A bump in the road for a broad-spectrum flu vaccine?

Posted on March 12, 2014   by Benjamin Thompson

An estimated three to five million cases of severe flu infection are reported each year. This isn’t the same as the “I can’t come to work today, I’ve got ‘flu’…” type of illness (often caused by the common cold), but one that can have severe consequences. The WHO estimate that up to 500,000 people – particularly those in high-risk groups – die each year as a result of the virus and the problem isn’t going away – it’s called ‘seasonal flu’ for a reason.

Seasonal flu can cause localised epidemics, while global pandemics can be caused by flu strains hopping the species barrier from birds and animals into humans. These include the new H7N9 avian flu strain that is causing some concern among researchers.

Although flu vaccines can take time to produce, they are the best way to prevent an infection. The problem is that today’s vaccines only protect against a specific viral strain and its close relatives. The strains constantly change, so repeated immunisation are required.

The key to this change is Haemagglutinin, a molecule densely packed on the surface of the flu virus. Haemagglutinin is a glycoprotein that spikes out from the virus’s surface, helping it attach to, and enter, a host cell. The molecule’s globular ‘head’ is one of the first parts of the virus that defensive antibodies can attach to, making it an excellent vaccine target. Indeed, current vaccines mainly promote antibodies that attack the Haemagglutinin head.

Because the head is under attack, it is also under pressure to change. Mistakes made while the virus replicates its genetic material can lead to changes in the amino acid sequence of Haemagglutinin. These changes may prevent antibodies from binding, allowing the virus to escape from the antibodies.

While the majority of antibodies that attach to Haemagglutinin bind to the head of the molecule, this is not the only location available to them. Samples from humans taken after an influenza infection reveal that there are a group of antibodies that bind to Haemagglutinin’s stalk region, rather than its head. These antibodies are fewer in number, as the stalk is closer to the virus’s surface and less accessible on the densely packed viral surface.

This inaccessibility to antibodies means that the region is under less genetic pressure to mutate. The stem’s stability, coupled with its apparent similarity between strains, has led many researchers to suggest that it might be an ideal target for a broad-range vaccine, effective against a variety range of flu strains.

But how stable is the stem region of the Haemagglutinin molecule really? From a vaccine development perspective, there’s a huge difference between being stable (unable to change) and being ‘stable’ (reluctant to change). One could ultimately save millions of lives, the other waste millions of dollars.

New research from Birkbeck, University of London, published in the Journal of General Virology, has investigated how stable this area of Haemagglutinin actually is, by studying the genetic sequences of 3,440 samples of a subtype of influenza called H3N2. Also known as the ‘Hong Kong flu’, H3N2 caused a pandemic outbreak in 1968 that killed an estimated one million people. The collected samples give a continuous record of how the virus has changed between 1968 and 2009, when the samples were collected.

Given the inherent instability of the flu genome, which frequently introduces mistakes while replicating its genome, how can you identify mutations that are caused by evolutionary pressure rather than those that are caused randomly?

The researchers behind the study, led by Dr Adrian Shepherd, looked for mutations in the stalk region that suddenly appeared and spread rapidly throughout the viral population. These were described as ‘fixations’ and probably represent advantageous mutations that helped the virus escape antibody attachment.

To add further weight to their argument that these changes were helping the virus escape, the team added a spatial aspect to their research, looking at the 3D structure of the Haemagglutinin stalk. The researchers could see whether multiple mutations in the amino acid sequence of Haemagglutinin occurred close enough to each other in 3D space that they could fit within the binding ‘footprint’ of an antibody. If so, this would provide more evidence that the mutations were advantageous and helped the virus to escape from antibodies found in the human population.

The work suggests that there have been two incidences where mutations in the Haemagglutinin stalk have prevented antibodies from attaching. These areas are in the region where broadly neutralising antibodies are known to bind, suggesting that the region is not as stable as previously thought.

This finding creates a potential problem for plans to create a stalk-based flu vaccine that is effective against a broad range of strains. A worldwide vaccination campaign that targets the stalk would put a much higher pressure on the virus to change, allowing it to escape the immune system. It’s hard to imagine that a pharmaceutical company would invest the time and considerable capital necessary to produce such a vaccine if its lifespan would be limited.

However, Dr Shepherd is more optimistic: ‘Even if this is true, it may take a long time for resistance to occur, so a vaccine would still be beneficial. We’re also saying that there’s a specific section of the Haemagglutinin stalk to avoid, which will be an important part of vaccine development.

‘I’m not expecting influenza to be wiped out in my lifetime: it’s persistent, has animal reservoirs and mutates rapidly. It may be that using a multivalent vaccine that targets both the head and the stalk is necessary, in a similar way that multidrug therapy targets other infections.’

Lees WD, Moss DS, & Shepherd AJ (2014). Evolution in the influenza A H3 stalk – a challenge for broad-spectrum vaccines? The Journal of General Virology, 95, 317-24 PMID: 24187015