Douglas C. Rees
Douglas C. Rees
Douglas C. Rees
Roscoe Gilkey Dickinson Professor of Chemistry;
Investigator, Howard Hughes Medical Institute
By David Zierler, Director of the Caltech Heritage Project
September 17, September 29, October 1, October 13, and October 20, 2021
DAVID ZIERLER: OK, this is David Zierler, Director of the Caltech Heritage Project. It is Friday, September 17, 2021. I'm delighted to be here with Professor Douglas C. Rees. Doug, it's great to see you. Thank you for joining me today.
DOUGLAS REES: Oh, it's a real pleasure, David. I'm excited about this experience, and I've certainly enjoyed the oral histories over the time I've been at Caltech and have been reading through them.
ZIERLER: That's wonderful to hear. To start, would you please tell me your title and institutional affiliation?
REES: I'm the Roscoe Gilkey Dickinson Professor of Chemistry, and I'm also an Investigator of the Howard Hughes Medical Institute.
ZIERLER: This is such a special opportunity. So many faculty have a tenuous, at best, connection with the person that they're named for in terms of their chair. For you, of course, this connection goes back all the way to Caltech's foundational years, and academically and scientifically, there's a very interesting academic lineage as well. So let's go all the way back for you. What is the significance for you for Roscoe Gilkey Dickinson?
REES: Well, I'd say there are several. One is, in the academic lineage, Roscoe Gilkey Dickinson is my academic great grandfather. My academic father, my graduate advisor, William Lipscomb, his advisor, Linus Pauling, and his advisor Roscoe Gilkey Dickinson, all received their PhDs in chemistry at Caltech. Although, I think technically, Dickinson's was in chemistry and crystallography, the exact title of his degree. That was one part, but it also is the case that Roscoe Gilkey Dickinson received the very first PhD at Caltech. So this was sort of doubly significant for me, especially when I served as Dean of Graduate Studies and was reflecting on the contribution of the graduate program to academic life at Caltech.
ZIERLER: A really basic question that gets right to the heart of your research at the convergence of so many scientific disciplines. There is biochemistry, biology, biophysics. You seem to touch on all of it. Can you explain broadly in your research career where these distinctions are simply academic nomenclature, and where there are actual distinctions as it relates to your research?
REES: That's a great question. We do tend to have sometimes highly defined or siloed sort of structures in our academic life, but the reality is that research often spans many different disciplines. And in my case, I was always broadly interested in science, especially on the biology side, how life works at a molecular level. But to have some sort of description requires not only the biology, but also, for the fundamentals, the chemistry and physics.
And so, as an undergraduate, I majored in molecular biophysics and biochemistry at Yale. But my program really emphasized physical chemistry, and the basic upper-level physics classes in electricity and magnetism, and quantum mechanics. I think in a sense, they were connected by math, and I've always been interested in how one can model experimental data with equations. So this was really, I'd say, a reflection of my interests that seem to span these different disciplines.
When I went to graduate school, I didn't have a particularly strong interest in a specific field and so I applied to a range of programs. I applied to the biology department at Caltech, biophysics at Harvard, and chemistry departments at Berkeley and UCSD. But again, I didn't have a strong specific interest in what I was interested in doing. Through some really fortunate events, basically a first-year course in graduate school at Harvard on protein structure and function that was taught by Steve Harrison and Don Wiley, I discovered protein structure. I had gone into graduate school with, I'd say, a negative interest in structure.
And to my surprise, I found that all these things I liked to do were really sort of embraced by structural biology. Structural biology could be drawing a picture of a tree or whatever, but in this context, I mean protein crystallography, specifically X-ray diffraction. And I discovered that all the things I liked to do were contained in this field. And it was well-suited for me because you basically had a license to work on all sorts of approaches without being an expert maybe in any of them. So I've always enjoyed having this opportunity to work in different areas, and they were connected by this structural theme. But I wasn't the world's expert in any one of these areas.
ZIERLER: You're, of course, a professor of chemistry. But at the end of the day, if you had to pick one, would you say that you're a chemist?
REES: The short answer would be I think of myself as a biochemist in a chemistry department. The thing I like about chemistry, or why I think chemistry is the proper umbrella, is that chemists really are striving for molecular mechanisms. And of course, they do other things, there are synthetic chemists and so on. But at the core, I think chemists want to understand how nature works at a molecular level in terms of reactivity. And I think that's sort of an appropriate umbrella for what I do. Now, usually, you think of synthetic chemists making things from coal, air, and water. And certainly, that doesn't describe biological systems. But I've always felt that my colleagues in chemistry appreciated why what we were doing is interesting. And I think in part, it was because we're all trying to work out mechanisms. It's really been a great environment for me to do biochemistry, since my background and strengths aren't in mechanisms. So being able to talk to real chemists has been quite valuable.
ZIERLER: The oft-used metaphor of a pyramid, where physics is at the bottom, and then chemistry is one up from that, and biology is one above that, for the research questions that are most important to you, either in terms of your approach or a molecular perspective, is that a metaphor that resonates with you?
REES: Like a lot of things, there's some element of truth, but it's, of course, hugely oversimplified and misleading. I think the thing that really drives most research programs are the questions that you're asking. At least in my world, these are almost entirely biological questions. And so, in that sense, knowing where the frontier areas are in biology or biochemistry is essential. Working on a problem because physicists would find it interesting in what I do would not be a great career strategy. The main consumers of what we do are biochemists, and biologists, and biomedical types. I'd say in terms of the choice of problems and the importance of the problem, that's purely a reflection of the biological interests.
When it comes to solving problems, then, in fact, the foundation in physics and chemistry, at least the way we approach things, is essential. Maybe another way to think about this is, if I think back on my undergraduate courses in physics, chemistry, and math, my guess is those lectures probably haven't changed much in almost 50 years. The biology lectures have changed completely. And it's not because what we were thinking about or were taught then was wrong, there's just been this explosion of information. And of course, there are exciting frontier areas in chemistry, math, and physics as well. But there's such a well-developed foundation that you have to work through. So in that sense, I think the pyramid for what I do makes sense, but not in terms of the findings or scientific opportunities.
ZIERLER: To stay at that foundational level, given that the origins of your education were in biophysics, what are the concepts in physics, either at an observational or theoretical level, that are most formative for your research agenda? In other words, is it statistical mechanics, is it Brownian motion? What are the things from the worlds of physics that are most important to you?
REES: That's a great question. I think fundamentally, it's the quantitative thinking. My colleague, Rob Phillips has this mantra that quantitative data require quantitative models. Biochemists, to hugely overgeneralize, talk about how an enzyme works. And we say, "Well, we've got this group here, and the substrate comes in, and there's some interaction that lowers the transition state. And that's how it works. And so, we say, "If we change this residue, then it should work less well," but we don't typically make a quantitative prediction that we expect the effect to be 10% or 100%. We may make the mutation and find that, in fact, the enzyme's more active. And then, we go, "Oh, well, that's even more consistent with my model," and all this stuff. So I think it's really the importance of the quantitative modeling. And there is an underlying physics, physical chemistry foundation that you can use to try to understand the effects that are going on.
ZIERLER: I'd like to ask an overall question about the relationship between observation and theory in your research agenda, and this also requires a historical appreciation for where you see the relative maturity of the fields that you're most active in. To take out of the playbook from physics, which is certainly mature in the way that theory and experimentation have bifurcated, certainly since the middle of the 20th century, in chemistry and biology, where do you serve as your own theorist, where you're prognosticating ideas and then go to the lab and see what works, and to what extent are those fields sufficiently well-developed, where there are dedicated theorists who are working in that realm, and they're responsive to what's happening in an observational or experimental framework? How do those distinctions work in your research agenda?
REES: Well, we discussed my academic lineage, and I would say both Lipscomb, my advisor, and Pauling were unusual in that they really combined both theory and experiment. And I don't think I really, fully appreciated, at least for my advisor, how strongly the theory part was because I always thought of him as this remarkable crystallographer. It was only in later years where I realized that the crystallography provided a structural foundation that he could then use to begin understanding bonding in these compounds. And whether it was boranes that he won the Nobel Prize for or enzymes, like carboxypeptidase mechanisms, I think he was truly driven by the bonding piece. And theory and experiment were just intimately linked for him. I would say for a long time in chemistry and biology, those have been distinct disciplines.
In my own work, I'm not a theoretician, I appreciate what they do. I liked trying to have simple models that we could then compare against experiment. Probably the level of the theory as I understand it is now 50 or 75 years out of date. But one of my early papers, as a graduate student–and Lipscomb would let us do projects and publish on our own–was trying to understand, if you change a fundamental property of certain proteins, namely their charge, does their affinity for electrons change? And of course, electrons are negatively charged. If you change the overall charge on the protein, you could imagine that that will then affect the relative affinity for electrons. And it does.
So there was a simple model when I was a graduate student, but also at that time cutting edge, state-of-the-art, developed by John Kirkwood, who was a famous theorist who'd actually been in the Caltech chemistry department before he moved to Yale, that would let you, as a function of one parameter, the internal dielectric constant, evaluate the relationship between charge and affinity. And so, I was able to take other people's data and model what this dielectric constant might be. And for me, this was sort of the perfect level of theory. It was Coulomb's law, 1 over R scaled by an effective dielectric constant. You could plot data against the model, there'd be a line, and you could use the slope to give you some insight. That's sort of the level I work at. When I started out, I wouldn't say it was state-of-the-art, but it was acceptable. Now, of course, things have become much more sophisticated, and while I like to think at that level of analysis– it is not anything that you would think about publishing. So in that sense theory now involves more accurate and realistic calculations and practitioners need to have to have quite a bit more background.
More recently, there has been an important coupling between theory and experiment that goes back to the Lipscomb and Pauling style. I think this was really pioneered by Steve Mayo in our BBE Division –and he also has an appointment in chemistry and we're happy to have him as a biochemistry colleague–in his protein design work.
For a long time, the theorists would make predictions, say, "We think this protein should be more stable." Steve revolutionized this field by actually making these molecules, seeing if they were more stable, and then refining his forcefield. And so, it was that coupling between experiment and theory that propelled the field forward. Now, with artificial intelligence and protein structure prediction, I think, again, the theorists and computational chemists have really leaped ahead. Sometimes, you need one person or a few people to do both theory and experiment to move the field forward. And then, I think the specialists can really take over.
ZIERLER: A really broad question as it relates to X-ray crystallography, which of course, goes all the way back for your academic lineage as well. So with this rich century-plus heritage in this field, what is it about X-ray crystallography that makes it still relevant today, and where does that fit in with more recent advances in microscopy?
REES: This is an important question. Certainly, now, I've been in the field long enough to realize you can't just keep doing the same thing for an entire career. There were a number of things that attracted me to X-ray crystallography. Some, as I mentioned, were just that I liked all the different pieces, so I could, as it turned out, make a living doing those things. The broader piece is, if you want to know how something works, at some level, you have to know what it looks like, whether it's a protein molecule, whether it's a computer chip, an elevator, an airplane, whatever. If you're explaining to someone about how something works, the first thing you often do is draw a picture. And so, you have to know where the pieces are located relative to each other. And when it comes molecular structure, you need atomic resolution, and it's just hard to beat X-ray crystallography.
Now, how X-ray crystallography has been used over time has changed quite a bit. When Lipscomb was a graduate student at Caltech in the 40s, his thesis included the X-ray crystal structure of methylammonium chloride - so essentially three heavy atoms, carbon, nitrogen, chlorine, because Pauling needed to know the C-N single-bond distance for his resonance theory of peptide bond structure. This was actually the second structure that was ever refined by least squares methods to obtain an accurate C-N bond distance. There was also some electron diffraction structures in Lipscomb's thesis, so in that sense, his graduate research was an example of "back to the future" because it was not uncommon at that time for Caltech graduate students studying molecular structure to combine both X-ray and electron scattering, just as they might today. Finally, it is worth noting that half of Lipscomb's thesis was classified because he was engaged in war-related research.
In Lipscomb's time, if you could solve a small molecule structure and establish the atomic positions, this was an employable skill. Within a few decades, however, while the importance of X-ray crystallography didn't change, at least in chemistry departments, you were no longer getting faculty jobs. X-ray crystallography was becoming more of a core analytical function. And even today, most chemistry departments have an X-ray crystallography facility because the synthetic chemists and other chemists need to know the actual structures of the molecules they are studying, even though the theory of covalent bonding in organic compounds is well understood. So the importance of X-ray crystallography hasn't changed in chemistry departments, but how it is used has changed.
X-ray and electron scattering have been at the core of structural biology since the earliest days. My entree into structural biology when I rotated in the laboratory of Steve Harrison as a graduate student, started out trying to take images of viruses by electron microscopy. And I was, I'd say, singularly unable to get the images in focus. I think I was traumatized by that experience, maybe another reason I moved to crystallography. [laugh] Until recently, the main challenge with electron microscopy for structural biology was getting atomic resolution data. This is not a fundamental limitation of electron microscopy, however, since for materials, you can get high resolution structures, but there is too much radiation damage with biological samples, which would never let you get atomic resolution structure. That's just changed completely in the last ten years, however, due primarily to improved detectors that reduce the electron dose required to get high resolution data. As a consequence, many of the frontline structural studies for moderate-sized and large macromolecular structures are how being done by electron microscopy. I have to say, I failed to anticipate this transition, but fortunately we have a great cryo-electron microscopy facility at Caltech and as a result have been able to enjoy some of the real advantages of this technology.
These developments have critical implications for the role of X-ray crystallography in structural biology, especially for larger structures. These were often problematic systems to study, starting with challenges in the purification and crystallization that is essential for crystallography. One might approach large systems by trying to break them into smaller pieces and find units that would crystallize, but now you can just put the intact system in the electron microscope for structural analysis. Electron microscopy now makes a lot more sense for large systems than trying to crystallize them. But there are still several areas where X-ray crystallography is, at least at the moment, unparalleled, and that is for getting very high-resolution structures, sub one Ångstrom resolution, and also for smaller systems under maybe 50 kilodaltons in molecular weight that are, at the moment, still too small for electron microscopy.
ZIERLER: Can you explain that on a technical level? What is it about X-ray crystallography that has allowed it to accomplish these imaging miracles that even the most incredible electron microscopes can't?
REES: I'd say it's not because of any fundamental restrictions on electron microscopy, it's about the nature of the samples that one's investigating. A real advantage of electron microscopy is that electrons are scattered something like four or five orders of magnitude more effectively than X-rays. So for a given sample size, you get a stronger signal from electron scattering. But if you're able to crystallize a sample so you now have, I don't know, a ten-micron-sized crystal or even micron-sized, you would have enough copies of that molecule that they would give it an appreciable scattering signal as long as the intrinsic order of the lattice is OK. Even though X-rays are scattered less effectively than electrons, the interference pattern generated by the scattering of molecules in a lattice effectively amplifies the signal. Now, of course, you can do electron crystallography as well. And that can use even smaller crystals than in X-ray crystallography.
But in many cases, the challenge is getting the crystals in the first place. A real advantage of electron microscopy is in so-called single-particle work, where you deposit molecules in a thin layer of ice on a grid. In this layer are random orientations of individual molecules. By imaging that data in an electron microscope, with a lot of computer power, you can figure out how these molecules are oriented with respect to each other and essentially take, say, 100,000 random images, and basically do what the lattice is doing in a crystallography experiment.
But if you look at these raw images, it's like there's no signal there. It's still almost like a miracle to me that one can obtain atomic resolution structures. Except, of course, it's not a miracle, it's physics and data analysis. But there are limitations on how well you can align the individual particles because of the noise in the images. And that's sort of what's currently limiting the resolution of these single-particle EM studies. The big advance was in detector technology. It turns out if you have your sample on an EM grid, when the electrons are coming through the sample, the electrons are sufficiently energetic that they will interact with your sample and knock out electrons. And you get, I think, these funny electrostatic effects with the result that the sample starts moving over the grid. And so, if you're taking a still picture of a moving sample, the image becomes blurred. These new detectors are basically able to record movies, where instead of having one long exposure, you now have a number of shorter exposures, and you can move things around in the processing so you get a motion-corrected image. And that sets the current resolution limit. But with time, clearly, the current resolution limits will be smashed, and we'll be doing more and more amazing things.
ZIERLER: It does sound like, at least for the foreseeable future, there's a certain level of job security in the field of X-ray crystallography. This is a technology that's not going anywhere any time soon.
REES: Well, I'd say yes and no. There's always a need for crystallographers, but the way in which it's used has changed quite a bit. For example, I think for faculty positions, universities are now hiring electron microscopists. They're not looking to hire, in general, crystallographers. Although, many people are doing both, so it's not quite that black and white. I think there are other demands. For example, we have the Molecular Observatory in the Beckman Institute that provides access to X-ray sources for members of the Caltech community, we have a great staff scientist, Jens Kaiser, who's running that facility, and he's very much in-demand. But the crystallography by itself is a tool to get structures. The main consumers of structures, as we were discussing in this area, are typically biologists or biochemists. The actual details, the sausage making of solving structures, is not something that really is of such interest to the consumers of these structures.
ZIERLER: Broadly conceived, where do you see the historical, and even intellectual, development of structural biology vis-a-vis molecular biology? How do you see these things developing?
REES: I should say I'm not a historian of science, so this will reflect a certain ignorance or selective focus. With that caveat, I would say, at some level, the origins are closely coupled. Again, at the risk of oversimplification, there were two major centers of development of structural biology, and they were also centers of development of molecular biology. And those were Caltech and Cambridge, England. And especially in Cambridge, I think molecular biology really was structural biology. Lawrence Bragg, the son in the father-son team that won the Nobel Prize in Physics for their early contributions to X-ray diffraction, realized that a frontier in crystallography was to understand biological structure at the molecular level. This effort eventually led to the establishment of the Laboratory for Molecular Biology at the Medical Research Council Laboratories in Cambridge. The LMB was the birthplace of protein crystallography. That's where the DNA structure was solved, and the first protein structure, myoglobin, was solved, and Perutz worked out the allosteric mechanism of hemoglobin. So a lot of the major home runs. And what's amazing is how they've continued their impact up to the present. At Caltech, Pauling saw protein structure as sort of the natural progression of his work on molecular structures. The initial efforts on protein structure began with the structures of amino acids. Amino acids are the constituents of proteins, and the crystallographic studies at Caltech focused on the molecular structures of amino acids and how amino acids hooked together through peptide bonds to form polypeptides. (Understanding the properties of the peptide bond provided the motivation for Lipscomb's thesis research to determine the C-N single bond distance).
In the culmination of that era of structural biology at Caltech, Pauling's group worked out how peptides could form so-called secondary structures, alpha helices and beta sheets. It's interesting to compare and contrast the Pasadena and Cambridge, England schools of protein structure. The Cambridge school really saw this problem from a top-down perspective. "If you want to solve the structure of a protein, you need to get crystals of the protein." Whereas the Caltech school approach was, "If you want to understand the structure of a protein, you have to understand the structure of the amino acids, how they're combined to form peptides. What are the regular conformations of these peptides? And then, how are the amino acids packed together to form a protein?"
For Pauling and his group, the pinnacle of structural biology was identification of the alpha helix and the beta sheet. They never made the connection to the three-dimensional fold of proteins, although there were efforts underway by R. B. Corey working on lysozyme and other structures that were not successful. The Cambridge group also had some misfires. They, too, were trying to predict the structure of the alpha helix, but didn't realize, as Pauling had, that the peptide bond is planar, and you don't have free rotation around it. So, when they were coming up with helical structures, they had more degrees of freedom and proposed a number of structures, none of which were correct. However, the Cambridge group ultimately succeeded in first solving the structure of an entire protein, myoglobin. As fortune would have it, myoglobin has a significant number of alpha helices, so that first protein structure captured the efforts of both Cambridge and Pasadena schools.
Now, while these developments were occurring in structural biology in Pasadena and Cambridge, Max Delbrück was also at Caltech pioneering the foundations of molecular biology. His work spanned physics and biology without, it seems to me, much of a role of chemistry in that process. A lot of the work that was done by Delbrück on mutagenesis was sort of thinking about the statistics of mutations and so on, using small viruses, bacteriophages, as the test bed. Eventually, those efforts developed into a school of molecular biology school without much of a role for structure in it. So at Caltech, you had these two efforts going on side by side, until the structural biology part reached its culmination by the early 1950s while the molecular biology continued. In Cambridge, these threads were connected for a much longer period of time.
ZIERLER: A purely speculative thought experiment kind of question, but one that might serve as a real window into some of the advances in the field. Perhaps going back to Dickinson is too far, but if you magically had the good fortune of inviting Linus Pauling into your lab today, what are the questions that he would immediately recognize, and where have things advanced so far that you'd have to sort of take him back to 101 with the things that you're working on?
REES: Yes, great question. We're building on the heritage of the Cambridge and Pasadena schools. In 1939, Linus Pauling wrote a paper entitled, The Structure of Proteins, and J. D. Bernal, who was affiliated with the developments in Britain but wasn't at Cambridge, also wrote a paper on protein structure with the same title. These papers are remarkable reads today, even after 80-plus years. While nothing was known about the atomic resolution structures of proteins at that time, there was an excellent understanding of the basic molecular forces and interactions that would stabilize protein structures. Pauling was focusing on hydrogen bonding, while Bernal was talking about the hydrophobic effect.
An important distinction in the approaches of Pauling and Bernal to the "protein structure problem" is that Bernal's student, Dorothy Crowfoot Hodgkin (who subsequently won a Nobel Prize for her structures of vitamin B12, cholesterol and penicillin), had crystallized a protein in 1934 and discovered that if you kept it hydrated, the diffraction pattern was maintained. And so, Bernal, knew by 1939 it was possible to solve the structure of a protein by X-ray crystallography. He didn't know how this could be done, but the existence of a diffraction pattern from a protein crystal meant it would be possible to get the structure. At the same time, Pauling, who'd been working on his bottom-up approach, starting with amino acids, peptides, and so on, was saying, "It took us so long to do the structure of a dipeptide (two linked amino acids), I don't see how we'll ever solve the structure of a protein."
A number of protein structures had been solved by the time of Pauling's passing in 1994, but the qualitative understanding of the forces that hold proteins together, and the existence of the alpha helix and beta sheet in protein structures, had been appreciated since the 1950s. So, I don't think these features would have seemed remarkable to him. But I think the pace at which structures are now being solved and our understanding of enormous assemblies of proteins such as the nuclear pore complex solved by my colleague André Hoelz, would, I think have to be astonishing to Pauling. It's astonishing to me. [laugh] Of course, Pauling was a visionary so he may have foreseen that electron microscopy would let you do all this stuff. The technical advances in DNA sequencing, genomes and all these things – I think our ability to get the basic structures, the sequence foundation, the technical advances, are extraordinary.
But there are still a lot of questions about how cellular processes are regulated during development and with changes in their environment. Pauling was quite interested in immunology and had very broad interests. And I think at those levels, there are still a lot of questions. And so, maybe he would be surprised that, knowing all this structural information, we still have big holes in the detailed understanding of the mechanics of cells and organisms.
ZIERLER: Staying on the Caltech lineage, you've written about how Bill Lipscomb really instilled in you a sense of the values in the scientific approach at Caltech. What were those values, and how do you incorporate them into your own work?
REES: Of course, I had this view of Caltech before I ever got here, certainly from Pauling's general chemistry textbook, Feynman's physics books, Apostol's book on calculus, and so on. It's really an extraordinary place. Growing up, several of my neighbors in Lexington, Kentucky had been either graduate students or postdocs here and that also contributed to this sense that Caltech was a special place. But the opportunity to see this in action was when I was a graduate student at Harvard, working with Bill Lipscomb. Lipscomb had an informal manner and a great sense of humor that was definitely not the characteristic I would use to describe the Harvard chemistry department or Harvard when I was there. Lipscomb would wear a string tie and didn't want to be called Professor Lipscomb but rather "The Colonel" because he was from Kentucky. When I first met with him, I addressed him "Professor Lipscomb" and his response was "No, call me The Colonel." In addition to this informality, there was also a sense that – and it's not unique to Caltech, but rare –it was important to be working on really interesting problems. Just turning the crank to do something was not a sufficient reason to be working on a problem. This view was captured in his saying that the worst thing that could happen was not to be wrong, but to be working on something that wasn't interesting. That said, Lipscomb did not like to be wrong, and of course what is important and interesting can vary quite a bit from person to person. [laugh] But there was this sense that people should be working on projects that, when you solved them, you would learn something new and important, however that was defined. Another part of that view, which I didn't associate with Caltech at the time, but I've seen it now, is that the way Lipscomb ran his group. We basically had full responsibility for our projects. We weren't just working on one little piece of this big project, but we had ownership from beginning to end.
One of my first experiences with this culture was when I went into the lab one day, and found that the X-ray generator was broken. I'd never fixed any X-ray equipment before, and I had no idea what the protocol was. So I went to Lipscomb, and reported "Colonel, the X-ray set's broken," or something like that. And I don't remember exactly what he said, but I clearly remember when I left his office, that it was my responsibility to get the X-ray set fixed. I hadn't worked on anything mechanical before, I'd never taken cars apart or any of these things. Lipscomb had remarkable trust that we would figure it out–he gave us a lot of trust. I would say for many of us, it was not justified initially. I'd like to think over time, though, it became more justified. But I think we quickly realized that it was our responsibility to solve the problems that came up in our research. In my group at Caltech, I try not to make it quite so much being thrown in the deep end of the pool, but there is an aspect, in terms of doing academic research and doing a graduate PhD, where you're trying to work out new things, and there is a lot of uncertainty and exploration, and you need to discover these things.
Lipscomb had a well-developed sense of humor, which I would say was at least atypical for Harvard at the time I was there. I came to associate this trait with Caltech; while it's not like everyone at Caltech is funny, there is a sort of informality and a lightness, while at the same time, the underlying work is quite serious.
For me, Caltech finds a way of threading that needle where you hopefully don't feel like you're suffering while you're doing all of this challenging work, and that you can also have a good time while doing things that are really hard. That was one of the main impressions I got of Caltech through Lipscomb.
Lipscomb came to Caltech in the physics graduate program. It was Pauling's lectures on chemistry and the chemical bond that inspired him, and he switched to chemistry. Today, I see that many of the graduate options at Caltech do not have a really highly structured program, and graduate students can work for any faculty member, whether they're in the option or not, as long as they satisfy their degree requirements. That sort of flexibility, I also see to be part of the Caltech experience.
ZIERLER: As you well know, Linus Pauling had a very colorful and even public political identity. I wonder what your sense is in what he transmitted to Lipscomb in terms of those kinds of questions about comporting yourself as a public intellectual, talking about politics, when to bring it into the classroom, when not to, and how those ideas may or may not have been transmitted to you. What's your sense of that aspect of this genealogy?
REES: One of the things that still remains a mystery to me is what Lipscomb's political views were. I don't think he was either super liberal or super conservative. I don't think he was in the Pauling tradition in that way. By the time I got to graduate school, the draft had ended. But before that, there must have been in his group a number of students dealing with the draft. I didn't ever get a strong sense of what his views were on the Vietnam war and the draft, for example. More broadly, I never did get a sense of what his political views were like.
Now, building on his sense of humor, Lipscomb did become very active in the Ig Nobel Prize ceremonies, which has a very public face, and would do things I would be too inhibited to do. Like performing in the ballet "The Interpretive Dance of the Electrons" or serving as the prize in the first "Win a Date with a Nobel Laureate" contest. And he was really happy to do these things. Lipscomb liked to perform and was quite an accomplished musician. In fact, he went to the University of Kentucky on a music scholarship and played clarinet. Every summer, the one vacation he took was to go to a music camp in Vermont. So in that way, certainly, he wasn't just cloistered in an ivory tower. But his politics, I have no idea.
ZIERLER: I want to ask you to reflect on Caltech values not that you received by osmosis as a student, but in your own experience as a faculty member, and that is the sense that at Caltech, the distinctions between scientific disciplines, departmentally, really are not that important. There's a convergence, where you can go and talk to anybody wherever the research might lead. I wonder if you might reflect on that sense of how science happens at Caltech.
REES: I'd say the defining traits of Caltech are the small size and the exclusive focus on research and education in science and engineering. There's no other research university like this in the world. There may be other small places like Rockefeller University that focus on the biomedical sciences, but trying to have this umbrella approach to science and engineering with a faculty of 300 is extraordinary. And I think that one of the consequences of that is, by and large, the research interests of faculty are quite broad. After all, we grandiosely cover all of science and engineering with 300 faculty. And one of the consequences of that is that you're not in a place where there may be six or ten other people working in the same field, and you don't have to think about, "Well, if I work on this, I may be stepping on someone's toes."
So that's one part. Another is that there are 1,300 graduate students for 300 faculty. Graduate students soak up this culture of informality. At least, this is my idealized view of things, which, even after five years in the graduate office, I'd say has survived more or less intact. And so, it's not unusual for students, at least in chemistry and biology that I'm most familiar with, for students to be interested in studying a research problem going on in one group with a technique that's going on in another group. The structure of the candidacy and thesis committees with different faculty getting to know a student also reinforces this. Since the research interests of the faculty tend to be broad, and I think the graduate students also tend to feel empowered that they can shape their thesis, they often see interesting areas at the interface between groups.
I think this effect is a powerful driver of new projects launching, not necessarily through collaborative efforts, but rather catalyzed, motivated, and fertilized, by this environment. In principle, the small size of university should mean that you run into your colleagues more frequently, at least in non-pandemic times. One of the things I worry about is that everyone's just become so busy that, in fact, you don't see your colleagues any more frequently than you might see non-colleagues in your field. And I worry how these informal connections for faculty - just talking about things and so on - can be maintained when people are pulled in so many different directions and don't have time to sit around and chat with their colleagues, or students, or whatever for several hours a day.
ZIERLER: You touched on that, and it's a question we're all dealing with right now. For your research, for the way that you interact with your colleagues, how have these past 18 months in the pandemic been for your scientific life? What has been difficult in terms of social isolation? What may have been productive just because you're not going in as often, you might have more bandwidth to work on papers? How have these things played out for you?
REES: I'd like to back up a bit to where I am in my career. On October 1 of 2020, I finished my tenure in the graduate office, where I'd been serving as Dean of Graduate Studies for five years and three months. This was an enormously rewarding experience, but really quite time consuming. I was at a point in my career where I was as comfortable as one could be with that tradeoff.
ZIERLER: You mean having many more administrative responsibilities than research capacity?
REES: Yes. And partly, it's connected to how I am viewing the end game for my career as a Caltech faculty member. I have had the incredible opportunity, as every Caltech faculty member does, of running a research group. Caltech faculty play an important role in the running of the Institute but these administrative responsibilities also take a lot of time, I am at a point in my academic career where I felt it was important to participate in some administrative capacity, if given the opportunity, since the time away from research can be so much more devastating to the research careers of younger faculty. I certainly found as graduate dean that the time I had to focus on research was severely diminished. And even more so, the time I had to interact with my colleagues. That preceded the pandemic, but of course, the pandemic accelerated that. During this time, the size of my research group has been decreasing, again, by design. Overall, I've been surprised by how much work we've been able to get done during the pandemic, but it certainly has impacted my ability to spend time with the group, and certainly to spend time with my colleagues. Zoom has helped more than I anticipated, but I think we all benefit from the in-person connections. We're starting to get that back as we're coming back to campus a bit but the in-person connections are still impacted. But I'd say that's one of the long-term negative aspects of this period.
ZIERLER: Let's take a broader view of your overall research agenda even before your administrative work at Caltech over the past five years. Let's start first at the most basic level. I asked before about a certain binary in science between theory and experimentation. Another one would be, for you, between basic science and applied research. In other words, just figuring out how stuff works, and then sequentially thinking about what these things might be applied to, either in therapies, in products, or things like that. For you, obviously, it's mostly a basic science approach. Where are the applications, either for you specifically or in others who have used your research as a launching pad for those applications?
REES: I'd say society invests in science not because we as scientists want to solve some problem in principle that we are interested in, but because through this trickle-down chain, there are real benefits to society of basic research, which I think have really been realized in this country. Through much of my career, the focus of universities has been on the basic questions and the training of students and postdocs. Certainly, when I was a graduate student, the idea, especially in biology, that there would be direct applications of research, starting companies and stuff by faculty, was outlandish. There was a really sort of, looking down on organic chemists who might consult for chemical companies, or whatever. What I have concluded is that biologists are quick learners, and there was a quick realization that you could have a company or companies based on technology being developed in your group. One example I think about is DNA sequencing, which was developed by Maxam and Gilbert at Harvard just before I got to graduate school (and also by Sanger in Cambridge, England). As it turned out, I rotated in Gilbert's lab and worked for Allan Maxam, and I could see this transition going on. I would never have realized that you could have a company, or an entire industry based on the molecular biology technology that was being developed in academic labs around the world. It's just extraordinary.
My own interests have been fueled by the basic questions and a belief that the answers to interesting, whatever that means, biological questions will have impact on society. Of the two major projects that we've worked on, one is biological nitrogen fixation, the process of which has enormous practical ramifications. Until 100 years ago, essentially all of the nitrogen that was needed for fertilizer and life came from the action of nitrogen fixing micro-organisms. At that time, the Haber-Bosch process was developed, and today, roughly half of the fixed nitrogen that goes into agriculture goes through the Haber-Bosch process. This whole process has fueled an enormous expansion in the increase in the world's population and so on. And it also takes quite a bit of energy, whereas the biological process, as it turns out, is also pretty energy intensive, but you can get it from sunlight and other sort of "green" processes.
Much of our motivation in trying to understand how the biological process takes place has been to understand the basic catalytic principles of this system. With this understanding, instead of running the industrial nitrogen fixation reaction at about 500 degrees Centigrade and 200 atmospheres pressure, it might be possible to run the reaction under the milder conditions of the biological system at 20 degrees Centigrade and one atmosphere pressure which should be less energy intensive. I've been working on this problem now for 40 years, and it's still a great problem. We've learned a lot., but we still don't know fully how it works. I have concluded that we're not going to change the Haber-Bosch catalyst to the biological catalyst because whatever the key catalytic principles are, the biological system is too complicated to do things on an industrial scale. But, there may yet be some lessons in terms of the basic catalytic mechanisms that will have an impact.
The latter emphasis overlaps with the area of small molecule catalysts for nitrogen fixation that has been a focus of research in the CCE division, starting with John Bercaw, and then Jonas Peters and now Theo Agapie, who have been working to develop chemical models that maybe in some sense, in the big picture, were inspired by the biological process, but will be working differently. Caltech has provided an outstanding environment for studying the chemistry and biochemistry of nitrogen fixation and I am hopeful that these efforts will ultimately lead to some important practical developments.
Our other major research interest has focused on the structures of membrane proteins. Anything that goes in and out of a cell has to go through a membrane, and the passage across the membrane is facilitated by membrane proteins. When we started, there were no structures of membrane proteins, and we were primarily trying to understand the basic principles – what do they look like and how might they function in the cell membrane? We eventually did a fair amount of work on specialized membrane proteins called ABC transporters, for ATP binding cassette transporters. ABC transporters use the cellular energy currency, ATP, to pump molecules in and out of cells. ABC transporters are important, among other reasons, because a major class of drug efflux pumps are in this family that pump therapeutics out of cells. If tumor cells make more of these drug efflux pumps, it serves to make the cells less sensitive to therapeutics by analogy to more quickly bailing a leaking boat. Another disease that involves ABC transporters is cystic fibrosis. The development of therapeutics targeted to ABC transporters hasn't been the focus of our work, but structural information is useful to various biotech and pharmaceutical companies trying to design inhibitors and so on that can activate or block some of these transporters.
I don't have any involvement with biotech companies, not out of any sense of purity or whatever, but I just don't have enough time as it is. I remember visiting one big pharmaceutical company about 25 years ago, which had a big drug discovery effort with targets that were membrane proteins. But working on membrane protein structures at that time was just too long term. I think now with some of the technologies that have been worked out, it actually is possible for this type of research to be done in these environments. I see the role of academic research is to work out ways of driving developments in fields that then can be utilized more broadly by the community. Obviously, there's no incompatibility between being in universities and being involved in companies, and a number of our colleagues have been pioneers in these areas. And I think students, especially now, are really quite interested in seeing how what they're doing can be applied to various aspects of society. That is an important motivation for graduate students and postdocs now.
ZIERLER: On that point, the way that your research does translate at some level to these advances, what has been the overall value in having the affiliation with the Howard Hughes Medical Institute? What has it done for you to be an investigator for them?
REES: Being only slightly hyperbolic, it was transformative. Let me answer your question about HHMI by first providing some context.
Reflecting over my career, I'd say in many ways, it's really been a remarkable period to be a scientist. And I hope everyone starting out that goes into science ends up, when they reflect back, saying, "Boy, I was lucky to be a scientist during this period." I feel really lucky being a scientist during this period as there were a lot of untapped problems at the chemistry – biology interface to study. One of the things I think is fundamental to how I see the best of science working is –it's important to have goals and that these goals are reviewed by outside parties.
Although grant writing is time consuming, there is a real value in writing grants and saying, "Here's what we want to do, and here's how we're going to approach it." Being reviewed regularly is essential to make sure that the financial resources of a funding agency are being used appropriately. The part that I find concerning, however, is that it the time frame for writing grants and getting funding works against quickly responding to new opportunities. In my view, one of the best ways in which universities work are, we've got these really smart, motivated, driven students, faculty, postdocs working together, and if some new opportunity comes up, we should be able to say "I just heard about this new problem – lets work on it" or "I just talked to my colleague the other day, and she said something really exciting about a possible collaborative project. I think we should start working on this." I think that's the way that university research works best.
The problem is, now, people are worried about how to fund something like this - "Gee, I'd like to work on this problem, but all my funding is for these other specific projects, and these students are working on specific aim one, these students are working on specific aim two, these students are working on specific aim three. It would be a great to start this new project, but I can't… I've got to make progress on my current grant, or it won't get it renewed, and I'll lose my funding," etc., etc., etc.
So, with that background, the important thing about the Howard Hughes Medical Institute is that they fund the investigator, not the project. Everyone should have this opportunity. The way in which HHMI does this is that you have to propose specific research projects that are subject to periodic reviews - currently on a seven-year cycle. If you're not renewed, you lose HHMI funding going forward. Now even though HHMI investigators propose specific research projects, you are reviewed on what you actually did, and not what you said you would do. HHMI really made question-driven research in my group possible.
My completely speculative view is that I don't think Howard Hughes (the person) realized that he'd left his estate to the Howard Hughes Medical Institute when he died. I don't think he was paying careful attention to how his estate was set up. For some time after Hughes' passing, HHMI had a modest impact on biomedical research. But, when General Motors bought Hughes's aircraft in the mid-1980s, the Hughes estate, meaning HHMI, became the beneficiary of a significant amount of money. The history may not be completely right in details here, but I think it's close. Basically, the IRS said, "With this level of funding, you have to have a real structure, review panels, all this stuff." The IRS required that HHMI had to give away a certain percentage a year, which was a significant increase over what they were doing. With this funding and an invigorated administrative process, HHMI transformed a number of biomedical areas, including structural biology. The way in which HHMI works, which I think has been really great, is, rather than building a campus, where everyone worked, they supported investigators in their own universities.
There were initially, I believe, five universities that had HHMI investigators in structural biology. And so these groups, some of which had a long track record in structural biology, but others that had less of a track record up to that time, all of a sudden were able to buy new equipment and computers, hire postdocs, etc. I was at UCLA in the mid-1980s with my colleague, Dave Eisenberg and we really tried hard to figure out a way to get Hughes funding but were unsuccessful.
Eventually, I was awarded an HHMI appointment in 1997 after moving to Caltech in 1989. HHMI funding really transformed our ability to work on membrane proteins, which at the time was a wide-open field, but was just too risky for NIH to fund. Reflecting the risk, research in this area involved significant expenses in terms of the supplies, and reagent, and personnel. Given that postdocs and students working in this area were putting their careers on the line, one didn't want to start working on membrane protein structure without the ability to actually fund this research. The HHMI program really made that work possible.
ZIERLER: Is that to say that your interactions with the NIH happened within that broader NIH, Howard Hughes framework? Or you have an independent funding relationship with NIH?
REES: Those are separate. In my case, I've had HHMI funding since 1997, but I also just finished nearly 30 years of NIH funding for nitrogen fixation. Over the years, we were also getting funding for various membrane protein projects. Once we did the initial work in terms of getting a structural program launched with HHMI funding, then we could write competitive NIH grants. Geographically, NIH and HHMI headquarters are within about five miles of each other. There are institutional interactions, but they are really separate programs.
ZIERLER: Do you have a sense of why the focus was on structural biology for Howard Hughes, in terms of all of the projects they could've focused on, all of the research?
REES: It wasn't the exclusive focus, but it was one of several areas HHMI initially selected, and I think this was probably the vision of Max Cowan, who recognized that structural biology was a resource-limited area that was poised for significant developments. Max got it right. Over time, of course, other areas have come up. And now, I think there's a sense that HHMI has put a lot of support into structural biology, and there are other areas that need funding, such as neurobiology and the brain. Also, some perhaps more surprising, but clearly important, subjects like plant biology. I think HHMI tries to have a diversified portfolio. They have roughly 250 to 300 investigators. The number of investigators is roughly the same as the faculty size of Caltech.
ZIERLER: A technology question as it relates to computers. How have computers changed both the day-to-day in structural biology and even, at a more fundamental level, the most basic questions that can be raised?
REES: There's been, of course, a huge revolution in computation. Back when I was a graduate student starting in 1975, we had our own lab computer, which was a PDP-11/20, and it had 28K of memory. And of course, at that time, that was great. Probably one of the most remarkable accomplishments I've seen in my career involved another graduate student at Harvard, Jim Crawford, who had been an undergraduate at Caltech. Jim wrote a film scanning program because we were collecting X-ray diffraction data on film, and we had to densitometer the films to measure the intensities. In the old days, with small molecule crystal structures, that was done by hand. But with macromolecular projects, you couldn't measure the diffraction data by hand. So we had this drum scanner by Optronics that I think was probably developed for digitizing spy photographs from flights over the Soviet Union.
And Jim wrote, in 28K, a program that would allow you to do all the things necessary to processing diffraction data. The level of programming and crystallographic knowledge required to do this still seems just extraordinary to me. But we couldn't do a lot of calculations on that computer. We had another computer, a PDP-11/45, that was in the computer room at the Harvard chemistry department off the lobby where we could do some bigger calculations. There was also a card reader that would let us run jobs remotely. Lab funding was much smaller than we have today and we had to buy computer time, so we ended up computing remotely at Columbia University, even though the Harvard computing center was right there, because it was too expensive. But to get the cheap rates, we had to work nights.
We were really on a nighttime schedule. And we'd submit the job, and the operator at Columbia would have to mount a disk, so there was a lot of time waiting. And I think, frankly, this was probably one of the most important parts of my education because this was the only computer room for the whole Harvard chemistry department. And there were three structure groups together with all the computational groups. And this included Martin Karplus, who also got his PhD here at Caltech working with Pauling. So Karplus's group was developing the molecular mechanics stuff that he won the Nobel Prize for.
So we'd submit the jobs, then you'd have to wait. There was a lot of time just to talk about stuff. Being there, just listening to what was going on in the development of the molecular mechanics was interesting. I learned a lot of crystallography. I'm not a good programmer, but there, I really learned a lot. And also, a lot about operating systems. The Columbia University computer was an IBM 360 and the Job Control Language was esoteric for me. But there were people there who were just incredible experts. To me, this was probably the most valuable part of my graduate career, sitting in the computer room, basically being forced to talk to people because we were just waiting for these jobs to run. I learned a lot of crystallography, computers, science, life, whatever.
Subsequently, every place I've been at, I've tried to replicate this computer room because I had such fond memories. But as the computing power increased, there was no need to sit in a room with other people because you could sit at your desk, or you could do this at home, or whatever. Certainly, when I moved to Caltech, the computing resources were a limitation on our ability to solve and refine structures. One of the many surprises I've had in my career was when a new postdoc, Oliver Einsle, came to the group and was working on this project that I considered to be computationally intensive. He was telling me about some of the work he was doing, and as I rewound the conversation, I realized he had mentioned that he was doing this all on his laptop. And I thought, [laugh] "How can you do this on a laptop? We've got this big computer over here."
So at that point, the computational needs of X-ray crystallography 20 years ago were not the rate limiting step. Maybe disk storage for data collection at the synchrotron, but not for computation. That's really changed with electron microscopy, both disk space and computation are really resource-limited. Even just the process of moving files around becomes a real challenge. Crystallography definitely benefitted from the advances in computation, probably up to about 20 years ago. And of course, it's always good to run things faster. But I'd say that wasn't such an issue for us. We didn't try to have our own computing cluster for X-ray crystallography once we could use laptops. But now, with electron microscopy, we're back to clusters to process data and solve structures, and it takes a lot of time to transfer files, and to archive and store all of this data. It's a real challenge.
ZIERLER: You mentioned earlier simulation and even AI. Where in your research have you found value in decoupling, to some degree, your life in the laboratory, your life in direct observation, and allowing the computers to explore some of these questions?
REES: In my own personal research, that's not a direction that I've pursued. And not because I think it's offensive or outrageous. I've been pretty much absorbed in the analysis of nitrogen fixation, membrane proteins, electron microscopy. I should say, under the hood for the electron microscopy data processing programs, there's a lot going on. And there may be AI and machine learning is a part of that. But that's something that's just completely opaque to me. I don't really, unfortunately, understand how these programs work and what developments they're based on.
One thing that's sort of interesting, and maybe this shows a generational gap, but my colleague, Bil Clemons and I joined our groups and will be moving into new space in six months or so. He's really been excited about the DeepMind AlphaFold program for modeling and predicting protein structures and is building this into his research program in the sense that we don't need to solve structures if good models are now available. And I guess I look at this and say, "Well, these structures might be right, but they may not be." If they look like other structures, then I'm happy to do this. But if these are de novo structures, I guess I'm not yet mentally able to make that leap and say, "Yes, this might be right. This is almost certainly close enough that there's a lot of benefit to using this as a guide for our experiments."
ZIERLER: Looking even farther afield, and I would assume there would be even more hesitancy in this regard, but to the extent that we are in a quantum revolution right now, and Caltech is central to this effort, do you see any clear limitations in classical computing for which quantum computing would unleash all kinds of new avenues of research in structural biology?
REES: Again, my grasp of the fundamentals of quantum computing is even more tenuous than artificial intelligence or machine learning. I understand from my colleague, Garnet Chan, who's involved in these efforts, that the iron molybdenum cofactor of nitrogenase where biological nitrogen fixation takes place is considered one of the grand challenges for quantum computing. Now, certainly, I think no matter how optimistic or chauvinistic I am about the value of experimental work, when it comes to looking at intermediates in the nitrogen reduction cycle catalyzed during biological nitrogen fixation by the enzyme we work on, I don't think we'll ever be able to model experimentally all the different intermediates that occur. And this will certainly be achieved, ultimately, by computation.
So again, it's not that I have a disbelief in this, but at this point, I think there are definitely computational limitations in the realism of the models that one can use to really distinguish the right solution from the wrong solution. A lot of quantum chemistry has been done on the iron molybdenum cofactor to try to understand how nitrogen will bind. And the best you can say from all these studies is, they can't all be right. A lot of very different models have been proposed. But clearly, there is a limitation to the models that can be treated, and perhaps quantum computing will provide a way to crack that. That would be fantastic.
ZIERLER: For the last part of our talk today, and one that I think will be very useful to set the stage for subsequent discussions, is if we can take sort of a definitional tour of the research areas that are most important to you now and over the course of your career. So first, at a most basic level, can you explain why protein structure is such an enormous field? And for you personally, what is the allure of protein structure?
REES: Almost all of the chemistry of life is made possible by protein catalysts. Inside our cells, there are a large number of chemical reactions taking place that are essential for our metabolism, replication, information processing, interacting with the environment, growth, division, all this stuff. If you just took those chemicals and put them in a bag, you wouldn't get a living system. And many of these chemicals are actually very similar to each other. There are many secrets of life, but the one that I've been most interested in is the nature of these protein catalysts that are able to selectively mediate the transformation of one molecule into another that's part of all these processes that enable us to eat food, extract metabolic energy, synthesize the building blocks to build more protein, nucleic acids, and copy all of these things.
In fact, in one of Kurt Vonnegut's books, Cat's Cradle, there's a scene in a bar where the characters are discussing that "science was going to discover the basic secret of life someday", The bartender then recalls that he just read about this in the paper – and the secret of life turns out to be "Protein". This book was published in 1963, several years after the structure of myoglobin had been solved. Of course, there are a lot of secrets. Like many complicated systems, all the pieces have to work. So nucleic acids, membranes, all these things are also secrets. But I've really been interested in the nature of the protein catalysts. And one of the interesting features is that proteins are usually considered to be large, but the metabolites are often very small molecules. Why do you need this large molecule to interact with these tiny molecules? And how does catalysis work?
When I was a graduate student, there were various theories of catalysis, which in a general sense were probably correct, but it was challenging to apply them to enzymes in the absence of structures. I went to graduate school not thinking I wanted to do structure. I had this negative interest in structure from my undergraduate experience, which is ironic because Yale, where I was an undergraduate, actually was a hotbed of crystallography when I was there. But I wasn't that interested in structure. There's a lot about how I ended up at my particular graduate school and worked on structure, but I don't know if this is the time to discuss that or not. But to answer your question on the allure of protein structure, in graduate school, I took Biochemistry 112 taught by Don Wiley and Steve Harrison that was an extraordinary course. (as an aside, –Don subsequently was the thesis advisor of my colleague Pamela Bjorkman. At Caltech, Pamela and I introduced a course on protein structure and function, BMB 170, that was modeled after Biochemistry 112.)
As one of the topics covered in Biochemistry 112, I learned about protein folding from Don and Steve, which was, to me, the most fascinating problem. I'd never heard of the folding problem before. But although this was a great problem, I didn't see any way to do experiments. So, I came to the protein structure by thinking, "Well, I don't know how to understand the protein folding problem, but if I know what the final product of folding looks like, maybe we can understand protein folding." And that turned out not to be true. But that did get me into protein structure, and I ended up, my second term, rotating in Steve Harrison's lab. Steve got me interested in crystallography. I still have his handwritten notes where he wrote out books to read as an introduction to crystallography At the time, Steve was working on the structure of tomato bushy stunt virus. While rotating in Steve's group, I realized that all these things I loved to do were contained in crystallography. However, solving the structure of an entire virus seemed like it would take a lifetime (it actually was solved in less than 5 years). So I wasn't so excited about virus structure. Of course, now, we're really interested in virus structure. [laugh] Not every decision you make when you look back at, you say, "Yeah, I really nailed that one."
As part of my work in the Harrison group, I was scanning the X-ray films recording the high-resolution diffraction data from crystals of tomato bushy stunt virus. The film scanner was in Lipscomb space, and I'd scan films in the evening when they weren't using it. Lipscomb was often in his office in the evening, and I would walk by his office to use the film scanner. And he would tell me about all these projects that they were working on. One of those projects was carboxypeptidase A, one of the first structures solved in the US. Carboxypeptidase A was a zinc enzyme, and since I had taken several inorganic chemistry courses as an undergraduate and had been interested in metals and spectroscopy, I thought, "Maybe I should work on metalloproteins."
I had met Lipscomb once before, as it turned out, through an unusual set of circumstances. Lipscomb's sister Helen had been Becky's (my wife-to-be) piano teacher in Kentucky when we were growing up. And unfortunately, Helen Lipscomb had died I think in her 40s of polio. I took a year off after college when Becky and I got married, and she had to finish college at the University of Kentucky. Helen passed away during this time and Lipscomb came to play his clarinet at her memorial. And so, I went up and introduced myself. I had applied to Harvard, but wasn't interested in working with him since I had no interest in structure at the time. But ultimately after getting to graduate school, Lipscomb got me interested in metalloprotein structure. And at the end of my first year of graduate school, when I decided I wasn't that interested in viruses, I moved from Steve's group to Lipscomb's group. When I would talk to graduate students at Caltech as Dean of Graduate Studies about changing groups and related issues, I know it's traumatic because this was something that I went through.
After joining Lipscomb's group, I needed to find a research project. The main project at that time was the structure of an allosteric enzyme aspartate transcarbamoylase, that also contained zinc. This project had been initiated by Tom Steitz, a Lipscomb graduate student who subsequently won a Nobel Prize for his work on the ribosome. After Tom graduated, Don Wiley took over the project and he made great progress determining the low-resolution structure. On the basis of this structure, Don got a faculty position at Harvard, and then the next crew of Lipscomb graduate students took over. As was often the case at that time, getting a structure at low resolution was hard, but not like the challenges of getting the high-resolution data and solving the structure at atomic resolution. And the people in Lipscomb's group were just stuck and worked like crazy for years. Both the graduate students and Lipscomb were really, really frustrated at the slow progress. It was not a good situation.
Given the challenges of the aspartate transcarbamoylase project and the number of students already working on it, I decided to work on carboxypeptidase that Lipscomb's group had solved earlier. I was attracted to this project since it was a metalloenzyme with a zinc ion at the active center. And again, in terms of thinking about catalysis and so on–there were a lot of controversies over how it worked, and we thought that by looking at how different ligands, inhibitors, substrates bound to carboxypeptidase, we might understand the catalytic mechanism. Although there were a lot of technical challenges with my project, I look back at that time fondly. I was able to learn a lot of crystallography from my colleagues, especially in the computer room. There was a slow pace, very different from today. It might take us a year or several years to solve the structure in what now might take an afternoon. While it's great to be able to solve projects faster, I really enjoyed the opportunity to learn crystallography. Graduate school confirmed that I liked the combination of biochemistry, chemistry, math, physics, equipment, computers needed to do crystallography. We could do some theory, even though we weren't theorists. And it just blended all these things together in trying to understand some basic process that's important for living systems.
ZIERLER: On to the definitional side, the word energetics, what does that mean in the context of protein structure?
REES: That's a great question. It's sort of like, "What does the word dynamics mean?" They're not so well-defined, in my opinion. I think structure has really benefitted from having a clear definition. "Here's the structure." Now, is there really the structure? Of course not. Things are moving around in their equilibrium positions, or sometimes there are far excursions, but we can define a structure in an average sense. And on average, we can say, "These groups form a hydrogen bond. They're 2.8 Ångstroms apart. This hydrogen bond stabilizes this intermediate species." I would say that's not completely well-defined, but at least we have a mental image of what structure means.
What energetics and what dynamics mean are more nebulous in my view. And of course, I recognize that I describe my program as structural bioenergetics. I see bioenergetics as trying to understand the fundamentals of energy metabolism in a living system. It can include things like how membrane potentials or ion gradients are set up across membranes. It can also mean how ATP, sometimes called the energy currency of the cell, is used as a stable, portable form of energy. And so, the part of bioenergetics I really think about is, how is this ATP molecule used to drive unidirectional processes? We've focused in my research on these two basic systems –nitrogen fixation and membrane transport. While they may seem completely different, they share the common feature that both systems couple protein structural changes to the binding and utilization of ATP. The important property of ATP is that it can be hydrolyzed (attacked by water) in a highly energetically favorable process to form the products ADP and water. If you have an energetically unfavorable process, you can make it energetically favorable by coupling it to an even more favorable process (such as ATP hydrolysis). The role of the protein is to make this process unidirectional by suppressing the back reaction or side-reactions that would occur in the absence of ATP. The protein goes through a series of conformational changes coupled to the binding and hydrolysis of ATP that provides this directionality. I find these conformational changes just completely fascinating. As one example, we are working on nitrogen fixation, which is a hard chemical problem, in part because there are a number of other potentially reducible substrates in a cell, most notably protons, which can be reduced to hydrogen, that would essentially short-circuit this whole process. So one of my working hypotheses for why biological nitrogen fixation is so complicated is that not just the chemistry of reducing the N-N bond, but you also can't be reducing anything else in the cell. I think the role of ATP basically is to design the electron transfer pathway in such a way that electrons end up to in ammonia where they need to be, and not some other species. The ATP makes the process unidirectionally by transiently providing a connection for electrons to go from one place to another. And once the electrons get to their destination, then the system gets unplugged before the electrons can flow to the wrong place. It's the role of the ATP to facilitate the "plugging-" and "unplugging" so you can get this unidirectional translocation of electrons. We've been interested in understanding the role of ATP in these processes, and so that's where the bioenergetics piece comes from.
ZIERLER: Finally, to get a sense of the overall structure of your lab and the major questions you're asking, where does metalloproteins occupy its own distinct world, and membrane proteins occupy its own distinct world, and where is the overlap that gets to those overarching questions that you've been after in your career?
REES: At one level, where they overlap, I guess, is in my head. The biochemical technologies that you need in these two areas are distinct. In the case of nitrogen fixation, the purification is old school biochemistry, before the use of over-expression technologies, with the twist that the proteins are oxygen sensitive. As a postdoc, I worked in a biochemistry lab with my postdoctoral advisor, Jim Howard, who got me interested in nitrogen fixation when he was on sabbatical at Lipscomb's lab. After talking to Jim for most of that year, he offered me a postdoc position in his lab at the University of Minnesota. I thought "Nitrogen fixation sounds great. I'll work on this." I didn't have this grand, highly developed plan for my career, but I liked Jim, nitrogen fixation was a metalloprotein, and it also involved oxygen-sensitive protein which added some experimental complexities.
Membrane proteins provide a distinct set of experimental challenges, beginning with that they are designed to work in a membrane. And so, you have to always have detergents or some membrane mimetic with membrane proteins, since otherwise they precipitate and aggregate. We had not previously worked with detergents, and so we had to adapt and develop some of the biochemical technology for using detergents to solubilize and stabilize membrane proteins. We also needed to work out how to over-express membrane proteins since the amount of space that's available for membrane proteins in a membrane is just much smaller than the cytoplasm. The technologies for using detergents and over-expression systems for membrane proteins is quite distinct from those we needed for nitrogenase.
One general connection between the projects we have studied is that at the time I started working on nitrogenase and membrane proteins, there was little structural information available about them. You would say these were target-rich projects. Membrane proteins constitute roughly a fourth of the proteins encoded in genomes. And when I started, there were no membrane protein structures. Anything you worked on was interesting. But there was a reason why few people had worked on these projects and that was there were also high technical barriers to their study. So that was another piece that made these projects interesting to me. These were just problems where a structure would provide a great foundation for a field of knowledge that had grown up without any structural information because of technical challenges.
At the time I started working on nitrogenase and membrane proteins, I had this complete clarity of vision, or at least I thought I did, that these were important problems to work on. I wish I could identify those new fields today.
ZIERLER: And just for fun, given how fundamental protein structure is to understanding life itself, more broadly, in a research life steeped in protein structure, what might your expertise yield in greater understanding for the origin of life here on earth or maybe even in an astrobiological sense as excitement builds in exoplanet research?
REES: When I was starting out, my view of the origin of life originated with the Soviet scientist Oparin, who envisioned a primordial soup. This was my main vision as to how life might originate. I think about this problem now, and I don't know whether to admire or just have sympathy for people who are tackling these problems. There are so many things that represent high barriers to overcome. Again, I'm sort of at the 50,000-foot level. But what you see as the critical step, really, I think, reflects what you're most interested in. For Oparin, and Miller and Urey, it was making some of the basic metabolites. Wachterhauser proposed that life started on catalytic iron sulfide surfaces, which I like because they remind me of iron sulfur proteins. The RNA school focused on information metabolism. Jack Szostak has emphasized the role of membranes so that there is the inside and the outside of a system, raising the questions of how molecules get across the membrane. The part that I'm most interested in is at what point did fixed nitrogen become an essential nutrient? Trying to work out how you get information transfer, compartmentalization, protein synthesis, etc. is not for the faint of heart. I don't have a problem at all that this happened since billions of years are a long time for life to originate and evolve. But it is hard to imagine which came first between the chicken and the egg.
ZIERLER: And on that, we'll pick it up for next time.
[End of Recording]
ZIERLER: OK, this is David Zierler, Director of the Caltech Heritage Project. It's Wednesday, September 29, 2021. Once again, I'm so happy to be back with Professor Douglas C. Rees. Doug, it's great to see you. Thank you for joining me again.
REES: Yes, it's my pleasure to be here. I'm looking forward to today's discussion.
ZIERLER: Last time, we did a wonderful grand tour of your approach to science and your appreciation of all of the advances in your field in historical context. Today, we're going to go all the way back for you in your personal narrative to the beginning. Let's start, first, with your parents. Tell me about them.
REES: My father was Earl Douglas Rees, but was also known as Doug. He was born in Cleveland, Ohio in 1928. And my grandparents, his parents, never went to college, and I don't think they finished high school. My father was born was right before the start of the Depression and grew up in Findlay, Ohio with his sister, my Aunt Annie. It seemed like they had a fairly challenging, but in hindsight, happy childhood. My father was really interested in sports and science. He went to Harvard and majored in biochemistry, as well as played on the junior varsity football team. My mother, Mary Alice Klingensmith, was born in Parkersburg, West Virginia in 1929. On her side of the family, there was a much longer tradition of higher education. My maternal grandfather had a PhD. This was also a fairly scholarly or academic family, but they clearly enjoyed a lot of different activities. From Parkersburg, they moved to Athens in southern West Virginia, where my grandfather was the principal of Athens High. It seems like a fairly remote area, but curiously, at the same time, the neighboring city of Bluefield, West Virginia, John Nash, of A Beautiful Mind, was also growing up. Nash would have been a contemporary of my mother and her brothers, my Uncle Dave and Uncle Walt. Unfortunately, by time I realized this, I wasn't able to see if they had ever met since they must've been in high school at the same time. My mother went to Radcliffe. Her brother (my Uncle Walt) went to Harvard and was my father's roommate in college. This is how my parents met. Of course, I wasn't there at the time, but I guess love blossomed.
After my father graduated from college, my parents got married. My mother still had a year of college to finish. My father started at Yale Medical School, so my mother took the train from New Haven up to Cambridge for a year, graduated, and then they started a family. There were four of us in five years while my father was finishing medical school. Four in five years. My mother was wonderful, very creative, imaginative. In hindsight, I don't know if she quite expected that this was what she would be doing at such an early age, taking care of all of these children while my father was finishing medical school. My father did finish in 1954, and then he spent one year as a postdoc in the Yale Chemistry Department, working with S.J. Singer. Singer had previously been a Senior Research Fellow at Caltech with Pauling, working on the project that led to the identification of sickle cell hemoglobin as a molecular disease. Singer worked out that there were chemical differences between the wild-type hemoglobin and the sickle cell hemoglobin, which they realized was due to a change in the amino acid composition. In hindsight, I think that time in the Yale Chemistry Department as a postdoc was one of the most stimulating, enjoyable periods that my father had. There were giants in the Yale Chemistry Department then, including Lars Onsager, who later won a Nobel Prize in physics, and John Kirkwood, who moved to Yale from Caltech to serve as the chair of the chemistry department.
Another postdoc at that time who worked with Kirkwood was Ignacio Tinoco, who subsequently became an eminent biophysical chemist at Berkeley. Apparently, he and his wife would babysit me when I was growing up. Of course, I don't remember any of this. My father loved his time at Yale, and he really liked his research in the chemistry department. His research with Singer involved studying the properties of proteins in non-aqueous solvents. Singer was one of the pioneers in establishing the physical chemistry of membrane and membrane proteins. So this was, I guess, always something in the back of my mind, that membrane proteins were really fascinating proteins and ultimately would be great proteins to do structural work on.
I should also say my father had been in the Reserve Officers Training Corp while in college and medical school and so then served in the Army. From New Haven, we moved to Rockville, Maryland, where my father worked at the Armed Forces Institute of Pathology. Subsequently, we moved to Chicago, where he was (I believe) a lecturer at the University of Chicago, working with a scientist, Charles Huggins, who won a Nobel Prize for hormonal treatment of prostate cancer.
In 1960, the medical school opened at the University of Kentucky, and my father got a faculty position in the Department of Medicine. So we moved to Lexington in the summer before starting 3rd grade. My father had clinical responsibilities, but he also had a lab. I would always want to go in and work in his lab. I also had a lab at home where I could also do experiments. Probably not as dramatic as some of the ones I hear from my colleagues in terms of the smoke, and smells, and explosions. But still, it was something I was always interested in. And I think, certainly just thinking back in a nonscientific way about that particular time, purely driven by an accident of birth, with the launching of Sputnik and the realization that we needed, as a country, to build our scientific and engineering infrastructure, there was a lot of emphasis on science. Teaching curricula in high schools were reinvigorated through the Physical Sciences Study Committee, the Biological Sciences Curriculum Study, the Chemical Bonding Approach Project and New Math. So the ways these topics had been taught were being revised in exciting ways. Chemistry had a lot more emphasis on chemical bonding and molecular structure. Of course, I never did the control experiment to know if these concepts helped or hurt my scientific development. [laugh] But it was just an exciting time to be interested in science. I was interested in astronomy, I started grinding a mirror for a reflecting telescope, which I never finished. In fact, I still have the mirror here in my garage. I hope to finish it someday. Computers were just starting to really take off. The University of Kentucky had a group where we could go over on Sundays and learn how to program on an IBM 360 computer. These sorts of interests were encouraged.
Since I was interested in science, I didn't need any encouragement. I had really great teachers at Tates Creek High School, and wonderful classmates. Reflecting on the opportunities that we had, I think Kentucky, in terms of their public education, was quite progressive. I felt like I got an outstanding background in the fundamentals. My parents were loving and supportive, and worked hard to keep us balanced. We would go camping around Kentucky, including several epic and memorable trips to visit some of the great National Parks in the western US, and canoeing in the Boundary Waters in northern Minnesota. I played different sports in high school. We all learned to play the viola. There were four of us playing the viola. And in this case, it was really driven because the boys, my brother and I, were expected to play football. I think there was some concern that if we played a band instrument, this might provide an escape route from football. So we ended up playing the viola, but that opened up an interest in music. At the time, it was mostly classical music. I was completely uninterested in rock ‘n roll and blues in high school, to my later regret, thinking of some of the bands I could've seen if I'd been more interested.
I was really fortunate to have an interest in science, for whatever reason, from an early age. My father loved books; my mother loved books. We would go down to the University of Kentucky bookstore, and bookstores would have these great technical book sections that we would go through. The books were $1.50 or $2.50, 50-plus years ago, but I still have some that have become classic texts in various fields.
ZIERLER: On both sides of the family, how many generations back can you trace your family in the United States?
REES: Well, it really varies. It's possible to do it farther than I will be able to reconstruct off the top of my head, because both of my parents had people in their families, aunts, grandparents, and so on, who provided this information. On my father's side, again, it depends on what branch. I'd say that they generally came over later than on my mother's side. My grandfather was born in the US, and I think his father was as well. But some of his relatives would speak German at home and he learned the language as well, rather to my surprise when I realized decades later that he could speak German.
On my mother's side, there are branches that go back to the 1700s in the US, and I have the genealogy on my maternal grandmother's side that even includes Richard Henry Dana of Dana Point as a distant relative. So that side of the family had been in the US a lot longer. But that's something I guess I need to go back over all the notes I have in storage to try to reconstruct this. On my mother's side, there was much more of a tradition of going to universities and higher education than on my father's side.
ZIERLER: So your father, in his pursuit of higher education, was more of a pathbreaker in his family, in a sense, than your mother was in hers?
REES: Absolutely. For what my father did, it's hard for me to appreciate because it wasn't a question of whether we were going to college or not. But for him, especially coming out of the Depression, for a family with very modest means, it must've been difficult. He ended up getting a scholarship to Harvard, which I'm sure is what made it all possible financially. On my mother's side, she had two older brothers. Uncle Dave went to the Coast Guard Academy, and Uncle Walt went to Harvard, where he was roommate to my father, and then my mother went to Radcliffe. Uncle Walt ended up staying in West Virginia, while Uncle Dave moved back there after he retired from the Coast Guard. And my mother lived in Kentucky, at least while we were growing up. They were very much small-town people.
ZIERLER: Given how important research was for your father, do you have a sense of why he went to medical school and did not stay on the biochemistry PhD path?
REES: Yes, I believe it was all because of the Depression and the paramount importance of having an employable job skill. When I was in college, my rebellion, such as it was, was not being pre-med. Of course, I loved all the courses that you take in pre-med, chemistry, biology, and all this, but I just wasn't interested in medical school. And I think my father felt this was a big mistake. Eventually, he felt it wasn't a mistake, that things had worked out fine. But it was, I think, the experience of going through the Depression and having to have an employable skill. Being a doctor was seen as, I guess, about as secure an employment skill as one could have.
ZIERLER: Now, did he gear his research toward clinical or therapeutic applications? Or did he operate in a basic science environment himself when he had the opportunity?
REES: His research was geared towards cancer, but he eventually had little time to spend on research with his other responsibilities. To me, one of the most unexpected developments–this was after I'd left home–was when he left the University of Kentucky and went into private practice. For someone who had loved universities so much–he'd been chair of the Academic Senate of the University of Kentucky –he just got tired of academic politics and so on. He was doing mostly clinical work anyway, so he went into a private practice, which he really enjoyed. His research work involved looking at chromosome abnormalities due to exposure to compounds found in tobacco smoke. Which, for various reasons, was a focus in Kentucky.
ZIERLER: That's tobacco country, and I'm sure a lot of people smoked.
REES: A lot of people smoked at that time. Interestingly, the University of Kentucky has a smoke-free campus, which Caltech does not have. [laugh] But my parents split right after I started college, and my father moved out to a farm, where he grew tobacco, ended up having a tobacco acreage. And my wife and I ended up living out there for a year and were involved in overseeing this process. His research focused on some polycyclic aromatic compounds that are found in tobacco smoke and how they could lead to chromosome abnormalities. His research work involved several staff, I don't think many students. But while he was at the University of Kentucky, he always had a research project, and I think doing research was always something that was important to him. And again, what I recall was just how he would talk about his time as a postdoc at Yale. And he clearly just loved this experience.
ZIERLER: Did your mom reenter the workforce after the kids were older?
REES: Yes. This is one of the complicated things. I sometimes wish you could go back in time and try to better understand things that were going on. Certainly, I would say, the social contract at that time, overgeneralizing, was that the husband was the breadwinner, and the wife took care of the kids, which my mother did. And as I mentioned, my parents separated when my father wanted a divorce. This was right after I went off to college and it was a highly traumatic and from my perspective, a completely unanticipated, development. So my mother found herself in a very difficult, I would say, situation. And she ended up getting a teaching credential, then teaching I think elementary school until she retired. So she did go back to the workforce. But I think this was partly motivated by financial necessity.
ZIERLER: Where are you in the birth order?
REES: I am the oldest.
ZIERLER: And did your siblings pursue scientific careers as well?
REES: No. I've decided, based on a very non-systematic survey, that having a scientist for a parent definitely has an influence on your career, but the sign can vary. [laugh] My sense is that with my interest in science, I ended up checking that box for my father. My siblings have had very interesting careers, but not in science; my brother Dave was in the Army, my sister Kathy is an artist and my sister Jennie has worked as a journalist in horse-racing. So, all very different.
ZIERLER: Would your dad allow you to just sort of roam and explore in his world? Did he encourage you along those ends? Were you the instigator of wanting to go? How did all of that work for you?
REES: I don't remember the specifics, but I do remember times when I wanted to go into the lab, and I obviously had to be supervised, and it wouldn't work, and I would be distraught. I wouldn't be able to go into the lab. But he could also bring chemicals and some glassware home and would get me books on chemistry experiments that I could perform, including crystallizations of various salts. One experiment I still remember was isolating chloroplasts from spinach and measuring oxygen evolution when they were illuminated. For this, I could use manometers that he had in his lab. His interest in manometers was related to the Warburg effect, where cancer cells have a high respiration rate and end up burning a lot of glucose. At that time, the way one would measure the consumption or production of oxygen was with a manometer. It was really the ideal gas law in action. I was able to look at oxygen evolution upon illuminating chloroplasts. I don't think I could do those experiments now, because it was old technology and I don't have that specific equipment.
We moved to Lexington because the Medical School opened at the University of Kentucky. We moved into a new subdivision, and a lot of the people there were also at the University of Kentucky. One of my neighbors, Prof. Richard Schweet, had been a postdoc in Biology at Caltech. I would mow their lawn in the summer, and we would have these discussions about nitrogen fixation, which is the first time I remember thinking about nitrogen fixation. This is while mowing the yard. [laugh] Dr. Schweet was an exceptional biologist who made a big breakthrough in protein synthesis at the City of Hope after he left Caltech. It's one of the things that really struck me about the people I worked with at the University of Kentucky, what outstanding scientists they were. I've come to realize that a lot of what we're able to do at Caltech is because we have access to these incredible students that are actually doing the work. But there are a lot of great people in our universities. I also ended up working several summers on lipid biochemistry in another lab at the University of Kentucky with Prof. Robert Lester, who had been a graduate student in Biology at Caltech.
From these experiences, I was getting the general vibe about Caltech as a special time in the lives of our neighbors, friends, and acquaintances. The Schweets lived a block away and Mrs. Schweet was an amazing photographer. And their house, to me, just seemed very artsy. It made me feel like, "Wow, this must be California." It also gave this impression that there was something special about Caltech and California. But at the same time, it just seemed like things were crazy in California. Lexington was not at the cutting edge of the political developments in the 1960s.
Although, maybe not as much as I subsequently realize. Lexington's schools were effectively segregated when we moved there, since the neighborhoods, which the schools were based on, were still essentially segregated. You don't think of Lexington in the same breath as Birmingham, Alabama and so on, but there was a lot going on that was just not on my radar growing up.
ZIERLER: Did you have a sense of what your parents' politics were when you were a kid? Would politics be a matter of discussion at the dinner table? Kennedy-Nixon debates, the Vietnam War, as it was ramping up? Did you have a sense of where they landed on these kinds of issues?
REES: Well, they were both Democrats. One of the really fascinating things is that Kentucky was a hugely Democratic state when we lived there. It was a long time ago. The miners, the unions, it was heavily Democratic. And my parents were Democratic. I remember that they took us to see John Kennedy when he visited Lexington during the 1960 election campaign. My father's father had been a union organizer at an oil refinery in Ohio. And that certainly influenced us in that trying to provide a more equitable society was an important part of their values. And to me, this was probably most reflected in our religious upbringing. I didn't realize until later that there was probably a lot more religious divergence between my parents than I knew at the time. My father, I think, was not very religious, whereas my mother was a lot more religious than I appreciated. And the compromise was that we went to the Unitarian church in Lexington, which certainly provided a more liberal environment for our religious education. Part of the values of the Unitarian church were supporting civil rights and what we would now call social justice. I don't remember how much of the discussion was talk and how much of that was action. But certainly, these were important values.
At that time, a lot of Democrats were supporting the Vietnam War as well. You could see the generation gap because my father supported the War in Vietnam. My brother ended up going to West Point and serving in the Army. My father had been in the Army, and my grandfather had been in the Marines after he'd left high school. And I don't know how far back that goes, but also not having any military service was an unusual part of my background for my family.
ZIERLER: Growing up, given the amount of academic heritage that you had from both Harvard and Yale, where did they rank in your mind hierarchically? Where was the lore? Was one considered more prestigious than the other? And did that influence your decisions when it came time to think about those things?
REES: I recall applying to four colleges: Yale, Harvard, Case Western, and Caltech. So I did apply to Caltech. But my father had loved his time at Yale. He'd been an undergraduate at Harvard, and he was an undergraduate right after World War II ended. At that time, there was a huge influx of returning veterans. So the student population was actually quite large. And I don't know how much this influenced things, but I would say his overall experience at Harvard was not that positive. Actually, he'd wanted to be a veterinarian. And you could imagine the reception one would have, talking to your advisor at Harvard and saying, "Oh, I want to go to vet school." So he was steered toward medical school. As it turns out, my father really enjoyed his time at Yale. So, I think in hindsight, if I got into Yale–that was my role in this process–I think there was little doubt that I would go there. That would've been a huge act of rebellion, to not go there if I had the chance. Certainly, I now see how complicated these rankings are. I don't feel like I can really rank schools now. But I ended up feeling like I'd been fortunate in my choice of schools.
ZIERLER: Between your father's lab and your own high school courses, were you always sort of going to look between biology and chemistry? Or were you going to follow your father's interests in biochemistry? Was that the plan from the beginning at Yale?
REES: I guess so. My major was molecular biophysics and biochemistry (MB&B).
ZIERLER: And where was biophysics at that point? In other words, famously, at a place like Princeton, biophysics wasn't real physics. Was there that sense at Yale as well? Or that was not the case?
REES: I think that was just a general phenomenon. In fact, a lot of places, biophysics would be in the medical school, as a Department of Physiology and Biophysics. And they might be interested in things like muscle contraction, or propagation of electrical signals. The MB&B program, I can't remember the history now, but it did involve a merger of different departments. If I have my history correct, there was someone (Pollard) in the Yale physics department who had been interested in viruses and biological physics. But in general, I would say this was something that was not seen as real physics but was more involved in the application of physical methods to the study of biological systems.
ZIERLER: Let's start on the social scene. You arrived at Yale in 1970?
REES: It was September of 1970.
ZIERLER: There's so much to discuss here. You mentioned that your schools were effectively segregated, although you were largely oblivious to it as a younger boy. What was the Civil Rights scene like, both on campus and in New Haven, the town and gown relations as they related to race?
REES: Well, it's complicated. I should qualify this and make it clear that I was pretty oblivious to a lot of these issues, even in college. I talked to my contemporaries who grew up in Los Angeles about their public school experiences and my public school education, the Vietnam War, drugs, music all these things. Kentucky, to me, seemed to be several years behind LA. I went to college, and I had never had a drink before. And with one exception, I'd never traveled on my own before.
ZIERLER: So this is a pretty sheltered existence up to this point.
REES: It was very sheltered. And, you could argue, since then as well. [laugh] I can't remember if I flew to New Haven or to New York, but most of my stuff for college was in several trunks that went by train to get to New Haven. And they would take a week to show up. So I arrived at Yale at night with a suitcase, it's dark, finally find the entryway to the residence hall I'm in, and it turns out one of my roommates is drinking beer. And it was just like, "Whoa. [laugh] Where am I?" There were a lot of new things to deal with.
In the spring of 1970, while I was finishing high school, there had been in the New Haven the trial of Bobby Seale, a Black Panther, that had brought in a lot of people and protests. Yale ended up closing the campus and canceled classes several weeks before the end of the semester. And so, this, then, was also in the background. Of course, they'd reopened by the fall when I arrived. And this is one of the things it'd be interesting to go back in time and revisit. My very superficial sense was that as of the spring of 1970, students wanted to change the world, but starting in the fall of 1970, all of a sudden, students wanted to go to professional schools and make money. It really just seemed like a switch had flipped. I participated in one anti-war march. But even that seemed to be dying out. And where I could really see this was in the pre-med population. There weren't so many pre-meds when I was starting. But by the end, this became a large program. And it really seemed to me that that period was a transition between social change, the War, and Civil Rights to where more people were thinking about themselves.
ZIERLER: Was your program a dual major in molecular biophysics and biochemistry?
REES: No, it was a single program. I think there had been probably a biochemistry program and a molecular biophysics program, but they merged. So it was one degree.
ZIERLER: Had that merger been a recent development?
REES: It probably had been in the 60s, but I don't recall.
ZIERLER: Who were some of the key professors for you as mentors, intellectually introducing you to new things at your time at Yale?
REES: Ironically, it wasn't so much in the MB&B program. As it turns out, one of the real strengths of the MB&B program was in structural biology. But at the time, I had a negative interest in it. I came out in college with this negative interest. I didn't have a lot of close interactions with many MB&B faculty. The two most formative interactions, one was with Prof. Carolyn Slayman, who was in the physiology department at the Medical School. I worked in her lab starting as a freshman and stayed there throughout my entire college career.
ZIERLER: And what was the overall framework of her lab? What were some of the major research questions that were going on?
REES: I got into her lab because my TA for introductory biology was a graduate student named Alan Lambowitz. Alan was an MB&B graduate student. As I recall, he gave us a choice between doing some project or doing research. [laugh] To me, that was a no-brainer. I ended up working with him for a bit, and then he graduated. He's at UT Austin now and has had a very distinguished career. That's how I got into Carolyn's lab. It wasn't because I went around saying, "These are the areas I'm interested in." She was interested in Neurospora, the bread mold that George Beadle eventually at Caltech and Edward Tatum at Rockefeller made famous; Tatum had been Carolyn's thesis advisor. There are mutations in Neurospora that lead to respiratory defects, and she was interested in trying to understand the basis of some of these defects. What I was doing was not really biochemistry, but it was measuring respiration rates of Neurospora suspensions with oxygen electrodes. So, I've now evolved from manometers to oxygen electrodes that would measure the respiration rate. And then, I would add different inhibitors that block the respiratory chain at different places, trying to figure out where different mutations might be influencing respiration. Carolyn was also working on various ATPase proteins that would have fit in well with the sort of things I subsequently did. But my particular project was the one I just fell into, in a lab that I happened to have the great fortune to work.
Carolyn was an incredible mentor. At the time I joined her group, she was an assistant professor, one of the first women hired in the basic sciences in the Yale Medical School. She was subsequently appointed as head of the department of human genetics, the first woman to head a department at the Yale Medical School. While I was there, she had children, I believe was coming up for tenure, and always seemed calm and unflappable. I think now about all of the pressures she must've been facing, but as an undergraduate working in her lab, I was completely oblivious to this. Occasionally, we would be discussing one of my projects, and I'd have a proposal for how something might work. In hindsight, I realized that she really thought my ideas were way off-base, but she had a very gentle way of pointing out that maybe we should be thinking about other explanations for what's going on. So for research, it was a great experience.
Looking back at my time at Yale, you realize these random events that you could never predict would have an impact on one's career. I had wonderful instructors in high school. But there's a limit to how much math and other topics that you actually cover in high school. I'd had a year of calculus, but we didn't cover series. And the first year of calculus in the math department at Yale went through series. But because of the closure of the Yale campus due to the Bobby Seale trial, they never covered series in first year calculus during the Spring of 1970. And as a result, they had to start in the Fall 1970 with series, in what would be the third semester of calculus. So I was able to get into what would've been a second-year math class because they had to cover the material I hadn't had before. That turned out to be important because then, I'd finished the calculus classes after the fall of my freshman year. My father really loved to look through college catalogues, so he would know more about the courses being offered at Yale than I did. He discovered an applied math sequence that was being taught in the engineering program by Prof. George Veronis, who was in geophysics.
That applied math sequence was really transformative for me, and George ended up being a second mentor for me. He started the geophysical fluid dynamics course at Woods Hole and taught it for over 50 years. And this was the first time he had taught this particular class to undergraduates at Yale. It was a sequence of three semesters that covered more series, differential equations, and complex variables, which was the coolest thing I'd ever seen.
ZIERLER: Why were they so cool?
REES: Well, it was like discovering magic. I thought the conformal mappings were completely fascinating – but sadly I've forgotten most of what I had learned. With conformal mapping, you can take a plane where you can solve a fluid flow problem, and then you transform it into an airplane wing, and then you can transform the flow so you can, through this sort of magic, take these really complicated problems and make them simpler, though they wouldn't be for me now. And it was just something I had never seen before. Just really cool. So George and Carolyn, I would say, were the two biggest influences I had in college. I've always wanted to find a problem I could solve with conformal mapping. I've talked to Rob Phillips–this is one of my goals maybe for the next few years, relearn this stuff and find some problem I can solve with it because it was just really amazing.
ZIERLER: Did you stick around New Haven for the summers, or did you go back home?
REES: No. I look at our undergraduates now who spend summers on campus doing research. I have positive memories of college, but it was stressful. I never ever considered working over the summer there. I would go back and work as a lab technician at the University of Kentucky. Mainly, when I was in college, I worked summers in the lab of Robert Lester, who had been a graduate student at Caltech and was interested in lipid metabolism. We would purify various types of lipids from baker's yeast, where we'd get these big blocks of yeast and extract these lipids. It was a really interesting group of people that I got to know over the summer. I never considered spending the summers in New Haven. I was happy to get back home.
ZIERLER: Where were computers, even in their primitive form, in their undergraduate years?
REES: I hardly used computers at all. All the problem sets were with slide rules. And one of the things that I do remember in the science library was a very primitive calculator. So you would go to the library just to use the calculator. All of the papers were written on typewriters. I do remember getting some instruction in Basic, perhaps for a problem set. But computers were not yet having an impact on the curriculum or even really in research, at least at the level that I was doing things.
ZIERLER: When did you really start to feel that pressure about medical school or not medical school?
REES: That was in the freshman year. I actually did declare as a pre-med my first year, but that was it. I never followed up.
ZIERLER: That made your father happy, at least at that point?
REES: I guess. I turned 18 when I was in New Haven, so I had to register for the draft in New Haven, and it may have been about the same time I signed up for pre-med. But I don't remember if there was anything where you would opt out, or if you'd just quit, how it all worked. Because I don't remember saying, "I'm not going to be pre-med," at least officially. But I never ended up pursuing that.
ZIERLER: What kind of advice did you get about graduate programs from your professors? And was there a culture at Yale that it was better to move on, or that for some select students, it was best to stay?
REES: My interests were so diffuse that there wasn't one particular field where I said, "This is what I'm going to do." As a result, I applied to Caltech in the biology division, to Harvard in biophysics. At this point, I was interested in getting to California for whatever reason. I applied to Caltech, I applied to UCSD, and I applied to Berkeley, the latter two in chemistry. It was the end of the Vietnam War, there were a lot of changes in funding. I think it was a bad time if you were a graduate student to be looking for jobs. One of the programs I applied to at Berkeley basically said–I have to see if I still have this letter–"Do you really want to go to graduate school? The job market's not good." There was the sense that the job market was terrible for graduate students.
So I applied to graduate schools in the Fall of 1973 during the my senior year of college. But then, several things happened. My wife-to-be, Becky, though we weren't married at the time, was at the University of Kentucky, and her father worked for IBM. Her parents initially moved to Lexington when IBM moved from Kansas City to Lexington. And then, her parents were going to be transferred to Boulder, Colorado in 1974. We met in seventh grade and had been dating on and off since high school, and it sort of seemed, in hindsight, I guess we decided if we didn't get married, we'd never see each other again. Of course, there's some middle ground. [laugh] It's not like we just disappear. But we decided that we really wanted to get married. But Becky still had a year of college at the University of Kentucky. So I decided to take a year off. They weren't called gap years then, but I took a year off. Of the four schools I applied to, the only one that would defer my admission a year was Harvard. The biophysics program at Harvard was incredibly accommodating. Having that sort of assurance that I hadn't completely cut off my career options was really important to me–my father was worried about my game plan. I took the year off and Becky and I got married. I worked at the University of Kentucky with S.K. Chan on plasma proteins while Becky finished her undergraduate degree in education.
ZIERLER: At this juncture, it's so important to delve a little deeper into this negatively defined interest in structural biology up to this point. What's the story there? Was it something that you specifically did not want to get involved in until a later point? Or where were you with structural biology at this point?
REES: Well, I definitely had a negative interest. Of course, like a lot of things, I didn't have any experience with what I was talking about, but the impression I had from my classes was that crystallography just seemed like a lot of mindless work, collecting data, screening crystals for diffraction quality, etc. This didn't seem nearly as interesting as doing real experiments and getting real results. The irony is that Yale did have a great structural biology program, including Tom Steitz, who got his PhD with Lipscomb and ended up winning a Nobel Prize. Tom was at Yale then as an assistant professor doing really exciting things. But from what little I knew, I thought crystallography just involved the mindless labor of turning a crank. What's the fun of this?
And in some ways, I have to admit that impression was pretty accurate, but one of the things I subsequently learned was I actually like that sort of thing. Anyway, that was all part of my thinking. When I was applying to graduate schools, again, I wasn't quite sure what I was interested in. I applied to biology at Caltech. There was crystallography in the chemistry department, but I didn't apply there. The protein crystallographer at Caltech then was Richard Dickerson, who was one of Lipscomb's first graduate students after he left Caltech and went to Minnesota. But I had broad interests, and so my choices of graduate schools weren't really driven by a particular area of interest, but more that a lot of different types of research were going on.
ZIERLER: Was it always biophysics, all the programs you applied to?
REES: No. I remember while we were living out on my father's farm at rural Kentucky, I talked to a faculty member at one of these schools about my interest in low-temperature chemistry. This involved studies such as calorimetric measurements of residual entropy at absolute zero. My interests were all over the place. So it wasn't just biophysics. I don't remember how biophysics got on my radar in the first place because the biophysics program at Harvard was not designed for people like me. It was really designed for people coming from a physics background to get into biology. It was started by A. K. Solomon at the Harvard Medical School who studied red blood cell biophysics involving a lot of work with radioactive tracers to measure ion fluxes across membranes.
The biophysics program was really quite remarkable, and Prof. Solomon must have had supernatural administrative abilities to get it approved through both the medical school and the Cambridge campus. The program let you work with a range of faculty on both campuses and required six course that weren't specified. But the most important feature, which I didn't realize at the time I applied to graduate school, was that it also had you do rotations. When I applied to graduate school during my senior year in college, Harvard biophysics was the only program that would defer my admissions. I really wanted to go to California, though, so, I applied again to graduate schools while we lived in Kentucky. I particularly wanted to go to Caltech since I'd been hearing so much about it from neighbors, the Feynman lecture books, Pauling's General Chemistry, Apostol's calculus book, etc. Just for the education in the basic sciences, Caltech had this aura.
But as it turned out, Caltech didn't admit me straight away, as they probably felt like they'd been fooled the first time. They ultimately decided to accept me after I was awarded an NSF graduate fellowship, and I had put Caltech as my top choice. At that point, then, I heard from the biology graduate option that they had admitted me. But, as I remember, that didn't leave a good impression on me. So I committed to going to Harvard.
ZIERLER: Coming into the biophysics program, as you say, it was geared towards physicists. How much physics did you have at Yale as an undergraduate?
REES: More than most molecular biophysics and biochemistry majors, but much, much, much less than you would as a physics major. MB&B was a fairly flexible program, as I recall, at least in some ways. I ended up taking the yearlong quantum mechanics course for physics majors, and I took the electricity and magnetism semester course for physics majors. These were for the undergraduate physics majors, not for the graduate students. Then, the statistical mechanics and thermodynamics were through chemistry courses. And I think about it, I was really fortunate with the instructors that I had, both in chemistry and in physics. The instructor I had for quantum mechanics was a theoretician named Richard Slansky. He was really interested in the conceptual and philosophical framework of quantum mechanics. And my mind just doesn't work that way. I could sort of see these things, but at some point, I guess I just like to do calculations. And that was certainly something that indicated to me, probably incorrectly, that a career in physics wasn't for me because I wasn't so interested in all these different interpretations, and paradoxes, and related topics. I didn't really understand at a deep level the problems, the issues, or how to resolve these issues.
ZIERLER: What was the interface in the biophysics program with the giants at Harvard in physics at the time? Steve Weinberg, Shelly Glashow, did you have any interface at all with these people?
REES: No. I sat in on one course that Nicholaas Bloembergen, who subsequently won a Nobel Prize, was teaching on electricity and magnetism. But this was a graduate course, and one of the things I learned in graduate school was, especially for the advanced courses for advanced graduate students, was these were not for dilettantes. These were for people who wanted to be the next Weinberg or whoever. They weren't really intended for someone like me, who just wanted to increase my exposure to certain fields.
ZIERLER: It sounds like the biophysics program was pretty self-contained in that regard.
REES: Well, in hindsight, I would say that this was a biophysics program, and not a program between physics and biology. I don't really recall that the physics department was playing any role in the administration of this program. In fact, for a long time, that was my perception of biophysics. And it actually wasn't until I was at Caltech, and we had established a biochemistry graduate program, and we wanted to change it to Biochemistry and Molecular Biophysics. And with the name change, it was like, "Wait a minute. This option has physics in the name." [laugh] But at that time, I don't think physics, from my little, tiny corner, seemed like had any ownership of the biophysics program.
ZIERLER: Do you have a clear memory of when you first met Bill Lipscomb?
REES: Absolutely. In fact, several. But the very first time I met Lipscomb was in an unusual situation. While growing up, Becky took piano lessons from his sister, Helen Lipscomb., and Becky's mother was good friends with Lipscomb's mother, Edna Lipscomb. Becky's mother would be telling me, "Oh, I have this friend whose son is a chemist at Harvard." And it's like, "Yeah, right" [laugh] I would be a little skeptical of some of the things I would hear. It all turned out to be true, that Lipscomb grew up in Lexington, he was known as The Colonel, he went to college on a music scholarship at the University of Kentucky. His sister, Helen, was an accomplished musician and composer. Sadly, she contracted polio and was confined to a wheelchair. Becky would go to the Lipscomb home for piano lessons with Helen and as a result became acquainted with Lipscomb's parents. After we got married and were living in Kentucky, Helen passed away, and there was a memorial service for her at the University of Kentucky. And The Colonel, Bill Lipscomb, came and played at it. By that time, we had already decided to go to Harvard. After the memorial service, I went up and introduced myself to Lipscomb and said I was going to the biophysics program. Lipscomb, at this time, was not in recruiting mode and commented "There are a lot of people to work for".
ZIERLER: And what was his primary appointment?
REES: In chemistry. Purely in chemistry. In fact, biophysics was a graduate program, so it was not a department.
ZIERLER: So everybody there had their home departments.
REES: Exactly. And one of the things that is typical about many graduate programs is that you can only work for certain faculty who are affiliated with the department that administers that program. But the way the biophysics program was set up, it was only a degree-granting program, and not a department, and you could work for nearly anyone. That was incredible flexibility for someone like me who hadn't decided yet on what I wanted to spend my life doing. Lipscomb was in chemistry. That's where I first met him, in that context.
ZIERLER: What were the big ideas? What was exciting in biophysics at Harvard at that time? What did you have to choose from as you were thinking about narrowing the research for what would ultimately become a thesis?
REES: When I started graduate school, I was interested in was membranes – and still am. There wasn't a lot of membrane research going on at Harvard, but I did find one group. We had to do rotations and I picked my first rotation working in this membrane group. I think we were on the semester system, so maybe these rotations were half of a semester. But most biophysics graduate students took a course on protein structure and function that was taught by Steve Harrison and Don Wiley. And they were in the biochemistry department in the Cambridge campus, but they'd both been biophysics graduate students. Don had worked for Lipscomb. And Pamela Bjorkman, my colleague, ended up working for Don.
So I took this course, Biochemistry 112. And I took another course in membrane biophysics. The rotation working on membrane proteins didn't really click. But this course, Biochemistry 112, was like no course I had ever seen before. It was really transformative for me, just like complex variables. But I'll try to hold it together describing Biochemistry 112. Instead of developing a field in a systematic way, which even now, I tend to do, it basically covered a set of topics. The lecture might provide some background, but there'd be these really long problem sets that involved a lot of reading of the original literature. I would produce 10-15 handwritten pages as the answers to each problem set. One of the topics that was covered was the protein folding problem. Somehow, I'd never heard of this before, or at least, it hadn't registered. And it just struck me as one of the most fascinating problems there is. How can you take this long polymer, add water, and it folds into this well-defined, very specific structure that confers on the protein the ability to perform all these remarkable activities?
ZIERLER: What was the history of protein folding up to that point? In other words, who had these original ideas of adding water and recognizing the remarkable reaction that happened?
REES: Again, like a lot of things, there are a few people who end up being recognized as the heroes of this story. And what I will say will be reflecting that as well, which, of course, is a huge oversimplification of what was going on. The structure of a protein reflects the end of the folding process. How do you get from the initial point to this final stage? I think for a lot of people, probably, they didn't even appreciate it as a problem. I think just to take the highlights, one was a Danish physical chemist–we now would say biophysical chemist–Linderstrøm-Lang at the Carlsberg Institute who pioneered the use of H2O-D2O mixtures to look at the exchange of Hs for Ds in proteins. He noticed that there were different rates of this exchange. Some hydrogens didn't seem to exchange, while others did. Lang recognized that this implied that the native structure of a protein wasn't like a rock, but there must be some dynamics that allows some of the Hs to exchange with solvent protons, but not others.
The classic experiment that really suggested that the intrinsic ability of proteins to fold was encoded in the protein sequence was the work of Christian Anfinsen using the protein ribonuclease. A lot of the early work was based on proteins like ribonuclease, because you didn't have over-expression systems, had to be done on proteins that naturally occur in large abundance. For example, proteins like hemoglobin from blood, or myoglobin from these diving mammals, or lysozyme from eggs. Another great source was digestive enzymes from cows. So from slaughterhouses, you'd get these pancreases that make a lot of the digestive enzymes. Some of these were proteins like carboxypeptidase, which I ended up doing my thesis on, and another was ribonuclease. Ribonuclease ends up having eight residues called cysteines that are able to form these stable, under some conditions, covalent linkages with each other called disulfide bonds. These eight cysteines form four disulfide bonds in a unique partner in ribonuclease. And you can break those bonds. If you just look the probability of eight of these that formed four pairs, there are 105 different ways at random that you can join them together. And the thought was, if you break those bridges, and then you just let them reform, the active enzymes should be about 1%, one out of 105, if this was all done randomly.
And of course, that's too simplistic because the probability of forming these bridges will depend on how far apart the two residues are, the polymer physics part of it. But the upshot is, if it's just a random sort of thing, a very small fraction of the protein, if you tried to reform these bridges, would be active. In fact, what Anfinsen found is 100% recovery of activity. And this showed that the native state of the protein must be the most thermodynamically stable form. And it's interesting because other people had done similar experiments, but didn't appreciate the significance of this result. Fortunately for Anfinsen, the Nobel Committee thought it was absolutely pathbreaking.
This, then, led to a lot of interest in how this folding process might occur. There were various models. Maybe it's like a ball of yarn that starts folding up in the middle, and you keep adding more yarn (peptide) around the outside of the ball. The first structures showed, however, that wasn't true because you would have certain structures, where you have two parts that might be close together in the sequence, and then one from a very distant end that would be synthesized at the end, and yet these were all together. Or folding could be like a jigsaw puzzle where there are many different ways of assembling a jigsaw puzzle, but you might start primarily with the edges and gradually fill in recognizable features in the interior. Or maybe these analogies are all irrelevant for how proteins really fold. To me, protein folding seemed like such an interesting problem, but I really couldn't think of any experiments to understand how proteins might really fold.
Most folding experiments at that time involved unfolding the protein, and then measuring the rate of refolding, so these were kinetic experiments. They didn't really tell you anything about the molecular mechanism. Now, when I started hearing about protein structure in this Biochemistry 112 class, I thought, "Maybe if you understand the native state, you would understand how it was acquired during the folding process." Unfortunately, that's not true either. But I then rotated in Steve Harrison's lab and realized that all these things I thought I would find completely off-putting in crystallography, I really enjoyed. I liked collecting data. We got to work with computers, we got to work with proteins, we got to play with equations. And Steve was very patient in guiding me through crystallography. I started realizing that there was a lot you could do with diffraction. And that completely changed my thinking and I realized I really liked crystallography, and more generally, structural biology.
I was lucky because, of course, I had not picked graduate school on the basis of where exciting structural biology was taking place. In retrospect, I couldn't have made a better choice. Even at the time I started graduate school it wasn't clear, because the groups at Harvard were all stuck trying to solve these really challenging problems. It's just as I was finishing graduate school that all these projects came together, primarily through the work of Don and Steve, and I couldn't have been at a more exciting place. So I made the choice for graduate school that I think was the right choice for in some sense the wrong reasons, and also not having done any control experiments. But you have to have some reason for making a decision. Whether things work out or not often depends on things you just didn't know at the time you make the decision. To me, this was almost an example of–I don't know why this came to me after our last session–directed serendipity. I went to Harvard for certain reasons – namely the flexibility of the Biophysics program and the general range of research that was being conducted, and not because there was great structural biology research.
That era of structural biology at Harvard starting in the late 1970s was really exciting. Don and Steve were really driving this transition. Lipscomb, of course, played an important role in the evolution of structural biology in the US. He made the transition from doing small molecule structures like boranes and other small molecules, and realizing with the same equipment, you could be doing protein structure. And this impressed upon me the power of structure, that with the same experimental approach, you could be working on small molecules, big molecules, you could be doing all sorts of things. Lipscomb did not have to retool his lab to start working on protein crystallography.
The first protein structure solved in Lipscomb's group was carboxypeptidase A. And then, they started working on the aspartate transcarbamoylase enzyme, which was an allosteric enzyme. Carboxypeptidase is a digestive enzyme. You give it a substrate, and it will start chewing off amino acids. That's it. It just works. Aspartate transcarbamoylase is a much more mysterious system in that it carried out this important reaction, making a precursor for nucleic acids. But sometimes, you need certain nucleic acids, other times, you don't. So there was this whole elaborate regulatory mechanism or allosterism, where the enzyme could be in either an inactive or active form. And Don Wiley, as a graduate student in Lipscomb's lab, got a low-resolution structure of aspartate transcarbamoylase. Don got a faculty position in Harvard, same building, no postdoc. Lipscomb's graduate students continued to work on aspartate transcarbamoylase to get a high-resolution structure. And it was a really hard problem, a really dark time.
After my first year of graduate school, I starting working for Steve. Steve was interested in viruses, plant virus structures. Of course, we now know how important virus structures are. But at the time, it seemed to me, "It will take forever to solve the structure of a virus." And I would talk to Lipscomb, having to use his film scanner as I mentioned previously, and Lipscomb would say, "Oh, we're working on all these projects, etc etc." And I was thinking, "Enzymes. That sounds interesting." He was also interested in metalloenzymes. From my undergraduate chemistry classes, I'd taken some spectroscopy classes, and the spectroscopy of transition metals always seemed fascinating to me. So I thought, "Oh, metalloproteins, that sounds interesting." I ended up switching into Lipscomb's group, but I decided not to work on aspartate transcarbamoylase because they were just having all these problems. So I started working on carboxypeptidase, which no one was working on in the Lipscomb group at the time, and ended up figuring out how various ligands bind to it. In hindsight, it was a pretty pedestrian project. But it was a great opportunity to really learn crystallography. Because of the computer room, I would get to spend a lot of time with the other graduate students and postdocs in the crystallography labs and also in the computational groups. I ended up learning a lot from that whole environment.
During that time, I had a birds-eye view of the structure determinations of some incredible structures from the Harrison and Wiley groups. In my view, this work transformed the field of structural biology from a somewhat esoteric field to an indispensable tool for solving important biological and biomedical problems. A major driver of this transformation was the graduate work of Pamela Bjorkman, who was a few years behind me in graduate school. I served as her TA in Biochemistry 112, and we subsequently TAd this class together. Pamela's structural work truly transformed our understanding of how self- versus non-self are distinguished by the immune system. Her structural findings revealed a groove where peptide antigens could bind to be presented to the rest of the immune system. Her structure illustrated in molecular terms how this recognition could occur, so that biologists, immunologists and biomedical researchers then were like, "We have to have crystallographers doing structure." It was no longer just chemists and physicists, but the biomedical community, who wanted these structures. And to be in that environment and see what was going on at the time was just incredible.
ZIERLER: Between simple intellectual maturity and advances in instrumentation, where does structural biology hit home for you in graduate school in a way that it never could in undergraduate?
REES: Obviously, I didn't know much about crystallography as an undergraduate. There is this very mathematical part of crystallography, which should have appealed to me because of my interest in applied math. Initially, though, it just seemed like there were these equations, but they didn't connect with me in any way. I mentioned the conceptual framework for quantum mechanics went right over my head. I found some similarities between that and learning crystallography – I could read about crystallography, but it was pretty abstract until I started working in the lab and could actually see what happened when crystals were put in an X-ray beam. At that time, we were using film as an X-ray detector and it was a much slower process. We'd have the crystal in the X-ray beam exposing the film, and after developing the film you'd see this set of spots. We'd be trying to align the crystal in a certain way, realizing, as you rotate the crystal, the pattern changes in a precisely defined way that took advantage of a lot of these mathematical concepts, vectors, basis sets, all these things. At any particular orientation, only certain reflections would be observed, and if you move the crystal, then those reflections change. Just realizing that you could make all these predictions and could actually calculate what you would see when a crystal was placed in the X-ray beam was eye-opening. What does reciprocal space mean? It seems like an abstract concept, yet you could use that framework, move the crystal, see these changes, and it was all in a completely deterministic way. And I think just seeing that sort of connection, in some sense, made it all real. And frankly, crystals are just fascinating. You look at these crystals, and they're just so cool. I don't know how to explain it. I think it was really that.
Now, to solve a crystal structure, you needed to measure not only the intensities of the diffracted reflections, but you also needed to recover the missing phase information and that was a big deal. Solving the phase problem would take years. But there was also, I realized, sort of a rhythm to this process. I'd come in in the morning, develop a film, put a new crystal on, expose it for 10 or 12 hours, develop that film, put on another one, etc. You'd sort of get in this groove, and there was a lot of time that you could just spend learning crystallography. And it turned out I liked that pace. I liked having time to think about things. Not many people were doing this, we thought it would take a decade or more to solve structures. So it wasn't, "Push, push, push. I'm going to get scooped. We've got to get stuff out now," that is a part of life now that to me, interferes with the enjoyment of the process.
ZIERLER: What was Lipscomb like as a graduate advisor? Would you work with him one-to-one? Would you have more formal pre-scheduled meetings? How did that work?
REES: It was interesting. Lipscomb was surprisingly available, given what I now know about faculty schedules. He was around quite a bit. At the end of my first year in graduate school, I joined his group. Within a month, he had won the Nobel Prize in Chemistry.
ZIERLER: Was the buzz there? Was that sort of out of left field?
REES: From my perspective, it was completely out of left field. There were three of us who went in that morning and told Lipscomb that he had won the Nobel Prize. Now, that was the story I believed for a number of years. I couldn't quite figure out, though, why Lipscomb happened to be in so early that morning. Of course, now I know he knew. He came in to work and was waiting for people to show up and tell him he had won the Nobel Prize. I heard the news in a roundabout way in those pre-internet days. I was in an office with two other people. It must've been before 8 o'clock. One of the others was also a biophysics graduate student, Paul Kuttner, and we were good friends.
The third person in the office took a phone call. And just listening to the one conversation, because there's one phone in the office and you could hear these things, it was clear he was talking about the Nobel Prize. But he didn't say anything to us. And I said to my friend Paul, "Who won the Nobel Prize?" Paul was always a man of action. Of course, now, you would already know these things. But he calls the Boston Globe and asks, "Who won the Nobel Prize in Chemistry?" And they tell him, "Professor Lipscomb of Harvard University." So, we run into Lipscomb's office and say, "Colonel, congratulations. You've won the Nobel Prize in Chemistry." And his response was, "Are you sure?" Now, knowing what I now know, I would never have gone into his office until I had triple- or quadruple-checked this–because if you tell someone they'd won the Nobel Prize, and they haven't, it's the only thing worse than not winning it. But at the time, it's like, "Well, my friend called the Boston Globe, and they said yes." I can't remember what we said. But fortunately, it was true.
This was a very exciting time, celebrating the occasion and watching the interactions. I remember one professor who I'd had for a class, really distinguished, but who didn't win a Nobel Prize for work related to enzyme mechanisms. Should have, in my opinion. And I'm just watching as he asked Lipscomb what he'd won the Nobel Prize for. Which, again, shows Lipscomb's versatility that there would be a question asked like this. Lipscomb felt there were three things he could've won the Nobel Prize for. One was boranes, one was protein crystallography, and I recall the third was boron NMR. Lipscomb replied "boranes". And I remember just being struck by what seemed like this sense of relief from his colleague that the Nobel Prize was for boranes and not for enzymes. Now, I realize, his colleague must have been thinking, "I still have a chance." If Lipscomb had won it for enzymes, then maybe he wouldn't have a chance. Again, these are things it'd be fun to go back to see, knowing what I now know about a lot of these things and all the undercurrents. The group went out to lunch that day to celebrate at Legal Sea Foods, which was then in Inman Square, as I recall, and sat together at a long table in a noisy restaurant. It was really quite exciting.
Lipscomb was gone a fair amount that year. Otherwise, he was around quite a bit. In general, Lipscomb did not have regular meetings. Now, I go on my Google Calendar and make appointments to see my students, but Lipscomb would often be available. In fact, he would frequently, several times a day, walk through the lab and find out what was going on. He would typically whistle, which we could pick up. My office was in one of these old buildings, so if I wanted to talk to him, I could stay there, and I had a few minutes to prepare what I was going to say. But if I didn't want to talk to Lipscomb, I could escape through a back door and wait. And one door was to the dark room, so we could always go in the dark room and close both doors. Then, we were pretty safe.
In hindsight, Lipscomb did an amazing job of keeping apprised of what we were doing. But what I found fascinating was I don't really ever recall looking at primary data with him. He'd ask what was going on, we'd discuss it, then we'd go over drafts of papers. I think at some point, he made a decision whether he trusted you or not. I heard this described as follows: when you first joined his group, you start out at the top, and everything's great. And then, you do start working and encounter problems, and you drop to the bottom of the list. Over the next few years, you spend your time working your way back up the list. When you get back to the top, then you graduate. Lipscomb had a remarkable record of mentoring, training about 100 graduate students and maybe that many postdocs. Before my time, I understand that Lipscomb had been completely hands-on with research and working with the group; that wasn't so much true when I was in the group, but he was really available for discussion.
ZIERLER: And that didn't change after the Prize?
REES: No, sort of surprisingly, he didn't travel so much. He did go on one sabbatical. He was a big believer in sabbaticals to learn new things. His sabbatical was in Munich towards the end of my time in graduate school. I pretty much at that point knew what I needed to do. I just had to get it to work. And of course, at the time, I was thinking, "Well, Lipscomb's gone, and now I'm making more progress." I don't like to think about that now concerning my role in my group. Sometimes, I do think advisers get in the way of students making progress.
Lipscomb had really broad interests – in many ways, he was a one-person science department. I had encountered some disorder in my crystals, and I was trying to model the disorder in various ways to see what effect it had on the intensities of the diffraction pattern. I'd calculated these different distributions that described the intensities, and one involved hypergeometric functions, which I had just learned about. I mentioned this to Lipscomb, thinking that I would finally have something that he didn't know anything about. To my astonishment, he started telling me all about them based on his time as a graduate student at Caltech, taking an applied math course I believe from Erdelyi. Hypergeometric functions arose in the context of certain integrals, and at that time, your ability to do these types of integrals was really paramount to being able to solve many different types of problems. Having programs such as Mathematica do them for you was not an option, like today. And Lipscomb remembered all this stuff. So I don't think I could ever come up with something about crystallography that he didn't already know. There was one time where I did ask him "Where do you think this inhibitor binds to carboxypeptidase A?" I actually had collected the data and so I knew the answer. And in that case, I did stump him and could tell him, "Here's where it binds." But in crystallography, he was just incredible.
ZIERLER: Between all of the data-taking and analysis, when and how did you decide that you were ready to defend? Was there a specific goal or benchmark that you wanted to reach, at which point you knew you could write the thesis?
REES: I think about this now with my students. It's not like when you're an undergraduate, and you have completed so many units and satisfied these requirements of taking these classes, you're done, you get your degree. The PhD thesis is really different, and there's a window where I think it's acceptable to defend. But it's not well-defined. And I think in some sense, my friend Paul, who I mentioned previously, and I decided that we had completed an acceptable amount of work and were ready to finish being graduate students. I had solved the structure of a complex of carboxypeptidase with an inhibitor protein, but I hadn't yet finished the refinement. I had a good friend Mitch Lewis who was a postdoc with Steve Harrison, and we had found that we could collect, for the time, really high-resolution data on carboxypeptidase. So we were working out a lot of refinement methods as well. There were just a lot of projects going on, and I think in the ensemble, there was sufficient work for a thesis. Probably the main thesis project itself wasn't quite finished, but there was enough work that was completed for an acceptable thesis. I finished the remaining work as a postdoc with Lipscomb.
One side of Lipscomb that was truly generous, which I think also reflected Pauling's influence, was that there were certain projects I was doing that Lipscomb let me publish on my own. So I actually had a fair number of papers as a graduate student. I think Lipscomb generally enjoyed having students publish projects on their own. That said, another part of me always wondered if he wasn't sure if those papers were correct, and he didn't want to be associated with the papers if there was some problem - but I don't really think that was the case.
ZIERLER: To the extent that there was an overarching conclusion to all of the papers that would come together to form your thesis, what was that conclusion, and how did you see that in terms of your contributions as you're thinking about your next steps?
REES: Don't get me wrong, I'm proud of what I did for my thesis, but I would never call it transformative. The main lesson, which was not really original, was that you could use crystallography not only to solve structures, but to start looking at substrate complexes as a tool to study mechanism. This wasn't original, but I was able, with carboxypeptidase, to determine the structure of a number of complexes of substrate-like molecules, to understand how they bound to the active site. I also included my analysis of the effective dielectric constant of proteins I mentioned earlier, as well as my treatment of the crystal disorder that involved hypergeometric functions. Ultimately, what I learned from the latter, besides the fun of deriving intensity distributions, was that the best way to deal with disorder is to eliminate it. Which actually was a good thing to learn, but it didn't build on my analyses.
I really distinguish my thesis from someone like Pamela Bjorkman, whose thesis was truly transformative in the way that immunologists viewed how the immune system works and how they viewed structural biology, that just had a tremendous impact. There wasn't a big, global paradigm shift in my thesis. From my perspective, I was able to learn a lot of crystallography, and I just got really excited about the field. And those were probably the most important things that I got out of graduate school.
ZIERLER: Last question for today, who was on your thesis committee besides obviously Lipscomb, and was there anything memorable from the oral discussion or defense?
REES: There was Steve Harrison, Don Wiley and Frank Westheimer, a physical organic chemist whose thesis with Kirkwood at the University of Chicago originally solved the dielectric constant problem I mentioned earlier. What I most clearly remember was going into Lipscomb's office. Lipscomb had been in Munich on sabbatical, and there was the Fasching celebration where people wear costumes. Apparently, one of the typical costumes was a little hat with horns. So I walked in, and there's Lipscomb with this hat, which looks sort of like the devil. I didn't know if this was a good sign, a bad sign, an omen. Lipscomb had a really great sense of humor and was, I think, just trying to lighten things. But of course, you're pretty stressed when you go into these things. There wasn't a formal seminar, just a discussion with the committee members. As a result, I had never given a research talk by the time I'd finished graduate school. It was the way the system worked.
ZIERLER: This continues the theme of your sheltered life.
REES: And also, I think, in the chemistry graduate program, there may not have even been a committee. The advisor decided when you finished. There was a committee in Biophysics, and we had a thesis defense.
ZIERLER: It turned out OK, obviously.
REES: It did turn out OK, yes.
ZIERLER: Well, that's a great place to pick up next time, where we'll start in on your post-graduate career.
[End of Recording]
ZIERLER: OK, this is David Zierler, Director of the Caltech Heritage Project. It's Friday, October 1, 2021. Once again, it's my great pleasure to be back with Professor Douglas C. Rees. Doug, great to see you again.
REES: Yes, well, good morning. It's great to see you, and thanks for going through all this.
ZIERLER: Absolutely. So the year is 1980, you are a newly minted PhD thinking about your next opportunity. What comes next for you?
REES: Like a lot of things in my career, I would say that my post-graduate plans were not well-thought-out in advance. As I recall back to that era, I didn't have a strong sense of what I wanted to do when I grew up. I think that was fairly typical of my cohort of graduate students and postdocs. Partly, I would say, this was because the academic job market at that time was really quite tight. For the people who were a year or two ahead of me, there were zero academic jobs in structural biology.
ZIERLER: If I can interject on that point, obviously, the late 1970s, early 1980s is a time of economic malaise during the Carter years, but specifically in your field in terms of the funding sources, can we assume that this is a tight budget environment for NIH and NSF specifically as it relates to opportunities that would come your way?
REES: Honestly, I wasn't paying a lot of attention to funding at that level. I think there were at least, from my recollection and vantage point, maybe two effects going on. One was that the first structures were determined in the early to mid-1960s, the myoglobin structure by John Kendrew and then hemoglobin of Max Perutz. And there was a lot of excitement over those structures, followed by lysozyme by David Phillips, carboxypeptidase in Lipscomb's group, and so on. And that led to an initial set of hiring of structural biologists. But I think as those crystallographers were starting, the community as a whole was maybe thinking, "This is a fairly limited technique, it takes decades to solve structures, and you have to have a lot of a particular protein to do protein crystallography." Many proteins of biological interests were not available in large quantities and hence could not be studied by crystallography. This was really an assessment of the future of protein crystallography, that it was sort of a niche technique. I also think after the end of the Vietnam War, there was a lot of funding that was going into the physical sciences that if not ended, really started to dampen down. So it was a tough job market, and crystallography was a field that I think was seen not to have too much of a future.
ZIERLER: In terms of coming right off of your PhD, to the extent that there was an academic niche that you were looking to fill in various departments, where were you focusing your efforts in terms of lab opportunities, teaching opportunities, that kind of thing?
REES: Again, it would be interesting to go back and sort of see if what actually was going on fits in with my recollection. My recollection was that I didn't have a particularly strong sense that I would only look for academic jobs. I did go to a couple of industrial lab interviews, which were disasters. We didn't have career development offices, and I think I just went in thinking that they would be telling me about how wonderful their companies were. And instead, of course, they wanted to know what I was interested in, and I was completely unprepared. Not surprisingly, I did not get those jobs. [laugh]
ZIERLER: What kinds of labs? Pharmaceuticals? Materials? What were you looking at?
REES: As I remember, one, for sure, was GE. And I think they were more sort of physical sciences, so not pharmaceutical. Pharmaceutical companies, I don't think, had any interest in structural biology at that particular point. This was more, again, like, GE. I don't remember if another was IBM or Bell Labs. [laugh] Whatever aspirations I might have had for industry, they were pretty short-lived. Of course, in a way, the default option was doing a postdoc. Lipscomb arranged for me to do a postdoc with a really great crystallography lab in Germany, which I had accepted, but I have to say, my wife and I were, I guess, not completely sure that this was the right move. I don't think, except for Canada, we'd ever been out of the country before. We'd never traveled or lived in another country.
So we were a little tentative about that. But also, research is pretty intense in terms of the time I was spending in the lab. And it wasn't really clear what Becky would be doing during this period if we lived in Germany. That was an issue that–again, we'd been married a few years–we didn't really know how to address. But that was the initial plan. And two things happened that really changed that trajectory, although I only realized this in hindsight. One was that Jim Howard from the University of Minnesota came to Lipscomb's lab to do a sabbatical. Jim was interested in the biochemistry of nitrogen fixation. And these are oxygen-sensitive proteins, and he brought a fair amount of equipment with him to work with these proteins in Lipscomb's lab with the goal of trying to crystallize one particular protein.
ZIERLER: If you can explain scientifically, what is an oxygen-sensitive protein? What does that mean?
REES: Proteins are sensitive to their environment, and there's a certain class of proteins that will react with oxygen and become inactivated. In our world, since we live in an oxygen-containing environment, our proteins are largely insensitive to oxygen. But there are a group of proteins, and they often have reactive metal sites that react with oxygen in deleterious ways. If you purify those proteins by the normal biochemical methods, which is working in the presence of an oxygen-containing atmosphere, then the proteins inactivate. There's a specialized technology to remove oxygen from all of the solutions you use and eliminate it from the purification to maintain the activity of the proteins. In biochemistry, this is a pretty specialized area. In chemistry, this is not as unusual, especially in inorganic chemistry, where people work with compounds that are reactive with oxygen or with water, or both. Because of this additional technology that one has to use to study oxygen sensitive proteins, they were generally not being addressed in most biochemistry labs, and especially in structural labs. Jim came to Harvard with all this equipment to work with oxygen sensitive proteins. When he arrived, I didn't have any idea who he was.
ZIERLER: Had he collaborated with Lipscomb? What was the Lipscomb connection?
REES: No, not previously. Jim is someone who really took sabbaticals seriously. And he used the sabbaticals as an opportunity to learn new technologies and to work in labs where he didn't previously have a connection. Jim realized that to advance our understanding of the biochemistry of nitrogen fixation, you needed the structures of the responsible proteins. He, then, decided to spend a sabbatical in a structural biology lab. I think his choice of spending a sabbatical with Lipscomb may also have been motivated because after Lipscomb left graduate school at Caltech, he started his career at Minnesota. And that's where his Nobel Prize-winning work on boranes was done, the University of Minnesota. Jim also liked to sail, and Boston is closer to the coast than Minneapolis. I think those were probably the criteria for his choice of sabbatical: structural biology, the University of Minnesota connection, and proximity to the Atlantic Ocean.
After Jim arrived, we started talking. We worked in an older building with an interior stairwell that went up through the middle with all these offices off it. So when I would go to my office, I'd often pass where Jim was, and we would just talk. And over the course of this, I really found Jim to be fascinating with a lot of interesting insights. He would tell me about nitrogenase, the enzyme responsible for biological nitrogen fixation, and it just sounded like an interesting protein. One day, he offered me a postdoc in his lab to learn how to work in nitrogenase and try to crystallize it. And I talked to Becky about this, and we thought, "Yeah, this sounds good."
ZIERLER: Better than Germany.
REES: Yes. It addressed several issues. We knew the language, I liked Jim, and nitrogenase sounded like an interesting protein. I'd been interested in metalloproteins. The carboxypeptidase protein I did my thesis on contained zinc. But I didn't have any strong ideas of what that might mean. So it sounded like this would be a good system, the structure might be interesting. And with that level of analysis, we decided to postdoc in Minnesota, not realizing that this was sort of going to launch me in a direction that would last for 40-plus years.
That was one thing. The other thing that happened, I believe while Jim was there, but I don't really have the chronology straight in my mind. I had mentioned there had been zero faculty positions in crystallography or structural biology. Sometime in this period, which would've been in late 1980, early 1981–after I defended my PhD - two job announcements appeared. One was at Johns Hopkins University, and the other was at UCLA, both looking for structural biologists. And this was really extraordinary because there'd been zero before. I think about some of my predecessors in Lipscomb's lab, who included some of the most talented crystallographers I've ever known, like Jim Crawford who I mentioned wrote this film scanning program. They ended up doing other things because academics was not an option.
ZIERLER: Do you think having two in one year was more a fluke or that departments were looking to increase their presence in structural biology because of its importance?
REES: Well, it was both. I didn't get an interview at Johns Hopkins, so I don't know what they were thinking, but I know UCLA really recognized that there was enormous future in structural biology. This was the vision of Dave Eisenberg at UCLA. And as part of this, UCLA, at this time, late '80, early '81, hired Richard Dickerson from Caltech to move to UCLA. Dickerson had been one of Lipscomb's first graduate students at the University of Minnesota and worked on the boranes, which was the Nobel Prize-winning work. He then went to Cambridge, England, where he worked with John Kendrew on the structure determination of myoglobin, which was the first protein structure determined, and also Nobel Prize-winning work. After a faculty position at Illinois, Dickerson moved to Caltech in the 60s to do protein crystallography. I obviously wasn't there at the time, but Dickerson ended up being extremely frustrated with what he saw as the lack of support for crystallography at Caltech. David Eisenberg had been a postdoc with Dickerson at Caltech before starting his independent career at UCLA. I don't know the whole timing of this, but UCLA offered Dickerson a position, and he moved from Caltech. Dickerson still lives not far from the Athenaeum.
Eisenberg recognized that there was a future in structural biology, and UCLA was actively pursuing this future. It was really, it turned out, the leading edge. Most universities were not willing to make this commitment, including Caltech. At that time, Caltech was completely uninterested in offering me or anyone else a job in crystallography. So UCLA and Hopkins had two openings for structural biologists. In our computer room, there were six or eight senior graduate students and postdocs, and we had a close group. We really enjoyed interacting with each other. And as I recall, I'm sure incorrectly, none of us had been thinking too much about what was next. One day, one of this cohort came in and said they were applying for these jobs, which I believe was advertised in those pre-internet days in the newsletter of the American Crystallographic Association. I recall we all sort of looked around and thought, "What are we doing about jobs?" I think most of us ended up applying for these two jobs. It led to this strange situation where we were really good friends, but now we were also, in some sense, competing for the same rare commodity. As it turns out, the person who came in and said, "There are these jobs," got the job at Johns Hopkins, and I got the job at UCLA.
ZIERLER: Before we leave Cambridge altogether, in what ways did your postdoctoral research put you on a different path than you might otherwise had been on had you gotten a job straight from the dissertation?
REES: Well, this was a combination of the two because I was still at Harvard. My work as a postdoc in Lipscomb's group involved finishing up the work I did as a graduate student. I had a postdoc arranged with Jim Howard at the University of Minnesota, but that had not yet started. So when I interviewed for the job at UCLA, I was still in Lipscomb's group, and I hadn't started working on nitrogen fixation and didn't know how that would turn out. Getting the position at UCLA was based on the work I had done as a graduate student. One other piece of this, before I interviewed at UCLA, Dave Eisenberg visited Harvard and gave a seminar. I ended up having half an hour to talk to Dave. And I really enjoyed the conversation. As graduate students, we didn't get a lot of experience–all these people would come to visit Lipscomb, but generally, he wouldn't introduce them to the group. I didn't have much experience talking to faculty, I'd never prepared a seminar before. But I do remember thinking how much I enjoyed talking to Dave.
I remember going out for the interview at UCLA and not being so interested in actually getting the position, since I was thinking I had a good postdoc lined up, which was paying an annual salary of $18,000, which was a lot of money at the time. As it turns out, not much less than what I made as a faculty member at UCLA. And I thought, "Eh, I don't really know if I want to be at UCLA." So I wasn't so stressed about the interview. It wasn't until I was flying back from the interview that I realized, "Wow, I really liked what I saw at UCLA, I liked my interactions with Dave Eisenberg, I liked talking to the other faculty there." And at that point, I really started getting nervous, hoping I hadn't screwed up the interview in some way. Because I had never given an actual talk before. I gave one practice talk before my interview that was just a disaster. I had read my talk, I was really uncomfortable, and Don Wiley took me aside afterwards and gave me, I would say, a very direct summary of all the things that I needed to work on for the real talk, which was enormously helpful. He also talked about some of his experiences where he had also had a disaster of a seminar visit which was very helpful for me to hear. So that was my preparation going to interview at UCLA. The one positive, looking back, was that I didn't feel that stressed about having to get this job at UCLA until I was returning and really realized how much I had enjoyed my visit.
ZIERLER: How much of this was simply butterflies in the stomach, how much of it was what we would call Impostor Syndrome, and how much of it was difficulty articulating what your contributions would be at this stage in your career?
REES: I think it was all these things. I'd much rather study scientific systems than try to understand what's going on in people's heads. I think Impostor Syndrome is something that many people experience, and it's a funny combination of being completely insecure coupled with at times being overly confident in one's abilities. It's a funny dynamic. And I've certainly experienced that range of emotions in my career. When I was in graduate school, there was little emphasis on professional training and the sort of skills that one would need to find a job or to be a faculty member. This is still, in a way, one of the odd things about being a faculty member, you get hired on the basis of your ability to do research, but at some fundamental level, that skill becomes almost useless or irrelevant compared to everything else you need to do to run a group. Especially working with a group and the administrative responsibilities. Certainly, the scientific vision is critical, but that's not necessarily the same as being able to do an experiment well. I had never thought about giving a talk, I didn't know what was involved. I'd been to plenty of seminars but never really thought about what was involved.
ZIERLER: And in this field, a talk means what? Is it the opportunity for you to sort of give a grand synthesis of your research up to that point? Or is it more forward-looking in terms of, "This is what I did, and this is what my research agenda will be if I join this community"?
REES: I think it's a combination of those two for a job talk. A lot of seminars are more like, "Here's the problem, and here's what we've done, and here's what we've learned." But in the job talk, one also needs to describe what one's going to do as an independent scientist. And in an odd way, I think the bar was pretty low at that point in time for doing that sort of presentation. In retrospect, an important consideration was the technical mastery of crystallography. Because a lot of the talk was spent on, "Here's how you solve the crystal structure." If you were to talk about that now, immediately, everyone would be thinking, "This interview's over." Things have changed so much. So at that time, the technical part of crystallography was seen as more of the challenge. A crystallographer would typically get proteins from a biochemist and then the craft of the crystallographer was to crystallize the protein, solve the structure, and then be the hero. I think places wanted to hire someone who understood crystallography. Again, because I had spent most of my time learning crystallography, and Lipscomb had let me publish papers on my own, I think I could demonstrate that.
At the time I interviewed, I hadn't started working on nitrogen fixation because I hadn't started my postdoc in Minnesota. My job application included a page or two of proteins I had started working on or had tried to work on as a graduate student. There was one copper-containing protein called laccase that was being studied in Sweden. Lipscomb had a connection, so they sent me some protein, which I never got anywhere with, but I think, "This is a protein that reacts with oxygen but is a substrate and has metals. Wouldn't this be interesting?" was sort of the level of analysis I went into, thinking about these independent projects.
In the summer of '81, I was still at Harvard, and UCLA gave me the offer. Becky and I discussed this. Becky had never seen UCLA. We thought about the offer, it seemed reasonable, and so I accepted the offer the next day without any negotiations. There was $15,000 for my startup, I think $21,000 for salary. I was looking forward to my postdoc in Minnesota, but by this point, I realized how much I enjoyed my interactions with Dave Eisenberg. So I thought this seemed like a great opportunity.
ZIERLER: Now, here's the question. On $20,000, can you buy a house in LA in 1981?
REES: No. Jim Howard had been a graduate student at UCLA. He had a wonderful time, but kept saying, "Do you know how expensive it is in Los Angeles?" He showed me the real estate ads from LA, probably the LA Times. I'd look at these ads, and my conclusion was, "Oh, they just don't list the cheap houses in the paper." So I was completely oblivious to what was involved in getting a house. I did get UCLA to agree that they should fly Becky and me out so that Becky could see UCLA and Los Angeles. I can't remember when this was, but we had our life savings with us, in the form of a $10,000 cashier's check. We just imagined going into the bank, depositing this, and then going out and putting a down payment on a house. I think we were in Los Angeles for five days, and the first thing they say at the bank is, "Oh, this will take two weeks to clear." [laugh] But then, we started realizing that housing was a major problem. But like many things, we were already committed by the time the ramifications of that came through.
ZIERLER: Now, on the family front, was Becky looking for a job? Were you looking to start a family at that point?
REES: Certainly, Becky would look for a job. We do have two children now but we didn't at the time. It's something we never really thought much about whether or when we would have children. So it wasn't something that was on our radar either way. But certainly, a job was on our radar, especially as the cost of living in Los Angeles became apparent. I think there's something to be said for being naive or not completely informed that you make decisions without knowing all the facts because I wonder, if we did know all the facts, if we would have done what we did. But I'm glad we did. When I started, July 1, 1982, as a faculty member, I did get temporary housing through UCLA, in student housing. UCLA had an apartment on Venice Boulevard, away from campus, where we'd gotten a one room efficiency. There was a school bus that would go from the Venice Berry Apartments to UCLA. We had one car, so I would take this bus with students to get to UCLA. Becky was teaching, and one of the odd things is that although she was credentialed in Kentucky, Massachusetts and Minnesota, she was not credentialed in California. So whereas I could teach in a state university without any teaching experience or credentials, she needed a credential. Through a contact of the Eisenbergs, she was able to get a job teaching in a private school in West Los Angeles. She needed to drive to that, so I would take the school bus.
Occasionally, we would look for housing, but it was pretty bleak, not only because of the price of housing, but also the interest rates at that time were on the order of 14%. UCLA had a program for new faculty in our situation that had a 12% rate. I was able to get one of these loans. We looked at housing and concluded that we wouldn't be able to find anything, even with this 12% loan. So I wrote to the dean and said, "Thanks, but we can't use this program." And to my amazement, he called me up. He wanted to understand why the program to help new faculty obtain housing wasn't working. Not that he could change any of the economics, but it impressed me that he was an administrator who was concerned. A related issue that I hadn't realized was the idea of a nine-month salary, which is what I had. I had never negotiated for a summer salary, or any of this stuff. I was oblivious to this. Given the housing prices, that summer salary was really important, and I had not planned on this. Another dean granted a month or two summer salary which was really appreciated. And then, the departmental administrative secretary realized the situation and wanted to know why I wasn't using my startup funds for summer salary.
ZIERLER: But the startup funds would be for building up the lab.
REES: That's right, so it probably wasn't startup funds, but I had other funding from starter awards. Maybe this was in my second year. There was some starter grant I probably received, and some of it, I guess, could be used for summer salary. But she realized that I did have access to funding that I hadn't realized could be used for this purpose, which helped out. Interestingly, I remember having discussions with other people asking if you could do this and getting the indication that you couldn't. So I realized that it was important on complicated matters to talk to more than one person. Because the first person may give you advice that's accurate but maybe not quite relevant to your own situation. So there were a lot of adjustments in moving to UCLA and life as an assistant professor.
ZIERLER: Was there a delay in actually setting your lab up? Or did you get to the work of building it really right away?
REES: When I look back at my experiences, and I compare this to the situation today, I actually think I was lucky to be starting when I did. Because the startup funds, let's just say, were modest. But what that meant was, you didn't have to spend a lot of time ordering all this equipment and trying to hire lots of people. I was given my lab space, and there was obviously not enough money to order a lot of new equipment. This wasn't anything particular to UCLA, but rather the way things were at that time everywhere. You basically made do with the equipment that was available. And in a funny way, you could actually focus on your science as opposed to building up the lab. Also, it seems to me that the expectations for what you are supposed to do have scaled with the startup costs. Startup costs have exploded. No one would get $15,000 startup costs now. It's probably at least two orders of magnitude more now for startup funds. But then, you immediately become a manager. My job was to scrounge for equipment. And again, everyone was doing this. Dave Eisenberg was doing crystallography, so we had the X-ray equipment at UCLA. It didn't always work, but I was used to that since, I'd learned how to fix X-ray generators as a graduate student.
The expectations in terms of what we were supposed to accomplish as an assistant professor, I think, were also modest. I'm not sure I would trade what I went through with what people have to go through now because people are expected to do incredibly impactful things. We all want to do incredibly impactful things, but this can take a while. I remember some of the projects that I was working on. I thought they were interesting, but you wouldn't call them earthshaking or pathbreaking. It was just, "I wonder if we could measure this," without some deeper biological significance.
ZIERLER: As you said in your job talk, the emphasis was on the technical capability of solving the structure and not necessarily responsive to those deeper or more fundamental questions. How much of that do you ascribe to your own personality or work style, and how much of that serves as a window into where the field was at that point in time?
REES: In a sense, I bridged the older way and the newer way of doing structural biology. My training was in crystallography, but then, I did my postdoc in a nitrogen fixing laboratory. It was not a crystallography lab. The postdoc in Germany would have been the traditional route where I'd go and work in a crystallography lab to learn, really learn, how to do crystallography. Instead, I went to a biochemistry lab, and I learned how to work with oxygen-sensitive proteins. In the course of my postdoc, I was able to crystallize one of the two proteins involved in nitrogen fixation. My postdoc at Minnesota ended up being about ten months, which again was more like the earlier way of doing structural biology, compared to today when postdocs can be 4-6 years or more.
During my postdoc, I learned how to work with proteins anaerobically, in the absence of oxygen. When I moved to UCLA, most of my lab setup was involved in assembling the technology for anaerobic biochemistry. I needed a glovebox and various manifolds to manipulate samples anaerobically. For this purpose, I needed a glass shop to make specialized glassware, and chemistry departments, at least at that time, always had great glass blowers. So my $15,000 in startup funds was used primarily to set up this technology. In hindsight, I spanned the era from where a crystallographer would do crystallography, and they'd get the proteins from other people, to now, where it's all about the system that you're studying, and everyone assumes you can do the crystallography.
I was fortunate when I started my position at UCLA that I had a specific project to work on, namely the nitrogenase protein I had crystallized during my postdoc. That allowed me to start writing for grants, which was important. I got great feedback from my colleagues at UCLA about research, teaching, running a lab, etc. I started teaching Introductory Biochemistry to 250 students, many of whom were pre-meds. I'd taken one biochemistry course as a junior in college, and so I was not well-prepared to teach this class. But many of my colleagues had taught the course before and were very generous with their lecture notes. In the biochemistry group at UCLA were a number of colleagues who really understood metabolism, including Paul Boyer, who would receive a Nobel Prize for his work in bioenergetics.
As much as I loved doing the crystallography, my motivation for the structural work was to understand how these proteins functioned in the grand scheme of things. I think my more biochemical colleagues could appreciate that attitude and were very supportive. I was fortunate to have outstanding mentors in my colleagues at UCLA. At Harvard, to overgeneralize, every lab was its own fiefdom. When I got to UCLA, the generosity was just incredible. One senior colleague, who wasn't even in my department, wanted to know if I needed any supplies because he would order them for me. Another had an instrument he was using and apologetically told me, "I'm sorry, you can't take this into your lab because we're using this, but you're more than welcome to use it in our lab." I had never experienced this before. It was really remarkable.
ZIERLER: You needed to be deprogrammed from the East Coast is what it sounds like.
REES: Absolutely. I haven't thought of it in those terms, but I think that's exactly what it was. It took me until I was well into my time at Caltech before I realized, "Wait a minute, I really like it here."
ZIERLER: Beyond your colleagues at UCLA–I know that crystallography is a pretty tight-knit community–were professional societies like the ACA important for you to be plugged into?
REES: The American Crystallographic Association did play a major role. Subscribing to Acta Crystallographica was important. The way I found out about the job at UCLA and probably Johns Hopkins was through the ACA newsletter. I had joined the ACA as a graduate student and subscribed to Acta Crystallographica. My goal was to get a paper published in Acta Crystallographica. At UCLA, besides the protein crystallographers, there was a distinguished group of small-molecule crystallographers. Ken Trueblood and Chuck Strauss, who had been very active in the American Crystallographic Association.
As much as I like interdisciplinary research, at some point, you have to have what I'd call a union card. There's got to be community that you feel like you're a part of. And for me, that was, at least starting out, the crystallographic community. You'd go to the Gordon Conference on Diffraction or ACA meetings, these sorts of things. There was more of a sense, which is quite different now, that although we were studying different proteins, we were united through our common interest in crystallography. This attitude to me is a testament to the power of crystallography and structure. Now, however, you need to be connected with the biological community that's interested in your structural studies, more so than going to the more technique based meetings.
ZIERLER: Did you take on graduate students right away? Or was that a gradual process for you?
REES: I did. I did pick up one student in the first year, and then several in the next year. And I think it's probably fair to say I learned more from them, in many ways, than they learned from me. Obviously, I had never had any experience mentoring. I don't think I ever had an undergraduate work for me. I'd always worked in the lab. And so, one of the things I think you end up doing is thinking everyone is sort of like you. My first graduate student, John Kornuc, had been to Ohio State where he majored in microbiology. And in his application, amazingly, it said something about nitrogen fixation. I wasn't at UCLA at the time, but he ended up working in my group, and he was great in the lab.
John loved to do preps, loved to set up crystallizations. And that was mostly what we were doing at the time. We finally got to the point where we had collected some diffraction data and I was starting to analyze it. I'd managed to work out how the molecules were packed in the crystal lattice. I was describing this analysis to John. I remember this so clearly - I was explaining how these molecules were packed, what this meant for the diffraction pattern, etc. And as I was droning on and on, I looked up at John, and I noticed he was looking at the window, like, 1,000 miles away. And it just dawned on me, "He's not interested in this." And it was a revelation because I'd always assumed, "Why would you be doing crystallography if you weren't interested in crystallography?"
I realized that people have their own skills and interests, and your responsibility as an advisor was to make sure their project can fit with their skillset. I could try to force John to do all this crystallography, but it would've been a disaster. He'd be unhappy, I'd be unhappy. But instead, he really loved doing experiments. Growing the bacteria, doing the purification, growing up crystals, doing all the sort of stuff that was really an incredible barrier to getting the structure solved. Go and collect data. The first time we went to a synchrotron was to Hamburg at the DESY source. John was just an incredible person to have along because he would work hard and get the job done. But he wasn't going to be figuring out crystal packing and writing programs. So that was, to me, a really important realization, that you can't just say, "You need to do this and here's how I want you to do it." Because you have to understand where they're coming from as well.
The next year, I got two students, Todd Yeates and Barbara Hsu. And they were both remarkable. Todd is now a faculty member at UCLA. When I was thinking of moving to Caltech, people would say, "Oh, the students at Caltech would be better." There are great students at both places. I've been really fortunate to have some remarkable students. So I now had multiple students, and they started out working on nitrogen fixation because that was the one project in the group. But then, I realized that while you learn a lot when you have your first student, you start learning even more when you get multiple students because there's not only the interaction between you and the student, but now also the interactions they have with each other. There can be some complicated dynamics when someone has been working on a project for a while and then a new student comes in.
I was looking for other projects, and I was doing all sorts of other things as well because I didn't feel at the time that I had to solve these great problems. Could we collect data and see where ions bind in carboxypeptidase? And could this tell us something about the electrostatic potential? These were problems we started and never finished, but looking back, I still think would be interesting pursue, especially now, with the accuracy we collect data with. There was a lot of learning and trying to figure out what I was interested in, in terms of projects and so on.
ZIERLER: An instrumentation question about the 1980s. I asked you this in our first discussion overall. But at that point, where was spectroscopy and microscopy vis-a-vis crystallography? Was the arms race already on at that point? Or was crystallography, in terms of the things that you were interested in, clearly best application to understand these things?
REES: From my perspective, there was no arms race. Crystallography was the only way. For projects that we were working on in my group and the others at UCLA at that time, crystallography was the only way in which you could get atomic information. Electron microscopy, at the time, was just not at all competitive for high-resolution structures. In fact, it was more at the level of blobs. So it didn't really seem like something you would ever imagine would be blowing crystallography out of the water for certain projects in terms of getting high-resolution structures.
ZIERLER: Maybe, again, if you had to give a job talk after you got the job, you would still emphasize the technical aspects. But in retrospect, at UCLA, what were some of the big questions you were looking to answer, given the fact that you had this ability to create your own lab and were calling the shots in terms of your research agenda?
REES: I would say, again, this is the serendipity part of this, I think I mentioned that I'd had this general interest in membrane proteins. But at the time, the biochemical challenge of getting membrane proteins for structural studies seemed impossible. They're generally not present in the high abundance needed for crystallography, and then, you had to use detergents to solubilize them out of the membrane. Unlike the proteins we usually work on, which are soluble in water, membrane proteins function in a greasy environment in the membrane. And to extract them from the membrane so that you can try to purify and crystallize them, we use detergent, just like dishwashing detergents, to extract the proteins out of the membrane.
And that was just a huge biochemical challenge. I knew membrane proteins were important, but I didn't even know how to start working on them. Then, out of the blue at UCLA in 1984, Dave Eisenberg and I got a letter from George Feher at UCSD telling us that a postdoc in his group, Jim Allen, had crystallized a bacterial photosynthetic reaction center. And he wanted to know if we'd be interested in working on it. It's like, "Would you like to win the lottery?" This was a no-brainer. It was just this incredible stroke of good fortune. George was a physicist who came from solid state physics, invented ENDOR and other techniques when he was at Bell Labs, but realized that he was interested in biological questions and that photosynthesis was a great problem for a physicist to be studying. After moving to UCSD, George started retooling himself to learn how to work with bacteria that carry out photosynthesis. George recognized that to understand how photosynthesis works, you needed to have a crystal structure. It was the same idea that Jim Howard had with nitrogenase. In both cases, they are coming at this from a functional perspective. So, George's group purified and crystallized a bacterial photosynthetic reaction center. Dave and I couldn't believe a project like this would just show up at our doorstep. More amazingly to me, we had not been the first crystallographers that George had tried. He was essentially working his way up the coast. Most crystallographers apparently had said, "These crystals aren't good enough," which I think is a reflection that for crystallographers to do crystallography, you have to have good crystals, as opposed to the structural biologists, who are like, "Wow, this is a really interesting biological system. Let's see what we can do this."
So George drove up with Jim Allen, we met at UCLA, and Dave and I agreed to work on the reaction center project. Over time, Dave, very graciously, let me take over the project. At this point, there's no FedEx, so Jim Allen would drive up from UCSD once a week or every two weeks with a load of crystals, and we would screen them. Now, we have a project that's light-sensitive, to complement our nitrogenase project that is oxygen-sensitive. During our initial efforts on the reaction center, we were really concerned about light activation. So, we would cover the X-ray equipment with this canopy and use green light because reaction centers don't absorb much light at that wavelength. Jim would be working away on improving the crystallization conditions – he would set up these crystallization trays, and every well would have one crystal that was probably a millimeter or more. You could see them with the eye. And of course, at that point, they're black because the chromophore concentration's so high, they just absorb all the light. He would work away, and over time, the resolution got better and better.
One of the technological developments that Eisenberg realized would be transformative for crystallography was an electronic detector for collecting X-ray data. At Harvard, we used film. I remember thinking, "What do you need electronic detector for? Film is such a great detector." I quickly realized the error of my ways and that having this rapid readout, where you would immediately integrate the data without having to integrate the film, was the way to do things. So we were able to collect diffraction data on the area detectors and Todd Yeates and Jim Allen started making progress on the structure determination. In the meantime, in Germany, there was this group that was interested in membrane protein structure and started with bacteriorhodopsin, then moved to bacterial reaction centers. They also had crystals of a related reaction center. Ironically, the crystallographic work was being done in the lab where I could've been a postdoc but didn't. And who knows, maybe I could've been working on this project if I had postdoc-ed in Germany, although I don't think the timing quite worked out.
ZIERLER: Could you explain coming at this from a functional perspective?
REES: Photosynthetic centers absorb light and convert it into, basically, electricity. There is a separation of charges following the absorption of light and then the subsequent movement of charges away from each other. The separation of these positive and negative charges is the way that the energy of the light is captured chemically. So the question is, how does that happen? George had worked out, from spectroscopic measurements using EPR, electron paramagnetic resonance, of which he was the world's expert, that the primary donor, was a pair of chlorophyll molecules. He realized that this meant that the primary event of absorbing light and leading to this separation of charges involved two chlorophyll molecules. The challenge, then, is that once you excite this pair of chlorophyll, then the electron in the excited orbital needs to be ferried away from where the absorbance takes place to prevent the competing reaction where the light is absorbed, you get an electron in an excited state, and it then returns to the ground state. So all you've done in this case is basically absorbed light and made heat, which is not a way of capturing energy in a biologically usable form. The way to capture the energy is, once this electron gets excited, to move the electron away from the original pair of chlorophylls that absorbed the light and not just return the system to the initial state where this whole process started.
The question then was, how are the chlorophylls organized that allow this absorption of light and the subsequent movement of electrons, so that most of the energy is conserved in a biologically usable form? Our interest was to establish the structure of a bacterial reaction center, to understand how the chlorophylls and the associated chromophores were arranged to facilitate this process. The underlying chemistry and physics of electron transfer reactions had already been worked out by Rudy Marcus, who subsequently won a Nobel Prize for this work. Some of the key features of this ability for electrons to move in one direction had been anticipated by Rudy's analysis of the electron transfer problem. And so, there was also a lot of interest in the structure, energetics, the thermodynamics, and the kinetics of this process to see how well they were described by Rudy's treatment.
My motivation for studying the structure of the reaction center was twofold. One was to understand how the chromophores were arranged that would allow this very specific transfer of electrons such that the energy would be conserved in a biologically usable form. The other was to understand the principles governing the structures of membrane proteins since at that time, no membrane protein structure had been determined. The reaction center combined these two goals. It was a membrane protein, and it carried out this fundamental process of photosynthesis that was important for sustaining life. So that was my motivation for this work. We were incredibly fortunate that George Feher had realized the importance of the reaction center and started working on it. Unfortunately for George, the group in Germany had better crystals of a different reaction center and were able to solve the so-called phase problem and consequently solve the first structure. Our reaction center was the second structure of a membrane protein, along with the reaction center structure independently solved by a group at Argonne National Laboratory. The first group, with Michel, Deisenhofer, and Huber–Huber was the crystallographer I was going to postdoc with–ended up sharing the Nobel Prize for their work. George had the vision of the importance of bacterial reaction centers, and the necessity of a crystal structure, and so I felt like he should have received a Nobel Prize at some point for all his contributions to spectroscopy and photosynthesis.
ZIERLER: When did you start to get interested in homology?
REES: Well, this was always a focus from the beginning of protein crystallography. The first protein structure to be solved was of myoglobin by John Kendrew, but Max Perutz was also working on hemoglobin, and they were clearly related structures. Even though the protein sequences for myoglobin and the two chains of hemoglobin were distinct, there were similarities between them. That was one of the very important concepts in the field of protein structure, that different sequences can fold to the same structure. There are many sequences that can fold to the same structure. Now, we know that some sequences can fold to multiple structures. But still, there's this degeneracy that many sequences can exhibit similar structures. What are the key features that must be providing whatever information is needed to specify the structure? The German group worked on the reaction center from the bacteria Rhodopseudomonas viridis while we worked on the one from Rhodobacter sphaeroides. While these were distinct sequences, they had the same structure. And we could use the structure of the Rps. viridis reaction center to phase our structure. By comparing these structures, we could start looking at the principles underlying the structures of membrane proteins.
As a starting point, we examined families of related membrane proteins, so-called homologous proteins, starting with different reaction centers and looked at the patterns of sequence conservation. Not surprisingly, one finds that some sequence positions are more variable than other sequence positions. What became apparent once we started examining the reaction center structure is that residues that were on the outside of the membrane protein facing the membrane were less well-conserved than the residues that were packing on the inside interacting with each other. And the same is true with water-soluble proteins, that the surface residues are more variable than the buried residues. And in a way, that's not so surprising. You'd think that the buried residues have to have these very specific interactions to pack well with their mates in the interior of the protein, and that's true. What was a little bit, perhaps, unexpected was that the interior of membrane proteins to first approximation looks like the interior of water-soluble proteins in terms of the types of residues that were there. Once we had more structures of membrane proteins, we could start looking at characteristic features of membrane proteins to try to understand what features were more universal and what might be unique to, say, photosynthesis.
ZIERLER: I asked about instrumentation in the 1980s. What about computation? Were computers reaching a point where simulation would be a viable tool in your research agenda?
REES: Yes and no. Certainly, as a refinement tool. You have the experimental data, and from that, eventually you get a structural model. From the model, one can calculate the diffraction pattern, and then you can start a least squares type refinement process, where you adjust the coordinates of the model to improve the fit to the observed diffraction pattern. This type of refinement was originally introduced into small molecule crystallography by Eddie Hughes, a senior research fellow at Caltech under Pauling. The second structure refined by least squares was Lipscomb's graduate work on the structure of methyl ammonium chloride. Now for macromolecular structures, if you have 10,000 atoms in a protein, and they each have X, Y, Z, with some other parameters, the least squares problem is overdetermined, but not by much. So there's a real problem of getting trapped in local minima and overfitting the data. There were advances in computation, where instead of just moving the atoms the improve the least-squares fit, you'd combine this with molecular dynamics. In this case, the coordinates would be effectively bouncing around, so that in principle, you could imagine hopping out of local minima and converging to something closer to a global minima. Getting trapped in local minima is real concern when you're using the structure of a related protein to solve your protein, where there may be some features of the initial structure that are not present in the other structure. With the standard refinement, one would just get trapped in something that looked closer to the original structure. So molecular dynamics was definitely becoming part of the toolkit of improving the convergence of refinement and trying to minimize the likelihood that you would be trapped in local or false minima.
In the 1970s and 80s, molecular dynamics were also being used for computational studies of protein function, but those applications were generally distinct from the crystallographic applications. There wasn't much overlap between the two applications except in this one area of refinement. Ironically, I was mentioning how valuable the computer room had been at Harvard because of the colocalization of crystallographers and Karplus's group who were developing molecular mechanics. Some of the molecular dynamics refinement programs ended up growing out of that environment. I have to confess, however, that even though we were working on ways of trying to incorporate stereochemical restraints into refinements, it never dawned on me that you could try to combine molecular dynamics and crystallographic refinement. It just goes to show, even when you're in the right environment, you don't always make connections that seem obvious in hindsight. It's just like electron microscopy today where I think "Why didn't I start this ten years ago?" [laugh] Fundamentally, I just never made the connection that this could be useful. I think that was in part because I tend to think the existing methods are good enough, there's only so many things you can work on at one time, and you have to decide what's important. And I was trying to solve structures, and so, advances in refinement and electron microscopy didn't seem like they would help me overcome more quickly the problems that needed to be solved.
ZIERLER: An overall budgetary question as we round out the 1980s. From that initial seed money of $15,000, by the end of your time at UCLA, what was your budget like? Who was supporting you? How big had you grown in those seven years?
REES: Actually, I have those numbers somewhere. I don't have them off the top of my head, but there were…
ZIERLER: I guess what I'm asking is, was the $15,000 just the nudge to get you going, and then we're talking about much bigger money?
REES: Absolutely. I have to admit I still don't really understand how research finances work out, but somehow, they do - at least if you're able to stay in business long enough. There were several sources of funding that became available. One is, as I mentioned, I had crystals of the nitrogenase iron protein. I wrote an NIH grant on this structure, and it was funded. There are also various starter grants, which were just starting at the time. My department at UCLA nominated me for the Searle Scholars and the NSF Presidential Young Investigators Awards and I received those awards. I can't remember my group size, I think I had two postdocs and maybe five graduate students, something like that at the time. We were working on the reaction center, and we were working on nitrogenase. Furthermore, we had this wonderful group of structural biologists at UCLA with Dave Eisenberg, Richard Dickerson, and me. We received an NIH program project grant directed towards the structural biology of HIV. My part focused on using structures to try to potentially identify docking sites for different ligands that might be a way of doing a very primitive drug screen. It was incredibly naive and simplistic in hindsight.
ZIERLER: In thinking about your collaborations at UCLA, it's at the periphery, and to some extent, the terminology only gets you so far, but where is your background in biophysics? Where is it relevant, both to the people that you're working with, the questions that you're asking, and what you're able to see?
REES: At the time, protein crystallography was definitely recognized as a legitimate form of biophysics. And by using these physical methods, we were definitely considered biophysicists. Dave Eisenberg had co-authored with Don Crothers this highly influential book on Physical Chemistry: With Applications to the Life Sciences that transformed the teaching of physical chemistry for the life sciences because the examples in physical chemistry were based on living systems. I was in the rotation for teaching the biophysical chemistry class that used this text, which was very important for me because while I'd had physical chemistry courses, I hadn't studied a lot of the material useful in biophysical chemistry. Teaching this course, just like teaching biochemistry, really forced me to get a stronger foundation in the fundamentals to the craft.
There were interesting sociological dynamics I first encountered at UCLA, but which I emphasize are not unique to UCLA. It was not uncommon for there to be real tension between biochemists and chemists because chemists, at the risk of over-simplifying and over-generalization, didn't feel that biochemists were real chemists. A lot of what one studies in biochemistry department, could also be done a biology department, or in a biological chemistry department in a medical school. Before I got to UCLA, there had been a departmental retreat the year before or so that ended up in changing the name of the department to Chemistry and Biochemistry, among other changes, and the tensions had eased somewhat. But in a sense, the two views were epitomized by Don Cram and Paul Boyer. They both won Nobel Prizes for work they had done at UCLA and were major scientific figures. But there was just this real cultural difference. But the thing I appreciated is that the biochemists really did feel that we were chemists. We were interested in molecular mechanisms and so chemistry was our natural home.
I think we were talking in the first session about the hierarchy, where physics fits in. I think every discipline has the counterpart of, "What is a real chemist?" In addition, within biochemistry, there were also tensions between so-called "real" biochemists and biophysicists. And again, I can tell you this is not unique to UCLA. My view has been that these are all destructive attitudes. There are a lot of things we can be focusing on, and these internal feuds are not helping anything. In a real sense, they provided my first exposure to departmental politics. Prior to becoming a faculty member, I hadn't thought much about what goes on between the faculty in a department, but I quickly learned that the dynamics can be complex. The closest I had come to thinking about such matters was while I was a graduate student, working in the computer room which at that time was off the lobby of the Mallinckrodt chemistry laboratories at Harvard. Once a month, I'd notice all these faculty members going into a nearby meeting room. And I remember thinking, "Wow, I wonder what they discuss in there. What great pedagogical concepts and key issues are they discussing?" I'd just have this observation, all these people going on, "Wow, what are they talking about? It must be interesting." I remember after my first faculty meeting, it was like, "I can't believe I wondered what was going on in these meetings." [laugh] It's a little harsh, but still sort of like, "Wow."
ZIERLER: You punctured the aura.
REES: It really did.
ZIERLER: If you can give sort of an organizational overview to give sort of color to the departmental politics, at UCLA, how did that break down? What the school? What were the departments? How did you fit in in terms of biochemistry?
REES: Chemistry departments, for better and for worse, have a very strong sense of subgroups. It's different than biology. There are organic chemists, physical chemists, inorganic chemists, biochemists. They're not completely autonomous groups, but still, those are the groups typically involved in doing teaching assignments, recruiting graduate students, running their own seminar programs, etc. Usually there is a single chemistry graduate program for the organic, physical and inorganic chemists, and then there's a separate biochemistry graduate program. At some universities, especially with medical schools, there may be other biochemistry graduate programs.
At UCLA, the Department of Chemistry and Biochemistry was in the School of Physical Sciences. The dean who asked me, "Why isn't this real estate program working for you?" was the Dean of Physical Sciences. In addition, there was the life sciences, and then this huge species, the medical school, that was mostly clinical people, but also included a number of basic science departments. The building where my labs were located, which is now Boyer Hall but at that time was the Molecular Biology Institute, was extraordinary in that it housed faculty from both the Medical School as well as faculty from the College of Letters and Science. The director of the Molecular Biology Institute, who was originally Paul Boyer, oversaw the MBI, which included a graduate program. So there were a number of different administrative levels. And in my own little niche, I was in the Biochemistry group in the Department of Chemistry and Biochemistry, but I was also an associate member of the Molecular Biology Institute. I was part of these groups, but most of my life really focused on just Chemistry and Biochemistry.
ZIERLER: Did you have opportunity at all to collaborate with the hospital? Was that an asset to you at all, some of the research that was going on there?
REES: I never collaborated with the medical school, but through the Molecular Biology Institute, there were regular interactions–in fact, the person who asked if he could purchase any supplies for me was in a medical school department and housed in the Molecular Biology Institute. And some of the teaching was done jointly for graduate students in different graduate programs. I didn't really collaborate with other groups at UCLA when I was there. I had two senior colleagues doing structural biology, Dave Eisenberg and Dick Dickerson. And I was a junior faculty, and I had the projects I was working on. I think, the natural inclination is, if you were interested in a collaborative project, you would probably talk to Dave Eisenberg because he had been there the longest and was a very collaborative and interactive person. I would hear about these things, but I wasn't involved in any collaborations. The notable exception was with George Feher from UCSD, who initially reached out to both Dave and myself about working on the photosynthetic reaction center.
ZIERLER: So your relative aloofness, can we assume there that ideas about translational applications, that's beyond your radar at this point?
REES: I don't think it ever has been a fundamental part of my radar. Now, at the time, biotech was starting.
ZIERLER: Just to go back to an earlier comment about when you were initially on the job market, pharmaceutical companies did not care about structural biology. And obviously, that changed in a big way.
REES: It did. I met a postdoc at Harvard who had been a graduate student in structure at Caltech and then did a postdoc in molecular biology at Harvard. I remember talking to him, and he told me he had this job at this new company that we would now call biotech. And to me, even though there were no academic jobs, it sounded hopeless. I was like, "This poor guy. There's just no future." The company was Genentech, and he ended up becoming a vice president, starting other companies, all this sort of stuff. So again, I both had no interest and saw no future in these sorts of things. At UCLA, I wasn't involved in biotech, but there were some faculty who had started companies. For whatever reason, LA was never much of a biotech hub.
But there was one company I think in Santa Monica that I visited. They were interested in the structure of a protein called thaumatin that is intensely sweet. But it had some undesirable properties, including a strong, sweet aftertaste. A crystallographic group elsewhere had solved and published the structure of thaumatin. The company's idea was if you knew the structure, you could start doing mutagenesis and perhaps the undesirable properties could be mutated out. The structure was solved, but as was not unusual at the time, the coordinates were not deposited in the Protein Data Bank. Often at that time, structural biologists didn't want other groups to be taking their coordinates, as it was seen, even though they were published, and using them for other purposes. There was stereo diagram of the structure published in the paper, and there was a program that someone wrote to try to reconstruct the three-dimensional coordinates from the right and left images of the stereo pair. I consulted one time with this company to try to extract the three-dimensional coordinates of this protein from the published stereo pair. So I did have a little bit of involvement in the local biotech. But it was very limited.
ZIERLER: To set the stage for your transfer to Caltech, just to get an overall sense of your motivations and sense of opportunities, first of all, you were happy at UCLA, it was a great research environment, you had great collaborators, your lab was running on all cylinders, good graduate students. So you could have been quite happy to spend an entire career at UCLA.
REES: Absolutely. I would have been happy to spend my entire career at UCLA.
ZIERLER: Had you achieved tenure at UCLA at that point?
REES: I was promoted to Associate Professor with tenure at UCLA in 1986. But around this time, the climate for doing crystallography changed dramatically. I was describing how the graduate students and postdocs ahead of me in the structural biology job market had zero opportunities. But then, several things happened. One was recombinant DNA methods came out, so you didn't have to work on an enzyme from a cow pancreas just because there was a lot of it. You could work on other proteins of biological interest. And I really think it was Don Wiley's structures, and specifically the influenza virus hemagglutinin structure and Pamela Bjorkman's structure of the major histocompatibilty antigen, MHC, that attracted the attention of biomedical scientists. Structural biology can help you understand how these very complicated systems are working. So not only could we now start to express any protein, but the community actually started caring about protein structure. In Pamela's case, it was how the immune system can recognize a self-antigen from non-self because of course, you only want to make antibodies against foreign material, not against your own proteins. And Pamela's structure provided the first clue. What that meant in this trickle-down sense is, all of a sudden, the community's attitude towards crystallographers went from, "Who cares about the solid state of structures of digestive enzymes?" to, "Holy smokes, we need to have a crystallographer, and if we can't hire one, because they like to stick together, we'll hire two or three." So the job market for crystallographers went from completely underappreciated to, frankly, completely over-appreciated.
UCLA had been unusual in appreciating the value of structural biology prior to this time, and had hired me at a time when only one other academic job had been available. I had launched my research program there, had incredible colleagues, and great students.
ZIERLER: It's also, by way of contrast, a big university.
REES: It's a big university. I had incredible undergraduates, really interested in working in the lab. Housing was just this huge problem. It was so expensive to live in West Los Angeles. Frankly, it's a problem for all of Los Angeles, but it was magnified in West Los Angeles. And UCLA would try to come up with things, but it was just expensive, and I really worried that even if my housing problem was solved, what about everyone else?
ZIERLER: Were you still taking the bus circa 1988?
REES: We ended up buying a house in 1983, using the loan program that I had discussed with the Dean of Physical Sciences. We found this one area that we could afford, which, as it turned out, was near the bridge on the I-10 that collapsed in the 1994 Northridge Earthquake. So then, I ended up driving. I recognize I am coming from a position of privilege, but our home was near the freeway with a lot of noise and not a great area. By this time, our son John had been born. So we did start a family. And my wife was teaching, and we lived in this little place with two bedrooms. We were looking for a three-bedroom place so we could have a study. At the time in the late 80s, the price of a three-bedroom home in West LA was $800,000. $800,000, sadly, sounds cheap now, which is unbelievable. Housing was a major source of frustration. I had already considered an offer at another university.
ZIERLER: Outside of LA, you mean.
REES: Outside of California. I'd been really intrigued. It was closer to Kentucky, where we were from. We flew my mother down from Maine to look after our son, John, who was an infant, less than a year old, while we visited this other university. The first day was like, "Ah, this is great." By day four, it was like, "I love LA." [laugh]
ZIERLER: There's a reason the houses cost so much.
REES: Absolutely. And this potential position was in a medical school, where there was a view, "Sure, we'll give you the money to purchase an X-ray set, but we expect you to write grants and pay all this money back." So it got me appreciating the environment I was in. But housing was still a problem. That was a major problem.
In addition to housing, there was another aspect of moving that was specific to Caltech, and that was my academic heritage. Lipscomb had been at Caltech, along with Pauling and these other giants like Feynman. I did hear Feynman once when I was at UCLA - he came over to give a talk, which was non-scientific. I can't remember what the connection was. There was just this aura about Caltech. And in comparison to UCLA, it was tiny.
ZIERLER: He was a national figure at this point also.
REES: Absolutely. A large state university has a lot of appeal in many ways – there are lots of students and you really feel that you are helping fulfill the educational mission of society. But there is so much going on that you also feel like what you are doing is just a little piece of an enormous operation. That stands in contrast to a small place like Caltech, where the only purpose is research and education in science and engineering. And I think there was also a little piece of me that was wondering, "Could I really start something on my own?" I'd been so lucky at UCLA, which, in a way, was essential because my startup was minimal. But that was OK, because frankly, all the equipment was present at UCLA that I would need. I didn't have to set up a lab. Dave Eisenberg was an incredible resource. So yes, it was an incredibly stressful period.
ZIERLER: Had you stayed in touch with Lipscomb as a junior faculty member?
REES: Yes, but not regularly. I would talk to him from time to time. And he was writing letters of recommendation for me.
It's interesting, what happened at Caltech–I've sort of pieced this together. Dickerson left because he wanted to make a senior research fellow appointment for one of his people to run his X-ray crystallography lab which takes a lot of expertise. You can't just run labs with graduate students, and postdocs. You can do a lot, but Dickerson wanted someone who was really more senior, with significant academic credentials, to keep his X-ray lab running. One of the aftermaths from when Pauling left Caltech, was that he had a number of senior research fellows, and the CCE Division felt, I think, in some ways, unfairly, that they were stuck with all these people. So the CCE Division said, "We're not going to approve this senior research fellow appointment." During one faculty meeting that must've been one for the ages, Dickerson stormed out and basically said, "I'm out of here." That's when UCLA, that had the foresight about the importance of structural biology, capitalized on Dickerson's unhappiness by recruiting him. The reaction at Caltech was interesting. They realized that they actually did want someone doing structure. But it was specifically someone who was working on proteins that bound to DNA. Dickerson had been working on DNA structure. In fact, he solved the first high-resolution structure of the B-form DNA helix. But proteins binding to DNA is at the core of how a lot of regulation works in biological systems. And so, Caltech wanted to hire someone working on protein DNA binding interactions.
Caltech considered a number of people working in this area. But in the meantime, the job market had changed from, "If you want a job, you're out of luck," to a situation where, with not complete exaggeration, anyone who understood Bragg's law, they were like, "We'll hire you." By this time, everyone working in this area was getting job offers because this was a really hot area. Again, I wasn't here at the time, so I don't know exactly what transpired, but Caltech tried a number of people and wasn't able to hire anyone in that area. They interviewed a good friend of mine I had been very close to in graduate school, Mitch Lewis, who had been working on DNA binding proteins. The out of state university where I had previously considered moving, Mitch and I had seriously discussed whether the two of us could move there together. Mitch told Caltech, "Oh, you ought to consider Doug Rees." And of course, I was not working on DNA binding proteins. But I was doing structural biology.
ZIERLER: Just to get a sense of your own place in this, had you put out feelers? Were the right people aware that you would be open to offers? Or you're learning about all of this after the fact?
REES: I'm learning about this after the fact.
ZIERLER: Because you're happy, and you've got a good thing going. You're not out there talking to people in this way.
REES: Exactly. I came over to Pasadena to pick Mitch up from the restaurant where he'd gone out with the faculty members after his seminar. At that time, I got to meet some of the Caltech faculty at dinner just in the course of picking up my friend. But that was primarily, "Nice to meet you". There was nothing more to it. Jack Richards was the senior biochemist and was the key person leading this search. At some point, they decided that they weren't making any progress with getting someone in protein DNA interactions. But they had my name from Mitch. At this point, I'd published nothing on nitrogen fixation, but the reaction center work had just come out. And so, they interviewed me.
ZIERLER: Who at Caltech was really driving the process of recruiting you?
REES: Well, Jack Richards was the key person in biochemistry. But my research was closely related to bioinorganic chemistry and transition metal chemistry, where Caltech had a significant effort. Harry Gray, Sunney Chan and John Bercaw were interested in what I was doing. And obviously, I was just in awe of what they were doing. Within biochemistry, I'm not quite sure. There really had been a desire to recruit someone working on protein DNA interactions, which was not what I was working on. There was sort of an odd dynamic that goes back to what I noticed at UCLA between "biochemists" and "biophysicists" - the "Are you one of us, or aren't you?" So that was sort of an interesting dynamic. But I did get the job offer.
ZIERLER: Before that moment of decision, how well aware were you of the relevant research that was happening at Caltech and some of the organizational peculiarities in terms of how Caltech was put together at the division level?
REES: I was completely unaware of the organization peculiarities. Of course, I was really aware of the electron transfer work because of the research we'd been doing in photosynthesis and nitrogen fixation. Rudy Marcus, Harry Gray, Sunney Chan, John Bercaw, these were giants. I was very aware of that. Now, at the same time, and I don't remember the precise chronology, but biology had been trying to hire Pamela Bjorkman. Of course, I knew Pamela from my time as a graduate student.
Before my time, you were a crystallographer first, and then you happened to work on some types of proteins. In my case, I was certainly a crystallographer, but I got really interested in nitrogen fixation and actually postdoced in a nitrogenase lab to learn the biochemistry. Pamela was a great crystallographer, but she was also an immunologist and worked in an immunology lab on T-cell receptors at Stanford as a postdoc. She was hired by the Biology division because of her expertise in immunology. The fact that she was working on structure was important as well. So that was another part of the dynamic about whether or not to move to Caltech. The searches in chemistry and biology were independent, but obviously, Pamela and I knew each other and liked each other.
The decision was one of the most stressful periods of my life because for all the things that you noted–I liked UCLA and Dave Eisenberg, the structural biology group and my colleagues at UCLA were incredible. Becky and I decided to make a list with all the relevant parameters so we could add up the pluses and minuses to help us make the decision. "Smog, definitely worse at Caltech +1 for UCLA". "You couldn't have better colleagues than Eisenberg and the biochemistry group +1 for UCLA.". "Becky likes her teaching job in West LA +1 for UCLA" "Housing opportunities in the Pasadena area +1 for Caltech", etc. UCLA would always win in this analysis. It was like, "Why would we leave UCLA?" But yet, in thinking about this decision, there was something in my gut feeling that wanted to move. This is, I think, the way I finally thought about this. Moving is always painful and the transition is very stressful. But I was trying to imagine two scenarios at some point in future. One, sitting in my office at Caltech saying, "Wow, I could've stayed at UCLA." The other was sitting in my office at UCLA saying, "Wow, I could've been at Caltech." There was just something about the legacy of structure at Caltech, and I think also the challenge in seeing if I could set up a lab on my own. Of course, it wasn't completely on my own because I would be doing this with Pamela Bjorkman. I also felt our housing situation, there was just no comparison in what we would be able to get in the Pasadena area as opposed to what we would be able to get at UCLA.
The rational brain was saying to me, "You should stay at UCLA because this is a great situation, and there's always that uncertainty, 'Maybe I'll be a flop'" if I move. I'd always worked in a lab where things were already set up. I'd never set up a lab before. I'd never set up an X-ray lab before. It was more than just having the money to do these things. But the other side of that was, "I've got to see if I can do this." And I really liked working with Pamela. Ultimately, I made that decision. But Becky and I went back and forth on what to do. And I told Caltech several times that I wasn't coming, I was declining the offer. I just remember telling this to Fred Anson, who was the CCE Division chair at the time the chair, and he goes, "Wait! Don't make the final decision." And I can't remember what all happened now. After we finally made the decision to move, because it took a year for the Caltech lab to be built, I was like a lame duck at UCLA. And that was just an incredibly awkward year. I think my colleagues understood it was a professional decision. We're still friends with the Eisenbergs and other UCLA colleagues. But I still feel conflicted about this decision. Even though I wouldn't change the decision, having good friends is really important.
One aspect that I really value about UCLA was that it was a great place to be a junior faculty member. Dave Eisenberg and my colleagues in biochemistry really showed me the ropes and helped me through all these things that I just was completely oblivious to in terms being a faculty member. So that certainly made leaving even more difficult. But it was also a time where places started to realize, "Crystallography's great. We have to have it." Just like electron microscopy now. It was a weird time to be in, where you go from few people caring about what you are doing to all sorts of places wanting to hire you. Staying grounded through all this, I found, to be difficult.
ZIERLER: Last question for today. In assessing the constellation of considerations you had in making this decision, given how deeply you respected Pamela's work, to what extent did her decision serve as a proxy for you intuiting, at least on some basic level, that there's a certain magic to the way research happens at Caltech, where the departmental distinctions, which are very clear at a place like UCLA, might not be as prevalent? Was that a deciding factor because it would be so beneficial to your research?
REES: I started on July 1, 1989, and I think she started on January 1. She had made the decision before I did. I can't remember where I was in the cycle of yes, no, yes. It was impressive to me how my discussions about moving involved both the CCE and Biology divisions. Lee Hood was the chair of Biology at that time and so he would be describing a sweeping vision for biology and the interface with technology. The space for my labs was in the Braun basement, which was split between chemistry and biology. Pamela's labs were in the basement of Church. While our labs weren't immediately adjacent, they were connected by a tunnel, so it was fairly convenient. The way this all happened across various administrative entities seemed seamless to me. At the time, though, I have to admit I wasn't thinking in any detail about the administrative structure at Caltech.
There was one experience that looking back really did make a significant impression on me about the culture at Caltech. In the course of my negotiations, I met with Barclay Kamb, the then-provost (and son-in-law of Linus Pauling). I remember sitting in this office, and we weren't talking about things "Well, we can't do $50,000 for this expense, but we might be able to do $40,000." Instead, Barclay wanted to know about my research. We didn't explicitly discuss how much startup money I needed, even though, of course, in some sense, the conversation was all about money. But in another sense, it was not about money. It was like, "What is your vision? What do you want to do in your career? How can we help you?"
ZIERLER: It was an academic discussion in the purer sense of the word.
REES: It was an academic discussion with a potential colleague who was a geologist with incredible breadth, including crystallography and solving the structure of an ice phase. It really was a memorable discussion for me. I remember afterwards thinking "Wow, what a place! If you say, 'Here's my vision and here's what I need,' they're like, 'How can we help you execute it?'" I'd forgotten about this until now. That conversation, for me, captured the institutional tone of Caltech. Of course, a lot of things have changed over the last 30-plus years. Unfortunately, Caltech doesn't just give you the money, but if you have a vision, in my experience they'll try to work with you to bring things to fruition. I think that's the advantage of a small place like Caltech. It's really focused on research and education. They want us to be doing not just stuff, but interesting stuff – just as Lipscomb learned as a graduate student here with Pauling.
ZIERLER: Well, that's a perfect place to pick up for next time.
[End of Recording]
ZIERLER: OK, this is David Zierler, Director of the Caltech Heritage Project. It's Wednesday, October 13, 2021. Once again, it's my great pleasure to be back with Professor Douglas C. Rees. Doug, it's great to see you. Thanks for joining me again.
REES: David, it's great to be here. This has been a really interesting process, sort of going through this timeline of how we got to this point.
ZIERLER: Absolutely. That's the historian's job. So the year is 1989. You've already made this big decision that you're going to join the faculty at Caltech. And I'll return to this really important point that you emphasized last time that you were so energized that in your discussions with the provost, you were talking about the science. And that it gave you a jolt in terms of how you were going to start things up. Perhaps just by way of comparison, obviously, this is a different stage in your career from UCLA. But how would you compare the differences in terms of starting up the lab seven years earlier at UCLA versus what you were facing at Caltech in 1989?
REES: There was a tremendous difference. I would say maybe at the core of this, when I had originally been looking for jobs in the early 80s, the job market was bleak. In fact, UCLA was extraordinary in having a position in structural biology at all, which is something only one other school in the country had. Caltech and these other places did not see protein crystallography or structural biology as a direction for the future. And what happened over these seven or so years is that through technological advances, especially in cloning, recombinant DNA work, it became possible to start making sufficient quantities of all sorts of proteins of interest where one could now envision structural studies, whereas previously, one was limited to studying proteins that occurred naturally in large abundance. All of a sudden, all these proteins that were off limits became tractable. In the meantime, there were also several high-profile structural accomplishments that I've described previously that changed how the biological community viewed structural biology. As a result of these two events, protein crystallography went from being something that didn't seem to have much of a future to a becoming a very hot field. And this was seen in that a number of places were trying to hire crystallographers. They would often try to hire several crystallographers. The whole landscape in terms of how people were seeing the future of this field changed during that seven-year period.
ZIERLER: And for your own research at the time, was your research agenda at all in a period of transition? Was that relevant in terms of making the switch to Caltech? Or were you looking to basically move the whole operation wholesale and continue on the trajectory that you were on?
REES: I'd say more of the latter in that I didn't really see this as, "Now, I have a chance to completely retool what I'm doing." But I was working on several projects. I think we discussed last time that the main one was nitrogen fixation, which had been a slow, painful slog. I'd been working on that for seven years at UCLA, and we'd been making some progress but were still some ways from the structure. And then, the photosynthetic reaction center, the membrane protein that I'd been working on in collaboration with George Feher from UCSD, where we actually had a structure. The reaction center was, I'd say, seen as sort of a real hot project. But there were still plenty of things to do with the reaction center once we had the structure. I wasn't seeing the move as an opportunity to start getting into new areas as much as, I think, wondering if I could actually succeed in setting up a lab. Pamela Bjorkman started her faculty position in biology at Caltech a little bit before I got there. It was about seeing whether we could actually set up a thriving structural biology operation at Caltech where there was this tremendous historical tradition, but with nothing happening at present. So that was both exhilarating and also a little intimidating.
ZIERLER: Bjorkman's coming in at the same time. Once you got the lay of the land, and given your appreciation for the spirit of multidisciplinary collaboration, across divisions and departments, who was at Caltech at that time who would've been really interesting to collaborate with or even just bounce ideas off of?
REES: There was a strong group in the chemistry department that I didn't really know before my arrival but enjoyed interactions with, especially in the metalloprotein area. That included Harry Gray and Sunney Chan, who had been working on metalloproteins and membrane proteins involved in electron transfer, respiration, and so on. They were both interested in the detailed chemical reactions taking place in these proteins, where we were more interested in defining the structure of these whole assemblies, of which the metals are a small part. But there was certainly an environment where there was a lot of interest in structures.
And then there were people like John Bercaw, who was an inorganic chemist that earlier in his career had been very interested in transition metal nitrogen chemistry. Those colleagues were very supportive of what we were hoping to do. On the more biochemical side, Jack Richards, who spearheaded the search, was a visionary in protein chemistry, enzymology, and even protein engineering, and was also incredibly supportive.
ZIERLER: What was Gray working on at that point?
REES: Harry Gray's group was working to map out the precise details of electron transfer, electrons going from a donor group to an acceptor group in proteins, how long that took and how well these experimental results were described by the Marcus equation that Rudy Marcus had developed. The structure determination of the photosynthetic reaction center provided an atomic resolution structure that defined the spatial arrangement of the various groups involved in electron transfer. Importantly, the rates of electron transfer between these different species had been measured so there was quite a bit of interest in trying to understand how well Rudy's equation mapped onto the experimental observations. The Marcus equation related the chemical driving force to the rate of electrons transfer. It's sort of like Ohm's law, where for a given resistance, if you increase the voltage, then the current will increase. And that's true in some regimes in proteins. But once you get to a sufficiently strong driving force, then that behavior breaks down. It turns out that photosynthetic reaction centers exhibit this so-called Marcus inverted region that had been predicted by Marcus. This effect is critical for the ability of photosynthetic reaction centers to absorb light, generate a high-energy species, and then allow the negative electron and the positive hole to move in opposite directions without just recombining and converting the light into heat. The Marcus inverted region was a non-intuitive prediction and so there was a lot of interest in trying to characterize this in protein systems. The Gray group were the pioneers in engineering experimental systems to make the necessary measurements to confirm this effect in proteins.
ZIERLER: I'm sure you've heard this idea that with Caltech being so small, one of the ways it punches above its weight is not in attempting comprehensive coverage in any general field of research, and that hires at Caltech are hired because they're top rate, and they're doing interesting stuff. With that in mind, was your sense that you were brought on not because Caltech was looking to strengthen its work in structural biology generally, but simply that they liked what you were doing, and they wanted to host that research?
REES: Of course, I wasn't here at the time those deliberations were taking place, but actually, I would say, my hire conflicted with that principle. But from what I've seen subsequently, when we're making appointments, we generally don't look for someone in a very narrow research area. We might be looking for someone in biochemistry, or chemical physics, or inorganic chemistry.
ZIERLER: But just to interject, you're historicizing the narrowness of the field circa 1989. Right now, it's much bigger. What you're saying is, with your hire, structural biology was very much considered a niche area of research at the time.
REES: Right. My UCLA colleague, Richard Dickerson, shortly before I arrived, moved to UCLA from Caltech. Dickerson solved the structure cytochrome c and also the first atomic resolution structure of B-form DNA. Dickerson moved to UCLA due to feeling a lack of support for structural biology at Caltech, which ironically created the hole that I then moved into. At the time Dick left, I think it was true that crystallography was seen as a field that had a limited future. But in those intervening years, it really changed. Subsequently, the biochemists and the CCE division realized that they wanted someone not just in structural biology, but someone working specifically on the structural biology of the interactions of proteins with DNA. At the time they were starting that search, I believe the first structure of a protein DNA-complex had already been solved. Several years were spent trying to find someone in this area. Ultimately, and I guess I would say fortunately, they were not successful, and they decided that they wanted someone more generally in structure, but it didn't have to be in the protein DNA area. And that's when my appointment was considered. But still, I would say, they were looking to hire someone doing a specific technique. That is much more targeted than I've seen searches since I've been here, at least in chemistry.
ZIERLER: And in terms of the searches, and again, this goes to Caltech's very unique administrative structure, people were making these decisions that were not specifically in your department or field, but the search committees were more widely conceived beyond just chemistry.
REES: I don't know who was on the search committee. I think it was probably more within chemistry. The search that led to the hiring of Pamela Bjorkman was in biology at the time. Now, those two searches were independent in that they were started not with the thought of, "Let's try to hire someone(s) to interface between chemistry and biology." But biology had identified Pamela, and chemistry had identified me. We knew each other from graduate school, and at least speaking for myself, that had a lot of attraction in terms of working jointly with Pamela. So the searches themselves were probably siloed, but when it came to actually setting up our labs at Caltech, this process included the two divisions, the Beckman Institute, and the Institute. And I don't know for sure where all the funding came from, but there were multiple sources contributing to the startup.
ZIERLER: Now, did you bring instruments and infrastructure with you from UCLA? Or this was from scratch, everything new at Caltech?
REES: Almost everything was new. Again, as was typical at that time, my startup at UCLA had been modest. I did have some equipment, which was being used more broadly by the structural biology community, which I left at UCLA because it was being used by others. Basically, we started things at Caltech from scratch.
ZIERLER: What were some of the advances in instrumentation, given the fact that you had this opportunity to buy new? What were you able to purchase that might not have been feasible had you stayed at UCLA?
REES: I can't really answer the latter part, but certainly one thing that had happened in the meantime was that there were significant advances in the development of electronic detectors for X-rays, two-dimensional position-sensitive detectors. Through the foresight of Dave Eisenberg, we had this large multi-wire detector assembly at UCLA, which really got me to appreciate being able to read out the diffraction pattern in real time without having to develop film and then integrate it, which could take hours or days. It was clear that electronic area detectors were transformative. At Caltech, we were able to get the most recent detector available at the time. And we put in performance criteria that were tailored towards our nitrogenase projects that ensured we were able to measure data of the quality we needed to solve the projects we were working on.
ZIERLER: What was some of your research in membrane proteins at the time?
REES: My research in membrane proteins at the time exclusively focused on the photosynthetic reaction center. Our responsibility for this project was doing the crystallography. The cell growth, the purifications, the crystallizations, etc. were all done at UCSD in the Feher group. Post-docs in his group, including Jim Allen and Herb Axelrod, would come up here with crystals, and we would collect the diffraction data and work on the structural analysis. It was still too early for my group to make that leap into trying to purify and work with membrane proteins on our own.
ZIERLER: And if you can explain more broadly, why was nitrogenase so important for your research generally at this time?
REES: I had started working on this project as a postdoc with Jim Howard at Minnesota and managed to get crystals of one of the two component proteins. There are two proteins involved in nitrogenase that are given not very imaginative names based on their metal composition. There's the molybdenum-iron protein, or MoFe protein, and there's the iron protein, or Fe protein, and these two proteins work together to carry out the reduction or the conversion of nitrogen in the air to the bioavailable form of ammonia. These two proteins have essential but complementary roles. The MoFe protein has the active site for converting nitrogen into ammonia involving an unusual metallocluster known as the FeMo cofactor. The Fe protein is responsible for transferring the electrons to the MoFe protein that are needed to convert nitrogen to ammonia, and it couples this transfer to the binding and hydrolysis of ATP. The iron protein has many similarities to a number of other biochemical systems that have nothing to do with electron transfer or nitrogen fixation. It was an example of a protein system where, if we learned about the structure and function in one context, it would also illuminate and inform our understanding of the structure and function of other distinct, but homologous, proteins. The Fe protein closely resembles a large family of proteins that use ATP or GTP that were involved in a number of different phenomena, including signal transduction, protein targeting and other contexts. I started working on the Fe protein as a postdoc at Minnesota, and then pursued the crystallography of this protein for seven years at UCLA. We were able to collect good data, but had not been able to crack the structure. This is the protein, in the acceptance criteria of our area detector, we wanted to make sure that we could continue to collect this good data.
The other protein, the MoFe protein, has two types of extraordinary, and at that time, structurally uncharacterized, metal clusters, each containing eight metals and seven to nine sulfurs that attracted a lot of attention from chemists because these clusters were responsible for the conversion of nitrogen to ammonia under ambient conditions. Industrially, the reaction is carried out in the Haber-Bosch process, where using high temperatures and high pressures with a catalyst, the rate of the reaction is fast enough to convert nitrogen to ammonia on a large scale. But unfortunately, at these high temperatures, the equilibrium disfavors ammonia production. At room temperature, ammonia production is highly favorable, but the reaction rate is really slow. So there was a lot of interest in trying to understand the nature of these metallocenters and the catalytic strategies that allow them to work at room temperature, instead of requiring these high temperatures and high pressures to force the reaction in the unfavorable direction under those conditions.
ZIERLER: A teaching question. Coming from UCLA, Caltech is a very different undergraduate body. It's a huge university, a much more diverse undergraduate population coming from a range of academic abilities and geographic placements. What were some of the challenges, and what were some of the assets in interacting with Caltech undergraduates?
REES: I had no prior experience with Caltech students. I'm not an alum, and I had never been here before joining the faculty. I would say several things stood out. Instead of teaching a biochemistry class with 250 students, mostly pre-med, or a biophysical chemistry class with maybe 50 students, now, I was teaching courses–at least until I started teaching Chem 1c, which did have 250 students–where a class of 20 would be considered good-sized. With the smallness of the undergraduate population, we get phenomenal support from graduate teaching assistants (TAs).
There were aspects about teaching at Caltech that reflects the academic culture, about which I was completely clueless. It's been about 30 years, and I can't remember all the specifics, but teaching my first course at Caltech was a real experience. Probably while distributing the syllabus, I said that the mid-term exam was going to be say, October 29, and it'd be an in-class exam. One of the students then tells me, "You can't do that," ie – you can't have an in-class exam. And I'm thinking, "What do you mean I can't do that? I'm the instructor. Of course, I can do this." And so, I hadn't appreciated this whole culture of the honor code which includes take-home exams so you don't have students take an exam in class. Today, we have orientations, not only for students but also for faculty, where we go into some of these things. At that time, Caltech didn't, and I was completely dumbfounded that all the exams were take-home because I had never seen anything like this before.
My memory may not have things completely right, because I do remember talking to Harry Gray when I was thinking of moving to Caltech, and he was telling me about the honor code and this phenomenon of students having specified time limits for exams. I don't know if I appreciated or not that these were take-home exams, however. The students would be working on an exam and would sometimes get to the point where they reached the time limit while in the middle of a problem. They would draw a line at that point in the exam and write, "Time's up." Then, they'd continue working on the problem because they were interested in solving it. At UCLA, exams were in-class and we would sit there and proctor – keep in mind that UCLA students were wonderful, so I'm not making any value judgment here – but we were worried about cheating. So, the idea that students have take-home exams on the honor code, and when they reached the time limit, they would indicate that, but they would still keep working to solve the problem, was just remarkable to me.
A similar issue also came up, again likely when distributing my first syllabus, when a student asked, "What's the collaboration policy?" I thought, "What do you mean, what's the collaboration policy? You do the work. What's so hard about that?" Again, not realizing, especially for the undergraduates, how this culture of working together really is an important part of how they learn material.
Another major change on moving from UCLA to Caltech was that at UCLA, I was teaching classes that would be offered multiple times a year. Introductory Biochemistry at UCLA might've been offered five times a year. I would teach it one term, but there'd be four other instructors. So there was a well-defined curriculum that everyone stuck to because students would take this class at different times. And if I had questions, my colleagues were incredibly helpful and supportive. They'd share their lecture notes with me, which was invaluable because I had never previously taught the courses I was teaching at UCLA, mainly Introductory Biochemistry and Biophysical Chemistry. And not only would they share lecture notes, but also exams, problem sets, and all this stuff. There was a collective sense of community engagement and responsibility for making sure that our biochemistry courses went well, and my colleagues were wonderful. I don't remember the first course I taught at Caltech, but that course was only taught once a year. I was surprised to realize that we had a lot of flexibility to cover whatever material we wanted to teach. We didn't sit down as a division or subgroup and say, "Here are the core things our students need to learn from our classes", but instead, we would each decide what to teach. That was a bit jarring to me. I do think maybe for advanced courses, the ability of an instructor to put their unique stamp on a course can be valuable. But for some of the more fundamental or even upper-division courses, it is important to systematically cover the basic foundational material. But at Caltech, the instructor, at least outside of the core classes, has a lot of freedom and flexibility to tailor the course the way they want to – for better and for worse.
ZIERLER: Staying on the topic of undergraduates, not necessarily specifically related to your research, but it would be interesting to get your take as a newcomer in the late 80s and early 90s, as I'm sure you're aware, 50 years ago at Caltech, the dominant major for undergraduates was physics. That was the hot thing. And then, fast-forward to today, that's really been supplanted by computer science. I'm very interested in historicizing this shift, when it happened. When you arrived at Caltech, what was your sense of the big things that the undergraduates were interested in, and how did it relate to chemistry?
REES: This is my impression off the top of my head, as of October 13, 2021, based on my experience. When I arrived at Caltech, I initially taught upper-division classes in chemistry or biophysics. At some point, fairly early on, though, I started teaching one term of Ch 1. At that time, there were three terms of introductory chemistry, fall, winter, and spring, Ch 1a, 1b and 1c. What was absolutely clear to me was, by and large, students did not come to Caltech because they were interested in chemistry. They were interested in the physics and math core classes but chemistry was a subject that in general, students weren't so happy about taking. And that was before there was a biology requirement. There was this sort of hierarchical thinking that physics and math were more rigorous than chemistry and biology.
I do remember the first time I taught Ch 1c in the spring term. For some reason, I was feeling sort of evangelical to get all these students excited about chemistry, and then we'd get all these chemistry majors. Not many students place out of Ch 1, but the ones who do, at that time, were the chemistry majors. In fact, there weren't a lot of people in that class who were considering majoring in chemistry. I learned pretty quickly that hoping to convert this population was not something that was going to have a high degree of success. But if a couple of people thought, "Yes, maybe there is something useful in chemistry," I decided that was a win.
Also at that time, though, as I recall, there was very little interaction between the undergraduates and the rest of campus. We were in orthogonal spaces. We would be in the classroom at the same time, but there wasn't much overlap. One of the things, I think, that has been a positive development over the years is working to have more interaction between undergraduates and the rest of the campus community. I think that's great. I can't remember if there was a SURF program when I arrived. One thing, though, that was striking was to me was that I had a number of undergraduates at UCLA who were interested in working in my group during the academic year. Caltech's a lot smaller but in general, I found students here, especially when I arrived, were not as interested in academic year research as I had expected. In part I think this was related to a culture of taking a lot of classes and not focusing on other types of activities. And one of the things I think has happened more recently is to shift the emphasis from, "How many classes can you take at one time?" to encouraging undergraduates to have a much broader, and I would say healthier educational experience.
ZIERLER: When you joined, was it already the Division of Chemistry and Chemical Engineering? Or that came later?
REES: I think it's been the Division of Chemistry and Chemical Engineering from the beginning. It's not like Biology, which changed to Biology and Biological Engineering in the last 10 years. I think if you look at the earliest catalogues, it was Chemistry and Chemical Engineering. I don't know what triggered that initially.
ZIERLER: Was that administrative breakdown, with having the chemical engineers in the same division, uniquely useful to you at all?
REES: I think at the time, especially when I moved here, the chemical engineers were more separate. And since chemistry departments tend to have a strong subgroup flavor, in some sense, it seemed to me they were another subgroup. Organic chemistry, inorganic chemistry, chemical physics, biochemistry, and chemical engineering. Chemical Engineering was a bit more independent, since there are both undergraduate and graduate options in chemical engineering and so on. Subsequently, there have been a number of hires where now, there's a lot of intellectual and research overlap between faculty in chemical engineering and in chemistry. Frances Arnold, Dave Tirrell and Mikhail Shapiro really exemplified that. I'm not sure what a chemical engineer is, but I wouldn't have necessarily thought of them of chemical engineers. They have such broad interests and perspectives that it seems like a natural fit to be in one division.
ZIERLER: I asked about the undergraduates, but on the graduate side, how did you go about building up your research group with graduate students and postdocs? Was that a gradual process?
REES: Like a lot of things, it wasn't specifically planned. There wasn't a Biochemistry or BMB graduate program when I arrived. In Chemistry, graduate students tend to join research groups during the first term. In the first two weeks of the Fall term, there are six evenings where different chemistry faculty talk to the incoming graduate students and describe their research. If you're interested in taking students, as I certainly was when I arrived at Caltech, your presence at these research orientations would be the indication that you were open for business and looking for students. When the entering chemistry graduate students arrive here, they're not associated with a specific lab, and we don't have a room where first-year graduate students sit. Instead, they get assigned desks and so are distributed through the different labs. I indicated that I would be able to accommodate a student. In my first year here, I did get one student, Jongsun Kim, through this mechanism. He was assigned a desk in my lab, at least until he officially joined a group. I remember he made it very clear that he was in my lab because that's where his desk was assigned, but he wasn't making any commitments to working for me. I was like, "That's cool."
Fortunately, Jongsun decided to join my group, so he was my first Caltech graduate student. And I think five people moved with me from UCLA, so I already had a group of graduate students and postdocs. I would say by and large, my UCLA group wasn't particularly happy to have moved. They liked UCLA. Jongsun started working on nitrogenase, and I had several people working on the Fe protein and several working on the photosynthetic reaction center. Those were the two major projects in the group. I don't remember the exact timing, but one day, Jongsun came to me. He had been working on the Fe protein because that was the nitrogenase protein that we were working on, but there was also the molybdenum iron protein.
When we purify nitrogenase, we grow the cells and extract the proteins, and the MoFe and Fe proteins come out together. So you always purified the MoFe protein as part of purifying the Fe protein. We do this because you need both proteins to measure the activity, to make sure you were purifying active protein. One day, I remember, Jongsun asked me if I was interested in working on the MoFe protein. I said, "Well, I don't know." I knew that other groups were already working on it. At that time, I was a little reluctant to get into any sort of competition because it took so long to solve structures. I probably gave some vague answer, and then asked, "Why are you asking?" And he said he was asking because he had crystals of the MoFe protein.
This was just unbelievable. Jongsun had these incredible crystals, which changed the whole course of the group and to me, really exemplified the best features of a Caltech student. Jongsun was working on this project, he saw an interesting new direction, and he pursued it. And then, he asked me if I was interested in this direction. The upshot was, he ended up solving the structure of the MoFe protein in 1992, which gave the first insight into the structures of these really extraordinary metalloclusters that, frankly, has propelled my career ever since. To me, it just illustrates a lot of points. Perhaps the most important is was that this outcome wasn't because I had a great plan to work on the MoFe protein. In research, it's essential to have great students and let them feel that they can follow their nose in terms of pursuing things. Jongsun finished the structure and graduated in four years. At his commencement, he received the Clauser Award for the best thesis. The best thesis is always a subjective term, but Jongsun really got my group off to a great start at Caltech. That was my introduction to Caltech graduate students.
ZIERLER: As a benefit to our non-specialist audience out there, the phrase, "Solving the structure," as you use in your field, what would be a metaphor that would really help visualize the triumph of solving the structure as a way to convey, A, how difficult it is, and B, what it looks like to solve something that was previously not solved?
REES: Well, that's a great question. There are several aspects to that. By solving the structure, and we only approximate this, I mean figuring out the spatial positions of the atoms in a structure in a molecule of interest, in this case, the protein, to determine their location, their coordinates in space. Let's say there are 10,000 atoms in the structure. Each one has an X, Y, and Z coordinate, so there are 30,000 coordinates, and we end up with a table that has the coordinates. It's basically like a map, and if you wanted to know, "How do I get to Altadena from here?" you could look at a map and find out. "Here are the coordinates of Altadena, and I'm at these coordinates." The map shows where all the different places are, and the roads that connect them. When I say, "Solving the structure of a protein," I mean that we know the location of the atoms in the three-dimensional structure of this protein. When done properly, with some major qualifications, there's a definitiveness to the structure determination process. You have either solved the structure, or you haven't. It's like climbing to the top of a mountain and planting a flag that says, "I made it to the top of the mountain." Of course, if you have 10,000 atoms, what does it mean if some are in the wrong place, or you don't see them all, or whatever? There are some ambiguities in what it means to have solved the structure of a protein and planted this flag.
ZIERLER: And that's what you mean by approximation because you can't account for all of the coordinates.
REES: That's right. And the word structure implies a certain definitiveness. Of course, everything's in motion, so it's not like there's a precise coordinate. All the atoms are jiggling around. - that's just part of the laws of physics. There can also be parts of the molecule that maybe have multiple conformations, so they exist in different positions. Or, in fact, say they occupy a range of conformations, and you really can't see them at all because it's blurred out. It's a lot like a photograph. And you can think of this like a photograph with atomic resolution, and you can see where all the pieces are. I think of Pauling, Bragg, and the Cambridge, England group as framing many of these issues as the "structure – function" paradigm – to know how something works, you have to know what it looks like. Feynman famously captured this view as "It is very easy to answer many of these fundamental biological questions; you just look at the thing!"
For example, I think if you were talking about how a plane flies, you wouldn't typically start with equations and say, "Here's the way in which lift is generated," but you'd almost certainly be drawing pictures of wings and describing the features of the three-dimensional structure of the airplane that are associated with lift and flight. Or if you want to talk about how something works like a city or a building, you would draw pictures and indicate how traffic moves on streets and highways, where people live, shop, etc. What we are doing is the counterpart of this, but at the atomic level.
In the case of nitrogenase, what we wanted to understand, and we still are motivated by this, is how nitrogenase is able to catalyze the conversion of this inert gas around us, dinitrogen, to ammonia. To accomplish this, we need to know what the catalyst looks like to try to understand how it might interact with nitrogen and facilitate the conversion to ammonia. Those considerations were the main driving force for this work. As a related issue, no one knew the structures of these metalloclusters with remarkable catalytic properties, although there were some clues about specific features of the clusters and we were also motivated by solving that chemical puzzle.
ZIERLER: To recap, it sounds like the act or process of solving the structure evokes more of a cartographic exercise than, say, a mathematical proof or even a Rubik's cube.
REES: Exactly. It's nothing at all like a mathematical proof, where you say, "I've proved this theorem." and that means it has been proved, not that it is proved except for a few missing pieces. The structure is not as well-defined as a mathematical proof. Operationally, today it means that the structure has been published and the atomic coordinates have been deposited in a public database, where they are available for other people to look at and use, and also to find out if you've screwed up.
ZIERLER: If there is an unavoidable level of approximation, as you say, how, then, can you strive to remove ambiguity from the process so that an independent researcher working with the same materials would come to the same conclusion?
REES: If you're working on a protein structure, you now will know the sequence of amino acids that make up this protein. A protein is a polymer, and from the DNA gene sequence that encodes this protein, we know which amino acids are present at each position. The structure you solve can use that information and then try to position where all these different amino acids are in space. And that's important because sometimes, the quality of experimental data is not sufficient to reliably locate all the atoms in a particular amino acid at one position in the protein. But the fact is, you know that a particular amino acid is at a certain position in the sequence, so you can build it in. The way in which you can get around this is by getting better experimental data at higher resolution. Resolution is something like if you're painting a picture. Let's say it's on an 8.5x11" piece of paper, so if your brush has a thickness of four inches, you're not going to capture very fine detail in your painting. But if you had brush that was a millimeter, then, in principle, you could provide a lot more detail. By preparing better, more highly ordered crystals for our X-ray crystallography, we can start getting more refined views of the structure.
Now, the challenge, the thrill, and the big uncertainty in the case of the MoFe protein were these two metal centers, since the structures weren't known. If we were at a sufficiently high resolution where we could truly resolve atoms from each other, then that would solve the structure unambiguously. But our brush was too thick to separate these individual atoms. We had these blobs of electron density that we were trying to interpret in terms of detailed structures. For that, we had to integrate other information from chemistry and spectroscopy. Now, I think, we do have an accurate structure. But it took ten years, from 1992 to 2002, to get to a point through the work of postdoctoral fellow Oliver Einsle where I felt we had really defined all the atomic positions of the metalloclusters, with the possible exception of hydrogen atoms that are too light to confidently recognize.
ZIERLER: In the simplest terms, better resolution, less ambiguity, and the march continues in that direction.
REES: Yes. it's like with a camera where the resolution is related to how many pixels it has. A higher resolution corresponds to more pixels. Each time a new phone comes out, you can get more pixels and higher resolution data, and thus, more detail in your images. Another way of describing it would be to think about this in terms of pixels. In our case, it's not necessarily so much the number of pixels, but how small the pixels are in relation to the molecule that we're trying to describe. The typical unit of length in a molecule is the Ångstrom (Å), which is 10 -10 meters, which is roughly the distance between covalently bonded atoms. You would like to have your pixels being a fraction of an Ångstrom, and then you would have great detail. When we start out, the resolution is typically more like 2 Å, which means, in fact, the pixel sizes are larger than the atomic spacing. And so, we can't really individually identify atoms, but we know roughly the shape of each amino acid residue, and we know something about their chemical character.
ZIERLER: I can't help but ask, in the world of crystallography, is there a Moore's Law of resolution?
REES: I don't know about resolution, but certainly in terms of the pace at which structures are being solved, I think, you could convert into a Moore's Law type of relationship. It turns out that, ironically, the very first structures that were solved, back when all these things were quite laborious and difficult, were of high resolution. If everything wasn't almost perfect, you would fail to solve the structure because there was no way you could improve the structure by computational methods. As we got more information on protein structures and computational methods improved, the resolution started to get worse because you could start interpreting maps that were not as good quality as the initial ones. And so, I don't think there is a Moore's Law for resolution.
ZIERLER: Going back to your very first graduate student at Caltech and bringing that narrative all the way to the present, a general, even retrospective question. I asked about the motivations and interests of undergraduates. What about the changing motivations and interests of your graduate students in terms of their desire to go into academia, to go into government, to go into industry? Has that more or less been stable over the years? Or have there been changes in those trends?
REES: I think over my 30 years here, what I've seen from students and postdocs is the first 15 years, there was definitely more of an emphasis on academics and academia. I think that was maybe in part because, especially in industry, there were fewer opportunities for employment in this area at that time.
ZIERLER: But embedded in there, of course, is the larger story of the growing importance of biotechnology in industry.
REES: Right. In the last 15 years, though, I think it's flipped. It's not 100% to 0 and then 0 to 100%. But I think, also, that is a reflection of the opportunities in industry for people with these skills who want to directly contribute to something that helps society. In a funny way, I think you can do that maybe more readily in industry than in academia because you're working on projects that are intended to be directly used by society.
ZIERLER: Not to be blunt about it, but also to earn a salary that, in many cases, dwarfs what academics can make.
REES: Yes. I really enjoy academia, but I would never tell anyone that this is the only job, that everything's perfect, each day is a blast, etc etc. There are a lot of really wonderful aspects to academia, but there are a lot of challenges and hard parts as well. I think people sometimes look at faculty and make the decision, "Is this the lifestyle that I want or not?" I think it's complicated. And obviously not every student, in any event, could go into academia. I think Caltech alone probably would turn out as many graduate students a year as you would have in research university positions. So I think it's important that students and postdocs consider a range of options.
ZIERLER: To that point, is there a specific time in your research career when the questions you were asking became more and more relevant to translational research in human health?
REES: I would say as we moved into membrane proteins, that connection was definitely stronger. And I haven't thought about it this way, but maybe it's not been unexpected that, at least off the top of my head, people in my group who were working on membrane proteins may have been more likely to go into industry, biotech or pharma. Again, that direct connection is not what floats my boat, but I think it's an important reason why society is supporting this research. I was interested in the basic protein structure underlying these phenomena, which then could be developed in a number of different directions.
ZIERLER: A nomenclature question as it relates to the science. Is crystal morphology the end goal, and crystallography is the means to understand crystal forms? How does that work?
REES: Crystal morphology would refer to the shape of the crystal, and sometimes it has an aesthetic quality as well. I would say that it is essentially irrelevant to what we're trying to do, which is using the technique of X-ray crystallography to infer where the atoms and electrons are located that scatter the X-rays. The important property, which we just touched on, is resolution. And resolution is not, in any way I've ever been able to understand, related to morphology, how nice the crystals look. We've had some beautiful-looking crystals–and this is not so uncommon–that are ordered on some macroscopic level, but microscopically, the order's just not there to scatter the X-rays to sufficiently high resolution. And we have other crystals–and this is rarer–that look pretty bad but, in fact, scatter X-rays well. We have to have the crystals, and I think part of the appeal of the whole process is producing and getting to look at these crystals, which I think are pretty spectacular. But that's sort of irrelevant for how well they may or may not scatter X-rays.
ZIERLER: As would be expected, the titles of your articles are obviously highly technical and dry. One that sticks out is a 1995 article you wrote with Michael Adams, Hyperthermophiles: Taking the Heat and Loving It. The first question there is, why the playful title? And what were you trying to convey about hyperthermophiles?
REES: The period of the 90s into the early 2000s was really a golden area in structural biology, especially of metalloproteins and membrane proteins. The technology for solving structures by X-ray crystallography was getting better and better, but it still was sufficiently difficult that it wasn't so widespread. Many groups were working on other types of proteins that had some biomedical relevance, such as interacting with DNA, or enzymes that catalyzed some desirable reaction. We were interested in metalloproteins, which is a smaller but not insignificant community. Harry Gray and Sunney Chan are part of the community. Mike Adams was (and still is) at the University of Georgia, and he works on metalloproteins isolated from a fascinating organisms known as Pyrococcus furiosus. P. furiosus is a hyperthermophile that grows at very high temperatures, biologically speaking, such as found near hydrothermal vents in the bottom of the ocean. So they grow at temperatures near the boiling point of water. When one boils an egg, the egg white protein denatures and loses its native, defined three-dimensional structure. That is typical behavior for many proteins when heated to the boiling point of water. However, for hyperthermophiles living at 100 degrees C, the proteins are obviously working just fine under these conditions. The question was, what did these proteins look like? Could we understand the origins of this high-temperature stability?
Equally important to me is that Mike was also interested in metalloproteins, and he identified a set of proteins that had some fascinating metal centers. And we worked on several, but the one that came to us was initially called RTP for red tungsten protein because it was red and it contained tungsten. What does a protein look like that had tungsten? I remember John Bercaw, my chemistry colleague, saying, "tungsten can do everything molybdenum does, but it needs a higher temperature." There are two ways molybdenum can be used in the active site of enzymes. One was in the MoFe protein of nitrogenase, but it turns out that's the only example that has this particular type of molybdenum containing center. More common is what's known as the molybdenum cofactor, or Moco. Occasionally, this species was found in the tungsten-containing form including the red tungsten protein. At the time we started working on this structure of this protein, there were no structures for any Moco enzymes. The postdoc working on this project, Michael Chan, was Sunney Chan's son. Michael solved the structure, and we were able to figure out what the active center molybdenum cofactor, looked like, at least in the tungsten containing form.
In addition, Mike Adams had isolated other proteins from P. furiosus, and we solved several of those structures. In the article you mentioned, we reviewed these structures and while there were several likely contributors to their stability, it wasn't immediately obvious why they were so much more stable than proteins from more typical organisms. I think if we didn't know they were more stable, we wouldn't have thought there was anything unusual in the structure. Whatever the origin of the stability, there was significant biotech interest in proteins from hyperthermophiles because you could put them in laundry detergents to digest proteins and help remove stains at high temperatures in washing machines. There have also been applications to the polymerase chain reaction, or PCR, tests that we've been hearing a lot about lately. PCR involves running reactions at high temperatures, so having a DNA polymerase that will function at higher temperature was a key part of that ability for those assays to work.
ZIERLER: Onto the administrative side, in 1996, you're named executive officer of the biochemistry graduate option. Tell me about that role. What were the circumstances of becoming executive officer, and what were some of your responsibilities?
REES: To try to provide some perspective, Caltech is a small place with a compact administrative structure. The graduate options, historically, cover a broad range of topics. The Chemistry graduate option and the Biology graduate option, for example, were the two primary PhD programs that would be of interest to a prospective student interested in biochemistry (remember I applied to the Biology option as a prospective graduate student). But one of the consequences of this for those of us in biochemistry, was, if you were a perspective graduate student interested in biochemistry, you may not be interested in either a chemistry program or a biology program. While I was able to get outstanding students through chemistry, and occasionally through biology, collectively we felt that there was a pool of students who would be interested in a specific biochemistry graduate program that we were missing because they didn't want either a chemistry or biology degree. With a Biochemistry program we could design a curriculum to train biochemists that didn't necessarily fit well with either the chemistry or biology graduate programs. In addition, there's a big cultural difference, between chemistry and the biosciences, having to do with graduate rotations. At the time we were discussing the Biochemistry option, chemistry graduate programs generally wanted their students to join labs during the first term, whereas first year rotations were an important part of biology graduate programs. In my graduate experience, the rotation I did in structural biology was transformative.
There was a group of us in the CCE and Biology divisions who were interested in establishing a graduate program in biochemistry, supported by the chairs of CCE and Biology, Peter Dervan and Mel Simon, respectively. Ultimately, that effort was successful, but it was a painful process for a number of reasons. I also learned an important lesson about administration in this process, that you never go into a major meeting without knowing what the outcome was going to be. [laugh] There were some real false starts in terms of getting the Biochemistry option going, that I thought were unnecessary, together with a lot of important, but also frustrating, dynamics between faculty.
Part of the academic culture that contributes to Caltech's success, but is also a point of frustration, is that you essentially need to have everyone on board when you're making faculty hires or major administrative decisions. Faculty have a lot of views on things, so it's important to really work through these. It's probably no different than many other activities in the world. And so, the process of creating the Biochemistry option was pretty involved, and I felt there was a lot of friction. But ultimately, I think it ended up being a really successful graduate option. And even though I don't like the idea of creating more administrative structures, sometimes this is necessary. And certainly, bioengineering, neurobiology, medical engineering and other graduate programs at Caltech have subsequently gone through the same sort of considerations. They're spanning important interfaces, which is a strength, and while we also don't want to cut the pie too finely, sometimes the current pieces are inadequate.
After the Biochemistry option was established, I ended up as Executive Officer and Pamela Bjorkman served as the Option Rep. The administration of the option was split between CCE and Biology. We switched at one point, and so Pamela then became Executive Officer and I became the Option Rep until we handed off to the next group.
ZIERLER: And just to get a sense of the administrative functions and the hierarchy, because Caltech is so idiosyncratic in the way it's organized, were you essentially, as executive officer, serving as a department chair for the biochemistry graduate students? Is that perhaps one way to think about it?
REES: These issues are still relevant in 2021. [laugh] Subsequently, Biochemistry became the Biochemistry and Molecular Biophysics (BMB) option. In the CCE division, there is a biochemistry subgroup. Technically, the executive officer is an administrative function within a division. while the option rep has responsibility for overseeing the administration of a graduate option. There were more details and nuances when Pamela and I were doing this, and we worked together to administer the Biochemistry, and subsequently BMB, graduate option.
ZIERLER: One aspect of your research trajectory that I've been quite curious about, and we've now finally worked up in the historical chronology to, is your decision to become an adjunct professor back at UCLA in the department of physiology. Lots of questions there. First, what was the magnetic pull back to UCLA? Why serve in an adjunct professor capacity? And why physiology?
REES: Great questions. To answer those, there's one piece I need to fill in, and that is getting appointed as an investigator of the Howard Hughes Medical Institute.
ZIERLER: That was my next question. I assumed that they were related, but I wasn't sure.
REES: They were absolutely related. But in the sequencing, the adjunct appointment would have been irrelevant without an HHMI appointment. I had mentioned at UCLA, Dave Eisenberg and I tried to get HHMI support for structural biology, but we were not successful. I don't remember the complete history, but Caltech did have HHMI investigators at that time. When Pamela Bjorkman came here, she was also appointed as an HHMI investigator. It was becoming clear to me, especially as we wanted to get into membrane protein structural biology, that resources were fundamental, just like politics and money. Money is critical. And a lot of the top structural biology groups had HHMI funding. At that time, Hughes appointments were several-stage processes. By the mid-1990s, there were periodic competitions for new HHMI investigators, starting with being nominated by your university. And each university could only make a limited number of nominations. I was nominated by Caltech, survived the competition, and was subsequently appointed by HHMI as an investigator in 1997.
As part of that process, because HHMI is a Medical Institute, the interpretation was that you had to have an appointment with a medical school. I think this was how they were interpreting guidance from the IRS in terms of how HHMI could supporting medical research. You needed a medical school appointment. My colleagues at Caltech who had Hughes appointments were all going through the USC Medical School. But having been at UCLA, that was a hard step to take. On a more serious note, I also didn't know anyone at USC, but I did have some connections to the UCLA Medical School. The closest connection was with Ron Kaback, who worked on transporters and arrived at UCLA as an HHMI investigator shortly after I had moved to Caltech. Ron was in the physiology department, and he discussed this situation with his department chair, Ernie Wright, who made this happen. An appointment like this would be very difficult to generate at Caltech. But between Ernie Wright and Ron Kaback, they shepherded my adjunct appointment through at UCLA. It worked out well since I had common interests with Ron, Ernie and other members of the Physiology department, and so it was a natural fit.
ZIERLER: A chicken and the egg question. Were you onto a set of research for which HHMI was just an ideal partner? Or were you looking to get into something, and HHMI served as a launchpad, for which it might not have been otherwise possible?
REES: Well, it was both. I think of the mid-90s as the golden era of structural biology, specifically macromolecular crystallography. I had always been interested in membrane proteins. But the problem was, first, you have to get the membrane proteins. Some membrane proteins, such as photosystems from plants or photosynthetic bacteria, or respiratory enzymes from mitochondria, can be isolated in sufficient quantities from natural sources. But most membrane proteins are only present in small amounts and so we needed to make them in larger quantities using recombinant DNA tricks. When I first got to Caltech, the technology was too challenging for my group. But that started to change in the 1990s.
A key breakthrough for our work was a 1994 publication in Nature by Ching Kung, from the University of Wisconsin. Kung had identified a stretch or mechanosensitive channel in the membranes of E. coli. An incredible paper, and I thought about this a lot when the Nobel Prizes were announced recently in Physiology or Medicine because they were for the discovery of mechanosensitive channels in humans. These were tremendous accomplishments. But Kung had originally done this in the mid-90s in bacteria, and in the process established how mechanosensitive channels could be identified.
The channel that Kung and his group identified was a membrane protein he named the mechanosensitive channel of large conductance, or MscL. MscL was a membrane protein with the remarkable property than when you stretch the membrane, the channel would open and conduct ions, and when you relaxed the membrane, the channel would close. So that was exciting because the function of the channel reflected a change in structure in response to applied mechanical tension. But even more importantly, the E. coli MscL protein only had 136 residues, so it was a rather small protein. In fact, originally, I was thinking that while cloning might be too hard for my group, perhaps we could make this channel synthetically. I even discussed this with Jack Richards. In hindsight, this was an idiotic idea. It would've been essentially impossible at that time to synthesize MscL. But cloning these channels wasn't impossible for postdocs with a suitable background in molecular biology.
At this point in the membrane protein structure field, the reaction center had been solved, but there were no structures of channels or most other membrane proteins. So I started getting terrific postdocs who wanted to work on membrane proteins. Two postdocs, Geoffrey Chang and Rob Spencer, joined my group and I hired a technician, Allen Lee, who's still working with me, to try to express, purify, crystallize and ultimately solve the structure of MscL. But now, unlike the work we had been doing on nitrogen fixation, where we can grow the bacteria on sugar and a few minerals, now, we're growing E. coli, and you have to provide it with a more complicated medium. The materials required for cloning and working with membrane proteins were expensive and so our supply budget started to expand. [laugh] The molecular biology was something that you could parallelize as well, so the number of targets kept expanding. And, our daughter Abby was born around this time, so it was a very exciting period, but also very intense, both inside and outside of the lab.
So, the overarching practical question I needed to solve was how to fund this juggernaut? The answer to that question was provided by my appointment as an HHMI investigator in 1997. This allowed me to support more postdoctoral fellows, as well to make staff appointments that were essential to the efficient operation of an increasingly large research group. I was fortunate to be able to appoint Phoebe Ray as my administrative assistant – Phoebe just retired from Caltech after 38 years of service to the Institute, including over 30 years working with me to keep things running smoothly. I was also able to appoint superb technical staff, Allen Lee and subsequently Jeff Lai, and then Welison Floriano as our computer manager.
The HHMI funding came at a propitious time as it allowed me to appoint a great cohort of postdoctoral fellows, which was essential because at the same time, my lab was losing its appeal to graduate students.
ZIERLER: What's the broader story there? Why would it become less attractive to graduate students but more attractive to postdocs?
REES: That's one of many mysteries that I don't understand. I mentioned my first Caltech graduate student, Jongsun Kim, and in the two years after that, I think I got six more students. So when I first arrived at Caltech, my group seemed pretty popular with graduate students. I think partly–I saw this when I was in the graduate office–that when you come to Caltech from the outside, you're a blank slate. Whereas if you've been at Caltech for a while, students will have an opinion about your group. You might think coming from the outside would create uncertainty, but in this case, I think students are willing to take a chance on the unknown because they think they know too much about the available options for advisors.
So my lab was attractive to students when I first arrived at Caltech. After a few years, I had 10 people in the group – mostly graduate students with three postdocs, Pinak Chakrabarti, Hiromi Komiya and Leemor-Joshua Tor, who were outstanding crystallographers and essential for both the research work and the training of graduate students. By 1992, several of the projects came to fruition, including Jongsun's nitrogenase MoFe protein structures, but also the nitrogenase Fe protein spearheaded by UCLA graduate student Millie Georgiadis (this is the project I had started studying as a postdoctoral fellow with Jim Howard), and the structure of the P. furiosus rubredoxin protein solved by Michael Day, a Caltech graduate student, in collaboration with Mike Adams. At that point, for reasons I don't understand, graduate students essentially stopped joining my group. I don't know if this was a reflection of something I was doing that was unattractive to students, or if students weren't interested in my research, or if it was just the statistics of small numbers. But in the meantime, postdocs were seeing what we were accomplishing, and they were very interested in my group. I think of my group as a certain size and would give priority to graduate students, but since they weren't joining my group, I made up the difference with postdocs. And of course, it was easy, then, just starting to think about all the projects we could do with experienced postdocs. That was very exciting.
But, this meant the group was growing, and with more expensive postdocs needing more expensive supplies. Even though people in academics don't get paid much, salaries constitute about 75% of a research budget, and I didn't know how to support all these people. So HHMI was an obvious way of doing this. Not so much for nitrogen fixing and metalloproteins, but because we wanted to work on membrane proteins. It was easy at the time to make the argument that membrane proteins were a target-rich environment. We were one of the few groups that had experience working on membrane protein structure with the reaction center, even though in a way, it wasn't so relevant because the crystals literally were driven up here from San Diego. But we had worked on membrane proteins and established a name and reputation in that field. So I would say HHMI was essential as part of getting support for this research.
ZIERLER: To go back to the affiliation with UCLA, the department of physiology would not be my first guess to be your affiliation. Why physiology? What can be gleaned from that?
REES: Actually, in fact, it was a natural home. But the immediate answer is, I went there because my friend and colleague Ron Kaback was there. Physiology departments were often physiology and biophysics departments, with significant interests in channels and transporters, the flow of ions across membranes, potassium channels, sodium channels, all these uptake systems, and so on.
ZIERLER: Was this a teaching appointment as well?
REES: No, I would say it was by courtesy. They really helped me out. I would go to some seminars, give a lecture, visit the department maybe once or twice a year, but there were no teaching responsibilities.
ZIERLER: To go back to the broad issue of translational research with human health, did this get you more involved in that general area at all?
REES: No. I would interact with especially Ron Kaback, Ernie Wright, and others, but this was a basic science department and our discussions focused on basic questions of membrane protein biology and membrane transporters.
ZIERLER: I know this research is important for your overall agenda, but from an institutional perspective, why HHMI, absent UCLA, absent the department of physiology was this something that you could not pursue with the infrastructure you had at hand at Caltech? Is it a funding question, an intellectual partnership question?
REES: It was primarily a funding issue. Again, Caltech was great in setting up a lab. But once you get things running, it's your responsibility to keep things running. There are things where Caltech has made enormous investments, like the synchrotron beam line used by the Molecular Observatory or the Cryo-EM facility. The Molecular Observatory is a great example of this, since the support of the Gordon and Betty Moore Foundation to Caltech made possible the construction of a remarkable beamline at the Stanford Synchrotron Radiation Laboratory that was a game changer for us.
When it comes to supporting graduate students, there are institutional and divisional fellowships, and there's the institutional TA budget, that help to support graduate students, especially during the first year or two. But at some point, the responsibility for supporting students and postdocs rests on the faculty member. Traditional funding mechanisms work better when you have some accomplishments. What I was doing at this point in our work on membrane proteins was trying to move into a new direction that I felt was absolutely at the right time. But, I obviously didn't have any accomplishments, so I couldn't be writing successful NIH grants to support my graduate students and postdocs. Because at that time, they would all ask, "Do you have crystals?" And we'd say, "No, we're working on it. We're optimistic. We're trying all these things. etc. etc." And the study sections would say, "This is an important project, but it's just too uncertain until you have crystals." It was really one of these things that would've been impossible to work on at other places. Caltech provided a great intellectual environment. I had the equipment and all the other stuff I needed for this research. But at some point, you have to pay people. When you reach a certain critical mass, you have to find a way of bringing in the revenue stream.
ZIERLER: I'm not sure if you're aware, but Caltech trustee Charlie Trimble has this wonderful program of what he calls funding risky research at Caltech, research that's so provisional, it's not ready for the primetime, but it's really quite exciting and promising. Was there anything like that 20, 25 years ago, where you could tap into sources before you were ready for NIH?
REES: I don't want to paint an overly stark or bleak view. We are fortunate at Caltech to have a number of resources. Again, there were fellowships, a lot of my students were supported in TA-ships and/or had external fellowships like the NSF graduate research fellowships, there were other institutional supports like the Beckman Institute. But the burn rate and the rate of time that's required to see a project through fruition – I didn't know how long it would last until we got something to work. I think Caltech is great at working with you to find resources and so on. But this was a pretty large effort. And of course, I didn't know how long it would take before it worked. And frankly, what one person says is risky, transformative work, "We've got to support it," is something another says is crazy and will never work. That's part of the problem with these risky projects. If they work, they could be transformative, but there's no guarantee. But at some point, you do have to believe it'll work because if you don't, the funding people are not going to put up the money. This was the type of problem that HHMI was supporting. And it was not unprecedented at Caltech because a number of my colleagues in the Biology division were already Hughes investigators.
ZIERLER: This notion of believing it's going to work harkens back to an earlier discussion we had about the role of theory or even intuition in your research. At this juncture, before you have the crystals, but you have the vision, how does that work in terms of your research? These are not willy-nilly ideas. There is a theoretical grounding for where you see this going. Let's take it back to that question about how you see the role of theory or intuition before the crystals are a reality.
REES: Frankly, I think intuition is critical. We've talked a bit about my decision-making at several critical junctures. I got into structural biology not because of any purposeful intent, but through a course and a rotation I took with Steve Harrison. I ended up working on nitrogenase because of this chance encounter with James Howard while he was on sabbatical. In both cases, these intuitively seemed like the right decisions to me, without going through a detailed analysis of the options. The one goal where I felt like I did have a long-term vision for what I needed to be studying was the structural biology of membrane proteins. But for the first 15 years of my faculty career, it just didn't seem like it was the right time with the skillsets that were in the group, starting with my expertise, or lack thereof. What I realized around the mid-90s was that this was the time. And Kung's 1994 paper on MscL helped me realize that this was something we could actually do, even though we didn't have any experience with the technology for making membrane proteins. I think the fact that MscL was a small protein, it was from E. coli that was easier to grow, and we had a fermenter, so we could grow E. coli on large scale convinced me that this wasn't such a crazy thing to do.
It seemed to me like it was exactly the time to be working on membrane proteins. This was the one time in my career where my intuition was crystal clear about the science I needed to do next. I wish I had that clarity for what's next right now. But I just knew in the mid 1990s without having done a detailed critical analysis of all the options that it was the time to work on membrane proteins. And that window would pass, and if I didn't take advantage of it, I'd miss it. In that case, intuition was really important.
ZIERLER: What came of the research? Did you end up getting the crystals you were looking for?
REES: We did. Rob, Geoffrey and Allen solved the structure of MscL and we published it in 1998. It was the same year that Rod MacKinnon published the structure of the KcsA potassium channel for which he received the Nobel Prize. And quite appropriately – while KcsA is also a bacterial channel, like MscL, unlike MscL, it tied into decades of mammalian and human physiology and helped understand a tremendous body of work. Our contribution was not at that level, but I was really proud of what we accomplished with MscL- we had established a workable approach to solving membrane protein structures, and as a result the lab was getting great postdocs who wanted to take these approaches in new directions.
ZIERLER: Another nomenclature question. Conformational variability. I would guess that this would be in an advanced math textbook. What does it mean in your field?
REES: It means that there can be variation in the conformation of a protein. I think it's probably like a lot of terminology in that it's sort of hand-wave-y. Basically, it denotes the behavior of a protein that can exist in distinct conformational states. This is usually recognized by capturing one protein in different conformations–for example, perhaps you'll find one domain in different positions–or it can also refer to a set of related proteins that have distinct conformations. In a way, maybe it's sort of like LEGOs. You can take certain pieces but combine them in different ways so that they have different structures with the same basic materials. I don't know if that's a great analogy, but it sort of captures what we're interested in. A lot of times, proteins, to carry out their function, will need to adopt multiple conformations so they can do different types of things. One way we try to get at that is by looking at the range of structural space that's accessible to a protein. A good example of this with nitrogenase was provided by the changes in the Fe protein structure associated with complex formation to the MoFe protein, in structures that were solved by postdocs Hermann Schindelin and Caroline Kisker, using biochemistry developed by Jim Howard.
ZIERLER: Do you have a general memory of when bioenergetics became really front and center in your research agenda? Is it in this golden age of the late 1990s?
REES: Bioenergetics arises more from the physiology side in terms of how energy is stored in cells. In the mid-90s, we were working on both metalloproteins and on membrane proteins that are often studied by distinct communities. But in my group, they were connected because we were interested in the structures to understand their functions, and often they interacted with ATP, this basic energy currency. So that was a unifying thread for a number of our projects.
ZIERLER: And what about DNA research? Is this happening in tandem?
REES: Well, this is a great example of what graduate students, postdocs and your colleagues bring to a research program. The research interests of faculty at Caltech are quite broad. We have the small size of the faculty covering grandiosely all science and engineering, and we have a fair number of graduate students and postdocs. So the research interests of groups tend to be pretty broad and we often do things that are pushed by interests of our colleagues. Jackie Barton's group was interested in a new type of metallointercalators and how they bound to DNA in a sequence-specific way. The first postdoc who joined my group at Caltech was Leemor Joshua-Tor who had this tremendous crystallographic experience working on DNA structure. She also knew how to freeze macromolecular crystals, which was a rare skill at the time. I also had a visiting scientist from the University of Kansas, Kristin Mertes, who was interested in DNA and Jackie's work. Between Leemor and Kristin, we'd made some efforts with Jackie's group to try to crystallize DNA bound to her metallointercalators. I have to confess this wasn't really my top priority, but it was of interest of people in my group, and I enjoyed working on interesting projects in a colleague's lab. Those efforts led to crystals to crystals that diffracted really well. But for whatever reason, we were unable to solve the structure, which was quite frustrating. Subsequently, a graduate student, Clara Kielkopf, joined my group, and she was also interested in DNA, and started working with both Jackie's group and Peter Dervan's group on their small molecules bound to DNA. Clara was able to get these really nice, beautiful structures of DNA with their compounds, rhodium intercalators, and these polyamide groups binding to DNA that were really exciting. Eventually, Clara also got a very high-resolution DNA structure, where you could see the hydrogens in the base pairs that was pretty cool. All this grew out of the collaborative, collegial atmosphere of Caltech, the interests of people in my group, and working with my colleagues' groups to use structure as a tool to tackle really important problems in their group.
ZIERLER: Looking into the 21st century, you wrote about nitrogenase at a crossroads. I wonder if this is connected to the golden age of biochemistry coming to an end in some ways.
REES: I think in a way, maybe it was. But that's not what I was recognizing. I'm not sure I was successful, but I was trying to refer to the crossroads in the sense of Robert Johnson making a pact with the devil to acquire superb guitar skills. In the case of nitrogenase, I was trying to draw some parallels that although we had quite a bit of information on nitrogenase, there were still mysterious features of the mechanism that we just did not understand that we thought the structure would reveal. And the structure didn't, at least immediately, give us that insight. What I was trying to refer to was whether one would make a pact with the devil to figure out how nitrogenase works. I was writing a number of reviews at that time and maybe trying to find some different way to cast the problem. But it did have that sort of element that, "We have all this information about nitrogenase, but we're floundering at understanding how it works, and maybe I would sell my soul to achieve that level of enlightenment."
ZIERLER: What was the mystery exactly? What were some of the key unresolved questions at that time?
REES: How nitrogen binds and gets reduced, which is still unresolved. The basic catalytic principles are still, I'd say, not fully resolved yet, in part because there are these intermediate states. We're looking at a non-catalytically active form of nitrogenase and trying to understand what's happening in the transient state during turnover. So we don't know, in my view, a lot of basic features about how nitrogen can bind to the active sites. I think at some level, we may have some general understanding, but there are a lot of dots between dinitrogen and ammonia that we haven't yet have filled in.
ZIERLER: One interesting area of inquiry you pursued around this time was the role of iron sulfur metalloclusters serving, if I understand correctly, as a bridge between the inorganic and the biological world. Why iron sulfur metalloclusters, and what's the significance of the bridge?
REES: I think there's always been a fascination with the origin of life and how the basic chemistry evolved. An enzyme like nitrogenase just doesn't spring up overnight. In looking at the ways in which one can imagine some of the initial chemistry taking place, there are, I think, parallels between iron sulfur clusters found in proteins and iron and sulfur containing minerals. There is a framework by Gunther Wachterhauser for thinking about the origins of some of the basic chemical reactions that would be needed to sustain life that highlight chemical reactions taking place on the surface of pyrite, which is an iron sulfur containing mineral. One can imagine that, early on, you had a protein where the active element was provided by some inorganic components that may have been co-opted from the inorganic part of earth. These iron sulfur clusters do seem to show the overlap or interaction between the inorganic world, iron sulfur minerals, and the more organic chemical species found in proteins. And for whatever reason, it may just be the nature of the reactions, but some of the basic reactions involving nitrogen metabolism, sulfur metabolism, and even carbon metabolism involve iron sulfur clusters.
ZIERLER: I'm curious if this line of research pulled you into that large and quite interesting community of origins of life researchers generally.
REES: A little bit, but not so much. I think that's a fascinating area, but it's also one that I find under-determined. There are a lot of models that can fit the facts.
ZIERLER: You're a lab guy is what you're saying!
REES: Yes. [laugh] I've been to some meetings where these questions have been discussed, and it's fascinating. I think the key questions relevant to understanding the evolution of biological nitrogen fixation concern at what point in the earth's history did it become necessary to fix nitrogen? What was the composition of the earth's atmosphere when it was formed, and how does that compare to what you see on other planets? An understanding of early planetary history is important because presumably, there was enough reduced nitrogen in the early stages that it wasn't limiting, but at some point, it absolutely became necessary to fix nitrogen. So trying to understand when and how nitrogenase evolved would be a really fascinating thing to understand.
ZIERLER: Obviously, it's not your field, and it's too theoretical for you, but given the fact that we're on the topic and that you've thought about this, does your focus on iron sulfur metalloclusters suggest to you that there's a more plausible origin of life story than another? For example, fungus, asteroids, some brew in the ocean? Do you have any sense that this line of research yields any greater insight into these fundamental questions?
REES: No, I don't. And I think partly, that's because there are so many pieces to life. And all the pieces have to work together for life to be possible. Some people focus on RNA and information transfer. Other people focus on the basic metabolic reactions, how to build the building blocks. Other people are focused on bioenergetics and how to store energy. Other people focus on formation of the compartment that separates the inside of a cell, where you can have chemistry in a compartmentalized fashion, from the outside. Those are all important. A cell has to have all those things working. I guess I've focused more on metabolism or the membrane part because that's more interesting to me. But these are all really important pieces. I just don't see that there's enough data for me to be able to weigh in intelligently.
ZIERLER: Last question for today. Given how important Oxford and Cambridge are historically in your field, tell me about your visiting professorship to Oxford. What was that like, and what were the circumstances of that appointment?
REES: This is an interesting story, at least to me, about the Vallee Visiting Professor program that allowed me to spend six weeks at Oxford with few formal responsibilities. I'd never been to Oxford. The thing that was sort of ironic was, the foundation that supported this professorship was the Bert L. and Natalie K. Vallee Foundation. Bert Vallee was a biochemist at the Harvard Medical School, also interested in metals, and at one point had been working with my advisor, Lipscomb, on the enzyme carboxypeptidase, that had been the subject of my PhD thesis. At some point in that project, Lipscomb and Vallee had a falling out. I, for a long time, only saw Lipscomb's side of this. Vallee, to oversimplify things, had done some beautiful biochemistry on carboxypeptidase, but he also focused on why the crystal structure was not relevant to the solution structure – this was a common concern in the early days of protein crystallography. I think Lipscomb liked my PhD work because it provided him an opportunity to fire a few salvos back on why the crystal structure of carboxypeptidase was absolutely relevant. So when I was invited to be a Vallee Visiting Professor, I was a little bit nervous. When I accepted, I said, "Oh, by the way, if you don't know, I did get my PhD with William Lipscomb." And they said, "Oh, we know all that. It's OK."
So that was a huge relief. I went to Oxford for two three-week periods. When I got there, I didn't know many people. I had never been to Oxford or Cambridge, even though they were so central to the origins of structural biology. I'd never even spent much time in England. My hosts were Louise Johnson and Fraser Armstrong. I just went, and talked to people, and worked on my own thing. I think I had to give a seminar or two, which was enjoyable. I had a wonderful time.
In terms of energetics, I did get to meet R. J. P. Williams, who, with Vallee, had earlier proposed the entatic state hypothesis that was a concept about enzyme active sites that Lipscomb couldn't stand. Williams was incredibly bright and opinionated - he always seemed to have a strong view and took one side or another. I asked him about that and he explained that his father told him that the middle of the road was the most dangerous place to be in. [laugh] Anyway, I got to meet a number of people I hadn't met before, as well as visit Cambridge for the first time. It was really a wonderful experience.
ZIERLER: That's a great place to pick up for next time.
[End of Recording]
ZIERLER: OK, this is David Zierler, Director of the Caltech Heritage Project. It's Wednesday, October 20, 2021. Once again, it's my great pleasure to be back with Professor Douglas C. Rees. Doug, great to see you again, as always.
REES: Great to be here. Looking forward to getting through the Caltech years today. [laugh]
ZIERLER: We'll start with the science and the chronology in the 2003, 2004 area. My mind and I always focus on the intriguing titles in publication lists. I'd like to ask you about Breaching the Barrier, in 2003, the Science article you wrote. What was the barrier, and what needed to be breached?
REES: This was not a research article, but rather a perspective or review that was written with postdoctoral fellows Kaspar Locher and Randal Bass to accompany the publication of a pair of articles in Science related to the structure of membrane proteins known as transporters. All cells are surrounded by membranes, and everything that gets in or out of a cell has to go through a membrane. It's sort of like the border control. There are specialized proteins known as transporters in the membrane that allow the passage of the right molecules in and out of cells. We were writing about two papers describing the structures of related transporters that are involved in getting small metabolites, like lactose and glycerol phosphate, into cells. In a sense, breaching the barrier.
Cells are surrounded by a barrier, but there are ways in which molecules can get in and out of cells mediated by these transporters. We were asked to write this article because the preceding year, we had published the structure of a transporters that was the first member of an important class of transporters known as ATP binding cassette, or ABC transporters, that use ATP to move molecules in or out of cells. Kaspar Locher, a postdoc working with one of my staff, Allen Lee, who's still here, succeeded in determining the structure of the transporter from the bacteria E. coli that's involved in bringing vitamin B12 into cells. And that was a really exciting advance that got us into the field of transporters. We were then asked to write a perspective on these two transporters in the subsequent year that are involved in getting other molecules in the cells.
ZIERLER: Was this work that was related to other large things that you were doing at the time?
REES: In the mid-90s, there was a real opportunity to get into membrane proteins. We'd originally started by working on the photosynthetic reaction center with George Feher and fumarate reductase with Gary Cecchini and my then graduate student Tina Iverson. These were protein systems that were being functionally studied in other groups, and we then were involved in a collaboration to determine the structures by X-ray crystallography. Starting in the mid to later 90s, we realized that we could actually do all the biochemistry in our group. We started out with the mechanosensitive channel MscL. We were focusing on membrane protein families where there were no currently available structures.
The first structure of any member of a larger family, say, ABC transporters, would really illuminate a whole body of biochemical and biological literature on the entire family. We worked out a so-called funnel system for trying to solve the structure of one (or more) members of a family of related proteins. To do X-ray crystallography, you obviously need crystals. Just like salt, sugar, or other molecules, we can crystallize proteins, but they're a bit more finicky. And our experience had been that some proteins, not that they want to crystallize, but some proteins crystallize more readily than others, or we can more readily find the conditions that lead to suitable crystals. Our idea was that we would try a lot of related proteins to find the one that works – effectively we were trying to increase the number of shots on goal to find some way of getting at least one shot in the goal.
There are two limiting extremes for the structural characterization of a system. One is, you can work on one protein and just do whatever it takes to solve the structure of that protein. Alternatively, you can try lots of different proteins, trying to find the one that works. That's the shots on goal or what we might also call the speed dating approach. You want to try a lot of possibilities to hopefully quickly find one that works. Our work on MscL, a second mechanosensitive channel MscS and subsequently this ABC transporter for vitamin B12, were the results of such screens. And it allowed us to get structural information about families of membrane proteins where there was no prior structural information.
It was a very exciting period in my group. It's like exploring a new continent. It may not matter exactly where you land, [laugh] but that is the starting point for subsequent exploration and making observations that people haven't made before. Eventually, however, a lot of the territory's explored and discovered. And then, in fact, you have to be very strategic to find areas that may have been overlooked. But there's something about that first opportunity to do that exploration that's a real thrill.
ZIERLER: On the administrative side, we already talked about your work as executive officer for biochemistry graduate option. In the mid to late 1990s, you had a bit of a trifecta as executive officer. You were executive officer for Chemistry overall 2002 to 2006 and then a longer term as executive officer, Biochemistry and Molecular Biophysics graduate option. So between the graduate option, the length of the term, or the overall chemistry designation in the middle tenure for executive officer, I wonder if you can compare and contrast these three different roles and your responsibilities in them.
REES: I think there are some clear distinctions. The titles, in some ways, were a little misleading for the responsibilities. In general, the executive officer I would describe as like a vice chair or assistant chair.
ZIERLER: And that would be a chair of a department at a non-Caltech kind of place that has departments.
REES: Exactly. Not like a professorial chair, but an administrative chair, like the Division chair of CCE or BBE, for example. The Biochemistry graduate option started out as Biochemistry, but then the name changed to Biochemistry and Molecular Biophysics. As an executive officer, you basically have the responsibility to do what the chair of the division asks you to do, and that varies quite a bit between the divisions. In chemistry, Dave Tirrell, our current provost, was the chair of CCE when I was executive officer in chemistry. He was super organized, great to work under. And that's still true. The main responsibility was coming up with the teaching assignments and matters related to the educational program.
Once a year, I would work with the different subgroups in chemistry to come up with the teaching assignments. We would also meet quarterly to get feedback on the undergraduate chemistry courses. Courses have student ombudspersons who would meet with the executive officer and the option rep for the undergraduate major in chemistry to get feedback in the middle of the term about how courses were going, which we would then relay to the instructors. The executive officer of chemistry, at least as was defined by the chair of CCE, was really focused on the chemistry teaching program.
ZIERLER: And that's undergraduate and graduate for that role?
REES: That was for primarily for the undergraduate teaching program. The operation of the graduate programs, of course, is very important to the divisions, but there's a separate structure with the graduate dean and graduate studies committee that has oversight of those programs. Now, my time as executive officer for Biochemistry and Molecular Biophysics, in a sense, was sort of a hybrid of responsibilities since it involved coming up with teaching assignments for the biochemistry part of chemistry, but it also ended up having a fair amount to do working with the Biochemistry and Molecular Biophysics option rep, helping run and administer the graduate program.
That responsibility is actually somewhat unusual for an executive officer since its usually the graduate option representative who has that responsibility. But as we discussed previously, BMB, originally biochemistry, subsequently Biochemistry and Molecular Biophysics, was born out of an interdivisional effort between chemistry, CCE and biology. We generally had an option rep from one division and the executive officer from the other. The idea was that they would work together to provide administrative oversight for the operation of that graduate program.
ZIERLER: I'm always interested in when a name change, with regard to molecular biophysics, is simply an administrative distinction or if there's something substantive there in terms of who's thinking about or doing biophysics. What does the name change here suggest?
REES: It was always clear that this graduate option would focus on biochemistry and related areas. My PhD is in biophysics, structural biology, meaning macromolecular crystallography, X-ray crystallography, electron microscopy, which were always (to me, at least) seen as biophysics. When the Biochemistry graduate option started, naming it with a title containing Biophysics when this was being shepherded by the CCE and BBE divisions that didn't have physics in their title, I think there was some concern, as I understand, from the PMA division. I wasn't a part of these discussions, but apparently over the next few years, those concerns were eased. And so, we changed to Biochemistry and Molecular Biophysics, which then, I think, more accurately reflected the focus of the option. This includes structural biology which has traditionally been viewed as molecular biophysics.
Over the last 20 years, however, I think things have changed, and structural biology is now seen as a mainstream part of biochemistry. And biophysics now is more physics at the interface with biology. Of course, that's what the name suggests. But a lot of physicists have started working on biological problems, including single molecule work and the sort of things that Rob Phillips does, the quantitative modeling and so on. I think that's something where what we meant by biophysics 20 years ago has changed with time.
ZIERLER: And where is that change being driven by the theory, and where is it being driven by advances in instrumentation and what we can simply see?
REES: Again, this is just my own personal perspective, but I think part of it was driven by an appreciation in physics that there are a lot of interesting problems in biology that benefit from the mindset of the importance of quantitative modeling. The way I think Rob would describe this, quantitative data demands quantitative models. Essentially, if you have experimental data and a model to describe the data, you should be testing the model to within the accuracy of the data. And of course, there is a theory component because you're coming up with quantitative models. But I guess that's how I see this. These disciplines are all quite broad without well-defined boundaries, fortunately. But this is how I see the center of mass of biophysics today.
ZIERLER: Now, in the middle of all of this, in 2004, of course, you're named to the Roscoe Gilkey Dickinson Professorship. I'm curious, at that time, given your appreciation of the intellectual heritage and academic lineage that really takes it all the way back to the origins of these studies at Caltech, if you saw in that honor responsibility to think about Dickinson, to go back and read some of the sources, were there any remarks you prepared, did you talk to anybody? How did you respond to being named to this in real time?
REES: Like most things, the story is a bit more complicated than this. Ideally, chairs are endowed, so there's an endowment, and then there's a payout that offsets part of one's salary. Since I had the HHMI position, which pays 100% of my salary…
ZIERLER: But is that as an endowment? Or is that soft money?
REES: Hughes has a substantial endowment, but it's not tenured, so you have to keep getting renewed. [laugh] In that sense, it's not a lifetime guarantee.
ZIERLER: As opposed to an industrialist endowing a chair, and then that chair is endowed in perpetuity from the interest of the principal, essentially.
REES: Correct. That's how it's supposed to work in principle.
ZIERLER: But when you're named after an eminent professor, where the honor is in the name and not necessarily the family's financial resources, it's a different arrangement.
REES: Right. In fact, I knew John Hopfield had been the Roscoe Gilkey Dickinson Professor of Chemistry. That's where I first became aware of this chair. And we overlapped a bit. Hopfield left early in my time at Caltech. But because of my Hughes position, there wasn't actually a specific chair that was assigned when I was offered a chair because there wasn't this financial piece that was required. It was unusual in that case. Because of the role of Dickinson, both in my academic lineage, but also as the first PhD here, I asked to be named the Roscoe Gilkey Dickinson Professor of Chemistry. And since it is an un-endowed chair, that was OK. Because of the HHMI connection, there's a different process than what you would normally have in this. Of course, I was really honored to be considered to be a holder of a chair, and frankly, I really appreciated the opportunity to be able to say, "I'd like to be the Dickinson professor." That's the backstory to that situation.
ZIERLER: Does Dickinson have any descendants who are involved? Have you ever had any chance to interact with the family?
REES: Unfortunately, he died in 1945 at age 51. I understand there is family, but I have never had the opportunity to meet with them.
ZIERLER: Back to the science. With James Howard, in the Academy proceedings, you wrote an article, How Many Metals Does it Take to Fix N2? I'm always interested in the theory. What is the theoretical proposition that N2 needs to be fixed? What does fixing mean in this context? And how many metals did it take?
REES: Our atmosphere is 78% nitrogen. It's not a noble gas, which are, except for a few exceptions, truly inert chemically and un-reactive. But nitrogen gas is pretty inert, which is good because otherwise, if it was more reactive, like oxygen, this would lead to all sorts of other problems. The challenge is that nitrogen is an essential element for biological systems. It's part of proteins, amino acids, vitamins. Humans and most other organisms, however, are unable to directly access this abundant reservoir of atmospheric nitrogen to serve their metabolic needs. That process by which nitrogen gets converted into a more chemically usable form is called nitrogen fixation. There's a similar process of carbon dioxide fixation that plants use when they make biomaterial out of atmospheric CO2. There was some fixed nitrogen in the early atmosphere of earth that living systems could use to make different biomolecules, but as the biomass increased, all that fixed nitrogen was eventually sequestered in the biomass. So at that point, growing organisms were faced with the problem of how to get more nitrogen.
This problem was not unique to early life forms but was also faced by agriculture and the growing global population. In the early 1900s, the British scientist William Crookes called this "The Wheat Problem". There are a lot of parallels to analyses that were being done a decade or two ago about the oil supply. Crookes made an analysis of how much wheat needed to be produced to support the populations and realized that fertilizer was the limiting nutrient, specifically nitrogen in the fertilizers. A major source of fixed nitrogen at that time was from saltpeter, sodium nitrate deposits, in Chile. Crooks estimated how much fixed nitrogen was left and the rate at which fertilizer was needed and realized that society was running out of fertilizer. He concluded at the end of this analysis that civilized countries were in danger of starvation and that chemists had to come to the rescue of these societies. Of course, Crooks had a very, let's say, non-inclusive view of what he meant by civilized societies, which focused on Europe and the East Coast of the United States. I don't think he was so concerned about the West Coast, and certainly not Asia, Africa and South America. Crooks identified a number of different ways of fixing nitrogen.
As it happened, within ten years, the problem was solved by two German scientists, Fritz Haber and Carl Bosch. This is now called the Haber-Bosch process, where atmospheric nitrogen is directly reacted with hydrogen that comes from natural gas in the presence of an iron catalyst at high temperature and pressure. Under these conditions, the kinetic barrier to reactivity is low enough that you can get the direct reduction of nitrogen to make ammonia. The Haber-Bosch process is now one of the largest chemical industries in the world and led to the increased global population by billions over the last century due to the increased agriculture. Today, the industrial method makes as much fixed nitrogen as the biological method, but it requires significant energy resources.
Our interest has been trying to understand how the biological system fixes nitrogen to work out the catalytic strategy that allows nitrogenase to work at room temperature. Almost certainly, although this is still perhaps not conclusively established, the active site is formed by irons. The question is, which I think is still a good one, does the nitrogen bind to one iron, two irons, four irons, six irons? The purpose of my article with Jim How Many Metals Does it Take to Fix N2 was to highlight that we don't actually yet understand the detailed chemistry by which nitrogen is reduced by nitrogenase. Ultimately, we concluded there are 20 unique metals in nitrogenase, and it takes them all. At least practically, you need all these metals to fix nitrogen. It was really a way of framing the problem in terms of what we know and don't know.
ZIERLER: The following year, you were recognized for your teaching by ASCIT, which is…
REES: The Associated Students of the California Institute of Technology.
ZIERLER: What is ASCIT? And more broadly, for all of the ways you've been recognized, in what way was it special to be recognized not just for research but your commitment to students?
REES: ASCIT is an organization of Caltech undergraduates that does a number of things, including sponsoring teaching awards. This was for my teaching of biophysical chemistry, and I have to say, it was one of the most personally meaningful forms of recognition because I think we all view ourselves, fundamentally, as teachers who are preparing students for the future. Caltech has had an incredible tradition of inspired instructors. While I don't consider myself an inspired teacher, teaching is important, and I try to be competent. I found the ASCIT award to be personally meaningful because teaching is not something that I feel comes naturally to me, and that there's always more I can do to improve. The Center for Teaching, Learning, and Outreach has been an incredible resource that has really helped me, other faculty, teaching assistants, and so on.
ZIERLER: The following year, an award for your research, the Dorothy Crowfoot Hodgkin Award. Tell me about that.
REES: This was an award of the Protein Society. Hodgkin has a very special role in structural biology and, I think, in the history of science. She was an incredible crystallographer who won the Nobel Prize for her structure determination of biochemically important molecules, including vitamin B12, penicillin and cholesterol. In all these cases, crystallography was not confirming what chemists had already known, but rather was the decisive approach in working out the structure of important but chemically enigmatic chemical species coming from biological systems. Usually, in protein crystallography, we don't know the three-dimensional structure, but we do know the covalent connection between the amino acids from the sequence. Covalent bonding is not a surprise in these structures. But occasionally, we get to work on things that are chemically uncharacterized, and certainly, the metalloclusters of the nitrogenase MoFe protein were an example of where we were able to use structural techniques to understand the structures of otherwise uncharacterized chemical species.
ZIERLER: Around this time, you were doing work on ABC transporters. I wonder if you can explain what they are.
REES: ABC transporters are ATP Binding Cassette transporters that use this energy currency, ATP, to move molecules from one side of the membrane to the other. When we started working on membrane proteins in the mid-90s, we were really interested in looking at proteins that were members of families where there was no structural characterization. ABC transporters form a large family of transporters found in all forms of life. In humans, they have roles like the cystic fibrosis transmembrane conductance regulator that's mutated in cystic fibrosis. Several of the so-called drug resistance pumps, that tumor cells will over-express during chemotherapy to reduce the effectiveness of drugs, are also ABC transporters.
ABC transporters come in different flavors, and we started working on them with this funnel approach to get a structure. We were working in bacteria, since for technical reasons, it was easier at the time to make proteins in bacteria. Kaspar Locher was the postdoc in the group who came in and cracked the problem with Allen Lee by solving the structure of the ABC transporter responsible for importing vitamin B12 into E. coli. We were ultimately able to get several structures of ABC transporters, but the structure of the vitamin B12 transporter was really quite exciting to help us understand how the different parts of the ABC transporter fit together in the intact transporter structure.
ZIERLER: Why was E. coli interesting for you to work on?
REES: E. coli has been the workhorse of molecular biology. It's a common organism in our intestinal tract. Early on–I don't really recall the full history–it was used as one of the main tools for many developments in molecular biology. There was a famous quote from Monod (paraphrased), "What's true for E. coli is true for an elephant." The basic biochemical processes are generally found throughout many different lifeforms, and it's easier to study these in E. coli. It has a doubling time of about 20 minutes under optimal conditions, it is well characterized in terms of knowing the genome, the different proteins, transporters, and so on. Obviously, it's easier to do these sorts of studies in bacteria than in an elephant.
ZIERLER: I'm not sure if you've ever noticed this, but if you were to put your publications on a bar graph, 2013, 2014, 2015, fantastically productive times for you. I wonder, did you have a sense that you were going to become dean of graduate studies, and you wanted to get all of this in before administrative responsibilities pulled you away?
REES: Not in such a deterministic way. On this general subject, when we teach biochemistry or kinetics, we talk about the steady state, where the rate of formation of a compound equals the rate of breakdown, so that the concentrations remain constant over time. I think we sort of imagine that what happens one year will be like what happened the year before or will happen the next year. But reflecting the statistics of small numbers, there are quite significant departures from the average, so it's not like I always publish five papers a year, and there are definitely fluctuations. I do think around 2002 and around 2014 were major years for me in terms of publications. I think the underlying sociology, maybe even the psychology of this, is complicated.
Certainly, the 90s, up until the early 2000s was a special time in protein crystallography. I think we were well-positioned in that we had worked out the basic pipeline for solving structures. We then started putting in the pipeline for working on membrane proteins. During the 90s, we were able to really, I'd say, make a lot of progress on membrane proteins and nitrogenase. It was just an exciting time. This reflected not only where the field was, but also the group of students and postdocs that were attracted to work in this area. 2014 was another bump, and my group was large at the time. Again, we'd been doing a lot of work on both membrane proteins and nitrogen fixation. There are special times when everything comes together on a project, and you are able to solve problems and understand things.
And then, there is a period, which in my case does seem to be about ten years, where we're regrouping and gearing up for the next challenge. And it takes some time. Not everything we've worked on has been successful. But fortunately, over that period, enough projects seemed to come together. I think the large number of publications in certain years was a reflection of how projects came together after some period of time. It was not a reflection that I wanted to go into administration.
But the administrative side was a reflection of several factors, starting with, I guess, just getting older, among the factors one has no control over. My whole professional career has been spent in a university. These are just incredible organizations, but they don't work just by magic.
ZIERLER: You're saying that there's a real sense of service here.
REES: Again, just like with teaching, I don't think that I was an inspired administrator or would be an inspired administrator, but especially at Caltech, we think of ourselves as a faculty-run institute. We expect that our president and provost are real scientists, not someone who has gotten a PhD in the past and then been full-time administrators. The division chairs, everyone. These are major commitments. Personally, I never felt that I had what it would take to contribute to administration at the level of a division chair, much less provost or whatever, but I'd always enjoyed working with students. Serving as a dean was a five-year commitment, and I was thinking about my timeline.
When I turned 60, which was in 2012, I realized that this was the time that if I was going to do a five-year commitment, I couldn't wait ten years or even maybe five years. I was certainly open if the opportunity came to serve as dean. If I didn't get that opportunity, that would've also been fine. I didn't really care if it was undergraduate or graduate dean. I just thought we as faculty have a responsibility to keep the organization running. Universities are incredible, students are wonderful 99% of the time to work with. I was sort of prepared. And through my work with the BMB option, I got to know Joe Shepherd, who was the graduate dean at the time. In my case, my willingness to serve as dean reflected where I was in life more than where I was in terms of my vision for my research career.
ZIERLER: Before we get into what you accomplished as dean of graduate studies, I did want to ask about that same year in 2015, when you were awarded with the F. A. Cotton Medal. First of all, can you talk a little bit about Cotton and the dinner that was held in your honor? Because in your research world, this is a major, major recognition.
REES: Cotton was one of these larger-than-life inorganic chemists. Besides being a prolific researcher, he was famous for his inorganic chemistry book by Cotton and Wilkinson. That was one component, Cotton's stature. Another thing, a sort of personal thing, my graduate advisor, William Lipscomb, had a strained relationship with Cotton. I heard Lipscomb's side. I did get a chance to meet Cotton, but we didn't discuss the specifics directly. [laugh] And again, we had a fine conversation. But it was a little sort of unusual, given the history of my advisor and Cotton, to have this opportunity. Becky and I were able to go to College Station and visit for several days. We had a symposium with Marcetta Darensbourg from Texas A&M and a former nitrogenase postdoc from my group, Akif Tezcan now at UCSD. It was really a wonderful, wonderful time.
ZIERLER: Did you have a chance to prepare remarks? What did you talk about?
REES: It was all on nitrogen fixation. I should say Cotton, unfortunately, had passed away prior to that. I had met him on another occasion when I visited College Station.
ZIERLER: Back to the dean of graduate studies position, one of the running themes in these talks is the idiosyncratic way that Caltech organizes itself relative to other institutes of higher education. So dean is a position that would be recognizable at other universities. How do you understand that within the context of Caltech? What does dean of graduate studies mean here? And is it basically any different from a dean of graduate studies elsewhere?
REES: The way Caltech is organized, there's only one. [laugh] There's nothing quite like it. Being Dean of Graduate Studies at Caltech obviously has a lot in terms of responsibility for the administration and the oversight of the graduate program. That would be common to any graduate dean. But where the dean fits into the administrative structure and some of the broader responsibilities varies significantly between universities. I went to one national meeting of graduate deans, and I was listening to the presentations. And one dean started out and said, "Well, if you're the graduate dean, you're talking to the vice president for research every day." And I was thinking, "No, I'm not. We don't even have a vice president for research."
The oversight of the graduate program is common to any graduate dean. At Caltech, the Dean of Graduate Studies reports to the Vice President of Student Affairs, so we're outside the direct chain of academic command, which is the Provost and the six Division Chairs. There is a dotted line in the organization chart that goes from the Dean to the Provost. [laugh] It's sort of an odd arrangement. But since the Vice President for Student Affairs has typically (but not always) been a faculty member, which is certainly true with Joe, Kevin Gilmartin, and Anneila Sargent, the Dean is still reporting to a faculty member, which we think is important. I don't believe the arrangement we have at Caltech has any direct counterparts at other universities.
ZIERLER: Beyond chemistry and biology and the kinds of graduate students you would interact with in your world as a professor, at this broader view, what were some of the things that you learned about Caltech graduate students, why they came to Caltech, some of the struggles they had here, their interests, things like that generally about Caltech graduate students?
REES: By some standards, we have a very narrowly focused graduate program. There are 30 graduate options and they're all essentially STEM fields. We don't have a graduate program in French, or History, or a business school, these sorts of things. In that sense, the focus is largely on STEM, engineering, and research. And most faculty have an active research program, so the graduate program engages the whole faculty. From my experiences as a faculty member, I was familiar with chemistry graduate program, I was familiar with BMB, I was fairly familiar with biology. As a result, it was fascinating in the Dean's office to learn about the different cultures of the various options, how PMA, or EAS, or GPS, or HSS, etc. do things.
Even though all the graduate options are focused on graduate education with certain common elements, such as a candidacy process, a thesis and a thesis defense, there were a lot of differences in the way that the graduate programs were organized. Caltech is a very decentralized place. In a way, it's like the difference between the federal government and states' rights. If the Dean were to go in and tell an option, "I don't like this requirement you have. I'm going to put in a mandate, and from now on, you have to do candidacy by the end of the second year," this would be a nonstarter. It was important to appreciate the local flavors that have evolved over the last 100 years of our graduate program.
Working with the students and the Graduate Student Council, working through various situations, was both highly rewarding and challenging. My time as graduate dean was bookended by the Christian Ott situation and the pandemic, with a lot happening in between. I have to say, I was, in many ways, unprepared for the range of issues that came across my desk. But it was, I think, also part of the reason I did the job to take on new challenges where I had to work outside my comfort zone.
And, of course, I had quite an extensive support network to help me through these challenges. During my tenure as graduate dean, I greatly valued my weekly one-on-one meetings with Joe Shepherd – not only to discuss various issues relevant the graduate program, but also science and Caltech more generally. I really benefited from the support of the staff in the Graduate Office, Student Affairs offices, Student Wellness Services, the Registrar, the Caltech Center for Inclusion and Diversity, the Office of General Counsel, the International Student Office, and the Office of Research Integrity, as well as the Division and Institute staff more broadly. In many ways, I felt like I was taking an advanced tutorial in contemporary issues in higher education. And I'd get multiple problem sets a week. [laugh] At least the collaboration policy would let me use any resources. Some problem sets would be trivial. Others, I had to get them done in a week or so. Some, frankly, I started and never finished in the time I was in the office. I could read things in the newspaper about various issues, and they would be playing out with the students at Caltech. And so, it was really educational and highly rewarding.
I left the graduate office with several overarching impressions that I would summarize as follows:
The role of Student Affairs and other campus offices in supporting our students, both graduate and undergraduate, is largely below the radar of the Caltech faculty, yet their role is vital to our educational program. I was particularly struck by the importance of our graduate and undergraduate Associate Deans on the front line of student support. I was extremely fortunate to have worked with Felicia Hunt and Kate McAnulty and experienced first-hand their dedication and commitment to our graduate students.
Student mental health was a constant focus, and I tremendously valued the remarkable job that Jennifer Howes, the Student Wellness Services staff and the CARE team did both before and especially during the the pandemic.
We have a significant international student component, so trying to understand how some of the current political issues impact the experience of our international students, and getting to work with the International Student Office, was critical.
ZIERLER: On the specific question of international students and their importance in the graduate student body, when the Trump Administration made studying in the United States more onerous, more difficult through various immigration and other policies, how was Caltech affected? And how did you specifically respond to those political changes?
REES: Well, there was definitely a significant impact and anxiety arising from immigration and other policies. We have an incredible international student office that Ilana Smith runs. And they serve as the main point of communication with international students. I learned early on that there were a lot of facets related to the visas, Curricular Practical Training, CPT, that I'd never heard of before, and Ilana and others were very patient in helping me try to get up to speed on these complicated issues. I would generally make sure that I would talk to Ilana and our Office for General Counsel when I was working through a situation. This was highly rewarding. I learned a lot from their commitment, their understanding of these issues, and the importance of regular messaging, communication about issues. Even, I think, in hindsight, if we didn't know how to address something, it was important just to say, "We're working on it." The International Student Program was very good at having regular communication, and I think that was a main contributor to keeping our international students informed and reassured as much as possible. It wasn't always possible to provide definitive information because we just didn't know what the ramifications of some of these policies would be.
ZIERLER: I'll just take one example, since there are many to choose from, but if we're in a situation now where the best students from China want to come here to places like Caltech, Harvard, Stanford, or what have you, in this role as graduate dean, did you look more widely at how Caltech could remain competitive at the very top echelon, where we're not even talking about competition within the United States, but where maybe in a generation, graduate students in China might not feel compelled to pursue the best education in the United States because they could get it right in China?
REES: One of the responsibilities of the graduate office is to oversee the admissions process. In this process, each option has their own admissions committee making the actual decisions. A number of schools in the US reported declines in international student applications and enrollment. We did not see this.
ZIERLER: You mean during the Trump Administration?
REES: Correct. Now, a lot of this was driven by master's students at other universities, and we are primarily a PhD granting institution, although with a small, but important, master's program. We kept seeing increasing numbers of international applications. I think that the attractiveness of Caltech remains as a place for talented students from around the world. But I also think your point is absolutely correct. We cannot assume that the current attractiveness for international students will continue for a couple of reasons. One is that universities in Europe have, of course, always been strong, but I think are becoming more attractive for international students who previously would only have considered the US for graduate school. And, the research programs in China have also been getting very strong. So I don't think that we can assume that everyone will continue to recognize Caltech as a top place. We have to recognize that international students play a vital role in our research program, both at Caltech and across the US, as graduate students and postdocs, and the fact that this is under assault is of concern.
ZIERLER: In light of your receipt of the Fred Shair Award for Programming Diversity from Caltech's Center for Diversity, that's probably a great opportunity to put in historical context what was happening in 2020 with regard to the renaming issue. Obviously, all of these things have a history to them. In your role as Dean of Graduate Studies, because so many graduate students in particular were vocal in expressing their concerns, when did you first start to get a sense that people had concerns about the tallest building on campus being named after Millikan, for example?
REES: One of the great features of being in a university is the student population. Students see things from a different perspective than those of us who have been around for a while. There's been a recognition that Caltech has really struggled in terms of issues of diversity of our student, postdoc and faculty populations, and that this is a reflection, whether intentional or not, of a number of policies and the attitudes of people, past and present, here.
I did see that an important responsibility of the graduate office was, through admissions, to promote diversification of the graduate program by reaching out to communities that have been historically underrepresented in the graduate program. In my own career, I was never told, "You can't do this," or "You can't do that." I had many role models and so on, but unfortunately that is not a universal experience. It struck me at one point when I was on the faculty at UCLA that one of my biochemistry colleagues, John Jordan, was the first Black professor I had known in college, graduate school, or as a postdoc. At Caltech, the first Black faculty to be appointed and tenured was Steve Mayo with his primary appointment in Biology. Bil Clemons was the first Black faculty member appointed in CCE. It is troubling that these developments have only been in the past few decades of Caltech's history.
We really have a responsibility to try to make sure that we're bringing in the most talented students and postdocs for careers in science and engineering. I think one of the challenges is, there's a long learning curve to get up to speed in science. Again, I was fortunate from the beginning to be funneled into this pipeline. If one decides in college that you're interested in science, it's really difficult to master all the material. Not impossible, but really difficult. I believe that we have a responsibility to try to help support students coming from different backgrounds. After all, a research career will be, let's say, 40 or 50 years, but we make all these decisions on who's going to participate in this process based on performance in the first ten years. It's like if you're trying to judge who's going to win a marathon after the first five or so miles. There may be some correlation, but it's hardly exact.
During my time in the graduate office, I was impressed with the Diversity Center's role in supporting our campus community. I more fully appreciated that we have a responsibility to make sure that we're trying to find and support students that may not have come through in a straight pathway or had great instructors encouraging them to go into science and engineering. Growing out of these experiences, I was fortunate to get involved with our NSF supported Alliance for Graduate Education and the Professoriate (AGEP) with UC Berkeley and other universities focused on developing approaches to diversify the professoriate in the physical sciences. My role in the AGEP program also led to my participation in Presidential Postdoctoral Fellowship program initiated by President Rosenbaum to support a more diverse and inclusive postdoctoral community at Caltech.
ZIERLER: In promoting this atmosphere, when did students start communicating ideas that having these names on buildings of people who were associated with the eugenics movements was antithetical to that sentiment of creating a more diverse place? How did those comments manifest, and when did you start to hear them?
REES: I recall reading an article some time ago in Engineering and Science (I believe), which was a Caltech publication, about Millikan's eugenic views. I think it was sort of bubbling there below the surface. After the George Floyd murder, there was a recognition that there have been a number of policies and views over the years that have been antithetical to what we strive for at Caltech. The Black Scientists and Engineers at Caltech, BSEC, took the lead in enumerating a number of policies and aspects of life at Caltech that were harmful. I think that was the event that really propelled this issue to the surface and forced the university to take this issue seriously.
ZIERLER: Long before the concern over people like Millikan and their support for eugenics manifested in the context of increasing a more diverse and inclusive student body and campus environment in general, what about the problem with it because eugenics is such lousy science?
REES: I think one reason why these sorts of experiences are so important for scientists to understand is that science is used to justify offensive, incorrect, harmful policies. Certainly, scientific racism had a group of adherents in the US and elsewhere. Science is perverted in different ways. And we see this now outside of eugenics with climate, pandemic policies, and so on. This is also something where scientists don't like to get involved in things, but I think we have a responsibility. Now, we like to think of scientists as being objective seekers of truth, rational, dispassionate, all this stuff. It may be more true than the population as a whole, but the fact is that scientists are people. We all have ideas and prejudices that we bring with us as scientists, and the consequences can be catastrophic when the underlying motivations are distorted.
ZIERLER: Now, during your tenure as dean, you did keep up a robust publication rate, although obviously, your bandwidth for doing the research was diminished because of your administrative responsibilities. I wonder if, to some extent, your tenure as dean gave you a chance to sort of go through a winnowing process of the research that was most precious to you at that time. That you had less time to work on it, and in light of that, "Here's what I'm going to focus on." Or to what extent were your collaborators and students really driving those topics while your concerns were in the administrative realm to some extent?
REES: I think the way I've been viewing this, the main driver right now has been time, chronology, age. [laugh] Thinking about what's the appropriate time and role for a faculty member? Being graduate dean was supposed to be a half-time position, and I found it to be more than a full-time position. Partly, that was due to my inability to delegate well and so on. But it also reflected what I liked about the position. I really enjoyed the personal interactions and working with students. And there are times when you just can't say, "Oh, yes, I know you're having a crisis, but my schedule is full for the next two weeks." [laugh] You just have to drop everything. It definitely impacted my time that was available for research.
I wasn't planning on having a large, active research group when I left the graduate office. So I quit taking students while I was in the graduate office. I have appointed several postdocs, but basically, I was letting my research wind down. I have pretty much ended my independent work on membrane proteins, and I'm focusing on nitrogenase now. We don't have a retirement age, and I don't think there's a set time in everyone's life where it's appropriate to step down. But, I think it is important that this happens. Universities really depend on having an influx of new, young faculty that have been thinking about things in new and different ways, about what's important. And, retiring is not incompatible with doing research. I think of Seymour Benzer, who retired but was active in the lab until the very end, coming up with fresh and original ideas. In many ways, being emeritus seems like the best type of faculty position as you can focus more on the things that you think are important.
Caltech has, essentially, a fixed faculty size, and so the ability to appoint people depends on people leaving the professoriate. If the total number's a constant, then the rate going in equals the rate going out. [laugh] I personally feel that we have a responsibility to move on, which doesn't mean that we can't be actively involved in research and the life of the university. But I think the vibrancy of the university really depends on having new, young faculty. The other part is, again, related to diversity. The faculty's not very diverse. The lifetime of a faculty member at a university is probably 30 or 40 years. If the half-life of a faculty member is 30 years, it really impacts the ability to change the composition, whether it's in terms of gender, ethnicity, race, or research interests. And I think that's a real problem.
For these reasons, I've been feeling that I need to ramp down my research group. Now, one way in which I'm trying to accomplish this, while also achieving these other aims that I've been trying to describe and having no theory to guide me in this transition, is that I've merged my research group with Bil Clemons' group. Bil has a really active group and exciting research program in membrane proteins and other areas. While I'm no longer serving as advisor for new students, I'm open to co-advising students with Bil. I've been struck by not only how much work it is to run a university, but how much work it is to run a research group. It is a thrill to do research at the Caltech level, but it is also really intense and requires a lot of time by the faculty – not just for the science but for everything else that is required to do run a university. And I do think that by tapering down some of my own research, I have more time to try to help Bil and others with the operations of their group, in ways where experience is still valuable. I've been fortunate that my postdoctoral advisor, Jim Howard, has spent a significant amount of time in Pasadena since he "retired" helping me with research and life. There are a lot of issues wrapped up in these considerations, but definitely, the time axis is the main driver in doing these transitions.
Thinking about the time axis also reinforces the idea that time marches on and we don't have an infinite amount of time. The relentlessness and intensity of research is great, but it also requires a lot of time. Becky and I have been married since before I started graduate school. Science has opened up incredible opportunities for us – including moving to Southern California and getting to work at UCLA and Caltech – but it also comes at a price in terms of the time that is required. So, while Becky and I can, there are a number of things we would like to do and experience together, such as visit with our children John and Abby who live in Paso Robles, travel, and work on improving our musical skills.
ZIERLER: I can't say on a happier note, but for 2020, when COVID hit, from your vantage point in terms of thinking about the graduate students, when you were dean and were involved in those high-level discussions, what were some of the most urgent things that Caltech needed to consider at the time they said, "We need to shut down and go remote"?
REES: I guess there was one immediate consequence and one longer term. Immediately, there was an incredible amount of anxiety over the infection process, whether you could work together and if it was safe to be in the lab with someone else. "Is it safe to be in the Catalinas with a roommate?" The undergraduates had to leave. Graduate students largely stayed on-site in the Catalinas, which is more like apartment living as opposed to…
ZIERLER: A dormitory.
REES: Yes. People were really, really anxious. What if your roommate goes out to meet with people and comes back? And of course, in the labs, everything completely shut down, with the exception of a very few groups. We were trying to understand the parameters and to answer questions. And this didn't involve just the graduate program but everyone. The administration really was under the gun to come up with guidelines that would help reassure to some extent that we understood and were trying to control the risks.
There was the three-month period when people were out of the labs, and then we came back in June of 2020 with reduced density, trying to structure people's schedules so that they could continue to work but not feel like they were putting their lives in danger by being in the lab. That was one part. The other part that was coming up was that we had a number of incoming international graduate students, and it wasn't clear if they would be able to come into the country or not. So, there was the whole immigration piece, and at the same time, during that summer of 2020, there was a big spike in COVID. We had to ask if it made sense to be bringing people to Pasadena if there was a big spike in COVID.
We worked through a number of options, and ultimately, we decided we would allow the students to matriculate remotely, so they could be either in Pasadena, the US, or out of the country. There were a number of ramifications about registration status and so on, but we worked through these issues with the Registrar, International Students office, Office of General Counsel, even the Credit Union to work through how to pay students who were out of the country. And at the same time, students wanted information. "What are we doing?" I remember a four-page letter we sent to the students that tried to outline what we were working on, what we knew, and what we didn't know. And I honestly thought, "I should put this letter on my CV," because I think I spent more time on that letter to the community than many of my research papers. But of course, these were all just works in process, and things would change. But it was really just trying to understand what was going on, what we could and couldn't do, and just all this anxiety and understandable concern.
ZIERLER: I'm curious if Caltech, institutionally, at this moment of anxiety and confusion, relied internally on its own resources and not just listening to what the FDA, CDC, and Dr. Fauci had to say. Because right here on campus, we have world-class immunologists, we have world-class fluid dynamicists, we have all of the brainpower we need in-house to make these decisions. I wonder if you can comment on that at all.
REES: I was sort of at the periphery of these, but absolutely, there's a lot of insight and firepower. But what I've learned is that intelligence and expertise are important, but there's also this whole broader picture, and it takes some time to get up to speed. I think there's a certain limitation in terms of what you can really do in a short period of time. Certainly, there were many internal discussions, but that ultimately, we have to respond to the LA County and Pasadena public health agencies. Some of these things, we can't just come up with our own policies.
ZIERLER: When it was time to start thinking about stepping down as graduate dean, what were the things that you really were proud that you accomplished, that you wanted to do and actually did? What were the things that were in train that you were going to hand off to your successor because these are issues that go beyond any one term? And what are those challenges that you hoped to accomplish but for whatever reason, weren't able to get to?
REES: Those are great questions. I'm probably not the best person to answer them in any objective way. My view on administration is that everyone has strengths and weaknesses in terms of their ability to do things. The organization benefits by having turnover in these positions. For me, I was done by the end of my five-year term, and not just because of the pandemic. The part of the job that I really found rewarding was the one-on-one interactions with students. It really takes advantage of the strength of Caltech, which is the small size. We collectively would spend a lot of time working to help support students so that they could be successful in the program. I would say we had some success. It wasn't 100%, but I know that there is nothing more satisfying than seeing someone get a degree at graduation and thinking - not that we did it for them - but we helped them. They were struggling, and we were able to help them get over a rough time, and they end up with their degree. It's hard to beat that sort of feeling.
I did not have a strong vision of what the graduate programs needed to be doing. This is a bottom-up place. The options have their strong views from experience on how things should be organized. There aren't many sort of high-level things that you can say, "Well, everyone at a graduate program at Caltech needs to do X, Y, and Z." So I didn't have any strong visions for changing academic policies. I saw our job as helping students be successful and getting them through. One area where I think we made some progress but definitely had a lot more to do was diversifying the student population and helping support students with different backgrounds be successful, making this a welcoming place. That's one where I feel like we made some progress, but it wasn't like a switch was flipped and everything was great going forward.
ZIERLER: Once you had an opportunity to take a deep breath after your tenure as dean of graduate studies, was there anything that you wanted to do in the lab that just sort of exerted a magnetic pull on you that you weren't able to do, but then you could get back in there and get to it?
REES: In an odd way, it took me about a year, I'd say, to decompress from this experience.
ZIERLER: Plus, there's probably also a muscle memory that you have to regain as well.
REES: Yes. I was and still am behind on papers that students and postdocs have written. I'm really just trying to get back on top of the research projects of my students and postdocs. I feel like I've been playing catch-up.
One of the things I always enjoyed doing is deriving equations and fitting numbers to those equations. When I started out, the students that were attracted to my research, I think, tended to be wired in this particular way because there was a lot of computation and math, or there could be. Now, over time, as the technology became more mature, crystallography could be used by many more people, and the problems we were solving were much more focused on the biochemistry of the system. Over the years, I'd work out different problems that we encountered in our research. Some were just trying to understand derivations that were well-known but I had never worked out, and others were trying to really understand different sorts of effects. These notes were in a lot of different places – in file folders, in notebooks, in papers and in computer files. Over the last few months, I've really worked hard at trying to put these notes together in a cohesive and comprehensive way. I do think this process has allowed me get back to more just thinking, as well as highlighting certain areas where I need to learn more, especially related to electron microscopy. Not that these are necessarily deep problems at all, but ones where we're able to take some sort of basic framework in terms of thinking about how experiments are done and the type of data we collect, and to see if there are new types of information we can extract.
ZIERLER: At the very beginning of our interview sessions, we talked about what you're doing right now circa October 2021, and so we've reached a point of completion in terms of the historical narrative. What I'd like to do for the last part of our talk today and for the last part of our interview sessions is to ask some broadly retrospective questions about your work and career. While we're on the theme of graduate students, for yours in particular, I certainly don't want to burden you with naming the ones that you're most proud of. There are so many, of course. But I wonder if you can sort of reflect broadly, of all of the students who have gone on to such great success, not just in the United States but around the world, and not just in academia, but in industry as well, what are some of the commonalities that you could look back on when you see their work in the laboratory, the way they interact with their fellow students and professors? What are some of those commonalities where you might have said, X number of years ago, "So-And-So is going on to great things?"
REES: I think when I started out, I was primarily seeing things in terms of research accomplishments. Reflecting on my experiences, I feel I was very fortunate to have found a line of work that I have really found personally satisfying. Over time, I've realized that my main role is to help students and postdocs find something that they're passionate about. We have very bright, talented students and postdocs at Caltech. Some of them come in and know exactly what they're going to do, and from the very beginning, you know that they will be doing wonderful things. I think in that case, my main responsibility is to stay out of the way and not screw things up. I guess a related issue is that everyone, not just administrators, has their own strengths and areas of less strength. The way in which a student or postdoc will approach a problem will likely be different than the way I would approach it. When the students start out, they're trying to find their scientific voice, how they like to approach a problem and work on it.
A very rewarding experience has been working with students or postdocs who have come in, and they think they're interested in something, but it's not really clicking. But then in the course of trying different projects, they find what they're passionate about, and you can just see that growth in the course of their time in the group and beyond, whether in industry or academia. And I feel like in these cases, we were able to influence the trajectory, which is a wonderful feeling. Perhaps more importantly, we didn't get in the way. And there's a lot of satisfaction when people come through your group, then go off and do wonderful things.
There's also a group of people who I don't think have found their passion. And life's complicated. There are many choices. I think this is part of that process of self-discovery, that in the course of learning what you don't like, it helps you understand and find maybe what you do like. And I consider that finding what you don't want to do to be an important contribution as well. As a result of these experiences, I've switched from thinking that our most important outcome is our research and research papers, to thinking that our real products are our students and postdocs. They're the ones that will go out and change the world or do things that impact people's lives. And the research projects, in an odd way, are really just sort of the vehicle we use to train students, to help them discover what they like to do, and prepare them for their careers. The research, of course, is important to me. I really am fascinated with all these projects. But in some way, it's almost a secondary thing. It's important that people discover what they're interested in.
ZIERLER: On the undergraduate side, you've been at Caltech long enough where you have some real historical perspective on some basic questions about the things that undergraduate students are interested in, the tools that they use to succeed, their motivations in undergraduate for what they want to go on to accomplish afterwards. What have been some of the big themes among undergraduates, both in terms of what's remained constant and what has changed over the years?
REES: Well, Caltech is, again, a unique place with its small size and focus on STEM. I think that this place has really been attractive to students who have an interest in physics and math, and now in computers. I think maybe generally less of an interest in chemistry and biology, based on my experiences having taught Chem 1b and 1c. I think that's changing a bit, although computer science is so dominating to undergraduate life that that may be swamping out biology and chemistry majors.
When I first got here, I found there was little interaction between the undergraduates and the rest of campus, at least that I could see. One of the things I found really surprising was the lack of interest in undergraduates doing research, at least in chemistry, during the school year. Partly, it's because of the intense course load, and I think there was a period where overloading and taking more courses was a sign of, in some ranking scheme, a quantity that people were trying to maximize. Over time, there's been a lot more engagement and interaction in both directions, between the campus community with undergraduates and vice versa. Undergraduate life is not just about how many courses you're taking, but it's an important opportunity to be exploring other things, whether it's research in the lab, athletics, or the activities of the Caltech Y, getting involved in student government, and so on. It's still a very intense place, but I think we now see a somewhat more balanced distribution of time and activities.
ZIERLER: In 2020, when the Royal Swedish Academy recognized you with the Aminoff Prize, because this came at a time when, as you said, you were in a period of winding down your research agenda, to what extent did you see the Aminoff Prize as sort of an overall achievement in your contributions in structural biology and chemistry?
REES: Well, it was really exciting. [laugh] It's a reflection of our work on nitrogen fixation, which I'm quite proud of. If I look historically at the people who have been recognized by this award, there are a number of my scientific heroes. It's pretty cool to be in that group. Ironically, the symposium and presentation were supposed to be held in March of 2020. It's been rescheduled multiple times, and now been moved now to the spring of 2022. I'm hoping that Becky and I will be able to go and experience this in person. But yes, it was nice. You don't do this for the recognition, but I wouldn't be completely truthful if I didn't say, "Wow, that was sort of cool."
ZIERLER: In our first talk, I asked you sort of the broad overarching questions about your field. Now, what I'd like to do, in light of you having discussed some of your contributions, to use your powers of extrapolation to think about these contributions to see where the field might be headed next for the next generation of scholars out there. So let's start, first, with nitrogenase. Over the course of your career, and in light of your contributions, where are things headed for nitrogenase research?
REES: It's just astonishing to me how much we have accomplished with nitrogenase, if I could have a moment of immodesty. But at the same time, the fundamental question of how it works, I think, remains incompletely answered. In this case, the real outstanding question is the chemical mechanism, the detailed chemistry by which the NN triple bond is broken and ammonia is formed. There are still contributions that structural biology will make, especially with electron microscopy, which we are focused on going forward. It's important to have relevant chemical models like my colleagues Jonas Peters and Theo Agapie are making, to guide our chemical intuition about catalytic strategies. Ultimately, I believe the catalytic mechanism will be defined in detail by computational chemistry. My colleague Bill Goddard has been helping us with calculations on this system, so we can start getting involved in that part of this process.
ZIERLER: Because your research on mechanosensitive channels emphasizes the commonality of how these forces work in all kinds of life forms, to the extent that the future of this field is going to emphasize this convergence, to what extent do you think that will lead us to deeper existential questions about the commonality of life and how it forms and evolves?
REES: It's sort of like, in a sense, the statement, "What's true for E. coli is true for an elephant." Yes and no. [laugh] I think that the great thing about bacteria, relative to higher forms, even including yeast, is that they're more tractable in terms of biochemistry. How does a force-sensitive channel respond to tension applied to a membrane? That's just a basic question that can be addressed with bacterial channels like MscL that is a small-sized channel with a large conductance. But MscL is only found in microorganisms, and so the specifics don't apply directly to mammalian systems. The human mechanosensitive channels are quite different, such as the Piezo channel, discovered by Ardem Patapoutian (a former graduate student here with Barbara Wold) who was recognized for the Nobel Prize this year; Piezo is a massive channel with a small conductance.
The detailed tension-sensing mechanisms are very different between MscL and Piezo. But in a way, I like to think about the relationship between these channels in terms of the hydrogen atom and more complex atoms. You can more easily solve Schroedinger's equation for the hydrogen atom than for other atoms. So, while the hydrogen atom is useful to hone your intuition, that doesn't mean that every other atom is exactly like the hydrogen atom.
ZIERLER: In structural biology, to the extent that advances in instrumentation will continue to enhance our understanding of molecular structure, do you think that the instrumentation and our ability to understand structure will get to a point where structural biology will no longer be a distinct field within biology?
REES: My gut feeling is that structural biology will keep reinventing itself in ways so that it will always be a part of biology and chemistry. If I draw a picture of a tree, in some sense, that's structural biology, so structural biology isn't just restricted to molecular structures. The ways in which certain types of structural biology have been used–let's say X-ray crystallography-will certainly change. There will always be a demand for X-ray crystallography, but it may not be something where universities will widely hire faculty members who do crystallography as their primary experimental approach. Electron microscopy is really the growth field at the moment. The state of electron microscopy today reminds me of X-ray crystallography in the late 1980s. There are plenty of opportunities for advances in sample preparation, detector development, and the way that experiments are done, so that electron microscopy will continue to be a frontier area for structural biology for the next few decades. Experiments that seemed impossible when I started are now so routine that you couldn't even imagine getting a thesis for something like it. We are heading towards the time when with electron microscopy, we can get high-resolution structural information on components of cells in their natural environment, without any sort of purification.
I've spent my entire career pursuing a reductionist approach, purifying proteins, crystallizing them, getting structural information, and figuring out how they work. The way you'd really like to be able to do this is to characterize their structure and function in the cell, where they're in the native environment, they're with all their partners, and so on. No more purifications. I don't think that's just science fiction but we'll be seeing an era of structural or mechanistic cell biology when these sorts of studies are conducted in the native environment. Structural biology will be a big piece of this effort.
Light microscopy is also a part of these developments to provide spatial information on molecules in living cells, for subsequent high-resolution structural characterization by cryo-EM. Cryo-EM can't be done on living cells because of the cryo part and the radiation damage. I see these fields morphing together into a broader continuum. When I started as a graduate student, I would never have imagined that crystallography and light microscopy, or even electron microscopy and light microscopy, would be something that you would envision as part of one project. But that's definitely where things are headed.
ZIERLER: I asked you in the professional chronology about the allure of protein structure in the sense that there was such fundamental work to be done at the beginning of your career. To the extent that you would give advice to students who were looking at that timeframe in the next 30, 40, even 50 years of their career, does that hold today with regard to protein structure? Is there fundamental work that can be done that will keep the field busy for the next half century?
REES: In my view, a structural foundation is always valuable. I was really struck by the realization I had as a graduate student that the same type of equipment that allowed my advisor, William Lipscomb, to do structures of boranes and other small molecule compounds also allowed him to determine the structure of one of the first proteins. There are not many experimental techniques with such versatility that let you study such a diverse set of systems. If you want to understand how things work, you have to know the structures – whether a molecule, cell, organism, society, the universe. Now, when I started out, there weren't many protein structures. So each structure, it was like, "Oh, wow. This is unbelievable. This is so cool." The same was true also when I was starting out that there weren't any genome sequences. It is hard to believe that determining a ~20 base sequence of the lac operon was a huge accomplishment at one time.
Crystallography (and sequencing) has changed dramatically over the past 4 plus decades and in many ways has become much more routine, but it still is an essential technique. We will always need people like Jens Kaiser of the Molecular Observatory, who understand how crystallography works and can train the next generation of structural biologists. My students are doing so many different things in the course of their research that they cannot master in detail all of the different things they're doing. It's just impossible. So we still need people who understand crystallography and how crystallographic facilities operate, whether in-house or at synchrotrons.
ZIERLER: A question to bring it all together for your career at Caltech, which also has been a theme of our conversations and is a hallmark of Caltech, not just for the specific interactions that lead to research collaborations and publications, but just the sense at Caltech that you can talk to anybody about anything that's interesting, what has been the value of that research culture at Caltech, the sense that multidisciplinary means it's all interesting, it's all relevant? What has been the value of that for the most important questions that you have pursued in your career?
REES: I think the small size, and the informal culture of interactions means that we're able to do things at Caltech where at other places, we might say, "Well, there are already five people working in this area. I can't do that because I'll step on toes." I think my decision to move into new areas was facilitated by being able to talk to colleagues who were experts in cloning, or DNA structure, and so on. These sorts of interactions are particularly valuable when you're thinking about future directions or "the next big thing". And, the value is not just the disciplinary knowledge, but also the generational perspective from interacting with younger colleagues. I interact extensively with Bil Clemons, and with the two decades' difference in our careers, he sees things differently. One example is program AlphaFold that uses machine learning algorithms to predict protein structures. My first reaction was, "Oh, yeah, who knows if this really works?" whereas he's like, "Oh, yeah, this is this great tool, and we need to build it into our research." Another great example is Rebecca Voorhees' work at the interface of cell biology and electron microscopy. We all have our own views and perspectives.
I feel like I left graduate school with a great toolbox for doing crystallography. I've kept using this toolbox throughout my career, although when I started out, I was working on the equivalent of Model-Ts, then maybe progressed over the next few decades to sports cars. But I can also use this toolbox for other purposes, to continue the car analogy, to study electric cars, at least in some ways, not the internal combustion part. For people coming along later, they will start out from the beginning working on electric cars so that provides their reference. One of the great aspects of being at Caltech is that people here are involved in so many different things, and not just on the science and teaching part, but also in terms of their interactions more broadly in society. It's really been illuminating. I think this helps keep you alive and excited about the prospects for the future.
ZIERLER: On that point, last question, finally. Looking to the future, to the extent that now, you're at a stage in your life and research where you're thinking about winding down, and eventually you'll go emeritus, of course–we can emphasize that emeritus does not mean retired, it does not mean detached, it does not mean separating yourself from the field. When that time comes, and you no longer have the burdens of the administrative responsibility, the burdens of funding, the burdens of worrying about the graduate students, and you could just peel all of that away and go right back to the things that were most interesting and curious to you as a graduate student, now, a year, five years into the future, what are those things you want to pursue just because they're so fundamentally fascinating to you?
REES: I think in some ways, emeritus seems like the best position because you can be active, you can keep a research group, and you can pick and choose how you want to be involved. Of course, there are some things that you are cut off from - you're no longer involved in staffing decisions and other administrative functions, but you can continue to have a research group.
I've been grappling with what we should be studying going forward – but this is not a new consideration but rather an on-going process throughout one's career. It is easy to get caught up in the day-to-day challenges and lose track of the broader objectives. It also takes time to acquire the resources one needs to change research directions. As part of this process, Bil and I have been holding regular "brainstorming" sessions with our groups as a way to think more broadly about the present and future research goals. This has provided a great opportunity to reflect on priorities and interests.
Personally, I like deriving equations and fitting these equations to numbers. But ironically, the way my research group has evolved, that has not been a very useful skill of late. It was useful in the past, but as our projects have progressed, a lot of what we're trying to do now is trap certain biochemical systems in a given state. This is typically messy and challenging biochemistry, although incredibly important for our research goals. I like crystallography, and I would like to go back to the basics and delve more deeply into certain aspects of crystallography. I certainly want to better understand what happens in an electron scattering experiment and what sort of non-standard ways we might be able to do experiments to teach us new things. And, last but not least, as I mentioned earlier, I would like to devise some problem I can solve with conformal mapping.
Science is incredibly exciting. And it's an unbelievable privilege to be able to do this at a high level at a place like Caltech. But it's also relentless, and there's a lot of responsibility running a group. If you have students, it's important to keep up with the literature and all these things. Caltech needs people who are all in, who are committed to making sure their research programs are operating at a high level, doing all the things required in terms of administration and so on. There are parts of those things I'd be happy to help out with as emeritus, but I need a break from the relentless side and the responsibilities of the past 40 years. It's sort of like being an administrator. If you haven't gotten something done in ten years, you're unlikely to get it done in 15. It's not a perfect analogy, but the things I've been working on, I've had 40 years. And while there are still things I'd like to do, but I also feel like I've taken my best shot. And it's important for the research enterprise in the universities themselves that you keep having fresh, eager, excited people who are all in.
ZIERLER: It sounds like what you're saying for the next stage, there's still, for you, a lot to do, and you'll have fun doing it.
REES: Absolutely. Universities are just incredible. I think the contribution of universities to society, and the world, and knowledge–they're unique. But they're complicated, they take a lot of time, and maybe the thing I like, which is also frustrating, is that we're often just working on individual solutions. You're stuck on a problem, and you've got to solve that problem, or support a student, or whatever. Those are the sorts of things I like. I feel like there are still plenty of things to do, universities are still special places, and I look forward to keeping my connection and being engaged.
ZIERLER: Doug, this has been an epic journey we've been on. I'd like to thank you so much for spending all this time with me and for sharing your insight and perspective. And as you said in an earlier comment, scientists are people with all of the struggles, all of the mistakes sometimes, all of the learning, all of the joys, and this has just been a remarkable illustration of all of those things. I'd like to thank you so much.
REES: Well, thank you. It's been a real pleasure, and this has gotten me to think about things at a higher level than I normally would about a lot of these issues. So thank you.
ZIERLER: Well, that's great. Mission accomplished.