be pushed forward. 09:23.0 So, using that kind of very fundamental 09:26.0 thermodynamic, energetic argument, 09:27.2 Terrell Hill and Marc Kirschner in 1982 09:29.2 wrote this fabulous 125 page manifesto 09:32.3 describing all the different ways that protein polymerization 09:34.3 and depolymerization 09:36.1 can be used to generate force. 09:37.1 And from their calculations 09:39.0 they came up with this simple rule 09:41.0 about the max amount of force 09:43.0 that can be generated by polymerization of a filament. 09:45.1 That depends on the size of the monomer, ÃŽÂ´, 09:48.0 and depends on the concentration of free monomers 09:50.1 in solution, 09:51.2 relative to this critical concentration, 09:53.2 or the equilibrium binding constant. 09:56.2 So, if you calculate for actin inside of a living cell 09:59.2 what those numbers actually should turn out to be, 10:01.1 we estimate that the amount of force that 10:03.1 you should be able to get from a single actin filament 10:04.2 is on the order of actually 5-10 picoNewtons, 10:08.0 which surprisingly is just about the same amount of force 10:10.2 that you get from a real molecular motor 10:12.1 like myosin or kinesin. 10:14.0 So, although this kind of polymerization-driven motion 10:16.1 is maybe not as familiar 10:17.2 and not as intuitive, 10:18.3 it's actually a perfectly good motor 10:20.1 and it can generate just as much force 10:21.3 as something like myosin. 10:24.2 Now, thinking about how fast processes 10:27.1 can be driven in this way 10:29.0 is a little bit different from thinking about the energetics. 10:30.2 In order to understand how fast things go, 10:32.2 we have to have some sort of kinetic model, 10:34.1 and in particular we're thinking about a filament 10:36.2 that's polymerizing up against a barrier... 10:38.1 we have to have some idea in mind 10:40.3 about how that space can be opened up 10:42.2 for another monomer to come and sneak 10:44.2 onto the end of the filament, 10:46.1 and extend the filament, 10:47.1 and then push the barrier. 10:48.2 And George Oster and his colleagues 10:49.2 have done a lot of interesting calculations, 10:51.2 assuming different kinds of thermal flexibility 10:53.2 and different components of the system. 10:55.2 For example, you can imagine the barrier 10:57.3 is able to move back and forth, 10:59.0 or you could imagine the filament 11:00.1 is able to flex up and down a little bit, 11:01.3 or you could imagine the crosslinks between the filaments 11:04.2 maybe can move their angles. 11:06.1 All of those things can give you 11:09.1 enough space to allow actin monomers to come in 11:10.2 and continue the polymerization reaction 11:12.0 and generate force to 11:14.0 physically move the barrier forward, 11:16.1 and doing some reasonable calculations 11:17.2 about how fast those kinds of things 11:19.1 should be able to happen in the context of a living cell, 11:21.2 they've been able to estimate that this very non-intuitive reaction 11:24.3 is enough, 11:26.1 is fast enough to drive processes 11:27.2 even like the extension of that neutrophil 11:29.1 as it's chasing a bacterium. 11:33.1 So, those are calculations... 11:34.1 of course, it's always very satisfying 11:35.2 to be able to see an experimental result 11:37.3 that confirms the models 11:40.1 that have been put forth, 11:41.2 and a few years ago 11:43.2 Matt Footer in my lab, 11:44.2 together with our collaborators 11:46.2 Marileen Dogterom and Jacob Kerssemakers 11:48.1 in Amsterdam, 11:49.1 were able to actually measure the force of actin polymerization directly. 11:53.1 And the way that they did this was 11:55.2 to prepare a very stiff, 11:58.1 rigid actin bundle 12:00.1 from the sperm of the horseshoe crab 12:01.1 -- it's called an acrosome -- 12:02.2 and those acrosomes 12:04.0 can grow actin filaments off of one tip. 12:07.0 So, after preparing the acrosomes 12:08.1 and making sure they were able to polymerize actin, 12:11.1 Matt and Jacob put those acrosomes 12:13.1 into a special kind of optical trap 12:15.2 that had been invented in Marileen's lab, 12:17.2 in order to pull this structure, 12:19.3 where here is the filament bundle 12:23.0 and then there's a bead that's being held 12:25.0 onto by the optical trap, 12:26.3 and bring it up against a microfabricated wall. 12:30.0 In that configuration, then, 12:31.2 when the bead is pushing up against the wall, 12:35.2 when we add actin monomers to that mixture, 12:37.2 the actin monomers, 12:39.0 given the thermal fluctuations of the bead and the trap, 12:41.3 are able to sneak in between the end of the acrosome and the wall 12:45.3 and actually extend that bundle of actin filaments 12:48.2 and push the bead out of the trap. 12:51.1 We're able to measure, very precisely, 12:53.1 exactly how far the bead has moved out of the trap 12:55.3 and use that to measure the amount of force 12:57.3 that's generated in this experiment, 12:59.0 and what they were able to find 13:00.2 was that the actin filament 13:02.0 will continue to grow, 13:03.1 as you see in this trace, 13:05.0 until it reaches a stall, 13:06.2 where the force is exactly balanced 13:09.2 by the pushing against the wall. 13:11.1 And the force that they can measure 13:13.1 for these small bundles of actin filaments 13:15.0 is on the order of a few picoNewtons, 13:16.2 1 or 2 picoNewtons, 13:17.3 which is very close to what 13:20.0 that thermodynamic prediction 13:21.1 would have guessed 13:23.1 under the conditions that we used. 13:26.1 So, it's very satisfying to see here 13:27.3 that we've got one filament 13:28.3 that is able to actually directly generate a force 13:31.0 that we can measure, 13:32.1 but if we want to think about how this works in the context 13:34.1 of cell motility, 13:35.1 we have to actually think about many filaments working together, 13:38.1 and this kind of experiment 13:39.2 is not really going to help us understand 13:41.2 what the rules are 13:43.1 that govern the cooperation 13:45.0 among a whole gang of filaments 13:46.2 that are operating over a very large area of space. 13:50.2 In order to get some mechanistic insight 13:52.2 into this question of 13:54.2 What happens when many actin filaments are trying to work together?, 13:56.2 we've actually been able to learn a tremendous amount 13:58.3 from a completely different area of biology 14:01.1 and that's actually the field of bacterial pathogenesis. 14:04.3 So, it turns out that infectious bacteria, 14:07.1 and also viruses, 14:08.2 that live inside of mammalian hosts 14:11.1 and exploit the resources 14:13.0 of their mammalian hosts' body 14:14.1 in order to replicate and in order to spread, 14:16.3 actually turn out to be excellent cell biologists. 14:19.3 They know all the little details, 14:21.1 all the little ins and outs of exactly how mammalian cells work, 14:25.1 exactly what the most important aspects are 14:27.1 for things like their replication or their motion, 14:28.3 so that the pathogens can take advantage 14:31.2 of those properties of their hosts' cells 14:33.2 in order to perform whatever it is the pathogen 14:35.1 wishes to perform. 14:37.2 One of the examples of a pathogen, 14:39.3 a bacterial pathogen, 14:41.1 that's turned out to be a really outstanding cell biologist 14:43.1 is this organism here, Listeria monocytogenes. 14:46.3 Now, Listeria is a common soil organism 14:49.1 and all of us eat a little bit of it every day 14:52.0 when we eat a salad or something, 14:53.3 but when we consume food 14:56.1 that is very heavily contaminated with Listeria 14:58.2 -- this could be something like a cantaloupe, 15:00.1 or a chicken, 15:01.2 or sometimes ice cream -- 15:03.0 then the bacteria are actually able 15:07.1 to invade the mammalian host cells 15:09.0 that line our intestines. 15:11.0 And for most healthy adults, 15:13.1 this infection is self-limiting, 15:14.1 you'd never even know you had it, 15:15.2 but for people who are immunocompromised 15:17.1 or for pregnant women, 15:18.2 the infection can actually spread throughout the entire body. 15:21.1 And what we're looking at here 15:23.3 in the micrograph 15:25.1 is a single mammalian host cell 15:26.3 that has been grown in tissue culture 15:29.1 and that was infected 15:31.0 with probably one bacterium 15:32.2 about five hours before this movie was made. 15:34.2 And you can see the bacteria have been replicating in the cytoplasm 15:38.2 -- they're each of these little dark 15:40.0 bullet-shaped objects that you see -- 15:42.0 and what I hope that you'll appreciate 15:43.2 when I start to play the movie 15:44.3 is that these bacteria are able 15:46.2 to not only replicate in the most cell cytoplasm, 15:48.1 but they're also able to do this absolutely extraordinary thing, 15:50.3 which is they can cruise around 15:53.1 like little speedboats. 15:55.2 And in fact the motion of these bacteria 15:58.1 inside the cytoplasm of the host cell 16:00.0 is driven by the assembly of host cell actin filaments. 16:04.3 So, behind each of the little bacteria that's moving 16:07.2 you can see a phase-dense streak, 16:09.0 and that's actually what's called a comet tail, 16:11.3 that is the actin left behind the bacterium as it moved. 16:17.0 Now, we can see for example in this fluorescence micrograph 16:19.2 that those filaments, 16:21.1 those comet tails that are made of host cell actin, 16:24.0 where here the bacteria have been labeled red with a fluorophore 16:26.3 and then the actin filaments of the host cell 16:28.2 have been labeled in green. 16:30.0 And you can see every bacterium 16:31.2 that's inside the cell 16:33.1 is associated either with a little cloud 16:34.2 or else with one of these comet tails 16:36.1 of actin filaments. 16:39.1 And looking a little more closely 16:40.2 at the level of the electron microscope, 16:42.1 you can see the way that those filaments are arranged 16:45.1 in this incredibly dense crosslinked structure 16:47.2 that traces behind the bacterium 16:50.0 and records the path it traveled over. 16:52.2 Now, having analyzed some of the dynamic behavior 16:54.3 of the actin filaments associated with bacteria comet tails, 16:58.0 what I can tell you is that the assembly of all of the filaments in this comet tail 17:01.2 took place right at the bacterial surface, 17:04.1 and it was actually the assembly of those actin filaments 17:06.2 up against that smooth cell wall of the bacterium 17:09.3 that generated the force 17:11.3 to push the bacterium through the cytoplasm of this host cell. 17:16.0 As we move further back in the comet tail, 17:18.1 the actin filaments stay crosslinked to one another 17:20.2 to make this very nice, tight, 17:22.1 characteristic columnated line, 17:24.0 and the filaments actually remain stationary in the cytoplasm, 17:26.2 like the wake behind a boat, 17:28.2 as a history of where a bacterium has moved. 17:31.2 And then, finally, 17:33.0 at the back end of the comet tail, 17:34.1 the old filaments fall apart. 17:35.2 Because their ATP has been hydrolyzed, 17:37.0 they get disassembled by accessory factors 17:39.2 in order to regenerate actin monomers 17:41.2 that can then rejoin the actin monomer pool, 17:44.1 diffuse around the cell, 17:45.2 and come back to the front, 17:46.2 where they can continue to generate force. 17:50.0 So, part of the reason that this particular organism 17:52.2 has been so appealing as a way to 17:55.3 study the cooperation among many actin filaments 17:57.2 is because it is particularly amenable 17:59.0 to both biophysical and biochemical manipulation. 18:02.3 So, for example, on the biochemical side, 18:04.2 we were able to reconstitute the movement of this bacterium 18:07.1 in cytoplasmic extracts, 18:09.1 more than 20 years ago now. 18:10.3 And then a series of biochemists 18:13.1 started fractionating extracts 18:15.1 and trying combinations of ideas of different proteins 18:17.2 that they thought might contribute to this process, 18:20.1 until in 1999 a really heroic piece of work 18:22.2 from Marie-France Carlier's lab 18:24.1 was able to demonstrate 18:26.1 motility of these bacteria 18:27.3 using a mixture of only purified proteins 18:29.1 that had all been completely identified. 18:32.3 Now, in the meantime, 18:34.1 we were actually also able to get rid of the bacterium 18:37.0 and replace it with a polystyrene bead, 18:38.3 which you see... sorry, which you see right here. 18:42.0 That polystyrene bead is coated with 18:43.2 actually just a single protein from the bacterium, 18:46.1 which is enough to initiate this whole reaction 18:48.2 and grow this beautiful comet tail 18:50.2 that looks very much like the comet tail associated with a real bacterium. 18:53.2 And in fact when we put those beads 18:55.2 under the microscope, 18:57.0 what we can see, here with labeled actin, 18:59.1 is that those beads actually jet around 19:01.1 and in fact look exactly like the bacteria do. 19:04.1 So, this is something that be reconstituted 19:06.2 both biochemically and biophysically 19:08.0 with these artificial substrates. 19:10.2 So, based on those kind of experiments, 19:12.1 over a period of about 10 years 19:14.1 about 20 different labs 19:16.1 contributed to understanding the roles 19:18.1 of all the different proteins associated with this form of motility, 19:21.0 identifying the proteins, 19:22.1 figuring out their roles, 19:23.2 and figuring out how they all work together. 19:25.1 So, to summarize all this work, 19:27.0 we're gonna focus right on the surface of the bacterium, 19:28.2 which is where all the action takes place, 19:30.2 and the first step in motility 19:31.3 is that the bacterium has to express 19:34.0 a particular protein. 19:35.1 For Listeria monocytogenes, 19:36.2 it happens to be called ActA; 19:38.1 for other bacteria that do the same kind of trick, 19:39.2 they express other proteins 19:41.2 that interestingly enough actually evolved 19:44.0 independently from ActA, 19:45.1 so it seems like this mechanism 19:47.1 for pathogen actin-based motility 19:48.3 has actually appeared multiple times 19:50.3 in evolution 19:52.2 in apparently completely unrelated strains of bacteria. 19:56.0 When that protein is presented 19:58.0 on the surface of the bacterial cell, 19:59.1 it's able to bind to particular factors 20:01.3 in the host cell cytoplasm 20:03.2 that are critical for nucleation of actin filaments, 20:06.1 and remember that nucleation is the rate-limiting step. 20:09.2 So, for the ActA protein 20:11.3 from Listeria monocytogenes, 20:14.2 it binds directly to the nucleating complex, 20:16.1 called the Arp2/3 complex, 20:18.0 that is then able to first bind 20:20.1 to the side of a preexisting actin filament 20:22.1 that's already in the tail, 20:23.2 and then nucleate the growth of the new actin filament 20:26.2 in a branch off of the side of that old actin filament. 20:30.1 And you'll remember, 20:31.2 you saw branches like that in the electron micrographs 20:33.2 at the leading edge of a crawling cell - 20:36.1 it's thought to work by a very similar process. 20:38.2 Now, as those filament are nucleated, 20:40.1 they grow by addition of actin monomers 20:43.1 that are present in the cytoplasm, 20:45.3 just floating around, 20:47.1 and landing on the ends of those filaments 20:48.2 by diffusion. 20:50.0 And that growth, as I said, 20:52.0 pushes the bacterium through the cytoplasm 20:53.2 using this force generation mechanism 20:55.3 that I described. 20:57.2 Now, this whole thing doesn't require 20:59.2 any classical molecular motors 21:00.3 - there's no involvement of myosin 21:02.1 or any of its relatives in this process. 21:05.2 As the bacterium moves forward, 21:07.0 it leaves those actin filaments behind, 21:09.2 including leaving the branch junctions behind, 21:11.2 and those old filaments, 21:13.3 as they're ripped off the surface of the bacterium 21:16.2 based on the motion of the bacterium, 21:18.1 get capped by proteins like CapZ or gelsolin, 21:20.2 so that they don't continue to grow out of control, 21:22.3 and that keeps the comet tail in its nice, 21:24.2 characteristic narrow shape. 21:27.1 And then finally, depolymerizing proteins, 21:29.1 things like cofilin and ADF, 21:31.0 will come in and disassemble those filaments, 21:32.3 tear them apart into their native monomers, 21:36.2 so that the whole thing, again, 21:38.2 can continue indefinitely, 21:39.3 as long as the bacterium is present 21:41.2 in a cytoplasm that has these factors in there. 21:46.0 So, thinking about this mechanism, 21:47.3 which we understand in a lot of molecular detail, 21:49.3 there are several things about it that I still find absolutely astonishing. 21:53.2 So, one is just that it works so incredibly fast, 21:57.0 and to put numbers on that, 21:58.2 here's an image of Listeria monocytogenes, 22:00.3 this organism about 2 microns long, 22:03.2 and its typical speed is about 0.2 microns/second, 22:07.1 and it's moving at that rate 22:09.0 by piling up all these tiny little actin monomers 22:10.3 that are only about 4 nanometers in diameter. 22:13.2 Now, it's hard to really judge... 22:15.3 is that fast, is that slow? 22:17.1 So let's compare it to some macroscopic thing 22:20.0 for which we have some physical intuition 22:21.2 about what's fast or slow. 22:23.2 And my former student Fred Soo 22:25.3 pointed out that the geometry of the Listeria moving 22:28.3 with its comet tail 22:30.0 is actually very similar to the geometry 22:31.2 of the Ohio class nuclear submarine. 22:33.2 Just like the bacterium, it's a cylinder 22:36.2 that's capped on two ends with hemispheres. 22:38.0 As it moves through its medium, 22:39.3 it leaves behind these characteristics curving patterns behind it. 22:43.1 And with the nuclear submarine, 22:45.1 we know that the length of that 22:47.1 is about 560 feet 22:49.0 and its typical cruising speed 22:50.2 is about 30 feet/second. 22:52.0 So, what would it mean if we scaled up Listeria 22:54.3 to be as big as the submarine. 22:56.3 Well, in order to do that, 22:58.2 we'd have to have the actin monomers go from being 4 nanometers 23:01.1 to being about the size of a basketball, 23:02.2 or about a foot across. 23:04.1 And if we do that scaling 23:05.3 and then see how does that apply 23:07.1 to the other numbers I've shown you 23:10.0 here for length and speed, 23:11.1 it turns out that if actin monomers 23:13.1 were the size of a basketball, 23:14.2 then the length of the bacterium 23:16.1 would be about 500 feet, 23:17.2 so very comparable to the submarine, 23:19.2 and its speed would be about 50 feet/second. 23:22.1 So, it can just as fast as the submarine. 23:24.1 And the astonishing thing about this fact is... 23:27.1 what it means is if you were in a satellite 23:30.0 watching submarines cruising around 23:32.1 on the surface of the ocean, 23:34.1 the movies of that would look exactly identical 23:35.3 to the movie that I showed you before of Listeria motility. 23:38.2 It's really just the same thing, 23:40.0 just scaled up massively. 23:42.1 But the thing that's most amazing 23:43.3 is the submarine is just moving through water. 23:46.3 But the bacterium is moving through cytoplasm, 23:50.1 and cytoplasm is dense, 23:52.0 cytoplasm is filled with all of these cytoskeletal filaments, 23:55.1 it's chock full of organelles, 23:57.0 there's all sorts of things in the way, 23:58.2 and if you actually try to calculate what the equivalent would be, 24:00.2 it's not like a submarine moving through water. 24:03.3 It's actually a lot closer to a submarine 24:06.0 moving through concrete. 24:08.1 So, it must be the case that 24:10.3 the bacterium is generating a tremendous amount of force 24:13.1 with this coordinated assembly of the actin filaments 24:16.1 in its comet tail. 24:18.1 Okay, again, 24:19.2 it would be nice to see some experimental data 24:21.0 that can tell us something about 24:22.2 how much force is actually being generated. 24:24.1 Well, one way you can get a bit of an intuitive feel for that 24:27.1 is by looking at these beautiful movies 24:28.3 made by my former student, Catherine Lacayo, 24:31.0 where she has labeled the mitochondria 24:33.1 in a host cell with a red dye, 24:36.0 and the bacteria are expressing GFP, 24:37.3 so you see them in green. 24:39.3 And as the movie plays, 24:41.1 we can watch the bacteria move around 24:42.2 and can see, actually, 24:44.2 what happens when the run into a mitochondria, 24:46.2 and remember, mitochondria are about the same size as bacteria. 24:49.2 Long in the distant past, 24:51.0 they actually used to be bacteria, 24:52.2 so this is a pretty equal battle. 24:54.2 And if we zoom in 24:57.2 and you see the bacterium is gonna... 24:59.1 as the movie loops back, you'll see it come in from the top... h 25:02.0 ere it comes, 25:04.1 and it just slices its way through that big pile of mitochondria. 25:05.3 It shoves them aside without really 25:07.2 even slowing down at all. 25:09.0 So, this has got to be a lot of force 25:10.2 to be able to just push all the organelles in the cell 25:12.3 out of the way. 25:15.1 Now, Dan Fletcher, 25:16.2 when he was a postdoc in my lab 25:18.0 -- he's now a professor at UC Berkeley -- 25:19.2 he worked out this really very clever technique 25:22.0 for measuring large forces 25:24.2 using this kind of actin polymerization geometry, 25:27.2 and what he did was he took advantage of the fact 25:29.2 that we could reconstitute motility 25:31.2 on those little polystyrene beads 25:33.2 and then just took one of those little ActA-coated polystyrene beads 25:36.2 and stuck it on the end of an AFM cantilever. 25:39.1 And what this meant was he could then 25:41.2 bring that cantilever down to a glass surface, 25:44.1 allow the actin polymerization 25:46.1 to occur in between the tip of the cantilever and the glass, 25:48.1 and that would push the cantilever upward 25:50.2 in a way that he could measure the displacement 25:52.1 and therefore also measure the force. 25:54.2 And looking at these traces, 25:56.2 there's a couple of things that are very astonishing about them. 25:58.3 The most astonishing one actually being the stall force, 26:01.2 which you see up here, 26:03.2 for a bead attached to the end of a cantilever, 26:05.2 is on the order of about 300 nanoNewtons. 26:09.0 Now remember, before we were talking about picoNewtons, 26:10.3 so it's really clear that when you get 26:12.2 a whole bunch of actin filaments together, 26:14.1 they're able to cooperate with one another 26:15.2 in order to generate tremendous amounts of force. 26:18.3 And based on estimates that Dan's been able to do 26:21.0 of the density of the actin filaments in that gel, 26:23.1 if you then estimate 26:25.2 how much force comes from a single actin filament, 26:27.0 then again we get down to numbers 26:29.0 that are on the order of about 26:30.2 a few picoNewtons per filament, 26:32.3 which is again what was calculated 26:35.0 originally from the thermodynamic argument, 26:36.3 what we measured for single filaments, 26:38.2 and what also seems to be fulfilled, now, 26:40.1 in the context of this branching growth 26:42.1 of a network 26:43.3 against the surface of, in this case, a cantilever, 26:46.0 or in the cell against the surface of a bacterium. 26:50.0 So, overall, 26:51.3 what I've told you about 26:53.2 is a set of processes 26:55.1 that are essentially based simply on the self-organization 26:57.2 and self-assembly 26:59.0 of tiny little nanometer-scale protein subunits 27:00.3 in order to make an ensemble 27:03.0 that is much, much greater than the sum of its parts, 27:05.1 in terms of its ability to do real physical work, 27:07.3 such as pushing a bacterium around inside of a cell. 27:10.1 And part of the reason that this is interesting 27:13.2 is because it applies not only to the context 27:15.1 of the bacterium moving around, 27:16.2 but also to what's happening actually 27:18.2 at the leading edge of a crawling cell. 27:21.1 Basically, you can think of the bacterium 27:22.2 as imitating a little fragment 27:25.1 of a crawling cell's plasma membrane, 27:27.3 and the growth of these branching actin filaments 27:30.1 up against that membrane 27:31.2 are pushing the edge of the cell forward 27:33.2 in the same way that they push the bacteria around 27:36.2 inside of the cells or in extracts. 27:39.1 Now, there's obviously a lot more to say 27:41.1 about how all these ideas apply 27:42.2 to the actual problem of cell motility, 27:44.0 and if you'd like to hear more about that, 27:46.0 please come back for part 2. 27:47.3 Finally, I'd like to end 27:49.3 by acknowledging the absolutely amazing team of colleagues 27:53.1 that I've had working on this project, now, for many years. 27:55.1 I've listed here the members of my group 27:57.2 who have participated in various different aspects 27:59.2 of characterization of the cell biology and biophysics of Listeria, 28:03.1 and in particular today I showed experiments 28:04.2 that were done by Matthew Footer, 28:06.1 by Lisa Cameron, 28:08.0 by Catherine Lacayo, 28:11.0 and also by Dan Fletcher. 28:13.1 And we've also had the privilege 28:15.0 of just absolutely wonderful collaborators. 28:16.2 The most important one I want to draw attention to 28:19.0 is Dan Portnoy at UC Berkeley, 28:20.1 who's been a close collaborator for almost 25 years now 28:23.3 on all of these processes associated with Listeria motility. 28:27.0 Thank you very much for your attention.