RAS Chat: An Interview with Kevan Shokat and Ziyang Zhang

June 26, 2023, by Megan Rigby

Megan: We’re really excited to talk to both of you today, Ziyang and Kevan, thank you so much for agreeing to chat with us. So, my first question is for Ziyang. You're a young, really talented researcher—you did your bachelor’s at Peking University and then your PhD at Harvard where you worked on synthesizing new macrolide antibiotics. What attracted you to Dr. Shokat’s lab to do a postdoc?

Ziyang: That's an excellent question, one I ask myself sometimes. When I started doing research, I knew I wanted to do chemical synthesis because I just liked the idea of making molecules, especially making molecules that may or may not have existed in nature at all. Then I kept asking myself this question of why do I make these molecules? Eventually I realized that I really just want to make things that can affect our health in a positive way. Treating diseases is one obvious way to do that, so that's actually why I decided to do a PhD with Andy at Harvard where I made macrolide antibiotics, because I thought it was a real problem, and I wanted to do something that would make a real impact. So, you can see how I set my trajectory to Kevan already. Moving to the next step in my career, I really wanted to see how I could further use chemical reactivity to guide our design approach to the treatment of cancer, and part of me was also interested in the immune system. Immune oncology was just reviving at the time when I started my postdoc—well, it actually had been reviving for some time, but it was new to me—and I also wanted to see if there was a way to use chemistry to help that aspect, and that drew me to Kevan’s lab. That’s a long answer to your question.

Tommy: That’s alright, we like long answers. What are some lessons you learned in your graduate and postdoc work?

Ziyang: I think there are two things: one is always be optimistic. When something doesn’t work, go back to it. You don’t have to go back to it all at once, but go back someday. I can tell you that most of my projects in Kevan’s lab worked this way. It’s sort of a tradition in Kevan’s lab that many projects work when you’re ready to move on to the next stage of your career. And the second thing is always be science-driven. I get most of my energy and drive from thinking about science. That’s what gets me up early in the morning, and gives me the energy to keep doing experiments, thinking of new ideas, etc.

Megan: Were you always curious about science, even as a kid?

Ziyang: I mean, I’ve always been a nerd. But I think a lot of it has to do with the environment—both Kevan and Andy’s lab are populated with great people who are also excited about science, and they want to help each other. I think this is the really nice feature of chemical biology/chemistry in general. You can have your own specialty, but we teach each other how to do things. I think the same can be said about the RAS field. We learn from RAS biology how each of the mutants have their own personalities, and their own vulnerabilities. 

Kevan: That’s a good point, I feel like RAS has been such a big challenge that there was a much more collaborative effort across labs—even people who didn’t collaborate were working together on it—because it was just such a huge challenge. I think that’s pretty amazing. 

Megan: Kevan, what made you throw your hat in the ring to target RAS initially back when it was deemed undruggable?

Kevan: It was Frank McCormick back when I first met him in 1999 when I was getting recruited to UCSF. He told me back then that RAS is really the most important oncogene, that we don't have a way to treat it, and to “come here and work on it with us”. I was working on protein kinases at Princeton at the time, and when I got here, I really didn't have a great idea about how to work on RAS. But Frank would bring people in the RAS field like Fred Wittinghoefer and Roger Goody to give talks at the Cancer Center. It was around 30-40 people, a lot of swearing, and some great science and biochemistry. It was a lot of looking at the problem, seeing opportunities, hitting challenges, then—no success. It felt like this big wheel-of-fortune—you find something, you take advantage of it, and then you go back to the beginning. I had worked on PI3Kinase because it was an important oncogene, and then we worked on mTOR, but the reason I moved to RAS was because in about the mid 2000s, the RAF kinase inhibitors were tested in the clinic. Because Raf is right downstream of Ras, I kind of thought “oh you know what, that's gonna take care of the RAS that Frank kept telling me about. You still can't drug it, but this should take care of it.” But then there was this complete bolt of surprise when the RAF inhibitors would actually make Ras tumors grow faster. So, I thought “holy crap, this is totally not working, this is terrible.” So, we worked on that for a bit and helped figure out what was happening there. Then we said “OK, we’ve got to work on RAS now,” and then that was kind of the launching point and around that time. Also, people began to be more tolerant of covalent drugs around that time, and Jim Wells moved to UCSF, and before that, he had the special screening technology that we could use…so a lot of things just coalesced. But overall, the push came from Frank. 

Megan: I wanted to talk about your work targeting the less common mutations of G12R and G12S. In your G12S paper, you were inspired by a family of natural products with a strained beta lactone that you ended up using as your electrophile. How did you find this family of natural products and can you speak more on your interest in using natural products as building blocks? 

Ziyang: Yeah, I can I take this one. I’ll start by saying that people tend to think of these two as rare mutations, but thinking about the base population of patients that have KRAS mutations, these mutations actually happen in a lot of patients, so we took it as a serious problem. Going back to the question of where do we run into these natural products, I think part of it was because I worked on antibiotics. There is a family of antibiotics called beta lactams that has a four membered ring structure. The funny thing is that I worked on the synthesis of beta lactones when I was an undergrad at Peking University, back when I was still learning reactivity. I went back to my group meeting slides and I was like “wait a minute, this is basically a serine-reactive electrophile.” It's got the right strain, it's got the right reactivity and the really amazing thing is there is already an approved drug that has the beta lactone in it, so we thought “that has many boxes checked, we should just try making some.” And the rest of the story was, we made a lot of them, and none of them worked. And then we kind of took a hiatus, especially during the pandemic, but then right after that we started looking into old projects—things that we tried that didn't work. I looked back at it and thought “there's no reason this doesn’t work.” So, I made two more compounds and one of them reacted with G12S and that got it all started. Every single component we made moved us a little bit forward, and eventually we got a crystal structure where we saw the actual ester bonding. So that was the eureka moment when we knew that this was going to work.

Megan: That’s when you break out the lab champagne.

Kevan: Right. We worked on it and then just nothing worked, and then and all of a sudden, a slide shows up with that data on it and I was like “oh wow, what on earth?” That was when I was catapulted into the future, into what’s possible. It was very cool. And the two molecules were so similar, and one showed reactivity and the other one not at all, so it just taught us so much about these less reactive amino acids—that everything is on a knife’s edge.

Megan: With your G12S compound, did you see any cross-reactivity with HRAS G12S and would you be interested in modifying that compound to look at therapies for Costello syndrome?

Ziyang: Wow, you should advertise for my lab…definitely, that’s one of the reasons we wanted to target G12S. I do want to say that with the compounds that we made, part of the structure encodes specificity for KRAS and we know that HRAS doesn’t have a histidine at amino acid 95, so that will eliminate a lot of the binding affinity, so we don’t think that the compounds we have now will have any affinity, but like you said, there’s no reason not to target HRAS G12S, and we should be able to figure it out

Megan: That would be really helpful for a lot of people

Kevan: I know, it would be so perfect, because you could get the allele that’s in every cell and not the wildtype…it would be so exciting

Megan: Definitely. Your G12R paper also uses some clever chemistry, but it’s limited by its affinity for the GDP-bound protein, which is a problem because G12R exists almost exclusively in the GTP-bound state. With drugs that target KRAS in the GTP-bound state, are you optimistic that you could modify your existing compounds to effectively target GTP-bound KRAS G12R?

Ziyang: I think so. We realize our G12R work has a lot of limitations, and a big part of it is because of the biology of that particular mutant. We think the chemistry will be portable once there are more scaffolds out there, so yes.

Megan: Yeah, the work that you’re doing is just so important. Just providing a starting block for industry to jump off of—just proving the concept—is a huge step in the right direction.

Kevan: Yeah, and the one thing that we’re starting to appreciate about the GDP and the GTP state is that it seems like when you move from cysteine to serine, you’re still stuck with the GDP state binders, and those two amino acids are basically isosteric, they have the same kind of methylene and then the heteroatom, sulfur or oxygen, but when you get to aspartate or arginine, you’re farther away, so I keep hoping that that difference will allow us some more plasticity of the GTP state. G12R is yet its own special beast, but it’s just amazing, and you at the RAS initiative do so much with each different allele. That’s what’s been so special over the last 10 years is just seeing that. Ziyang said it, too, there are just so many patients with each specific mutation. We call it “rare,” but in the RAS world that’s still more common than all these kinase inhibitors that have been targeted to fusions and things. So, it’s just amazing the difference in biochemistry and the signaling. The RAS Initiative is just helping us, and everyone really dive into the rich biology of these differences.

Megan: I’m just excited to be a part of it. There’s so much momentum in the field now. Are you surprised at the current pace of research from where it was ten years ago—is it beyond your wildest dreams from that time?

Kevan: Yeah, I’m just blown away. I mean, I tell the lab all the time you never know what your work will lead to. About 12 months ago, from some data that Ziyang generated and some papers we saw coming out, I was just teleported 10 years ahead to my most optimistic point. I thought “it would be so great if in 10 years we could be finding these compounds,” because I didn't know if it was possible. I didn't know if we could get the affinity, I didn’t know if we could get the reactions. I thought, “if it's possible, it probably will start to be clear in like 10 years,” and it was one year from that point that it just started to all come together.

Megan: What are you most excited about for the future of RAS research?

Kevan: Covalent G12D inhibitors for pancreatic cancer. I mean that's what we've been after from the beginning. It's just the prototypical RAS disease. Almost everyone knows someone who’s died from it, and it’s such a horrible cancer. Maybe a G12D inhibitor won't be a cure, but it will be the foundation because everything else we tried has just not moved the needle. 

Tommy: I have a question for you, Ziyang. As you establish your own research efforts as an independent investigator—what is the direction that you want to take things? What is your vision?

Ziyang: I think there’s still a lot of room for chemistry there. A big part of my project is expanding the serine- and arginine-targeting chemistry to other cancer targets, like EGFR cysteine-to-serine mutation, or the BTK cysteine 481S mutation. Those mutations tend to emerge under the selective pressure of a covalent therapy. I think we should be prepared for this type of mutation because they are already observed in the clinic.

Megan: Kevan, your own companies and partnerships with companies is what helped push the G12C inhibitors to the clinic and now to FDA approval. Do you think there’s enough momentum in the field now that drug companies will take on these less common mutants like G12S and G12R?

Kevan: Yes, I already know of that happening. It was interesting, when we were submitting the paper, one of the reviewers said “this is not a common mutant, this is not medically important.” Not medically important…that was just tragic. But I got an email in the review process from a physician in Israel that said “I know you work on G12C and we have those drugs, but a cousin of mine has G12S. Do you think there’s any way to translate what we know about G12C to G12S?” This was before the paper was even out on bioarchive. Then later on, a person I knew at one of the cancer institutes was diagnosed with cancer, and they have the G12S mutation, and reached out after the paper came out. So, I just think that once something’s doable, the number of patients should not hold the companies back. There’s many, many cases where ‘rare’ disease becomes sufficiently commercially viable…

That reminds me of last week when I was in the University of Wisconsin in Madison giving a talk that wasn’t like my usual talks, it was the surgery grand rounds. And physicians were telling me how grateful their patients were to know that there are people who are working on their disease even if they didn’t have the pill yet. I was really amazed by that, that it’s a real comfort for people to know that somebody is working on the problem. And we are, we think about it night and day, we’re working on it even when we’re not working on it.

Megan: That’s got to be a really satisfying part of your work, hearing from patients that have been on sotorasib and done well…

Kevan: It’s very cool, yeah, I think that happened to me the first time back in about 2009. One of our companies made a PI3Kinase delta inhibitor, the drug that became duvelisib, that went into patients, and one of the chief medical officers said “Kevan, let me tell you about one of the patients that was on the drug.” This older CLL patient played a lot of tennis. His health was declining so he couldn’t play tennis on the top court. So, when he enrolled in the trial, his goal was to get back to the top court. After one cycle, he was there. I was like “WOW, I want more of this.” That was the first time I’d heard about a patient on a drug that we helped somewhere along the line

Megan: That’s so cool, that’s what it’s all about…

Kevan: It makes you, like Ziyang said, make the next molecule. It’s not just playing around; you realize this is people’s lives.

Megan: What combination therapies do you see as being effective for the different targeted RAS drugs?

Ziyang: I think there are a lot of combination trials out there, especially for the G12C inhibitors, and we know that growth factor signaling is an important branch that would benefit from combination therapies. I think it’s all about the biological insight that guides us to the next wave of combination therapies. Kevan’s lab had done a CRISPR screen to look at which genes would synergize with a G12C inhibitor, and that became a really good resource. There are extensions or similar work being done here and there that I think is a really powerful way to interrogate the biology. If you inhibit RAS, what else do you need to do to make the therapy more effective? Another thing I think is that we should keep a tight watch on the resistance mechanisms that are emerging from patients that receive these inhibitors. Kevan can probably comment more on this, but I think that’s another direction to go with combo therapy. A third branch I think are combination therapies that help mobilize the immune system. Sure, there are checkpoint inhibitor combo trials going on already with mixed results, but I feel like we haven’t hit that sweet spot yet.

Megan: Yeah, I’m super excited about your paper on bispecific antibodies targeting MHC I and covalent inhibitor-bound KRAS G12C peptides. It seems as though that’s going to be a crucial part of eradicating these cancers.

Kevan: Yeah, we hope so. It seems like to get to a cure, you really need the immune system engaged. And the fact that we’ve been so focused on signaling—it’s just so adaptive, with so many bypass mechanisms. The immune system just looks at things very differently. It doesn’t care too much about the signaling in the target cell—if the flag is up, it will come after it.

Megan: So, as you’re looking at more mutant-specific covalent inhibitors, like for G12R, G12S, and G12D, are you planning on finding more bispecific antibodies for those inhibitors as well, are those in the pipeline?

Kevan: Yeah, it will probably be focusing more on the cysteine side for now, because you have to wait to have a clinical compound before you can dive into that.

Megan: Of course. Kevan, what’s something you’re excited to see come out of Ziyang’s lab, and vice versa, Ziyang, what’s something you’re excited to see come out of Kevan’s lab?

Kevan: Oh, great question. The easy one is HRAS G12S inhibitors for Costello syndrome. I mean, that could just transform that whole disease. It’s a significant proportion of patients that have that mutation. I mean, that would be incredible. To have the potential to transform the disease when 5 years ago nobody was even thinking about a way to hit that particular allele.

Ziyang: This question is harder, because Kevan has so many things cooking and every single one is exciting. I think I’m excited to see Kevan come up with a general design for GTPase inhibitors. He’s done that for kinases, but GTPases are really different. I think it will happen, and I suspect it will happen pretty soon…

Megan: We’ll definitely have to look out for that.

Tommy: When you think about the RAS family, which is pretty large, we focus obviously on the ones that cause diseases, but there’s so much biology there. Do you think of your compounds, also from a classical chemical genetics perspective, as tools to study that biology? Do you think sometimes about expanding to some of these less studied GTPases?

Kevan: Precisely, yes, that’s exactly what we’d love to have: a way to program an inhibitor against any GTPase in a cell so you could understand the biology. And they have all the features that kinases have: they’re a big family, they’re similar, they each do their own special role. Being able to do that precisely would be so valuable. I think for GTPases, it’s going to be a richer pharmacological insight than genetics because of the way they’re switching. If you take out both the GDP and the GTP state, you have a real loss of function, whereas if you can inhibit, then you can block things. 

Tommy: Right, so with a small molecule you could shift that ratio of GDP/GTP ratio with timing and dose, and tune the biology in that pathway.

Kevan: Exactly. These rotation students, that’s their whole project.

Megan: Excellent, well thank you both so much for sharing your valuable time with us! We will be looking forward to seeing what breakthroughs come out of both of your labs next!

Ziyang: Thank you, we’re really happy to be able to speak to you.