Raymond J. Deshaies
Senior Vice President for Global Research, Amgen
By David Zierler, Director of the Caltech Heritage Project
March 24, 2023
DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It is Friday, March 24th, 2023. I'm delighted to be here with Dr. Raymond J. Deshaies. Ray, it's great to be with you. Thank you for having me in your house.
RAYMOND J. DESHAIES: Sure. This'll be fun.
ZIERLER: To start, Ray, would you please tell me your title and institutional affiliation?
DESHAIES: I'm Senior Vice President for Global Research at the biotechnology company, Amgen.
ZIERLER: Tell me about some of your key responsibilities in that role.
DESHAIES: I'm responsible for developing molecules that enter human phase one clinical studies, and I'm responsible across all therapeutic areas and across all modalities up until the filing of an IND – investigational new drug – application, to the Food & Drug Administration, which is the prerequisite for human clinical testing. We continue to support programs after they enter human clinical testing, but I would say 80% of our effort is up until the filing of an IND application.
ZIERLER: The term "modalities," what does that mean in this context?
DESHAIES: "Modalities" means the actual agent that is being introduced to interdict the disease process. A modality could be a small molecule, a chemical like aspirin. It could be an antibody like Herceptin. It could be a cell therapy like Gilead's Yescarta. It could be a gene therapy like Novartis's Zolgensma.
ZIERLER: The term "global research," does that suggest there are international partners that you work with?
DESHAIES: No. It means that we have a global footprint. I have staff that are in Munich, I have staff that are in Copenhagen, and I have staff that are in Burnaby, British Columbia. When I first joined, I had some staff as well in Shanghai. It's just that we operate globally. To better answer your question, we also have multiple international partners but that is not the reason for the "Global" in our name.
ZIERLER: How much interface do you have in Washington, D.C., with the FDA but, more generally, with health policy?
DESHAIES: I, myself, very little. We have a regulatory organization in R&D that deals extensively with the FDA. We have a policy group in Washington, D.C., as well. But I, myself, not really.
ZIERLER: Ray, how much of your work is fundamental research? Are you doing anything that's simply curiosity-driven these days?
DESHAIES: Yes and no. [laugh] I still have a postdoc who followed me from Caltech, who's doing really fundamental research on a particular protein that is actually of interest to Amgen as a drug target. I would say there's some fundamental research that goes on within Amgen but, by and large, not a lot.
ZIERLER: How much are you working with attorneys in an intellectual property context?
DESHAIES: We have attorneys that work with people in my organization to file IP around new molecules we develop. I, myself, am essentially not involved in that process at all.
ZIERLER: On any given day, how much of your time is spent on science, and how much of it is administrative?
DESHAIES: It depends on how you define—
ZIERLER: It's a blurry line sometimes?
DESHAIES: Yeah. I would say it might be a third evaluating science. I spend a fair amount of time just reading about developments in the field, tracking the field. I don't know if you'd call that spending time on science. It's keeping abreast of developments. The one-third cited above, I'm counting mainly the time that is spent evaluating candidates in our pipeline. We have various governance forums where we talk about program advancement, we review programs, and so forth.
ZIERLER: Ray, do you have a sense of the overall history of Amgen, sort of its founding mission, and how it's changed over the years?
DESHAIES: A reasonable sense of that, yes.
ZIERLER: What is its origin story?
DESHAIES: Amgen started as a biotechnology company in the 1980s. We're now entering our 43rd year. It was originally a professor at UCLA, Winston Salzer, who conceived of forming this biotechnology company, and he was having trouble getting funding is my understanding. He had enlisted people from Caltech and UC Santa Barbara, to help with the founding. Lee Hood was involved as an early advisor when he was at Caltech. Norman Davidson was also involved as an early advisor. I think the main person at UCSB was John Carbon. I know a lot of this because the first company I started involved one of the very early employees of Amgen, and he told me a lot of war stories. This war story in particular was Winston was having trouble raising money. Winston was a typical academic in many ways: not terribly pragmatic [laugh]; big visions; not a lot of pragmatism. He was all over the map in terms of things he wanted to pursue. In the early days, they were pursuing things like growth hormones for chickens. They were pursuing recombinant indigo dyes. [laugh] They're all over the map. He needed an experienced industry veteran to help him.
One of his former graduate students was this guy named Philip Whitcome, who had gotten a PhD in molecular biology at UCLA, and then decided he didn't want to continue working in bench science. He went and got an MBA at Wharton. Then after he got the MBA at Wharton, he went to Abbott in Chicago area, and his boss was the guy who became the first CEO of Amgen, George Rathmann. One day, Winston calls Phil up, his former graduate student, figuring that Phil is going to be able to help him out because Phil has now got an MBA, and he's at Abbott. Winston says, "I'm trying to start this company. I'm having a lot of problems. The investors want a CEO. Do you have any ideas for me?" Philip approaches his boss, George Rathmann, and George is like, "I'm interested in that." George flew out to California and met Salzer. Fast forward, George became the first employee of Amgen and the founding CEO. He was CEO for the first 10 years of Amgen. About a year later, Philip joined George at Amgen. Phillip wore many hats during his tenure at Amgen. He was also at Amgen for about 10 years or so before then going on into other biotechnology companies. In any event, all of the original stuff they were working on – chicken hormones, indigo dye – blew up. Eventually what happened is they were working with a scientist at the University of Chicago who had identified erythropoietin. That's the protein that stimulates production of red blood cells. Amgen wanted to clone the gene that encodes erythropoietin. At the time, it wasn't so trivial cloning genes from protein sequence. What you would do is you'd make an oligonucleotide that would correspond to that protein sequence. But because of redundancy in the code, you can't just make one oligonucleotide because certain amino acids like alanine can be coded for by three, four, five, six codons, so you have to make a degenerate series of oligonucleotides. In the early 1980s—that's when I was just starting graduate school—making oligonucleotides was highly nontrivial.
There was a guy at Boulder, and now I'm forgetting his name, who was also one of the key early advisors of Amgen. It's Marvin Caruthers. He had figured out the chemistry for making oligonucleotides. He was making them in his lab. They approached him. He was already an advisor, and he made the oligonucleotides that were used to go fishing by hybridization for the gene that encodes erythropoietin. Amgen cloned erythropoietin. They expressed it in CHO cells. They could purify the recombinant protein. That became their first product. There was another company, a famous biotechnology company, started by very well-known Harvard professors, called Genetics Institute. They were also producing erythropoietin, but they were using a different production method for the erythropoietin. There then was a very famous legal case that was really transformative that was decided in 1991. I'm forgetting who sued who, but it was Genetics Institute and Amgen battling over who really owns erythropoietin. Amgen won that lawsuit. That was absolutely key, because erythropoietin turned into a major drug. It's the drug that—what's his name? —the guy who was cheating in the Tour de France and was eventually—
ZIERLER: Lance Armstrong?
DESHAIES: Yeah, Lance Armstrong. That's the drug that Lance Armstrong was using – erythropoietin. A lot of elite athletes—
ZIERLER: As a performance enhancer?
DESHAIES: Yeah, because it stimulates red blood cell production, so you have more oxygen-carrying capacity in your blood. That oxygen-carrying capacity is a limit on muscle performance. It became an important drug for kidney dialysis, because kidney dialysis patients are anemic. It also for a period of time was being used in oncology, because some of the oncology drugs make you anemic. But then it turns out that it had a negative impact on overall survival of cancer patients, so that indication got withdrawn from the market. But erythropoietin and Aranesp, which was an extended-duration version of erythropoietin, essentially, those both became multi-billion-dollar drugs, and really fueled the rapid growth of Amgen during the '90s, after they won that lawsuit with Genetics Institute, and they really owned outright the exclusive ability to market erythropoietin as a drug.
ZIERLER: Amgen took off from there, essentially?
DESHAIES: Amgen really took off, and Genetics Institute was sold. American Home Products acquired 60% of Genetics Institute for $666 million in 1991.
ZIERLER: Now, for you, in graduate school, postdoc, early faculty position, are you tracking these developments? Are these biotech advances registering with you?
DESHAIES: No. But I have a funny story about that because my best friend in graduate school is a guy named Kevin Jones. He's just retired last year. A professor at University of Colorado Boulder. He was very actively following the stock market as a graduate student. Probably we're like second-year graduate students. I had some money. I had a little bit of excess cash. I think it was like $700 or $1,000. He's like, "You should buy this Amgen stock. I've been following this. This is a great company [laugh], and you're going to make money on this. You should buy this Amgen stock." I was like, "Yeah." I had gotten into cycling. We were in Berkeley, and I was cycling into Berkeley Hills. "Yeah, Kevin." I spent the money on a bicycle.
DESHAIES: It was a Trek 670, and I still have the bike. It was about $700, $800. Kevin, I was out visiting him—I don't know—about two years ago, and I stayed over at his house. He said, "I hope you really got a lot of value out of that bike"—
DESHAIES: —[laugh] "because I sat down and did a calculation, and the $700 you spent on that bike today would be worth about a quarter million dollars in Amgen stock." [laugh] He said, "I hope you still have that bike." I said, "I still have that bike." [laugh]
ZIERLER: Ray, is this to say that your research career, at least in the early portion, you were not overly interested in translational research?
DESHAIES: Not at all, zero. Zero.
ZIERLER: We'll get to what changes later on. Do you retain an affiliation with Caltech? Is that relevant for what you do now?
DESHAIES: I'm still a member of the Athenaeum. [laugh]
ZIERLER: You're listed as a visiting associate.
ZIERLER: Is that just an honorific?
DESHAIES: Kind of. I'm not really doing anything at Caltech. I have friends at Caltech. I hang out with some of the biology faculty members. I have connections, but I'm not really doing anything at Caltech.
ZIERLER: Some overall questions about the field. Biochemistry, undergraduate major, focus in graduate school, this has been a theme in your career. Has the term changed over the years in light of all of the advances in biotech, in computation, and things like that? What aspects of biochemistry retain what they meant for you when you first started, and what's totally different now?
DESHAIES: When I took biochemistry, there was that famous chart that people would have on their wall from the Albert Lehninger textbook—which is the textbook I used—of all the metabolic pathways, like the Krebs cycle and all that stuff, glycolysis. Probably what's changed the most is just an increasing interdigitation of all of the molecular sciences, if you will. The boundaries became really fuzzy between what was considered biochemistry versus molecular biology versus cell biology. You could even see it in departmental names. At Berkeley, biochemistry became subsumed into a larger department, as I was leaving, called MCB, molecular & cellular biology. That was a trend across all the graduate programs. You saw these mergers of individual departments into super departments that usually had some combination of molecular, cellular, and/or developmental biology, or Caltech has a biochemistry & molecular biology graduate program, and so just a blending of the boundaries. In the early 1980s, you had your biochemistry department, you had your molecular biology department, you had your cell biology department, and those boundaries broke down.
ZIERLER: Has your research gotten more computational over the years due to those advances?
DESHAIES: Yeah. At the end, when I was at Caltech, we were doing a lot of proteomics, which inherently was fairly computational in nature, because you have to computationally link these signals you see in a mass spectrometer to proteins encoded in the genome that they are derived from. I had a person in my lab—we just published a paper actually—who did a lot of computational analysis, one of my postdocs, so the work did become more computational.
ZIERLER: Of course, AI and machine learning are very much in the news these days. What have those advances meant for you, and how has Amgen embraced AI and machine learning?
DESHAIES: So far, a relatively modest affect. The most noticeable thing is we entered a collaboration a year ago with a company called Generate Biomedicines, based in the Boston area. They're using computational methods to go from protein sequence to protein structure. We're using computation increasingly to try to link not just from protein sequence to protein structure but also to protein function and physical properties. For example, we've developed an algorithm, and this sounds boring, but it ends up being important. We've developed an algorithm that accurately predicts viscosity. Viscosity for any antibody is important. We'll deliver Repatha, one milliliter, in a self-administered syringe, at 140 mgs per ml. You get a protein up to a concentration of 140 mgs per ml, viscosity becomes a huge issue, because if it's the viscosity of molasses, you can't force it through a 26-gauge needle. The pressure wave will crack the glass casing. You want it to be much more like water [laugh] and not like molasses. Being able to predict that computationally in advance, and discard candidates that are going to have high viscosity, has a lot of value, just from a manufacturing and commercial perspective.
ZIERLER: You mentioned early in your career, you were not really tracking biotech. You were not overly concerned with translational research. What changed more broadly in the field that got you involved in those realms?
DESHAIES: It's nothing that changed in the field; it was really in me. Even when I got to Caltech, I had no interest in translation. By training, I was a yeast geneticist, and then I started doing biochemistry with yeast extracts. When I came to Caltech, I was merging biochemistry and genetics, and molecular genetics and molecular biology to address cell biological questions. I was particularly interested, when I came to Caltech, in cell cycle, the control of cell division. One of my projects was very genetic, looking for genes that regulate the process . Another project was very biochemical, identifying an enzyme that catalyzed the key step in the process . That enzyme turned out to be the archetype or the progenitor of a huge family of enzymes. We discovered it in yeast, but we now know that it's the progenitor of a family of enzymes in human that's 250 or so members. That enzyme is a ubiquitin ligase. What it does is it catalyzes the transfer of a molecule of ubiquitin, which is a small protein, onto a substrate protein, and it's a covalent linkage. You take this little ubiquitin protein, which looks like a lollipop, and you attach it to a substrate protein in the cell. The enzyme doesn't stop there. It keeps adding lollipops, and it adds lollipops to the lollipop itself, to build a chain of them. That chain of ubiquitins is a signal for a cellular protease to grab onto that chain of ubiquitins, and destroy the protein that's attached to it. This is the way that cells get rid of proteins. That's called the ubiquitin-proteasome system. When we made the discovery of the ubiquitin ligase enzyme, we got really interested in it because: it was conserved to human, number one . Number two, it plays important roles in regulatory processes like cell division. Number three, it's the progenitor of a huge family. Number four, it had really interesting regulatory properties.
This was not a dumb enzyme that just runs around attaching ubiquitin to everything. It very specifically attaches ubiquitins to proteins involved in the cell cycle that were restraining the cell from dividing. It attaches ubiquitin to those proteins, they go away, and now the cell can advance and divide. This was a very highly regulated process, and so I was like, "Wow, this is important." I started switching my studies from cell division to ubiquitin-dependent turnover. This started happening around 2000. We reported the enzyme in 1997 , so it really would've started happening in '97, and accelerated around 2000. In 1998, I met this guy, Craig Crews. We had the same fellowship. It was the Burroughs Wellcome Award in the Basic Pharmacological Sciences. We had to show posters at a meeting for fellows, and he had a poster there. He had done a postdoc in a chemistry lab, Stu Schreiber at Harvard. He was using small molecules, which are just like chemicals with a molecular weight, usually less than 1,000. How much of a background do you have in biology?
ZIERLER: I'm tracking with you.
DESHAIES: OK. There's this thing known as the two-hybrid system, which was very popular 20 years ago, where you take a transcription factor, and you split it in two. Usually there's two parts of a transcription factor: DNA-binding domain and transcriptional activation domain. The DNA-binding domain brings the transcription factor to the gene. It binds the DNA where the coding sequence is. The activation domain recruits the RNA polymerase to transcribe the gene. The two hybrid system, you broke those two pieces apart, and then you fuse them independently to two different chunks of protein. If those two chunks of protein naturally interact, they reconstitute the transcription factor. The two pieces of the transcription factor – the DNA-binding domain and the activation domain – they won't find each other by themselves. But if they're attached to things that naturally interact, now you reconstitute that activity. That was a very popular assay for identifying interacting protein pairs.
Craig was generalizing this and using it to find small molecules that would join two proteins together, serve as a bridge. He had this poster, and we had just discovered this ubiquitin ligase enzyme. I'm looking at his poster, and I thought, "Gee, that's a cool idea. I wonder if you could use a small molecule to bring together a protein in the cell, any protein, and link it to a ubiquitin ligase enzyme. If you could do that with a small molecule, that small molecule would program the degradation, the attachment of ubiquitin and degradation of the protein that you're bringing to the ubiquitin ligase. That could be an interesting way, an interesting therapeutic approach of programmed elimination of any protein from the cell." Craig and I start talking about this idea. We're building on the idea in our conversations. Then when I came back to Caltech, I immediately wrote a Gates Grubstake grant. I got one of those Gates Grubstake Awards, and we began working on it. I had a graduate student in my lab, who was an MD and was interested in doing something more applied. Craig had a chemist in his lab, who started making these compounds. We began this collaboration. Through that interaction, I then became aware that there was a chemical that Bristol-Myers had been working on. They had a research outfit out in Tokyo. The chemical killed cancer cells, but they had no idea why. At some point they closed that research institute in Tokyo, and they just put everything in the public domain. They published. It's like, "This compound epoxomicin kills cancer cells. We don't know why." Craig's postdoc in the Schreiber lab, that's what that lab specialized in, finding compounds that kill cells or have an interesting therapeutic property but where the mechanism was unknown. What they would do is they would synthesize the compound with a biotin attached. Biotin is like a handle that you can grab. Then they would throw it on cells, and they would then grab the biotin handle, pull it out, and see what stuck. Craig did that with epoxomicin, and he figured out the target was the proteosome. The proteosome is the thing that degrades proteins that have ubiquitin attached. This was interesting.
Craig and I were working on a targeted degradation idea of bringing things to ubiquitin ligases. He's got this compound in his lab that, just by chance, inhibited the proteosome. Meanwhile, I had a staff scientist in my lab, Rati Verma, who had discovered a new activity in the proteosome that was essential for the function of the proteasomes, known as Rpn11 . Now, why was this interesting? This was interesting because there was a company, Millennium Therapeutics, that was developing an inhibitor of the proteosome for therapy of multiple myeloma, a blood cell cancer. It's what Ahmed Zewail died of. That inhibitor was showing really promising data in clinical development. Craig raised the idea of starting a company with me. I was not entrepreneurial at this time. We got this little collaborative thing, which is kind of like therapeutically oriented. But that was the only applied project I had in my lab.
I was not thinking entrepreneurially at all. But the guy Craig had trained with, Stu Schreiber, he's famous for being really entrepreneurial. He's launched multiple companies. Craig had much more of an entrepreneurial mindset. He calls me one day in 2000, and he said, "Why don't we launch a company? We got these three things going on. I got epoxomicin, which is a proteosome inhibitor, and we know that Millennium's data on proteasome inhibitors is looking positive . You've got that new target of the proteosome, Rpn11. We're doing this thing of small molecules that link substrates to ubiquitin ligases. We're collaborating on that project. We could put this all together, and start a company."
Initially, I was like, "I don't know anything about starting companies. How do you start a company? What do you do?" I had tenure at this time, and Craig didn't have tenure, and so wanted me to take the lead because he was focusing on getting tenure. I went to the Tech Transfer Office, and there was this guy, Scott Carter. He had been a graduate student of Peter Dervan. After Peter Dervan, he went to law school, got a law degree, came back to work in the Tech Transfer Office for a couple years, and then ended up going into one of the big firms, Latham & Watkins. He was still very young, and he wanted to get involved in something. He got all excited about this, and he said, "I'll help you out." I was like, "I don't know what to do." He's like, "That's OK. I'll help you out." We put together a business plan, and we put together a pitch slide deck, and then they started introducing Craig and I to relevant folks. Number one, we got a lawyer. This is amazing. Scott hooked me up with Alan Mendelson, who just passed away. But Alan was the lawyer. He was the lead outside counsel on the Amgen versus Genetics Institute trial.
ZIERLER: The one that Amgen won?
DESHAIES: Yes. Alan was the most famous lawyer in the biotech industry. I looked him up. He was in American Lawyer magazine as one of the top 100 lawyers in the United States. How many lawyers are there in the United States?
DESHAIES: Like hundreds of thousands. [laugh] He was in the top 100, all domains of law; not biotech law. He was a huge fish. He agreed to represent us. Now, who did Alan connect us with? I fly up to San Francisco to have dinner with Alan. At this time, I've just gotten tenure. It's probably 2000. I came to Caltech in 1994 as an assistant professor. Alan's like, "You guys, like, you're a yeast geneticist. You need adult supervision."
DESHAIES: I said, "Yeah, I agree with that."
DESHAIES: He said, I got just the guy for you." That guy was Philip Whitcome. Philip Whitcome, as I mentioned earlier, was the graduate student of Winston Salzer, who introduced Winston to George Rathmann, who became the first CEO of Amgen. That's how I know all these Amgen war stories from the founding of Amgen, because I hung out extensively with Phil Whitcome for a year and a half, raising venture capital to start Proteolix. That's the whole thing condensed. Craig, myself, Philip Whitcome, we brought on board Susan Molineaux, who became the chief scientist, and then we formed Proteolix. We had those three different things I told you: the epoxomicin, Rpn11, the new proteosome target, and what we were calling PROTACs, these proteolysis-targeting chimeric molecules that link a substrate to a ubiquitin ligase [5-7]. We had those three things. The investors were like, "We'll fund the epoxomicin—that'll be like 90% of the money—10% of the money on this new target Rpn11, nothing on PROTACs." We closed at the end of 2003. We had started in 2001 to really raise the money. By then—who was the prime minister?—the Blair-Clinton press conference, which trashed the biotech industry in March of 2000, had happened. The whole NASDAQ crash had happened. One investor said, "If this was 1999, I'd give you $15 million for PROTACs. But it's not 1999, and I'm not giving you a penny for PROTACs." [laugh] We raised the money. That was when I became involved in commercialization and entrepreneurialism.
What I learned from that, number one, I learned that really interested me, that whole intersection of science and business and finance and law. How do you pull everything together to start a company to take ideas from the lab and commercialize them? I just found that whole aspect fascinating. Then I was a board observer. I was not a board member; Craig and I were board observers. I went to all the board meetings. Just seeing that whole process, interacting with the people and the company, [laugh] I got bit. I got bit. But I didn't change fundamentally. From then forward, I always had a translational-oriented project in my lab but only one or two. Out of a lab of 14 people, the vast majority of what we were doing was fundamental and had no commercial orientation whatsoever. But I consistently had a project or two that was more oriented towards can I find a drug? Can I find a drug first against Rpn11, then against this other related protein known as CSN5, and against p97/VCP? We always had something in the background, one or two people working on something like that. But it really all went back to the founding of that first company, Proteolix, and that whole experience of raising capital. Then Proteolix ended up being successful.
We developed a drug, and we got acquired. Before the drug got approved, we got acquired, and then the acquirer got the drug approved by the FDA. Amgen owns that drug now, and it's based on Craig's epoxomicin. It's called carfilzomib, Kyprolis. It's in multiple myeloma. Amgen's selling ~$1.3 billion a year of that drug. It's a reasonably successful drug. That whole experience of having an impact, There's two days I really remember in all of this: one was the day it got approved by the FDA—that was July 20th, 2012, and then a year later, on August 25, 2013, Onyx got acquired by Amgen, and that was a big acquisition. Wall Street attributed seven billion of the purchase price to Kyprolis. That made the Wall Street Journal. [laugh] It's a different kind of impact. As a scientist at Caltech, you're measuring your impact based on a Nature paper. You get a Nature paper, you're all excited, but the reality is 99.9% of the world is not paying attention. But The Wall Street Journal, that's a different kind of impact. Feeling like, hey, I started that company that triggered this acquisition that's featured in The Wall Street Journal, that's impact. That's seductive. It's addictive.
ZIERLER: A question about—with your career superimposed, Caltech's increasing orientation toward entrepreneurialism, start-up culture, and biology. If we were to simplify the narrative and use Lee Hood as the starting point where he starts to get interested in these things and, culturally, Caltech was not there alongside him. You may have heard famously Murph Goldberger was disapproving of what Lee Hood was doing. Then on the other side, the bookend where someone like Sarkis Mazmanian can happily combine both worlds. He can have his start-ups, he can have all of this translational work, and he can do his fundamental research in the lab, and they work seamlessly. For your career, where do you see and where do you fit in that overall institutional evolution?
DESHAIES: That evolution was happening while I was at Caltech. There was no one defining moment. But I very much remember a faculty meeting, and we were having a discussion about partnering with UCLA on an MD PhD program. We ended up forming two partnerships, with UCLA and with USC, on MD PhD programs. But there was a discussion. This would've been the late '90s. I don't remember if Baltimore had arrived yet as president. But the discussion was acrimonious. There were very strong viewpoints from some of the very senior—by "senior," I mean older faculty members—who were the real dyed-in-the-wool Caltech purists, who were negative about having these joint MD PhD programs. It was polluting the pure fundamental science orientation with pragmatic considerations.
ZIERLER: Who would've represented the old guard? Who are you thinking? Like Ray Owen?
DESHAIES: Ray wasn't coming to faculty meetings at that time. No. The person I remember the most—and this is ironic because he was one of Lee's chief lieutenants when Lee was there—was Eric Davidson, who I remember making an argument against this, which again is very ironic because Eric was Lee's second-in-command when Lee was chairman. He was the executive officer. I really remember Eric making that argument. Probably Seymour Benzer as well, I think he might have been on that side as well of being negative about it. It's so long ago, I just don't remember enough. But there were people who were already engaged with industry. Norman had been intimately involved with Amgen over the years, and had retained that connection from the early 2000s until he passed away, so he was clearly thinking that way. Mel Simon, who was chairman at the time—I don't know if it was John or Mel who was chairman at that time because when I arrived, John Abelson was chairman. Then Mel Simon was chairman. They had jointly launched a company named Agouron that had developed one of the HIV protease inhibitors called Viracept, which was the top-selling protease inhibitor in 1999. Both of them had both gotten wealthy out of that, made quite a bit of money. John and Mel weren't doing applied research in their labs. They were both doing very fundamental research. There were multiple people like that who showed that you could be doing fundamental research in your lab, yet have these activities on the outside that were more entrepreneurial. What's different about when you get to somebody like Sarkis is now you have somebody where the research in the lab is more directly applicable to the companies that are being formed. When I was there, especially, this transition is happening in the '90s. In the ‘90s, you did not have people in the department who were doing work in their lab that I would call directly translatable. But there were still nevertheless people like John Abelson and Mel who—
ZIERLER: Who had these ventures?
DESHAIES: Yeah. Probably the person who would be most relevant to doing work that was translatable was Steve Mayo, because he was doing protein design. Although, in reality, people have barely still yet figured out how you translate protein design into products. But that seemed like you could design proteins. You could design therapeutics. We were all naïve in 1998. He launched Xencor at that time. It was like, "They're going to be making drugs." Xencor still hasn't developed a drug of their own that has made it to the market. [laugh] But that just felt more applied, that you're designing. He's not primarily trying to figure out things about nature. He's trying to design new proteins, and that's inherently pragmatic. You're trying to design widgets to do things, as opposed to I'm trying to understand some deep secret in nature. It was a process, and that transpired over time.
ZIERLER: How relevant was David Baltimore's presidency to these developments?
DESHAIES: Not very, I think. David was a proponent of that, but I don't think it would've been any different if some totally random other person was president at the same time. I don't think anything would've been different. David bought the St. Luke's Hospital, which is on the western side of Pasadena. He bought it with the intention of creating an incubator there where people could incubate ideas from their labs that were translational on the way to company generation. It just never went anywhere. The incubator thing never happened.
ZIERLER: You think because it's just too orthogonal to what Caltech is?
DESHAIES: I don't know why it didn't happen. I don't know if it's because he couldn't get the investment, because he was going to have to refurbish the space, and it was going to cost money. Caltech sold it right before the burst of the real estate bubble, and actually made quite a bit of money on it, is my understanding. But it got sold because it just wasn't going anywhere. Meanwhile, there were commercial developers like Alexandria, which is based in Pasadena. They're the biggest real estate company in biotech. They had built the Alexandria Innovation Center on Hill Street. So biotech was happening in Pasadena. I don't know all the details of what happened with St. Luke's, but it just didn't happen. There was another thing David tried to do later. He had enlisted me as his deputy in this later effort. He went to Eli Broad, and he tried doing something that we ended up christening Broad West. It was modeled on the Broad Institute. Now you have to go back before the Broad Institute was founded, in 2003, which was right in the middle of David's time as president of Caltech. Eli Broad, who was a trustee, wanted to make a big investment in medical research, and he approached David. David's like, "I want to introduce you to somebody." He introduces him to Eric Lander. I don't know if you've ever met Eric Lander.
ZIERLER: He's an intense guy.
DESHAIES: Yeah. Eric completely sold Eli on this vision of what he wanted to do. Eric's brilliant, and he is extremely persuasive, a very compelling person, and a very broad visionary thinker. Eli was smitten, and he's like, "I got to do this." He wanted to do it in LA, but then Eric said, "No. The only way I'm doing this is if we do this in Boston."
ZIERLER: Because LA's not a biotech hub?
DESHAIES: Because LA's not a biotech hub.
ZIERLER: He didn't want to be a pioneer?
DESHAIES: LA had natural disadvantages. Caltech doesn't have a medical school. UCLA to Caltech to UCLA is—
ZIERLER: It's a schlep.
DESHAIES: It's a schlep. USC is also a schlep. Where would you put the Broad Institute? Somewhere in the middle? Geographically, it's not ideal, and it just didn't have the same center of gravity. There was not a biotech industry established here. Eric was right. Eric is like, "Cambridge is the center of gravity. It's got to be in Cambridge." Eli capitulated, and they built it in Cambridge. Fast forward, David's president, Eli has expressed interest in doing more in the LA area, so David spins up a proposal, and I was his lieutenant in helping him draft this, called Broad West. It ended up being very contentious. There was a meeting I missed, but I heard about it. I was traveling. But this became a famous meeting, because David had a vision, and his vision was cell therapy. He had these collaborations going on at UCLA at the time in cell therapy with James Economou and—I forget who the other ones were—Toni Ribas perhaps. He wanted to focus it on cell therapy as being the next-century medicines. What that meant is it cut the chemists out. When it became known more broadly within Caltech that David was out trying to raise $300 million from Eli Broad for this institute, the chemists requested a meeting. This is the meeting I missed. Apparently at this meeting—Mayo told me about this—Zewail is at this meeting, and Dervan, and all the top people in chemistry. At some point, Zewail asks David to describe his vision, and David describes his vision. Then Zewail goes, "That's fine, but can you really tell me what your vision is?"
ZIERLER: Oh, no.
DESHAIES: [laugh] Apparently this meeting did not end well. But, at the end, Broad gave David $5 million, which just went into David's lab. There was no Broad West. It was like, "Here's five million bucks," and that was the end of it. But that was David's big effort to really drive a translational medicine effort at Caltech. When that fell apart, then this whole thing went into hibernation. Essentially nothing happened until this recent—I think it's about $100–$150 million commitment from what's his name?
DESHAIES: Yeah, from Merkin. Increasingly, you have a lot of people at Caltech doing medically-relevant work. Historically, there were always people in other departments that were doing stuff that was translational…like Axel Scherer, who's got I think a hundred patents or something; Mory Gharib, who is doing devices and engineering. You got all these people at Caltech in other divisions that were doing stuff that was relevant to biomedicine. There was also some stuff going on in chemistry; not so much bio. Dervan had done some stuff, and had a start-up. Mark Davis had some start-ups. You had distributed activity in different divisions at Caltech, but it just never really came together into a real juggernaut. Part of it's just size. Caltech is small. We don't have a medical school. There's partnerships with City of Hope. There's been some activity there but not a tremendous amount of activity. Caltech is more of a fundamental research place. I could do translational things, but—
ZIERLER: Institutionally, the core is foundational? It always has, always will, kind of thing?
DESHAIES: Yeah. The irony is what happened with the whole PROTAC thing. I think the PROTAC thing is really interesting. I don't know how much you know about it. PROTACs is like a huge thing now, huge in industry. I would say the amount of investment in the industry now is approaching $20+ billion. There have been ~10 start-ups launched around PROTACs. Every major company has an internal program. We have one in Amgen. It is viewed as the next big potential frontier in small molecule-based drugs. Whether that ends up turning out that way or not remains to be seen. But there's a huge amount of investment and excitement around this approach. When Craig and I came up with this, I told you the story about Jim Blair from Domain Associates who wouldn't put a penny into it because it was too blue-sky. I tried raising money for it. The DOD had these congressionally-directed medical research programs – CDMRP. They have one in breast cancer. They were giving out big awards, like $5 million. I tried twice getting one of those, an innovator award, and I wasn't able to get the funding. At some point, Craig and I split ways because Craig was doing both chemistry and biology in his lab, and I was only doing biology. They were making the compounds, and then we were testing them. Not surprisingly, the people in his lab were like, "Why are we making this stuff, which is the boring part, and then giving it to these guys at Caltech to do the interesting stuff?" Craig was like, "Yeah, that's a good point." [laugh]
DESHAIES: We parted ways, amicably. Then I was looking for a way to continue the work, but I'm not a chemist, so I took two approaches. One is I went into the chemistry department, and I tried to convince people in the chemistry department to collaborate. Nobody was interested – even the people in chemistry who ended up being more entrepreneurial. I went to Brian, I remember. At the time, he was not doing entrepreneurial stuff. I think he became more so later on. The chemistry wasn't interesting. They're all about, "We want to discover new reaction mechanisms like Bob Grubbs and metathesis." He won a Nobel Prize for metathesis. That was the orientation of people in chemistry. I'm going around to people in chemistry, saying, "Hey, we got this cool technology to degrade proteins." They're not interested. Then I tried raising money, but I figured I needed a lot of money—not NIH money—because I couldn't bring in chemistry graduate students because I couldn't train them in chemistry. I was going to go hire people who were medicinal chemists out of industry, and then I would need to pay industry-competitive wages. I needed like a $5 million grant to do this. I tried a couple times. I couldn't get the money. I got turned down. Then I just stopped working on it. Then years later, this whole thing explodes, and now it's this massive part of industry. But I couldn't get any traction on this 15 years ago.
ZIERLER: You were just ahead of the curve?
DESHAIES: A little too far ahead of the curve, it would appear.
ZIERLER: [laugh] Ray, let's go back and establish some personal history now. As an undergraduate at Cornell, how did you come to biochemistry?
DESHAIES: I'm not 100% sure. I went to Cornell. I was in the Ag school. I ended up in the Ag school because one of my hobbies in high school, of all things, was growing plants, like ornamentals, flowering plants, growing plants in pots and baskets. I lived in an inner-city ghetto. It was probably my escape from that. I lived in a three-story inner-city house, and the tenant on the third floor, we had a very close relationship with. She said, "You should go to Cornell because they have an Ag school, and maybe you can major in horticulture." I was like, "Oh, OK." [laugh]
DESHAIES: I applied. I got in. I went to Cornell as a horticulture major. You had to take a lot of pre-reqs. You had to take chemistry and physics and calculus and biology. As I was taking the pre-reqs, I was gravitating towards more molecular things, so away from horticulture. The things that really interested me in the courses were the things that were very molecular. Just by a gradual process of evolution of my interests, I ended up as a biochemistry major with no idea of what you do. Technically I was premed, because if you're a biology major as an undergrad, it's almost by default you're premed. But I never did the premed stuff. They had a premed club at Cornell. I never joined the premed club. I never worked in a hospital. I never did the premed kind of things. My GPA was OK. But I didn't ace all my chemistry courses, and I was worried like, "Wow, will I be competitive for medical school?" When I was a junior probably is when I started worrying where's all this going. I had gone home the first two summers after freshman and sophomore year. I decided during junior year to seek out a research appointment for a summer at Cornell, and I got into a lab, a real biochemistry lab but actually in the plant physiology department, but with a real biochemist. I did research in that lab.
ZIERLER: Who was the professor?
DESHAIES: André Jagendorf. He was in the National Academy [of Sciences], famous for proving, early in his career, Mitchell's chemiosmotic hypothesis for how cells make ATP. I did the research over the summer. I really enjoyed that. Then I decided during the summer after junior year to apply for graduate school. I literally decided to apply for graduate school a few months before the [laugh] applications were due. Part of that was like if you're a major in biochemistry, and you're not going to medical school, what do you do? I had no idea. Nobody in my family had ever gone in the sciences. I was like, "I guess you go to graduate school." What else are you going to do? [laugh] I applied to graduate schools. Fortunately, I got in to a really good graduate school. That's what set me on the path. I went to Berkeley. I had the good fortune of getting in a really amazing lab. I got in the lab of this guy, Randy Schekman, who went out to win a Nobel Prize. It was a good environment. [laugh]
ZIERLER: Was he doing the work that led to the Nobel Prize when you were there?
ZIERLER: What was the work?
DESHAIES: Before I got there, they had isolated a bunch of mutants that block secretion from yeast cells. But Randy himself is a real biochemist, and the isolation of mutants was really a means to an end. It was to understand the molecular basis of protein secretion. While I was there, a very key development happened. While I was there, there was a graduate student, David Baker, who's gone on to become incredibly famous. He's the founder of the Institute for Protein Design at University of Washington. He is the leader in protein design. Steve might not be happy to hear that, but I think Steve would acknowledge that.
ZIERLER: Steve Mayo, you mean?
DESHAIES: Yeah, because they were competitors early in their career. David had reconstituted trafficking between what's called the endoplasmic reticulum and the Golgi, the first step in secretion of proteins from the cell. When proteins are made, they're inserted into the ER membrane, then they're transferred to the Golgi, and then from the Golgi to the cell surface, where they're released from the cell. David had reconstituted that first step in a test tube, not with purified proteins but with crude extracts. But that assay was the basis for Randy's lab going on to figure out how to get that first step to work with purified proteins. That took another 12 years, I think [laugh], to get that to work with purified proteins. But it was that sequence of events that gave him the Nobel Prize. The isolation of the mutants, which enabled all the subsequent research, happened before I got there. But the reconstitution, which was then the key to driving the molecular unraveling of the process, happened while I was there. But early on while I was there I really had no conception of what the career possibilities were, what the trajectories were. I just had no idea. I went to graduate school very naïve, and with significant opposition within my family, actually. I had a brother who was working at NIH at the time. He was a finance guy, and he had a friend who was a biochemist who had left bench science to become a program manager at NIH, who was very negative about science like, "Oh, it's competitive. It's impossible to get a job. It's impossible to get grants," just negative. My brother was like, "I don't understand why you're doing this. You should go to medical school. This is ridiculous. What's this bullshit? You're going to get a PhD in biochemistry? This friend of mine is telling me it's a loser's game. [laugh] There's no future in it." He talked my parents into this. My parents were not educated, and they didn't know because they were just like, "But Roger's telling us"—
DESHAIES: —"that you should [laugh] go to medical school." I went to Berkeley. I figured, look, it's a top graduate program, and so it can't all be bad, and I'll figure it out. Either I'm going to like it and I'll figure it out, or I'm not going to like it and I'll do something else.
ZIERLER: Was biotechnology a career path at that point?
DESHAIES: No. I wasn't even aware of the existence of biotechnology when I went there. Even though by '83, when I went to Berkeley, Amgen had already been formed, Genentech had already been formed, I was pretty much unaware of the existence of these things.
ZIERLER: It's basically professor or bust is your purview?
DESHAIES: Yeah. When I got to Berkeley, I then became aware that there were biotechnology companies, because some people from Berkeley, postdocs were going into biotechnology companies. But, at that time, the people who went into biotechnology companies were perceived as the losers. "These are the people who couldn't get an academic job, so they go to biotech." There was a real jaundiced view of like, "The really good people go on to become professors, and the people who aren't so good"—I was affected by that. I never once thought of going into the biotechnology industry. Once I realized that I was good enough, which I probably didn't realize until I was in my second or third year of graduate school—at least my second year—I could measure myself up against other people in the department, and I'd see postdocs going on and getting jobs in academia. That's when I realized there's a path here. There's an endgame, and I could play this game. I'm competitive in this game. By the end of second year, it was very clear to me that I wanted to go that traditional academic route. My goal was fixed. My mind was fixed on that.
ZIERLER: How did you develop your thesis topic?
DESHAIES: [laugh] In a state of complete inebriation.
DESHAIES: No doubt I was stoned and inebriated simultaneously. I was dating my wife, who's just off in the other room somewhere on the first floor. We were both graduate students. We knew each other as undergrads. She was a year younger. She had come out to Berkeley a year after me. It was my second year, her first year, and she was doing a rotation. You do these rotations as first-year graduate student. She was rotating in Schekman's lab. She ended up doing her PhD in Schekman's lab as well. It was the fall, it was Halloween, and we were dating. We had started dating right when she arrived at Berkeley. We'd gone to this Halloween party together, and we left the Halloween party together, and I think we were going to stay at her place. I forget where we were staying. I had another project at the time. I was technically working on the reconstitution step that David Baker ended up getting to go. That was technically my project. I was following up on the work of a German postdoc in the lab who had some data where he thought he was reconstituting ER to Golgi transport but, in retrospect, it probably wasn't real. I had had a series of mishaps that fall. Number one, I had to serve as a teaching assistant, which took a lot of time. Number two, my parents had visited, and so I was probably out of the lab for a week or two or for at least a week when they were visiting. Then I got really sick. I got flu, and I remember I must have lost like 15 pounds, like 10% of my body weight, at the time. I got very sick. I remember it was terrible. I remember at one point, I literally crapped my pants, and I was at work. What do you do [laugh] when you're at work, and you've crapped your pants? I didn't have spare underwear. I had a really bad fall semester. That would've been 1984. We go to this party, Halloween, and no doubt I had drank and smoked. Then I'm leaving with Linda, and she announces, "I got to go to the lab, and inoculate some cultures." [laugh] I'm like, "Oh, great." [laugh]
DESHAIES: Then she's like, "It's only going to be like five minutes." Of course, five minutes turns into 40 minutes. We go to the lab. It's late. They had a break room. I'm like, "What do I do?" It's clear it's going to take way longer than five minutes. I started thumbing through some journals. There had been a graduate student in the lab who had thought she had isolated a mutant in protein translocation, how proteins start the process by being inserted into the ER membrane, through a protein channel. That protein channel had been posited by Günter Blobel in his famous ‘signal hypothesis'. There was a signal at the N terminus of proteins (the ‘signal sequence') that guides them to this channel and across the membrane, and that's how they begin the process of secretion, like for an antibody. Günter went on to win a Nobel Prize for his signal hypothesis. But he had identified by that point what's called signal recognition particle, which is the thing that recognizes the signal sequence. But that's in the cytosol, and it binds the nascent-polypeptide as it's being translated by the ribosome, and it takes it to the ER membrane. But then nobody had identified the actual channel of the ER membrane.
This graduate student that Randy had before I got there, she thought she had a mutant in that channel. But while I was in Randy's lab, there was a French postdoc who showed that her results were misinterpreted, and this mutant was not in the channel; it was actually in the apparatus that puts sugar chains on proteins as they're going across the membrane. My rotation had involved investigating that problem, so I was interested in that problem. I'm thumbing through a journal, and she's now gone on to do a postdoc, and there was a paper from her postdoc in that journal issue that I was thumbing through. That paper was the trigger for me to think about her PhD work, which was this effort to find mutants in the early secretory pathway, and how they thought they had a mutant in the channel, but it turns out it was misinterpreted. That got me to thinking, gee, couldn't one look more directly for the mutants that she thought she had? Her way of looking for it wasn't all that direct. They thought they had this kind of mutant, and they were all excited about it, and it turns out they were wrong, "they" being her and Randy. But it was a happenstance how they got there, so why couldn't one just really explicitly look for that kind of mutant, and design a screen explicitly to find that kind of mutant?
I conceptualized an approach for how you would do that, and I wrote it down on a piece of paper. Randy left the key to his office in the lab, and I go and I unlock his office door, and I leave the piece of paper on his desk. Then I proceed not to hear anything from him except, at some point later on, he had expressed his extreme unhappiness with my productivity. [laugh] We had a very difficult conversation in which he expressed his unhappiness, and I dug my heels in because, again, I had had a very difficult fall semester, and there was a good reason I wasn't productive. He says nothing about this piece of paper I'd left on his desk, for weeks. Then in December, mid-December, so this is a month and a half later, he goes to a seminar, and the seminar triggers an idea in him that reminds him of this note I left on his desk. It suggested a practical way to achieve what I was describing. Then he summons me into his office, and I'm like, "Oh, shit, he's going to drop the ax on me." But he's all excited. I'm like [laugh], "What the hell's going on?" We were not having a good relationship at this time. [laugh] He said, "I went to this seminar, and they were describing da-da-da, and then I remembered your idea, and that this could be a way of practically applying your idea." [laugh] I'm trying to compute all of this. I said, "Why are you telling me all this [laugh]?" He said, "Do you want to do it? I said, "Now, seriously?" He's like, "Yeah." I was like, "Starting today?" [laugh] He's like, "Yeah." I said, "Done." I walked out of his office.
DESHAIES: I started the project. That idea ended up being successful. I literally identified the channel protein that Blobel had posited in his signal hypothesis. I didn't win a Nobel Prize for it though. [laugh] But that's how I started my PhD project.
ZIERLER: What did the lab look like where you worked, the instrumentation, the research atmosphere? What was it like?
DESHAIES: The instrumentation was funny because, when you think in retrospect, when I started there, it was 1984. Randy had started in '76, so he was a good eight years in. But he still had all this used equipment that he inherited from prior labs. There'd been prior labs that he had inherited the space of. He was not super well-funded in the early part of his career, and so the lab had a lot of shitty, old, crappy equipment, and was very parsimonious, because he had gone through his first years without a lot of money. He had started the project. He had no experience in yeast. NIH grant people didn't think he could do what he was proposing to do. It was very frugal, but the environment was phenomenal. He had great people in the lab. There were some really smart, creative people in the lab. But, more importantly, I think, he fostered an atmosphere that really encouraged people to develop ideas. Randy didn't feel like he needed to intellectually dominate the lab. I think that was really important. He was very secure. He had a strong sense of self-worth, and was very secure in his skin. He didn't feel a need to be overbearing, and dominate the lab. It was a real freewheeling intellectual environment. It was a very stimulating environment to be in. I thought it was a great place to develop as a scientist.
ZIERLER: What did you see as your contributions with this research?
DESHAIES: Two things. Number one, just on a purely technical level, I discovered core components of the eukaryotic translocation machinery that we know are conserved all the way to human. I discovered an important piece of eukaryotic cell biology, how proteins cross the ER membrane. I discovered that piece of machinery. Secondly, it grew to be, for a significant period of time, one of the focal areas of the lab. I was the first person to start the project. By the time I left, there were three or four people working on that: myself, David Feldheim, Sylvia Sanders, Colin Stirling, Jonathan Rothblatt. It was like a little subgroup in the laboratory. I felt like that project had a pretty big impact on the lab during those years, and it continued for some time after I left. I don't think he's still working in that area, but he worked on it for quite a while, and I was the one who really launched that.
ZIERLER: What were some of the postdocs you were looking at?
DESHAIES: Gosh, I no longer remember all the ones. I didn't look at too many postdocs, in part, because there was a constraint. Linda and I were still together, and we decided to look for postdocs together, so that put a constraint on where we could look.
ZIERLER: The Bay Area was definitely a focus?
DESHAIES: The Bay Area was totally a focus. I got interested while I was in my graduate work in cell cycle, cell division control, and that area really started taking off while I was wrapping up my time at Berkeley. I left Berkeley in 1990, and '88–'89 were years that whole thing started taking off. Paul Nurse, the paper that won him the Nobel Prize would've been like 1988. I was really interested in Marc Kirschner's work across the bay at San Francisco.
ZIERLER: What was he doing?
DESHAIES: He had been working for years on trying to identify what was called MPF, maturation promotion factor, which was an activity that you can inject into Xenopus oocytes, and it would drive them into meiosis. We know, in retrospect, that it's a cyclin-CDK complex that drives mitosis and meiosis. But, at the time, it was just cytosol that you inject into an oocyte, and it causes it to enter the next phase of the cell cycle. I was really interested in that. They had reconstituted the cell cycle in a Xenopus egg extract, crushed eggs, just pure cytoplasm. They could show the extract cycled through interphase and mitosis. They had shown that you need a cyclin for this, and the cyclin had to be degraded. If it wasn't degraded, the extract got stuck in mitosis. That connection between cell cycle and degradation ended up being the focus of my research. At the time, I didn't know it was going to become the focus of my research. But that connection between cell cycle and protein degradation by the ubiquitin pathway had emerged in Mark's lab while I was in his lab, and it was something I was really excited about.
ZIERLER: What aspects of your graduate research did you see as a continuation in the postdoc, and what were the really new directions for you?
DESHAIES: I continued working in the same system but using completely different methods and on a completely different problem. When I went to Mark's lab, I went with the goal of reconstituting the transition from the G1 phase of the cell cycle to the S phase of the cell cycle in extracts of budding yeast. Budding yeast is what I worked on in Randy's lab, Saccharomyces cerevisiae. In Randy's lab, I primarily used genetics to discover the machinery involved in protein insertion into the ER membrane. In Mark's lab, I switched problems, not protein biogenesis in the secretory pathway but cell cycle control, and switched methods, primarily using biochemical reconstitution in extracts, as opposed to genetics and molecular genetics. But I stayed with the same organism: budding yeast. This is no doubt influenced by the success of David Baker in reconstituting that early step in yeast secretion in Randy's lab.
What I could see was you could merge genetics and biochemistry, because what David did to set up his system is he used mutants. Once he had reconstituted the process with wild type cell lysates, he could show that if he made lysate from a mutant—the first mutant he looked at was sec23—that lysate didn't work. There was a graduate student, Linda Hicke, who was working on that gene, and they showed together that if you used the sec23 mutant, the reaction wouldn't work. That automatically created an n-1 situation where you had a reaction that was specifically defective in one factor, and then you could use that as a recipient to purify that factor. I saw the power of that, that if you could reconstitute the process, and then show it was defective in a mutant, you could use that as an assay to purify the functional gene product. And if you have a functional assay for your gene product, then you could work out the mechanism. These were all principles I learned in Randy's lab. Mark is also a biochemist. But this provided a path where you could go from genetics, which tells you this particular gene is important for entering into S phase, but now you have an assay that allows you to purify the active protein encoded by that gene. It's a functional assay, so it allows you to work out mechanism.
ZIERLER: Did Mark and Randy collaborate? Did they work together ever?
DESHAIES: No. They knew each other, of course. But they worked on such different problems.
ZIERLER: What do you see as the most important things you've accomplished during your postdoc?
DESHAIES: [laugh] I largely felt my postdoc was a failure. But what I did do is I established an extract system that I was able to exploit later. The core reaction in the G1/S transition in yeast, in all organisms, is activation of a cyclin-CDK complex, a G1-cyclin, G1-CDK complex that drives that transition. I went there to reconstitute that, and I succeeded in doing that. Mark's lab was doing this in frog extracts, frog egg extracts, looking at the cell cycle transition from G2 to mitosis, where it's a mitotic cyclin, mitotic CDK enzyme that catalyzes the transition. What they had shown is that that process is complex. There's multiple layers of post-translational regulation during that process, multiple different phosphorylation-dephosphorylation events that are occurring on the CDK enzyme that are triggered by the cyclin. Normally, the transition into mitosis is regulated, but it's not nearly as much of a watershed event as the transition into S phase. The transition into S phase is regulated by a much more complex set of signals. The entry to mitosis is more straightforward, arguably. I felt that, if there's all this post-translational regulatory complexity for the entry to mitosis, it's going to be that multiplied by five for the transition from G1 to S phase. If I can reconstitute this, I'm going to have a bonanza of things that I can study. That was completely wrong. There is indeed a lot of regulation but most of it is transcriptional in yeast. It's not post-translational covalent modification; it's transcriptional regulation.
ZIERLER: You're finding this out in real time, or you're retrospectively thinking back?
DESHAIES: I'm finding this out in real time in two ways. Number one, while I was underway with reconstituting the process—and that took a long time. It took me over a year to reconstitute the process. It took me nearly two years to reconstitute it—Fred Cross at Rockefeller was showing that there was this complex transcriptional feedback loop where cyclins stimulated their own transcription. The key thing is when you activate a CDK, you get an abrupt activation. It's like a step function. In frog going into mitosis in frog egg, that step function is driven by post-translational phosphorylation-dephosphorylation. What Fred showed was that in yeast, that step function was driven by positive feedback and transcription.
ZIERLER: Which tells you what?
DESHAIES: Which tells you that you don't necessarily need post-transcriptional controls or post-translational controls. You're getting the step function from transcription alone.
ZIERLER: This is an efficiency?
DESHAIES: It didn't preclude that there would be post-translational controls, but it suggested you might not need them. I was aware of Fred's data. But I was committed to the project, so I kept working on it. When you add cyclin to a frog egg extract in G2, you see two phenomena. You had a threshold. You keep adding cyclin, nothing, nothing, nothing, nothing. Then all of a sudden, Boom, CDK turns on and the extract enters mitosis. It wasn't that you add cyclin, and you had a linear response of CDK activity. If it was simple, you'd expect one unit of cyclin, one unit of CDK. Two units of cyclin gives you two units of CDK. You'd expect a linear relationship. No. There the curve went like this. Da-da-da-da, nothing, nothing, nothing, nothing, then all of a sudden, everything. There was a threshold phenomenon. That's the first thing that suggested something complicated is going on. There's a threshold, and it's a step function. The second thing was a lag phase. You'd expect, oh, you add cyclin, and then you have a somewhat linear time course of activation of CDK. There's twice as much activity at two minutes as at one minute da-da-da-da-da. You have a linear time-dependent activation. But in frog egg extract, there was a lag phase. No activity, no activity, no activity, no activity, no activity, then all of a sudden boom, activity. There was a step function to concentration, and a step function to time. That told you that there's a switch. It's like a light switch. It's not a graded response like a volume knob. It's a light switch. It's on-off. To get on-off, you need complex post-translational events that create that switch-like behavior. At the time, none of this was known. All that was known is you had this complex behavior. How it worked was not known. Now that's known. When I made my yeast extracts, I used genetics to make them cyclin-deficient, and there is a CDK in the extract. I added G1 cyclin to the extract. The CDK turned on. But there was no lag. No threshold. The amount of activity I got was strictly proportional to the amount of G1 cyclin I added, and the amount of activation was strictly proportional to the amount of time. There was no threshold. There was no lag. Everything was just perfectly linear. Then the killer experiment, the absolute killer is I didn't even have to have ATP. I could just add the cyclin. No ATP, so no energy requirement, and the CDK would just activate.
All I was observing was literally the binding of the cyclin to the CDK. I was literally just measuring binding of cyclin to CDK. The CDK was already in a state where it was ready to be active. It was already phosphorylated. In frog egg extract, the phosphorylation that activates it doesn't happen until the CDK binds. The switch-like behavior comes from a phosphorylation-dephosphorylation cycle at an inhibitory residue that when you phosphorylate it, it inhibits the CDK, and then you have to dephosphorylate it. In frog, you had an inhibitory and an activating phosphorylation going on after you added the cyclin. In yeast, nothing. It was just like, literally, it's already phosphorylated. There's no inhibitory phosphorylation. All that has to happen is the cyclin binds to CDK. I realized I've been working on this for two years, and I have nothing, literally nothing.
ZIERLER: But you're confirming something as well. The nothing is significant in and of itself. It wasn't known before.
DESHAIES: Yes, that wasn't known before. The whole point of a postdoc is to have a project that has legs that you can go and start your own lab, and you have phenomena that you can investigate. I had no phenomena. I showed there was nothing interesting going on post-translationally in the yeast G1/S transition, literally nothing interesting. All I was measuring was cyclin biding to the CDK. I could publish a paper saying, "Oh, guess what? Yeast G1 to S transition is really different from frog egg G2 to mitosis transition." [laugh] But this was not going to get anybody's pulse racing. I went through a period where I was depressed. It's fair to say I went through a period—probably one of the few periods in my life when I was depressed.
I had tried different ways of adding cyclin to the extract. The first way was making it in bacteria, but it didn't express well in bacteria. It was all broken down. Then I had tried making it in retic lysate where you make lysates from reticulocytes, which are immature red blood cells, and they'll translate mRNA that you add to the lysates. I would add the G1 cyclin mRNA, and it was translated into the G1 cyclin protein. I would add the G1 cyclin to my cyclin-deficient yeast extract. It worked. It would turn on the CDK. But the key thing there is when you translate in retic lysate, the protein is radioactive. You label it with [35S] methionine. That's how you follow the translation of the mRNA. The key thing there is then I could observe what happened to the cyclin when I added it to the yeast extract. I observed two things. Number one, it got phosphorylated. Upon electrophoresis on an SDS-polyacrylamide gel, it shifted in molecular weight due to extensive phosphorylation. But then after it got phosphorylated, I could see high molecular weight smears, and those high molecular weight smears are caused by the building of these lollipop chains of ubiquitin on the cyclin, because the extent of modification of each molecule of G1 cyclin is heterogeneous. Some molecules get six, some molecules get seven, some molecules get two chains. Some molecules only get one. Some molecules might get three chains. Because of this heterogeneous modification with ubiquitin the cyclin goes from being a single band on the gel to a smear. We knew from yeast genetics that the next step after you turn the CDK on was that there was an E2 enzyme—that's an enzyme that transfers ubiquitin—called CDC34, that was required to enter into S phase [Ed. Note: yeast proteins are referred to in non-italicized upper case.] But nobody knew why you needed that. While I was a postdoc, this guy in Vienna, Kim Nasmyth, an amazing scientist, published just an incredible paper, which argued that you needed CDC34 to antagonize an inhibitor of the cyclin-CDK complex that drives S phase. You have a G1 cyclin-CDK, which is what I was working with, and then immediately downstream you have an S phase cyclin-CDK, which triggers initiation of DNA replication. Kim showed that there's an inhibitor of S phase cyclin-CDK called SIC1 that keeps the S phase cyclin-CDK off, but then SIC1 disappears, and that allows the S phase cyclin-CDK to turn on and activate DNA replication.
The disappearance required CDC34, which encodes an enzyme that transfers ubiquitin [Ed. Note: wild type yeast genes are denoted in uppercase and italicized, whereas mutant genes are lowercase and italicized.] He showed it also required two other genes that were known to be required for entering into S phase: CDC4 and CDC53. But he didn't show, and he didn't know what was triggering all of this. Why does SIC1 all of a sudden get degraded? What's regulating the temporal nature of that? Is the role of CDC34 direct, and what are these other proteins, CDC4 and CDC53 doing, because they're also required for SIC1 to go away? It was just genetics, pure molecular genetics. I then immediately looked at my reconstituted reaction in which I observed attachment of ubiquitin to G1 cyclin. Now, I'm not looking at SIC1, I'm looking at G1 cyclin, but conjugation of ubiquitin to SIC1 required CDC34. I can make a lysate from that mutant. The cdc34 lysate was defective. It didn't make these smears. I could make CDC34 in bacteria add it to the lysate and the smears come back. Lysate from cdc4 cells had the same property: no smears. cdc53 had the same property: no smears. All of these are required to conjugate ubiquitin to, in that case, the G1 cyclin. But then when I got to Caltech, I showed the same thing was true for SIC1, which, as Nasymth showed, is the key target that explains the genetic requirement for these factors to promote entry into S phase. It just so happens that they're also promoting G1 cyclin degradation at the same time. That sort of reaffirmed the whole reason that I did my project, the ability to reconstitute a process, and then use the mutants to create an n-1 situation where you have a functional requirement for a gene product, you can make that gene product in bacteria, add it back, and reanimate the lysate that's defective. Except I thought I was going to be using my reconstituted system to study G1 CDK activation by G1 cyclin. What I instead ended up using it to study was G1 CDK-dependent activation of ubiquitin transfer, because it turns out that the signal that turns on this ubiquitin transfer is phosphorylation by G1 CDK. I said earlier that when I added G1 cyclin to the lysate, it shifted by phosphorylation. It turns out that the G1 CDK is phosphorylating the G1 cyclin that is bound to it. That generates a signal. That signal is bound by CDC4, which is a component of the ubiquitin ligase enzyme. CDC53 is also a component of that enzyme. They recruit CDC34, which is the carrier for ubiquitin to transfer it to the substrate.
The same thing is true for SIC1. The G1 CDK puts a phosphate on SIC1. We figured that out at Caltech. That's a signal for SIC1 to bind CDC4. We also figured that out at Caltech. Then CDC34 is recruited, and transfers ubiquitin onto SIC1, and then SIC1 is degraded. We reconstituted that entire process in a 2001 Molecular Cell paper . Based on the work I did as a postdoc, when I got to Caltech, what we did is we started expressing CDC4 and CDC53. You couldn't make them in E. coli. You had to make them with baculoviruses, which is an insect cell system. We didn't know, but we were missing two components. The first one was discovered by another lab while I was at Caltech. It's called SKP1, and it's in between CDC53 and CDC4. It links them together. CDC4 binds the substrate when it has a phosphate on it. SKP1 binds CDC53 and recruits it to the complex. CDC53 then binds another protein that we didn't know about but we discovered when I was at Caltech that is now called RBX1 . RBX1 recruits CDC34, the ubiquitin-transferring enzyme. The genetics identified three of the five components. It didn't get SKP1, and it didn't get RBX1. That whole thing came together while I was at Caltech. The first full reconstitution reported in 1997 was actually missing RBX1, which is the thing that recruits the ubiquitin-transferring enzyme. But, unbeknownst to us, when we made the yeast CDC53 protein in insect cells, it was recruiting the insect cell RBX1.
The homology is so high that the yeast proteins could co-assemble with the insect cell protein to make the active complex. We didn't figure that out until a few years later when we realized there was another protein in there. Then when we added the yeast RBX1, the activity went up like an order of magnitude. What we were seeing, which we thought was great in 1997, ended up being pathetic [laugh] once we had all the yeast proteins together. But that was the first of what are now called cullin-RING ubiquitin ligases. CDC53 is now known as CUL1. There's CUL1, CUL2, CUL3, CUL4A, CUL4B, CUL5, CUL7. Each one of them has a different adapter and a series of substrate receptors. The CDC4 part in human, there's 69 different proteins that can fill that role. In work my lab did at the end of my time at Caltech, we did a census of all those CUL1 complexes, each with a different receptor, and we quantified them, and we showed their relative abundance, and we showed how they exchange, like, how do you exchange those receptors out . The signal is substrate. If you have a substrate, that stabilizes a complex. If you don't have a substrate, the complex is unstable, and the receptor gets exchanged out. It's a really elegant system – the composition, the repertoire composition is determined by the substrates that are present in the cell . Each different cell could have a different composition of CUL1-based ligases, depending on the substrates that that cell possesses, because the substrates determine the stability of the complex.
ZIERLER: Ray, to go back to the idea that the concern as a postdoc, the research is a failure because you have nothing to build on for a lab as a faculty member, clearly, Caltech took issue with that self-assessment. Externally, obviously others did not feel the same as you.
DESHAIES: I was depressed. I went through that period of depression a year before I was looking for a job.
ZIERLER: You feel like you turned things around?
DESHAIES: I turned things around in my last year as a postdoc. When I was looking for a job at Caltech, I knew I had this assay where I could see the CDC34 dependence of ubiquitin conjugation on G1 cyclin, and I think I also had CDC4 dependence. I had a system. I had an assay to monitor transfer of ubiquitin onto a substrate. What I was selling, if you will, when I was applying to Caltech, was I was going to use this system to unravel the mechanism of these ubiquitin transfers. I was going to use that assay to deconstruct in a reductionist way how this whole thing worked.
ZIERLER: Did you have any interaction with Alex Varshavsky before Caltech?
ZIERLER: What about since? How big is the ubiquitin community?
DESHAIES: Still minimal. I get along with Alex. I literally have an email from him this morning. He just nominated me for a Prize.
DESHAIES: Alex and I get along, but Alex—
ZIERLER: He operates in his own world?
DESHAIES: Alex is a lone wolf. He's not a collaborator. He operates in his own world, and that's how he wants it. There are some scientists who really like to collaborate. Alex is not one of those scientists. He wants to be doing his own thing. He wants complete creative control over what he does. I interacted with Alex, but we never collaborated. We interacted. We had similar interests. He would send me his manuscripts. I would send him my manuscripts. We communicated, and we are aware of what each other was doing. We have a lot of respect for each other. We always got along, but we didn't have deep interaction in a collaborative sense.
ZIERLER: Now, were you on the job market widely? Did Caltech recruit you, and curtail that?
DESHAIES: [laugh] There's a funny story behind that. [laugh] I ended up at Caltech almost by accident. I'd applied to about a dozen jobs. There was a guy working in the same room as me in Kirschner's lab at UCSF, Tim Stearns, who's now I think dean of the graduate program at Rockefeller. He was chairman of biology at Stanford until a few years ago. It was ridiculous, in retrospect. We were the hot commodities on the job market that year. I had secured an offer at MIT, at Berkeley, at Princeton, and there were multiple others if I pursued them. I'd interviewed at Duke, but then I didn't really pursue it. But I had offers at those places. Once you have an offer at MIT and Berkeley, Princeton [laugh], how much more do you need to look for? Tim had secured an offer at Yale, Harvard, Stanford, Caltech, Princeton. Princeton was making multiple offers. We both had an offer at Princeton. Tim did a little better than I did in the job market but rightfully so because I had no papers. This is unusual. When I was on the job market, I had zero papers from my postdoc.
ZIERLER: Because it's only that last year, and that's when you would've been in the middle of writing papers?
DESHAIES: No. I wrote the papers when I got to Caltech.
ZIERLER: Oh, wow. OK.
DESHAIES: I was still finishing up the research. I published my two postdoc papers while I was at Caltech. I had zero papers. It is not advisable to go on the market when you have zero papers. But I had to go on the market for a variety of reasons. Tim had two papers, both in Cell, which is the hottest journal. Tim was a much better prospect than I was. He did a little bit better on the job market, but the irony is I got the two jobs that he perhaps wanted the most [laugh]: MIT and Berkeley. He was interested in those jobs, because he had been a graduate student at MIT, and one of his best friends was a professor at Berkeley, and they would've had labs next door to each other. Of the jobs he had in hand, the one he wanted the most was Caltech; between Yale and Harvard and Stanford, he really wanted the Caltech job. His wife drew a line in the sand and refused to move to Caltech. She didn't want to move to Southern California. They were in the Bay Area. She's like, "Take the Stanford job." He held out until the end, because he kept hoping a job at Berkeley would come through, because Berkeley was going to make two offers. But they ended up making the second offer to somebody else on the specious argument that it would be intellectual—
DESHAIES: Canalization, meaning, how ideas can be guided into narrow channels, with the resulting concern being that you're taking two people from the Kirschner lab. It would be a huge mistake because it's like groupthink. The second offer went to a different person. When that happened Tim took the Stanford job. Meanwhile, I'm trying to decide between Berkeley and MIT. Linda was also looking for a job. She was making a big career transition from being a researcher. She didn't want to run a research lab. She wanted to go into undergraduate teaching as a professor at a four-year college. The problem is, as a trainee at a place like Berkeley, and then she did her postdoc at Stanford, you don't get teaching experience. You TA two courses during your five years as a PhD student. That's it. You don't do any teaching as a postdoc. But she was applying for these teaching jobs.
Normally in that field, what people do is they'll go and they'll become like a sabbatical replacement professor or an adjunct or whatever to build up a teaching résumé. She hadn't done any of that. She's looking for a teaching job with literally no credential other than TA-ing five years earlier. She was having difficulty getting traction. What was happening was the places where I had offers were making her offers to come in like, "You can run our lab courses or something like that." But these were contingent offers based on the fact that they were making me an offer. She wasn't going to have an independent tenure track faculty position. These are subordinate positions, and that was much, much less appealing to her. At the 11th hour, she gets an interview from University of Redlands, which is out in Redlands. They bring her down. I think she did two visits in a short period of time. They make her an offer. I have no offer in Southern California. I have nothing. Tim has the offer at Caltech, and I'm trying to decide between MIT and Berkeley.
ZIERLER: It's almost like you could swap. [laugh]
DESHAIES: Then what happens is the thing unravels at Berkeley for Tim, and he says no to Caltech. It must have been days after my wife gets the offer at Redlands. Caltech calls me up after Tim turns them down, and the irony is Mel Simon is calling the same phone number for Tim and me but doesn't realize it. Mel Simon's trying to tell me a bullshit story. He's not telling me the truth, obviously, which is that they were trying to get Tim, and Tim turned them down, so now they were turning to me. He's telling me some bullshit story. I'm like, "Mel, it's OK. Did you recognize the number you just called is the same number you called a few hours ago?"
DESHAIES: There's silence on the phone. Then he's clearly really embarrassed. I was like, "No." Tim and I got along really well, and I respected him tremendously, and it was obvious to me that he had done better as a postdoc. He had two Cell papers, and he had a really exciting problem. I had promising-looking stuff, but Tim was more—
ZIERLER: A much more proven entity?
DESHAIES: He was more of an obvious choice than I was at that point. Also, Tim was bringing a whole new research area to Caltech—microtubules—and I was overlapping with Alex. I already knew Bill Dunphy because I was working on cell cycle and ubiquitin, so I wasn't going to bring a new research area to Caltech. I was like, "Mel, it's fine. I understand. If I were you guys, I would've made the offer to Tim." Literally, I come down to Caltech, I think, three or four days after that conversation on the phone, I got the offer within about a week.
ZIERLER: Because Mel was insistent or someone else came in to make the case for you?
DESHAIES: No, because I think when they were deciding between Tim and me, we were the top two people on the search, and there was a lot of debate about which one to make the offer to. Tim came out ahead of me, but the faculty had already decided that both Tim and I crossed the bar for what they were looking for but that Tim was the more obvious first choice. They only had one position, so they went after Tim. The minute Tim turned it down, literally within hours, in the same day, they pirouetted, and approached me. This had already been predetermined. They had predetermined if we don't get Tim, we're going to go after Ray. The minute Tim turned it down, they came after me within hours. I was primed to accept because my wife had just gotten the offer at Redlands.
ZIERLER: This was an easy yes for you?
DESHAIES: I preferred to go to Berkeley or MIT, but this was hard to argue with, because the Caltech offer was a very generous offer. It was the most generous offer I had, and it enabled both my wife and I to have independent positions doing what we wanted to do.
ZIERLER: The generosity of the offer, were you aware of Caltech's reputation of really supporting junior faculty?
DESHAIES: Yes. I knew that from Tim. Tim and I had talked about our offers. Tim was raving about his Caltech offer because it was the most physical space. It was the biggest dollar start-up. It was the highest salary. All the quantitative metrics, it was his best offer, so he was like, "I want to take that Caltech offer." [laugh] But his wife was like, "I'm not moving to Southern California."
ZIERLER: What was the game plan for you when you got here, setting up the lab?
DESHAIES: What do you mean, what was the game plan?
ZIERLER: How do you set it up? Where do you start?
DESHAIES: My first year was horrible. It was miserable. [laugh] It was terrible. That was my second phase of depression. That first year here was very difficult. UCSF was very collaborative. I got here, and I quickly realized that the two people whose research was most closely related—Bill Dunphy and Alex—neither one of them was eager to collaborate. I envisioned having joint group meetings. While Alex was happy to interact with me, he didn't want to do joint group meetings. It quickly became clear to me that I was going to have to completely build my own intellectual environment for my lab. I wasn't going to be able to do this jointly with other labs.
ZIERLER: But isn't part of the tenure success story demonstrating that you could stand on your own two feet?
DESHAIES: Yeah. But you could stand on your own two feet with—I was doing different problems than those guys, so it wasn't like I was going to be relying on them. When you start a lab, you don't have anybody. It's a little more stimulating if you have an intellectual environment that you're operating in. It's not a necessity. I had been in a place where that intellectual environment was vibrant, especially in cell cycle research. There were multiple labs. There was a big joint meeting between all the labs. It was just radically different. That was hard. But then the other hard thing, I had difficulty recruiting people when I got here. In my first year, the first people I recruited to the lab, they were challenging. I didn't have what I felt was a serious and productive environment.
ZIERLER: You also have papers that you need to write, on top of everything else.
DESHAIES: I'm writing grants. I'm writing papers. I had a bunch of rotation students, some of whom were better than others. I didn't have a superstar, go-getter person at the beginning. I showed up in January, and literally the first 12 months, there's not much going on. No progress. In fact, there's regression, because I had a freezer, an upright -80 freezer, and the door had not been properly latched, and it thawed.
ZIERLER: Oh no…
DESHAIES: All the extracts I had made at UCSF were ruined.
ZIERLER: Oh no!
DESHAIES: I had months' worth of extracts, all the different mutants and stuff for reconstituting the reactions I told you about. All gone. All no good. Purified proteins all gone. Everything gone. It was a fucking disaster. That happened probably in August. Then I spent the Fall in the lab, remaking all those extracts.
ZIERLER: Is there an opportunity to do it better second time around—
ZIERLER: —or you're just replicating it?
DESHAIES: Just remaking them. Then the real turnaround point came. Judy Campbell had a senior postdoc in her lab, Rati Verma, and that was Rati's second postdoc. She had done a first postdoc with Art Pardee at Dana-Farber, had come out here, was in a postdoc with Judy, and had been with Judy like five years. She had foregone, for her husband's benefit, taking an academic position a couple years earlier that she had been offered, I think, in Texas. She was trapped, a senior person, and she asked to join my lab. [laugh] Initially, I went to Mel, asking for more money, and Mel counseled me against it and, I think, completely rightfully because, as he said, these things never turn out well. These things just never turn out well. But this one turned out great.
Despite having made those compromises, and being stuck in a dead end, she had not lost her passion. I offered her a job. I had been at Caltech 12 months. I started in January '94, and she either came in December of '94 or January of '95. She just hit the ground running. A biochemist. I made these extracts, and she then took the extracts, and started doing experiments literally from day one. I'd gone from having no data in my first year, nothing, no data, to all of a sudden data was coming out. I was feeding her. I was making reagents, extracts, I was making proteins, and I was feeding them to her, and she was doing the experiments. She was mature, and she had a work ethic [laugh], which my lab was lacking. She single-handedly helped turn the whole environment in the lab around, because I couldn't be in the lab all the time. I had other stuff. I was having to do a lot of writing. I just had other things that I had to do. That's when the lab really started turning around, and it just built from there. I didn't have my first paper until August of '97 . I had been at Caltech over three and a half years before my first paper, because I started January '94.
ZIERLER: Was your three-year review dicey because of that?
DESHAIES: No, I don't think so. [laugh] Mel was exhorting me to publish papers. I clearly remember that. [laugh] Every time I would see Mal—he was chairman—he would go, "Publish papers. Publish papers." [laugh]
ZIERLER: [laugh] Because he was looking out for your tenure prospects?
DESHAIES: Right, exactly. I think he was starting to get worried because it's three and a half years in, and I don't have any papers. But my first year was a complete wash. Just nothing literally got accomplished in the first year. Then what happened, in October of that year—I'd been at Caltech by then three years and 10 months—is I published two papers within a week of each other, one in Science  and one in Cell . Those are two home runs. All of a sudden, I went from nowhere to doing well. From there, things progressively got better, and that's when my career really started taking off. That Cell paper was describing the progenitor of this cullin-RING ligase family, and so that was a big paper, and one of my most highly cited papers to this day.
ZIERLER: When things start to take off for you, where do you orient entrepreneurialism, translational interests? Are these dual-tracked? Are they intersecting? What's your sense of the narrative?
DESHAIES: I'm not thinking about translation at all, zero. My first idea that relates to something translational was the PROTAC idea I hatched with Craig while we're at that Burroughs Wellcome meeting. But I'm not sure. I can't put myself in my shoes of 1998. But I'm not sure that even when it was hatched that it was hatched from a wellspring of entrepreneurialism. I've always been interested in methodology, and I think it was more coming from that wellspring of, gee, wouldn't it be cool if you could bring a substrate to ubiquitin ligase at will to trigger its degradation. I wasn't necessarily pragmatically thinking that this would be a route to therapy. For example, I wasn't out there thinking about starting a company around it. That was really Craig that was motivating that. [laugh] I have a funny thing in my office at work. It was a card my wife had given me at the time, and it shows two bulldogs on chairs that are next to each other. One bulldog is just looking straight ahead with a sense of purpose, and the other bulldog is looking towards the first bulldog like [laugh], "Wah."
DESHAIES: The bulldog looking straight ahead is saying, "How does the G1/S transition work?" The other bulldog looking at the first bulldog is [laugh] saying, "Do you love me?" [laugh]
ZIERLER: [laugh] That's great.
DESHAIES: I think that card accurately captures [laugh] where my head was at at the time. I was very focused on trying to understand this research problem that I was building my lab around. Translation did not exist as a thought.
ZIERLER: Tell me how the Proteome Exploration Laboratory got started.
DESHAIES: As I said, I was always into methodology. One thing I was proud of—and that's, I think, the wellspring the PROTACs came out of—one thing I was proud of is that I was an early adopter of technology. I mentioned the Science paper in 1997, and I mentioned that the signal for SIC1 getting ubiquitin conjugated to it, and getting degraded was attachment of phosphate, and the phosphates are then recognized by CDC4. How did we figure that out? I collaborated with these guys Steve Carr and Roland Annan, who were at GSK. They were some of the first people to use mass spectrometry to map phosphorylation sites in proteins. Up to that point in time, when people were mapping phosphorylation sites in proteins with mass spec, they were doing it, initially, in recombinant proteins, like some protein made in E. coli, and then you'd throw a bucketload of kinase at it, and it gets phosphorylated. Then you can use mass spectrometry to map those phosphorylations. That's highly unphysiological. Technically, you can do it but BFD. People at the time were primarily using this other methodology involving two-dimensional separation of peptides by chromatography to map phosphorylation sites. We were working out the biology of the cullin-RING ligase. It was called SCF-CDC4: "S" for SKP1; "C" for CDC53; "F" for F-box—that's the substrate receptor; and CDC4 is the particular F-box substrate receptor that was being used. We were trying to understand the biology of how the SIC1 gets phosphorylated, and then how that serves as a recognition signal for CDC4 to come in and bind it and ubiquitylate it and thereby trigger its degradation.
To try to prove the model, we wanted to map the phosphorylation sites, and then mutate them, and show that the SIC1 would no longer be degraded, and you'd no longer enter S phase, because you need to degrade the SIC1 to enter S phase. That was the model. I contacted Steve & Roland. Now, this was not proteins purified out of bacteria. We mapped the phosphorylation sites of SIC1 from yeast in a physiological situation where SIC1 was restraining the G1/S transition. My god, we mapped 12 phosphorylation sites with Steve & Roland on this real protein—not some bogus thing made in E. coli—and we mutated those sites, and we showed that that molecule was no longer ubiquitylated, it was no longer degraded, and it blocked the G1/S transition. That was the Science paper. That was a big deal. That was the first time anybody had ever used mass spectrometry mapping of phosphorylation sites to solve a real biological problem. I was very proud of that work. I then got the mass spec bug, I was bit by the mass spec bug, and I entered into another collaboration with the same person. This involved Rati Verma again. She led the Science paper, and then a couple of years later—
ZIERLER: Was she like a staff scientist at this point?
DESHAIES: Yeah, I ended up making her a staff scientist. She worked for me until I left Caltech, and now she's at Amgen. She followed me to Amgen.
ZIERLER: Oh, wow.
DESHAIES: There was a guy at University of Washington who then moved to Scripps, John Yates. We had done this other mass spec stuff for a while with this guy Andrej Shevchenko in Dresden, at the Max Planck Institute in Dresden. He was the best person in the world. If you had a gel that you stained with silver, and you had bands on it that were unknown proteins, he could cut them out, and sequence them by mass spec, and it was more sensitive than the other methods at the time, like Edman degradation, which was the classic way you would do this. The mass spec was way more sensitive. Shevchenko made an important discovery for us using that methodology of cutting bands out of a silver stain gel, and that was a Science paper we had in 2001 on a key regulatory factor for cullin-RING ligases, a discovery of a key regulatory factor . But, meanwhile, we started collaborating, Rati and I, with this guy at University of Washington, John Yates. He had this crazy idea, which is you don't run a gel. You just take the immunoprecipitation that you would separate on a gel, and you throw the whole thing into a mass spectrometer; the entire thing. No separation. You could have 30 proteins in there. It doesn't matter. It would sequence them all. I'd given a seminar up there, around 1997, and he had told me about this, and I'm like, "Wow, this is phenomenal," because when you run a gel, you have to extract the bands from the gel. You always lose most of the material. If you don't have to run a gel, your sensitivity goes up dramatically. This guy told me, "I could purify ribosomes. I can inject the purified ribosomes in. I identified every single subunit of the ribosome." I'm like, "Holy shit, this is amazing." Rati had worked out a way to purify proteosome by affinity purification using epitope tags. You pull it out of yeast, and you run a gel, and you have all the bands that represent the different subunits of the proteasome. This guy at Harvard, Dan Finley, had been doing this for years, and he had systematically just finished the year before. He had just finished identifying all the proteins, and it had literally taken him five or six years to do this, to systematically purify enough proteosome the old-fashioned way, and to cut bands out of gels, and do Edman degradation. He had already published this, so there was nothing new here. But we showed that in literally two days of work, not only could we recapitulate the last several years of what he had done, but we identified two new proteins .
Literally, Rati in one afternoon, affinity purified proteasome with an epitope tag, not having to go through conventional purification, and then we give the IP to John, and he runs it in mass spec, and he identifies every single subunit, including two that Dan did not identify: two new subunits of the proteosome. I'm like, "This is the cat's pajamas. This band cutting shit is out the window." [laugh] Then I became obsessed with doing this, what's called multidimensional mass spectrometry, where you no longer do separations on gels. All the separation is happening chromatographically on an HPLC column, and then you're feeding directly into a mass spectrometer, and you could identify all the proteins in a complex sample. Barbara Wold then stepped in and really helped out. She had been asked—gosh, I'm going to forget the detail. She had access to some pool of grant money, for technology development and omics, because Barbara's always been doing genomic kind of stuff, and proteomics would be covered under this. She had heard about what we were doing, and she said, "I've access to this pool of capital." We did a grant together, Barbara and I, and we got the money to buy our own mass spectrometer for doing this. The way we set this up so that we could do it is I recruited a graduate student who wanted to do this. The new mass spectrometer was sent to John Yates's lab in San Diego, the guy we had done the collaboration with on the proteosome, and my student went down there for a year, and he learned how to run the machine. Then once the year was up, the mass spectrometer was shipped to Caltech, and Johannes Graumann, my student, came back to Caltech. Then he started doing this work in my lab. What we did is I think we picked 20 or 30 proteins of interest in the cell cycle, and we tagged them all, and purified them all, and did this multidimensional mass spec on them all. We published a paper . Basically, the message of the paper was multidimensional mass spectrometry for the masses. This technology had gotten to a point where it was no longer a specialist technology. You could use this in a basic molecular cell biology lab like my lab. I was not collaborating with John Yates to do this any longer. I wasn't going to a mass spec facility. I was doing this in my lab and. I was making the argument that this is the way forward. This is ready for prime time.
ZIERLER: Is that related to your initial ideas about a start-up, and where does the proteome lab fit in?
DESHAIES: No. This is completely independent. Johannes was doing samples for other people in the department, and it was clear there was a lot of interest, particularly Erin Schuman's lab, who's now at the Max Planck in Frankfurt. She was very interested, and started collaborating with us. Even though we had done this in my lab, and I made a big deal of that, I realized that long-term, number one, I couldn't be a service provider for the department. Number two, they were coming out with new mass specs every two years, and they were doubling in price. While Barbara and I could get this mass spec ~ 2000 for a quarter million dollars, five, six years later, the entry fee was like a million for an Orbitrap. This is like a whole NIH grant. I realized to really do this, you need to buy expensive instruments that are beyond the means of an individual lab. Gordon and Betty Moore had made their gift to Caltech, and Elliot was chairman, and Elliot—
DESHAIES: Yeah. Elliot offered to me, he said, "We can put this up for Gordon and Betty Moore," because Caltech was running these internal programs. He said, "Why don't you put in for a facility at Caltech, that would do this high-end mass spec, make it broadly available in the community?." I wrote the grant, the grant got awarded, I hired Sonja Hess out of NIH to run it, and then that proceeded to become the Proteome Exploration Laboratory, which was broadly used. Erin Schuman's lab used it a lot, my lab, Kai Zinn's lab, a lot of labs. Mitch Guttman made one of his big early discoveries at that facility. A number of other labs, including Dave Tirrell's lab in Chemistry, used it quite a bit. Different labs over the years used it, but it really grew out of this work in my lab, because I was the first lab at Caltech to start doing multidimensional mass spectrometry. I think the PROTACs came out of that same interest of technology development. I always had a little bit of interest in technology development.
ZIERLER: Where does the start-up story begin?
DESHAIES: The start-up story begins when Craig and I were collaborating on the PROTACs [5-7]. We had started in '98. I'd gotten a grant, the Gates Grubstake grant that I mentioned previously. Kathy Sakamoto, my MD graduate student, was working on PROTACs. Craig was making them, Kathy was assaying them, showing their effects in lysates and cells. Meanwhile, as I mentioned earlier, Craig had epoxomicin, and he showed it was a proteosome inhibitor. We had discovered Rpn11—Rati had—as a new activity in the proteosome that was essential for the proteosome, a new target to inhibit . Craig had just called me up one day, and said, "Why don't we start a company? We can propose three different things. We could propose epoxomicin as a new proteosome inhibitor, based on the success of the Millennium compound in the clinic. We think ours is better than Millennium's compound. You have Rpn11, a new target in the proteosome, and then we have the PROTACs, which is this whole new way of drugging targets. Why don't we pull this all together into a proposal to start a company?" That's when I was like, "Er," and I went to Scott. I went to Tech Transfer, and I met up with Scott Carter. Scott Carter was like, "This is great. I'll help you guys out." I was like, "I don't know what to do. How do you do this? I have no idea."
ZIERLER: To go back to the earlier part of the conversation about Caltech's evolution, how was this regarded by your colleagues? Was this normal at this point? Were you doing something that was perceived as out there or different?
DESHAIES: I don't think my colleagues were really aware. The people who would've known about it were more my peers, the people I was interacting a lot with anyway. This wasn't the kind of stuff that I would be talking about typically when I would make my presentations at like the Friday lunch faculty meetings every year. You'd talk about something going on in your lab.
ZIERLER: Because you saw it as separate from your bread and butter at Caltech?
DESHAIES: Yeah, as I said, I had 16 or 17 people in my lab at the time. I have this one person, Kathy, doing PROTACs. David Baltimore was on her committee, and David was always a little bit more entrepreneurial anyway. But I think most people at Caltech would not have been aware. When I launched Proteolix, I think most people at Caltech wouldn't even know that I had done that. Still, most people at Caltech wouldn't know that there's a drug on the market that's selling over a billion dollars a year that came out of that company. People at Caltech are not following that sort of stuff. They don't know what's going on in the pharma industry.
ZIERLER: Did that plant a seed for you, to set the story?
DESHAIES: Here's another thing that's interesting. It would be interesting to get Fred Farina's take on this. I felt that when I started Proteolix, we pursued a fundamentally different model than what was normally happening at Caltech. What was normally happening at Caltech was people were raising—and this is going back now to the early 2000s—people were launching companies on a shoestring. A lot of the companies coming out of the non-biology parts of Caltech, you don't need a lot of capital, because the costs are not high. If you're developing a piece of software or even a device, you don't have the regulatory environment that you have with the FDA. Even in chemistry, Bob Grubbs, who started Materia that's just off San Gabriel Boulevard, there's nothing regulated about that. They were making materials to make golf clubs and shit. The amount of money, the capital you need is just different. When people at Caltech were starting companies, like, there was the Athenaeum Fund affiliated with Baldeschwieler, who was a chemistry professor. You give people a quarter million dollars to start a company. We went a completely different way, with Scott's help, and actually somewhat contrarian, because Yale was in the same ‘start small' mindset. You start with a few hundred thousand dollars. You rent a little lab space next door to the professor. The postdocs from the lab become the first employees. It's all very rinky-dink. That was my impression of a lot of what the companies coming out of Caltech were at the time. It was rinky-dink. We went by a very different model. I should note that an early effort to found Proteolix, which was spearheaded by Yale's technology transfer department, was unsuccessful. After that, we went out, and we were like, we need $15 million. I had a disupte with Caltech's Tech Transfer because they wanted a certain percentage of the company. I forget what percentage it was, like 5%. I'm like, "Dudes, you can have 5% of your $300,000 company, but you're not getting 5% of a $15 million company. I'm sorry, it's just not going to work that way." But that was the model. The model was people did these things small, but we came with a very different approach right out of the box of saying, "We want big dollars. We have a molecule we think we can turn into a clinical candidate, and take into the clinic, and it's going to cost money." We asked for $15 million. We got 18 in the end.
For perspective, in the recent past of 2020-2022, $15 million is small potatoes. Many startups raised north of $50 million. But in 2003, on the heels of the 2001 Nasdaq crash, $18 million was a princely sum for a preclinical play like Proteolix. Maybe when Lee was there, he had done stuff on a bigger scale. I have no idea. But in the interval, in the 10 years preceding our launch of Proteolix, I was not aware of Caltech companies launching with that scale of A-round money, right out of the block. We never did the Baldeschwieler Athenaeum Fund thing, the quarter million dollar financing. A couple of VCs offered us like a million dollars to demonstrate proof of concept with our molecule. Hire a CRO to do some work. We said, "No thanks". We wanted $15 million out of the blocks, and we got it. It'd be interesting to go back in the books of Tech Transfer. But, as far as I was aware, that was not common at that time.
ZIERLER: Do you think that experience planted a seed for ultimately your decision to leave Caltech?
DESHAIES: Oh, totally. I didn't leave Caltech because I wanted to leave Caltech. I would've happily stayed. I came close to leaving Caltech in 2010. I'd been recruited by Novartis to run a site they have in San Diego called GNF, Genomics Foundation of Novartis. That whole thing unraveled at the last minute.
ZIERLER: You were prepared to go?
ZIERLER: You were prepared to go?
DESHAIES: The salary offer was on the table. We were talking about how much they would help with housing. We were negotiating details of how much money they would put into my lab to help wind it down. We were all prepared to go. I almost went in 2010. Then when that didn't happen—and it didn't happen for complicated reasons that I can get into—but when that didn't happen, I assumed that I was going to finish my career at Caltech. Part of it was my wife really liked her job. She's in a very good environment. It's a very collegial department. People in those kind of jobs, like the college teaching jobs, they're not mobile. It's not like academics at a place like Caltech, because your reputation in a place like Caltech is national and international.
Your reputation in a four-year teaching college is local, and people just don't move around. Not only do people not move around that much, but she had no desire to move because she was very happy in her role. Those jobs are very difficult to get, and she didn't want to leave it because the chances are you never find another one. There's all these people in that environment that are permanent adjuncts. They never have a tenured position. She was happy in her role. She didn't want to leave her job. That always put a constraint on. If I took the Novartis thing, she was going to leave the job, but that would've been difficult. In the end, it didn't happen, and so I figured I was happy. I was an HHMI investigator. I was very well supported. The lab was going well, and I figured I'm going to finish my career at Caltech. I was very happy, despite the problems I had early in my career, my first year. Amgen reached out to me in December 2016. I was getting contacted frequently for leadership positions, either in academia or in industry.
ZIERLER: Like Deans and VPs type of thing?
DESHAIES: Yeah. I was getting at least a few of those a year, those kinds of inquiries. Now I'm getting those weekly. But back then, I wasn't getting them weekly, but I was getting them with reasonable frequency, a few times a year. I'd started two companies. I started another company in 2011. I think especially among people in industry, it was known that I'd started a company that was successful. That developed a drug that made it to market. You get a reputation for being somebody who's entrepreneurially minded but also a successful entrepreneur.
ZIERLER: What was the second company?
DESHAIES: It's called Cleave Therapeutics. They still exist, but they struggled. We put a molecule in the clinic, and the molecule had a toxic effect, a completely unanticipated toxic effect. The company then collapsed for a while but never went away. We brought it back. I found a path to bring it back that involved a public-private partnership with NIH, where NIH funded to get a second-generation molecule ready for the clinic. We put that molecule in the clinic, but we're still not seeing a clinical benefit. We haven't hit the dose that might be required for that. But the company's running out of money. Even though they dialed out some of the toxicity we saw with the first molecule, they're starting to see early signs of that toxicity again, and so there's a concern that even though we dialed some of it out, maybe we didn't dial enough of it out. We may not be able to have a therapeutic index. The likelihood is it's not going to make it. But it's still around. It still exists. They're in the clinic. We're in phase 1. But when I got approached in December 2016 it was not by a headhunter – usually it's by a headhunter. But this was directly by Amgen. The person who would become my boss, who I didn't know at the time, actually called me personally, and asked me to think about taking the job. This was Sean Harper, the executive vice president of R&D.
ZIERLER: This is what your current job is, same job?
DESHAIES: The job I have now is bigger, so the job expanded.
ZIERLER: You're still in your original role, it's just expanded its purview?
DESHAIES: Yeah. The role grew in scope. For me at the time, I had already been through this process six or seven years earlier with Novartis. Some of the things I was concerned about with Novartis were obviated here.
ZIERLER: Is Amgen more academic just from its founding, its origin story?
DESHAIES: No. But I did feel that Amgen would be a better fit, because Amgen was more biotech, and Novartis is more old line, small molecule drugs. I felt like Amgen, in that way, felt like a better fit, because it's more biotech.
ZIERLER: It's also local. Bingo! [laugh]
DESHAIES: Bingo. That's number two. But number three, which was very important, one of the biggest worries I had about Novartis was I was going to be running a 500-person site, and my boss was going to be on the other side of the United States in Boston, and everybody at that site was going to be subordinate to me. While I had been involved in starting two biotechs—
ZIERLER: This is serious management?
DESHAIES: This is serious management, and this is serious drug development. I'm not a drug developer. I'm still not a drug developer. I've never been hands-on in drug development. I've been watching from the sidelines, but that's not the same thing. I had a hard time wrapping my head around like what do I do on day one? I was trying to envision, OK, I show up at Novartis. I'm in charge of this 500-person site. What do I do? What's the first thing I do? Will I get up in front of 500 people, and I have to say something? What do I say? I just was like, "I'm going to have nobody, and my boss is on the other side of the country, and he's got a huge operation he's overseeing. He's not going to be able to hold my hand. Who's going to help me in this job? Who's going to mentor me? I'm going to need mentorship. Where am I going to get that?" I was really worried about that. I was really worried about who is going to help enable me to be successful? Plus, I was taking over in the footsteps of the person who my future boss just fired – – Pete Schultz, who's the president of Scripps, TSRI, in San Diego. Pete's kind of a larger-than-life figure. The folks in senior roles at Novartis in La Jolla were his acolytes.
I'm going into this situation where everybody directly under me has loyalty to Pete Schultz, who's this larger-than-life figure. I have no experience in drug development. I have nobody to mentor me. I was worried this was not setting me up for success. I felt very differently about Amgen, because I was going to be a member on the senior leadership team from R&D where all the other people were going to be the same rank as me, and they're all experienced drug developers. In fact, the guy who's now my boss, David Reese, had been running a unit that was between research and development, called translational sciences. But when my predecessor left, Reese had to take over the research part, the discovery part. My predecessor was Alexander (Sasha) Kamb, son of former Caltech Provost Barclay Kamb. Sasha left Amgen and his role was temporarily taken over by David Reese until I arrived. Reese was like chomping at the bit for somebody to come and relieve him of that responsibility. I also knew that he would be somebody that I could rely on to help mentor me. He had come from academia as well. He's an MD. He had come from UCLA about 12 years earlier. He had a lot of experience in industry. In addition I didn't have to move, and it was an environment where I would have peers that I could learn from and get mentorship from, and it just felt right. It felt better. When Sean called me up, I'm like, "No, not interested. I'm happy at Caltech. I got Howard Hughes Medical Institute (HHMI) funding. I just got renewed a year and a half ago. Why would I do this, honestly?" He persuaded me to at least think about it.
ZIERLER: This is a big pay boost too, I assume.
DESHAIES: [laugh] Yeah.
ZIERLER: Orders of magnitude? [laugh]
DESHAIES: Not an order of magnitude, but it's a lot. Yeah. It's a lot. It's a huge pay boost not so much in base salary but because I get bonus and stock. Now, compared to what I was making at Caltech when I left, probably if I factor everything in, it's seven or eight times more. It's a lot. That was a factor too, because I had just had a brother die and another brother diagnosed with metastatic melanoma, who I thought was probably going to die. It turns out he's still alive, but I thought at the time, "He's not going to make it. He's got metastatic melanoma." I'd up to then thought I was going to live forever, because my dad died at 99. But then after having one brother die at 65, another one with metastatic melanoma and—
ZIERLER: Time have to put some money in the bank?
DESHAIES: I want to put some money in the bank so that I can step down when I'm still young and in good health. I used to think I'm going to live forever. Now, I'm not thinking I'm going to live forever anymore. All those things factored in, including the fact that I didn't have to move. Sean had called me on the recommendation of this guy, Robert Tjian, who had been the president of HHMI. But I knew Tij since 1983. He was on my thesis committee at Berkeley. I immediately called him up, and he encouraged me. The other one was David Baltimore, who was on the board of directors of Amgen. Of course, when Tjian put my name in the ring, when he had mentioned me to Sean, and Sean started doing due diligence, Sean went to David and asked about me. Then after Sean had approached me, I went to Tjian, and then I also went to David, just to sound them out, and they were both encouraging. I thought about it, and I guess where I landed with it is I was 55, and I had just gotten renewed by HHMI, but HHMI is also famous for cutting people loose later in their careers. The reality is I'd been through three renewals, but even people like Pamela Bjorkman got cut. Paul Sternberg got cut. It's not like they were over the hill not doing anything. I have to be realistic. The likelihood is, next time, I might not get renewed. If I don't get renewed, I'm going to have to let go of half my lab. I was getting direct funding of about a million a year. It's not easy to fill a million a year hole in your funding. That was one. Second is, I had taken my research program pretty far down the line. There was more that I could do, but it was clear that it was probably more diminishing returns, and it was going in directions that were probably more suited to people with different skill sets.
ZIERLER: There's also a moral dimension about how well you can serve graduate students and postdocs?
DESHAIES: It wasn't that so much; it was more because grad students and postdocs will stop coming, so there's not really a moral dimension there.
ZIERLER: The market will define that for you?
DESHAIES: Yeah. It's more that you'll become less and less relevant, and you'll attract less and less qualified people. I felt if I stayed on, I would probably eventually have to shift research focus. There was a lot of interesting stuff going on in the lab, and I felt like we're still doing really good stuff. I was still at the top of my game in the ubiquitin field. I was getting invited to all the good meetings. I was probably considered one of the top four or five people in the field, and I was one of those people. I was at the top of my game. I had really good funding. I had really cool projects going on. I was happy with the lab. But I felt like it's going to be hard to maintain that position. It's not easy to stay in that position. I had to acknowledge that, eventually, I would probably slip a little bit from that position.
ZIERLER: Go out on top?
DESHAIES: Right. I can go out on top, number one, because if I stay, what's going to change? I'm going to have more students. I'm going to have more postdocs. I'm going to publish more papers. I'm going to get more grants…I'm going to do all the things I've already done. I know the game. I know the routine, and the routine's not going to change. I wasn't that interested in becoming a chairman. I could do a leadership position but, fuck, if I'm in a leadership position, I might as well go and make real money. I might as well be in a real leadership position in industry, because I wasn't that interested in leadership. I liked running the lab. Being a chairman or being a dean or being a provost…if I'm going to do that, I might as well go do that where I'm not herding cats, like, I have real power and I have real resources. I figured, why not? I knew there was a risk, a significant risk, because how am I going to scale? I'm running a lab of 15 people. At the time, my responsibility at Amgen was going to be 750 people. Now it's about 1,100 people. I can't scale leadership from 15 to 750. It's a fundamentally different ball game. I was acknowledging that it might not work, that this might fail, and then how was I going to cope with that? Amgen helped in the respect that my severance package was two years of salary and bonus. Since I felt that even if I was a disaster I would last one year, I looked at it as if they were guaranteeing me three years of salary and bonus, so if anything happened in the first year or two, I would have three full years of salary and bonus. That really diminishes the risk because if I'm going to flame out, it's probably going to happen relatively early. Or if I really hate it, it's going to happen relatively early.
ZIERLER: Although, in fairness, you flamed out relatively early as a postdoc. You flamed out relatively early as a faculty member.
ZIERLER: Maybe that's not exactly the best extrapolation.
DESHAIES: Yeah. [laugh] I figured if it didn't work out, number one, I had a financial cushion. When Proteolix got bought by Onyx, I didn't make it all in that day because there were milestones that paid out over time. But the milestones that eventually paid out, I realized, that's 20 years of my Caltech salary at that time. I have a level of financial comfort from the purchase of Proteolix that allows me to take a risk. That was number one. Number two, Amgen – we had negotiated that generous severance arrangement where I felt like, no matter what, I got three years of salary and bonus, which is multiples of what I would make a Caltech in those three years. Third, I figured if things didn't work out, my marketability, if anything, would only increase because now I had some industry experience, and the NIH was moving in the direction of translation. It was clear there was a lot of energy around translational science, and Caltech was interested in translational science. I figured there's actually a pretty damn good chance I could land back at Caltech if things didn't work out. In fact, Caltech has come after me multiple times to run the Merkin. It just seemed like—
ZIERLER: Go for it.
DESHAIES: Go for it. As I said earlier, that whole intersection between science and business interested me. It's very different working with peers because when you're running a lab, you spend most of your time with trainees who are less knowledgeable than you are. Now I spend most of my time with peers who, in their domains, are more knowledgeable than I am. It's a very different feel to it. I had gotten a dose of that in the start-up experience, and I found that stimulating, interacting with people more at a peer level. Not that I didn't find it stimulating interacting with trainees, but it's very different. It's very different interacting with people that you could learn from as opposed to people that you're teaching. I figured I'd learn a lot. Learning is always interesting. The hardest part, by far, was winding down, because you can't just walk out. I had a lab of 13 or 14 people. That was the hardest part.
ZIERLER: But you're not staying on until your very last student defense? You can maintain that mentorship?
DESHAIES: Right. What I did is I negotiated an arrangement with Amgen whereby they put about $2 million into Caltech, because the HHMI money would just stop almost immediately. The NIH money, which was a small part of my budget, would keep going. But the HHMI money would stop on a dime. Caltech would allow people to stay in that space for up to two years.
ZIERLER: This backfills the shortfall from HHMI?
DESHAIES: Right. I had a plan, and I knew I could take people with me to Amgen. There were, in fact, five people that followed me to Amgen to finish their postdocs, one of whom is still there. The other ones have all moved on and gotten jobs of one sort or another. I reasoned that I could make this work because I was still going to be living in Pasadena. The first year, I was going in once a week to meet with my people at Caltech that were still working there. It gave me a path to wind things down in a way that maintained integrity for my people. That was by far the most difficult aspect of it, how do you go from running a lab to not running a lab without really harming the people that you've recruited?
ZIERLER: Ray, for the last part of our talk, you've already laid out all of the factors that compelled you to leave Caltech. There was a known entity. In the years since you've been at Amgen, the opposite side of that, what's worked out in terms of all of the pull factors that validated that decision?
DESHAIES: It's funny because when I try to think of it in this sense, it's always less obvious. I find, on a day-to-day level, I feel fulfilled in the sense that there's always a lot going on. The job is very transactional and, because there's always a lot going on, it makes it very stimulating. I like the transactional piece. I feel like I have an impact, and it's a bigger impact. When you're in a place like Caltech, you have an impact on your lab, but I didn't really feel like I was having an impact on the institution. I had an impact on my lab, and I had an impact in my field. I would say that's where I felt I had the impact: in my field. I could look at things like the multidimensional mass spec, and I was the early adapter of mapping phosphorylation sites. Then those kinds of things spread. Cullin-RING ligases became their own little subfield, and now PROTACs have become their own little subfield. I felt like I had an impact in my field, not so much in my institution, pretty much not at all in my institution. I never felt like I had that much of an impact at Caltech. At Amgen, I definitely feel like I have an impact. I have a different way of looking at things. I've had an impact at the company. I feel like I've learned a lot. It's very stimulating. It's exciting when we bring drugs forward, when we get positive clinical results. Although, on a personal level, those achievements are far removed from me. It's not—
ZIERLER: You're part of a system that makes that happen?
DESHAIES: Yeah, and I'm an important part of the system, but it's not as personal. But it's still fun. That part is still fun, but it's not as personal in the same way as it is when you're running a lab, and your lab makes a discovery. There was something I was thinking of, and then I forgot what it was.
ZIERLER: Either because it's professionally relevant or for your enjoyment, are you keeping up with the literature?
ZIERLER: It's just not feasible?
DESHAIES: Just not feasible. I built a group at Amgen—but I don't run it. I recruited a guy to run it—that's going after PROTACs, so having the opportunity to do that, to build up this capability at Amgen. I guess the most satisfaction I get from Amgen is the ability to influence a $150 billion corporation. Being able to have an influence on a $150 billion corporation, that can be satisfying. I don't do the earnings calls, the quarterly earnings calls. My boss does those, and Bob Bradway. But hearing Bob Bradway or my boss say things on an earnings call that I know they got from me, that's very satisfying. It's very satisfying to look and say, this is not a Nature paper. This is one of the biggest employers in Southern California. This is a 24,000-person corporation, $150 billion market cap, and I'm having a discernible impact. Bob is getting up in an earnings call, and talking about initiatives that I launched, and that's very satisfying.
ZIERLER: This goes back to The Wall Street Journal front page feeling that you got?
DESHAIES: Yeah. It's the impact thing. This is something I would explain to people at Caltech, and I think it's hard to understand. I would say, "I have more impact and less autonomy."
ZIERLER: Yeah. [laugh]
ZIERLER: Yeah. That's well put. [laugh]
DESHAIES: I can't walk across Amgen campus without multiple people saying, "Hi, Ray," and I have no idea who they are—no idea. In a corporation of 24,000 people, I am very visible because, number one, even though the company sells drugs and runs clinical trials, the stuff that they put out there is the discovery piece. When they're trying to get people excited, they have me do a podcast, and all this stuff. Discovery research is a big driver of media for them. They really emphasize that. They put that front and center for the media. That's what they believe excites people more than details of clinical trial results. I'm the face of that, so I am very well known within Amgen. Like I said, I can't walk across campus, without people saying, "Hi, Ray," and I have no idea who they are. It's always a little bit embarrassing because I don't know what to do. I could walk across Caltech campus, and nobody knows who I am, when I was a professor there. If I walk by a chemistry faculty member, they know who I am. But faculty members in other divisions – most of the faculty who were not in biology or chemistry had no idea who I was. Students and postdocs in those departments, undergrads, they had no idea who I was. My impact and visibility within Caltech was much lower than my impact within Amgen. But I could walk in tomorrow at Caltech as a faculty member, and say, "I'm tired of doing ubiquitin, and I want to work on sleep," and I could do it. It probably wouldn't be smart to do that, but you could. You could in theory. I could walk in on any given day, and just do whatever the hell I wanted to do. I most certainly can't do that at Amgen.
ZIERLER: There's no rediscovering yourself in a corporate environment? [laugh]
DESHAIES: I can't walk into Amgen, and say, "You know what? I'm tired of oncology."
DESHAIES: "We're stopping all our discovery oncology programs." I couldn't do that.
ZIERLER: Is the Caltech professor thing part of your professional persona at Amgen? Is that something that you lean into?
DESHAIES: Absolutely. It allows me to get away with certain stuff.
DESHAIES: I'm very non-corporate. I get made fun of a little bit. I'll make fun of Bob Bradway from the stage. Nobody at Amgen does that. I'm the only person in the entire corporation who does that. There are people of higher rank than me. I'm not an EVP; I'm an SVP. But I'll make fun of Bob Bradway.
DESHAIES: I've shown slides with cartoon depictions of him on the slide. Nobody at Amgen does that sort thing.
DESHAIES: I have a reputation. My boss will often make a sardonic comment about trying to manage me. "At least you don't have to manage Ray," that sort of thing. I am still very much under wraps, but compared to the normal ‘under wraps' at Amgen, I am a little bit more freestyle. That is widely known and acknowledged and commented on. To some extent, I'm given a pass because it's like—
ZIERLER: Caltech professor.
DESHAIES: —he's a Caltech professor, and what do you expect?
ZIERLER: Obviously it works for you and it works for the company.
ZIERLER: If I may, for my last question, let's look to the future. Professors, they have that emeritus thing that they could look to when they say, "I can shed all of my admin responsibilities. I don't have to teach. I can just focus on the research." Is there an emeritus path for you at Amgen, or what does that look like, because you're still going to be a scientist even after you retire?
DESHAIES: That's a good question, and I'm actually trying to figure out the answer to that myself. I've announced my intention to retire at Amgen, and I've given them a date. I've communicated to them that after I step down from my major administrative responsibility, I would be interested in staying on for some time in some sort of advisory-like consulting role. I don't know if they're going to take me up on that or not. People in my role, what do they do when they step down? They become members of scientific advisory boards, or boards of directors of small biotech companies, or they go consult for venture capitalists. I am in constant communication with venture capitalists. Multiple of them know that I intend to leave a direct line management role.
ZIERLER: Maybe there's another company in the cards for you.
DESHAIES: I don't want to go into another operational role. I've made that clear.
ZIERLER: But certainly, in an advisory capacity for one of these things?
DESHAIES: I flew up to San Francisco a couple weeks ago for a dinner that Google Ventures organized around me. This was in part because they know I'm thinking of stepping down and, when I step down, I'm thinking of becoming involved with venture capital. Versant arranges a call with me every quarter for the very same reason. There's multiple venture capital firms I'm interacting with because they all ask. They're constantly coming after me. They want me to run a portfolio company, and I'm like, "No, not doing that." But I said I am potentially interested in having a relationship with venture capital when I step down. Whether I maintain a relationship with Amgen, whether I get involved with venture capital, or with start-up biotechnology companies, I'm interested in remaining intellectually active. The way I frame it to people is I want a job where if I want to go trekking in the Himalaya for a month, I can do it. The kind of job I have now, you can't do that.
DESHAIES: I want a job where it frees me up to travel. Yes, I'll have occasional responsibilities, but they're not super time-sensitive, and the world won't stop if I'm not there to sign a requisition or whatever. Whatever way that turns out, that's what I'm thinking, something where I can use my knowledge of basic science, my knowledge of industry, in a productive way.
ZIERLER: I said last question, but this is really going to be the last question. Not keeping up with the literature, losing the muscle memory of being in the lab day in and day out, have you made your peace with the fact that the emeritus status means you can hang out in the lab? Are you OK with that?
DESHAIES: OK with not going back to the lab?
DESHAIES: Oh, yeah. Going back to the lab just seems so implausible to me. Define "being in the lab," because when I was a postdoc, everybody was like, "You're going to miss the lab when you're a faculty member." I did my last experiment in 2000. I tried. I kept doing stuff in the lab. What I found is the graph would be like when I start at Caltech, I'm in the lab here, and then it just went like this. By 2000, it got to zero.
ZIERLER: That trajectory was already set even before all of this?
DESHAIES: I haven't been in lab in 23 years. But I was still very much engaged with experimental design, experimental interpretation, writing grants, reading papers, all that stuff. Now I'm not even doing that. But I realized I'm not going to go back in the lab like that again. That just doesn't seem plausible. When I left the lab, I remember thinking I loved working in the lab. I'm good with my hands. I do a lot of repair work around the house. I'm handy, and I like working with my hands. I liked working in the lab, and I was good. I was a good experimentalist. I was a very good experimentalist. But I was a little bit obsessive about it. I realized, as a postdoc, I can't do this my whole life. I'm too obsessive about it. It'll ruin me. I can't do this my whole life, because I really couldn't control myself. I would just keep pushing to get the result. There was that. The second thing is I realized I could get more done when I became a faculty member by thinking, and then getting other people to execute—and this was hard in the first few years, because I knew I was better than all the people under me. I was just a better experimentalist. No, there was one person who came to my lab after I was there three years, who I felt was as good or if not better than I was as an experimentalist [laugh].
At the end of the day, by ~2000 I figured I'm going to get more done by not doing experiments and working through other people. Now it's just that at a much higher level. I'm going to get more done by setting a course for this organization to go in a certain direction than I could get done by trying to manage three projects or whatever. If I went into venture capital, arguably, I can get more done by helping to launch a biotech that developed a new medicine than I could by trying to manage three postdocs or whatever. The other thing is it's really hard to go back. Going back to Caltech is out the question, because you have to build up grant support, and you got to recruit people. [laugh] Starting that in my '60s, that's a pretty big hill to climb. I have fantasized at times, but it's really a fantasy. I could go to Amgen and say, "Hey, I could go from being the senior vice president of global research to running a little lab with two or three people." But even that, do I really want to find a research problem, and spin up a research area, and recruit people, hire people? [laugh] Then it's just going to go back into that whole obsessive thing. I'm not going to be able to go and manage it.
ZIERLER: If you do that, you're not going to be able to turn off and go to the Himalayas for a month.
ZIERLER: You'll figure it out. [laugh]
DESHAIES: I've done a lot of things in science. I've made discoveries. I've started companies that have put drugs on the market. I've led large organizations. But there's more to life than that, and I don't want that to be my whole life. But the one thing I find, especially when I was at Caltech—and that's one thing I like about the job now—the job now is just not personal in the same way, because when I leave, Amgen's just going to put somebody in my place. I could walk away. They'll put somebody in my place. Five years from now, they'll have forgotten about me. It's not personal.
ZIERLER: That's liberating to you, in a way?
DESHAIES: It's liberating. As a PI at Caltech, the wonderful thing about it but the burden of it is it's really personal. It's your lab. When you leave, they're not going to put somebody in your place, running the lab.
DESHAIES: The lab is gone.
ZIERLER: That's right.
DESHAIES: It's your lab, and it becomes your identity. This is what's hard to explain to people. People at Amgen just don't get this. I'm like, "People retire? People at Caltech don't retire."
ZIERLER: Exactly. [laugh]
DESHAIES: "They work into their '80s," and they're like, "What?" [laugh] I'm like, "No, but it's not like you're working for Caltech."
ZIERLER: Yeah. It's you.
DESHAIES: It's like you're working for Amgen. It's not like that. It's your thing. It's you, it's your identity. That's simultaneously one of the biggest positives and, I always felt, one of the biggest negatives because it becomes very personal. Somebody scoops you, and it's very personal, and it's devastating. It's why people don't stop working. I was like, "That's not how I see the latter half of my life playing out."
ZIERLER: You got to experience both perspectives.
DESHAIES: I always remember one of the things that really struck me is when Frederick Sanger died, they had an obituary in Nature. Until this year, Frederick—well, no, because Marie Curie had two as well. I think there's only three people in history that have won two Nobel Prizes in science. I think Marie Curie was one, and Frederick Sanger was the second, and now—what's his name?—Barry Sharpless, the guy from Scripps who just won this past year. The obituary said Fred Sanger, number one, he didn't have trainees. He worked in the lab himself. Fred Sanger developed protein sequencing, and he sequenced the allele of hemoglobin that causes sickle cell, and he also developed DNA sequencing, the dideoxy method of DNA sequencing. Those are his two Nobel Prizes. In the lab himself; one technician. In England, he was a government employee. Standard retirement: 65. Literally, on that day, he pushed in his desk chair, walked away, never came back.
ZIERLER: That was it.
DESHAIES: That was it. He worked until his retirement day, and then just walked away, and gardened in his cottage in Cambridge, and never came back. I remember reading that, and it just blew me away. You can't say that somebody who won two Nobel Prizes wasn't passionate.
DESHAIES: You can't say that they weren't fully invested…because most people would look at that and be like, "That's not possible. How could you do that? He must not have cared or whatever." Dude, the guy won two Nobel Prizes [laugh] working alone with a technician. It's that he was able to have other outlets. It wasn't just the science; there was more to him than that.
ZIERLER: This is something that you can only recognize from the outside looking in.
DESHAIES: No, I felt that when I was at Caltech.
ZIERLER: You did? In fact, maybe that was one of the factors that made you think about Amgen?
DESHAIES: Totally, but not until right before. It was having a brother die, and another one get metastatic melanoma in a short period of time that drove that point home to me, because I recognize if I stayed at Caltech, I'd be like the rest of them. I would just keep going, because that's what they do.
ZIERLER: On that note, this has been a phenomenal conversation, and I want to thank you for doing this.
DESHAIES: All right.
- Shou, W., et al., Exit from mitosis is triggered by Tem1-dependent release of the protein phosphatase Cdc14 from nucleolar RENT complex. Cell, 1999. 97(2): p. 233-44.
- Feldman, R.M., et al., A complex of Cdc4p, Skp1p, and Cdc53p/cullin catalyzes ubiquitination of the phosphorylated CDK inhibitor Sic1p. Cell, 1997. 91(2): p. 221-30.
- Lyapina, S.A., et al., Human CUL1 forms an evolutionarily conserved ubiquitin ligase complex (SCF) with SKP1 and an F-box protein. Proc Natl Acad Sci U S A, 1998. 95(13): p. 7451-6.
- Verma, R., et al., Role of Rpn11 metalloprotease in deubiquitination and degradation by the 26S proteasome. Science, 2002. 298(5593): p. 611-5.
- Sakamoto, K.M., et al., Protacs: chimeric molecules that target proteins to the Skp1-Cullin-F box complex for ubiquitination and degradation. Proc Natl Acad Sci U S A, 2001. 98(15): p. 8554-9.
- Sakamoto, K.M., et al., Development of Protacs to target cancer-promoting proteins for ubiquitination and degradation. Mol Cell Proteomics, 2003. 2(12): p. 1350-8.
- Schneekloth, J.S., Jr., et al., Chemical genetic control of protein levels: selective in vivo targeted degradation. J Am Chem Soc, 2004. 126(12): p. 3748-54.
- Verma, R., et al., Selective degradation of ubiquitinated Sic1 by purified 26S proteasome yields active S phase cyclin-Cdk. Mol Cell, 2001. 8(2): p. 439-48.
- Seol, J.H., et al., Cdc53/cullin and the essential Hrt1 RING-H2 subunit of SCF define a ubiquitin ligase module that activates the E2 enzyme Cdc34. Genes Dev, 1999. 13(12): p. 1614-26.
- Reitsma, J.M., et al., Composition and Regulation of the Cellular Repertoire of SCF Ubiquitin Ligases. Cell, 2017. 171(6): p. 1326-1339 e14.
- Liu, X., et al., Cand1-Mediated Adaptive Exchange Mechanism Enables Variation in F-Box Protein Expression. Mol Cell, 2018. 69(5): p. 773-786 e6.
- Verma, R., R.M. Feldman, and R.J. Deshaies, SIC1 is ubiquitinated in vitro by a pathway that requires CDC4, CDC34, and cyclin/CDK activities. Mol Biol Cell, 1997. 8(8): p. 1427-37.
- Verma, R., et al., Phosphorylation of Sic1p by G1 Cdk required for its degradation and entry into S phase. Science, 1997. 278(5337): p. 455-60.
- Lyapina, S., et al., Promotion of NEDD-CUL1 conjugate cleavage by COP9 signalosome. Science, 2001. 292(5520): p. 1382-5.
- Verma, R., et al., Proteasomal proteomics: identification of nucleotide-sensitive proteasome-interacting proteins by mass spectrometric analysis of affinity-purified proteasomes. Mol Biol Cell, 2000. 11(10): p. 3425-39.
- Graumann, J., et al., Applicability of tandem affinity purification MudPIT to pathway proteomics in yeast. Mol Cell Proteomics, 2004. 3(3): p. 226-37.