专访：蛋白质折叠与神经退行性疾病的研究

Could you please introduce yourself to us? I'm a member of the Whitehead Institute for Biomedical Research. I'm also a professor at MIT and a member of the Howard Hughes Medical Institute.

I'm gonna talk to you today about some very interesting new developments in the world of protein folding. What is the protein folding problem? Protein folding sounds like a rather esoteric subject, but it's actually really central to biology.

The, the basic problem is this, the linear information that codes for us, codes for everything, every organism on the planet is DNA and it's one long string of code and it doesn't do anything terribly interesting. What it does is it codes for proteins and basically proteins are us, we're all made up of proteins and all of the things that we do that are, that are interesting primarily are due to either proteins directly or the action of proteins. But the difficulty is that the proteins are very complicated structures when they're first decoded from the genetic information, the DNA, they start as out as long strings of linear amino acids, pro polypeptide, and that does nothing.

It has to fold in this very, very complicated sort of, or game-like way, do a very, very specific structure for it to function. And getting that done in this incredibly crowded environment of the cell is actually quite difficult. And so there are very large number of human diseases, for example, that are due to problems in protein folding.

And my work has moved from, from doing really very, very basic biological experiments and biochemical experiments in test tubes aimed at trying to understand really the very basics of how proteins fold and why they misfold and what goes wrong to now working on, on much more complicated systems that are involved in, in human diseases with the eventual aim of hoping to make a difference in, in how on disease therapeutics, What are the diseases caused by problems of protein folding? The breadth of problems that are caused by protein folding difficulties is, is absolutely enormous. It includes everything from diseases like cystic fibrosis to Alzheimer's disease to emphysema.

Increasingly recognized that adult onset diabetes is caused by problems in protein folding. Many cancers are caused by problems in protein folding because this protein might accumulate a mutation. And if that protein would only fold properly, it could function properly and, and not go about telling cells to grow, grow, grow inappropriately, but the mutation prevents it from folding properly and, and being inhibited in its activity, regulated in its activity.

Protein folding is also plays a big role in the way in which organisms can evolve and that means that it also plays a role in the evolution of drug resistance and things like this. So our, our research spans a a really very broad range and it covers genetics and biochemistry. It covers a wide number of organisms.

We work with yeast, we work with fruit flies, we've worked with mice, we work with tissue culture cells in in culture, and we've worked with rabbit opsis, a little small mustard plant. That's a, it's a very good model for certain kinds of biological problems. What are the mechanisms of pry diseases?

Prions are a very specific type of protein folding problem. At first it was thought to be a very, very bizarre and unusual aspect of protein folding. And now we are beginning to appreciate it that it much might be much broader.

PR are, are actually most famous, at least in, in the public mind for causing a disease called mad cow disease or quotes called ya of disease in humans goes by various names, but it's due to the misfolding of a protein. And the protein has the very peculiar property that when it misfolds, it gets other proteins of the same type to misfold in the same way. So you end up getting this confirmational chain reaction that changes one protein after another after another.

Somehow we don't still understand exactly how, but somehow or other that process with this particular protein, I think it's in part because the protein doesn't, doesn't alter its folding efficiently enough. It's dwelling in these intermediate states which are toxic. You wind up getting disease.

And the odd thing about it is because the protein has this self perpetuating ability to transmit a different fold to another protein of the same type, it's infectious. So the idea that a prion a protein could be that cause of a, of a infectious disease was very bizarre at the time. And Stan prisoner had did, and, and there were many other people involved too, but done wonderful, wonderful work on really taking a a, a hypothesis that everybody thought was nonsense and realizing that and creating the, the work that demonstrated that it was true.

How could a protein be infectious? Well, it turns out it's really pretty simple. It's one of our own proteins.

It's a protein on our, in our, on our nerve cells. And it's there all the time, but it's not normally misfolded. So what happens is if we get some of this protein misfolded, and that can happen by sheer accident once in a while, luckily that's pretty rare or it can happen because we've eaten some infected meat or it can happen because we have a mutation in the protein that makes it more likely to change from this shape to this shape.

And but once that protein appears and that altered the shape, it then will start changing more and more and more the other proteins in the cell and that both winds up causing this toxic, toxic cascade toxic species. It also winds up being able to generate more of itself in effect. cause you start out with just a very small amount of this protein in this fold.

Most of it's in this fold pretty soon it sweeps across the brain and and changes the folding of all the proteins. And so, so that winds up creating more and more and more and more infectious material. So when, when a neuron comes in contact with that infectious material from a variety of different sources, it can cause disease.

Could you please explain how protein can act as a genetic material? It winds up being a very fascinating phenomenon that you have this self perpetuating change in protein folding. cause as you can imagine, the protein's gonna have different functions in these two different states.

So if a cell has protein in this state, it will behave in one way. It have one set of characteristics. All the proteins in this state or most of the proteins in that state will have a different set of characteristics.

And it turns out that that becomes a genetic material. The protein can actually cause an heritable new trait to appear in the cells because of that change in function. And it's heritable because when the mother's cell buds off and starts and to create a daughter cell, she passes some of that protein to the daughter's cell.

So then when the daughter starts to make her own proteins, instead of going into this shape, they'll go into this shape. And so that will keep on for generation after generation after generation. And you actually have a mechanism for inheriting a alter altered trait, an altered phenotype.

It's due to the inheritance of a protein that changes the way it folds. And there's no change in the nucleic acid in the cell at all. So they're very different just as the prion diseases were a very different type of disease.

The PreOn in terms of these genetic elements are, are very different kind of genetic element. How did you choose yeast as a model organism to investigate neurodegenerative diseases? So when using yeast cells then to start studying these difficult protein folding problems that have to do with the brain, we realized that it might be a reasonable idea to take some of the proteins that are known to misfold in various neurodegenerative diseases of man, put them into yeast and see if their misfolding would cause problems for the yeast cell as well.

Now I will have to say that when we first decided to do this, a lot of people thought we were crazy, but it's turning out that it's working pretty darn well. The motive of course, is that you can grow yeast cells very easily and cheaply. And moreover, starting from a hundred years ago when man was interested in making better and more reliable types of beer, geneticists have been working on yeast and making over the, over the a hundred years, they have made one advance after another, after another.

So we can now do things in yeast in terms of genetic manipulations more rapidly than we can in any other organism. It's just plain amazing. So, and the advantage is that yeast, our cells, believe it or not, are pretty much like us.

They have, unlike bacteria, they have membrane, all them say membrane bounded compartments that we have in our cells. They have a membrane bounded nucleus, they have membrane bounded vesicles inside the cell. They have an endoplasmic reticulum.

They move things from the endoplasmic reticulum out to secretion from outside the cell via, via a whole series of membrane bounded compartments. And all of the ways in which that works, the ways in which that's regulated are, are really rig done the same way in yeast cells as as they are in us. So there's a lot of what we call hemology, genetic hemology between the, the basic underpinnings of, of, of cell biology in a yeast cell and the underpinnings of cell biology in us.

Now of course we realize that there are gonna be aspects of the biology of neurodegenerative diseases that we won't be able to ever touch. The idea was that for some of these diseases that are due to protein misfolding, well that's a pretty ancient problem too. And in fact the ways in which yeast cells deal with problems in protein folding we've known for years are really pretty similar to the ways in which higher organisms do too.

So we thought it might be worth trying to take. Some of these proteins, which misfold and are, are, are studied by neurobiologists and are known to be their misfolding is known to be associated or misfunctioning is known to be associated with disease, might be able to put them into yeast and learn much more rapidly. Something about why are they toxic?

Why, why are they suddenly causing toxicities? Now if the protein is doing a nerve specific function and interacting with other nerve specific proteins to cause toxicity, we'll never be able to study it needs. But in those cases where the protein is doing something really fundamental and basic and cell biology, we should be able to study it in yeast and, and then move rapidly from yeast cells, whatever we learn in yeast cells.

The I idea would be then to move quickly, as quickly as you can into nerve cells and see whether or not you've in fact been, you know, been wasting your time or whether in fact you've learned something that is of useful knowledge. How do you apply high throughput screening in yeast? Recently the things that we have done are, for example, to do, to take advantage of high throughput screening mechanisms, which have been developed for yeast cells and much easier to do in yeast than to do in neurons.

And so we have been able to take every single open reading frame in the yeast genome, we can test every single gene for whether it make the, makes the cells worse or it makes them better. So we can do a genome wide screen for genetic modifiers of the toxicity of alphas nib and date. And then take those genes that we've gotten from yeast and we look for homologous genes in, in humans.

We take the human gene and then we go into some other model systems where there are, where there are neurons and in fact where there are the specific type of neurons that are characteristically sick in Parkinson's. People with Parkinson's disease, sick and dying, and this is a type of neuron called the dopaminergic neurons. They transmitted a particular neurotransmitter called dopamine.

And the idea has been for a long time that in order to study this particular disease, you'd really have to study it in dopaminergic neurons. You couldn't start study it in yeast. We're thinking, well maybe there's some general aspect of cell biology that's going wrong and that maybe those dopaminergic neurons are just more sensitive to it.

And it's really a very general basic cell biology process. So we take the genes that we get out of, out of yeast cells and we then have been collaborating with some wonderful, wonderful collaborators to now look and examine neuronal systems. Can you tell us a bit more about the genetic studies in model organisms and your collaborations?

So we've collaborated with Caldwell Laboratory. They have a little model of Parkinson's disease and a little nematode, a little, this little worm. And in that worm they express the same protein in dopaminergic neurons.

The nematode has eight dopaminergic neurons and it always has eight dopaminergic neurons unless it's expressing off a nucle. And then many of the animals don't make the normal and don't have the normal number of neurons. They actually start out with a normal number of eight neurons, but they, in an age related way, those neurons will die.

And we found that the gene that, that the first gene that we found is a strong suppressor of toxicity in yeast would also suppress the toxicity of alpha Snu and this little nematode. And then we went to work with Nancy Bonini, who has a fruit fly model, a little bit more complicated nervous system, a little bit harder to, to study, but we, we found there too that this gene rescued the yeast cells. It also rescued the dopaminergic neurons in the fruit fly.

And finally we worked with Chris Che in Indiana who has a model for the selective toxicity to dopaminergic neurons that involves taking primary neurons from the mid brain region of rats. And these are a mixture of dopaminergic neurons and other neurons. He uses embryonic rats.

And when he plates them out, he has about six to 7%dopaminergic neurons in the, in the culture dish. And then he takes a virus that is expressing alpha synuclein and particularly a mutation that's that's known to cause early onset Parkinson's disease in people. And because it's prone to misfunctioning when it has that mutation and misfolding.

And we then, and what he has found is that if he infect all the cells in the culture dish with that virus, the dopaminergic neurons are particularly sensitive to it. And we've been able to to show that if we then express this gene that we first identified in yeas at the same time as expressed in the mutant alpha nucle, it suppresses the toxicity. So that was the first, first major result we had.

The things that we were doing in yeast cells could really translate into something meaningful in a, in a, in a neuron. Have you performed high throughput chemical screens? I don't think we've done has been a high throughput chemical screen.

And there are libraries of potential drug-like compounds that are available at various major research institute repositories. Harvard has one, the broad, which is located right next door has one, you've got a hundreds of thousands of comp, different types of compounds and they will, it's possible using simple robotics to get very small quantities of these, of every single one of those compounds and to test them in, in whatever your assay is. And these dishes that have lots of little wells in them.

And so we can take our yeast cells and put them out into these dishes and then we test one by one each of these compounds. And we were able to screen through well over a hundred thousand compounds in a matter of a couple of months. And that was even when we were just beginning at this.

And then we of course we had to take a lot of time to retest and retest, but we came up with about seven compounds in the end that were quite efficacious. And we've done most of our work with three of those that are, that are quite closely related to each other. So what we had asked for there was for the cells to return to life.

So you turn on this gene, so associated with neurodegenerative disease, it kills the cell. We're asking for the cells to return to life. The reason why we use that particular assay rather than, for example, looking at the folding of alpha sinin is this is a very complex disease.

And even in the yeast, the genetics is telling us that it's a very, very complex process. A lot of different things are going wrong, it's not just one thing going wrong. And so we wouldn't know really which protein to start with to find the right target for a drug.

And so what we decided to do was to let the cells tell us which compounds would make them feel better. And that would also mean that that compound had to be able to get into the cell and it would also have to be not toxic because if inherently, at least in the yeast cell, so it doesn't have any general toxicity. And we've taken those compounds now and we've again gone into the same models I just told you about with respect to the nematode and the rat primary neurons in culture.

And we've been able to show that our compounds will rescue the nematode and will rescue the rat primary neurons. So that tells us that the process that we're seeing in yeast is a pretty good mimic of some of the toxic, at least some of the toxicities that are occurring in neurons. We have begun to learn and understand what the, the office nucle is involved in protein trafficking for which there had been some hints, but we've established that far more clearly I think than had previously been known.

We've un, we've uncovered a lot of other genes so we know, know something about the complexity and it's very heartening to me to think that it might be possible that there could be compounds that could correct those toxicities. A lot of, lot of work to be done before get to a, get to a drug unfortunately, but at least this maybe is a, is a faster way of, of getting to identify some compounds that could be beneficial. In summary, what would you like to tell us?

These are some of the protein folding problems that we study in my lab. We actually study a lot of other protein folding problems. We started out in the, in the protein folding world quite early on when it, when it really seemed like it was kind of maybe a narrow topic, but it's turned out to be so central to to general cell biology that we wind up have, we have wound up studying things like the role of protein folding and stress responses in cancer biology in the way in which organisms evolve in drug resistance in many, many of the other aspects of, of biology that are important to mankind.

So hopefully using simple systems, our, our ideas have generally been to try to find the simplest place to start and then to move out into the more complex systems. Once you've got something, once you've got a hold of something.