Viral Infection: Virus Entry and Subsequent Steps
Transcript of Part 3: Open Sesame: Cell Entry and Vaccinia Virus
00:00:02.04 My name is Ari Helenius and I'm from the Institute of Biochemistry 00:00:06.08 at ETH Zurich in Switzerland. In the last part of this seminar, 00:00:15.22 which I have entitled "Open Sesame: Cell Entry of Vaccinia Virus" the focus 00:00:24.06 will be now on the single virus and the virus and its entry into cells. 00:00:29.21 It illustrates what I will talk about, it illustrates many of the things 00:00:33.24 I have talked about before binding, signaling, endocytosis and so on, 00:00:40.03 but in a very dramatic way. This virus is rather unusual because itâ€™s a 00:00:44.21 very big virus, itâ€™s the most complicated virus family 00:00:49.06 known for animal cells because it belongs to the poxviruses. 00:00:55.21 Poxviruses are these giant viruses right here, they are enveloped 00:01:02.04 animal viruses and many of them cause diseases in different species. 00:01:07.05 Perhaps the best known is smallpox, which has been a major pathogen 00:01:15.01 in humans. In fact, between 1914 and 1978, when this virus 00:01:22.06 was finally eradicated it is estimated that some 300 million people died 00:01:26.26 of this virus. But as I mentioned it is now eradicated due to a worldwide, 00:01:32.18 sorry smallpox is eradicated due to a worldwide vaccination program 00:01:38.23 and in fact it is the only human virus that has been completely 00:01:44.23 eliminated from the population. Now the particle I will talk about 00:01:52.23 is a so-called "mature virion particle", because this virus 00:01:57.13 in its complexity comes in many forms or several different forms. 00:02:01.15 The mature virus, which you see in the electron micrograph in the back here, 00:02:06.13 is of Vaccinia virus. This is the vaccine strain of virus used to eradicate smallpox. 00:02:12.01 And it is the most abundant form of the infectious virus that exists after infection. 00:02:20.06 What you should know that it is an enveloped animal virus. 00:02:24.26 It has unlike the other forms of the virus only one single membrane, 00:02:28.22 You can see it at this light band here. It is a single lipid envelope. 00:02:35.00 It's a DNA virus. It's very big as you saw compared 00:02:40.13 to many other viruses, and it replicates in the cytoplasm, even though this 00:02:45.24 is unusual for a DNA virus. Receptors for this virus have not been 00:02:50.29 characterized in great detail yet, but it is already known that this 00:02:54.10 like many other viruses uses heparan sulfate proteoglycans as 00:03:00.17 an attachment factor. It is endocytosed like many other viruses 00:03:07.15 in large vesicles, one can see those by electron microscopy and it is known to 00:03:12.21 be acid activated so it has a membrane fusion protein complex, 00:03:18.17 which needs acid to work. So it looks like a normal virus, only 00:03:23.23 it's much bigger and more complicated. 00:03:25.27 So just to go through the early stages, which we are interested in here, 00:03:30.19 The virus binds to the cell surface like other viruses, it is then internalized in 00:03:35.28 large vacuoles. The acid inside these vacuoles after they have formed 00:03:40.18 then induces membrane fusion and the capsid which is this dumbbell 00:03:45.26 shaped structure in the middle, is released into the cytosol 00:03:49.18 where replication or transcription of the first viral protein. 00:03:54.15 The messenger RNA starts inside the capsid, later on further ones occur 00:03:59.21 from the naked DNA. So we are interested in how this virus enters cells. 00:04:07.13 The work started when Jason Mercer, a post-doctoral fellow 00:04:11.17 from the United States arrived in the lab and he had a viral particle 00:04:15.21 which was labeled such that it's fluorescent. One of the core proteins, 00:04:20.19 capsid proteins called A5 had a GFP, green fluorescent protein tag. 00:04:25.28 And the virus is perfectly infectious but fluorescent. 00:04:29.17 So the first thing he saw was that this virus also surfs on filopodia. So the cell, 00:04:35.26 the host cell is up here and you can see the virus moving along filopodia 00:04:40.14 and for all we know this is also happening through actin retrograde flow 00:04:47.03 just like the papilloma virus, and this phenomenon as I said was first seen 00:04:53.18 by Walther Mothes and co-workers for other viruses. Here, however, 00:04:58.12 something else happens. Follow what happens when the virus 00:05:01.28 over here arrives at the cell surface. So this is speeded up, so after virus 00:05:13.03 arrival at the cell surface, a bleb forms from the plasma membrane 00:05:17.19 of the host cell. It grows for about 20-30 seconds and then in 00:05:22.09 the following 20-30 seconds it retracts. And the bleb forms first where 00:05:27.25 the virus is located and later on all along the cell surface. You can see 00:05:32.14 that spreading here again, the same virus is down here, and you can see 00:05:37.05 the blebs moving up the cell body. So the whole cell becomes globally modified. 00:05:42.18 The plasma membrane starts to blow out in blebs in many different places, 00:05:47.18 up to 120 blebs per cell. So obviously the virus here is inducing a change 00:05:53.20 in the bleb. It's probably inducing a signaling pathway activating 00:05:57.02 something which induces the formation of the membrane blebs. 00:06:01.04 Such blebs have been seen before in other situations 00:06:06.07 which do not involve viruses, but not for viruses. And what we know about 00:06:10.27 this blebbing is that it is very similar to what one finds during 00:06:14.01 cell division, apoptosis, cell motility. A new technique called FIB/SEM, 00:06:21.15 or electron microscopy has allowed our co-workers 00:06:24.21 at the ETH to look at this phenomenon in three dimension 00:06:28.09 using electron microscopy. And what you see here is going through 00:06:32.11 a thick section of cells with the viral particle shown 00:06:35.11 in quite the light color. It is better appreciated if we look 00:06:40.24 at it in three dimensions. The viruses have been pseudo-colored red here 00:06:45.24 and what you see is that the cell surface is modified, you have these blebs 00:06:51.07 forming, and the viral particles seem to be endocytosing on the back side 00:06:56.26 of the fat blebs into the cell. What you also can see here are the filopodia 00:07:01.10 which extend out along which the viruses move towards the cell. 00:07:05.17 So this is very unusual, but it enforces the idea that this virus 00:07:11.14 is triggering a change in its host cell, complex change. That change involves 00:07:17.09 actin, very clearly, because actin itself in this case RFP-labeled moves 00:07:23.16 into the blebs and so does many actin modifying 00:07:26.15 and actin regulating proteins. That is already known from bleb formation 00:07:31.11 in other systems. All we can conclude at the moment is that the virus 00:07:35.27 induces something that the cell also does under normal conditions. 00:07:39.20 At the same time as Jason was looking at this plasma membrane 00:07:48.05 phenomenology, he was collaborating with a group of Lucas Pelkmans in 00:07:54.02 our institute at the ETH in trying to identify the Vaccinia virus infectome, 00:08:00.28 as we call it. Basically to obtain a full list of cellular genes and 00:08:06.22 cellular proteins involved in assisting the virus through its replication cycle. 00:08:13.06 Particularly in this case through the early stages of the replication cycle. 00:08:17.19 So we wanted to have a full list of these Trojan people here who were 00:08:24.09 helping the virus, and we want to know exactly who they are, 00:08:27.03 what they are doing, and what their addresses are and so on. Find out 00:08:33.02 in what way the molecular assistance given to the incoming virus happens. 00:08:38.26 To get such information today is possible due to two completely 00:08:44.19 spectacular events that have happened in the last 10 years. 00:08:47.12 First of all, we have now access to the human genome sequence. 00:08:51.13 Basically we have the names of the Trojans, and we have a new technology 00:08:59.12 which allows us to eliminate one gene at a time and then test the cells 00:09:05.02 for whether they can be infected or not, and that is using 00:09:07.28 high-throughput siRNA silencing in human cells. A screen in which we 00:09:13.20 silence genes one by one and test for infectivity. So this work has 00:09:20.08 to be done in an automated setting, and it was done together with 00:09:25.08 Lucas Pelkmans, Berend Snijder, and Raphael Sacher at the ETH. 00:09:31.05 So basically you need a robot to deal with the technology 00:09:35.29 also the infection all the work is done automatically and then we need 00:09:43.05 a library of siRNAs, in this case we used a library of 7,000 'druggable genes'. 00:09:50.28 They are selected as being more amenable to future drug development 00:09:57.13 than the others. So itâ€™s not the whole genome. In 384 well plates 00:10:02.14 infection was tested, and we used 3 siRNAs per gene in triplicate. 00:10:08.24 Moreover, we expressed in this case mature virus of Vaccinia in which 00:10:16.04 the GFP was included so that it was expressed from 00:10:21.28 an early/intermediate promoter. So when the cell got infected, it became green. 00:10:27.03 And then our readout was using an automated fluorescence microscopy 00:10:31.26 set-up, where pictures were taken from all the wells and then you can see 00:10:36.22 in a moment how that can be used to detect infection. 00:10:41.13 So we defined hits as those which had 2 or 3 out of the 3 siRNAs 00:10:47.03 cause either 50% or less infection in cells compared to controls. This is how 00:10:55.15 it looks. Here is a cell culture, perhaps you can 00:10:59.11 see the blue Hoechst stained cell nuclei, and the green ones are the ones 00:11:05.08 which have been infected. This is in a control with a control siRNA 00:11:10.08 and about 10-15% of the cells are infected. Now in a few cases where we 00:11:15.22 looked at them, we found that there were more cells infected than 00:11:18.19 in the control, but this was not so common. In many cases the good hits, 00:11:24.08 the strong hits looked like this. You can still see a few infected cells, 00:11:28.00 but most of the cells are no longer infected in cells transfected 00:11:32.08 with this particular siRNA. In this particular case we have silenced 00:11:37.25 a kinase called PAK1 and I'll come back to that in a moment. 00:11:41.17 So in the hits we had 142 hits out of 7,000 with 3 siRNAs 00:11:52.19 and 284 with more than 2 or 3 of the siRNAs and then only 4 where 00:12:01.23 the infection was actually increased. Now what you get from this is 00:12:07.17 then a big list of gene names and there are many ways of looking at this. 00:12:12.11 One is to make tables and pi charts like this. You can see that when we 00:12:16.28 identify which families of proteins these hits fall into, there's practically 00:12:23.10 every family that cells have from transcription factors 00:12:27.26 to membrane transport factors, ion channels, 00:12:31.05 proteasome components, etc. So this is not particularly helpful except 00:12:36.24 it tells us that the background to this infection is complex. A little more 00:12:42.12 informative is if you combine a hit list with information known about 00:12:49.07 interactions between proteins and that can be done by 00:12:53.09 using this String.embl program, which allows you to take every hit, 00:12:58.21 which every little ball here is a hit and then see is that particular protein 00:13:03.13 known to interact with anything else. And now you can see already 00:13:06.29 some clusters forming. This cluster here has practically every subunit 00:13:12.24 of the proteasome complex. This one has many subunits of 00:13:17.10 the ribosome and then associated proteins and so on. So now we are 00:13:22.12 starting to see some contours of the biology, in 00:13:25.28 the intracellular virology of this virus. For every virus we have done this 00:13:31.22 it looks different. And we're coming back in a moment to this cluster 00:13:35.06 that has to do with tyrosine kinases and molecules 00:13:38.22 involved in regulating actin dynamics. But before that I'd like 00:13:44.10 to pause a bit and think about this approach because we 00:13:48.29 and others are taking it and I think it's going to mean a huge difference 00:13:53.08 in the infectious disease area. So we are now have been looking at 00:13:58.16 pathogen-host interactions using a systems biology approach and by doing 00:14:04.13 these screens and looking at which cellular components are important 00:14:07.29 in supporting the infection of a given pathogen, we arrive at a huge 00:14:12.24 information set which can if we can analyze it correctly, reveal critical 00:14:19.19 host factors and processes that support or inhibit infectious cycles 00:14:24.17 of a pathogen. We know already they look different 00:14:27.19 for different pathogens, but they are sometimes related. 00:14:32.08 This information also exposes new molecular mechanisms 00:14:36.29 in the replication cycle. We have never had a clue that the proteasome 00:14:42.10 was involved, as you saw there it's one of the major hits of the screen. 00:14:46.18 The data also provides a basis for a functional classification of pathogens. 00:14:53.04 We can now classify on the basis of these hit lists, viruses according 00:14:58.22 to their functions in the cell, different viruses. And it is also starting to give 00:15:05.11 us some information about cell and species tropism. Why are some cells 00:15:11.11 infected and others not. In addition, I think that the data 00:15:19.17 will start to show us new similarities between viruses and viral diseases. 00:15:27.05 And hopefully we will find common cellular factors involved which may 00:15:32.23 allow us to approach antiviral agents so they can work on more 00:15:38.06 than one virus at a time. So of course it provides us new approaches 00:15:43.19 to perturb infection and obviously new potential targets 00:15:48.15 for antiviral strategies. So I think this systems biology approach 00:15:55.02 will be a key to a new stage in infection and disease. However, where we now are, 00:16:01.01 after we have the screen, we have to go back and do the cell biology 00:16:05.14 and the molecular biology. We have to validate the different hits 00:16:11.07 and clusters. So what we decided to do was to focus on 00:16:15.23 one of the clusters which we called the PAK1/RAC1 clusters. These are 00:16:20.05 various hits. PAK1 is a kinase known to be involved in regulating 00:16:26.09 the dynamics of actin in cells, and RAC1 is a GTPase that activates PAK1. 00:16:33.03 These other hits all make sense and they regulate also these 00:16:37.21 two components. Now the first thing to show of course is, 00:16:42.02 and I won't through more than very superficially, you have to use 00:16:47.14 other methods to show that PAK1 for example is a real hit. 00:16:51.04 You have to use additional siRNAs, you have to use specific inhibitors 00:16:56.10 which we luckily had obtained. You can then use dominant negative 00:17:01.24 constructs to look at it. All this has been done. You can demonstrate 00:17:06.16 that when you add the virus this PAK1 actually gets activated, and so on. 00:17:11.07 The main point here is that you have to move back at this stage 00:17:15.21 to the cell biology and the biochemistry of the system and then 00:17:20.09 in a tedious set of experiments analyze it from that point of view. 00:17:24.17 Here is just an example of what happens when you add the virus. 00:17:28.18 Then the PAK1-GFP, which is shown in green, moves, 00:17:35.25 this is a non-infected cell that's mainly cytosolic, now it moves into the blebs. 00:17:42.01 And is there as long as the bleb exists. So PAK1 is indeed activated here. 00:17:47.16 RAC1 is also activated. One can follow its activation already five minutes 00:17:53.24 after addition of the virus, RAC1 is activated very heavily for about half an hour. 00:17:59.12 Now you can then also look at what the blebbing means. 00:18:06.18 Is it essential for infection or not? And you can see if you 00:18:10.28 use a dominant negative RAC1 construct you block blebbing. 00:18:14.29 You also block infection, which is not shown here. If you have a 00:18:19.00 constitutively active RAC1, you increase blebbing and also infection. 00:18:25.16 Another approach is to use chemicals. I mentioned this in an earlier lecture 00:18:33.15 That's very powerful today. Inhibitors have their problems but taken 00:18:38.09 as one of many approaches they are powerful. Here, we are looking at 00:18:41.29 inhibitors of actin dynamics and inhibitors of different kinases here. 00:18:47.09 And you can then get a picture of which components in the cell 00:18:51.09 are important, which are not. All this can then be fitted in into the known 00:18:57.28 signaling pathways, and without going through any of the details here 00:19:02.11 I will just show you here a cartoon of the various signaling pathways 00:19:06.08 known or the components known to regulate actin modifications 00:19:13.06 down here by interactions happening at the cell surface. 00:19:17.05 So that can go through three classes of receptors, 00:19:20.20 receptor tyrosine kinases here, integrins and trimeric G-protein receptors here, 00:19:27.23 which feed into this complex network of signaling factors. 00:19:32.07 I'll just show you here which of these we already know from our work 00:19:35.17 and other people's work are involved in the activation of the blebbing 00:19:40.23 and the infection by Vaccinia mature particles. So it looks like it's coming 00:19:48.23 probably from the receptor tyrosine kinase pathway towards the actin. 00:19:54.23 This is obviously a cellular signaling pathway. The viruses are simply 00:20:02.05 using it to trigger this cascade of events that eventually leads 00:20:06.15 to their uptake. There are bacteria that also manipulate this same pathway 00:20:11.18 but they very often inject effector molecules into the cell, 00:20:16.12 which then directly interact with these pathways and modify them 00:20:20.23 so that they can be helpful to the bacteria. Viruses have nothing to inject. 00:20:28.27 They have to simply take advantage of the existing machinery and 00:20:33.15 this I think is an example of that. So what is the pathway like? 00:20:37.26 We have a rudimentary picture. The receptors are not clear perhaps 00:20:42.05 they involve to some extent EGF receptor, because it gets activated then 00:20:46.27 there are serine/threonine kinases, PI3 kinases, sodium/proton exchanger 00:20:52.15 required at the next stage. Activation of RAC occurs, PAK is activated 00:20:57.27 and then another set of kinases and other factors are involved. 00:21:02.01 Eventually then actin and actin associating proteins get activated, 00:21:07.10 myosin II is involved. And then that leads to blebbing and blebbing leads to 00:21:12.14 endocytosis and that in turn leads to infection by Vaccinia virus. 00:21:20.01 Now what type of endocytosis is this? Is there anything similar known? 00:21:27.20 One clue is that when you add the virus to cells, it starts to gulp in fluid. 00:21:33.25 Just liquid from the outside. It is the yellow line here. Normally these cells 00:21:39.03 are all the time internalizing some liquid but when you add the virus or 00:21:44.24 macropinocytosis activator, PMA, you get this huge increase in 00:21:54.03 uptake of fluid. So that starts to argue that what we are looking at here is 00:22:00.09 macropinocytosis. An endocytic mechanism that has been described 00:22:05.09 already some time ago and is now more and more coming into focus. 00:22:09.22 So what is macropinocytosis? It's a signal induced transient 00:22:14.20 endocytic pathway in most cells. It only occurs for half an hour or so. 00:22:19.21 It involves ruffling and blebbing, in our case it's blebbing, 00:22:24.01 in other cases it's simply ruffling of the cell surface. Those lamellipodia, 00:22:28.20 you saw in previous lectures, are a part of this ruffling event. The ruffles 00:22:34.20 and the blebs when they come back to the membrane and encase 00:22:40.08 some fluid space vesicles so it's an increase fluid phase uptake. 00:22:46.18 All of it absolutely actin dependent. And it's characterized 00:22:51.07 by dependence on RAC1/Cdc42, and PAK1 and so on. 00:22:57.09 Many of these or all of these were also found in our studies. 00:23:01.06 So there is already a sort of diagnostic set of features, which we know. 00:23:06.03 And all of them fit with the uptake of Vaccinia virus. So just to go back 00:23:10.29 to this one here, you see here is the macropinocytosis uptake process 00:23:16.28 right here, and used by Vaccinia. We also know from other people's work 00:23:22.11 it is used by adenovirus B and perhaps in some cases even 00:23:26.20 herpes simplex virus. Okay. What is there in the virus particle 00:23:33.09 that triggers all this? How does it actually manage to fool the cell into 00:23:39.12 endocytosis by macropinocytic uptake? It turns out that 00:23:44.10 the plasma membrane of Vaccinia virus is extremely rich in a phospholipid 00:23:49.21 called phosphatidylserine. It's a lipid that is present in the plasma membrane 00:23:54.29 of all cells, but it's in the inner leaflet. It's not exposed to the outside. 00:23:59.09 It is known to be, however, exposed when cells undergo apoptosis. 00:24:05.23 Exposed on the remnants of the cell in the so-called apoptotic bodies 00:24:10.17 have exposed phosphatidylserine, and there it serves 00:24:14.01 as the eat-me signal. A signal that induces macropinocytotic 00:24:20.09 and sometimes phagocytic uptake of those remnants by neighboring cells. 00:24:25.06 So when a cell dies, the remnants of that cell are eaten up by 00:24:30.10 neighboring cells and that uptake process very often involves 00:24:34.00 macropinocytosis. So here the virus also seems to contain 00:24:39.04 phosphatidylserine and if one blocks the phosphatidylserine 00:24:42.20 with a phosphatidylserine binding protein, Annexin-5, 00:24:45.16 then the virus can no longer infect. And moreover, if you exchange 00:24:50.14 the virus lipids, which you can do by first solubilizing away the other one, 00:24:57.14 the original ones, and then add new ones back on the virus, you can see that 00:25:02.02 if you don't add any lipids you lose infectivity completely. If you add 00:25:08.07 lipids without phosphatidylserine present you have also practically 00:25:13.08 no infection, but if you put the phosphatidylserine back again 00:25:15.26 you can then practically rescue full infectivity again. So this and other 00:25:22.01 studies show that the phosphatidylserine is critical for inducing 00:25:26.19 the blebbing, for activating the PAK, etc., and also for infection. 00:25:31.13 So the conclusion from this, and I'm almost at the end, is that the virus 00:25:36.15 here enters by apoptotic mimicry. It is mimicking an apoptotic body 00:25:44.10 and in that way triggering the response in this case tissue culture cells 00:25:50.27 which they normally will have the apoptotic bodies, which would involve 00:25:54.28 all these complex events including blebbing and endocytosis. 00:25:59.04 So the virus mimics an apoptotic body. So if we go now through 00:26:04.17 step-by-step what happens here, the MVs, these are the mature virus particles 00:26:11.03 bind to filopodia and surf along those filopodia to the cell body, 00:26:15.22 probably using retrograde actin flow. The exposed phosphatidylserine 00:26:22.08 then activates a signaling pathway, which involves RAC1, PAK1, 00:26:27.11 and other components, and the result of that activation is 00:26:32.07 that blebbing occurs, first at the sight of the virus attachment but then 00:26:37.05 globally around the whole cell. And then as the blebs are retracting, 00:26:47.17 our view is that membrane is being internalized by macropinocytosis 00:26:52.29 and that brings in some of the viral particles. Once they are in 00:26:57.00 the macropinosome, the vacuole that is formed, acidification takes place 00:27:03.04 and then the acid activated fusion takes place in the macropinosome. 00:27:09.01 So here is an example of a virus that does pretty much what I've been 00:27:13.00 telling you. It activates the cell, without this activation endocytosis 00:27:17.22 would not occur and it basically uses complex cellular machinery 00:27:23.29 intended for something else to enter cells. It is also interesting 00:27:28.23 and important that macropinocytosis, unlike phagocytosis is known 00:27:32.27 to suppress native immune responses. That means that the immune 00:27:37.29 response for the incoming virus is probably reduced, and that of course 00:27:42.14 is in the interest of the virus. It's also maybe true that this virus 00:27:48.24 being so big has had to evolve this different entry pathway because 00:27:53.13 it cannot use clathrin coated pits and so on. Okay, so I'd like to finish off 00:28:00.22 by going through perhaps the most important implications 00:28:05.20 of what I have been saying. The fact that viruses are Trojan horses 00:28:10.20 means that they depend on the host cell for infection at every stage, from 00:28:16.02 the early stages to the synthesis and also for final assembly of new particles. 00:28:24.22 So what will happen if we would now start instead of using antiviral agents, 00:28:31.20 developing them against the viral components, try to focus on making drugs 00:28:38.06 which prevent the cell components from doing their job. We would then 00:28:42.29 be able to block infection like we did with many of these drugs shown here 00:28:49.07 and we would have a totally new situation, because viruses normally 00:28:56.03 circumvent the action of antiviral drugs simply by mutating and 00:29:02.29 becoming resistant to those drugs. It's clear here I think that it is very much 00:29:08.02 more difficult for the virus to mutate in such a way that it can escape 00:29:13.02 the use of critical cell components. Another good thing is that the cell 00:29:18.23 provides many targets, hundreds of different proteins are needed for any 00:29:23.04 given virus. Any one of those in principle could be a an antiviral target. 00:29:27.06 We already know that if we look at the lists of these infectomes 00:29:31.21 some drugs already exist to some of them, and they may be developed 00:29:35.19 further to have antiviral agents. Some of these targets are clearly present 00:29:42.05 in the infectomes of more than one virus so maybe it's possible to imagine 00:29:47.13 in the future drugs which block a whole group of viruses at the same time. 00:29:52.17 And it will also be possible to have drugs hopefully, which block each of 00:30:00.09 the entry pathways for example so that when a newly emerging virus 00:30:05.21 comes into the picture, the drug or set of drugs maybe already exists, 00:30:10.13 which blocks its entry pathway, because it will have to use one or the other 00:30:15.14 of the known pathways. So I will stop there and thank you very much.