Beating Pathogenic Viruses and Bacteria at Their Own Game

Guy S. Salvesen, PhD; Robert C. Liddington, PhD


September 17, 2012

Editorial Collaboration

Medscape &

This feature requires the newest version of Flash. You can download it here.

Guy S. Salvesen, PhD: Hello. I'm Dr. Guy Salvesen, Program Director in the NCI-designated Cancer Center at Sanford-Burnham Medical Research Institute. Welcome to this segment of Developments to Watch from Sanford-Burnham and Medscape. Joining me today is my colleague Dr. Robert Liddington, Director of the Infectious Disease Program in the Infectious and Inflammatory Disease Center.

Today's program will focus on key research efforts in identifying mechanisms by which toxins and viruses infect host cells and how this research will affect clinical practice.

Thank you for joining us today, Bob.

Robert C. Liddington, PhD: Thank you.

Dr. Salvesen: Of the millions or tens of millions of microbes that co-inhabit our environment, historically only a handful have caused the pandemics that have been devastating to human populations. Do we know why this is? What has research told us about what's special about this group of pathogens?

Dr. Liddington: We know many things about them, but we don't know everything about them. Historical records are imperfect. Certainly, a lot of the early pandemics were caused by bacteria, and in those cases, in general, we have antibiotics to treat those bacteria.

Dr. Salvesen: What is it about the interaction of viruses with the host cell that makes them so well adapted to attacking us and causing such devastating disease?

Dr. Liddington: Certain viruses have devised special ways in which to change the nature of their proteins and to change them rapidly so that our immune system, which obviously is there to fight and try to kill these invaders, is itself evaded by these viruses. It's their ability to change rapidly that is the key.

Dr. Salvesen: Your research is focused on developing an understanding of the proteins that interact with host cells at the atomic level, or structural biology. What is our understanding of the structure of the proteins that mediate the interaction between the virus and the host cell? What does understanding that lead you toward?

Dr. Liddington: The key element in this case has to do with the body's immune response. If we think in particular about the primary antibody response against the virus, antibodies recognize specific epitopes, or specific surfaces on the virus, and bind to them strongly. Many of those will neutralize the virus.

Dr. Salvesen: What kind of proteins would these antibodies be raised against?

Dr. Liddington: Typically, at least in the first stages, they'd be raised against a protein on the surface of the virus. These proteins are involved in binding and recognizing host cells, as well as the key process of actually getting the virus across the host membrane and inside the cell. Those are the 2 key processes that we're interested in.

Dr. Salvesen: You have atomic resolution models of these proteins, so you're able to better understand the interaction between the antibody and the antigen on the surface. Is that correct?

Dr. Liddington: Yes, that is correct. We need that level of resolution because antibodies have very, very fine specificity and would recognize single amino acid changes within a protein. If that change occurs in the virus, the antibody may fail to recognize, bind, or neutralize the virus.

Dr. Salvesen: Do you have a model that you could show us of the interaction or of the hemagglutinin itself?

Dr. Liddington: Yes, I do. This is a 3-dimensional jigsaw puzzle that is supposed to be a golf ball, but it will do very nicely as an analogy for one of the surface proteins. I'll talk about it as an example of the so-called hemagglutinin molecule of influenza.

Imagine there is the surface of the virus with an extruding stalk region and a large globular head, representing the receptor antigen. The surface of the receptor has many dimples on it, and single point mutations will change the shape of these dimples. An antibody that once recognized that dimple will then no longer recognize it, so the virus is able to enter the cell and the neutralization can no longer occur.

What is special about influenza virus, though, is that it can not only change the shapes of the little parts, but it can also change the shape of whole segments. It has a segmented genome, which is unusual among viruses. Those segmented genomes can actually swap between one virus and another, so you can get a mixing of these elements.

Under those circumstances, you can get a pandemic, because the nature and the shape of the protein changes so dramatically that there are no antibodies that recognize the new shape. That's what's special about influenza.

Dr. Salvesen: So the problem is that as the virus changes its surface properties, any immunity that was generated in previous infections is no good. Is that correct?

Dr. Liddington: That's what we believe is the case. That's correct.

Dr. Salvesen: This entry mechanism that you're telling us about -- is that a common feature of pathogens, or does each have a different way to infect us?

Dr. Liddington: A large majority of pathogens and most viruses, with the exception of such things as the common cold or poliovirus, have a membrane and have this kind of structure.

This protein does 2 things. It not only recognizes a receptor on the host protein, but it also takes a critical step: It enables or promotes the fusion of the membrane layer with the host cell membrane. In that way, it can escape inside the cell, where it can start doing its work.

Dr. Salvesen: What about bacteria? Are there similar entry mechanisms for some of the pathogenic bacteria?

Dr. Liddington: Some of the bacteria do, and certainly some of their toxins do. Anthrax toxin, for example, has a very similar mechanism. It has a toxin protein that recognizes a receptor, and it also fuses with the cellular membrane. There are many, many common pathways; HIV also utilizes a similar, related pathway.

If we could understand the nature of these kinds of membrane fusion proteins, as they're called, and how to stop them doing their work, then we could devise a host of antiviral agents.

Dr. Salvesen: That's a nice segue, because I want to ask you next about therapeutic approaches. You explained how a mutation in the major part of hemagglutinin, for example, defeats the immune system so that antibodies are no longer good against these new strains. Are there any conserved areas in the protein that are required and that do not mutate against which you could generate therapeutic antibodies?

Dr. Liddington: Yes. If we take the model again, the receptor region is very immunogenic -- antibodies are always raised against it. If you vaccinate someone with a virus, antibodies will bind and some will neutralize the virus. But it turns out that the key area, which we consider to be the Achilles heel of this virus, is actually down at that stalk, which would be the golf tee in this case.

The antibodies are too big to get down and hit the stalk region of the protein. Imagine that there is not a single golf ball on the surface of the virus, but there is a whole sea of golf balls, as if you're on a practice range. They're all packed closely together, so the antibodies can very rarely, if at all, get into the stalk region.

Dr. Salvesen: Let me interrupt you for a second. It seems to me that, in order to generate therapeutic antibodies to defeat this sort of virus, you want to go for a conserved region. Yet, you're saying the conserved region is very difficult to approach. So how do you overcome that?

Dr. Liddington: It's quite simple: You remove the protein from the virus. If you use the protein as the immunogen to attract antibodies, in this altered state, the conserved stalk region is now exposed.

We found that, by using a particular technique called phage display, we're able to attract antibodies to the stalk region.[1] It further turns out that the technique of phage display likes to bind to conserved, relatively rigid regions of the molecule, whereas when you have a normal immune response against viruses, the antibodies like to hit the flexible parts on the surface of the protein.

Dr. Salvesen: By generating synthetic immunogens in the way you described, you're able to create neutralizing antibodies that would recognize the actual live virus? Is that correct?

Dr. Liddington: Yes. When we go back and use these antibodies that we've generated, they are able to recognize the real virus and bind to these conserved sites.

Dr. Salvesen: What would be the benefit of producing therapeutic amounts of neutralizing antibodies? Who would be the population that would benefit from that?

Dr. Liddington: Let me first say that the beauty of these antibodies is that they hit a highly conserved region, which does not vary between viruses and virus strains. So while the outer part of the receptor antigen is mutating and evading the immune system, the stalk region does not change. In fact, parts of it simply cannot change because they are elements of this sophisticated molecular machine. A machine has moving parts, and moving parts demand that the amino acid residues must be very highly conserved.

What we found is that antibodies that we raised against the avian influenza H5 are able to neutralize nearly all hemagglutinin types. In principle, we have a universal therapy for influenza.[2]

Dr. Salvesen: What is the human population that would benefit most from these neutralizing antibodies once they become available?

Dr. Liddington: If you think about the last pandemic, which was the swine flu or H1N1, as you recall, there were certain populations that were particularly at risk, such as pregnant women and young children whose immune systems were not well enough developed, as well as people on immunosuppressant drugs and people underdoing cancer therapy. Those people certainly would be prime recipients of such therapeutics.

Dr. Salvesen: In terms of the targeted antibody or therapy that you've explained to us, what can clinicians look forward to in the coming years?

Dr. Liddington: Our first set of antibodies is now in preclinical trials with a pharmaceutical company, and we would hope that within the next few years, there will be products on the market. These are fully human antibodies, so we do not foresee particular safety problems, and they are directed against a pathogen; they're not directed against self.

Dr. Salvesen: Are there factors that might limit the introduction of this therapy into the broad, worldwide population?

Dr. Liddington: There are 2 concerns I can think of. One is cost. Antibodies are considerably more expensive than mass vaccinations. On the other hand, one could certainly imagine the strategic use of antibodies in specialized populations, such as young children and pregnant mothers.

The second concern is that if we were to provide this to everybody, then we might get ourselves into a situation like the one we have with antibiotic resistance and get the equivalent of methicillin-resistant Staphylococcus aureus, for example. We would start to generate resistant viruses. That would be a concern, so it would be better to limit the use of these therapeutics and to use them strategically when and as needed.

Dr. Salvesen: That's interesting -- the possible emergence of resistance to these antibodies. Something that the clinical community has learned over the years is how easy it is to generate antibiotic resistance. You're saying the same thing could happen with indiscriminate use of therapeutic antibodies as well.

Dr. Liddington: I'm sure it could. Our studies thus far have shown that it's very hard to get an escaped mutant of the virus that's actually a good virus. You can always make mutants, but they're weakened in some way. But to say we've never been able to find one doesn't mean that they don't exist.

Dr. Salvesen: Presumably, part of the ongoing research in this area is to determine from a structural point of view and a biochemical point of view whether the emergence of resistance is possible against this therapy.

Dr. Liddington: Absolutely right. That's where we come back to the requirement for the atomic resolution information. Our approach in this case is not to use a single antibody, but to use a cocktail of antibodies, to try to predict which kinds of mutations could occur in principle, and then see how each member of the cocktail of antibodies might be able to neutralize each of the different potential escape mutants.

Dr. Salvesen: That was a very interesting topic. Thanks, Bob, for participating in this program.

Dr. Liddington: My pleasure.

Dr. Salvesen: And we would like to thank you for joining us today. I hope you will join us for additional programs in the Developments to Watch series on Medscape.