"I'm particularly amazed by the electron microscopy work,
it is really an incredible advance"

Nicola G.A. Abrescia: When did you start with your virus interest?

Roger Burnett: Actually, when I was a graduate student in Michael Rossmann's lab.
I had a bachelor's degree in physics and I thought that a future in physics was not very wise because while I could sort of work my way through mathematics it was not something that came immediately easily to me and also I hated electronics and I thought that was not a very good combination for being a physicist. And I had thought about going back and doing a medical degree, becoming a doctor, because I had an interest in biology. But in England in those days, it was very difficult to change direction and I did not have a biology qualification of any kind.
Then, just by chance, a letter arrived from Michael Rossmann who had just moved from the Medical Research Council [Laboratory in Cambridge,] England, to set up a laboratory at Purdue University in the Midwest of America. This was in 1964. And this was a time when most Europeans had never been to the United States. We only knew the United States from movies. So, as a young man I thought it was very exciting because it offered the possibility to go to a biology department in America. I knew that a graduate student in America had the opportunity to take courses and this would give me an education in biology as well as doing research in crystallography.

I have to confess I did not really have any particular interest in crystallography. I knew nothing about crystallography. In fact this seems amazing now, but I had never even seen a computer. This was in 1964 and computers were not that common; and certainly not at the undergraduate level. And so I actually sailed by ship to America on the original Queen Elizabeth and then took a Greyhound bus from New York to West Lafayette, Indiana, and for the first six months I was in America I thought I was in a movie myself because it seemed so strange and exotic. Michael Rossmann met me from the Greyhound bus and took me immediately to the darkroom to show me his latest diffraction photographs and it turned out that he was very interested in using viruses as... well, he was not really interested in viruses I think, as such, but he was very interested in developing methods. There is a method called the molecular replacement method and he was interested in viruses because there are many copies of the same protein in the same structure and so he thought that was a good model or test bed for testing his methods where you could use molecular replacement. So, he had a small [virus] programme. In fact, I was the first person who he prevailed on to try to crystallise viruses and in fact I got small crystals of a virus. I think it was, I don't know, cucumber mosaic virus or something like this. But in some ways I was fortunate that the crystals didn't grow very large because otherwise I probably would still be doing my PhD in West Lafayette, so I did something else. This was sort of an introduction to the idea that you could study viruses by actually doing crystallography.
But in those days it was really almost impossible to think about getting a virus structure. Steve Harrison I think was really the only person who was seriously pursuing virus structures in those days. And then I did a post-doc working on a flavoprotein, a small protein molecule; which is actually amazingly enough… I think it's only number 8 or 9 of the structures that are solved in the database. And then, at that time it was 1973 when I was looking for a faculty position, and science in America and maybe elsewhere goes in very cyclical periods of funding and this was a down period. And so it was impossible to set up a laboratory in America because people viewed crystallography as exotic and expensive, which it was, and needing lots of instrumentation, lots of computing power; all expensive things. So I asked Steve Harrison, who I knew, if he knew of any openings in Europe. And he said, 'Actually there is a person at the Biozentrum in Basel who has crystallised the coat protein of adenovirus and I think that they are looking for a crystallographer'. And so I wrote to [Richard Franklin] at the Biozentrum asking if there was any possibility of a position there and I got a letter back offering me a job. So, I moved to Switzerland. This was a very interesting time, actually. In fact, when I am visiting here I've been making comparisons with the Biozentrum in those days, because the Biozentrum was a relatively new institute; it was open for maybe four years before I came. It was kind of similar to here. The university system in Switzerland, in Basel, had set up its institute as an independent institute that was aimed at bringing together investigators from all over the world and setting up a centre of excellence. It was sort of outside the usual bureaucratic demands of the university structure or the civil service or whatever. It was a very exciting place to be and it was very well-funded, lots of instrumentation, many similarities to here. It was a very long answer to your question.

Nicola G.A. Abrescia: No, no. Actually you answered my second question. You were born in the UK, and then you moved to the United States. Why didn't you come back to the UK? I mean, UK was like the cradle of crystallography at that time.

Roger Burnett: Whenever I moved, until recently, almost until the end of my career, I would always look for positions in England but there was never anything that was interesting or available or whatever. And I think, also, when you grow up in a place you have certain antibodies against that place. And so while it's comfortable for being at home, also it's actually very comfortable to live outside your own country, because you don't have these kind of controls.

Nicola G.A. Abrescia: Yes, this is interesting because nowadays science in Europe is mainly driven by the UK. I mean, the big competitor of the United States is the UK, maybe Germany. So, everybody sees the UK as the place to go or the United States. So I found that a little bit...

Roger Burnett: Certainly when I was at the Biozentrum the resources we had were far superior to what you could get in the United States, and certainly as a young person; it was the sort of European system where you are given a graduate student and a post-doc and, you know, it was very good. It was a very good place to start a project.

Nicola G.A. Abrescia: So, you started to work on adenovirus in Switzerland and you have been working almost all your career on adenovirus.

Roger Burnett: That's correct.

Nicola G.A. Abrescia: What's so fascinating about it?

Roger Burnett: To start with, people knew nothing about viruses and so it was the idea that if you looked at the coat protein of a virus, and adenovirus was unusual in that when the virus infects the cell it produces a large excess of structural proteins; and so one could harvest those proteins and crystallise them and determine the structure. The idea of solving the structure of the entire virus was absolutely out of the question in those days because cryo-electron microscopy had not been developed. For crystallography, it was far too large and also it had the complicating factor that it had long fibres, which probably would make it difficult to crystallise. And so, this was an idea to use a kind of a piecemeal approach to solve the structure of the individual coat protein and then try to bootstrap ones way up to looking at a larger assembly. And we thought it might be a model for two things: one, to give us some insight into how the virus was built and how you got self assembly and so forth; and the other was that it might be a way, a general way, of attacking large macromolecular complexes. In other words, if you could solve the structure of one piece, then you could maybe use that to bootstrap your way up to develop a model of the whole thing. And that has actually worked out.

Nicola G.A. Abrescia: This is my other question. Now it seems that the x-ray structure of the adenovirus has been solved and you told me that there is an electron microscopy structure at 3.6 Å resolution.... My point is, how do you feel about it? After working all your career on the same system now you...

Roger Burnett: This is very interesting because I think, having retired for three years, I have a certain perspective. So I think if I had heard this news when I was still working in a lab I might… I would hope that I would feel generous, but I think that I might have felt jealous or annoyed… I don't know what I might have felt but I was actually pleased to have the news. I guess that with the crystallography that was sort of a nice advance but I think I was particularly amazed by the electron microscopy work, because this is really an incredible advance in terms of what's possible by electron microscopy. And the fact they came in the same year. It is a nice unification because, as you know, we started trying to combine electron microscopy and crystallography really quite early, and so the fact that these two techniques have now come out. The only thing you don't have now is an NMR structure of a virus....

Nicola G.A. Abrescia: Roger, adenovirus has been a particularly useful system, because, you proposed –with Dennis Bamford and Dave Stuart– a sort of revolution in terms of how we classify viruses.

Roger Burnett: Yes.

Nicola G.A. Abrescia: This was a big conceptual revolution, because people on the street probably see viruses as one big thing. And you know that scientists have been classifying viruses according to their overall morphology, genome-type… at least until 2002.
Then you, D. Stuart and D. Bamford, proposed that maybe viruses could be grouped using a different principle: the structural similarity between those viral proteins that make the virus. This led to the idea of viral common ancestry. Does this mean that maybe viruses that infect bacteria and viruses that infect humans could be tackled using similar strategies? This is a revolution… so what should we learn by looking at viruses from the structural point of view?

Roger Burnett: I don't know whether it is a revolution exactly, but it's a kind of unifying principle. The idea that before... well, there wasn't really an idea. People saw groupings of virus families but they had no idea how they were related and so I think this is a way of sort of putting, or seeing, the relationships between them. And once you have relationships between the viruses, it kind of simplifies the way that you can think and also opens up other areas. Because, for example, if you know that the two viruses are related in some way, if you discover something for this virus you can maybe go and look for it in another virus. And, it opens up also the possibility of developing therapeutics. For example, you could... if you find an antiviral compound that interferes with the assembly of this virus family, one can maybe think to try it on that virus, so I think it is very useful.

Nicola G.A. Abrescia: This unifying view is very interesting because usually physicists look for unifying laws in Nature. So I think that there is a motive...

Roger Burnett: But actually the idea really dates back to when we first started looking at the adenovirus hexon structure, before we even had a molecular structure, when we saw that the low resolution structures showed that the molecule was pseudo-hexagonal. We could work out very painfully, by looking at negatively stained images of capsid fragments, the organisation of the virus. At that time, an electron microscopist named Nick Wrigley had made some very beautiful images of very large viruses like Tipula iridescent virus where they had many, many, many protein copies and we realised that the pseudo-hexagonal structure allowed sort of a close-packing hexagonal array on the facet, and that you wouldn't need very many different kinds of environment for that molecule to build a very large virus. A virus as large as you want it would only need five different kinds of packing. So before we had even realised the building block of the virus was the viral jelly-roll, we sort of appreciated that the outer shape was very, very powerful in its ability to allow complex shapes to be made, or rather large shapes to be made. They're actually simple shapes in a way.

Nicola G.A. Abrescia: You say "simple"… but if I look at the PDB (Protein Data Bank) and I compare the virus structures that have been deposited with the protein structures or even membrane protein structures,… there are not many virus structures. There are very few people interested in virus structures and I was wondering what are the limitations or why do more research groups not tackle viruses from the structural point of view...

Roger Burnett: I don't know. I think probably because a lot of people are interested in, and there is a lot of money available for, correlating proteins with disease; so the structure of a protein obviously allows you to attack disease questions, if there is a mutation in an enzyme for example it can cause disease. And I think the link is not so clear with a virus, because most of what one can attack with a virus, just with the virus structure, really doesn't tell you very much about how that virus is related to the disease it causes, so that's possibly one explanation. Also, of course, the explanation is an older one that the size of the virus has prevented much work on viruses, historically speaking.

Nicola G.A. Abrescia: So the methodology is still quite behind in terms of studying virus structure.

Roger Burnett: Not any longer. With synchrotron radiation and all the advantages you know about...

Nicola G.A. Abrescia: Yes, I found really remarkable the fact that there is a negative trend in terms of virus structures deposited in the PDB and a positive one for virus structures deposited in the electron microscopy data base (EMDB).

Roger Burnett: I was just about to say that. I think that electron microscopy in some ways is better at looking at virus structure; I mean the structure of the whole virus, than crystallography. In fact with this new structure coming out for adenovirus, it is clear that the [nominal] resolution of EM is going down, and this is the crossover point because the resolution of the EM and X-ray structures is about the same.

Nicola G.A. Abrescia: This is sort of a theoretical question: do you think that the folds of viral proteins are limited or unlimited? Do you think that we know all the folds that viruses could have?

Roger Burnett: Probably not. Because we don't know very many virus structures.

Nicola G.A. Abrescia: Do you think there is a reduced number of viruses that use the same principles?

Roger Burnett: If the idea is that there are certain pathways or lineages that link viruses, it seems likely that there are not that many. Let's say 10 or 20. And then it's likely that there are certain viral or protein folds that are common to each lineage.

Nicola G.A. Abrescia: I have the impression that our predictive tools for assessing the topological space of (viral) proteins are still not robust enough...

Roger Burnett: I think that it's becoming clear that the sort of protein folds universe is actually more limited than people had thought when I was a student, for example. Then I think people thought there really was an unlimited number, that each protein was probably folded differently. But that is plainly not the case.

Nicola G.A. Abrescia: Nowadays there is some pressure on basic science to become more and more translational and probably I feel this issue more being a young investigator. But, what do you think is the impact that structural virology has on vaccine and/or drug development? Is it true that the link between the two is not as direct as we imagine?

Roger Burnett: I think it is probably true, because probably structure doesn't play much of a role in vaccine development. I don't know much about this area, but it would seem to me that you don't need to know much about structure to make a vaccine.

Nicola G.A. Abrescia: Because with the current computing-power pharmaceutical companies can pre-screen in silico thousands and thousands of compounds for a specific target...but I still value the basic knowledge of a virus structure as a way to understand biological mechanisms that could lead to biomedical applications...

Roger Burnett: Exactly. But, you know, something like adenovirus has been used in a very different direction that maybe you are not aware about. I mean, it has been used for delivering genes. And, in fact, I was involved in a project at the Wistar Institute where people were using adenovirus to deliver genes coded for HIV proteins with the idea of developing an HIV vaccine, but delivering it with a gene rather than a protein. And, to do this, we were using a chimpanzee adenovirus that we had solved the structure of, and the idea was that you could use the chimpanzee virus as a delivery vector for the first vaccination because most people have not been exposed to chimpanzee viruses, unless they were zoo keepers; and then we would use our knowledge of the hexon structure to model changes on the loops to make a different virus that could then be used for delivering the second vaccine dose, or the booster dose.

Nicola G.A. Abrescia: Did it work?

Roger Burnett: In principle it worked.

Nicola G.A. Abrescia: Fantastic.

Roger Burnett: It worked in mice (laughs). It hasn't worked in terms of delivering HIV vaccines but that's a different question. But the idea of modifying the protein worked, yes.

Nicola G.A. Abrescia: Human and animal pathogenic viruses have a daily effect on our society. Some of the diseases that concern the World Health Organization are viruses like Ebola, Rift Valley, Influenza, and some of them are zoonotic emerging viruses. But I have the impression that a virus project becomes interesting or financeable only when the virus reaches rich countries. Is it just my impression?... What do you think? For example Dengue became a target when it reached the US...

Roger Burnett: Yes, I mean people in countries are basically selfish people; they tend to spend money for their own purposes. And so I think that, as you know, there is a movement to try to redress... Well, let's back off. Pharmaceutical companies presumably have to make money to finance their research and so, they are very unlikely to carry out research into diseases for which they can't recover the cost. But, there has been a movement, as you know, with the Gates Foundation to try to redress this problem and put money into areas where there is not conventional funding available. But I am not quite sure what you are asking exactly. Are you asking if people are selfish, are countries selfish?

Nicola G.A. Abrescia: No, no. I am aware that there are re-emerging zoonotic viruses and it seems that the focus is not fully on them… maybe Rift Valley Fever, Ebola… seem to have more resonance but, while they remain in Africa, Middle East, Asia or even Australia or New Zealand … we are not too much worried about...

Roger Burnett: There is also the practical problem. If you want to investigate viruses like that, then, as you well know, have to... it's much more expensive because you need containment facilities, there are also all sorts of regulatory issues to do with working on viruses, and so it's a much more complicated and expensive area. So for the individual investigator, it is presumably very difficult to do that because they have to justify why it's important to invest a lot of money in studying a disease that really in terms of the host country can only bring problems to it.

Nicola G.A. Abrescia: But, it's anyway fundamental to study them. Some of these viruses may reach through their vectors (c.f mosquitos) the southern Mediterranean countries for example...

Roger Burnett: Yes, but, for example like in Spain, where we are, with this institute, there is a remarkable investment in structural biology so I am sure there are problems that people in other countries haven't attacked, that may be attacked here. I don't know, Mediterranean diseases or...

Nicola G.A. Abrescia: Stepping now on the financial side of projects, we are facing a crisis. It is a global crisis and Spain as a nation is facing a quite harsh period. The Basque Country is in a slightly better position, but the bottom line is that we are facing cuts in research. If you were the person in charge of awarding grants, what line of research would you support? (now that you are retired you can freely express your preferences...)

Roger Burnett: That is a very, very difficult question, because I don't think anyone is really wise enough to allocate resources, except by a method that has really been shown to work and that's to put resources into funding the most imaginative and best investigators. But, then, how do you define the most imaginative and the best investigators. That's the question of, I think, judgement and taste. I was always struck when I sat on grant review panels and, one might think that each person on the panel would have their own taste in terms of what is good and what is worth supporting, but what was strange was the remarkable unanimity among people in recognising good projects and interesting things. You could argue, of course, that we are all very homogeneous and that we are not trained to be imaginative enough. And of course, there are certain waves of fashion in terms of what was interesting at the moment. But, in general, I think, if you have been trained well, and I think most good scientists are quite imaginative; I think that when you see something interesting, you can recognise it. So I think that if you put resources into people with a track record of doing good work, that's one way to allocate resources. The problem is, of course, that then one also needs to fund young investigators that don't have a track record. And the success rate there is going to be lower, because you can... it's difficult to tell at every stage whether somebody is going to go on to the next stage. I remember when I was in graduate school I would listen to lectures, and I would have my fellow students, and I think 'why would I bother to do research, because they are such smart people'; and then, not all these people would finish their PhD. And then you go to the next level, and there would be incredibly good post-docs, but not all of them would finish as a successful post-doc and publish papers. And then you have starting assistant professors, and some of them were brilliant, but not all of them would continue and push a project through. So I think, to have a successful research career and be productive and run the lab and be able to deal with people and so forth requires a range of skills that is quite broad and is maybe not immediately apparent at each stage beforehand. So, that said, I think that obviously no matter how tight the circumstances one should always put money into starting up labs for the young investigators because you don't know which one is going to be brilliant.

Nicola G.A. Abrescia: So, you don't have a preference between the lines of investigation?

Roger Burnett: No, I think it would be very foolish to do that. Because, if I said, the whole of Spain is going to work on virus structure, this would be nonsensical.

Nicola G.A. Abrescia: You retired in 2007 from active science...

Roger Burnett: Yes, end of 2006 actually.

Nicola G.A. Abrescia: Now, you have an impartial view, so, what do you think are the challenges that structural biology, not virology, has to face in the next twenty years?

Roger Burnett: That's a question. Let's take the different disciplines one by one. We'll talk about crystallography because that's the one I am most familiar with; I mean I was amazed when I visited the crystallography facility here, to realise that what we did when I retired was like something from the 19th century in terms of equipment. People were still doing crystallisation by hand and now there are robots and machines for doing these things; it's obviously becoming much more automated; the ability to crystallise in different conditions using small amounts of protein; obviously the ability to crystallise things and solve the structures has improved enormously. All that is good, so one can tackle more things. But obviously protein crystallography is sort of headed in a direction where it's becoming almost a service facility. In other words, people could say 'I've got an interesting protein, I want to know the structure' and you give it to the lab and they solve the structure. So, protein crystallography would then become like small molecule crystallography has become; so it's just an adjunct service, like in the pharmaceutical industry where they solve hundreds of compounds a day. So, the question is, what are people who are trained in x-ray crystallography... what is their existence going to be in the future? Are they just going to be technicians? Or... How are they going to continue to contribute in a more scientific and imaginative fashion to science? And, I don't know the answer to that question but, electron microscopy and NMR are probably further behind in this direction. In NMR, it's still a matter of a lot more manual interpretation perhaps being involved; certainly electron microscopy is that way. But all these techniques are headed to become mature techniques and so people who are interested in structure are going to have all these powerful tools available to them. Perhaps there are other tools that are now emerging like atomic force microscopy that gives very low resolution images at the moment that will become much more important in the future. But, it's hard to see into the future.

Nicola G.A. Abrescia: But, which system or pathway would you suggest to tackle, for example?

Roger Burnett: In what sense?

Nicola G.A. Abrescia: For example, should we study cell signaling, virus structures, chromatin regulation or…? What do you consider as hot topic? As example nowadays many labs are focussing their attention on membrane proteins, ion channels...

Roger Burnett: Yes, I sort of left out membrane proteins which is kind of still a difficult area. But I think somehow the advances in structure will come in trying to look at even larger things than say a defined macromolecular complex, maybe somehow imaging parts of cells. Or there are ways in which one can develop an x-ray microscope that is extremely low resolution at the moment that can image cellular organelles. So, of course this would be fantastic if one could do that and in the same way that one bootstrapped one's way up with x-ray structures of virus coat proteins to get the architecture of the whole thing, if one could use the tremendous number of structures we have at the moment that are known in detail, to try to put together some large organisations of organelles in cells, and that's really the new frontier, isn't it?

Nicola G.A. Abrescia: Now, coming back to the challenges of structural biology. Let's think that you're about to start a new career: a scientific career. What system would you target? You, personally. Let's say that you start at CICbioGUNE again.

Roger Burnett: I don't know.

Nicola G.A. Abrescia: Would you change the field of your interest?

Roger Burnett: I think I had a very satisfying career actually, because my goal of learning more biology, integrating that with physics; as a graduate student I became very interested in computing; as a physicist, one likes to tinker with things so one uses instrumentation. And, although I never fulfilled my goal of becoming a doctor in the beginning, I ended up as a professor in a medical school; so this is a very strange life, and so I feel very lucky, just by following what I thought was interesting, it was satisfying in terms of what we could do in terms of the science and also very rewarding personally. So it's hard to imagine being so lucky again. (laughs)

Nicola G.A. Abrescia: I was looking in Science last week, I don't know if you still have the habit to look through Science. Craig Venter published an article on the first cells with a synthetic genome… What do you think about the fact that we are able to produce synthetic cells?

Roger Burnett: Well, in a sense, we have been producing synthetic cells for a long time, because people have been tinkering with various genes and the fact that he has made a totally synthetic one… I am sure that he has based that upon functioning ones, because the chances of getting one right are probably very small, so, although it is synthetic, it probably resembles something that exists already. I haven't looked in detail into what that was but I am pretty sure I am right. Am I?

Nicola G.A. Abrescia: Yes. Now, you have been here for two and a half days and you had a look at what the BioGUNE is. My final question: What is your impression of the institute? As a member of the Scientific Advisory Board, what are your impressions about what you have seen? And the people that you have met?

Roger Burnett: Well, of course the people make the institute, you can have a beautiful building but without good people it would be a disaster. The building of course, or the physical layout, and the infrastructure are amazing. It's a very nice place in which to work, because as I said before it reminded me of the Biozentrum when I joined it in the seventies. It's very well equipped. Obviously it's well run as far as I can see in terms of allowing scientists to do their jobs; and then I have been very impressed by... it's a very young group of investigators and people are very enthusiastic. Everyone I've met enjoys being here, enjoys doing their work. So I think that good things could come out of it. Obviously, like any organisation it needs nurturing and needs to find a way to develop. But I think it's got great promise.

Nicola G.A. Abrescia: So you see high potential in it?

Roger Burnett: Yes, I think so, definitely.

Nicola G.A. Abrescia: I almost finished my interview Roger. We went through most of the questions that I was interested in asking you. So maybe we should stop here. Thank you.