Inhibition of MAPK Signaling Increases Autophagy in Mutant RAS Cancers
March 26, 2019, by Jim Hartley
On the second day of the Second RAS Initiative Symposium in December, 2017, Martin McMahon saw a poster title that prompted him to engage its lead author, Kirsten Bryant, in very interested conversation. Fifteen months later two important papers were published jointly in Nature Medicine, one by Dr. Bryant, Channing Der, and their colleagues from the University of North Carolina’s Lineberger Cancer Center, and the other by Dr. McMahon, Conan Kinsey, and their colleagues from the Huntsman Cancer Institute at the University of Utah. The papers reveal an important characteristic of KRAS-driven cancers, that they become remarkably reliant on autophagy in response to inhibition of the RAS-regulated ERK MAP kinase pathway. Recently RAS Dialogue editor Jim Hartley talked with Drs. Bryant, Der, and McMahon about the complementarity of their work and their path to joint publication. Dr. Kinsey’s medical duties prevented him from joining.
Hartley: Martin, what was it about Kirsten’s poster that caught your attention?
McMahon: When I saw Kirsten’s title across the room I immediately made a bee line over to her poster and was interrogating her in some depth about her work. And when I saw the full scope of what she had done it was really clear that both labs were about the same place in the story - maybe 75% along the way toward a publication, but there were still some critical things that had to be done by both labs. And in our case the patient who presented in HCI’s GI Malignancies clinic, in March or April of 2018, was something that we hadn’t anticipated happening. This was a patient who had gone through all the chemotherapy options and we were keen to offer him something but we did so with no expectation that he was necessarily going to respond. Not in our wildest dreams would we have ever expected a response the likes of which we saw. So that was my perspective on where we were at the time, these were very complementary studies but there was definitely work still to be done, which was largely completed over the course of the early part of 2018. There was some discussion about where to go publish this research, but I knew one of the editors at Nature Medicine, so I made the approach to see if they were interested in the manuscripts and we got a very positive response.
Hartley: Kirsten, how do you recall your interactions with Martin?
Bryant: I was excited that there was so much interest in my poster at the poster session, and my excitement continued at dinner that night when I realized that “Martin McMahon wants to sit with me and talk about science, this is great.” At that time, I didn’t know that Martin’s lab was studying the same topic, I just assumed he knew the field well and asked good questions. At the time that I presented we had the outline of the story but there were still some pieces that needed to be filled in. I guess maybe the week after the Initiative is when you gave us a call Martin to tell us you had similar stories. We both went through what we had at that point and outlined what we wanted to do over the coming months and that’s how we settled in on when we would submit to Nature Medicine.
McMahon: Conan wasn’t at the RAS Initiative meeting and I wanted to chat with him in person before we made the initial approach. We had discontinued our pancreas cancer research at UCSF in 2012 and were focusing mainly on lung cancer, melanoma and thyroid cancer for which we had invented genetically engineered mouse models. But when I came to Huntsman Cancer Institute a young physician-scientist knocked on my door one day, this was Conan Kinsey, and he said I want to work in your lab, and I said what do you want to work on, melanoma or lung cancer? And he said I don’t want to work on either of those, I want to work on pancreas cancer. He said “I know you’ve worked on pancreas before”, I said “You’re right”, I said ”Well I don’t have any funding for pancreas cancer research, but if this is where your passion lies then that’s what you should do.” Subsequently the penny dropped when I realized that in addition to being a scientist that Conan was seeing patients, predominantly pancreas cancer patients, throughout the time he was doing his fellowship and laboratory research, and that was why he was so keen to do something that would be in some way related to his clinical specialty.
Hartley: Kirsten, you’ve had some interactions with the media since your paper published, so I know you have a family interest in pancreatic cancer, which made a good fit with Channing’s lab. But where did autophagy come in?
Bryant: When I came to the Der lab, I was interested in studying cancer metabolism. This was a new area that Channing hadn’t considered before. When I came with this interest, Channing asked me to write a review to better understand the field and talk about potential directions to go. In doing that, we started looking into the autophagy literature. The people I would say that started the focus on autophagy in pancreas cancer were Alec Kimmelman and Eileen White.
McMahon: Totally agree with you, definitely Alec and Eileen.
Bryant: I had reviewed some of their papers, and looking at them from a RAS perspective, which is how this lab looks at things, we saw that they had reported this up-regulation of autophagy in pancreas cancer, and we hypothesized that KRAS must be driving this. So we started the project with the simple idea just to show that if we knocked down KRAS we would eliminate autophagy from the cells. And instead when we did that experiment, we got the complete opposite result. When we knocked down KRAS autophagy flux was actually enhanced. After we made that observation, we next showed that we could phenocopy it with an ERK inhibitor, so this suggested a possible treatment strategy. That finding led to the next experiments showing that ERK inhibitor treatment and inhibition of autophagy with chloroquine were synergistic, that’s where the study took off.
Hartley: Martin, how would you summarize how you and Conan came to a focus on autophagy?
McMahon: Three strands came together all at once. My lab also works on melanoma and a physician-scientist at U. Penn, Ravi Amaravadi, has been promoting the idea that there was induced autophagy in melanoma following pathway targeted inhibition of BRAF signaling. The second strand to this is kind of a nice story. When I graduated from Glasgow University with a biochemistry degree in 1981 one of my classmates was a guy by the name of Andrew Thorburn, and Andrew Thorburn is now the chair of Pharmacology at the University of Colorado in Denver. Around 2014 he called me up and said he had a young pediatric neuro-oncologist in his lab by the name of Jean Mulcahy-Levy who was interested in two things: pediatric brain tumors driven by BRAF, which was in my area of expertise, and autophagy, which is Andrew’s area of expertise. And so he asked me if I would be a co-mentor on Jean Mulcahy-Levy’s K08 award, which I agreed to do. Suffice it to say Jean had published a very nice paper in Cancer Discovery in 2014 where she showed that a child who had an inoperable ganglioglioma in the brain stem driven by BRAF(V600E) could be treated with a combination of vemurafenib and chloroquine, and the chloroquine appeared to restore the sensitivity of the brain tumor to the BRAF inhibitor. Although this was an N of 1 patient case report, the data were nonetheless quite compelling.
And the final strand to this, as a graduate student in Harmut Land’s lab in Rochester Conan worked on regulation of autophagy. So when Conan did the first experiment [in my lab] he brought along a western blot and he said “I think MEK inhibition is inducing autophagy”, and I looked at the western blot and I said “I don’t believe it for a minute, it’s just random lane-to-lane variation. If you want to convince me this is real you have go find a better, more robust way of measuring it.” And so, sure enough, Conan went out and identified this little fluorescent reporter in Addgene which allowed him to measure autophagic flux by the relative ratio of red to green fluorescence in the reporter. Conan put it into his initial MiaPaCa-2 experiments and then the data was very clear-cut. I remember he showed me the first in vivo experiments, it’s actually in the paper, where he took MiaPaCa xenografts and gave them either hydroxychloroquine or trametinib or the combination, and the single agents had no effect and the combination made the tumors shrink to nothing. I remember looking at Conan straight in the eye and going “Is this real?” Because we had been working on these MiaPaCa-2 cells for many years with MEK inhibition and had never seen anything more than static disease, had never seen any signs of regression before. And then he did another cell line and then he did two PDX models and he did some NRAS-mutant melanomas in partnership with other people in my lab and the observations were reproducible, they were robust, they were found in multiple different diseases.
And then of course I saw Kirsten’s poster across the room at the NCI RAS Initiative meeting and any doubt in my mind was completely dispelled because the data from an independent investigator, entirely sight unseen of what we were doing, was just the most incredible validation. It was a little nerve-wracking at the same time because when I went to Kirsten’s poster it wasn’t clear to me whether her work was already in press, you know? That was always a possibility, I think I might have even asked her, “Have you submitted this yet?” “No.” “Well thank goodness for that!”
Hartley: Channing and Kirsten have you had any follow-ups from the pharma side or from patient groups since your paper published?
Der: We have been talking to pharmaceutical companies for quite some time with regard to ERK inhibitor combinations, and one of the combinations we pushed was ERK plus hydroxychloroquine. There was hesitancy from the pharma companies, there was interest but hesitancy.
Bryant: Our work has been funded by PanCAN [Pancreatic Cancer Action Network] and so we’ve always had that connection with patients through them. In terms of specific pharma groups our paper pushes an ERK inhibitor, there aren’t any that are far enough through clinical trials to be used in combination, none are FDA-approved. We are going to open a clinical trial in collaboration with MD Anderson led by Dave Fogelman looking at a MEK inhibitor, binimetinib, in combination with hydroxychloroquine. But we should mention that Martin’s group at Huntsman has already opened a clinical trial and are already accruing patients, looking at trametinib and hydroxychloroquine.
McMahon: Yes, that’s correct. When we had the initial data on patient 1, we were able to get a meeting with the oncology decision makers at Novartis in conjunction with the 2018 ASCO meeting in Chicago. At that meeting the Novartis folks agreed to provide us with free drug which was quite important, because for patients it’s between $10,000 and $13,000 a month for trametinib, and there would be no guarantee that one would necessarily get insurance coverage for the use of trametinib in combination with hydroxychloroquine for an experimental study. We have an outstanding clinical trials office at HCI that assists with protocol development, and we were able to go from initiation of the protocol to the actual opening of the trial in just a bit more than a hundred days. It’s only for patients with pancreas cancer, and they have to have failed one line of conventional chemo, which is either going to be FOLFIRINOX or Gem[citabine]-Abraxane. And that’s the only major stipulation in terms of enrollment criteria.
One complication of these drugs is because they are FDA approved, any physician can in principle prescribe them for any cancer patient they like. And so we are already aware of physicians, oncologists in other places in the United States, who’ve been giving this combination of drugs to patients off trial and off label, in much the same way we were doing on a compassionate basis. And I can understand why they do it, because their patients are desperate, they will have heard about our paper in many cases, and so they’re keen to know if they can get access to these FDA-approved drugs. But the reality of this is you’ll only ever truly understand how good or otherwise this therapy might be by doing it in a proper clinical trial, and we went to some length in our paper to make that clear, that it really should be done in a trial.
Hartley: Channing, you and Kirsten have done a tremendous amount of work to understand the processes underpinning the autophagy dependence.
Der: All of our research is really geared toward trying to get ideas to the clinic, but the reason Kirsten’s paper is chock full of mechanistic analysis is because after Kirsten made the initial discovery, then she spent several years trying to answer the question Why does ERK inhibition lead to up-regulation of autophagy. And at that point I connected Kirsten with a lot my friends to do omics, and we threw a lot of omics at the issue, including RNA-Seq and RPPA analyses, and so collectively we have analysis that helped Kirsten figure out how ERK inhibition mechanistically caused increased autophagy.
McMahon: Can I chip in there and make a big plug for basic biochemistry. I was trained as a biochemist but now, in my position at the HCI as Director of Preclinical Translation, we do a lot of GEMM [genetically engineered mouse model] and PDX models. However, I’m a firm believer that you can’t do good preclinical translational research unless you can do really good basic research at the same time. Looking at Kirsten’s paper from my perspective, I really enjoyed the biochemistry and the depth of the analysis. Because the success of cancer therapy is going to be predicated on having a deep mechanistic understanding of how the cancer cell’s biochemistry and cell biology are wired. The hope will be that that deeper understanding will translate through preclinical models and ultimately into clinical trials. But I’m a firm believer that you can’t do the latter unless you’re absolutely rock solid on the former.
Hartley: What’s next in the lab?
McMahon: One of the challenges is that although this is a combination therapy of two drugs, we know that one of the drugs, chloroquine, at least in our hands doesn’t have any effect on its own. And so this response we believe relies critically on the blockade of the MAP kinase pathway. And from that standpoint you can therefore think of the emergence of drug resistance as being roughly equivalent to what you would expect for single agent cancer therapy. So the prediction is that patients that might respond will almost certainly develop drug resistant disease. So one of the key things we want to understand is what predicates for either up front drug resistance or sensitivity, and what also might drive the acquisition of resistance down the line. And then the second thing is to add additional agents onto the backbone of trametinib and hydroxychloroquine, what can we possibly add that would actually make the therapy more effective and increase the durability and decrease the likelihood of drug resistance.
Bryant: I think that pretty much summarizes the things we’re focused on as well, basically figuring out ways to improve the combination, as well as understanding if it can be ported over into other cancers.
McMahon: I can say that in data that is not in our paper, the ubiquity of the response of cells to blockade of the MAP kinase pathway by upregulating autophagy is quite a remarkable thing. We’ve been looking at this in KRAS and BRAF-mutant lung cancer, in KRAS- and BRAF-mutant melanomas, in GNAQ or GNA11-driven uveal melanomas, and obviously in pancreas, and it’s quite remarkable how robust this response is.
Hartley: Any final thoughts?
McMahon: On the one hand I reflect upon what has been a remarkable collaborative venture, and collegial venture, with Kirsten and Channing and their colleagues at UNC Chapel Hill and elsewhere, which is kind of a feel-good thing for us as scientists, because we were able to work this out in a manner where everyone came to benefit. But it’s also important to reflect back on the fact that we’re tackling a disease that’s a real silent killer, and it’s probably one of the most difficult diseases in the whole panoply of cancers to target. The feel-good factor of having prosecuted these two papers and getting them published in a high-profile journal is certainly a good thing, and I’ve felt good all week about this, but it’s also sobering as well when we think about the challenges ahead in terms of the numbers of patients that desperately need help.
Hartley: Channing and Kirsten, what have you been telling patients and patient organizations?
Bryant: We’ve been very careful to stress the point of cautious optimism. It is exciting to have a potential new direction to be going in, but we definitely do not want to give people false hope. It is still too early to know how this will impact patients. The feedback that we get from the members of the local [PanCAN] affiliate quite frequently is that they are happy to know that someone is working to try to make things better, and that’s certainly what we are striving to do.