How do you do, the things that you do, Mr. Flu?

The influenza A virus is a simple little thing. Just eight strands of RNA packed inside an envelope consisting of a host cell membrane. No metabolism, no fuzz, all very simple.

But, but, but – it is not very simple after all. In the flu business we still have large problems understanding many basic things this virus does. For instance, how does it survive in the environment? What governs host shifts? What makes a virus airborne, and what determines virulence properties? It is, in truth, a virus full of surprises.

This week, we published an article in Virology ( on a particular aspect of flu – the way it changes genotype and phenotype through reassortment. Does this sound dull? NO, this is extremely interesting – follow me and see. It is evolution in the fast lane!

One thing that we are interested in in my research group is how influenza A viruses evolve. Being a RNA-virus with a sloppy RNA polymerase (equivalent to the virus’ genome copy machine) it mutates a lot all the time, much more than eukaryotes (animals like you, me, amoebae, fish, insects and whales). Thus, natural selection has a lot of variation to act upon, and hence the virus’ evolutionary trajectories may spurt out in different directions. Sometimes very fast, other times more slowly, when balanced by evolutionary constraints imposed by the host.

Long-term evolution in hominids

Long-term evolution in hominids

Long-term evolution in equines

Long-term evolution in equines

But the influenza A virus has another tool in its toolbox when it comes to evolutionary change. The reassortment tool. Reassortment starts when two viruses infect the same cell. The cell machinery is hitchhiked and forced into making new copies of the virus genome and making new virus proteins. And we are talking many, many, many copies of virus genomes!

The newly borne RNA segments are embedded with certain viral proteins and migrate to the cell surface. There they are stuffed into small pockets of cell membrane and are released to the outside by the scissor-like protein neuraminidase, that cuts the last anchors binding the virus to the cell surface. Off they go to infect new cells! But, the process of stuffing RNA segments into the virus bag is a rather random process (at least we think it is), which means that RNA segments from one virus can suddenly get in a genome constellation with RNA segments from another virus.

Reassortment of two influenza A viruses in a duck

Reassortment of two influenza A viruses in a duck

This means that the new virus progeny can consist of segments (each coding for one, or two proteins; thus having a function) with completely different descent. Different evolutionary histories! This is also the process that has started several of the flu pandemics we have seen in humans during the last century. In this context, it is often talked about the pig as a mixing vessel where flu viruses with human origin can meet viruses with an avian origin, since the pig has receptors on its cells that allows both virus types to enter the cell. Coinfections in the pig could then give rise to reassortant viruses, sometimes with the mammalian transmission adaptions and novel antigenic properties from the avian side. Wham-bam – no prior immunity in the population of humans and loads of sick people! Thus, understanding reassortment is important also for our own health.

The picture below summarizes how drastically the genotype/phenotype could change with reassortment.

flu reassortment

Reassortment creates genome constellations where segments may have completely different descent.

But how common is reassortment in nature? We know it should be fairly common, but can we enumerate it? I have previously written a post on individual disease histories and immune responses in Mallards. That article, published in PLoS ONE in April (, followed 10 semi-domestic lure ducks in our duck trap at Ottenby for 1.5 years. The Virology article is a follow-up study on the viruses collected from these ducks in 2009.

It is a fairly complex story, and the most devoted readers of this blog should have a look at the original article (if you don’t have access contact me and we’ll sort it out). From the ten Mallards, we managed to retrieve 92 viruses of 15 different subtypes. With a detailed sampling scheme, where each bird was sampled daily, we could really determine the sequence of infection in each duck from primary infections, to secondary infections, and to study how different RNA segments were exchanged among viruses in different hosts across time.

In order to quantify reassortment, we used a combination of phylogeny-based and network-based tools (and we even needed to invent some novel ways of analyzing data, to make it happen). The level of reassortment was extremely high in our population of Mallards, and we estimated that >50% of the viruses were reassortants.

In a way this doesn’t make sense. If segments are exchanged so frequently, we would expect the influenza A virus gene pool to be in a panmictic stage, and that there shouldn’t be any linkage between segments. The solution, we believe, is that there is a strong selection afterwards, in the environment, or in transmission, where certain combinations of segments have a higher fitness. The discrepancy between the structured populations, where certain subtypes such as H4N6 or H1N1 are very common, and others combinations such as H4N1 or H1N6 are very rare, and the promiscuous panmictic reassortment levels suggest very strong selection pressures outside the host cells.

This is also the reason why unicorns are rare.

Jonas Waldenström