Effect Measure

The complexity of the spread of flu virus

When influenza viruses with different genetic make-up co-infect a cell there is the possibility that they will mix their genetic endowments. The influenza virus is designed to do only one thing: make a copy of itself. It does this by tricking the host’s protein manufacturing machinery to use the virus’s genetic blueprint to make a viral copy. Influenza genes come in eight discrete packages and at some point these genetic segments are naked in the cell. If the segments of two viruses are in the cell at the same time the segments can mix and match, with some of the segments of one virus being packaged in the progeny viruses with segments from the other virus. This is called reassortment of gene segments. It is a major mechanism of genetic variation in flu viruses. Two of the segments have coding regions for two proteins in the outer protein coat of the virus, hemagglutinin and neuraminidase. There are 16 different major versions (subtypes) of the hemagglutinin protein (each with lots of strain variation) and 9 neuraminidase subtypes. The hemagglutinin subtypes are numbered 1 to 16 and neuraminidase subtypes 1 to 9. The only subtypes that normally infect humans are H1, H2 and H3 and N1 and N2. Today the seasonal flu viruses are either H1N1 or H3N2. But evidence has been emerging for a few years that the actual situation is much more complex because there are 6 other gene segments and they also reassort. A paper from the Holmes group at Penn State and their collaborators has started to fill some of the details:

Researchers who conducted genetic analyses of hundreds of influenza viruses collected during the 2006-07 flu season found that many different variants circulated in the US at the same time, suggesting that the way each year’s epidemic spreads is more complicated than previously suspected.

The scientists, led by Martha I. Nelson of Pennsylvania State University as first author, found that several different clades, or lineages, of influenza A subtypes H1N1 and H3N2 circulated at the same time and even in the same localities, according to the report in PLoS Pathogens.

“Overall, the co-circulation of multiple viral clades during the 2006-2007 epidemic season revealed patterns of spatial spread that are far more complex than observed previously, and suggests a major role for both migration and reassortment in shaping the epidemiological dynamics of human influenza A virus,” the report states. It says the findings indicate that several viral strains were introduced separately into the country.

Among other things, the authors found that clades of the same subtype (H1N1 or H3N2) exchanged genetic material through reassortment in a number of instances. In one case, this gave rise to an H3N2 variant that was sensitive to the antiviral drugs amantadine and rimantadine, unlike most H3N2 strains in recent years. (Robert Roos, CIDRAP News)

There were nearly 300 isolates from a season that was dominated by H1N1, but this subtype (H1N1) was present in 8 different variants or lineages, representing different combinations of the other gene segments. There was much less variation in the H3N2 subtype that year (only two lineages). The same location always had more than variant (whenever the location had more than one isolate), so all sorts of variations of influenza virus are circulating simultaneously, sometimes in the same infected person. Houston, Texas, had six different lineages of H1N1 circulating at once.

Is there a pattern in the spread of the different variants? If there is, it isn’t evident at this point:

“Rather than a single viral lineage spreading across the US, multiple lineages of both A/H3N2 and A/H1N1 influenza virus were separately introduced and co-circulated, allowing for reassortment within subtypes and greatly complicating patterns of spatial-temporal spread,” the report states. “Given the extent of genetic diversity observed during this season, obtaining a strong signal for the spatial-temporal pattern of spread of multiple different lineages clearly would entail a large increase in sampling.” (quoted from Nelson MI, Edelman L, Spiro DJ, et al. Molecular epidemiology of A/H3N2 and AH1N1 influenza viruses during a single epidemic season in the United States, PLoS Pathogens 2008;4(8))

We shouldn’t be surprised at this. If the H and N segments reassort so do the others, and they aren’t “seen” as easily (or at all) by the immune system. It is only recently we have had the ability to analyze the full complement of genetic detail in hundreds of related isolates. Along with this new information will have to come new ways of mathematical analysis, an absolute necessity if we are to make sense of the fire hose quantity of raw data being produced by these techniques.

Lots to do and technically difficult.


  1. #1 anon
    September 5, 2008

    however HA and NA seem to reassort more than the other segments.
    Or probably better: reassortants which just only change
    the HA survive
    and spread more often than others.

    But since usually flu dies out in America in Summer and is reintroduced the next season
    newly from Asia, this is not really evolution. Just some playingground in a distant
    continent, while evolution happens in SE-Asia

  2. #2 revere
    September 5, 2008

    anon: What’s the evidence that HA and NA reassort more? I’m not aware of it, so I’m interested in what you base that on.

  3. #3 Ren
    September 5, 2008

    I was reading today about the “universal” flu vaccine. They promise it will be a “kill all” for the flu, regardless of H and N antigen make up. It seems plausible to me, but I am still a bit dubious.

  4. #4 revere
    September 5, 2008

    Ren: We’ve discussed it a number of times, most recently here and here. We waiting to see.

  5. #5 anon
    September 5, 2008

    more successful reassortment in HA and NA should
    already follow from the increased diversity in HA and NA.

    But I think there are also direct observations to
    indicate this. Alberta ducks, maybe.
    Also intrasubtype reassortments in humans, 1941,1947(H1N1),2003(H3N2), pandemics 1957,1968…

  6. #6 revere
    September 5, 2008

    anon: The underlying question relates to the pressures on the genotypes. Why do you assert there is “more successful” reassortment in HA and NA? What do you mean by this and what’s the evidence? What is the diversity of NS1 or PB2?

  7. #7 anon
    September 5, 2008

    16 types of HA, 9 types of NA and each of these is almost
    as diverse as any single-type of the other segments.
    checking here:
    I count 12,9,11,79,8,51,7,12 different groups of diversity in the 8 segments.
    (segment 4 is HA, segment 6 is NA)

    When you just search for reassortments in genome-sets
    by computer, I think you usually find overproportionally
    many (my experience) with just acquisition of another
    HA or NA or both. These types of reassortants seem to be more viable.
    I have no rigorously prepared data, but could do it –
    it just takes some time and effort.

  8. #8 revere
    September 5, 2008

    anon: Fair enough. Reasonable answer. Let me ponder it.

  9. #9 The Doctor
    September 5, 2008

    What about recombination? Reassortment has traditionally been restricted to the transfer of a complete gene between two strains infecting the same cell. Recombination on the other hand theorizes the transfer small segments of RNA between influenza virons much smaller than entire genes.

    The findings reported in this article could in part be due to recombination.

    Revere what is your opinion?

  10. #10 revere
    September 5, 2008

    The Doctor: I’ve kept an open mind about the possibility of homologous recombination, but the studies that have looked hard for it (e.g., from Holmes’s group) has shown it rarely occurs, the pattern that pertains to other negative sense RNA viruses. So for now I agree with the scientific consensus it isn’t a major driver of genetic variation. When published studies present contrary evidence I’ll consider it. I don’t mention it much here because it always causes a contentious argument I don’t think is very useful and has proved very distracting, not to say, destructive at other flu sites.

  11. #11 The Doctor
    September 6, 2008

    Thanks Revere.

  12. #12 Makina Park
    September 9, 2008

    industrial, industry, industrie, accident, equipment, lighting, packaging, property, safety, supplies, electronics, flooring, products, chemical, furniture, health, accidents, hygiene, plastic, automotive, welding, woodworking, cutting, grinding, design, drilling, filling, heavy, food, second hand, paper, simple, machinery, machine, machines, maschinen, business, mastercam, router, training, cae, cnc, shirink, machinery, machine, machines, maschinen

    Thanks a lot…

  13. #13 anon
    August 30, 2009