Fighting the Flu in the Future: Medicine's Next Big Thing--In-Depth Doctor's Interview
Tim Whitehead, PhD, Assistant Professor and Chemical Engineer in Material Science as well as Bio systems and Agricultural Engineering, Michigan State University talks about a protein that inhibits the flu virus.
How many degrees do you have?
Dr. Whitehead: I have Bachelors in Engineering from Vanderbilt and I have a Ph.D.
This research that you did seems pretty exciting to have a protein to basically render the flu virus ineffective. Is that basically the gist of it?
Dr. Whitehead: That’s exactly the gist of it.
So what’s the process like?
Dr. Whitehead: About two and a half years ago some exciting results from Scripts Institute as well as Harvard University identified a portion of the flu virus that was an Achilles heel for the virus. So it’s something that is conserved among many different subtypes of the virus, things as varied as swine flu to avian flu. And this Achilles heel these groups targeted with antibodies and they can show that these antibodies render them ineffective. So we asked a simple very broad scientific question at the outset. We know enough to actually design a protein on the computer to go and attack that same area and actually inhibit the flu virus by attacking and binding at that Achilles heel. So that’s what we set out to do and it turns out that we were successful for a couple of our designs, we tested many and a couple of these were successful in actually binding the site. What we showed just recently is that we could optimize these sequences using a pretty powerful protein engineering technique. And one of these candidates could actually neutralize the flu at therapeutically relevant doses.
So when you say engineered on a computer, talk to me about how that process works.
Dr. Whitehead: Sure. What we do with that is we take a surface of a molecule. In this case with the flu there’s a specific protein called hemagglutinin. You may have heard of H1N1 or H5N1, the hemagglutinin is the H portion of that. There’s a specific patch on that protein which if you bind there can render these viruses ineffective. We can take the three dimensional coordinates of that patch which is known and we can find protein sequences that can bind at that patch. And so this is a question about how do you get something to bind to another thing. You know Velcro binds or silly glue binds on things. At the molecular level you have between proteins you have surfaces that are geometrically aligned almost like a hand in a glove. Where you would have to get these two surfaces to intertwine and not only that but there are certain patches on that surface that have to interact correctly. So grease goes with grease watery things or more hydrophilic things go with watery things and you have to do this at atomic precision. Using a code called Rosetta developed by David Baker’s lab at the University of Washington in which this work was done we were able to uh—optimize protein sequences to bind this specific region. And this requires as you imagine super computers and a huge distributed competing network called Rosetta at home where hundreds of thousands of people worldwide download our computer software and actually run the code to determine these sequences for us. We can get these designs out. There are hundreds of them and we can actually test them in the lab. We can actually order synthetic genes, this is now as easy as just calling up a company and e-mailing a sequence and they can deliver them in a white vial in two weeks. And then we can test them.
So you came up with the design for the protein they thought would render the flu virus harmless and then you sent it away and had it made?
Dr. Whitehead: Right. So what we can do is we can send away the DNA encoding these proteins and then the DNA comes and then we can pass them pretty quickly in locked boxes. As you can imagine these are early days in designing proteins but a couple succeeded. So that’s where we are.
So it’s the DNA deep sequencing. Can you break that down for me?
Dr. Whitehead: Sure. What we did at that stage is we had a couple designs that worked, that bound these proteins. But they were very, very weak, okay. What we did then is we had to figure out how to make the binding as strong as possible. A weak binder is not going to neutralize flu; we need it strong as possible to neutralize the flu. So to do that we developed a method at the University of Washington and in conjunction with other groups to be able to see how every change on a protein sequence relates to its function.
Dr. Whitehead: Most people probably know the alphabet of DNA it four letters long. For protein it’s twenty letters long. So for every protein sequence you have twenty different letters you can do at each point. And what we did is we developed a method to be able to see how well each letter or each mutation does for binding. And we can do this in a very high thorough quick fashion at once. And what you’re looking at is the boxes, the black boxes means that binding is about the same as before, red means binding has improved and green means binding is worse. And we’re able to do this in a very, very high thorough quick fashion. And from that we’re able to pick and choose mutations that could confer improvements in binding and we can pick eight or ten of them at once. Our designs are not perfect at the front but very quickly using this DNA deep sequencing technology we’re able to identify the eight or ten mutations that when combined can actually make these molecules potent against things like flu.
So once you have the engineered protein what do you do with it?
Dr. Whitehead: When we had the engineered protein and we made sure that these combined the hemagglutinin molecule with binding characteristics, our collaborators at the Naval Health Research Center were able to go and test the ability of these molecules to block influenza infections in cell culture models. So that’s where we are right now and comparing them to the antibodies that bind the same region we’re at about the same level, maybe a little better in terms of concentration but only slightly. So we’re at approximately the same level. So the next step is to actually test this in animal models.
So another cell—
Dr. Whitehead: So you have human cells and then you infect them with flu and a lot of them die but when in the presence of our protein the cells aren’t infected.
How did that feel to know that actually happened?
Dr. Whitehead: We’re pretty excited. We thought that was going to work but it really was exciting to have these designs work as we intended them to do. This is a powerful new approach where you can actually design for first principles something that can go and neutralize something as complicated as a flu virus and get it to work in the computer and then rapidly optimize it using the approaches we developed and then actually have it neutralize as well.
Will this work on every strain or do you have to kind of keep redesigning for other strains?
Dr. Whitehead: The reason we attacked the portion of the molecule that we did is because it is common to most strains, including many of the pandemic strains of the virus. This portion isn’t common to all though. So for example the pandemic Hong Kong flu from 1968 does not share this epitope. So right now at the Institute for Protein Design our researchers are designing things also and I’m cautiously optimistic that they are going to be successful on that front. We tested it against two different strains of H1N1 in the live virus models and others have been tested but we haven’t published them yet so I can’t speak of them right now.
You seem excited talking about this, this could be definitely something to change the world.
Dr. Whitehead: Right. We’re—we’re really excited not only for flu but in general about designing proteins from scratch. In the case of flu, because of the research at Scripts and at Harvard, we are able to have a good target. If we have that target we’re pretty confident now that we can design proteins that can for example, bind that patch and then we can optimize this using deep sequencing to have a protein at the end of the day that can prevent things like pandemic viruses.
Would it be something that would be added to the flu vaccine or would it be in pill form down the road?
Dr. Whitehead: Of course there’s a long road ahead involving things far beyond academic labs. But if it were to get to that stage, some of these governments would stock pile hundreds of millions of doses and then in the awful event of a new pandemic strain this could be a treatment. So you can either administer it before infection or after infection is the idea. But it would be completely separate from a vaccine or a universal vaccine that other groups are working on.
And this could be used in other areas as well not just the flu?
Dr. Whitehead: We think this is great days for protein design. Where you can design a protein to have a function whether it can inhibit a flu virus or target cancer, which is what I’m working on right now. Or actually developing new enzymes or new proteins that can break down biomass which is why I’m here at Michigan State among many other things that proteins can do. And so we think that this general strategy of designing proteins computationally and then optimizing them very quickly and rapidly using deep sequencing is going to be a very promising approach in the future.
Have the proteins been used in the past, it seems like you don’t really hear people trying to manipulate proteins?
Dr. Whitehead: There’s a pretty large field of antibody engineering. Antibodies are just complicated proteins that the immune system uses to to attack and destroy pathogens. There’s a pretty large field of people engineering antibodies. Pfizer down the road in Kalamazoo, Michigan or in any number of sites like Boston or South San Francisco there are pharmaceutical companies working on those type of treatments.
But that is a single protein that you work for how long?
Dr. Whitehead: Most of the cancer therapies that actually exist -- Herceptin is a big example -- are their monoclonal antibodies. They’re a single protein that does this. But the strategy to develop these proteins is long and laborious over two years. Whereas we want to do this in two weeks and we think we can now. For this project it took two years and two post docs, two PI’s and a grad student to deliver these proteins. I was one of the post docs on that project. But in the future we want to do this in two weeks and we think we can. And we think that we can leap frog advances made by people doing monoclonal antibodies in the future using this joint experimental and computational approach.
So smallpox as well is something you’re looking at?
Dr. Whitehead: The Institute of Prtein Design at the University of Washington is looking at smallpox right now.
Is there anything I missed you think is important for people to know?
Dr. Whitehead: I think the credit goes to David Baker at the University of Washington, let’s make sure that’s mentioned. I think the Rosetta code is good and the other thing is the Rosetta at home project I’d light to highlight. So that is a distributive computing for the code where you can download this on your personal computer and have it run code that can design proteins to cure the flu.
So hundreds of thousands of people pitch in to try to determine what code is best to make the proteins to fight flu?
Dr. Whitehead: Yes. You design a computer program and you send it out to these different homes across the world. This is like the third or fourth biggest. I think SETI is the biggest.
So Rosetta code is the name of the program?
Dr. Whitehead: Yes, Rosetta.
Who can log on and do that?
Dr. Whitehead: Anyone can go and download the Rosetta at home it’s just a screen saver.
Is it closed now?
Dr. Whitehead: Rosetta At Home, no it’s still open. But what happened is the users wanted to actually play with the molecules themselves, the protein because you could see how the protein is changing etc. So there was a computer science professor that worked with David at University of Washington to develop a game called fold.it, so fold.it, and tens of thousands of people played this and it’s actually an interactive game where you can try to make the best protein structure. They try to design the best flu, the anti flu molecule. If we can get a hundred thousand designs out and we can cull and pick out the best ten that’s still better than having the best protein designers in the world at the fold.it game go and design us a couple. But they’re getting good, they’re getting really good.
So how much input did the people at home add to the protein project?
Dr. Whitehead: They put a lot of the resources in getting the infrastructure built and then we wrote the program for them to actually run. So with all this infrastructure that has been built up over the last ten years or so it has enabled us to do the kind of work that we’re doing.
This is it: Solve Puzzles for Science.
So you can call in and say how did you do that?
Dr. Whitehead: Yes and they’re really, really nice, they’re really excited to explain how they did it. So a lot of people have science backgrounds but not everyone, in fact some of the best players in the world have a high school education and they just see proteins better than anyone in my lab or myself.
END OF INTERVIEW
This information is intended for additional research purposes only. It is not to be used as a prescription or advice from Ivanhoe Broadcast News, Inc. or any medical professional interviewed. Ivanhoe Broadcast News, Inc. assumes no responsibility for the depth or accuracy of physician statements. Procedures or medicines apply to different people and medical factors; always consult your physician on medical matters.
Sign up for a free weekly e-mail on Medical Breakthroughs called
First to Know by clicking here.
If you would like more information, please contact:
Tim Whitehead, PhD
Michigan State University
To read the full report, Fighting the Flu in the Future: Medicine's Next Big Thing?, click here.