Spanish Lookalikes
Spanish Lookalikes
"On reading," by Simon Wain-Hobson, is a weekly discussion of scientific papers and news articles around gain of function research in virology.
Since January 2024, Dr. Wain-Hobson has written weekly essays discussing risky research in virology that were originally published on the Biosafety Now website.
We will republish these essays on our Substack every Friday, so the full archive will become available under the “On Reading” tab at the top of our Substack homepage.
What is the next pandemic flu virus strain going to look like? This question has been banging around flu virology for ages and pre-dates the modern genetic era.
The reservoir for flu viruses is aquatic birds, particularly ducks where flu is primarily an intestinal infection. Flu viruses have long been characterized by antibodies identifying two crucial proteins on the outside of the virus, the hemagglutinin (H) and neuraminidase (N). There are 16 distinct H and 9 N proteins making 144 possible combinations of H & N going from H1N1 to H16N9. Of these 135 (94%) have been found in the wild. Within many of these 135 H & N combinations there are myriads of individual strains.
Human flu virus pandemics have been provoked by only three called H1N1, H2N2 and H3N2. They are often characterized by a swap of the H protein — H1N1 Spanish flu was replaced by H2N2 Asian flu which was replaced by H3N2 Hong Kong flu. However, this is not a steadfast rule.
Occasionally spillover infections from birds to humans have been noted and when they occur, they can kill up to 60% of those infected. Such occasions are rare, less than several thousand documented symptomatic cases in over 20 years. Most fortunately, they are very rarely transmitted to other humans. This is why they are also referred to as dead-end infections. Typical culprits are designated H5N1 or H7N9 although there are at least 12 other combinations of H & N that can spillover.
The Fouchier and Kawaoka experiments that provoked the GOF controversy involved what is called a high pathogenic H5N1 virus from chickens that they adapted to efficient transmission between ferrets. By contrast, there are large numbers of low pathogenic flu viruses in the aquatic bird reservoir that only give mild symptoms.
The authors of the above paper took a low pathogenic flu virus of ducks and swapped the natural H gene of the virus for each of all 16 H genes derived from other, mainly, low pathogenic bird flu viruses. Although such mixing and matching was done in the lab, flu viruses can readily exchange genes in the wild, so this is akin to what occurs naturally — see aside.
They then tested the effects of these novel viruses in mice. Although they could not predict the outcome, they anticipated gain of function for some viruses, notably serious lung disease, compared to the parental virus and so performed all their work in an enhanced BSL-3 lab which was good. What did they find?
Five viruses proved to be as virulent as Spanish flu virus. No less! As the N component was always N1, these Spanish flu equivalents would be termed H1N1, H6N1, H7N1, H10N1 and H15N1. The H1N1 finding isn’t a surprise as it has created a number of human pandemics, notably Spanish flu as well as the much milder 2009 pandemic.
For H6N1 there have been less than a handful of documented spillovers, so the severity of the infection in mice is to be noted. To date no H7N1 or H10N1 spillovers have been recorded but that did not stop a group performing a GOF experiment of great concern on an H7N1 virus from an ostrich.
The H15N1 result is another surprise as no such virus has yet been identified in the aquatic bird reservoir, even ten years after this paper was published. H15 in combination with, for example N4 or N8 have been described, but not H15N1. And we learn from this work that the virus is highly virulent in lab mice.
Yet more surprises. The group were unable to make H4N1, H8N1 and H12N1 derivatives even though they exist in ducks. Negative results like these are notoriously difficult to fathom. Perhaps H4, H8 and H12 genes taken from other isolates might have worked. Yet it shows the difficulties in trying to anticipate flu viruses.
Conclusion: the authors made five very virulent GOF flu viruses for mice although they did not study their potential for transmission in the classic ferret model.
We read that The variable pathogenicity observed with these chimeric avian viruses did not correlate with SP-D hemagglutination inhibition, Simply put, SP-D is a natural molecule that interacts with the H protein and blocks part of its function. In so doing the results didn’t concord with expectations suggesting other mechanisms for the observed pathogenicity. As the paper was all about swapping H proteins there are clearly complex knock-on effects.
This suggests that the pathway for an avian influenza A virus to become a pandemic strain is an extremely complex, multifactorial process not easily achieved. Agreed. Although the authors talk up their findings somewhat, their message is: infections with avian influenza viruses bearing one of these hemagglutinins may cause enhanced disease in mammals, and therefore surveillance for human infections with these subtypes may be important in controlling future outbreaks. Everything is in the conditional.
Enhanced surveillance is important to characterize novel viruses at the animal-human interface, but ultimately, a broadly reactive “universal” influenza vaccine may offer the best protection against future pandemics. The ultimate goals are more surveillance and working towards an efficient universal vaccine, assuming that is a possibility.
This is a well-executed paper, carefully written, which turned up many surprises as well as illustrating the difficulties in fathoming the pathology of flu viruses in mice. They didn’t actively seek GOF viruses although the outcome was anticipated which was responsible. The working hypothesis was vague but plausible requiring experimentation to shed some light on it. The result was more questions than answers as happens in science.
Published only two years after the GOF controversy the findings are a far cry from those of Fouchier and Kawaoka papers which ignited the GOF controversy. They touted the next pandemic strain and with it, preventive drugs and vaccines. The paper of Qi and colleagues is the more surprising as the key authors are from the NIH which also funded the work Fouchier/Kawaoka work. Everything comes back to a hoped-for good vaccine and of course surveillance. Who could disagree?
This is the nature of flu virology where prediction and promises should be treated with great caution.
Aside 1
The prerequisite for mixing flu virus genes is an animal infected by two simultaneous infections. This leads to a few cells getting infected by the two viruses which is where the gene mixing occurs.
Aside 2
The body temperature of ducks is around 41°C while that of mice varies between 36.2 and 37.5°C. The authors used a mutant PB2 gene that encodes the enzyme that copies flu virus genome RNA. The mutant allowed better growth which makes sense when doing experiments in mice.
 | A guest post by
|