Hi, everyone. Thank you so much for joining us for today’s presentation in Salk’s Lecture series, the Power of Science. We are still operating remotely for these types of events. It’ll be great to have you back on campus eventually but we’re not quite there yet. I’m Cheryl Dean, Salk’s Planned Giving counsel and I’ll be your moderator today. The last time that we gave one of these presentations, I had several people asking me what I’m doing when I’m not putting on one of these science talks for you. My main role at Salk is working with people who have included Salk in their estate plans. So if any of you are considering including Salk as part of your legacy, I would be more than happy to speak with you and let you know a little bit more about that.
But that is not my role today. My role today is to introduce one of our fantastic scientists. But before we get to today’s talk, I wanted to mention Salk’s research efforts related to the Coronavirus. Here’s a link to our website where you’ll be able to find more information. And we’ll be sending you a link to today’s talk. You don’t have to take copious notes, and you’ll be able to get this information in a few days. I just wanted to mention that Salk scientists have been making some really incredible discoveries, including one just last week regarding the inflammatory role of the SARS-CoV2 spike proteins. But that’s not the subject for today’s talk. But you can certainly find more information about this on our website.
Before we get started, though, just a couple of tips for those of you who haven’t necessarily been on a webinar like this in recent times. At the bottom of your screen, you’ll be able to see a yellow arrow pointing to the closed caption button. And as you can see from the slide on the right side of the screen, you’ll need to click on Accessibility under Settings. And then you can adjust the size of the captioning. Hopefully, that’ll help for you. And the other thing you need to know is that some of you have already submitted questions when you register, thank you, I have those.
But if you think of new questions, then please hit where the yellow arrow is pointing, there’s a Q&A button on the Zoom window and type your question there. We’ll try to get to as many of them that you send in today as possible. And if we don’t get to yours today, then hopefully Dr. Stites will be able to get back to you before too long. Getting to today, we have The Power of Science, Mutants, Bad Influences, Loners, and Cancer. As I mentioned, today’s speaker is Dr. Ed Stites. He’s originally from the east coast, born on Shaw Air Force Base in South Carolina. And he also lived in New Jersey and Connecticut before his family settled down in Kentucky.
After getting his bachelor’s degree in mathematics from the University of Kentucky, he went to the University of Virginia to join their dual degree MD, PhD program. In that program, he went to medical school to become a medical doctor, and he also earned his PhD in biophysics, performing research that combined mathematics with the study of cancer causing gene mutations. Lots of studying. After completing both degrees, he went to the Translational Genomics Research Institute for two years, a research fellowship that focused on pancreatic cancer and genomic medicine.
He then completed his medical training at Washington University in St. Louis, where he completed his residency in clinical pathology. Joined the Salk Institute in 2017. And he is a fully licensed and fully board certified medical doctor, but he chooses to spend his time focusing his career on cancer research, as he believes he can make a bigger impact on patient outcomes through research. His laboratory studies cancer with both computational and experimental models, and his lab is currently funded to study colon cancer, lung cancer, and melanoma, and focuses on a gene family known as the RAS genes. On the rare occasions when Ed is not in the lab, he enjoys spending time with his family, traveling, playing games, and watching University of Kentucky basketball. Please help me welcome Dr. Ed Stites. Ed, you’re muted.
There we go. Like that. Thanks you all. I appreciate the very kind introduction. And thanks to everyone who’s here today to listen to us talk about the research work we’re doing, trying to make progress against cancer, combining these mathematical approaches with traditional cancer methods. This talk, which we call Mutants, Bad Influences, Loners, and Cancer talks about some of the most recent and most exciting things we’ve been doing in the lab. Now, thanks to pop culture, if you hear the word mutants, if you’re like a lot of people, you might start to think of things like this, comic book mutant characters that fill pop culture and movies and TV and comics. But this isn’t a talk …
Even though it’s a talk about mutants, it’s not a talk about the Teenage Mutant Ninja Turtles or the mutants from X-Men. But rather, we’re going to be talking about spelling errors in DNA, because that’s really what a mutant is. DNA, the blueprint of life, this gigantic, extremely long biomolecule built up, I mean, cytosine and guanine, the ATCG that spells the blueprint of life. Sometimes that blueprint, sometimes those instructions pick up a spelling error, and that’s a mutation. And the gene that that was coded by that stalling error, it’s now slightly different, that’s a mutant protein. Sometimes these mutant proteins are bad and damaging, sometimes they can cause diseases like cancer, but that’s not necessary. Some mutations may be beneficial, a lot will have no effect. Mutations and mutants are actually essential for us to be here. Mutations are the driving force of evolution.
Let’s see. There we go. We’re going to be focusing on one specific protein coated by the RAS genes. This is a molecular model of RAS. As far as proteins go, this is fairly nondescript. It’s kind of ball-shaped, there’s not a lot of interesting features. It’s not particularly big relative to proteins. It’s very small. Just for scale, those little circles is a single atom. It has these hundreds of thousands of atoms that build a single RAS protein that’s kind of shaped like a ball. Amazingly tiny. If you were inside of a cancer cell and you looked around and you saw one of these RAS proteins, you wouldn’t think much of it. It doesn’t look very different from any other protein in the cell.
You’d see a lot of them. There are about 200,000 RAS proteins in a cancer cell. So off on the right, there’s a picture of maybe 20 different colon cancer cells, and each one of them would have about 200,000 RAS proteins. That’s a lot. But there are millions and millions of proteins in each of those cells. RAS isn’t particularly abundant relative to other proteins, it doesn’t take up a disproportionate share of the volume. Again, if you’re that size and you’re looking around, you wouldn’t notice anything special about RAS. But RAS does a lot of really, really special things. Perhaps the most important thing is that it controls when a cell divides.
RAS can exist in two states. When it’s off state, there’s no solid rope. That cell’s happy, it stays where it is. But if RAS gets turned on, it will instruct the cell to grow. So it gets bigger and bigger, it divides itself in half, and now you have two cells instead of one. And if RAS stays on, it’s going to keep dividing and keep dividing, and the cells are going to keep doubling and doubling, and you’ll keep getting more and more cells. Honestly, this is a lot like what cancer is. Cancer is one single cell, picks up mutations that tell it to keep growing uncontrollably.
And it doesn’t take very many doublings for these cancer cells to become bigger and bigger and more and more massive. This growing massive cells or a tumor is cancer. And it can spread maybe to other parts of the body. Based off the Mass Effect, it could compress vital organs. Just by growing this much, it takes a lot of nutrition, and it can make the patient growing this tumor get really hungry and basically starve. This tumor is going to keep growing until it’s no longer compatible with the life of the person who’s growing into. Obviously, a terrible disease. RAS plays a really important role in the disease. And our goal really, is to study and understand it.
Now, after hearing about RAS, it might be natural to think, “That sounds terrifying.” But the reality is all of us have hundreds of thousands of copies of RAS in all the cells in our body. And for the most part, we’re totally fine. RAS isn’t doing anything it’s not supposed to do. Because normally RAS is mild-mannered. But RAS can transform into this form of mutant RAS that causes cancer. A lot of things can cause it to mutate. It might be ultraviolet radiation, maybe it’s [inaudible 00:09:42] radiation from our environment. It could be tobacco smoke. Sometimes alcohol can be associated with a higher risk of cancer and higher risk rates of mutation.
But once the gene picks up the spelling error, it might form a form of mutant RAS that’s going to drive the cancer, and tell the cancer to keep growing, keep growing, keep growing. A mutant RAS, it plays a common role in many forms of cancer. Almost all pancreas cancers are driven by a RAS mutation. It’s the gene that’s actively telling these cancer cells to keep growing when they shouldn’t. About a third to a half of colon cancers and lung cancers are also driven by a RAS mutation that’s telling these cancers to keep growing when they shouldn’t.
RAS has been known to be a really important player in cancer since the early ’80s, so almost four decades. And for the past four decades, outstanding scientists from around the world have been studying why does RAS behave this way, why is it active, why does it promote cancer. Pretty much any method that you can do scientifically has been applied to RAS. And just to give you one example, on the left, this thing that looks a bit like a wiring diagram is like a scientific wiring diagram, where scientists figure out what are the processes that turn RAS into this on state that promotes cancer, and what are the processes that turn it off.
In this diagram, RAS GTP, that would be the on state. And that’s because active RAS binds to another molecule called GTP. And that RAS GTP is a slightly different shape. And that slightly different shape instructs the cell to divide and proliferate. Another proteins in the cell, which here we’re just calling effector proteins, they bump into RAS, they recognize it in the on form, and they maybe change your activity to signal to the rest of the cell, “RAS is on, it’s time to grow.” If RAS is bound to GDP, that indicates it’s off state. And then if one of these effector proteins bumps into RAS, it says RAS is off and nothing happens. Because really big things can happen whether RAS is on or off, there are multiple processes that kind of together control whether or not RAS is on telling the cell to grow or if it’s off telling the cell not to grow.
Some of these are GAP proteins. They bind to active RAS, help it turn off. GAP proteins might bind to inactive RAS and turn it on. RAS might be able to turn itself off by itself. RAS can also turn itself on by itself. So all these processes have been studied. And scientists around the world, they’ll make this protein their lab, and they’ll measure all these reactions, all these arrows. They’ve quantified everything you can to quantify about it. How fast does it happen? What’s the propensity of these proteins to bind? How strongly do they bind? How important are they relative to each other? All these numbers have been found.
And that’s what these other tables around the slideshow, these are just samples of that data where a scientist studies one specific reaction. So the red arrow, the data for that is in the red box. And the purple arrows, data for those reactions are in the purple box. Someone measured all these different possible state transitions, how it happens in terms of chemical kinetics and biochemical kinetics, and they published the data. They’ve done this for wild-type RAS, the same type that’s in all the cells of our body, and also for the mutant forms of RAS that show up in a lot of cancers. Honestly, at the end of the ’90s, pretty much all this data was out there. That’s about when I started grad school.
So I was an undergrad, Cheryl mentioned I majored in mathematics, I was also a pre-med. And I was kind of unhappy with the fact that mathematics and medicine at the time didn’t seem to be very tightly connected. My math classes were very about numbers and equations, and there wasn’t very much biology. Maybe we talked about chemistry or physics, but we didn’t talk a lot about biology. And my physics and chemistry classes had a lot of math, but biology not so much. That kind of upset me, my giving up math if I’d pursue medicine.
But then towards the end of my undergraduate trajectory, I started to work in research labs. And I found there’s this growing area of biology research, which was actively trying to apply tools from that, to better understand cancer and other diseases, to understand these biological processes. And that was fascinating to me. So that’s why I applied to these dual degree MD, PhD programs so I could learn medicine, but also to pick up the research skills like apply math to biology research.
When I was looking for a research problem, RAS was really appealing. It plays a major role in human disease, we kind of know the wiring diagram, and we also know all the most important numbers for how these reactions occur. So these have just been sitting there in the scientific literature but no one had actually assembled it. So I asked the question, “Can we develop a mathematical model or a computational model where we can simulate all these reactions? Can we actually implement on a computer what we think happens in a cell, and actually do it in a manner that’s totally consistent with all the data we have and everything we think we understand?”
“And if we have this more complicated simulatable version of RAS on the computer, could we study that the same way we’d study a cancer cell or an animal model or human patient that has a RAS mutation? And if we study the computational model, can we start to infer, derive new things about RAS, that maybe we haven’t had the chance to measure yet, or maybe it would be really hard to measure if we don’t have the right tools. So can we use this as a tool for inference?” The model worked really well. Kind of like when you balance your checkbook. You, at the end of the month, you might make sure all the checks you wrote and the balance you have is the same one the bank has. And if those numbers add up, you feel good that you’ve been balancing it correctly.
And that’s kind of what happened for the RAS model. At first, we can see if we make mathematical equations for this wiring diagram, and bring in this data, when the model says, “Yeah, if you have a wild-type RAS, you should have this much in the active form. And if you have this specific mutant, you should have this much in the active form.” Those numbers lined up pretty well. So we felt good. It was a good check that everything was balancing out like a checkbook. But we don’t just want to balance our checkbook, we want to make progress. And the model started to tell us very new things we didn’t know. And some of these were controversial to the field.
So an example, and this is probably the most significant thing we discovered in the early model. The model told us that mutant RAS isn’t just bad, mutant RAS is also a bad influence. What do I mean by that? We’re showing mutant RAS here in green. So kind of like if the wild-type RAS is blue, the form that’s in our body, I kind of did a Dr. Jekyll and Mr. Hyde motif, where the blue RAS turns into the green, bad mutant RAS. We’ve been using the smoking and drinking as a sign of the bad habits of mutant RAS. And everyone is focused on mutant RAS because they knew that was the elephant in the room.
That’s new to this cancer cell, the cancer cell’s growing uncontrollably, mutant RAS is always on this on state telling the cell to divide. Everyone’s focused all their effort on there. But the computer model had told us, “Wait a second, mutant RAS is actually recruiting an army of wild-type RAS to help it do its job,” and that was unexpected. And no one had really looked about it because everyone focused on the elephant in the room. So we did some experiments and we found, sure enough, the model was telling us something correct. Mutant RAS is a bad influence. It’s telling wild-type RAS and the army of RAS proteins telling the cell to divide is much bigger than we realized.
But now, the next question is, “Well, how does mutant RAS influence non-mutant RAS?” And in the traditional experimental approach, you’d make this discovery experimentally, and then you go back to your lab and you’d think about the papers you’ve read, you’d think about your wiring diagram and your understanding, and you try to come up with new hypotheses. Then you’d systematically test those hypotheses to see what makes sense. But the nice thing about the math model is we had a totally different approach to study it. And the math model is already telling us based off what’s known, here’s this new unexpected behavior. Mutant RAS leads to wild-type RAS activation.
So we can actually go back to the model and systematically study the mathematics to figure out what are the key reactions, what are the key properties that were causing mutant RAS to activate wild-type RAS? What it told us was unexpected. I’m going to communicate it with an analogy that I think will make sense. Again, most of them wild-type RASs in all the cells of our body is not on it. It’s not telling the cell to divide. But every now and then, some of those individual proteins will turn off. They’ll do something they’re not supposed to.
And you can kind of think about this like children. I have a couple kids, they’re good kids, but sometimes they act up and they do things they shouldn’t. They stay up too late, they play computers when they should be doing homework, they might forget a homework assignment. But it’s not really a big deal. They’re good kids. So you kind of come and tell them, “Hey, don’t do that.” And good kids are redirected. Within the cell, there are proteins called GAP proteins. And their job is basically to take the RAS that turns on when it shouldn’t and turn it back off. Kind of like a parent just monitoring your kids and redirecting them.
But now these mutants that cause cancer, they can’t be turned off by the GAP proteins. Some kids are just going through a bad phase and they’re going to smoke and drink and not listen to their parents no matter what the parents say. The parent can try all they want, but that individual is not going to change their behavior at this time. And that was already known. That’s why these meetings are active, they can’t be turned off. What the mathematical model taught us is that this is what’s actually causing wild-type RAS to be active. Because if a parent had a lot of kids and one was causing a lot of trouble, the parents aren’t going to give up on that child. The parents are still going to try to help that child get back on the right track and improve their behavior.
When the parents are spending all their time focusing on the one getting out of line, but not able to go around and get the other kids back in line when they step out of line. And the parent might even think, “Hey, I have a bigger problem here. If they stay up and play too many video games tonight, it’s not the end of the world. I want to take care of this problem today.” But the end result is the wild-type RAS that normally could easily be turned off, it’s now in a higher state of activation. It’s also telling the cell, “Keep growing, proliferate, keep these cancers behaviors.” Because the protein is supposed to tell it not to is distracted by the mutant in the room.
That’s an example where we figured out that mathematical models, they can help us find out new things about RAS and how it promotes cancer. There was kind of a unexpected benefit of the models. RAS has been studied so much by so many people. When I started this project, part of me worried that it’d be impossible to find something new because it felt like every possible experiment had been done by a grad student or postdoc somewhere in the world. So it was really satisfying that we could find out new and important things about how RAS promotes cancer.
It was also satisfying because lot of people told us it was impossible that there’s no room for math and biology. And we’re starting to find out that, “No, you can apply math to biology.” But at the same time, skeptics, and it’s good to have skeptics so they keep you on your toes. The skeptic said, “Okay, that was interesting. But how are you going to make people with cancer live longer? How can you come up with better cancer outcomes?” It’s a great question, how can we use math to improve how cancer is treated? Since I’ve come to Salk, that’s really what we focused on. How can we apply this mathematical model and our ability to study cancer promoting mutations in a totally different way? We’ll use that to understand how different mutations should be treated and how we can better treat cancer patients.
So the next part of the talk is going to focus on the first major study we’ve published where we looked at colon cancer. Colon cancer and RAS have a very close relationship, because almost half of colon cancers are driven by a RAS mutation, and the other half or a little more than half do not. And even in those cancers that don’t have a RAS mutation, RAS is still playing a really important role. Because in those cancers, it’s still activated, and it can still be telling the cell to divide. And that’s because there are multiple ways RAS could be activated. It could be activated by a mutation but it can also be activated by proteins we showed in the wiring diagram called GEFS.
Coming up with a cartoon, I thought of a tobacco advertisement. That might be our schematic of a GEF. It’s something that will motivate RAS to adopt its behavior to this active state. Just to kind of show the power of science, once it was learned that RAS and colon cancer is active promoting cancer, and it’s driven by GEFS, they develop drugs that would stop that process. This drug is actually the drug that Martha Stewart famously invested in. But when you give that drug to a colon cancer patient, it shuts down the pathway that leads to the activation of RAS. RAS will shut down, the cancers proliferate slower. And patients who received this drug, they live longer.
But that drug doesn’t work on the mutant RAS. And that’s because these mutant proteins, they don’t care if the parent tells them not to smoke and they don’t need to advertise it. They’re going to have this bad behavior no matter what. So it was kind of figured out very early on that this new drug that works for these patients with no RAS mutation, it does not work if you have a RAS mutation. And that was actually extremely important in how cancer is treated nowadays, because you’ll hear topics like genomic medicine or personalized medicine or precision medicine, this is really one of the early examples of that.
Where if a colon cancer patient comes in, in the previous days, they’d all be treated very similar based off staging and how advanced the tumor was and whether or not you could operate. But now, there was a genetic reason to change how you treat the patients. If you have a RAS mutation, you do not get a drug. If you do not have the RAS mutation, you can get the drug. So doctors had to start sequencing the patient’s tumors. Is there a RAS mutation? Yes or no. And based off that, you do or do not get a treatment. And this has kind of been the model that keeps getting extended with more and more cancer drugs as they’re developed.
So that relationship was really satisfying. They figured out a pathway that led to a drug, understand the difference between muted and wild-type RAS, led to a really smart criteria for who does and doesn’t get the drug. Everything seemed really smooth. But there was a confusing exception. There was one RAS mutation that’s still active, it’s still no drinking or smoking, but for some reason, patients with that mutation seem to benefit from that drug. So it’s the one exception that made no sense. And it was really controversial because we know so much about RAS. How could this one mutation that behaves just like all the other mutations respond as if the mutation wasn’t even there? It didn’t make sense to anybody.
The evidence was really strong. It was phase three clinical trial data. That’s the gold standard of how you decide whether or not to give the patient a drug, that’s the best evidence you get. But clinicians have actually chosen to ignore that clinical trial data and say it doesn’t make sense. So they just treat those mutation patients like they have any other RAS mutation. So these patients don’t get the drug. Now, I understand part of the argument was we don’t have the mechanism, we can’t explain why they’re different. So it’s probably a statistical outlier in the data set. But we should do another study to prove it. The problem is these studies are really expensive, really time-consuming so no one’s ever actually done that next study in the past 10 years.
And honestly, the inaction on that study really bothers me for another reason. Because every year, there’s about 10,000 new colon cancer patients with this mutation, they don’t receive the drug. But some of them, they go through the standard treatments where they get radiation, they get surgery, they get chemotherapy, they get new targeted therapies, and the cancer still grows. They might still be in relatively good health but they know this cancer is going to keep growing and eventually will kill them, unfortunately. And they’ll say to their physician, “There has to be something else we can try.”
And they might go to a tumor board. There’s these new cutting edge tumor boards. There’s one at UCSD that I participate in where patient’s situation, they get their tumor sequenced, we figure out what mutations are driving their cancer and we think, “Are there any new options? Can we try something clever? Can we take an existing drug that’s approved for a different type of cancer or a different disease, and maybe it’ll benefit this patient based off what looks to be the driver of their cancer?” And even in those situations, people are so sure this exception can’t be true, they will ignore the phase three clinical trial data.
Again, it bothers me so I open my lab, I thought, “Well, maybe the math can explain this. If there really is a mechanism, maybe we can figure it out, and maybe that’ll lead to these 10,000 patients receiving this drug that’s ready to go for them. We were able to figure this out. So we’re proud of this study, it was published just before COVID. What we did is we took all the data there was for this unusual mutant that response to drug and the other mutants that don’t. When we ran it through our simulations, what we found is this mutant is different because it’s a loner. So where we said the other mutants are bad influences, where they motivate the wild-type non-mutant RAS to also adopt this bad behavior and promote cancer.
And not only that, they also kind of take the attention away from the parents so the parent can’t motivate the normally well-behaved RAS from acting up. This mutant, it’s still causing trouble. It’s still drinking and smoking but it’s going out into the woods where no one sees him or it’s going on behind the school so no one knows what’s … So it doesn’t influence the other kids into doing something bad, it doesn’t distract the parent. Sure, it’s still active. If you look at it individually, it’s just as bad as the other mutants. But if you look at how it’s affecting community, it has a very different effect.
This turns out to be really important in colon cancer. Because we mentioned before that a lot of colon cancer, more than half, have this property of having the wild-type RAS activated by these other factors, these GEF proteins, which we show like the advertisement. And then these other cancers that don’t have RAS mutation, there are so many of these other forces trying to activate RAS that overwhelms a parent’s ability to tell it not to. And if you don’t have the mutant RAS in those cancers, you shut that down, you help the cancer.
What we found in our studies is that it’s this exact same reason why patients with this rare mutation benefit, because since the loners are not a bad influence, when you give that other drug, it shuts down wild-type RAS. So our prediction based off the math was when you give this drug, you can shut down wild type RAS and the cancers of this exception, but you won’t be able to shut down wild-type RAS and the other cancers. So it’s a very specific testable prediction. It followed from the map, it followed from the data we had about how these mutants behave and what they do. But no one ever seen this or thought of this before. And we’ve now performed the experiments and we’ve confirmed this mechanism. So we published this just before COVID.
Actually, as [inaudible 00:26:43] was heading, I was traveling around giving talks at conferences trying to get cancer physicians to say, “Hey, if you have a G13D patient, consider giving this drug. You give it to your wild-type patients, it has about the same benefit. If you’d give it to your patients without a RAS mutation, try it for these patients. Or at least if they’ve advanced on treatment, you’re looking for something else, try it out.” What was really satisfying is the last conference I went to before COVID hit, an oncologist from a really prestigious East Coast Cancer Center came up to me, and she wanted to find me because in the previous three weeks, she had three different patients referred to her from the community who had colon cancer with the same mutation, KRAS G13D. And they were coming in with our papers saying, “Hey, can we get this drug?”
She was trying to figure out, “Does it make sense or not?” She found me after my talk and she wanted to go through it in detail to see, “Should I give the patient this drug or not?” We were all really excited, “Can we make a big difference on this?” COVID slowed us down but as we get back to the conference circuit, I think that’s going to be our real goal. Is can we motivate clinicians to try this out? Anyway, just to kind of big picture summary. By using these mathematical simulations of RAS regulation, we can reveal important new aspects of RAS biology. Experimental biology is still critical. That’s where our data comes from, that’s how we test our mechanisms. But now we have a new way to study the data, a new way to make new influences.
And one of the things we figured out was that some mutants are bad influences, some are loners, and figuring out which are which can have really important implications for cancer treatment. So actually, in our newest study, which we’re hoping finish up before the end of the year, we wondered could there be other loners. And if there are other loners, would patients with those loners respond to drugs? Actually, to date, we found about 10 more and we think eventhough these are all rarer than the one that’s already known, in total, it adds about another 10,000 patients a year. So we’re going to keep looking for more and more of these exceptional responders based off the fact that we cannot do experiments to see if you’re a loner or not loner.
We know what to look for. So now we have the mechanism, we think we can find more patients will benefit. And that’s really our goal is can we continue to improve cancer outcomes by finding patients and matching them to drugs better? Now, whenever I put up together a talk and I kind of go in comic sans serif font and it’s kind of cheesy, I feel like, “Am I not giving you guys enough science?” But we’ve actually covered a lot of science so we got to review what we’ve covered. Today, we talked about RAS, one of the most important players in cancer. And it exists in two forms, wild-type and mutant. Both forms, wild-type and mutant, can exist themselves in two different states, an inactive state and an active state. It’s this inactive state that is kind of benign doesn’t do much, the active state promotes growth. The RAS mutant is usually in the active state and the wild type RAS is usually in the inactive state.
Because active RAS can have so many bad effects, the cell has ways to kind of control its activation carefully and has also to turn it off. So the way wild-type RAS is usually activated are proteins called GEFS. So they can convert it from inactive to active, drugs that shut down this effect. Say the Ghostbuster type symbol up above, that’s the drug that Martha Stewart invested. That works for a lot of colon cancer patients. It helps shut down this activity, helps shut down RAS activation. GAPS are the proteins that turn RAS off.
That was all before we started to do our work. Where our work where the math helped us is figuring out how these mutations interact with these processes. So a lot of RAS mutations are bad influences. Not only are active but they also distract the GAPS, which leads to the activation of wild-type RAS. But then some are loners and they’re active by themselves, but they’re not interacting with other systems, which lets some of those cancers be more targetable. To finish up with some acknowledgments, the experimental work proven these loners is the mechanism is done by Tom McFall. In addition to being a great scientist, is also a great mountain biker.
The rest of the current members of the lab are shown, the papers that we have on this were done with a lot of collaborators. Some within Salk, some outside of Salk who’ve all given us different help on lots of different aspects. And we’re also really grateful for a variety of funding sources, both federal, foundational, and individuals that’s helped us, support our RAS program. And really now that we’ve been able to prove we can do this. We’re really just trying to do this for more genes, for more mutations, for more cancers, for more drugs. And we’ve been fortunate enough to get some support to help us try to expand our research program. That concludes the talk. I appreciate all of your attention. Thanks for the opportunity to talk to you all today. I’ll be happy to take questions.
Ed, thank you very much for an entertaining and quite educational and inspirational talk. I had not followed your latest because you do so many things in a lot of spheres, not just working on this colon cancer. But to share a little bit, a personal story, my younger brother in his 40s was diagnosed with late stage colon cancer a few years ago. He’s okay but he went through the horrible treatments that you’ve described. And I am so happy to know that there are going to be more people using something that’s more targeted, that is impactful and effective. Your work is going to save many lives. So thank you, personally.
This is just so fantastic to be able to share with our audience. Although Salk is known for our basic biomedical research, here’s young MD, PhD, Ed Stites showing you that it can really be transformational before long. Sometimes things take decades and decades and decades. And obviously, your work, as you mentioned, is based on work that many people had done for years before you started it. But having somebody with bright new insight and new tools to use, thank you.
I hope everybody else was as impressed with that as I was. But we do have some questions that I will get to. Let’s see, where to start? Sandy asks, “Is there a correlation between genetic mutation in cancer and genetics?”
Like personal germline genetics? Yeah, absolutely. Especially for colon cancer, there are some very well-known familial sources of colon cancer where you can develop colon cancer at a very early age. A lot of those are due to germline mutations that will cause you to have a higher mutation rate. So that’s one of the mechanism that you can get more cancer. There are other germline predispositions that have a very specific mutation that might be … One of the handful of mutations that can add to the total effect of cancer.
Usually, cancer requires about five mutations. Five to seven mutations is what they call the driver mutations that cause the disease. But if you’re born with one of them, you already have a headstart. So some of these are known to run in families. Then sometimes you just have a predispositioning factor that will just make you acquire mutations at a faster rate. You don’t correct the errors in your DNA as fast as others. So those are really important correlations.
This question comes from Joanne, “What role if, any, does epigenetics play?”
That’s another really important factor. I’ll give the simple answer and then a slightly more complicated. The simple answer is if there are say, five or six key steps that a cell must acquire to become cancerous, some of those could be caused by an epigenetic change. So that’s basically a modification to the DNA that might influence expression, but it might not be a spelling error in the DNA itself. It’s another type of mutation that’s going to affect how the DNA is interpreted or how its hold or how it’s processed.
That said, the epigenetic changes can be huge, because you have really global effects on gene expression. So I think the traditional answer in cancer, which is it takes five or seven events to become cancerous. I think that might be a big underestimate. So when you look at cancer genomes, there are a lot of mutations. I think there might be five or seven, what we call rate limiting steps, things that take a long time to acquire. But once you’ve acquired those steps, they’re actually probably a lot more other mutations that contribute to this cancerous effect. Epigenetics is one of those rate limiting steps that once you’ve dysregulated your epigenetics, you can rapidly pick up more mutations and influence gene expression.
Great. Rifka asks, “How do you measure the effect of inflammation on cancer development?”
Great question. So for us, we’re not modeling inflammation individually, we’re really just focusing on those effects. But inflammation can have a really big effect. They can affect the immune status of the cancer and the ability of the immune system to recognize a growing cancer and eliminate it or not. It can also have influences on whether or not these mutations are being picked up. It can have influences on the blood supply and the nutritional status of the cancer. So it’s definitely there. We’ve been focusing really only on the mathematical aspects that we can control well. We can make pretty good inferences from that but we’re not including it specifically in our models.
Again, I definitely want to let people know that while Ed is an MD, PhD, and just a brilliant all around person, he doesn’t have patients of his own to my knowledge right now, and can’t answer some of these specific questions, but he’s obviously doing pretty well so far. So James asks, “Is there any progress on using blood tests or similar means to detect the presence of cancer? And how about determining which cancers will grow rapidly or not?”
Absolutely. It’s a really tricky issue to do testing. Actually part of the COVID pandemic, you might remember, if someone got a COVID test when they’re likely to be sick, that test was really useful. Or if you pick up a COVID test when you’re not symptomatic, you might have a higher chance of a false positive. And that’s really the trick with cancer screening is if somebody has no symptoms, you pick up a lot of false positives. And this is true with sigmoid colonoscopies or colon cancer, it can be true for mammograms. You can get a lot of false positives. It doesn’t mean you shouldn’t do it if it’s recommended. That’s why they have these aging criteria, do it when you’re over 40 if you have a family history. Do it when you’re over 50 otherwise so you can kind of have more useful testing.
But what’s really changing cancer screening, which could be would be a lot of these blood genomics tests. You can measure the DNA from a cancer in the blood. Sometimes this DNA is freed up. It might even be a circulating tumor cell or it might just be DNA that’s kind of freed up from a cancer cell that was dying and now it’s in the bloodstream. I’m going to build a tech one of these KRAS mutations or other mutations in the bloodstream. It’s still a work in progress figuring out the right way to do this. I believe it’s more useful nowadays for following the cancer that’s already been detected to see how it’s responding to treatment and to see if it’s developing resistance to treatment, then for the noble detection. But I think it’s definitely on the rise. And could we find these blood tests, these blood biomarkers that will tell us what cancer is developing?
Great. Sherry asks, “Is your mathematical model based on theories of stochastic processes?”
Very great detailed question. Our model is actually … Because there’s so many RAS proteins, like we said, about 200,000, there’s definitely some stochasticity in the cell. That’s another word for stochasticity, for everyone else, it’s randomness. So these proteins are kind of just diffusing around the cell, bumping into each other or not. And people can make various stochastic models that account for this randomness. And we’ve done that for some problems. But there’s so many RAS proteins, we can kind of just look at the average behavior. So our models are actually deterministic. And the advantage of that is I can run it on this laptop here, it takes 20, 30 seconds to run a simulation and evaluate what’s happening in a specific state. Where if you start to do these bigger models that include the spatial and stochasticity, those could take much, much, much longer to run. So it’s a bit of a trade-off. And for these problems, we don’t think we need stochasticity at this time.
Okay. Well, that answer was a great segue to the next question. Marian asks, “Has the software you use in your cancer research been able to keep up with the computational power required to advance your research?”
Great question. I’d say both are really taking off. There’s a lot of software that’s really powerful for the type of modeling we do and for other types of cancer research. Computing power is also taken off. Because of my math background, though, I really prefer models that are simple enough where I can look at the equations and kind of understand them. I like them to be more complicated than we could think about without a computer, but I want them to be so complicated that I can look at all the equations and not understand what’s happening. I kind of like the sweet spot of models that are complicated. They go beyond what a really smart biologist could do in their head, but they’re not so complicated that we’ll never understand what’s going on. It’s really great for us to then go to lab to test it.
Honestly, I do notice when I buy a new laptop, that something that would have taken a few days might only take a day and a half and you notice things speeding up. But main time, I’m running things that take several days if I’m finishing up a project and making sure that these results will happen for a wide, wide, wide variety of conditions. If I’m just doing something exploratory, it might be like 0.2 seconds instead of 0.5 seconds for a quick [inaudible 00:40:26] or a handful of seconds versus 10 seconds versus 15 seconds. So it’s what has been limiting for the most part.
Great. Because you figured out how to ask the Goldilocks questions, not too hard and not too easy.
I’m a simple boy from Kentucky so I like it to be tractable enough.
Great. Well, we have a couple different people asking about breast cancer. Has RAS been studied in breast cancer? Can you talk to that at all?
Sure. Absolutely. Breast cancer is really interesting because it rarely has a RAS mutation. I think about among all breast cancers, maybe you’re talking one to 3% have a RAS mutation, which is remarkable, because the common statistic is that almost 25 to 30% of all cancers have a RAS mutation. I think that number is too high, largely because breast cancer, which is so common, does not have RAS mutations. That said, it doesn’t mean RAS isn’t important in breast cancer. Other genes, like the GAPS that turn those off, some of them are mutated in breast cancer. And when those are deleted, we basically to have a parent to tell the RAS to turn off so that causes high RAS activity.
The HER2 protein, which is targeted … One of these breast cancer drugs, Herceptin, that pathway can activate RAS. So RAS activation is still important to a lot of breast cancers. It’s just not doing it with a RAS mutation. And that’s one of these questions we’re really interested in. Is if that RAS activation is important to breast cancer and there’s other ways to activate it, why do they rarely get a RAS mutation? We have some ideas but we haven’t figured that one out yet.
Okay. Paul wants to know, he says, “Thank you for a great talk.” So I thought I should pass that along. There have been quite a few other thank you upstandings, I’m not going to tell you all of them because otherwise we won’t get to the rest of the questions. Paul wanted to know which hospitals do you participate in for your study and is that applicable to your work?
Great question. I’m kind of in this magic sweet spot where I don’t have any hospital affiliation at the moment. I hung up my pager when I was done with residency. It’s been very liberating and freeing. So I’m kind of in this lucky sweet spot where I will go over to UCSD and I know the conditions there, I go to some of the tumor boards but I’m not practicing medicine. So I’m not making any patient care decisions but I get to kind of go behind the curtain and find out what’s going on with their clinical trials, with their genomic medicine so I can stay current and I can talk to people who are shaping the field. I can have their ear and they can have my ear, but I’m not making any treatment decisions. That’s been really great so I can focus on research.
As far as the clinical trials, they’ve been really broad. For colon cancer, they’ve been pretty broad. I’ve talked to some physicians and other institutions besides UCSD about possibly testing some of these. Again, it was all before COVID. Now that even cancer care centers have been really devastated by COVID so many precautions had slowed down research. So now that we’re opening back up, I’m going to re-initiate those conversations and see if we can move forward. But I hate to put anybody on the spot by naming them specifically.
Fair enough. I just wanted to ask question and yup. All right. Paul actually has another question, and he’d like to know whether there are already studies on the side effects of the drugs targeting the loners.
Yes. I believe the side effect profile is fairly mild. I believe you can get a lot of rashes with this drug. But this drug is already used probably on colon cancer patients. So I’m guessing about 100,000 people a year are eligible for this drug. They all don’t receive it. It has a modest benefit like a lot of colon cancer drugs. So every physician decides what cocktail they want to treat their patients with. Well, there must be tens of thousands of patients a year receiving this drug for colon cancer, additional receive it for head and neck cancer. And the side effect profile is believed to generally be pretty manageable.
I do believe the patients have to go in for an infusion. They can’t take it orally. That can always cause some pain, discomfort and you can get rashes. But I’m not aware of any debilitating side effects. Looking at some of the chemo therapies that are effectively poisons that can have blood and bone marrow toxicity. I think it’s a much more manageable side effect profile.
Well, that’s promising to hear as well. Madison would like to know, she says, “You mentioned how mutated RAS can affect other RAS proteins in a negative way. Are there any other proteins or pathways that are often mutated when this form of RAS is active?”
Great question. This is one thing that’s really interesting is what we figured out was that since the loner can’t activate wild type RAS, it’s like something else wants to be causing wild type RAS to be activated. And we think it’s this whole pathway. It’s the EGFR pathway for those of you who follow biology. That leads to activation of this GEFS. But when you look at the cell lines in the lab, so in the lab, we can grow cancer cells basically in a dish. And all the ones that grow in the lab, they almost always have a gap mutation. So the ones that grow in a dish, they are also picking up the second mutation that basically takes away the parent. In the colon cancer, in the patient, we don’t see that mutation. They seem to be driving it by the GEFS, by that advertisement, but in a dish. And I think it’s because in a dish, we lose that exogenous, that microenvironment stimulation that activated the GEF.
So I think when you try to grow a cancer cell in a dish, you only select for those that have also lost the parent mutation. So there’s a big difference between what we see in the clinic and what we see in the cell cultures. And that’s honestly creating, I wouldn’t say a controversy but there’s another group that’s studying this problem. They focused on more what they’ve noticed in a dish and we’re focusing more on what you see in the patient. So I think we’re working that out. We have another paper where we hope we resolve the controversy. But there are some other co-occurring mutations.
Great. Rachel wants to know, “Would you please comment on how the drug works in the loner situation? What’s the mechanism of action?”
Sure. Basically, this drug, it shuts … On the surface of the cell, there’s a receptor that sticks out, EGFR. Normally, when that’s activated, it will lead to the activation of those GEF proteins. And those are the proteins that turn on wild-type RAS. So if you have a cell with no RAS mutation, you shut down this, you shut down the cascade, you shut down the activation of RAS. So that part was what people already knew.
People already assumed that if you treated this drug, you couldn’t turn off the mutant because the mutant, it’s always going to be bad no matter what you do. And that’s true for both the loner and the bad influence. What the difference is that if you hit this drug in a loner, you can turn off the wild-type RAS because the loner can’t turn on wild-type RAS. So all the wild-type RAS in that cell is only getting activated by EGFR. But if you take a cell that has a bad influence, you can shut down the GEF. You can take away the advertising but the wild-type RAS will still be activated by the bad influence mutant RAS. So it doesn’t matter if you get rid of the advertising or not. That charismatic leader is going to rally the rest of the wild-type RAS to be doing something bad.
I don’t think I mentioned this. But just for scale, most of the RAS in cell should be wild-type and not mutant. Because there’s three RAS genes, KRAS, NRAS, HRAS. They’re very, very similar. They just have very small differences. So we consider them all the same, we consider them RAS. But we also have two copies of each, we have one from mom and one from dad. So on our DNA, we have six different pieces of DNA that spell out a RAS protein. And only one of the six in general will be mutated so this mutant form that causes cancer and the other five are wild-type.
So most of the RAS genes in the cancer cell are still normal. Most of the RAS proteins in the cancer cell are still normal. So it’s whether or not this mutant can rally those wild-type RAS proteins to also misbehave or not is really the key. And if it’s wrong and it misbehave, the other processes won’t have a big effect. But if it can’t rally them, then these other processes will be in effect where you can target these proteins.
Thank you. Laura Anne now wants to know, “How does the CHEK2 mutation interact with or affect the RAS protein proliferation especially in pancreatic cancer?” Now, that’s a little bit of a specific question but …
That’s going to be way too specific for me. I’ve been kept up with CHEK2. And honestly, because of my … I’ll also admit, I have a lot of tunnel vision for RAS and for all the nuance of RAS so I looked at a really highly focused picture of RAS. And the further away we get, the more my knowledge quickly fades away.
Understood. I tell people you can’t answer absolutely every single question out there, but we’ll test you. You’ve done well so far. Here’s a simpler one. Excuse me. Someone asks, “Does RAS have a significant impact on prostate cancers?”
So that’s another cancer, like breast cancer, that’s really common, and that rarely has a RAS mutation. Again, it’s that statistic that a third of all cancers have a RAS mutation. Actually, I have a paper we’re about to send back where we recalculated it, we think it’s closer to 10% than 30%. Also, indirectly, we’re making the burden of RAS cancers lower by doing better math. But the reality is prostate cancer is luckily not been heavily driven by RAS mutations.
Great. Here’s an interesting one. Again, I don’t know if this is getting a little bit too far out of your field. But does any of your work look comparatively between animal and human cancers? Erin is the person asking this and says, “My dad and dog have bladder cancer right now and I am learning how much research is done correlating animal and human illnesses.” Erin, thank you very much for your time.
Thank you, Erin. It’s a great question. There’s a lot of work done in this. Honestly, to a degree, we’ve almost focused on a slightly different approach, which is animal models are extremely important to cancer because the cancer disease is similar between dogs and humans, between mice and humans. There are differences but there are also a lot of similarities so we learn a lot by studying cancer in a lot of different animals. But the same time, we want to minimize the amount of research that’s done on animals. And that’s one possible benefit of the math models is if we can come up with better hypotheses, if we can come up with an alternative system, maybe that’s a way we can reduce some of the experimental research and just learn more from animals.
Although I do think for the point of a specific cancer, I believe a lot of the veterinary cancer drugs are related to the human drugs. I believe I’ve seen at least for some of them, where it’s similar drug, different label. I believe that’s true for some of these. Because a lot of these proteins, a lot of the … If you look at the letters that spelled these genes out, there are definitely differences between humans and dogs and humans and mice, but they can be very similar. So I could perhaps take a human cancer cell, and if I express a mouse RAS protein in there, it’ll behave very similar. So a lot of the disease is very similar.
Thank you. That’s fascinating. Let’s see. We have two more questions for you. We’ve got a little bit of time. Okay. Paul then asks, “Is there any diagnostic way to decide whether a cancer is being launched via RAS, more specifically via loners? Your work is obviously mathematical models. But is there a diagnostic way to tell that?”
Absolutely. So say if a colon cancer patient is diagnosed and they take that biopsy, when they send it for sequencing of RAS, they’ll actually tell us the exact spelling error in RAS. So the most common spelling errors in RAS are actually … So proteins, they have another alphabet, the amino acids, slightly different letters, but the 12th letter’s usually a letter G, a glycine and the 13th letter also happens to be glycine G. And the most common mutations in cancer occurred either number 12 or number 13.
The G becomes a D or G becomes a B. Those are the two most common in colon cancer. If the G becomes a D at the 13th spot of the protein, that’s G13D, that’s the first loaner we discovered. We’re discovering other loners. So we can even say like, “Q61K, that looks to be a loner.” Or, “G12C looks to be alone.” Or, “G12F does not look to be a loner.” So we know these. When these pathology reports come out, we can actually compare that to our list and say, “These should be loners, you should be wild types.” Even though our model is a bit abstract or mathematical construct, we’re mapping it directly back to these specific mutations that are known, because we’re using the exact numbers that had been measured for those specific mutations. So the traditional sequencing relays exactly back to what we’re doing.
Wow. That I’m sure must give you some phenomenal data to work with after that.
There’s no shortage of data. That’s nice.
Okay. Well, the last question. I saved this one for last because … It comes from Lance and I thought that this is something that is very broad and has lots of applications, potential for the future, “In your opinion, will we ever get to the point where cancer treatment will not include devastating surgery, chemo, and radiation?” No pressure.
Great question. I’ll say this, I really enjoyed my medical school rotations on cancer surgery. Because some surgeries are definitely brutal. But on the other side, the ability to actually physically remove a cancer from a patient’s body is fantastic. So it really can be curative. But then sometimes the cancer spread really far and then it’s a game of catch up. Where you can move a chunk but you can’t move it all. And that can be really tough. Some of these cancers like pancreas cancer or glioblastoma, you can do a really complex surgery, but you might not be able get all the cancer, it still comes back and that’s devastating, too. I just want to point this out because there are some cancers where surgery is very safe, not very painful, the long term benefit is huge. There are other cancers where it is a much bigger ordeal.
So I just want to make sure I make that distinction. So someone doesn’t decide not to treat a very surgically resectable cancer because they know of another cancer where it’s really bad. I mean, I think chemotherapy, which really can be a poison can have bad side effects. It does seem like because of all these new treatments, the new immunotherapies especially, you’re having more and more cancers that used to be a death sentence, and all of a sudden are now cured or they’re trending much more towards curative. I think that’s the goal. The field that over time, more and more cancers that were uncurable become curable. More and more cancers that had a really short window of survival, that window keeps getting extended, extended so it becomes like a chronic disease.
I remember, when I was a young child, or my mom was a child, a heart attack was … You had a heart attack and you didn’t recover. And now people can have a heart attack and recover and go back and live a really robust life without any slowdown. I know when my grandparents had their heart attacks, it was a very different experience. They could survive but their ability to get around was much different than it would be if they had that same heart attack today. I think that’s the hope for cancer that as these treatments keep getting better and better, cancer becomes less of a … The prognosis just becomes extended and extended and extended where we have other ways to hit it when it comes back.
Thanks to all these different approaches, the new drugs, the new immunotherapies, the new targeted therapies, like the ones that work on the RAS pathway, as well as the conventional chemotherapy and surgery and radiation that we’ll be able to construct these cocktails that are less and less toxic to the patient. That the patient can enjoy their quality of life while keeping their cancer, if not cured, at least keep it at bay. It’s hard to say when that’s going to happen because cancer is so many different diseases. It’s not really fair to say cancer, even colon cancer, we can say RAS mutant colon cancer, RAS wild-type colon cancer. We can have microsatellite stable or microsatellite unstable. It’s amino cold or amino hot. There’s so many different ways to classify these.
Every cancer is like a snowflake. It’s a different collection of mutations. It really is a different disease in itself. But I think we’ll keep chipping away and some of it better and better. Testicular cancer used to be a death sentence and now it’s very high curates. Pediatric leukemia used to be, again, horrible outcomes. Now, pediatric leukemia outcomes are fantastic. I think we’re to see more and more cancers that used to be terrible, suddenly become not so terrible. And some that were not so terrible become cured and long term survivors.
Absolutely. That’s super hopeful message. The war on cancer was declared by President Nixon in the ’70s, I believe. And that was when it was just cancer, singular, to cover all of this. And now we know so much more in the decades since then. You’re right to point out, a ton of progress has been made, including by yourself personally. Thank you very much for all your work that you have done, and that you and your collaborators continue to do at Salk and around the world. Really appreciate it. I hope everybody found their past hour to be worthwhile. Thank you very much for joining us. We very much look forward to any feedback that you might have and any topics that you would like to hear in future sessions of our Power of Science webinars. Thank you for coming. Thank you, Ed. Really appreciate all of your time and energy. Bye-bye.
Thank you. Bye.