Articles, Blog

Susan Lindquist (Whitehead, MIT / HHMI) 1b: Protein Folding in Neurodegenerative Disease

August 18, 2019

Hi. I’m Susan Lindquist. I’m a member of the Howard Hughes Medical Institute and I work at the Whitehead Institute at MIT. I work on a variety of different protein folding problems and in my last lecture I gave you a broad introduction to the problem, told you how it manifested, at least a little bit, how it manifested in infectious diseases, and more broadly how it is used by cancers to drive the malignant state. In this lecture, I’d like to tell you about a different aspect of protein pathology, another equally devastating aspect of protein folding pathology — the neurodegenerative diseases — because all of these diseases are diseases of protein misfolding. This is an extremely vivid demonstration of the difference between the brain of a normal adult upon autopsy versus an adult who died of Alzheimer’s disease. It’s obviously a devastating disease and this is why the people who these diseases lose their memory, they lose control of functions. And these diseases are really terrible. Now, this is a graph of what’s happened to human longevity over the last couple of hundred years, and this is really, I think — the red and the black are just two different calculations… it’s not so easy going back to the older days to calculate exactly how old people lived on average, but these two different ways of doing it came out with the same answer — and you can see that there’s been this steady march of progress and it’s just been amazing. This has been, I think, one of the glories of mankind, to be able to do this and alter their own average lifespan, and it’s been due to many different factors: due to changes in public health and cleaner drinking water; due to refrigeration and preservation of food, and cooking; it’s due to the development of antibiotics, the development of vaccines, the development of anesthesia so you could do surgery on people and correct illnesses that way. So anyway, this wonderful steady, steady progress of mankind is unfortunately in some ways of thinking about it a Road to Ruin, because as we are curing these other diseases, as we are living longer and longer lives, we are finding that the incidences of neurodegenerative diseases are going up. These diseases used to be practically unheard of 100 years ago; now, there’s a very large fraction of people around the world that are suffering from these diseases, and as we extend lifespan it’s getting worse and worse. There are five million Americans suffering from Alzheimer’s disease alone, and this same increase in the disease is occurring for all of the neurodegenerative diseases across our globe. So, unfortunately, with respect to neurodegeneration and it being a Road to Ruin… this is why I say a Road to Ruin… we’re headed for neurodegeneration and right now there’s no exit. We do not have a single therapy that really fixes these problems. So, these are some of the common and uncommon neurodegenerative diseases you might have heard about: Alzheimer’s disease and Parkinson’s disease, frontotemporal dementia, Huntington’s, ALS, and Creutzfeldt-Jakob disease. And you can see these brown blobs inside of these cells, and those brown blobs are aggregated proteins like those aggregates of fried egg I showed you earlier. And, as I said, all of these neurodegenerative diseases are protein folding diseases and there’s not a single therapeutic strategy that cures the underlying protein pathology. We have some things that address some symptoms in some of these diseases, but for the most part we’re pretty helpless against them. So, I’ve been working on protein folding for a long time and I’ve worked on a lot of different organisms, and the one thing that my studies over the years has taught me is that this problem, as I mentioned earlier, is common to all organisms on Earth, and so we got the kind of crazy idea that, considering the Eukaryotic tree of life, you see plants, animals, and fungi actually split from each other not that long ago, in terms of evolution. So, we thought we might be able to take advantage of this similarity to study some of these really difficult, really complicated diseases. Yes, we will not be able to study many different aspects of protein folding in neurodegenerative disease in a simpler organism, but if we could study some aspects of the precipitating, initiating protein pathology, the cellular pathology, not the complexity of the disease as a whole, but just this initiating, precipitating pathology from those proteins in a simple organism, we might be able to move much more quickly than we would if we were confined solely to working on these more complex organisms. So, as I mentioned, one of the things that we have in common with yeast is a wide variety of systems for controlling the protein folding problem. So, we have chaperone proteins, which interact with highly reactive proteins that are not quite finished folding, and prevent, just like human chaperones prevent their charges from interacting inappropriately with other partners until they’re reading and mature, protein chaperones do the same thing. But we also have protein remodeling factors, things that can wrest those protein aggregates, when they start to appear, apart, we have osmolytes, we have things called the proteasome, which degrades proteins that are not properly folded, ubiquitin, ubiquitin ligases, and that entire system is just completely conserved from yeast to human cells. But it’s not just that. Lipid biology is actually quite highly conserved. There certainly are differences in the lipid biology of yeast and human cells, but, for example, cholesterol — yeast use a very closely related lipid called ergosterol for exactly the same reason that we use cholesterol: to control the fluidity of membranes and to control the movement and density of proteins within those membranes. And they move packages of membrane-bounded proteins around the cell in very highly orchestrated ways, really the same way that a nerve cell will move dopamine around, the yeast cells will move things like mating factors around. And lysosomes and peroxisomes, these are very complex organelles that are involved in very complicated functions, some of them are involved in degrading proteins, some of them are involved in a wide variety of metabolic actions that have to be segregated from the normal cytoplasm — yeast cells have both of those. They have autophagy. This is a process by which the cell actually directs its degradation and eating machinery actually to eat up protein aggregates and to get rid of them. Apoptosis, a programmed form of cell death. The cell cycle, a very complexly regulated cell cycle, regulated very, very differently in bacteria, but in yeast and humans, regulated in very much the same way. And in fact studies of that cell cycle were extremely important for our understanding of cancer and why cancer cells start to replicate uncontrollably. Studying them in yeast provides key insights. We have mitochondria, the energy factory of the cells, and mitochondria do amazing things in yeast and human cells, but they also are a place where reactive oxygen species are generated and can do a great deal of damage. And then there’s a whole variety of signal transduction pathways, again, these key pathways that control growth and development in us, but control responses to the environment and responses to other cells, and responses to internal and external stresses, those same signaling pathways have been controlled… have been preserved, rather, in yeast and higher eukaryotes. So, calcineurin is an example, MAP kinase is, G-protein coupled receptors… all of these were first developed long ago in eukaryotic life, and greatly, greatly elaborated in us, we have many, many more G-protein coupled receptors than a yeast cell has, for example, but the basic machinery and the basic concepts and the basic ways in which those signaling pathways drive processes inside the cell are similar. So, we got the idea that maybe we could use those yeast cells as our living test tube, and the reason that we would want to do that is there is no organism on Earth that we can manipulate and get to tell us its secrets better than yeast. It has an absolutely unrivaled toolkit and it really derives from brewers back about 150 years ago wanting to make better beer, and wanting to understand that organism and how to manipulate it, and it’s taken off from there and it’s just amazing… massive, massive numbers of people have been building and developing technologies that allow us to knock out every gene in the genome or overexpress every gene in the genome, make point mutants where we want in the genome, and so that’s just something that we can’t do in any other organism at this level today. So, here’s how we set things up. We have yeast cells that are growing on… in the top row, there… they’re growing on glucose medium. In the bottom portion of the panel they’re growing on galactose media, and we have a gene that will turn on whenever we give the cells galactose. And so we make a recombinant form of that gene that now will, instead of making the proteins that yeast cells use for galactose utilization, they make different proteins that misfold in human diseases, like α-synuclein, Aβ, TDP-43, Huntington FUS. And you can see that we’ve built… for α-synuclein here, I’ve shown you three different strains that are expressing the protein at different levels and are exhibiting different levels of toxicity, just by the fact that they can’t grow very well. And we’ve done that with all of these different disease proteins and we’ve matched them so that they have the same level of toxicity, so, same level of toxicity from different proteins. What it that? Is this just some non-specific protein aggregation mess? It turns out that it’s not, but when those proteins misfold inside of the yeast cell, they go into the cell, they interact with the same kinds of highly conserved constituents that they interact with in a neuron, and they do bad things in a very specific way. So, here’s an example of a phenotype. The black glob over there is protein nitration and it’s happening, although the cells have the same level of toxicity, the nitrational damage is happening really only in the cells that are expressing α-synuclein. That’s really interesting because in the human diseases that are known to be caused by the misfolding of α-synuclein — and that is Parkinson’s disease, multiple systems atrophy, Lewy body dementia, and neurodegeneration with brain iron accumulation — they too show very high levels of very specific protein aggregates with nitration. So, very unique and very specific cellular pathologies directly related to the human disease. So, here are our cells. We’ve got this gene that we can turn on with galactose, on and off with galactose. And we’ve hooked it up to GFP just so that we could see what’s happening to it in the cells as they were either healthy of dying, and when we had just one or two copies of the protein in the cells, they were fine and the protein went out to the membrane, which is where it belongs. And if we had more, just one extra copy, we started to see things going wrong, and if we had two extra copies it went even worse, this does not look good, these are protein conglomerates here and some type of aggregation. And then those cells grow fine, those cells grow slowly, and those cells die. Very, very strong dosage difference, and what’s really interesting about that is that’s true in man as well. Human beings that have one just one extra copy of the wild type α-synuclein protein will get early-onset Parkinson’s disease, and if they have two extra copies they’ll get an even earlier, more virulent form of the disease. So, this unusual, extreme sensitivity to exactly how much protein you’re making was certainly reminiscent of what was happening in man, so how can we get a better idea of what’s going on here, if there’s anything really deeper involved. Well, we do something called screening. We screen every gene in the genome for what makes cells better or worse, we can take… with yeast, we have libraries of every gene in the genome, we can turn them up or turn them down and see how that changes the disease manifestation. And in these cells that have the 4 copies, where they’re just plain frankly dying of the disease, we can screen for chemical compounds that might rescue them and studying those compounds might tell us something about the disease pathology. So, screening is a lot like panning for gold. You go through a whole lot of stuff and you look through it and you look through it and you look through it and you find nothing for awhile, and then all of a sudden you get these nuggets of gold. So, out of the 6,000 genes in the yeast genome that we studied, only about 60 or 70 of them in our initial study seemed to matter with respect to α-synuclein, and the genes that we got out of our α-synuclein screens were completely different from the genes that we got out of our Aβ screens, and completely different from the ones we got out of our Huntington’s screen. And they told us something about the biology because, for example, the largest class of genes we got were genes that are involved in vesicle trafficking, moving those membrane-bounded proteinaceous compartments around the cell. And so when I showed you these protein conglomerations, here, these aggregated forms of the protein in this cellular model of the α-synuclein pathology, it turns out that when we got that result, that the genes that saved the yeast cells from that pathology were genes that were involved in moving little vesicles around, we thought that, well, gee, I wonder if those things actually have something to do with vesicle trafficking? And so when we looked at them with the level of the electron microscope, which allows a much, much higher resolution view of the cell, you can see that, yes, these little vesicles that are packed with proteins in the cell, depending on how much of the α-synuclein we’re expressing, we get more and more of these protein aggregates. And then we did something called immunoelectron microscopy, where we attached a label to the antibody against the α-synuclein and against a protein involved in vesicle trafficking, and we found that they were there together. So, these blobs, these green blobs here, are actually blobs, not just of aggregated α-synuclein, but aggregated α-synuclein enmeshed in vesicles that are not moving around the cell and getting to the places they’re supposed to be. And when that happens in a nerve cell, it’s really disastrous, because that’s one of the major ways that nerve cells communicate with each other. That’s not good for a yeast cell, either. Anyway, this finding that α-synuclein blocks vesicle trafficking has now been corroborated by many other laboratories. And to cut a long story short and move on to the very final stage of this talk, we found that there were parallel effects, we moved back and forth between yeast and neurons, and we found that there were parallel effects on not just vesicle trafficking, but bursts of reactive nitrogen species, as I showed you in that protein blot, mitochondrial dysfunction, and perturbations in metal ion homeostasis. So at least at this early, very simple cellular level, there’s a lot of similarity there. But we really needed to be able to show that the genes we found in yeast and the genes that saved the yeast cells, that those same genes would matter to a neuron. So, we actually looked at this in a couple of different system initially. One was this nematode system, it’s a worm, it’s a simple little worm, but it’s got lots of different kinds of neurons and in fact it’s got the same kind of neurons, dopaminergic neurons, that are adversely affected in Parkinson’s disease. And we could actually peer through… wire up those cells to express α-synuclein and wired them up so that they’re green, they glowed green, we could actually study them in a living worm, and we could see that when the worms were expressing α-synuclein in those cells… you can see how some of them are disappearing over there? It’s a true neurodegenerative model in the nematode, and our genes that rescued the yeast cell also rescued that nematode. And the same thing happened when we took neurons from rat brains, the midbrain region of the rat, which is the corresponding region that’s affected in Parkinson’s in humans. So that was pretty encouraging. The next thing we did was to screen a chemical library, and this again is something that’s so much easier to do in yeast. We asked whether we could find compounds that would fix more than one problem — I told you, there are lots of things going on, there’s a cascade of pathology that gets kicked off by those misfolded proteins — and can we find compounds that can fix more than one of those problems? So… and the next question was, can we use yeast genetics to find the target? So, why would this matter? Well, we could screen through, and we did in fact screen through 500,000 chemical compounds, asking for which ones were able to rescue the yeast cells. That kind of a screen, which took us several months, would take, I don’t know, maybe 100,000 years if you were using a mouse, and also probably… I don’t know, billions and billions of dollars. We did it much more cheaply and much more easily and much more rapidly in yeast cells. The other reason why it mattered was that yeast cells offered, as I mentioned earlier, unparalleled genetics. And so we’re actually able to take advantage of those genetics to figure out what those compounds were doing to save the yeast cell, and then go back to neurons and ask whether those same compounds would work in the neurons, and whether those compounds would fix the same pathologies that are taking place in the neurons. And it worked. So, we screened 550,000 compounds, simply asked for restoring growth, we don’t… there are a lot of genes, we don’t know which one is the right one to try to go after, we just looked for something that would restore growth. And we’ve only dissected a few of these compounds so far, but they ameliorate vesicle trafficking defects, they ameliorate mitochondrial defects, and they work, the ones we’ve tested, in nematode, rat, and human neurons. So, the final piece of this story is to turn towards human iPS cells made from patients that have one of these diseases. This has been one of the most exciting aspects of… revolutionary aspects of biology in terms of being able to devise better treatments for patients, but you can take skin cells from a patient and actually dedifferentiate those skin cells into an embryonic sort of state, and then redifferentiate them into neurons. And another amazing technology that’s been developed recently by many other investigators has been the ability to surgically genetically edit those cells, such that you have corrected just the mutation that’s responsible for that person’s disease, and so you have absolutely genetically identical cell types here. The only difference between them is the difference that causes the disease, and so then you can ask whether or not you have any pathologies that are different between them and whether any of the things that you’ve discovered earlier work against those pathologies. Except… the cells looked pretty much identical. So, how do we figure out what pathologies might be happening because after all these pathologies only manifest themselves in terms of human disease, even in people who have these terrible mutations, they only manifest at the ages of 40, 50, 60, 70 years old. But we had the ability to go back to the yeast cells and remember what we had learned from the yeast cells and look for those same pathologies arising in those cells long before they started to die. So, long before they started to die, we saw the same problems in vesicle trafficking, the same problems in nitrosative damage, nitrational stress, and when… and they did not happen in our mutant corrected cells, so that allowed us to know that, yes, not only were those pathologies happening, but those pathologies were due to the mutation that was causing that person’s disease. And then we went back and we asked whether or not our compounds could rescue those cells and they could, at least the first few that we’ve tried have. And we then used yeast genetics, as I mentioned, a lot of complicated experiments I won’t take you all through, but we used yeast genetics to figure out what the target of the compound was. And lo and behold, it turned out to be a very, very highly conserved ubiquitin ligase, you can see this is a cartoon of the different domains of the ubiquitin ligase in the yeast cell, known in yeast cells as Rsp5 and known in human cells as Nedd4, and basically every domain is conserved. And this is really an interesting kind of protein to find because the ubiquitin ligases are a very large, complex family in humans. There are about 700 of them in humans; there are about 300 of them in yeast. And they’ve been very, very difficult for pharmaceutical companies to target when they take the protein out of cell and try to do what they have done traditionally over the years, which is try to find chemical compounds that will alter the function of the protein in a purified system. And that is because the biology of these proteins really only manifests within the context of a living cell, where all the proteins are very, very close together and crowded, and moving around, and changing their conformational states. That’s where you see that this protein particularly matters. And then this protein is a very complicated one and it would have been impossible to find without the kinds of simple chemical genetic methods that we used. So, it really demonstrates, I think pretty strongly, the power of phenotypic screens for looking at compounds that really have kind of very special properties, properties that can correct the disease pathology, and the power of chemical genetics to figure out how those targets work. Now, this is really only the beginning. Whether this will ever turn out to be a therapy or not, we don’t know, but what it has done is it has showed us that this ubiquitin ligase plays a very key central role in the way that pathology, that cellular pathology, is manifested in lots of different ways. And so it’s certainly a useful tool. We’re finding more and more of the genes that we found in yeast cells are useful tools in understanding the biology, and we hope one day, maybe, this will be a way of finding therapeutic compounds. But what I do believe is that we need… these diseases are very, very difficult, and we need to try every trick in the book. And so something as unconventional as this, maybe it could provide one key. So, we’re pursuing the same sort of strategy with Alzheimer’s and other neurodegenerative diseases. Many other people are pursuing other varied strategies, but the reason I concentrated on showing you what happens with one particular protein was that, remember that idea that maybe you could alter this heat shock response and soup up the heat shock in these neurodegenerative diseases and take care of the protein folding problem, well, our work with cancer told us that might not be a good idea, it might make the brain cells much more susceptible to cancer, so that’s why we’re going after individual proteins in these diseases. So, the first and final thing I’m going to say is that this work is all thanks to the extraordinary group of people in my laboratory who’ve been working with their whole hearts and souls over the last many years to try to understand the problems of protein folding and protein homeostasis, and try to figure out how understanding those problems might be able to make a difference to mankind. And I just want to say that I couldn’t be happier or more inspired than to have worked with these people, they’re just amazing. Each and every one of them has made a major contribution to this work. Thank you very much.


  • Reply Synat Keam October 26, 2017 at 10:23 am

    i really like her talk

  • Reply Huiting Li October 6, 2018 at 3:24 pm

    Omg I was just searching for videos related to my Final Year Project and found this, and when I watched half of it I suddenly realized she was just the author of a Nature Review I'm working on right at this moment!!! Thanks a lot!!!

  • Reply blasater October 26, 2018 at 7:19 pm

    Heat shock still may be the right answer. We have gotten used to suppressing this for comfort and maybe that is not a good thing.

  • Reply TheFenerbahceSK March 6, 2019 at 3:44 am

    Thank you very much for everything you've done.

  • Reply TheFenerbahceSK March 6, 2019 at 3:56 am

    How do you get extra copies of the gene?

  • Reply TheFenerbahceSK July 8, 2019 at 4:21 am

    14:20 Can someone please give me a link to the paper where she discusses the genes they got out of their Huntington's screen?

  • Leave a Reply