Kaposi’s sarcoma herpes virus (KSHV) nuclease: Viral Strategies to Manipulate Gene Expression in Host Cells

Part 1: Viruses Reveal the Secrets of Biology

00:00:08.00 Hello. I'm Britt Glaunsinger.
00:00:10.06 I'm a Professor of Virology
00:00:12.10 at the University of California, Berkeley
00:00:14.06 and an Investigator at the Howard Hughes Medical Institute.
00:00:17.16 And over the next couple of lectures,
00:00:19.10 I'm gonna be sharing with you
00:00:21.15 why it is that viruses have been invaluable teachers of biology.
00:00:25.10 Now, when most people think about viruses,
00:00:27.09 they conjure up images of those that are associated
00:00:30.17 with serious human diseases,
00:00:32.28 like this one: Ebola virus.
00:00:34.20 It's got a genome of RNA
00:00:36.28 that encodes only seven genes.
00:00:38.20 Yet those seven genes can devastate the human body.
00:00:42.05 Or this one: influenza virus.
00:00:44.11 It has eight segments of RNA in its genome,
00:00:47.17 and last year alone it sickened 48 million people
00:00:52.01 and killed nearly 80,000, just in the United States.
00:00:55.27 Or herpesviruses:
00:00:58.05 these are viruses that have double-stranded DNA genomes
00:01:00.27 and are among the most successful pathogens on the planet.
00:01:05.00 Nearly every human is infected with at least one herpes virus,
00:01:08.04 whether or not they know it.
00:01:10.16 Or HIV.
00:01:12.21 This is a virus whose RNA genome
00:01:15.10 reverses the central dogma of molecular biology,
00:01:18.10 and is reverse transcribed back into DNA
00:01:21.14 before it's integrated into the host's chromosomes,
00:01:24.06 and ultimately leads to depletion of cells critical for immune surveillance.
00:01:29.19 But what I also want you to appreciate is that
00:01:33.06 these viruses that are associated with human disease
00:01:36.02 are in fact only the tip of the iceberg
00:01:39.07 of viruses that are on us or around us on the planet.
00:01:42.19 We're coated with viruses on our skin,
00:01:45.12 or in our mucous membranes,
00:01:48.02 or in our gut, that cause no human disease.
00:01:51.21 And in fact, our own genomes are about 8% virus.
00:01:55.25 This is not virus that's actively replicating.
00:01:59.01 These are remnants of viruses
00:02:01.14 that invaded our genomes throughout the history of human evolution.
00:02:04.02 But it's now understood that some of these ancient viral remnants in our genome
00:02:08.17 have actually been exapted by our own cells,
00:02:10.21 and functionalized for things like
00:02:13.23 helping to control gene expression
00:02:15.25 or mounting appropriate immune defenses.
00:02:18.29 Viruses are also the most abundant
00:02:22.29 biological entities on the planet.
00:02:24.28 It's thought that they outnumber all other forms of life
00:02:27.02 by at least an order of magnitude.
00:02:29.15 It's been estimated that something like
00:02:31.27 10^24 infections are happening
00:02:34.12 every second of every minute of every hour of every day.
00:02:39.13 Now, if you also consider the fact that
00:02:42.00 our ocean's biomass is about 80% microbial,
00:02:45.15 and probably 20% of that is turned over on a daily basis by viruses,
00:02:50.07 it's not hard to appreciate that they're probably major players
00:02:53.03 in the carbon and oxygen cycles
00:02:55.29 that regulate our atmosphere,
00:02:57.22 and many other processes on the planet.
00:03:00.08 So, viruses in general
00:03:04.03 are playing huge roles beyond that of just disease.
00:03:07.10 And they have taught us an immeasurable amount
00:03:10.21 about biology in general.
00:03:12.23 In fact, much of what we know about how our own cells work
00:03:15.13 has been figured out by studying
00:03:18.14 how viruses interact with them.
00:03:20.11 And a key feature about viruses
00:03:23.25 that has been important in this line of discovery
00:03:26.18 is the fact that they are masters of genetic economy.
00:03:29.19 And I'm gonna highlight this point throughout this talk,
00:03:32.04 but, simply put, what it means is they've got really small genomes.
00:03:36.00 It's not just their genomes that are small,
00:03:38.06 but the capsids that house and protect their genomes
00:03:41.06 are also small.
00:03:43.04 Let's take the cause of the common cold -- rhinoviruses.
00:03:47.03 These viruses are about 30 nanometers in diameter.
00:03:50.18 So, to put that in perspective,
00:03:52.19 what that means is that it takes about 500 million rhinoviruses
00:03:56.15 in order to fit on the head of a pin.
00:04:00.08 Think about that next time you sneeze on someone,
00:04:02.22 or are sneezed on by someone,
00:04:05.11 in the movie theater or the grocery store.
00:04:09.29 Not all viruses are, of course,
00:04:12.04 as tiny as rhinoviruses.
00:04:14.00 In fact, there are some viruses that are relatively giant --
00:04:16.25 the mimiviruses and the mamaviruses that infect amoeba --
00:04:19.27 and whose sizes rival those of some of the smallest bacteria.
00:04:23.13 But on average, viruses are many orders of magnitude smaller
00:04:29.23 than the host cells that they infect.
00:04:31.20 So, if you took an average-sized virus
00:04:34.00 and set it next to a flea, for example,
00:04:36.05 that's about the same size differential
00:04:38.26 as you or me standing next to a mountain
00:04:42.01 that's twice as high as Mount Everest.
00:04:45.05 And it's not just their size of their...
00:04:48.18 capsids that are small.
00:04:50.27 Their genomes are remarkably compact.
00:04:55.10 So, most viruses have approximately a million times less genetic information
00:05:00.00 than the hosts that they invade.
00:05:03.01 And they have to do a remarkable amount
00:05:05.18 with those snippets of genetic information --
00:05:08.20 things like figuring out
00:05:11.26 how to attach to an appropriate host cell
00:05:14.18 that is going to allow that virus to replicate;
00:05:17.26 how to get inside that host cell,
00:05:20.06 past the cell's plasma membrane;
00:05:22.06 how to release the genome from its protective coating
00:05:24.13 of an envelope or a protein capsid;
00:05:26.15 and then, how to steal the host machinery
00:05:29.21 in a way that allows the virus to express its own genes
00:05:31.26 and to replicate its genome,
00:05:34.02 all the while fighting off and evading
00:05:37.14 an arsenal of host immune defenses
00:05:40.13 that are in place in order to stop viruses
00:05:43.29 from doing just that.
00:05:46.00 The virus, once it's replicated its genome,
00:05:48.05 then has to figure out how to package its genome
00:05:50.23 into a protective coat of protein and/or membrane,
00:05:53.12 and then it has to figure out how to get out of the cell
00:05:56.14 and disseminate to find a new host cell.
00:06:00.12 An important point is that,
00:06:03.14 outside of a host, viral particles are basically inert.
00:06:07.18 They are intimately reliant on their host cells
00:06:10.26 in order to perform pretty much all of these functions.
00:06:14.23 And that's an important point because
00:06:17.29 what it means is that if you can understand
00:06:20.15 how a virus is doing these things
00:06:22.04 -- how it's interacting with its host cell --
00:06:25.00 you'll of course learn something important
00:06:28.14 about what the virus needs to replicate,
00:06:30.10 but you will also learn something fundamental
00:06:32.25 about how the host functions,
00:06:34.24 be that host a human or an insect or a mosquito,
00:06:39.15 a crop plant, or another microbe.
00:06:43.15 And that is why viruses told us so much information
00:06:47.20 about the fundamentals of molecular biology,
00:06:50.19 and viral research has led to a number of
00:06:54.27 huge discoveries in biology.
00:06:56.23 This spans the gamut from the discovery of tumor suppressors
00:06:59.12 and oncogenes
00:07:01.18 to the basic understanding of how our DNA is replicated
00:07:04.05 and how our genes are expressed.
00:07:06.26 All of these things are critical for... for things like cancer.
00:07:10.07 To technological advances
00:07:13.21 and things directly related to human health,
00:07:15.16 like the development of vaccines and gene therapy vectors.
00:07:17.26 And of course, CRISPR,
00:07:19.27 which was originally in place, of course,
00:07:22.22 as an immune defense against viruses that invade bacteria.
00:07:27.07 And things like understanding evolution and immunology.
00:07:31.25 And I think it's important to note that
00:07:34.25 many of these discoveries were made,
00:07:37.24 and continue to be made,
00:07:39.23 using viruses that themselves cause little to no human disease.
00:07:43.07 And that really emphasizes the importance of
00:07:47.09 basic discovery-based research
00:07:49.13 for fueling clinical innovation.
00:07:52.01 So, let me try and put this in perspective
00:07:53.27 using gene regulation as an example.
00:07:58.05 You can think about gene expression
00:08:00.12 as being like a complicated circuit.
00:08:02.23 It's governed by many, many different regulators
00:08:05.26 or inputs,
00:08:07.22 which all have to coordinate to give a particular gene expression output.
00:08:11.26 And we know that viruses interface
00:08:14.27 with many of the central regulators of this pathway.
00:08:18.22 And that's because their gene expression strategies
00:08:21.19 closely mimic those of our own,
00:08:23.26 but because of this genetic economy
00:08:26.06 -- their small genomes --
00:08:28.05 they don't have the space to encode all of the factors
00:08:32.11 that they need to accomplish these tasks,
00:08:34.23 and that's why they have to steal them from a host.
00:08:37.01 And the other important point is that they evolve very rapidly,
00:08:41.11 particularly in relation to the hosts that they infect.
00:08:44.08 So, they can figure out quite quickly
00:08:46.25 what are the really central regulators of a pathway.
00:08:50.09 So, let's say, in the context of gene expression,
00:08:53.27 you're looking at a pathway that's got 10 or 20 or 30 components.
00:08:56.09 Well, if you track what the virus is targeting,
00:08:59.15 it will often lead you
00:09:03.05 directly to the center of the most important, or crux,
00:09:07.00 factors involved in this regulation.
00:09:09.29 So, I'm just gonna give you a couple of examples of this.
00:09:13.22 A classic example is the early studies
00:09:17.07 that were done with viruses
00:09:19.14 that really helped shape our understanding of events that lead to cancer.
00:09:22.15 Now, this comes from the fact that most DNA viruses
00:09:26.29 need to steal the host DNA replication machinery
00:09:31.00 in order to get their own genes expressed.
00:09:34.14 The problem for the virus is that this machinery
00:09:36.28 is generally only expressed in a cell
00:09:39.13 as it's undergoing S phase,
00:09:41.21 and not during all phases of the cell cycle.
00:09:44.09 And that's because transcription factors
00:09:46.20 that are required to turn on these genes
00:09:48.27 -- for example, the E2 family of transcription factors --
00:09:52.17 are held in an inactive state in the cytoplasm
00:09:55.13 by a tumor suppressor protein called retinoblastoma,
00:09:58.18 or Rb.
00:10:00.09 When the cell wants to enter the S phase, or the...
00:10:03.27 enter that phase of the cell cycle,
00:10:05.27 cyclin-dependent kinases will phosphorylate Rb
00:10:08.07 in a way that causes it to release E2F
00:10:10.29 so that factor can go into the nucleus
00:10:13.14 and transactivate those genes.
00:10:16.10 But if a virus is infecting a cell
00:10:18.10 that's not in S phase,
00:10:20.04 it needs to figure out how to get that cell
00:10:22.29 to turn on the machinery it needs
00:10:25.10 regardless of the cell cycle.
00:10:27.08 And so, it turns out that all viruses that have been studied,
00:10:30.10 that are DNA viruses that do this,
00:10:32.04 target the pathway in the same way.
00:10:35.08 And what they do is they each encode a protein
00:10:39.22 that binds retinoblastoma in a way
00:10:42.03 that forces it to release E2
00:10:44.22 so that that factor, or complex of factors,
00:10:47.03 can go into the nucleus and turn on these cell cycle genes.
00:10:49.27 Now, as an aside, E2 itself
00:10:52.25 was originally discovered
00:10:55.02 because it transactivates the early, or E, 2 region
00:10:59.03 of the adenovirus genome.
00:11:00.26 So, that also taught us something about cell cycle regulation.
00:11:04.16 Alright, now, this is one feature,
00:11:06.16 but it turns out that cells of course have ways of sensing
00:11:10.21 unscheduled entry into S phase
00:11:13.01 -- an unscheduled DNA synthesis --
00:11:15.02 and want to stop that,
00:11:17.21 because that's a hallmark of what will lead to cellular transformation.
00:11:20.25 And so, when this happens,
00:11:23.07 they activate a protein that's known as the guardian of the genome:
00:11:25.24 p53.
00:11:27.23 And p53's job is to stop entry into S phase,
00:11:30.28 and if it can't do that,
00:11:34.01 to lead the cell to a path of apoptosis, or cell death.
00:11:37.14 This of course would be detrimental to the virus,
00:11:39.11 because the virus needs to keep the cell alive
00:11:41.27 long enough to complete its replication cycle
00:11:44.23 and produce progeny virions.
00:11:46.21 So, not only do all of these viruses target retinoblastoma,
00:11:49.23 but all of these viruses
00:11:52.21 also have proteins that target and inactivate
00:11:56.11 the p53 tumor suppressor.
00:11:58.14 And in fact, p53 was originally discovered
00:12:00.26 as a 53 kiloDalton protein
00:12:03.01 that interacts with the large T antigen
00:12:07.03 from polyomavirus.
00:12:09.13 This is a protein that we now know...
00:12:11.23 or, a gene that we now know is mutated
00:12:13.26 in the majority of human cancers,
00:12:15.13 and is one of the most heavily studied genes in cancer biology.
00:12:20.06 And so, the fact that these small DNA tumor viruses,
00:12:23.15 even if they're viruses that themselves don't cause human cancer,
00:12:27.07 have taught us that inactivation of two key tumor suppressor proteins
00:12:31.13 is of fundamental importance
00:12:34.09 for cellular transformation
00:12:36.17 has really shed important light into cancer biology.
00:12:40.20 Another example is
00:12:44.16 viral manipulation of translational control.
00:12:47.27 So, every virus is absolutely dependent
00:12:52.05 on the host translation machinery
00:12:54.19 in order to get their genes translated into proteins.
00:12:57.18 No virus encodes genes that will make ribosomes,
00:13:02.06 which are key to protein synthesis.
00:13:04.20 The problem, though,
00:13:07.03 is that viruses have to compete
00:13:10.01 with the cellular messenger RNA
00:13:12.14 for access to these ribosomes.
00:13:14.28 And figuring out how they compete
00:13:17.13 has taught us a lot about plasticity in the translation complex,
00:13:21.21 or the translation apparatus.
00:13:23.21 So, they can do this for example by cleaving and inactivating cellular messenger RNAs
00:13:29.07 to essentially eliminate the competition.
00:13:31.12 Or they can selectively target factors that are part of the translation complex
00:13:36.07 that might be necessary for translation of many host RNAs,
00:13:40.16 but not necessary for translation of viral RNAs.
00:13:44.18 And I'm just showing you, here,
00:13:47.12 many examples of viral proteins from different viruses
00:13:50.19 that target individual components of the translation machinery.
00:13:53.20 Now, studying these inactivation events
00:13:56.10 has shed light on how different types of RNA
00:13:59.16 might be able to get targeted or translated
00:14:03.26 by distinct mechanisms.
00:14:05.29 For example, if an RNA
00:14:08.16 has a relatively unstructured 5' untranslated region, or UTR,
00:14:12.20 you might not need one of the key helicases
00:14:16.00 to unwind that RNA.
00:14:18.08 Or an RNA can get translated
00:14:20.25 without the 5' cap structure
00:14:23.23 if it has an internal ribosome entry site,
00:14:27.14 or IRIS element,
00:14:29.23 that allows direct recruitment of the ribosome subunits to the RNA
00:14:34.24 in the absence of a cap.
00:14:36.24 Or certain RNAs can be translated
00:14:39.12 by alternative or specialized ribosomes in a cell.
00:14:42.23 And it's notable that these are not things
00:14:46.08 that are happening only during viral infection.
00:14:49.10 You should think of viruses as thieves,
00:14:52.02 not inventors.
00:14:53.19 That means these things are happening normally in a cell,
00:14:55.29 and many of them have now been discovered
00:14:58.25 to be operational under certain cell contexts
00:15:01.01 like cell stress.
00:15:03.14 Finally, I just want to point out that
00:15:06.00 viruses and viral components
00:15:08.28 are centerpieces of many of the tools
00:15:12.18 that we use on a daily basis in molecular biology.
00:15:15.08 Let's just take a standard plasmid vector as an example.
00:15:21.07 This is a generic vector that you might use,
00:15:23.27 or scientists might use,
00:15:25.27 in order to express a gene of interest in cells,
00:15:29.12 by transfecting the vector into cells
00:15:31.24 or using a virus to transduce the vector into cells
00:15:35.11 to express their gene.
00:15:37.02 Well, if you've ever studied plasmid vector maps,
00:15:39.00 you might appreciate that they are
00:15:41.24 chock full of things that people have gotten from viruses.
00:15:45.26 For example, your gene of interest
00:15:48.24 is probably going to be transcribed by a promoter
00:15:52.26 that comes from a virus --
00:15:55.08 a really popular one is the cytomegalovirus immediate-early promoter.
00:15:59.07 Your vector might have promoter sequences
00:16:02.10 that allow you to generate RNA by in vitro transcription as well --
00:16:07.12 things like T7 and SP6 promoters.
00:16:10.25 Well, these are derived from bacteriophages
00:16:13.18 and are used in the lab to be transcribed
00:16:16.21 by T7 and SP6 polymerases,
00:16:19.03 which are also viral.
00:16:22.00 If you want to purify your protein from a cell, or detect it,
00:16:25.09 often you're gonna append an epitope tag to that protein.
00:16:29.03 Most of these epitope tags also come from viruses --
00:16:32.03 the FLAG tag, the HA tag,
00:16:35.07 these are from influenza;
00:16:37.04 the V5 tag from a paramyxovirus.
00:16:41.24 And if you want to then purify your protein and remove a tag,
00:16:45.24 you'll usually have a proteolytic cleavage site
00:16:48.03 that allows you to specifically remove that tag from the protein.
00:16:51.04 These tend to also come from viruses --
00:16:53.21 the TEV protease site, for example,
00:16:56.12 from tobacco etch virus.
00:16:59.11 Even the multiple cloning sequence
00:17:02.01 that you're gonna use to put your gene of interest into the vector...
00:17:04.27 well, restriction enzymes come from the immune system of bacteria,
00:17:09.22 that's used to fight off phage invaders.
00:17:13.14 And as your gene is being transcribed,
00:17:16.03 it's going to need to get the features added on to it
00:17:19.17 that are going to allow it to be recognized by the host translation machinery,
00:17:22.05 and one of these features is a poly(A) tail,
00:17:24.17 so you're probably gonna have a polyadenylation signal sequence,
00:17:27.15 and these were originally mapped in viruses as well.
00:17:31.05 So, something like the SV40 poly(A) sequence
00:17:34.04 is probably a component of these vectors.
00:17:36.17 And then finally,
00:17:39.01 do you ever wonder why 293T cells
00:17:41.11 are such workhorses for protein expression in biology?
00:17:43.25 It's because these cells express the large T antigen
00:17:48.05 from polyomavirus,
00:17:50.21 and that T antigen, if your vector has an SV40 origin of replication,
00:17:55.22 or ori,
00:17:56.29 will recognize that component or that sequence on your vector
00:18:00.27 and amplify the vector, and replicate it,
00:18:03.23 so that you can have a dramatic expansion of the number of templates
00:18:08.08 available in the cell for transcription.
00:18:10.18 And an added bonuses is, then,
00:18:12.24 as these 293T cells are dividing in culture,
00:18:14.26 your plasmid isn't getting diluted out,
00:18:17.00 because it's being continually amplified.
00:18:20.17 So, these and many other viral discoveries
00:18:25.10 have been recognized for their importance
00:18:28.05 through the receipt of many Nobel Prizes,
00:18:30.25 both for things that are
00:18:34.11 directly related to human health and of translational importance,
00:18:38.08 like vaccines,
00:18:40.09 but also more broadly for the spectrum of discoveries
00:18:42.19 that have been made using viruses
00:18:45.04 that have greatly enhanced our understanding of things
00:18:48.09 like gene expression and the immune system and cancer.
00:18:52.10 So, I'm now going to transition into an aspect
00:18:58.16 of virus host interactions
00:19:01.10 that is really the focus of my lab's research,
00:19:03.10 and it's gonna be something that I'm gonna be going into more detail on
00:19:06.18 in the second lecture,
00:19:08.21 and this is gene regulation,
00:19:10.17 and in particular, controlling the abundance of messenger RNA in cells.
00:19:14.28 Now, in this particular diagram,
00:19:16.17 if you think of messenger RNA as a pool of water,
00:19:20.04 and the amount of messenger RNA in this cell
00:19:23.02 is the amount of water
00:19:25.02 that's being held in this person's hands,
00:19:26.24 well, transcription of course is the process, that faucet,
00:19:30.22 that feeds that pool of messenger RNA.
00:19:33.15 And for many years,
00:19:36.05 people thought of degradation of RNA
00:19:38.08 as being this unfortunate consequence of loss of the RNA
00:19:40.21 -- the spilling of the water out of the pool --
00:19:42.24 which is why you needed to continually replenish it
00:19:46.19 through the act of transcription.
00:19:48.21 Now, however, it's well appreciated that RNA degradation
00:19:52.04 is a real focal point for gene expression control,
00:19:56.15 and that as many as half of the gene expression changes
00:19:59.29 that occur in a cell in response to particular stimuli
00:20:03.15 can be driven by alterations in RNA stability.
00:20:06.27 This is either at the individual transcript level
00:20:09.05 or to the bulk pool at large.
00:20:13.01 And for me as a virologist,
00:20:14.28 one of the things that really helps hammer home this point
00:20:17.11 is that many different viruses
00:20:21.02 that are unrelated to each other
00:20:22.28 -- from herpesviruses, to poxviruses,
00:20:25.28 to influenza viruses,
00:20:30.04 to SARS and MERS coronavirus --
00:20:33.12 all help reshape the gene expression landscape of a cell
00:20:38.06 during infection
00:20:40.23 by targeting the process of RNA decay.
00:20:44.06 They do this by a number of mechanisms,
00:20:46.15 but there are some overt similarities.
00:20:48.18 So, let me just summarize for you
00:20:50.11 some of the ways that the viruses do this.
00:20:52.10 Well, herpesviruses,
00:20:54.18 they encode factors that are able to internally cleave RNAs,
00:20:59.15 so these are called endonucleases.
00:21:01.27 And this inactivates the RNA, right away, for translation.
00:21:06.19 Influenza viruses, shown here in the nucleus of the cell,
00:21:08.25 also have an endonuclease
00:21:11.11 that preferentially targets RNAs undergoing splicing.
00:21:14.07 Poxviruses like vaccinia
00:21:17.15 have factors called decapping enzymes
00:21:19.22 that remove the protective cap
00:21:21.28 from the 5' end of an RNA.
00:21:24.17 And coronaviruses, like SARS and MERS coronavirus,
00:21:27.21 have factors that will cause a ribosome to stall on an RNA,
00:21:32.10 leading to an internal, or endonucleolytic, cleavage,
00:21:35.22 and inactivation of that RNA.
00:21:38.12 Now, an important point for all of these is that
00:21:43.09 you are rapidly inactivating the RNA
00:21:46.06 by cleaving it internally or removing its 5' protective cap.
00:21:50.03 And so, in doing so,
00:21:52.08 the RNA is immediately not able to be translated anymore.
00:21:55.13 But also, what you're doing
00:21:58.24 is you're exposing the ends of the RNA
00:22:01.09 to rapidly being accessed by host RNA degradation enzymes.
00:22:07.20 And these are called exonucleases
00:22:09.09 because they will degrade an RNA from either end,
00:22:12.04 rather than internally.
00:22:14.13 And major exonucleases in mammalian cells
00:22:18.25 are the XRN1 exonuclease
00:22:20.27 that chews things up from the 5' end
00:22:22.28 and things like DIS3L2 and the exosome,
00:22:26.05 which chew them up from the 3' end.
00:22:28.22 So, all of these viruses
00:22:31.02 are relying on this basal RNA decay machinery
00:22:33.12 to degrade the fragments
00:22:36.01 that are being created by the viral enzymes.
00:22:39.10 It's important to note that
00:22:42.28 this is reiterative targeting of the same pathway
00:22:45.13 by multiple different viruses.
00:22:48.07 And whenever you see this in virology,
00:22:50.24 it tells you that something about the way that they're doing it
00:22:54.20 must be very efficient,
00:22:56.21 because different viruses are solving the problem
00:22:59.10 by hitting on the same thing.
00:23:01.22 And why that must be efficient in an RNA degradation example
00:23:07.24 becomes obvious if we compare how viruses are doing this
00:23:10.19 -- or RNA decay during a viral infection --
00:23:13.06 to RNA degradation that's happening sort of normally --
00:23:18.24 basal RNA decay in our cells.
00:23:20.23 So, messenger RNAs are protected at their ends.
00:23:23.26 They have a 5' cap on one end
00:23:25.29 and a poly(A) tail on the other end.
00:23:29.13 And RNA decay is a very ordered process
00:23:33.13 that begins with gradual shortening of that poly(A) tail.
00:23:37.12 This happens by a number of enzymes that are called deadenylases,
00:23:42.08 and it's a rate-limiting and heavily regulated step in the cell.
00:23:46.08 Once the poly(A) tail has reached a certain shortness,
00:23:50.07 this licenses removal of the 5' cap from the RNA
00:23:55.00 by a decapping complex.
00:23:56.28 And now, only once you've got a headless, tailless RNA,
00:24:01.23 can that RNA be accessed
00:24:04.23 by these fast-acting exoribonucleases,
00:24:08.08 like XRN1 and the exosome.
00:24:11.09 So, that's what happens in an uninfected cell.
00:24:13.17 But in an infected cell,
00:24:15.29 viral nucleases, by internally cleaving the RNA,
00:24:19.27 are bypassing these rate-limiting, regulated steps of RNA decay
00:24:25.11 -- of deadenylation and decapping --
00:24:27.28 and creating RNA substrates that can be directly accessed
00:24:32.08 by these fast-acting exonucleases.
00:24:35.04 And because they're relying on the host machinery
00:24:37.22 to degrade those cleaved fragments,
00:24:41.06 the intermediates are going to look indistinguishable
00:24:43.19 from those that are normally produced in the cell,
00:24:45.22 and might not raise any red flags for the cell.
00:24:49.19 I should note that this strategy that the virus is using, of using an endonuclease...
00:24:55.00 well, there are of course examples of endonucleases
00:24:57.00 that are used in the host cell in an uninfected context.
00:25:00.17 Many of these are used in the context of quality control.
00:25:04.18 If you've got an RNA with an error,
00:25:06.18 you want to get rid of it quickly,
00:25:08.27 before it could get translated into a protein
00:25:11.06 that has a mutation
00:25:13.09 that might have a dominant-negative phenotype.
00:25:14.24 And so in these cases,
00:25:16.24 the cell uses its own endonucleases
00:25:19.07 to recognize and cleave and inactivate those RNAs,
00:25:23.12 ultimately following through this endonucleolytic followed by XRN1-mediated decay pathway
00:25:28.09 that the virus is using.
00:25:30.05 The difference, though, is that during viral infection,
00:25:33.15 cleavage is not limited to a few RNAs.
00:25:37.07 The scope is much, much broader.
00:25:40.13 So, I'll just close this portion of the talk
00:25:42.25 by emphasizing that this ability of viruses
00:25:45.26 to accelerate RNA decay
00:25:48.10 plays a variety of different roles in the context of viral infection.
00:25:53.12 A few examples that are notable
00:25:55.22 are that RNA degradation
00:25:58.16 is known to help viruses escape immune detection.
00:26:02.24 So, for example, in response to an infection,
00:26:06.05 a cell has many ways of sensing that infection,
00:26:09.08 and the cell is going to start to turn on factors
00:26:12.04 that are immune stimulatory and antiviral.
00:26:15.26 And so, as it produces these RNAs for translation into antiviral factors,
00:26:20.22 by cleaving all of these RNAs in the cell,
00:26:25.05 the virus can prevent production of those antiviral factors
00:26:28.09 and help escape immune detection.
00:26:31.23 Shown on the bottom, here,
00:26:34.07 is an example of translational control
00:26:36.17 that I already sort of touched on.
00:26:38.15 If the virus is cleaving many cellular RNAs,
00:26:41.12 then these RNAs are not competing with the virus
00:26:44.17 for access to ribosomes.
00:26:46.03 And this can, in a number of cases,
00:26:48.13 enhance the ability of viral genes to get translated
00:26:53.01 in an environment where cellular genes might normally be favored.
00:26:58.10 In certain viral examples, this RNA degradation
00:27:00.22 also helps with the ability of the virus to
00:27:04.13 transition between different gene expression cascades
00:27:06.15 that are needed in the course of infection.
00:27:09.06 And in in vivo mouse models,
00:27:11.08 it's been shown that this can help the virus
00:27:14.15 traffic to where it needs to go,
00:27:16.13 and in... in the animal,
00:27:18.26 and replicate in the appropriate set of cells.
00:27:22.07 So, what I'm gonna talk about in the next lecture
00:27:25.09 is more scientific detail on what we've learned
00:27:28.21 about RNA targeting by a particular viral endonuclease,
00:27:32.09 and how that's informed our knowledge
00:27:35.01 of some of the different connections
00:27:37.25 between different stages of the gene expression cascade.
00:27:41.24 So, I'd like to end this talk by acknowledging
00:27:44.20 the many generous funders
00:27:47.12 that have helped support our research over the years,
00:27:50.07 as well as the fantastic scientists
00:27:53.01 that are part of my lab.
00:27:54.27 The current members are shown here in these photos,
00:27:56.23 but we have a long history of fantastic scientists
00:28:00.00 that have worked with us on this.
00:28:02.02 So, thank you very much for listening.

Part 2: KSHV: Herpesviral Nucleases Impact Cellular RNA Destruction and Synthesis

00:00:07.21 Hello again. I'm Britt Glaunsinger.
00:00:10.17 I'm a Professor at the University of California, Berkeley
00:00:13.01 and an Investigator of the Howard Hughes Medical Institute.
00:00:16.09 And in my first lecture,
00:00:18.13 I gave some general information
00:00:20.25 about why viruses have been such fantastic teachers of biology.
00:00:24.26 And in this lecture,
00:00:26.14 I'm gonna go into some more specifics
00:00:28.20 on research from my own lab
00:00:31.05 about how herpesviral infection
00:00:33.10 alters the gene expression landscape of host cells,
00:00:35.01 and how we can use these viruses as tools
00:00:38.25 to better understand gene regulation in mammals.
00:00:42.18 So, herpesviruses are endemic in the human population.
00:00:47.10 Nearly every single human is infected with at least one,
00:00:51.14 and often more than one, type of herpes
00:00:54.29 that they start acquiring relatively early in life.
00:01:00.04 Now, we're not the only ones infected with herpes.
00:01:02.04 It's estimated that probably most vertebrates
00:01:04.00 have at least one herpesvirus associated with them.
00:01:06.14 And we now know that even invertebrates,
00:01:09.01 like oysters and coral,
00:01:10.29 can have herpesviruses.
00:01:13.00 Now, these are not the same herpesviruses as that...
00:01:15.27 those that infect humans
00:01:18.06 -- herpesviruses are very species-specific --
00:01:20.28 but nonetheless, it should give you a sense that herpesviruses
00:01:24.05 are really widespread on the planet.
00:01:27.18 And they're also really ancient.
00:01:29.18 It's estimated that these viruses
00:01:31.24 have been around and evolving
00:01:34.04 for close to 200 million years.
00:01:36.13 And that's one of the reasons
00:01:38.19 why we think that they make excellent tools
00:01:40.20 to understand virus-host interactions.
00:01:43.15 Now, the fact that most humans
00:01:47.10 are infected with at least one herpesvirus
00:01:49.06 might sound alarming...
00:01:51.20 except for the observation that in the majority of these cases
00:01:55.10 the infections are pretty much asymptomatic
00:01:58.02 or produce only mild disease.
00:02:00.04 And that's because your immune system
00:02:02.14 does an excellent job of keeping them at bay.
00:02:05.11 However, under conditions
00:02:08.02 where immune surveillance is impaired,
00:02:10.01 herpesviruses can be devastating.
00:02:12.21 So, this can occur, for example,
00:02:14.16 in the elderly, as immune surveillance begins to wane.
00:02:19.06 Individuals can have herpesviral reactivations,
00:02:21.21 for example, that cause shingles.
00:02:24.02 Or when a neonate is infected with a herpesvirus...
00:02:30.00 these types of congenital infections that occur
00:02:32.11 when an immune system is not yet fully developed
00:02:35.01 can cause extremely serious, lifelong diseases.
00:02:38.12 Or when an individual has another disease or disorder
00:02:42.09 that causes immune suppression,
00:02:45.10 for example HIV and AIDS.
00:02:47.20 Under these circumstances, herpesviruses are real killers.
00:02:51.18 My lab studies a subfamily of herpesviruses
00:02:54.05 called gammaherpesviruses.
00:02:56.22 These are illustrated in this picture, here.
00:02:59.18 They have two human virus members:
00:03:02.01 Epstein-Barr virus and Kaposi's sarcoma-associated herpesvirus,
00:03:06.18 or KSHV.
00:03:08.26 Both of these viruses cause human cancers.
00:03:11.13 They can cause types of B cell cancer.
00:03:13.23 And KSHV is the etiologic agent
00:03:16.17 of the main AIDS-defining illness of the 1980s,
00:03:18.23 Kaposi's sarcoma.
00:03:21.16 KSHV also remains a very common cause of cancer
00:03:25.14 in areas of sub-Saharan Africa
00:03:27.22 -- one of the dominant causes of cancer in this area --
00:03:29.23 due to the huge HIV endemicity there.
00:03:33.29 So, any herpesvirus infection begins with
00:03:40.10 the virus needing to deposit its double-stranded DNA genome into the nucleus.
00:03:42.17 In general, these viruses are not integrating
00:03:45.16 into the host genome like a retrovirus does.
00:03:47.19 But instead, the genome is being circularized,
00:03:50.26 into what's called an episome,
00:03:53.03 and tethered to the host chromatin.
00:03:56.00 In this state, they're relatively quiescent.
00:03:58.23 This is a state called latency.
00:04:00.23 And their ability to exist in this quiescent state
00:04:02.27 is the reason why if you're infected with a herpesvirus,
00:04:07.06 even early in life, you will never clear that virus.
00:04:10.19 That's because this quiescent state
00:04:12.27 is basically immunologically silent.
00:04:14.26 Your immune system doesn't know the viruses there,
00:04:17.12 so it doesn't know which cells to eliminate.
00:04:20.17 Now, in this quiescent state,
00:04:23.00 the virus is not actively replicating.
00:04:25.02 So, in order for it to spread
00:04:28.06 between cells or between hosts,
00:04:30.08 it needs to reactivate from this latent state
00:04:32.27 and engage in active, or lytic, replication.
00:04:37.07 This is the phase of the viral life cycle
00:04:40.11 when there are dramatic changes to the host gene expression environment
00:04:43.29 -- as the virus is commandeering and manipulating and hijacking host machinery
00:04:49.06 to get its own genes expressed, and its genome replicated
00:04:52.02 so that it can produce progeny virions that will go on
00:04:55.02 and spread to neighboring cells.
00:04:57.09 It's this phase of the viral life cycle
00:04:59.14 that my lab is focused on
00:05:02.13 and in understanding, what are some of these dramatic changes
00:05:04.14 and how can we use them to understand gene regulation
00:05:06.22 in human cells?
00:05:08.24 It turns out that early on we showed that
00:05:12.02 one of the relatively dramatic changes that is happening
00:05:15.24 during lytic gammaherpesvirus infection
00:05:17.27 is that there's widespread depletion of messenger RNA,
00:05:20.12 largely from the cytoplasm of these cells.
00:05:23.13 Something like 60-80% of messages
00:05:26.14 are depleted during lytic gammaherpesvirus infection.
00:05:29.24 And this is carried out by
00:05:32.16 a virally encoded nucleus
00:05:35.12 that in KSHV is called SOX.
00:05:37.09 SOX is a factor that plays
00:05:40.06 a number of important roles in the context of the viral life cycle,
00:05:43.07 and its RNA degradation activity
00:05:46.23 is important for helping the virus
00:05:49.09 maintain its normal gene expression cascade,
00:05:52.03 and escape the immune system,
00:05:54.09 and replicate in the context of an in vivo host.
00:05:57.21 And for these and other reasons,
00:05:59.22 we've been really interested in understanding
00:06:02.23 how SOX recognizes its targets
00:06:04.25 and how we can use that information to study gene regulation.
00:06:08.03 And so, my talk here will be broken into two sections.
00:06:13.00 The first is detailing our efforts
00:06:15.26 to understand this initial targeting event --
00:06:18.25 how does a viral nucleus target such a breadth of RNAs?
00:06:22.10 And in the second part of my talk,
00:06:24.25 we're gonna take a step back and say,
00:06:26.26 how do we use this information,
00:06:29.03 this nuclease as a tool,
00:06:30.24 to better understand how cells sense and respond
00:06:34.01 to large changes in messenger RNA abundance?
00:06:39.04 So, when thinking about SOX targeting,
00:06:43.16 we were faced with somewhat of a paradox.
00:06:46.06 We knew that the majority of messenger RNAs in a cell
00:06:50.11 could be targeted by this nuclease,
00:06:52.10 yet we also knew that targeting
00:06:55.29 seemed to occur at specific sites.
00:06:59.03 And contrary to what one might expect
00:07:01.20 for a nuclease that targets many RNAs,
00:07:05.09 you might think that the sites that it would recognize
00:07:07.24 are relatively simple --
00:07:09.27 maybe a preferred dinucleotide, for example.
00:07:11.29 But this was not the case at all.
00:07:14.08 When we mapped targeting sites on individual reporter RNAs for SOX,
00:07:18.27 we required a pretty large amount of sequence context
00:07:22.15 -- 25 nucleotides to sometimes 100 or more nucleotides --
00:07:26.23 that we had to insert into a new location in the RNA
00:07:31.08 to direct SOX cleavage there.
00:07:32.29 So, how do you have this seemingly complicated recognition element,
00:07:37.07 yet one that has to be present on most messenger RNAs?
00:07:40.25 The other unusual feature is that
00:07:44.18 while it wasn't recognizing an RNA
00:07:46.27 based on relative position to a landmark feature
00:07:49.27 -- like cleaving next to the 5' cap,
00:07:52.12 or at a splice junction, or near a termination codon --
00:07:55.17 we found that these recognition sites
00:07:58.21 could really be anywhere along the length of an RNA.
00:08:00.21 So, to understand this seemingly,
00:08:03.15 you know, paradoxical situation of specificity with breadth,
00:08:07.07 we reasoned that we needed to know,
00:08:10.12 at a much larger scale,
00:08:12.16 the exact cleavage sites that this nuclease
00:08:14.26 had on the cellular transcriptome.
00:08:17.14 Identifying cleavage intermediates
00:08:20.10 is notoriously challenging.
00:08:22.10 And that's because, as I mentioned in my first talk,
00:08:24.29 RNAs are protected at either end
00:08:27.01 by a cap and a poly(A) tail,
00:08:29.00 and once these are removed
00:08:31.15 RNAs are very rapidly degraded
00:08:33.20 by host exonucleases like XRN1.
00:08:35.28 So, if you're trying to trap a cleavage intermediate,
00:08:38.19 say, that SOX is generating,
00:08:40.14 as soon as that intermediate as gen... is generated,
00:08:44.17 these host exonucleases are gonna chew up that intermediate,
00:08:47.18 so it's gonna be hard to see that initial cleavage event.
00:08:50.13 However, we had known for our previous research
00:08:53.06 that XRN1 played a really important role
00:08:56.03 in getting rid of this 3' fragment
00:08:59.16 that is generated upon SOX cleavage,
00:09:01.19 so we found that if we deleted XRN1
00:09:04.05 or depleted it from cells,
00:09:06.06 we could stabilize this degradation intermediate
00:09:08.21 and then identify these cleavage intermediates
00:09:11.23 that represented the initial cleavage event genome-wide,
00:09:15.07 using a technique called parallel analysis of RNA ends, or PARE-seq.
00:09:20.17 In doing this analysis and then asking,
00:09:23.03 are there similarities in the sequences
00:09:26.19 surrounding all of these cleavage sites?,
00:09:28.08 what we found is that while it's true that the cleavage sites
00:09:32.04 seem to be relatively large,
00:09:34.09 they're anchored not by a strong, complicated sequence consensus
00:09:38.15 but instead by a few anchor residues.
00:09:42.07 And I'm gonna highlight a couple of these anchor residues
00:09:45.26 that we now know to be quite important
00:09:48.07 for targeting.
00:09:49.28 The first is a stretch of adenosines
00:09:53.08 that is just upstream of the cleavage site.
00:09:56.04 And our analysis of the folding prediction of RNA
00:10:00.26 surrounding this site
00:10:03.03 suggested that these adenosines needed to be unpaired --
00:10:05.25 so, not sort of in a base-paired configuration.
00:10:08.24 And the second was that right at the cleavage site
00:10:11.13 we could usually see nucleotides
00:10:14.12 that were C's or T's or sometimes A's,
00:10:16.24 but there was never a G there.
00:10:18.22 And so, it suggested to us that
00:10:21.26 there was probably a combination of sequence and structure
00:10:25.15 that went into recognition by this viral nuclease.
00:10:29.13 So, the first thing that we did is we took one of these RNAs
00:10:32.27 that we had identified from our PARE-seq analysis in cells
00:10:35.25 -- it's an RNA called LIMD1 --
00:10:38.21 that we knew had a really robust SOX targeting site.
00:10:42.01 And we used in-line probing to determine
00:10:44.29 the structure of this RNA.
00:10:46.22 And what we found was that, indeed,
00:10:49.12 as the predictions had suggested,
00:10:51.08 this stretch of A's, which I've circled here,
00:10:54.00 is indeed in this loop region
00:10:57.06 -- so, unpaired --
00:10:58.29 and just to orient you, the cut site here
00:11:00.24 is a couple of nucleotides downstream from that.
00:11:02.26 So, knowing this RNA's structure,
00:11:06.03 we could now go in and say,
00:11:08.05 why are these nucleotides important,
00:11:10.06 why are these sequences important for SOX targeting?
00:11:12.29 Or are they important for SOX targeting?
00:11:15.05 So, we made a series of mutations on this targeting site,
00:11:18.05 and three mutations in particular I'm gonna show you.
00:11:21.29 First, we made a mutation where we changed
00:11:26.10 that cut site to the non-preferred G,
00:11:28.13 just to see, does it really matter if you have a G there or not?
00:11:32.03 And the second two mutations had to do with testing
00:11:35.03 the importance of that stretch of A's.
00:11:37.07 We either mutated that stretch of A's
00:11:39.29 to a stretch of G's instead.
00:11:41.26 Or, what we did was we kept the A's there
00:11:46.12 but created a series of complementary base-pairing interactions
00:11:50.08 to zipper them up
00:11:53.09 so they were no longer present in an open conformation.
00:11:55.07 And this allowed us to distinguish was,
00:11:57.05 is the sequence important?
00:11:59.06 And/or, is their structural presentation important?
00:12:03.28 And so, what we did first was we asked,
00:12:06.08 how well does the enzyme recognize these RNAs
00:12:09.03 if we change these features?
00:12:11.00 What is its catalytically...
00:12:13.01 activity like on these RNAs?
00:12:15.08 And we mixed the RNA -- wild type or mutant substrates --
00:12:18.18 with purified SOX protein
00:12:20.19 and then measured its catalytic activity,
00:12:22.17 as shown in this graph here.
00:12:24.13 And what you can see from the first bar
00:12:26.14 -- this is the wild type RNA --
00:12:28.24 is that the enzyme has high catalytic activity on this substrate.
00:12:33.25 It cleaves the RNA very well.
00:12:35.24 However, for each of the mutants that we made,
00:12:38.02 there was a significant impairment
00:12:40.28 in SOX's catalytic activity on these RNAs,
00:12:43.25 indicating that, indeed,
00:12:46.07 these sequence features that we had pulled out of the sequence analysis
00:12:49.19 are key to the viral nuclease recognition.
00:12:53.23 So, what are they doing for recognition?
00:12:56.05 To answer that question,
00:12:58.22 we next wanted to evaluate how well SOX bound these RNA substrates.
00:13:02.20 Now, this might seem intuitive,
00:13:04.20 but in fact, in the field,
00:13:06.12 it had been assumed for many years that this enzyme
00:13:08.06 didn't have any robust RNA-binding activity.
00:13:12.28 And we thought, perhaps this was the case
00:13:15.08 because people weren't testing it
00:13:18.14 using a bona fide targeting RNA element.
00:13:21.01 And so, we measured the binding capacity of the enzyme
00:13:25.19 to this element
00:13:28.02 using a technique called bio-layer interferometry, or BLI,
00:13:30.06 which allows us to measure real time binding kinetics.
00:13:33.23 And these are measured as dissociation constants.
00:13:36.19 And the key to remember about Kd's
00:13:40.16 is that the smaller the number,
00:13:42.18 the tighter the binding.
00:13:44.13 So, when we measured SOX binding to the wild type RNA,
00:13:48.04 the first important thing that we noted,
00:13:51.08 which was in this dark blue bar here,
00:13:53.13 is that in fact it bound this RNA very tightly.
00:13:56.28 The dissociation constant was about 8 or 10 nanomolar.
00:14:00.02 So, indeed, it is a very robust RNA-binding enzyme.
00:14:04.21 If we mutated that cut site G,
00:14:07.08 which impairs its catalytic activity,
00:14:09.08 that's shown in the second bar,
00:14:11.16 binding is fine.
00:14:13.13 So, we can separate binding from catalysis here.
00:14:16.08 However, either of the mutants
00:14:19.05 where we changed the stretch of adenosines
00:14:21.20 or zippered them up
00:14:23.18 -- these are the third and the fourth bars here --
00:14:26.28 these mutants severely impaired binding,
00:14:29.27 pushing it into the micromolar range,
00:14:32.03 which indicates that this loop structure
00:14:35.26 with the adenosines presented in the loop,
00:14:38.07 serves as a landing pad,
00:14:40.19 and in fact a very specific landing pad,
00:14:44.06 for this enzyme.
00:14:47.13 And we've confirmed this using things like RNA footprinting analyses.
00:14:50.14 So, what this has told us is this type of motif,
00:14:53.21 where you have a structure
00:14:56.11 anchored by a few key residues,
00:14:58.15 probably helps resolve this paradox
00:15:02.09 of specificity with breadth.
00:15:04.01 Because it enables the enzyme
00:15:06.22 to target RNAs at very specific sites,
00:15:09.07 but because the structure itself
00:15:11.12 is not particularly complex,
00:15:13.04 and you don't require a large amount of sequences
00:15:15.18 that have to always be the same,
00:15:17.16 you can target a breadth of RNAs.
00:15:21.27 Why do we care about understanding
00:15:25.01 how a nuclease like this recognizes RNA?
00:15:28.19 There are a few reasons.
00:15:30.12 First, we know from viral infections
00:15:33.08 that not all RNAs are targeted
00:15:36.11 for cleavage by a virus.
00:15:38.07 In fact, you might imagine that there are RNAs
00:15:40.29 that the virus needs to keep around
00:15:43.06 to produce machinery that the virus needs to use.
00:15:46.10 And so, understanding how those are escaping
00:15:50.10 can't be done if you don't know how the nuclease
00:15:53.23 targets an RNA in the first place.
00:15:55.17 And we can now appreciate that probably
00:15:59.17 a large subset of these RNAs are escaping
00:16:02.01 because they don't form this correct sequence-structure combination.
00:16:06.11 However, we also know that
00:16:10.10 there are subsets of RNAs
00:16:12.16 that even if we give them a perfect SOX target recognition site,
00:16:15.05 they don't get cleaved.
00:16:17.21 And it turns out this is because
00:16:20.17 they have an extra protective sequence,
00:16:22.28 which we now call a nuclease escape element,
00:16:26.14 that is found in their 3' untranslated region.
00:16:30.05 And something about this sequence
00:16:32.12 actively guards the RNA against cleavage by SOX
00:16:36.00 and, as it turns out,
00:16:39.05 against cleavage by many other viral endonucleases.
00:16:42.27 So, understanding this interplay
00:16:46.28 is helping us understand how viruses can reshape
00:16:49.19 the gene expression environment,
00:16:51.13 and maybe how the cell can fight back against them.
00:16:55.10 So, I'm gonna to shift to the second portion of my talk now,
00:16:58.10 and tell you about how we've used SOX as a tool
00:17:02.11 to better understand
00:17:04.27 how cells sense and respond to these large changes in RNA abundance
00:17:07.15 that we know can happen
00:17:09.25 during a number of different viral infections.
00:17:12.10 And in thinking about this,
00:17:14.23 we were really intrigued by the idea
00:17:17.11 that there might be surprising connections
00:17:20.03 between different stages of the gene expression cascade.
00:17:23.05 Classically, people think about gene expression
00:17:26.12 as a pretty linear series of events,
00:17:28.06 beginning with RNA production
00:17:31.23 through transcription in the nucleus;
00:17:33.12 processing of that RNA by capping and splicing and polyadenylation;
00:17:37.13 and then export of that RNA into the cytoplasm,
00:17:40.28 where it's translated by ribosomes into proteins;
00:17:42.23 and then ultimately, at the end of the RNAs life,
00:17:45.16 it's degraded.
00:17:47.26 However, it's now appreciated
00:17:50.17 that in fact this is not really a linear series of event.
00:17:53.07 There's lots of crosstalk that can occur
00:17:55.07 between these different stages.
00:17:57.23 And crosstalk that I'm gonna tell you about today
00:18:01.06 connects, actually, the very last stage of the RNA's life,
00:18:03.29 that of RNA degradation,
00:18:05.27 with the very first stage of its life,
00:18:08.25 that of its production, or transcription in the nucleus.
00:18:11.23 Our thinking about this has been
00:18:15.12 really informed by lots of work that's been done
00:18:17.23 in the yeast system,
00:18:19.11 where it's been shown that, in fact,
00:18:21.07 cells have ways of sensing changes in RNA decay
00:18:26.00 or changes in the rate of RNA synthesis,
00:18:28.24 and altering one of those two rates
00:18:31.25 to compensate for those changes.
00:18:33.20 The idea here is that if you had, for example,
00:18:36.27 less RNA produced,
00:18:38.20 then you'd want to keep that RNA around longer
00:18:41.06 to have the same amount of genes
00:18:44.06 that you needed in the translation pool.
00:18:46.00 This is called the homeostatic model,
00:18:48.07 and there are examples of it in mammalian cells as well.
00:18:52.09 However, we thought that maintaining homeostasis
00:18:56.12 might not be the goal of the cell in the context of viral infection.
00:19:00.24 And so, what happens when you greatly accelerate RNA decay,
00:19:04.08 as many viruses do?
00:19:07.08 Does the cell try and compensate in a homeostatic way?
00:19:09.22 Or does it view this as a threat?
00:19:13.12 And so, to test this,
00:19:15.06 what we wanted to do was ask,
00:19:16.29 how does RNA degradation
00:19:19.25 influence transcription?
00:19:21.12 And we could do this just by expressing these viral nucleases.
00:19:25.07 So, an experiment that I'm gonna show you here
00:19:28.09 is one in which we take cells and we transfect them with an empty vector;
00:19:32.03 or a vector encoding a viral endonuclease, like SOX;
00:19:35.12 or, another one, vhs,
00:19:38.06 which is from herpes simplex virus,
00:19:40.09 also broadly cleaves RNA
00:19:42.16 but is not related to the SOX nuclease;
00:19:45.18 or a catalytically dead version of SOX.
00:19:48.28 And then we measure transcription in these cells,
00:19:52.06 either by looking at nascent RNA production
00:19:53.28 using a technique called 4-thiouridine labeling,
00:19:56.17 or 4sU,
00:19:58.02 which is shown in this experiment,
00:19:59.20 or by measuring occupancy of RNA polymerase at host promoters.
00:20:04.28 This is called RNA pol II ChIP,
00:20:07.19 and I'm gonna show you that in some later experiments.
00:20:10.12 So, what you should notice from this first experiment, though,
00:20:14.15 is that when we add SOX or vhs to cells,
00:20:19.07 that drives RNA degradation,
00:20:21.07 and we measure RNA synthesis,
00:20:23.13 we can see that in these cells RNA synthesis
00:20:25.28 is greatly reduced.
00:20:27.24 This relies on the catalytic activity of the enzyme,
00:20:30.13 because we don't see impaired RNA synthesis
00:20:33.19 if we add a catalytically dead version of the SOX gene to these cells.
00:20:38.29 So, this suggests that eliciting widespread RNA decay in the cytoplasm
00:20:43.28 does something to greatly reduce transcription in the nucleus.
00:20:49.10 How might this happen?
00:20:51.22 We thought of a couple of different possibilities, here,
00:20:54.18 which are not mutually exclusive.
00:20:57.02 One is that you can envision a scenario
00:21:00.20 where cleavage of particular RNAs matters.
00:21:03.04 For example, RNAs that themselves
00:21:06.03 are involved in transcriptional regulation
00:21:08.06 or transcriptional control.
00:21:09.25 You get rid of those; they can't be translated into protein;
00:21:12.14 if the protein that's already around in the cell is labile and gets degraded,
00:21:17.03 then it's just... its levels fall, and you impair transcription.
00:21:20.20 A second model is that cells have some way of actually sensing, or detecting,
00:21:28.09 the active degradation of RNAs in general --
00:21:32.21 the depletion of the messenger RNA pool.
00:21:34.24 And here, just cleavage of the RNA,
00:21:38.23 as the viral enzyme is doing,
00:21:41.14 might not be sufficient.
00:21:43.13 You might need to actually have degradation of the RNA,
00:21:45.15 which we know is carried out by the basal cellular enzymes.
00:21:49.19 So, we could distinguish between these possibilities
00:21:52.13 in the case where we're just expressing the viral nuclease
00:21:54.25 by depleting these cellular enzymes from cells
00:21:57.18 and asking what happens.
00:22:00.11 And what I'm gonna show you is that
00:22:03.02 in these experiments it appears
00:22:05.17 that there's evidence for cellular sensing of overall rates of decay.
00:22:09.18 We've done a number of experiments to test this,
00:22:12.10 and I'm just gonna show you one for simplicity.
00:22:15.01 In this experiment, what we're doing is we're asking,
00:22:18.23 what happens to RNA polymerase II occupancy in cells
00:22:21.26 when we express SOX?
00:22:24.03 And then, when we prevent those degrad...
00:22:26.23 those cleaved fragments from being degraded,
00:22:29.22 by getting rid of the cellular exonuclease XRN1,
00:22:34.00 using siRNAs?
00:22:36.01 And so, what you can see here is that
00:22:38.23 when we express SOX in cells that have XRN1
00:22:41.23 -- so, the RNA fragments are efficiently degraded --
00:22:44.13 there is a decrease in RNA pol II occupancy at promoters
00:22:48.12 , as I'd shown you previously
00:22:51.21 using RNA synthesis experiments.
00:22:53.27 And so, this is reflective of the fact
00:22:56.26 that RNA decay causes transcriptional repression.
00:22:59.00 We don't see that decrease if we use the catalytically inactive version,
00:23:02.04 which is shown in the white bar next to it.
00:23:04.03 However, in SOX-expressing cells,
00:23:07.04 if now we get rid of XRN1
00:23:10.07 -- so, the fragments are no longer degraded --
00:23:13.10 now RNA polymerase II occupancy
00:23:16.12 remains high at this representative promoter,
00:23:18.29 suggesting that it's relying on the degradation
00:23:22.03 of the already inactivated fragment
00:23:24.14 that's happening in cells
00:23:26.11 in order to send this signal from the cytoplasm to the nucleus.
00:23:31.08 So, how might this signal be sent,
00:23:35.02 or what might the signal be,
00:23:37.01 that allows cells to detect accelerated rates of RNA decay,
00:23:41.02 as can happen during infection?
00:23:45.14 Well, in thinking about this,
00:23:47.29 we reasoned that RNAs, of course,
00:23:49.21 are not naked in a cell --
00:23:51.25 they're coded by lots of RNA-binding proteins.
00:23:54.28 And if you envision what's happening with an RNA
00:23:58.11 as it's being degraded,
00:24:00.23 well, one of the things is that these RNA-binding proteins
00:24:03.15 are getting released from that RNA.
00:24:06.06 And we know that many of the RNA-binding proteins themselves
00:24:09.28 are shuttling proteins.
00:24:11.25 They can move in between the nucleus and the cytoplasm,
00:24:15.15 because of course RNA itself
00:24:17.22 starts out in one compartment and usually ends up in the other.
00:24:20.19 And so, you could envision that under a scenario
00:24:24.01 where you start to have lots of RNA decay,
00:24:27.02 the release of RNA-binding proteins
00:24:30.06 is such that you have many more unbound proteins
00:24:31.28 than bound proteins to RNA,
00:24:33.18 and that these proteins may start to differentially shuttle
00:24:37.10 between the two compartments,
00:24:38.27 sending a signal from the cytoplasm to the nucleus.
00:24:42.23 To test this hypothesis,
00:24:44.18 we set out to chart the scale of protein redistribution in cells
00:24:49.21 that occurs in response to RNA degradation.
00:24:53.24 And to do this, we partnered with Ileana Cristea's lab at Princeton University
00:24:58.17 because they're experts in measuring protein abundance by mass spectrometry,
00:25:02.24 also in the context of herpes viral infection.
00:25:06.02 The basic experimental setup here
00:25:09.01 was to take cells in which we were expressing empty vectors, there as a control,
00:25:13.26 our wild type SOX nuclease, shown in green here,
00:25:17.06 or a catalytically dead version of the nuclease,
00:25:20.14 shown in yellow.
00:25:21.29 Additionally, we engineered cells where we had knocked out,
00:25:24.27 using CRISPR technology,
00:25:28.24 XRN1,
00:25:30.16 because we knew that it was probably more than just that initial SOX cleavage event
00:25:33.11 that was important,
00:25:35.01 that it relied on the subsequent degradation of those RNAs.
00:25:37.22 So, we could look for things that were both SOX-dependent
00:25:41.05 and XRN1-dependent.
00:25:43.05 For each of these samples,
00:25:45.15 we separated the cells into the nuclear fraction
00:25:47.29 and the cytoplasmic fraction,
00:25:50.05 extracted proteins from both of those fractions,
00:25:53.29 and labeled them using isobaric tags called tandem mass tags.
00:25:57.28 These are tags that can be
00:26:01.03 identified during the process of mass spectrometry.
00:26:03.09 So, then we could measure protein abundance
00:26:05.19 in both of these compartments.
00:26:07.28 And what we were looking for here
00:26:10.08 were proteins whose abundance shifted
00:26:12.09 from one compartment to the other compartment
00:26:15.01 during RNA degradation,
00:26:17.01 for example going up in the nucleus
00:26:19.03 and going down in the cytoplasm.
00:26:21.21 And indeed, that's what we found.
00:26:23.18 So, if you look at this heatmap, here,
00:26:26.09 that's shown in blue,
00:26:28.29 the center bar shows what's happening during cells...
00:26:31.02 the context of cells expressing this viral nuclease.
00:26:34.07 Darker blue means proteins
00:26:36.28 that are becoming enriched in the nucleus
00:26:39.08 relative to the cytoplasm in these cells.
00:26:42.05 So, we see a whole collection of proteins
00:26:45.06 that look like they're shifting from one compartment to another.
00:26:48.11 They're not shifting in cells
00:26:51.00 where we're expressing the catalytically dead version of the nuclease,
00:26:54.01 shown in the first bar.
00:26:55.25 And gratifyingly, most of them, as it turns out,
00:26:59.09 about 60%,
00:27:01.06 don't shift in the cells where we have knocked out XRN1,
00:27:03.27 confirming that it's RNA degradation that's doing this.
00:27:08.04 Now, we looked at all proteins in a cell,
00:27:09.25 but when we did GO molecular function analysis,
00:27:12.11 we found that, as we might have predicted,
00:27:15.10 the signatures of the types of proteins
00:27:18.11 that are coming out here are things associated with poly(A) binding,
00:27:22.26 poly(U) binding,
00:27:24.16 messenger RNA 3' UTR binding...
00:27:26.18 just the sort of things you'd expect to get released from an RNA
00:27:28.20 as it's undergoing degradation.
00:27:31.05 We think probably a number of these impor...
00:27:34.23 proteins that are shifting are important for this phenotype.
00:27:37.21 I just want to draw your attention to two of them, though.
00:27:40.27 And these are the cytoplasmic poly(A)-binding proteins,
00:27:44.06 PABPC, or PABPC1,
00:27:47.22 and PABPC4.
00:27:49.09 These are proteins that consistently rise to the top,
00:27:53.20 or near the top, of our list.
00:27:55.11 And as you can see in these graphs below,
00:27:57.05 in both cases these are proteins
00:27:59.22 that in SOX-expressing cells,
00:28:01.17 the wild type nuclease-expressing cells,
00:28:04.00 increase in the nucleus
00:28:05.22 -- that's the purple dots here --
00:28:07.13 and decrease in the cytoplasm --
00:28:09.23 these are the teal dots.
00:28:11.12 You don't see changes in their abundance
00:28:13.24 in cells expressing the catalytically dead version.
00:28:16.16 Alright.
00:28:17.27 So, we would predict that if RNA-binding proteins
00:28:20.27 are things that are important
00:28:23.25 for conveying this information,
00:28:25.21 then if we depleted these RNA-binding proteins from cells,
00:28:27.29 we might now decrease that signal
00:28:31.09 that connects RNA decay in the cytoplasm
00:28:34.18 to transcription in the nucleus.
00:28:36.25 And this appears to be the case.
00:28:38.23 I'm showing you here depletion experiments
00:28:40.24 for these PABPC1 and PABPC4 proteins.
00:28:45.06 And when we deplete these from cells
00:28:47.10 and look at the relative occupancy of RNA polymerase II
00:28:51.21 at representative promoters
00:28:54.05 in the presence or absence of SOX,
00:28:56.05 you can see that, well,
00:28:58.11 when the proteins are around,
00:29:00.04 as we've shown before,
00:29:01.26 you get significantly less Pol II occupancy of promoters
00:29:05.23 because transcription is being repressed.
00:29:08.07 However, upon depletion of these factors,
00:29:10.08 now the relative difference
00:29:13.18 between cells lacking and containing the viral nuclease
00:29:17.01 is much different.
00:29:19.05 We don't think that these two proteins
00:29:21.25 are the only players in this pathway.
00:29:23.29 We think they're a component of perhaps several players
00:29:28.21 that are involved in this
00:29:30.10 that we're in the process of exploring further.
00:29:32.15 So, what I've shown you here, so far,
00:29:35.25 is that as we degrade RNAs in cells
00:29:40.12 initiated by a viral endonuclease like KSHV SOX,
00:29:44.11 these RNAs... once they're cleaved,
00:29:47.17 the fragments are degraded by cellular exonucleases like XRN1;
00:29:52.18 this leads to release of RNA-binding proteins
00:29:55.15 from these RNAs,
00:29:57.01 many of which we think can shuttle;
00:29:59.27 and once you reach a certain threshold of RNA decay,
00:30:02.09 these proteins, like the poly(A)-binding proteins,
00:30:05.22 start to accumulate in the nucleus.
00:30:09.05 And we suspect that this might be interpreted
00:30:11.22 as a stress signal by the cell.
00:30:13.27 And that leads to decreased transcription by RNA polymerase II,
00:30:18.01 of many host genes.
00:30:21.03 I just want to take the last minute to answer another question,
00:30:25.15 which is...
00:30:28.15 well, I've been telling you all about these host genes,
00:30:30.19 but as it turns out, of course,
00:30:32.15 herpesviruses need RNA polymerase II to transcribe their own genes.
00:30:35.20 So, why would they activate a pathway
00:30:39.21 that is going to decrease the activity
00:30:41.19 of a transcription complex
00:30:44.06 that they need for every single event of viral transcription?
00:30:48.22 They must be escaping this, right?
00:30:50.25 And indeed, they are.
00:30:52.24 So, if you look acro...
00:30:55.07 at RNA polymerase II occupancy across the viral genome
00:30:57.12 -- and here I'm just showing you a trace
00:30:59.17 from a particular region of the genome --
00:31:01.15 what I hope you can appreciate
00:31:04.00 is that there's not much of a difference
00:31:06.14 between the upper and the lower bar.
00:31:08.20 And this is cells infected with wild type virus
00:31:11.19 or a gammaherpesvirus
00:31:14.01 where the SOX gene is mutated.
00:31:16.07 And this is dramatically different
00:31:18.17 from what the host genome looks like,
00:31:20.11 where there's a clear difference in susceptibility.
00:31:21.25 That tells us that the virus is escaping
00:31:24.03 this transcriptional repression phenotype.
00:31:26.20 And the reason we think it's escaping this phenotype
00:31:29.11 is that viral replication
00:31:31.22 happens in what are called
00:31:34.23 replication compartments in the nucleus.
00:31:36.21 These are not membrane-bound,
00:31:38.14 but they are local concentrations of factors
00:31:41.06 that the virus needs to express its genes
00:31:43.14 and to replicate its genome.
00:31:46.08 And these replication factors,
00:31:49.13 we think, form a protective environment
00:31:53.12 for the virus to do things outside of regulation by the host.
00:31:56.01 And in fact, in these replication compartments,
00:31:59.20 if we put a promoter that is not from the virus
00:32:02.26 onto the viral genome,
00:32:04.24 it escapes transcriptional repression.
00:32:06.16 Conversely, if we take one of the viral promoters
00:32:09.15 outside of the virus,
00:32:11.13 and outside of the replication compartment,
00:32:13.22 and insert it into the host cell's genome,
00:32:16.23 it now becomes repressed just like the host genome is.
00:32:20.12 So, we think that these compartments
00:32:22.11 are the mechanism by which the virus
00:32:24.03 is able to activate this transcription repression pathway
00:32:27.12 that decreases host transcription,
00:32:30.13 but the virus is able to now use that polymerase
00:32:33.26 for its own purposes.
00:32:36.16 Okay.
00:32:38.02 So, in general,
00:32:39.25 what I've shown you during this lecture
00:32:41.21 is that viruses, like the gammaherpesviruses,
00:32:45.03 are able to dramatically reshape the gene expression environment
00:32:48.07 of a host cell during infection.
00:32:50.08 And they do this... one way that they do this, at least,
00:32:54.07 is through encoding a nuclease,
00:32:56.13 like the SOX endonuclease that I told you about today,
00:32:59.10 which is able to target a broad set of RNAs in a cell,
00:33:02.29 but with a high degree of specificity,
00:33:05.16 using a combination of sequence and structure
00:33:08.09 to make up its targeting element on these RNAs.
00:33:12.14 Once this enzyme cleaves these RNAs,
00:33:15.08 these inactivated fragments
00:33:17.14 are now fed into the cellular basal RNA decay pathway
00:33:21.12 and cleared by an enzyme called XRN1.
00:33:24.20 And during the process of XRN1
00:33:26.28 chewing up these fragments,
00:33:28.22 proteins are released from the RNA;
00:33:32.01 some of these proteins start to traffic into the nucleus;
00:33:35.09 and we think this is interpreted as some sort of a stress signal,
00:33:38.20 leading to broadly reduced transcription of host genes;
00:33:43.05 but the virus is able to escape this
00:33:46.13 through the formation of these replication compartments in the nucleus
00:33:49.12 that allow it to concentrate the polymerase in a cell
00:33:52.24 and transcribe viral genes.
00:33:56.05 Alright.
00:33:57.14 I'll end there by acknowledging
00:34:00.26 generous funding from a number of sources
00:34:03.04 that has contributed to these projects that are ongoing in my lab.
00:34:06.07 And I'd also like to thank all current and past members
00:34:09.20 of the Glaunsinger lab
00:34:11.21 that have contributed, broadly,
00:34:13.16 to the experiments that I showed you today,
00:34:15.12 and individual members, named here,
00:34:17.16 who have contributed to specific experiments
00:34:20.02 that I showed in this lecture.
00:34:22.17 Thank you very much for listening.

Videos in this Talk

Part 1: Viruses Reveal the Secrets of Biology
Audience:
- Student
- Researcher
- Educators
- Educators of H. School / Intro Undergrad
- Educators of Adv. Undergrad / Grad
Part 2: KSHV: Herpesviral Nucleases Impact Cellular RNA Destruction and Synthesis
Audience:
- Student
- Researcher
- Educators
- Educators of H. School / Intro Undergrad

Speaker: Britt Glaunsinger

Total Duration: 1:03:05

Recorded: October 2019

All Talks in Microbiology

Talk Overview

Dr. Britt Glaunsinger provides an overview of virology, and describes how the study of viruses has guided the understanding of many fundamental cellular processes, from gene expression to cancer. These insights arise from studying how viruses manipulate and hijack cellular machinery during infection and viral replication. Viruses can use the host machinery to their advantage, altering the gene expression landscape of the cell to create an environment favorable for the infection to progress. As Glaunsinger explains, one way viruses alter gene expression is by affecting the messenger RNA (mRNA) degradation process. By increasing decay of cellular messages, viruses can decrease competition for access to the translation machinery and dampen the expression of immune stimulatory elements that restrict viral replication.

In her second talk, Glaunsinger explores gene expression control by viruses like Kaposi’s sarcoma herpesvirus (KSHV), a frequent cause of cancer in AIDS patients. KSHV stimulates degradation of mRNA by encoding a nuclease, SOX, which is able to target a broad set of mRNAs for degradation yet cleaves them at specific sites recognized by a combination of RNA sequence and structure. Glaunsinger then describes how widespread mRNA degradation by viral nucleases such as SOX causes redistribution of RNA binding proteins in the cell and restricts mRNA transcription by RNA polymerase II. Thus, alterations to the rate of mRNA decay can have ripple effects in the cell that influence upstream events in gene expression.

Speaker Bio

Britt Glaunsinger

Dr. Britt Glaunsinger is a Professor at the departments of Plant and Microbial Biology and Molecular and Cellular Biology at the University of California, Berkeley, and a Howard Hughes Medical Institute Investigator. She earned her bachelor’s degree in Molecular and Cell Biology from the University of Arizona (1995), and completed her doctoral degree in Molecular… Continue Reading