SciencePodcast_110513

Science Magazine PodcastTranscript, 13 May 2011/science_podcast/SciencePodcast_110513.mp3 MusicHost – Stewart WillsGreetings, and welcome to the Science Podcast for May 13th, 2011. I'm Stewart Wills. This week: designing protein interactions from scratch; new hopes in the search for an AIDS cure; and comparing the efficiency of photosynthesis and photovoltaic cells. All this, plus a wrap-up of some of the latest science news from our online daily news site, Science NOW.PromoSupport for the Science Magazine Podcast is provided by AAAS: the American Association for the Advancement of Science. Advancing Science, Engineering, and Innovation throughout the World for the Benefit of All People. AAAS—the Science Society—at .Music endsHost – Stewart WillsA paper in Science this week describes a new method for designing from scratch, and on a computer, proteins that bind with high affinity to specific patches on other proteins -- in much the way that an antibody binds to its target. Using the method, the team was able to design two proteins that bound to a specific molecular surface on the 1918 pandemic influenza virus. It's not a cure for the flu -- but the paper's senior author, David Baker, suggests that the technique, once perfected, could prove useful in drug development. It's also potentially useful as a way to gain a deeper understanding of protein-protein interactions, which are crucial to a wide range of biological processes. Baker spoke with me about the work from his office at the University of Washington.Interviewee – David BakerProtein-protein interactions are really critical throughout biology; they’re basically how all the individual parts in our bodies interact with each other. People have been studying protein interactions for a long time, and a critical test of understanding is being able to make new ones. So, we wanted to know whether we understood enough about protein-protein interactions to design brand new interactions from scratch, and this has been the goal of the basic research we’ve been doing in my group for some time in this area. If one could design protein-protein interactions from scratch, there’d be many interesting and, perhaps, useful applications, and one possible one is described in this paper. Interviewer – Stewart WillsNow, when you say designing proteins from scratch, or protein-protein interactions from scratch, you’re talking about designing them on a computer, and I’m guessing there isn’t an app for that. How do you actually set something like that up?Interviewee – David BakerWell, that’s the hard part, right? You have to develop a computer algorithm which knows enough about the physics of protein-protein interactions to be able to create new ones from scratch, and so in the computer program, one encodes everything one knows, or thinks one knows, about what’s important in protein-protein interactions, and, then, when you actually go and make designed proteins - designed interactions - the ideas you put into your algorithm are validated or refuted by the success or failure of your designs. Interviewer – Stewart WillsAnd, presumably, that requires a great deal of computing power, as well.Interviewee – David BakerYeah, it does take a lot of computing power. We’re very fortunate to have volunteer computing from Rosetta@home volunteers who have been helping us with all aspects of our research.Interviewer – Stewart WillsOkay, well, let’s take a closer look at the actual work you did. You chose to look at a specific region on a protein from the 1918 pandemic influenza virus. Why that virus, and why this specific protein and region that you looked at?Interviewee – David BakerWell, if you were going to design a protein to bind to a target, you might pick a target that, where if you could make a binder, it would be useful, and obviously the influenza virus, especially the H1N1, is a target that one could imagine in a new flu outbreak would be very important to be able to bind and to block. We chose that epitope because there had been a structure determined fairly recently which showed that if you bound that epitope, you could block the virus, so we thought it was a good candidate for this approach.Interviewer – Stewart WillsSo, you selected that virus and that epitope. What did you do then? Tell me a little bit about the work.Interviewee – David BakerWhat our computer method does is – well, its goal is to be able to design a very tight binder to any surface patch on any protein, so, basically, we pointed the algorithm at this particular surface patch on the influenza hemagglutinin, and the way the first step in the procedure is to look for handholds and footholds, basically ways in which amino acid side chains can interact very strongly with that surface. At this stage, they’re freely floating, they’re disembodied, so you might find that one type of amino acid can fit into a nice groove on one side of the surface patch that you’re trying to target, another type of amino acid fits in somewhere else. Then, in the second stage is like connecting the dots; we have to find scaffolding that will hold all of those interactions we’ve identified in the first step in the right places.Interviewer – Stewart WillsSo you, in some sense, sort of start with the strongest interactions, match those up, and then sort of iterate to refine the fit?Interviewee – David BakerThat’s right, that’s right. So we start by – literally, it’s like having disembodied hands and feet, just trying to figure out how you could really hold onto that surface tightly, and then we try and place the body so that the hands and feet are connected.Interviewer – Stewart WillsSo what did this computation actually produce in the end?Interviewee – David BakerIt produced designed proteins for about 80 different candidate designs, each of which has a different amino acid sequence, and each one is predicted to bind to the surface of the virus. So the next thing we did, of course, is we had to experimentally test them, and the computer, our computational methods, which used the Rosetta program we’ve developed, are quite far from perfect, so we actually ended up testing about 80 different designs and only found two that bound significantly to the virus.Interviewer – Stewart WillsAnd how does that sort of test work actually?Interviewee – David BakerWell, the way we do it is, it takes advantages of real advances in synthetic biology. For each of these 80 designed proteins, they don’t exist anywhere in nature, they’re completely new. What we do is, essentially, fill out an order form to a gene synthesis company that makes genes, and we send it the sequences of these 80 proteins, and they send us back genes, and then we take these genes and we put them into yeast cells in such a way that the proteins, these designed proteins, appear on the outer surface of the yeast cell, and then to test whether they bind, we add fluorescently-labeled target protein and see whether the cells become fluorescent.Interviewer – Stewart WillsSo, in the course of this process, you ended up with computationally-derived proteins, and you tested those and found the ones that really bound tightly to this area of the influenza virus, so you have this method. What’s the sort of larger significance? What sort of applications might you envision of a technique like this?Interviewee – David BakerWell, there are many possible applications. One is for new drugs. We don’t know yet whether these small proteins that we’ve designed will actually block the infectivity of the virus, but all the indications we have so far from the biochemical experiments suggests that they will, and you could imagine generating many more such designed proteins that target different flu strains. You could imagine designing proteins, and we’re doing this,to other pathogens, like Ebola virus, or pretty much any pathogen, whether it’s a virus or a bacterium, has to get into cells somehow, and those interactions are almost always mediated by proteins, so if you could design small proteins to bind to those, you might have sort of a general approach to blocking pathogen entry. Also for diagnostics, if you wanted to detect something, you could design a protein that binds tightly to it. Another use is in molecular cell biology. Protein-protein interactions, as I described, play roles in almost all cellular processes, and it’s very hard to disrupt them specifically. But if there was a particular process you were interested in, and you had a designed protein which bound tightly to one of the partners to disrupt that interaction, then you could use these small designed proteins as tools to understand those interactions and their importance. Interviewer – Stewart WillsI gather, though, that in some sense, we have a proof-of-concept here, there’s still a lot to be done…Interviewee – David BakerYes, there’s a huge amount to be done. I can emphasize that in two ways; first of all, we had to try 80 different designs to get two that worked, and that obviously is far from desirable, we’d like to have methods which are much better which allow us to, you know, make it possible to design high-affinity binders in essentially every try. Another thing is that the initial binders, even those two, do not bind very tightly to the virus, so we had to use experimental methods to optimize them to become really high-affinity binders. So there’s a lot of room for improvement in the computational methods, and we’re working hard on that now. One thing we’re going to be trying is the power of protein design game players. We’ve developed an online video game called Foldit, in which the general public can try to solve hard science problems. And we’re going to be posting on Foldit these influenza-binding designs and trying to see if -- challenging people to improve them, and we’re then going to test them in the lab. So we’re very excited to get sort of the general public, people outside of our group, working on design problems.Interviewer – Stewart WillsWell, that sounds pretty cool. Where would I go to actually do that?Interviewee – David BakerWell, you would just look for Foldit on the Web. You’d Google Foldit, and it gives you instructions on how to sign up, and then you would see a certain selection of puzzles there now, and in not too long, you’ll start seeing flu virus design puzzles.Interviewer – Stewart WillsWell, David Baker, thanks very much for being with us.Interviewee – David BakerThank you.Host – Stewart WillsDavid Baker is the senior author of a new paper on designing protein-protein interactions from scratch via computer. You can read the paper, as well as a related Perspective article, in this week's Science.MusicHost - Stewart WillsSince the advent of antiretroviral therapies, most AIDS treatment efforts have focused on managing and minimizing infection by the HIV virus, rather than the apparently more quixotic goal of actually finding a cure. But recent events -- particularly the experience of one remarkable patient -- have breathed new life into the notion of an AIDS cure, and have set researchers on a race to explore new strategies toward that end. Science contributing correspondent Jon Cohen writes about this new race to cure HIV infections in a News Focus article in this week's Science; he spoke with me about the story from his office in San Diego.Interviewee – Jon CohenI think a lot of people assume that AIDS researchers have been working on a cure since the very beginning of the epidemic, and the truth is it’s just been far too complicated a challenge, and it’s only been in the past few years that it’s really become a serious research pursuit by many groups, and that’s what the story is about, and it’s also the story of one person who, in all likelihood, has actually been cured.Interviewer – Stewart WillsWell, before we get to that, what is it about HIV that has made it so resistant to a cure? Interviewee – Jon CohenHIV, when it infects the human cell, it weaves itself into our own chromosomes, and that means it actually, as long as it lies dormant, it actually looks human – you can’t even see it, the immune system has no indication that a cell’s infected. That’s been the biggest hurdle is that the virus isn’t even visible to the mechanisms that our body have to clear viruses.Interviewer – Stewart WillsWell, granted, it’s very difficult to cure, but there are therapies for AIDS already, these so-called antiretroviral drugs, that have been successful in at least keeping the disease in check. Why this push now, given that those drugs have been successful, to actually go beyond that and look for a cure?Interviewee – Jon CohenOne of the things that’s become apparent over the past few years is people who are on long-term treatment with these antiretroviral drugs, even if they have very low levels of the virus, they still are dying young of different diseases. Basically, the low level of virus causes inflammation probably that leads to aging processes – things like cancer and heart disease – so, the virus isn’t good for you, even at low levels. The drugs themselves are toxic and expensive, and there’s been a push to get the drugs to all of the poor people inthe world who need them, and there are an estimated 15 million people who need them, and the world has only been able to provide access to 5 million of those people. So we’re kind of backed up against the wall with the current drugs, as good as they are. Interviewer – Stewart WillsAnd, I guess, if I’m reading your story correctly, you know, as you mentioned, a lot of the new optimism that a cure might be achievable seems to stem from the experience of this one man, this so-called Berlin Patient. Could you talk just a bit about that? Interviewee – Jon CohenYes, that’s definitely thrown a lot of fuel on the fire and given people hope, because it’s a proof-of-concept. It’s not that people anticipate replicating what happened with him. He’s a very unusual case. He had a leukemia that threatened his life and he had to have a bone marrow transplant. And his doctor knew about this unusual feature in a small percentage of people who have immune cells that are highly resistant to HIV, and he suggested finding a donor who had that genetic composition to basically replace this guy’s immune system.Interviewer – Stewart WillsAnd what happened?Interviewee – Jon CohenWell, he first had to have his body conditioned to accept the transplant, and that was very severe chemotherapy, whole-body irradiation – it’s called ablation, they actually kill off his own immune system. He almost died from that. And they transferred the donor’s stem cells, and his leukemia returned, and he had to have it done a second time. But he stopped taking his HIV drugs, the antiretrovirals, from the first transplant, and the virus never returned. He hasn’t taken drugs since for his HIV, and it’s been 4-1/2 years, and nobody in any lab can find any HIV in any part of his body using the most sensitive tests available. So, there’s a growing consensus that he has actually been cured of HIV infection.Interviewer – Stewart WillsWell, that is a remarkable case, but it seems kind of like an outlier, I mean, in the sense that that kind of treatment path isn’t really open to everyone.Interviewee – Jon CohenThat’s right, it’s not, but you have to think about this in the broader perspective of -- 60 million or more people have been infected with this virus; no one on earth, other than this one man, has been cured in a clearly documented case. So it provides proof that the idea is possible. And there’s an old saying in science that if you have a talking dog, you don’t need a control group. You know, this is kind of a talking dog; it’s so remarkable that you can do something that many people thought was completely impossible. Well, it is possible. You don’t want to go down the route that this man went through. But there probably are clues, based on his case, to help people understand how to better engineer something that’s more widely applicable.Interviewer – Stewart WillsSo, given that experience – the experience of this patient – it’s raised people’s optimism on the cure front, and, at the same time, that his particular cure is not, you know, for lack of a better word, scalable, what are some of the other avenues that scientists are exploring, you know, to follow up on this?Interviewee – Jon CohenWell, what happened with this man – whose name is Timothy Brown; he’s gone public and he spoke with me at some length over many, many months – is that idea of what led to his cure, there is a way to modify that to make that very same idea more widely available, where you take a person’s own cells out of their body, and modify those cells so that they’re resistant, and put them back in. So, one avenue is to try gene therapy studies that do just that. Will that be widely applicable? Who knows? And will it work? Who knows? And it’s obviously an expensive way to do things. So there is the possibility of doing that. Something else that Timothy Brown’s case has brought to the fore is the notion that a cure need not remove all the virus from a person’s body. When his case was first reported, many people scoffed at it, but the supporters of it said, “Hey, wait, who cares whether there’s virus in his body? He’s not taking any antiretroviral drugs and the virus hasn’t returned.” In every other person who had gone off antiretroviral drugs, who had had a high level of virus in their body before the drugs, the virus came screaming back in a few weeks or months. So this created the notion of what’s called now a functional cure, which means you can live with the virus, but you’ve reduced it to such a level that your own immune system can handle it. So, several of the new approaches are aiming for a functional cure. They’re saying, okay, it’s okay if you have some virus left, but maybe there’s a way for us to reduce the amount of virus in people to the point that it no longer requires treatment with these expensive and toxic drugs.Interviewer – Stewart WillsWell, there was also an interesting thread in your article in a different direction about, you know, the fact that there are these latent viruses that are sort of hidden from how we might attack them, this sort of notion of, in some sense, trying to wake those viruses up, for lack of a better word.Interviewee – Jon CohenYeah. I think one of the important things that has happened over the past 15 years is – you know, 15 years ago, there were no good anti-HIV drugs. Then along came these cocktails that led to these Lazarus-like rebounds in people’s health, and completely changed the course of the epidemic. What those drugs did, and what they’ve done in the past 15 years, is, they’ve revealed that once you clean up most of the virus in a person, there still remains these pools of cells that have latent virus that aren’t popping out, so they’re not being detected by the immune system, and that becomes the challenge – how do you flush out those reservoirs of latent virus? So many of the new strategies are called purge strategies, or purge drugs, because they’re looking for ways to wake up those latent cells, make them produce the virus, and when they do that, when those cells produce thevirus, they will be killed by the immune system, or by, basically, self-destruction when they produce virus. The virus they produce can be mopped up by antiretroviral drugs, and over a period of time, researchers hope, you’ll eventually deplete the reservoirs to the vanishing point, or so low that they no longer pose a threat.Interviewer – Stewart WillsWell, it sounds like there’s a lot that’s really exciting going on. Is there any way to put even a vague number on how long it might be before one or more of these efforts actually pans out?Interviewee – Jon CohenI think what’s really going to surprise people is that many of the strategies being tested right now are FDA-approved drugs, so they can move into the clinic very quickly, they already have a good safety profile, and I would anticipate that there will be data showing progress – not showing cure, but showing progress – for example, reducing the size of a reservoir, much sooner than people expect. I would anticipate it within the next year or two we’ll actually see progress on this front, and millions of new dollars are going to this idea that once was in the realm of charlatans and quacks. These are now all the leading research groups are doing this now.Interviewer – Stewart WillsExciting stuff. Jon Cohen, thanks very much.Interviewee – Jon CohenThank you so much.Host - Stewart WillsScience contributing correspondent Jon Cohen writes about the emerging race to cure HIV infections in a News Focus in this week's issue.MusicHost - Stewart WillsArtificial solar cells, or photovoltaics, harvest energy from sunlight and convert it into electricity that can drive a variety of processes. Plants, of course, have been harvesting solar energy for millions of years, and using it to create chemical fuels and biomass, through the very different process of photosynthesis. But which process is more efficient? Robert Blankenship of the Washington University in St. Louis, and an international group of 17 other scientists in chemistry, biology, and physics, tackle that question in a review article in this week's Science. The answers to the efficiency question turn out to be not at all obvious -- but they have implications for building a sustainable energy infrastructure for society over the long haul. Here's Robert Blankenship. Interviewee – Robert BlankenshipThis paper grew out of a workshop that I organized on behalf of the Department of Energy to get at this question of what really is the efficiency of photosynthesis, and howdoes it compare to photovoltaic cells. Obviously, one wants to know the basic efficiencies of these processes as a way of making decisions about how to direct research funds, and also how to direct new developments.Interviewer – Stewart WillsAnd, I guess, the first question you need to ask here is just how do you, you know, what do you measure as efficiency in this context? We seem to be comparing two very different systems here.Interviewee – Robert BlankenshipRight. That was one of the challenges that we faced in this effort, in that even though both of these processes are solar energy storage processes, they work in rather different ways, and so it’s sort of an apples-to-oranges type of comparison. So we worked very hard to come up with consistent definitions of efficiency, make sure that we were comparing the same regions of the solar spectrum, and similar sorts of products that the two systems produced, so as to, as much as possible, come up with an apples-to-apples type comparison.Interviewer – Stewart WillsAnd when you say similar sorts of products, could you talk a little bit more about what that means?Interviewee – Robert BlankenshipPhotovoltaic cells produce electricity, and photosynthetic organisms produce molecules; fuels – sugars, or oils, or something of that sort – and so those are rather different forms of energy. And so we had to try to, in a sense, convert one into the other, and so the way we accomplished that was we mentally took the electricity produced by the photovoltaic cell and used that in an electrolysis system to convert water into hydrogen and oxygen, and then the energy content of the hydrogen, which can be burned as a fuel, can be compared to the energy content of the fuels that a photosynthetic organism produces. Interviewer – Stewart WillsSo, basically, you’re looking at the energy and efficiency associated with sort of a water-splitting reaction.Interviewee – Robert BlankenshipThat’s right, and the energy that a photosynthetic organism stores, in terms of the products that it produces, are actually at a very similar energy level to hydrogen, so that was actually a pretty close comparison that we could make there.Interviewer – Stewart WillsOkay, so when we define, you know, the target in this way, and define efficiency in this way, how does photosynthesis compare with photovoltaics, in terms of generating energy?Interviewee – Robert BlankenshipWell, that’s an interesting question. There’s a lot of misconceptions out there on that. I have a slide that I show that has a cornfield next to a big photovoltaic array, and I ask audiences which is more efficient and to vote on it, and the vote is usually something like 90% in favor of photosynthesis, even in professional science audiences. But the truth is actually the opposite, that the photovoltaic system is significantly more efficient, in terms of its actual conversion of solar energy into fuels.Interviewer – Stewart WillsAnd how much more efficient are we talking about?Interviewee – Robert BlankenshipWell, after you take into account various factors, in terms of the electrolysis and the efficiency of the device itself and so on, a sort of an average value for photovoltaics is around 10% efficiency, and that’s over the entire solar spectrum. For photosynthesis, a sort of average value maybe across all of agriculture would be something like 1%, and numbers up to 4-1/2 or 5% have been documented in certain situations, so it’s at least a factor of 2, or, perhaps, more than that, more efficient, in terms of the photovoltaics. Interviewer – Stewart WillsWell, I guess, in some ways, that’s not a surprise. I mean, plants didn’t evolve to produce energy for humans, just enough food for themselves. I guess they’re – are they’re, in some sense, just not tuned for this, or…Interviewee – Robert BlankenshipWell, I think that’s right, and we’re learning more about how plants work and how their systems are not necessarily set up for the highest efficiency. They have evolved, as you say, for reproductive success, not to produce the maximum amount of energy. So that we’ve identified, as part of our work, several places in the natural plant system that are particular sources of the inefficiency that are there, and some of those, we’re thinking about possible ways to improve those.Interviewer – Stewart WillsSo what are some of those factors?Interviewee – Robert BlankenshipWell, one of the things is that a plant will only use essentially the same part of the solar spectrum that we can see with our eyes, the so-called visible part of the solar spectrum. But there’s actually quite a lot of energy, especially in the longer wave-length infrared region of the spectrum, that the plant just doesn’t use. The photovoltaic cell will use a significantly larger fraction of that solar spectrum. So that’s one of the big areas of inefficiency in plants, is that right off the bat they throw away about half of the energy in the solar spectrum. So, there are a couple of possible approaches to that. One is to put in slightly different pigments that expand the plant absorption range, and that’s one of the projects we’re working on is to include additional types of pigments, and one that we’re using is called chlorophyll d, which pushes the absorption of the plant out about another 50 nanometers into the red region and gives you up to 19% additional photon flux thatyou can absorb there, so it can be a significant improvement. Other, more far-reaching ideas are to put in things like bacteria chlorophylls and pigments that absorbs much farther into the near infrared; that involves a more sort of global redesign of this system. Interviewer – Stewart WillsWell, you also mentioned in your paper something that you refer to as the legacy biochemistry of plants, and a particular protein in photosynthesis that seems to be part of the problem?Interviewee – Robert BlankenshipIt’s likely that photosynthetic organisms evolved after the sort of major metabolic cycles in most organisms had been invented. And, so, the ability to use light energy was, in a sense, grafted onto that, and, in some respects, there’s a mismatch between the amount of energy that’s present in a typical photon, and the amount of energy that would be present in many biochemical systems. So there is that aspect of it. And one of the places where the natural system really falls down is in the carbon fixation enzyme RuBisCO – ribulose bisphosphate carboxylase-oxygenase – and this is the first enzyme that takes carbon dioxide and fixes it into an organic compound. And this is sort of a famous enzyme for being inefficient and is inhibited by oxygen, and has a very sort of low turnover rate. And so there are lots of efforts to try to improve that, through a variety of mechanisms, to improve that particular part of the system.Interviewer – Stewart WillsWell, let’s step back and look at the bigger picture. Clearly, you know, our energy future is a big issue going forward; there’s obviously been a lot of talk about the role of plants and biofuels, you know, something to help manage the transition to an environment less dominated by fossil fuels. You’ve laid out some basic information in this paper. What do you think are the policy implications of what you lay out here?Interviewee – Robert BlankenshipWell, I think that many people agree that no single approach is going to be adequate to meet our energy needs, and so we really need to take all possible angles and look at them, including photovoltaics and including biofuels. And what this does is it allows us to focus on some of the parts of the biofuel picture that maybe can be improved using some more modern molecular biology, so-called synthetic biology techniques, and really make some significant improvements in how those systems work, rather than just accepting what nature has given us as the only possibility.Interviewee – Robert BlankenshipWell, Robert Blankenship, thanks very much.Interviewer – Stewart WillsYou’re welcome; it’s been a pleasure talking with you.Host - Stewart Wills。