Biophysics in Influenza A Drug Design

The 2015-2016 flu season is expected to peak this month. Though this year is expected to be comparatively mild, annual flu season claims 36,000 lives and leads to millions of hospitalizations in the United States.  A flu pandemic can result in a more catastrophic impact, as witnessed by the 1918 Spanish flu and the recent 2009 swine flu. We spoke with Jun Wang, University of Arizona College of Pharmacy, about his research on the M2 proton channel of influenza A viruses.

What is the connection between your research and influenza?

M2 proton channel is universally expressed in the viral membranes of all influenza A viruses. It is a multifunctional protein that is absolutely essential for the viral replication. Among the 97 residues, the transmembrane domain (25-46) forms a homo-tetrameric four-helix bundle which mediates selective proton conductance. This function is essential for the viral uncoating once the virus is engulfed in the endosome. M2 is the known drug target of amantadine. However, more than 95% of current circulating influenza A viruses carry mutated M2 channels which render them resistant to amantadine, among which S31N is the predominant mutant. Given the relevance of M2 as an antiviral drug target, we are interested in understanding the mechanism of amantadine in inhibiting the wild-type M2 channel. Once we are convinced we understand this process, we would like to apply our knowledge to a practical exercise which is to design novel channel blockers targeting the S31N mutant.


Why is your research important to those concerned about influenza?

Annual flu season claims 36,000 lives and leads to millions of hospitalizations in the United States.  Flu pandemic results in more catastrophic impact as witnessed by the 1918 Spanish flu and the recent 2009 swine flu. However we are limited in countermeasures in prophylaxis and treatment of flu infection: only one oral drug, Tamiflu, is still in use. Given the lessons we learned from antibiotics and antivirals, there is no doubt that with the increasing prescription of Tamiflu it is only a matter of time before a majority of the sensitive viruses will evolve to become resistant to it. The shocking reality is that a large number of Tamiflu-resistant strains have already been identified from human patients. Thus, there is a clear need for the next generation of novel antivirals. The S31N inhibitors we discovered represent the second line of defense should Tamiflu fail to confine an influenza A virus outbreak during the next flu pandemic. S31N inhibitors have been shown to be highly potent in inhibiting multidrug-resistant influenza A strains and have synergistic antiviral effect with Tamiflu. Thus, they can be used in combination with Tamiflu to decrease the pace of resistance evolution.

How did you get into this area of research?

I began studying the M2 proton channel as a graduate student at the University of Pennsylvania in the lab of Dr. William F. DeGrado. The DeGrado lab has a long standing reputation in de novo design of four-helix bundles with novel functions. As M2 is a natural four-helix bundle with profound proton selectivity; the DeGrado lab was interested in understanding the structure and function relationship of M2 as well as the drug inhibition mechanism of M2. For example, how conformational change is coupled with proton conductance?; why M2 selectively conducts proton in a unidirectional manner?; and how does amantadine block the M2 channel? The knowledge gathered from such studies are critical as they serve as invaluable guides not only for advancing our fundamental understanding, but also the design of novel channel blockers. I first started by addressing the question regarding where the pharmacologically relevant drug binding site is for amantadine in M2. This part of the work was done in close collaboration with Dr. Mei Hong, who is now a professor in chemistry at MIT. Other major contributors in the DeGrado lab working on this project include Dr. Rudresh Acharya, an assistant professor at the National Institute of Science Education and Research in India, and Dr. Yibing Wu, a senior specialist in the DeGrado Lab.

How long have you been working on it?

I have been working on the M2 proton channels for 10 years since I began my graduate research in 2006. I continue working on this target since I became a PI at the University of Arizona. The primary focus of my laboratory in this project is to further advance S31N inhibitors to the stage of filling an Investigational New Drug application. The DeGrado lab continues working on the biophysical aspects of M2.

Do you receive public funding for this work? If so, from what agency?

The drug discovery of M2-S31N inhibitors are funded by both the NIAID, NIH (AI119187) and the PhRMA foundation 2015 Research Starter Grant in Pharmacology and Toxicology. We are also particularly grateful to NIGMS for their support of DeGrado’s work on M2 through GM056423.

Have you had any surprise findings thus far?

M2-S31N was traditionally tagged as an undruggable target because decades of traditional medicinal chemistry campaign failed to yield a hit compound. Thus, the discovery of the first S31N inhibitor by itself was a surprise finding. With this tool compound in hand, we were able to solve the solution NMR structure of S31N mutant in the drug-bound form. The structure revealed that S31N inhibitor binds to the mutant channel in a reverted orientation compared with that of amantadine in wild-type M2. The drug-bound S31N structure represents the Openout-Closein conformation which was not captured by previous structures.

What is particularly interesting about the work from the perspective of other researchers?

First, we resolved the controversy regarding the pharmacologically relevant drug binding site of M2, which allows other researchers to focus their efforts on the more relevant channel pore for their drug design. Second, using molecular dynamics simulations, we identified three hot spots aligning along the channel pore where the positive charged ammonium from the M2 channel blockers bind to. This is a reminiscent of how potassium channel blockers work, although the detailed mechanisms are obviously very different. This mechanism can be applied to guide the design of inhibitors targeting other ion channels.

What is particularly interesting about the work from the perspective of the public?

From the public perspective, the S31N inhibitors we discovered offer an opportunity for the urgently needed next generation of antivirals. As S31N inhibitors have no overlapping drug resistance profile with Tamiflu, they can be used either alone to treat infections with Tamiflu-resistant virus or used in combination with Tamiflu to achieve better therapeutic outcome. Moreover as M2-S31N is prevalent among circulating influenza A strains, S31N inhibitors are expected to have broad-spectrum antiviral activity.


Molecular Dynamics Analysis of Antibody Recognition and Escape by Human H1N1 Influenza Hemagglutinin


How did you compose this image?

The cover art–composed of the 1918 antibody (Ig-2D1) binding to the 2009 hemagglutinin (09HA)–is generated from visual molecular dynamics and enhanced in GIMP (GNU image manipulation program). The protein at the bottom is the 09HA, colored by residue name, and the protein on top is the antibody, colored in grey. The protein system was solvated in a water box and simulated in NAMD. We aimed to study the interactions between the antibody and antigen under simulated physiological conditions. We were inspired by the idea that inside the body, everything is dark, but the antibody and antigen complex is our “super star” and our study of interest; we let it “glow” in the dark as a symbol of human understanding of the mechanism of the interactions. The protein complexes are typically surrounded by water molecules, signified by the bubbles or water droplets. The bubbles reflect the complex and echoes the advances in knowledge and celebration of science.

How does this image reflect your scientific research?

The human immune response to the influenza virus is a fascinating area of research, and the thrust of our research aims to provide detailed molecular-level understanding of the mechanism of interactions between antibodies and antigens through computational simulation. In this particular study, we have, to our knowledge, uncovered novel understandings of how mutations in the influenza HA protein, in particular, the 2009 pandemic H1N1 HA, may confer immune escape from human defense. In particular, water molecules may contribute to weakening the interactions between antibodies and antigens in conjunction with the mutations.

Can you please provide a few real-world examples of your research?

Computational simulation plays a key role in advancing our understanding of human-pathogen interaction, and molecular dynamics simulations have been our main tool of investigation into various aspects of the influenza virus. For example, these techniques have been applied in the design of new inhibitors against influenza neuraminidase, and the current study explores the use of molecular dynamic simulation to compare antibody binding affinities, and could have immediate impact on antibody design and engineering.

How does your research apply to those who are not working in your specific field?

Atomic-level physics-based simulations are accurate enough to provide predictive hypotheses for experimental validation. The techniques are now applicable to researchers studying protein-protein and protein-small molecule interactions in other fields. Continued advances in computing technology will continue to enhance the realism of the modeled experimental systems, driving our “computational microscope” toward better understanding of natural phenomena and ultimately leading to better and safer therapeutics for humans.

Do you have a website where our readers can view your recent research?

– Pek Ieong, Rommie E. Amaro, Wilfred W. Li

The Science Behind the Image Contest Winners: Influenza A Virus and Mammalian Plasma Membrane Models

We are starting off the week taking a look at the science and scientist behind the scenes of the influenza A virus model that won second place in the Society’s 2015 Image contest.  The image was created by Heidi Koldsø, a postdoctoral researcher in Mark Sansom’s group in the Department of Biochemistry at Oxford University. Koldsø moved to Oxford from Denmark three years ago on an independent fellowship from the Alfred Benzon foundation.  Her research is mainly focused on understanding in details how membrane proteins, which are the largest pharmaceutical target, interact with their surrounding environment. She notes that understanding of not only the membrane proteins but also the lipids surrounding them is of the utmost importance – for example understanding how antimicrobial peptides target bacterial membranes and not human ones. Also, as illustrated in this image, if it is possible to understand how a virus interacts with the cell membrane, scientists can hopefully come up with solutions for better antiviral pharmaceuticals.


The images of the molecular models that captured the attention (and votes) of the 2015 BPS annual meeting attendees were constructed using the molecular visualization program Visual Molecular Dynamics (VMD). Ambient occlusion shading and rendering techniques were applied to the images within VMD. The image depicts an influenza A virus model in close proximity to a coarse grained mammalian plasma membrane model. Due to the large number of particles within these models we used a powerful workstation with lots of memory and a recent graphics card to create the image.

I chose this particular image to submit to the contest because it encapsulates both the achievements we have already made and those that are to come and it was a natural choice as an image that represents the science we study, the hard work we put into it and the potential real world applications it has. Both the project on the outer envelope of the influenza A virus and the large scale cell membranes are very ambitious and time consuming projects and are the result of a lot of hard work from a number of people. This image is the culmination of those two separate endeavours but also the beginning of the next.

When looking at the image, I hope it inspires the viewer to appreciate the current advances in computational modelling and how far we have come in our efforts to probe very real and relevant problems. The image is not only a pretty representation of a simulation as being able to visualize systems at these length scales provide us with unique insight to how these elements might look in vivo. The image does not only give a first glimpse of the molecular details of virus interactions with the membrane but also hopefully provides scientists and the general public with something that they can relate to and will hopefully promote intelligent discussion.

Supporting Scientific Information

The image is a product of the combined efforts of two large ongoing projects in the Sansom research lab at Department of Biochemistry, University of Oxford. The specific research from which the structures are taken involves performing coarse grained molecular dynamics simulations that we run on super-computing resources.

During the last couple of years I have been working on methods to construct complex asymmetric membrane models that allows us to move towards more ‘in vivo’ like systems. Our initial results on studies of complex membranes and the correlation between membrane nano-domains and curvature was published last year (Koldsø et al. PLoS Comput Biol (2014) 10(10): e1003911. doi:10.1371/journal.pcbi.1003911). We have recently started to move toward studying ‘in vivo’ like plasma membrane models at experimental length scales, meaning we are running simulations of systems >100 nm in dimensions and composed or millions of particles.

A complete model of the outer envelope of influenza A virus has recently been revealed, this project was initiated by a previous Postdoc in the Sansom lab,. Daniel Parton (currently working with John Chodera at MSKCC in New York) and has been continued by Tyler Reddy during the last 3 years in Oxford. This Influenza A virion model was recently published in Structure (Reddy et al. (2015) Structure, 23, 584-597).

The image is a result of combining these two projects on the influenza A virus and large scale mammalian cell membranes. We are exploiting the large deformations and curvatures that we observe in the mammalian plasma membrane model to accommodate the virus. This system is extremely complex and large (> 10 M particles) and Tyler and I are currently in the process of equilibrating and optimizing this model. Computational studies of the influenza A virus in close proximity to the cell membrane will potentially provide us with valuable information regarding how viruses enters the human cell during infection.

More information about my research and publications can be found on my website  and the website of my collaborator, Tyler Reddy.