Daniel Ellis is a PhD candidate in the laboratory of Dr. Neil King at the University of Washington’s Institute for Protein Design. His research centers on designing a universal influenza vaccine using protein-based nanoparticles. We sat down with Daniel to discuss his research, including the Phase 1 clinical trial that has resulted from it.
It’s common knowledge that the efficacy of the seasonal flu vaccine varies from year to year. Why does this happen?
Good question. The flu vaccine is currently the most useful protection that we have against influenza. But, as you said, the problem is that its efficacy varies from year to year. The reason for this is that the vaccine needs to be designed before anyone knows what the major circulating influenza strains will be. And as with any prediction, there is room for error. The strains could mutate, or the predictions could just be wrong. Efficacy therefore ranges from about 10 to 60 percent each year.
Why do so many influenza strains exist?
Part of it is due to constant mutation of existing viruses. Another part is that, unlike many viruses that are composed of one large RNA strand, the influenza virus has a segmented genome made up of eight different RNA strands. Differences in strains are distinguished by differences in two of these RNA strands, which encode the surface proteins hemagglutinin (HA protein) and neuraminidase (NA protein). Whenever you hear influenza described as H1N1 or H3N2 or anything else, those H and N letters refer to the identity of these HA and NA proteins. There are 16 and nine major HA protein and NA protein subtypes, respectively, and although they all have the same general function, they’re recognized differently by the immune system. This is what differentiates strains. Influenza viruses unfortunately have the ability to trade these genes with each other to suddenly produce novel viruses, which directly increases virus diversity and can cause pandemics. If you do the math, there’s a huge number of potential combinations and of novel influenza viruses that could potentially exist.
One idea that has been proposed to solve this efficacy problem is a universal flu vaccine.
What exactly does this mean?
Technically, a universal flu vaccine is one that would protect against all strains of influenza, and therefore doesn’t need to be modified each year. While a perfect universal flu vaccine may take some time, the most immediate goal of this type of vaccine is to more generally raise the efficacy of the current vaccines, and lower the number of shots that one has to get in one’s lifetime.
What are some common strategies used to create a universal flu vaccine?
To answer this question, we need to take a step back and discuss the HA protein. It’s a tall, slender viral fusion protein located on the surface of the virus, and it’s composed of a head and a stalk region. The HA protein plays a vital role, as it’s responsible for binding to the surface of the host cell, and internalizing the virus. Interestingly, the head region is immunodominant. This means that the large majority of immune responses are mounted against it. Unfortunately, this makes vaccine development incredibly difficult, as the head region is constantly mutating. There are other sites on the virus that are less prone to mutation since they would sacrifice functionality, such as the stalk of the HA protein, but for some reason the immune system really loves focusing on the head region instead. Some vaccines have therefore been developed without the head region, or blocking it in some way, in order to increase the chances of an immune reaction mounted against the stem.
There are also a few other proteins that researchers are going after other than the stem. One is the NA protein, which is a quickly growing topic. Another is M2, which is a small transmembrane ion channel found in the viral envelope. Surprisingly, there are also small parts of the HA head that also have potential. But in terms of the things that look closest to being truly universal, the stem seems to be our best hope, at least for the moment.
You’re also interested in developing a universal flu vaccine. What’s your approach?
We’ve taken a number of different approaches, but the one that has stuck uses protein-based nanoparticles. Essentially, we take the DNA encoding the viral fusion protein and we append it to DNA encoding whatever proteins make up our nanoparticle. After expressing this DNA in cells, we can produce nanoparticles that display the viral protein on their surface. These big, synthetic protein particles look like viruses to the immune system, resulting in a strong immune response.
In terms of a flu vaccine, we’ve created nanoparticles that display the fusion proteins from the same seasonal flu strains that were selected for the commercial vaccine. Using these, we’ve observed immune responses not only to the head region, but also the stem region, which is incredibly exciting. We’ve tested the vaccine in mice, ferrets and non-human primates, and we’re seeing similar immune responses across the board. The NIH is funding a Phase 1 clinical trial that is aimed to start in the spring of 2021. It will provide us with information not only on the safety of the vaccine, but also further information on the efficacy.
Do you have a hypothesis to explain why these nanoparticles are able to elicit a stem response?
That’s a really good question, and we do not have a perfect answer yet. One of our collaborators, Masaru Kanekiyo, recently published a paper showing that a nanoparticle displaying similar but distinct antigens, such as distinct HA proteins from the different flu strains, can result in an immune system that gains a preference for going after the parts that are similar between the antigens. This could in theory help our vaccine focus immune responses on the stem region. We investigated this in our upcoming paper and it appears that there are other unanticipated factors that are aiding our stem responses instead. Right now, our leading hypothesis is that the exact shape of our nanoparticle vaccine is advantageous for guiding the immune system to recognize the stem, and we are studying this idea in greater detail. But regardless, we’re very excited about the stem-directed responses that we’re seeing.
What led you to use protein-based nanoparticles as your strategy of choice?
The idea of putting an antigen on a nanoparticle is not something that we came up with. It’s been around for a while, and our collaborators at the NIH, led by Masaru Kanekiyo and Barney Graham, made a lot of the early progress. Our contribution was figuring out how to design novel nanoparticles from scratch, and building the exact architecture desired, including the addition of specific features not found in natural nanoparticles. We figured out how to make nanoparticles by mixing two separately purified proteins in such a way that we can control exactly when assembly occurs. That may not sound particularly exciting, but it’s important in ensuring that some of the complex proteins displayed on the nanoparticles fold correctly. This allows us to make a number of vaccines that would not have been previously possible, and to do so with greater precision, which is vital for something going to the clinic.
Do you think protein nanoparticles can be used in the development of vaccines against other viruses?
Definitely. Our lab recently published a paper on a self-assembling protein nanoparticle that induces a response against respiratory syncytial virus, and we’ve also published on nanoparticles displaying HIV-1 proteins. There is another story from our lab on a vaccine against SARS-CoV-2 using similar approaches, which should be published in the near future as well. In theory, this technology could be used against many more different viruses.
Thank you for taking the time to discuss your research, Daniel! We wish you the best of luck as you continue your work!