Prof. James Spudich on Building Translational Research Ecosystem.

I have never had an introduction like that, and the stories are all true. I can't imagine where you found them. Unbelievable. So I've cut down my slides to the bare minimum because I knew the timing on the schedule was going to have me right up against lunch, and that's a problem. So I cut out a lot of data slides; you can go back and read papers. I'll show a little bit of data, but not very much. But I wanted to get the message across that if you want translational research, if you want to develop a society for translational research, then you have to invest in basic, very basic research. My research: I'm trained as a chemist initially, a chemist, a physicist, and had very little biology courses actually until Woods Hole, where we met. And my work has been very fundamental. As you'll see, I'll give you a little bit of that data having to do with very, very basic understanding of what the contractile system does in cells. And I'm going to tell you the personal story of how that basic research led to the biotech companies and then to medicines. So that's the lesson today. So when I started my laboratory many, many years ago, people did not know how all of these kinds of movements you're seeing in this slide occurred. Muscle had been studied, and myosin, the molecular motor in muscle, was known, although it was not known how it really worked. But it wasn't known whether there was myosin in non-muscle cells. So one of my starting projects, which has been consecutively funded for more than five decades by the NIH, was given total support to very, very fundamental research to understand how all of these movements occur. In the lower right, of course, you see a beating cardiomyocyte; that's what causes your heart to contract. But in the upper left, you see a cell migrating over a surface; in the middle, on the upper panel, you see a cell dividing into two daughter cells. All of these kinds of events require a motor protein that can cause this kind of force production to cause these kinds of movements in cell division. And the lower left is plant cells with cytoplasmic streaming. So life is movement, and it's driven by molecular motors. Since I started my lab, my own contributions plus a lot of my friends and colleagues in this field have shown that in fact myosins, but also another molecular motor called kinesins, are driving these movements. But how does it really work? That's what the basic research has been on my part for many, many years. So I'm going to tell you about one myosin type. They're different myosins; they look a little different, but they're pretty much like this: they have two heads with a coiled-coil tail. These movements and force production are really driven to a very large extent by this particular molecular motor, which is my favorite enzyme in biology. This myosin has to bind to another protein called actin, which makes these filaments colored in blue here, in order to produce the contraction. So it's an actin-myosin interaction driven by the hydrolysis of ATP, which is the fuel that drives the contractile event. Now, in early years, we studied an amoeba, as I mentioned in my little preface, called Dictyostelium discoideum. It's an amoeboid cell that actually looks very much and behaves very much like a leukocyte in your body. So we were studying it as a model system because we could manipulate it by genetics and structural biology and biochemistry. In this particular experiment, we genetically removed the myosin from this cell to see what the cell would not be able to do. And of course, it wasn't a complete surprise that when we eliminated the myosin, the cell could no longer divide into two daughter cells, because as I told you, we knew you needed a molecular motor to do that. By the way, we discovered that myosin exists in this cell; nobody knew that before. Having shown that, we provided the genetic proof that you needed this molecular motor for cell division. We were then able to use molecular genetic methods to put myosin back in using a plasmid, and rescue the cytokinetic event. Then you could start making mutations on the myosin and see how you can modulate it and still rescue, or whether the mutation failed to rescue. So there was a lot of work of that kind being done. This particular experiment was simply to put a fluorescent tag on the myosin molecule so you'd be able to see it in the cell. When you put the fluorescent tag on, you can then watch where the myosin is in the cell and where it's going, and you can see it accumulates in the middle of the cell, forms like a little muscle that's transient, and that divides the cell into two. So this is a decade of work with, I don't know, 50 to 100 papers describing much of this work. But this was very, very fundamental research, and at this time, if you told me that I'd someday be starting a biotech company, I would say that's highly unlikely. But this deep understanding from this model system that we got was essential to really go to where I'm going to take you now. Because having this deep understanding, there were textbook models of how muscle contracts, but as you all know, in the textbooks, those are largely models. There's often not a lot of data that proves the model. It's very hard to prove a model; you can mainly in the laboratory try to disprove your model. If you accumulate enough vast evidence, then you can pretty much believe the model. When I entered this field and looked at muscle, there was a lot of confusion even though people thought they knew how it worked in the textbooks. There was a lot of confusion about what was really going on if you were in the muscle meetings that we all attended. And what was missing, I thought, being trained as a hardcore biochemist, was you need to be able to reduce movement and force production, which was so important for muscle contraction and all these non-muscle movements, to purified proteins. So you need functional assays for movement and force production. It took us many years; it turned out to be very simple once you knew how to do it. But we reconstituted with purified actin and purified myosin a motility assay where we could watch the proteins move in relation to one another. One trick: the actin filament is only nine nanometers; a nanometer is one billionth of a meter, so it's tiny. You can't really see an actin filament without doing something, but you can put a fluorescent molecule on the actin, and then under a good fluorescence microscope, you see what you see in the black square here. You can see individual actin filaments which have landed onto a glass slide where we have placed a lot of these myosin molecules, which you can't see because they're not labeled with fluorescence. But the actin goes and it binds to those myosin molecules, and nothing's going to happen because there's no fuel. You need some chemical energy that's going to get transduced into mechanical motion. So when you add ATP, what you see is this, and this is almost a real-time movie. So you can imagine when my student Steve Kron first got this to work and ran into my office with such excitement, I can't tell you what he said; it was a four-letter word. Very excited he was. And it's a robust assay where you can actually measure the velocity. It turns out the velocity is very close to the maximum velocity you can contract your muscle or the maximum velocity that's occurring when your heart is contracting. So it replicates that. So this was fundamental and really changed the whole field because suddenly you could study the motion with a quantitative assay. But how about measuring the force? This was already mentioned to you. What we had to do was to combine physics with biochemistry, because in this case we needed to use the physics of laser trapping. As you were told in the introduction, Steve Chu, who was a physicist trapping atoms for which he got a Nobel Prize in 1997, had also discovered accidentally that you could trap larger particles. So I went to talk to Steve, and he agreed to help us build a laser trap, which was built actually by my student going over to Steve's physics department. And as you were told, this led to the development of Bio-X, which was very exciting for Stanford and a lot of other places that have replicated that approach. But what we were able to do was to use laser traps shown here in red, focusing them onto one-micron polystyrene little plastic beads. When you focus the laser trap on a particle like that, it actually traps it in space. So if you move the laser beam around, you can move the particle around. By putting a little bead on each end of a single actin filament and having two laser beams, you could take the actin filament and move it around and put it on top of this bump, which is on a glass slide, and that's just another plastic particle where we've managed to put a single myosin molecule. We were driven to have to do this incredible experiment because without doing this experiment, the field was at a standstill as to how this motor really transduced the chemical energy into mechanical motion. There were widely different views at this point, and this was the experiment that really showed what's going on because it demonstrated that you get a very small step in motion, and we were able using this single molecule approach to actually measure the force a single molecule can produce. That force is two piconewtons, really, really tiny, but you have billions of myosins in your heart. So when you take two piconewtons per myosin molecule, it actually adds up quite accurately to how much force you need for your heart to function. So lots of academic studies for many years were really essential to set the stage for translational research. About a dozen years ago, we decided to focus all of our efforts on studying a disease called hypertrophic cardiomyopathy, which is a big problem all over the world, certainly here in India as well. It affects like one out of 200 people; that means someone in this room undoubtedly has it. It's the most common cause of sudden cardiac death in individuals under the age of 35. I'm sure some of you in the room know someone who died that young suddenly, maybe without even realizing what they had; probably they had hypertrophic cardiomyopathy. It's a genetic disease, and about a third of the genetic mutations that cause hypertrophic cardiomyopathy result in mutations in this myosin molecular motor which is operating in your heart; it's called beta cardiac myosin. Importantly, those mutations are causing the myosin to be too good a motor. It's doing too good a job; the power output that this myosin is producing is more than you want it to be. So your heart is hypercontractile. I like to say it's as if you were out for a run when you're just sitting here listening to my lecture. So your heart is overworking; your body thinks you're exercising, so your heart starts to hypertrophy, which is what happens when you go out for a run. If you're an athlete, you have a larger heart than the rest of us, and your skeletal muscles also increase. That hypertrophy, if you stop exercising, will go back down. So that hypertrophy is normal if you're exercising. If you're carrying this mutation, it's as if you're exercising 24 hours a day, seven days a week; you're never stopping. Your heart continues to hypertrophy, and then you get HCM, shown here schematically, where your heart muscle has hypertrophied to such an extent that your left ventricle is diminished in size and is no longer able to push enough blood through your body. That's a very serious stage of the disease. So again, if you want to develop a small molecule therapeutic, which is the story that I'll finish with, you need to understand what these mutations are doing to the human cardiac myosin that cause it to be a better motor, better in the sense that it's producing more power output, not better for your health. So here's where I just took out all my slides in order to save time. Here's the answer. There were many possible explanations for what these mutations were doing, but the answer turned out to be totally different from what we and everybody in the field thought it was going to be. It turns out the answer is that as you're sitting here, about half of your myosins in your heart are in a state shown in the upper right, where the two heads are not sticking out where they can interact with actin and produce force, but the two heads are folded back onto their own tail, so they can't interact with actin. Half of the cardiac myosins in your heart are in that state. If you go out for a run, you have normal physiological responses that alter the myosin molecule, and more heads come out of that state and interact with actin. So this is your normal physiological way of increasing your power output. But in the case of HCM, it turns out, and I'm not showing you any of the data, you can read the papers, to all of our surprise, it turned out that what HCM mutations are actually doing is causing more of these heads, which are normally half the heads in this off state, to open up and interact with actin, and therefore you have a higher power output. So that's about an hour seminar worth of data which I'm not showing you. But understanding this, one can then try to think about: if that's the case, why can't you develop a small molecule that will bind to that myosin directly and cause it, if you can cause it to go back into the folded state, and just reverse the process? That's as good as you can do in the world of translational research. You have a myosin molecule where you know its mutations in that molecule that are causative of the disease, and now you're going to try to make a small therapeutic agent which binds directly to that defective myosin and basically reverses the hypercontractility back to normal. That was the goal, and it was with that goal we started a company called MyoKardia. Now let me emphasize with this slide that before MyoKardia, there was no treatment other than, if you get to a severe stage, the only thing you can do is open heart surgery and myectomy to cut away some of this excess muscle. There's another approach where you can inject alcohol, ethanol, into the heart, causing a heart attack, with some of that dead cardiac muscle then being eliminated. But both of these are horrendous and not really treatments you want to do. And this was all that was available for one out of 200 people suffering from this. So we then started MyoKardia and developed a small molecule called mavacamten. This is its structure. For those of you who don't understand what's involved in drug discovery, it's a very, very hard process. You have to have an assay, like the actin interaction with myosin could be the motility assay I showed you, or it can be the fact that actin activates the myosin ATPase, and you can use that in a screen of hundreds of thousands of small molecules. You look for what are called hits, that is, you look for a few molecules out of that library that actually, in this case, inhibit the actin-activated myosin ATPase or the velocity or whatever your assay is, to reduce the power output back to normal. You find such molecules; typically you find one out of about a thousand small molecules are hits. And then you're at a starting point, because you have to make that small molecule work in a potent way, much more potently than the hits typically are. Furthermore, you don't want that molecule inhibiting skeletal muscle or smooth muscle; you want it to inhibit only cardiac myosin. Cardiac and smooth and skeletal muscles are 95% identical, but that 5% difference enabled us to make a cardiac-specific drug. So you need potency, you need specificity, you need permeability, you need solubility, you need stability of the drug. All of these things are a very hard process which involve a lot of chemistry, dozens of chemists working full-time doing what's called SAR, structure-activity relationships, to increase the potency and so forth. So that's what you do. It turns out that what mavacamten does, which you can only discover by going back to basic research, so my laboratory at Stanford collaborated with MyoKardia, which is just 30 minutes away in the Bay Area, to carry out a series of studies. Those studies showed that what mavacamten does is basically reverse what the HCM mutations do. So the HCM is putting more heads in play so that you are hypercontractile, and if you use the right dose of mavacamten, you can take those excess heads that have been put into play, probably not the same ones, but that number of heads out of play, to reduce the power output back to normal. The hypothesis was that would very quickly help patients. The further sequence involved in drug discovery is you got to show it works in cells, you got to show it works in animals, and then you have to go through phase one toxicity studies, phase two efficacy studies, and then finally phase three studies. Most drugs fall out along this way, so it's a really tough process. MyoKardia took this all the way through phase three clinical trials. The company was founded in 2012; eight short years later, phase three clinical trials were completed that looked spectacular. So spectacular that Bristol-Myers Squibb just bought the company. Bristol-Myers Squibb immediately got FDA approval, and on the market for the last six months or so has been mavacamten, which somebody gave the name Camzyos; I'm not sure how that happened, but I guess it's catchy. This drug is now available. I know that Bristol-Myers Squibb is negotiating with India about getting this drug into this country, but it's being used now in the United States. I have one personal testimonial from one of the very first patients treated with this. Her name is Charnel Mumford, a 67-year-old woman. She was getting treated at Stanford. Within a few weeks of starting this therapy, she said, 'I noticed little things at first. One day I was able to walk up a hill,' which she had not been able to do; it was increasingly hard to do for her. But she noticed she was doing that much more easily. And then not long after that, she's now walking marathons. This has been transformative for her, and I hear more and more stories of this kind, which is incredibly satisfying. Furthermore, I'll tell you that patients who are being treated who were already hypertrophic, as she was, we are already seeing within a few months that the hypertrophy is decreasing. I'm not surprised by that, because as I told you, the hypertrophy is a result of being hypercontractile; your body thinks you're out for a run, so your heart hypertrophies. If you stop running, the hypertrophy decreases, and that's exactly what we're beginning to see. So conclusions: myosin is the most beautiful, important enzyme in biology. I have no prejudices about that. It's involved in all of these different biological functions. Really fundamental basic understanding of the molecule was really established in academic studies in my laboratory and others, without any idea in the early days that this was going to lead to drug discovery. So you have to invest in really hardcore basic research if you want translational research. This understanding made it possible for clinical studies in a biotech environment, which gave rise to this drug. With that, I'll thank current and recent members of my lab and others from MyoKardia who I don't have on this slide, collaborators, unending funding from NIH for five decades; I still have grants from them working on hypertrophic cardiomyopathy, where there's still more to do. Thank you.

I think if I heard you right, you're asking how many molecules did we have to go through with the SAR to get mavacamten in the end. Typically in drug discovery, it can be thousands. My wife and I have founded another company, Kinomics, actually targeting myosins in parasites for malaria treatment, and I can tell you we've already in our SAR in that company made over 1500 derivatives, and we still don't have a molecule that we think is close to being a drug. So it takes a very long process. It was quicker in the case of mavacamten; the initial hit already had pretty good selectivity for cardiac. We were lucky. It wasn't totally selective; that had to be improved. It was already like the potency was already better than typical, and the number of changes on the molecule that we had to make to make mavacamten was really minimal. In that case, it was only something in the order of less than a hundred different changes that had to be tried before we got the molecule we wanted.

You bring up a very important point, which of course I completely eliminated. In all of the companies that I've been involved in, the structure is really important. So we know where mavacamten binds; we know the pocket, and we knew that even with the earlier hit. So using modern AI approaches is really important in structural approaches. But what you do is you look at not mavacamten but an earlier precursor that's bound to that pocket, you look at what the interactions are, and you look at neighboring residues, of course. Then you can predict. The pocket's only so big, so there's no point in putting some big structure onto the molecule; it's not going to go in. So both size and where you have hydrophilic versus hydrophobic side chains become sort of obvious from the structure. So you go back and you modify according to that knowledge base. I think that's what you were getting at. Maybe one more? Oh, my wife has something she has to say.

It's very hard to do drug discovery in the way I described without having a high-resolution structure of the target protein of interest, because you really need to be able to do this SAR using that structural information. It's not always necessary; there are a lot of drugs that did not go through that process, but it's a lot harder and your success rate is much lower.

Your first question about basic research: that's why I went out of my way to emphasize that understanding your target, if you look at statistics in drug discovery, gives you a much higher probability of success. As I described in my case, understanding the target really involved fundamental research, starting with studies with amoebae. So I think there's no question that it's very difficult to just plunge into heart and say, 'I want to reduce power output.' How are you going to do that? You need to understand that molecule in real detail. There was no way to understand human cardiac myosin in that kind of detail; you couldn't get it to study. So that understanding came from other model systems. So very basic research is very important to support by whatever government is supporting it, no question about that. None of this screening was done in my lab. The understanding that I described of how the molecule works, the development of the laser trap, motility assay, so on, done in the lab, but the screening is done by the company. In the case of MyoKardia, it's not even done in-house; you can do that in India. The company that my wife and I now have, Kinomics, we have 12 full-time chemists here in India. So we tell them what molecules we want them to make, they make them, and they send them to us, and then we study them in our Kinomics lab. Your third question was about if you have a small molecule, how do you get the interest of a company like Merck or Glaxo, or if it's for heart failure, maybe Novartis. If you develop something like that in a small biotech company, the big company is going to be mostly interested in it if you've derisked it, meaning you've shown that it's been through all of these permeability issues and stability issues and it's ready to go in the clinic, and it actually works in animals. If you know that from your biotech company, then a big company would be interested in making a deal with you. In this case, MyoKardia, the CEO said, 'No, we're not going to make a deal at that stage. We're going to take it into phase one, two, and three ourselves as a small biotech company.' And then of course it's worth a lot more to the big company if they want it, and that's why they paid $13.1 billion to buy MyoKardia. I don't know if that answers your question.