Randy Schekman (2014) - Genes and Proteins that Control Secretion and Autophagy

Good morning. I am delighted to be here, to see all these bright young faces this morning and to have my first opportunity to speak at a Lindau conference. I hope the first of many as years go on. I’m going to talk this morning about a new research project in my laboratory having to do with the mechanism of protein degradation, Protein and organelle degradation in mammalian and in yeast cells. It begins with the question of how a particular membrane in a cell called the phagophore membrane gives rise to a structure called the autophagosome which is responsible for the turnover of macromolecules in cells that are subject to stressful conditions. So let’s have the first slide. This process was first revealed in mammalian tissues many years ago with the discovery that stress induces wholesale protein turnover in the cell. But the real breakthroughs came I would say with the development of a genetic approach using baker’s yeast to uncover the genes called ATG genes that are required for degradation of organelles and cytoplasmic components when yeast cells are starved for an essential nitrogen source. The pioneering work in this field was achieved by Osumi in Japan and Klionsky now in Michigan. Let me review for you what we’ve learned about this process in yeast and in mammalian cells. We know that the process begins with the creation of a membrane called the phagophore membrane that envelops organelles and cytoplasmic components that will be captured for delivery to the lysosome where they’re turned over by proteases and lipases. This structure, the phagophore membrane, grows eventually to enclose on itself creating a membrane with content and an outer membrane that serves to target the organelle by membrane fusion to the lysosome. The inner membrane and its content are then degraded and the amino acids and sugars are restored to the cytoplasm for renewed use in a new programme that allows cells to survive under conditions of stress. Now in the field of autophagy there’s been a vigorous discussion on the nature of this membrane and its origin in the cell. It’s a substantial membrane. And yet where it comes from in the cell has been a bit of a mystery. So much so that the literature has focused on a number of possible origins. Many investigators have argued that the membrane originates from the endoplasmic reticulum. But others have argued that it may arise from the mitochondrion or possibly a region of adhesion between the mitochondrion and the ER. And others still have suggested that its growth may be sustained by recruiting membrane from endosomes, from the Golgi apparatus and even from the plasma membrane. Now, most of the work in my laboratory has focused on the mechanism of protein traffic in the secretory pathway but about 3 years ago I had an inquiry from a wonderful postdoctoral fellow from Shanghai by the name of Liang Ge. And in discussing potential research projects with Liang we focused on this question of where the autophagosome membrane originates. And Liang, when he joined my laboratory, conceived of a really interesting strategy to using biochemistry to identify the origin of this organelle. Let me tell you the key molecular covalent reaction that he focused on for his investigation. That is the attachment of a lipid called phosphatidylethanolamine to a covalent linkage to glycine at the C terminus of a purely cytosolic protein in mammalian cells called LC31. Now, we know about the biochemistry of this conjugation to phosphatidylethanolamine through the work of Osumi in Japan who characterised a set of ATG genes that act similarly to the E1, E2, E3 cascade of events that allows ubiquitin to be transferred to proteins that are turned over by degradation in the cytoplasm. The same kind of scheme through the final action of this complex results in LC3 being converted to a lipidated species and then anchored to the phagophore membrane, one of the only, very few unique components of the phagophore that can be used in its diagnoses. Now we can follow this event in cultured mammalian cells because lipidation of LC3 results in a protein that migrates more rapidly on an SDS gel. So just at the onset of starvation most of the LC3 migrate slowly on this gel. Within 30 minutes much of it is now at this more rapidly migrating species. And the protein as it becomes lodged in the membrane collects on the concave face of the membrane such that when the phagophore closes on itself, much of the LC3 that’s lipidated is eventually delivered to the lysosome. And as you can see here it’s degraded over the next hours of this regimen of starvation. Liang decided to try to reproduce this event in vitro and specifically to recreate the conditions that would allow LC3 to be converted to its lipidated species in a phagophore that was maturing in an extract consisting of soluble proteins and membranes. And in very short order after he joined my lab he succeeded in the following experiment. He took membranes from a mouse embryonic fibroblast cell line taken from a mouse that had sustained a knockout of 1 of the essential lipidation genes, ATG5. This is an early embryonic lethal event in the mouse but embryos can be collected and fibroblast grown from these. These embryonic fibroblasts grow perfectly well in the laboratory but they are completely deficient in lipidation of LC3. And they are defective at a very early stage in the autophagic process. He harvested membranes from a gentle lysate of ATG5 cells and mixed the membranes with cytosol taken from normal cells that had been starved under conditions that induce autophagy. The incubation of this mixture with ATP resulted in the production of a species that by migration and other criteria is authentic lipidated LC3. This happened whether the cytosol was taken from mouse embryonic fibroblast or from a human embryonic kidney cell line. And importantly if the C terminal glycine to which the phosphatidylethanolamine is attached is mutated to alanine, this reaction is completely arrested. Now, this reaction though it seems to reconstitute the authentic lipidation event, could simply be lipidation without the overarching control elements that are imposed when yeast cells are starved for nitrogen or when mammalian cells are starved for sterol. So it was important to establish using other criteria that the reaction that we had reconstituted measured some significant element in the origin of the phagophore membrane. And Liang used 2 conditions that have now satisfied us that we’ve captured an essential element in this process. The first experiment was to harvest cytosol from cells that are starved under conditions that induce autophagy or unstarved cells we treated with 1 or another drug that initiates the autophagic process. Let’s look at starvation. If you compare the efficiency of lipidation with cytosol taken from unstarved cells with cytosol taken from starved cells, you see a 3-fold increase in that reaction dependent upon starvation. Likewise if cytosol is harvested from cells treated with a drug that induces autophagy, again a 3-fold stimulation in this event. Another essential physiologic condition to initiate autophagy and indeed to initiate lipidation is the activity of an enzyme, phosphatidylinositol 3-kinase that produces PI3P which is essential for the lipidation and autophagic pathway to proceed. Fortunately there are good inhibitors of that enzyme and so Liang used 2 drugs that block the formation of this lipid in vivo. And found, as you’ll see in the next slide, that lipidation is also prevented in vitro. So here is such a reaction. The 2 drugs are 3-methyladenine which acts in cells in the millimolar range and wortmannin which acts in the nanomolar range. At these concentrations the formation of PI3P is blocked in vivo and lipidation is blocked in vivo. And likewise as you see here there is a progressive inhibition in lipidation by this inhibitor. Even though the inhibitor doesn’t directly block the action of the lipidation enzymes, it blocks the necessary prelude to the activation of the lipidation enzymes. We can also ask if the product, PI3P, must be accessible to the proteins in the cytosol to initiate this pathway. And here is an experiment that tests that. It takes advantage of a protein that has a domain that recognises the head group on PI3P. It’s a domain called the 5 domain embedded within a recombinant protein that we could purify. This recombinant protein on a lipid strip shows great specificity for PI3P. And correspondingly it is a potent inhibitor of the lipidation event in vitro whereas the mutation that inactivates this protein for binding to the head group is without effect in vitro. So we felt comfortable with these results in suggesting that we had reconstituted a significant early event in the formation of the phagophore. And we then focused on the question that initiated this whole project and that is: What is the membrane template in a cell that promotes the lipidation event? Is it a membrane, any membrane in the cell that houses phosphatidylethanolamine? Or is it a special membrane that is uniquely equipped to initiate this reaction? We again chose a biochemical approach. And Liang devised a scheme for fractionating all of the membranes in a lysate of ATG5 mutant cells in an effort to identify a unique membrane that may serve as a template. Now, let me show you the scheme that he developed that was quite successful. He found that a crude homogenate could be subjected to rounds of differential centrifugation under conditions where at this speed of centrifugal force, 25,000 times G for 10 minutes, a membrane that is enriched in lipidation activity is produced. And that activity is not found in lower or in higher speed pellet fractions. This pellet was re-suspended and sedimented to equilibrium on a sucrose step gradient to achieve a fraction further enriched in this activity. And throughout the fractionation scheme in addition to measuring the activity of a membrane to promote lipidation Liang looked at the distribution of marker proteins characteristic of the various membranes in a cell. And even at this stage, after 2 steps, almost all of the endoplasmic reticulum membrane was gone. It was removed. It fractionated elsewhere. All of the mitochondrial membrane that we could detect in the lysate was gone at this point. He then took this fraction and achieved a further substantial purification by sedimenting it to equilibrium on a linear gradient of a dense material called opti-prep. And I’ll show you the data for that experiment that produces a quite nice resolution. So here again is the fractionation scheme. This gradient represents the last fractionation on an opti-prep gradient. The top of the tube is the low buoyant density fraction, the bottom of the tube higher buoyant density. Lipidation activity is more or less restricted to fractions 2, 3 and 4 in this gradient. And this activity coincides quite nicely with proteins that are known to mark an organelle called the ER Golgi intermediate compartment, a way station between the ER that shuttles vesicle material back and forth between these 2 organelles. In contrast other organelles that remain in this last step of the fractionation, the endosome, a membrane that may continue to maturation of the phagophore, the peroxisome, the lysosome and even the cis-Golgi membrane, are nicely resolved. Now a trivial explanation may be that this membrane, the ER Golgi intermediate compartment, may somehow be enriched in most of the phosphatidylethanolamine in the cell. Well that simply isn’t true. PE is distributed elsewhere and was fractionated away. But we could ask very specifically on this last gradient where the preponderance of PE localised. That data is shown here. We’ve quantified the lipidation activity and quantified the recovery of various marker proteins by immunoblot. And then we chemically or enzymatically measured the fractionation of the remaining phosphatidylethanolamine in this gradient. And the highest concentration of PE is in the higher buoyant density fractions clearly resolved from the lipidation activity. Well, this was a very surprising result because the ER Golgi intermediate compartment is a relatively minor membrane in the cell. So we sought an independent orthogonal procedure to enrich this membrane to see if it really did carry along with it the template activity for lipidation. And what Liang devised was an immunol isolation scheme using epitopes on proteins that are exposed on a cytoplasmic face of the ERGIC membrane allowing him to immobilise this membrane and to centrifuge it on beads. Here is 2 versions of that experiment. One of the antibodies he used was against the ERGIC membrane protein Sec22b. This antibody very effectively precipitates a membrane that promotes lipidation and correspondingly precipitates 2 proteins of the ERGIC membrane while not precipitating a marker of the endosome. If the immunoprecipitation is conducted in the presence of competing levels of the peptide against which the antibody is raised, the membrane, lipidation membrane is not precipitated nor is the ERGIC membrane. Equivalently and using another protein that marks this membrane a receptor that cycles between the ER and Golgi with a flag epitope, the membrane is very efficiently precipitated. The lipidation activity recovered nearly quantitatively. But it is not precipitated if the peptide, flag peptide is included as a competitor. Well, in the final experiments then we addressed in cells whether the ERGIC membrane really could be detected as a station in the development of the phagophore. And one strategy that Liang devised was to use one of 2 drugs that block the formation of the ERGIC by inhibiting the formation of transport vesicles, COPII vesicles, that bud from the ER. Here is one such experiment. The 2 drugs are H89 and Clofibrate which when exposed to normal cultured fibroblasts result in the dispersal of a marker that is diagnostic of this compartment. Clofibrate more dramatically so. A complete dispersal in just 20 minutes of drug treatment. If cells are incubated with the drug for just 20 minutes and lysate is prepared and evaluated by the lipidation assay, the 2 drugs dramatically remove the template activity that promotes lipidation. Whereas 2 inhibitors that block downstream in the secretory pathway have much lower effect. Now, Liang also observed the reversal of the drug effect when the drug is withdrawn from cells. They very quickly restore the ERGIC membrane and as you’ll see in the next slide. The ability of the template to be detected in a lysate of cells from which the drug has been withdrawn is restored. Very rapidly indeed. So after 0 time or 20 minutes of drug treatment the membrane template activity is gone. But within just 2 minutes and fully after 5 minutes the ability of a membrane detected in vitro is restored to these cells. The kinetics of reappearance of the activity in vitro parallels substantially the reappearance of the ERGIC membrane and precedes the reformation of the Golgi membrane as detected by quantitative immunofluorescence. Now in the last experiment, a simple one, we ask: Is there something special about this membrane that allows it to serve as a template? Is there some protein that fractionates along with the membrane that remains to be discovered, that can somehow serve as an anchor to initiate this pathway? A very simple experiment that Liang conducted, shown here, is to expose either crude membranes or purified ERGIC membranes to trypsin on ice. Under conditions where proteins exposed on the outside surface of the membrane are clipped or shaved. You see that with increasing concentrations of trypsin the ability of this membrane to promote lipidation is destroyed even though all of the core components for lipidation are in the cytosol and are provided by the addition of crude cytosol to the isolated membrane. The membrane itself is not disrupted by the trypsin treatment because this protein which is largely housed inside the organelle is resistant to trypsin whereas another protein that is exposed on the outside surface is very sensitive to proteolysis. Well, let me summarise the conclusions in 2 forms. First in the form of a diagram which shows the position of this organelle mediating flow between the ER and Golgi. A kind of a bus station that collects vesicles from the ER and delivers them to the Golgi. Nonetheless a station that exists in the steady state at least in mammalian cells and must house 1 or more proteins that serve to mark this organelle as a unique compartment. We’d love to know what that protein is. We’re actively pursuing using this biochemical approach. Well I’ll summarise the results before I move on to another topic of interest. And that is by telling you that we have achieved this cell-free reaction. We believe that it represents a very early event in the generation of this interesting membrane. The reaction is stimulated by conditions known to induce autophagy in yeast and in mammalian cells including the action of regulatory proteins, Protein kinase, a lipid kinase, that were established by genetic and physiologic studies. We’ve used this reaction to identify a novel compartment that mediates a very early stage in the creation of the phagophore. And we’ve shown in vivo that the formation of this organelle depends upon active vesicular flow from the ER. Now, I’m going to turn my attention in the remaining minutes to another question that you’re all interested in whether you’re interested in autophagy or not. And that is the important issue of where we seek to publish our most important work. I can tell by the response that you may anticipate some remarks on my part. I believe that our system, that you, the young scholars in this audience, are now subjected to is broken. There is a distortion of influences that forces your hand in where you choose to publish your work. You are increasingly driven to publish in a very few venues that have captured your imagination because they present to you a sense of exclusivity and the promise of great promotion of your work and even treasure. Let me show you an example of one of the distortions that I learned of last year in the form of a document published by the Chinese Academy of Sciences shown in my next slide. Now I’m afraid you can’t read this. But I have a number of Chinese scholars in my office, in my laboratory. And they’ve translated this for me. It is an authentic document that basically is a reward notice. Basically a bribe that offers scholars in China a cash reward. The equivalent of 33,000 US dollars, private spending money, not for your laboratory necessarily if you win the lottery and publish your work in the 3 most exclusive venues in the life sciences, Cell, Nature and Science. This I believe is a terrible distortion of the values that we hold as scholars in the life sciences. It reduces our achievements to a monetary prize. Now this is bad enough. But I was particularly offended by the next line. Because as the former editor of the proceedings of the National Academy of Sciences I was disturbed to learn that the bribe was considerably less than that offered for publication in these venues. Only 8,000 dollars. And finally if you lose the lottery and publish in some local journal you only get 100 bucks. This represents, I think, a misapplication of the values that we hold dear in scholarly achievement. And it creates an influence that forces some unfortunate members of our scholarly community to misrepresent their work. And I’ll give you one important example. Early this year 2 seemingly revolutionary papers were published in Nature that claim that adult cells could be induced to form embryonic stem cells by exposure to a simple medium of low pH. This made the evening news in the US, the front page of the New York Times, the Wall Street Journal. It was sensational. But we learned in the weeks following that laboratories skilled in this approach failed to reproduce these results. We subsequently learn that images in the paper were inversions of each other. The images were used over and over in different orientations to give the impression that it was a different sample of cells. Indeed language in the paper was borrowed from previous publications of the first author. Now every journal and every editor can have such troubling experience with papers in their journal. But I would submit to you that the journals that are most exclusive and that rely on for example the impact factor in deciding whether a paper is going to be important for publication in that venue, such journals and the editors of those journals are responsible for creating pressures on us that do not... are not subject to the normal scholarly recognition and achievement that we seek. Indeed many others have suggested that the system of publication at the high end and the review process is broken. Hidde Ploegh publishing ironically in Nature. Peer review of scientific papers and top journals get bogged down by unnecessary demands for extra lab work. Or Marty Raff, Sandy Johnson, Peter Walter. In science the stress associated with publishing experimental results can drain much of the joy of practicing science. Why is this? Journals increasingly rely on what I consider to be a flawed metric called the impact factor. Indeed probably the postdocs in the audience can recite the impact factors of the journals that they wish to publish into 3 significant figures. And they will use minor differences in these numbers to determine which journal they will seek for their publications. And this creates on the other side increasing pressure to be highly selective, to look for those articles that seem to make claims that are extraordinary. Very often unfortunately these claims turn out to be wrong. Now many of us in the community have tried to do something to fight the influence of this phoney number impact factor. Several years ago Bruce Albert, then the editor of science magazine, and a group of journal editors gathered together in San Francisco to promote a document called the DORA document, Declaration of Research Assessment. If you haven’t seen it, you could Google it under DORA San Francisco. It recommends that scholars, publishers, review committees, university administrators, government officials move away from the use of impact factor and develop other means of evaluating scholarship. Maybe even reading the papers. Many of us have signed this document. I would urge you to do so. If you haven’t there are now over 10,000 people who have signed on. Now let me leave you finally with what I’m doing about it. Of course it’s one thing to complain about it. It’s another thing to do something. And I was asked several years ago by the Howard Hughes Medical Institute, by the Max Planck Society and by the Welcome Trust to create a new open access journal that’s led and run by scientists, by active scientists where all the decisions at the editorial level are made by active scholars. Truth in advertising. As the editor in chief in spend half of my time on the journal. So in fact I am compensated, part of my salary comes from this journal. It’s a journal called eLife. It is completely funded by these organisations. It has open access. It is free for a submission. We have no page charges. At least for the time being. And all of the decisions are made by active scientists in a highly consultative manner. I’d be happy to talk to you about this later on. And I would encourage you when you have time to look us up and see what we’ve been publishing in the year and a half that we’ve been available. I thank you for your attention. Applause.

Randy Schekman (2014)

Genes and Proteins that Control Secretion and Autophagy

Randy Schekman (2014)

Genes and Proteins that Control Secretion and Autophagy

Abstract

The broad outlines of the secretory pathway were established by pioneering EM and cell fractionation experiments conducted by George Palade in the 1960s. Beginning in the mid 1970s and early 80s, my laboratory isolated a series of conditionally lethal, temperature-sensitive mutations that block secretion at one of several sequential stages along the pathway established by Palade. Concurrently, James Rothman’s laboratory established a cell-free reaction that reproduced vesicular traffic within the Golgi apparatus, and several of the proteins he isolated with this functional assay matched the Sec proteins we identified. Using a cell-free vesicle budding reaction, my laboratory isolated a complex of Sec proteins that comprise a coat, COPII, responsible for cargo vesicle traffic from the endoplasmic reticulum.

Autophagosomes mature by the addition of membrane material from various intracellular sources and the attachment of peripheral proteins that remain bound through a covalent lipidation reaction. However, the origin and the mechanism of generation of the pre-autophagic membrane are poorly understood. We addressed this question with the development and analysis of a cell-free reaction that reproduces the lipidation of a major peripheral autophagosomal protein, LC3. A crude membrane fraction isolated from cells deficient in lipidation was mixed with cytosol harvested from normal cells that were untreated or subjected to a stress regimen known to induce autophagy. On addition of ATP, incubation of the mixture resulted in the formation of lipidated LC3. The reaction requires both membranes and cytosol, and was found to be stimulated 2- to 5-fold when the cytosol was taken from stress-induced cells. Autophagosome maturation requires a class III PI-3 kinase (VPS34 homolog); LC3 lipidation in our cell-free reaction is inhibited by inhibitors of this kinase, and by the addition of a peptide containing a PI3P-bnding sequence, the FYVE domain. Using cell fractionation techniques we have identified the ER-Golgi intermediate (ERGIC) compartment as the major site for lipidation of LC-3. This cell-free reaction may now be used to understand the molecular mechanism of autophagosome maturation.

Schekman, R. (2002) SEC mutants and the secretory apparatus. Nature Medicine 8, 1055-1058.

Ge, L, Melville, D., Zhang, M. and Schekman, R. (2013) The ER–Golgi intermediate compartment is a key membrane source for the LC3 lipidation step of autophagosome biogenesis eLife DOI: http://dx.doi.org/10.7554/eLife.00947

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT


Related Content