Unfolding the UPR or the Unfolded Protein Response

Transcript of Part 1: Unfolding the UPR

00:00:29.15	Hello, I am Peter Walter.  I work at the University of California at San Francisco,
00:00:34.03	and we are basically cell biologists trying to figure out how cells work,
00:00:39.07	specializing on organelle biogenesis and how proteins
00:00:43.21	get to the right place and how they function.
00:00:45.21	And ultimately we would like to understand how proteins work as molecular machines
00:00:49.15	to make cells as wonderful and as complicated as they are.
00:00:53.20	So what I want to tell you today is a brief story of discovery
00:00:57.20	I think one of the most wonderful discoveries that was made in my lab in my career,
00:01:03.16	and it involves the reaction by which cells
00:01:08.03	decide how much of a particular organelle they have to have.
00:01:11.23	And the organelle that I'm going to tell you about is the endoplasmic reticulum,
00:01:16.15	which is the first way station which proteins enter
00:01:20.04	as they move through the secretory pathway to the surface of the cell,
00:01:26.05	and end up being secreted or inserted in the plasma membrane.
00:01:29.18	The endoplasmic reticulum then is a place where proteins have to...
00:01:35.21	They enter the ER; they have to fold;
00:01:38.18	they have to become assembled from multiple subunits; they become modified,
00:01:42.27	maybe a carbohydrate is added; disulfide bonds are formed.
00:01:47.02	And all of these reactions are important to produce properly functioning proteins.
00:01:51.13	And cells that are specialized, that make a lot of proteins and secrete them,
00:01:55.15	have to have a lot of endoplasmic reticulum
00:01:57.15	in order to carry out these processes with fidelity and with appropriately abundant machinery.
00:02:06.05	The proteins enter the endoplasmic reticulum
00:02:08.28	in an unfolded state as they come out of the ribosome
00:02:11.21	and then as they are in this organelle they basically have to mature.
00:02:16.00	And if they cannot fold properly,
00:02:17.25	then a cell basically puts proteins in its plasma membrane or secretes them
00:02:25.18	and because all the machinery that is required for a cell to know where it is in the body,
00:02:33.05	what it has to do, how it has to behave in the context of a multicellular organism,
00:02:37.25	all this information is being transmitted by proteins that are being secreted
00:02:42.27	or by the machineries that sit in the plasma membrane.
00:02:45.29	So it's very, very important that these machineries work properly, are properly assembled
00:02:51.01	because otherwise you may create a rogue cell that doesn't know when to divide,
00:02:55.26	when to differentiate, when to die, when to migrate to another place and so on and so forth.
00:03:02.02	So, for the organism, this is a very important process.
00:03:06.06	And this is sort of amplified in this slide
00:03:09.18	where you look at the differentiation of this cell, a precursor cell here
00:03:15.09	that turns into a professional secretory cell, in this case it's a plasma cell.
00:03:19.22	And the plasma cells are the cells that make antibodies that are being secreted in the bloodstream,
00:03:24.00	so this cell makes its own weight every day in antibody molecules.
00:03:27.21	And you see that this differentiation event goes hand in hand
00:03:30.27	with this vast proliferation of endoplasmic reticulum
00:03:34.01	that basically fills the cytosol wall to wall.
00:03:36.24	So, how does that cell know how much endoplasmic reticulum it needs?
00:03:41.09	So there must be signaling pathways in this cell that figure out how much of an organelle
00:03:46.21	we have and how much should be there according to the needs of the secretory load that the cell has.
00:03:55.11	And this signaling pathway that transmits that information,
00:03:59.28	which we've been studying for the last fifteen years or so,
00:04:03.29	is called the Unfolded Protein Response.
00:04:06.24	And it's called the Unfolded Protein Response because it initiates with an accumulation of
00:04:12.10	unfolded or misfolded proteins in the endoplasmic reticulum.
00:04:16.03	And these proteins, they are unfolded because there isn't enough capacity
00:04:20.26	to fold them properly from the machinery.
00:04:23.07	These proteins then create a signal that's being transmitted
00:04:25.29	across the membrane of the endoplasmic reticulum
00:04:28.14	that ends up eventually in the nuclear compartment
00:04:31.08	where it turns up a vast gene expression program
00:04:33.20	that makes, basically, more endoplasmic reticulum folding capacity, secretory capacity,
00:04:39.19	capacity to degrade proteins that cannot be folded properly and so on and so forth.
00:04:44.11	So it brings cells back into a homeostatic state that allows the load of secretory proteins
00:04:53.29	to be balanced with the machinery available to carry out their task.
00:04:58.25	To figure out how this pathway works two very adventurous graduate students in my lab,
00:05:06.10	Jeff Cox and Carolyn Shamu, started a project in which we tried to identify
00:05:13.29	these components genetically in the yeast system.
00:05:17.08	And basically, what they did is they built a reporter system based on an observation
00:05:21.23	made by Kazutoshi Mori in Mary-Jane Gething and Joe Sambrook's lab
00:05:25.25	that there is a small element in the promoters of the target genes of the response
00:05:30.26	that can be transplanted, can be put in front of a reporter gene,
00:05:35.05	and then we can isolate mutants in the cell where when we induce unfolded proteins in the ER
00:05:41.17	that no longer induce the Unfolded Protein Response by this induction of this reporter gene.
00:05:48.06	We can then take these mutants, we can clone the genes that have been mutated
00:05:52.18	by complementation and figure out what they do in the pathway.
00:05:57.12	And the nice thing is that the first gene, which we isolated this way, turns out
00:06:02.09	to be IRE1 and it encodes a transmembrane kinase.
00:06:06.02	So by virtue of it being a transmembrane kinase,
00:06:08.18	it already told us that this may be the signal transduction device
00:06:14.00	sitting in the ER membrane, figuring out in one end what's going on there,
00:06:18.05	and transmitting that information across the bilayer.
00:06:20.28	Very nicely, the second gene we isolated turns out to be HAC1,
00:06:25.12	and HAC1 encodes a transcription factor that binds to all these promoter elements.
00:06:30.01	So we then have the transmembrane kinase, we have a transcription factor,
00:06:34.12	and of course the way we are thinking about that
00:06:36.15	is very much in analogy to other transmembrane kinases,
00:06:39.22	like growth factor receptors in the plasma membrane of mammalian cells,
00:06:43.12	that this thing is activated and functions
00:06:46.04	by a process of oligomerization in the plane of the membrane
00:06:49.07	where as unfolded proteins accumulate, these kinase molecules come together,
00:06:55.02	they start bringing the kinases together on the other side of the membrane
00:07:00.12	where they are now juxtaposed so they can trans-autophosphorylate each other
00:07:04.20	and that somehow leads to phosphorylation cascade
00:07:07.07	downstream that activates the transcription factor.
00:07:09.29	But it turns out that nothing could be further from the truth.
00:07:14.21	This pathway is wired in a completely different way and a completely unexpected way.
00:07:20.09	And that was discovered, pretty much, by a series of control experiments that Jeff carried out.
00:07:25.20	And he was the first one, very simply, very naively,
00:07:30.17	just started looking for the transcription factor in cells
00:07:33.22	that are either induced for the response or that are not.
00:07:37.21	And as you can see here, the transcription factor is only present
00:07:41.10	in cells when the response is induced.
00:07:45.26	So that tells us two things right then and there, right?
00:07:49.05	It's either degraded when its not needed
00:07:51.05	or it is only synthesized when the response is induced.
00:07:56.01	And to distinguish between these two possibilities what Jeff did is
00:08:00.15	he just did a simple Northern blot analysis by which he asked,
00:08:04.12	"Does the messenger RNA encoding this transcription factor
00:08:07.04	change in its abundance when we induce unfolded proteins?"
00:08:11.09	As you see here, the messenger RNA doesn't really change much in abundance.
00:08:16.28	But what you see, what Jeff discovered here, is that we now have a band of a different size
00:08:23.21	that was completely unexpected.  So, the simple Northern blot then led to the discovery
00:08:29.19	that there is something happening to the messenger RNA and to make a long story short
00:08:33.24	it turns out that this messenger RNA becomes spliced.
00:08:37.02	An intron is being removed when the Unfolded Protein Response is induced.
00:08:44.25	So the idea then is, that the messenger RNA encoding
00:08:47.25	the transcription factor is initially encoded with this intron
00:08:52.07	and as the unfolded proteins accumulate, this intron is being removed,
00:08:57.22	producing the spliced messenger RNA which is then being translated
00:09:02.03	to produce the transcription factor that turns up the response.
00:09:06.28	Now this is highly unusual.  Normally, when cells decide to change their transcription
00:09:12.16	to their splicing program through some alternative splicing,
00:09:15.29	they make reasonably irreversible developmental decisions
00:09:20.24	so that they last a long time. They enter different cell fate with these decisions.
00:09:26.16	But this is bona fide signaling.
00:09:27.22	What's happening here is...strictly dependent on
00:09:31.29	the conditions inside the endoplasmic reticulum.
00:09:34.11	The signaling is on and off depending on whether you have unfolded proteins or not.
00:09:39.04	What's even more surprising is that this splicing reaction
00:09:42.16	followed none of the rules of normal messenger RNA splicing.
00:09:45.21	It's completely independent of the spliceosome. It's happening in the cytosol and
00:09:52.06	it's carried out by two enzymes and two enzymes only.
00:09:56.07	The first one turns out to be Ire1.
00:09:58.28	Our transmembrane kinase when activated, becomes a bifunctional protein that is not only a kinase,
00:10:05.28	but also a site specific endoribonuclease that cleaves the messenger RNA
00:10:11.02	encoding the transcription factor precisely at both splice junctions
00:10:15.23	and to the best of our knowledge, this is the only messenger RNA in the cell that it touches.
00:10:20.12	And then, along comes an enzyme called tRNA ligase
00:10:23.16	which was previously only known for its role in tRNA splicing
00:10:27.11	and we had isolated a mutation in that, another graduate student, Carmela Sidrauski,
00:10:32.04	and we couldn't make any sense out of that until Jeff had discovered
00:10:35.26	that there's this RNA processing step in the signaling pathway.
00:10:40.10	So the Unfolded Protein Response then transmits the signal by this unconventional,
00:10:46.24	completely unprecedented splicing pathway
00:10:51.03	and there's no phosphorylation cascade anywhere in sight here.
00:10:56.12	Let me just show you this, too.  We can make this messenger RNA in vitro,
00:11:01.14	we can incubate it with recombinantly produced Ire1
00:11:04.22	and you see that the messenger RNA gets cleaved,
00:11:07.13	producing the intron, 5' exon, 3' exon and then as we add also recombinantly produced,
00:11:13.19	purified tRNA ligase to this reaction the exons go away,
00:11:18.04	get ligated to form the product and the intron stays put.
00:11:22.26	So we can reconstitute this whole pathway
00:11:26.02	from two purified components with quite nice efficiency.
00:11:32.11	To make a long story short, this discovery then led to many other labs and it turned out that
00:11:39.11	pretty much everything...the cellular features that we've learned from
00:11:43.14	the simple yeast system hold true for mammalian and metazoan cells.
00:11:51.24	So this IRE1 exists in metazoan cells, it is involved in the splicing of a messenger RNA
00:11:58.11	encoding a transcription factor, XBP1, XPB1 here.
00:12:02.19	Things are more complicated, higher evolved eukaryotic cells have
00:12:08.01	added more bells and whistles to the pathway.
00:12:10.26	So we have three parallel pathways here that transmits information from the ER lumen
00:12:16.02	to the cytosol each leading to the activation of a transcription factor.
00:12:19.28	We have Ire1 working by this non-conventional mRNA splicing.
00:12:24.07	We have another transmembrane kinase here, PERK
00:12:27.13	that makes another transcription factor by a mechanism of translational control.
00:12:32.07	And finally, we have this exciting protein here
00:12:34.21	that sits in the ER membrane, ATF6, and then gets released,
00:12:39.10	a fragment of it gets released to become an active transcription factor
00:12:43.06	that moves into the nucleus and turns up the target genes only when
00:12:47.15	unfolded proteins are accumulating in the endoplasmic reticulum.
00:12:51.24	And the idea of the whole thing is the same as I told you before.
00:12:54.25	It is to establish, re-establish homeostasis so protein folding
00:12:59.23	in the endoplasmic reticulum can occur with fidelity.
00:13:03.04	So we have these three pathways here and they establish homeostasis.
00:13:07.10	But I also told you that there is a danger that if cells cannot achieve homeostasis
00:13:15.14	that they may make mistakes in protein folding and therefore turn into rogue cells
00:13:20.19	that endanger an organism. So there's a safety valve built in
00:13:24.11	that if this balance cannot be achieved again that cells, rather than
00:13:29.08	putting these misfolded proteins, and misfolded signaling machines on the cell surface
00:13:36.03	go down a pathway of apoptosis.
00:13:39.11	So rather than becoming rogue cells and endangering the organism,
00:13:43.11	they remove themselves by committing suicide.
00:13:45.18	And it's this point that the Unfolded Protein Response makes
00:13:50.16	life death decisions for the cell that puts this pathway in the midst of many different human diseases.
00:13:56.20	Some cancer cells are kept alive because the response gives them a growth advantage.
00:14:06.24	In diabetes, the beta cells in the pancreas may die through this apoptotic route here
00:14:13.27	by being over-committed to produce ever increasing amounts of insulin.
00:14:19.22	And there are neurodegenerative diseases where protein misfolding causes apoptosis
00:14:24.13	by turning up this response.
00:14:26.26	So, basically then, our very pioneering work of these very adventurous graduate students
00:14:35.18	has led us to understand a mechanism that we now, we and many other labs
00:14:41.27	have tried to utilize in therapeutic approaches to see if we can have
00:14:50.18	some ability of bettering mankind via interfering and modulating these pathways
00:14:58.10	and therefore affecting the outcome of human disease.
00:15:00.23	And I should close by giving credit to these really wonderful, adventurous pioneers:
00:15:09.06	Jeff Cox, who discovered, together with Carolyn Shamu, Ire1 and the Hac1 transcription factor
00:15:16.12	and Carmela Sidrauski, who discovered that Ire1 is a nuclease in the splicing reaction and
00:15:24.07	put the tRNA ligase on the map.
00:15:26.19