The headline was worrisome: “Bird flu becoming riskier for humans.” The story was about a new paper in PLoS Pathogens from Kawaoka’s lab that was said to identify “a specific change that could make bird flu grow in the upper respiratory tract of humans,” according to the lab leader.
Birds usually have a body temperature of 41 degrees Celsius, and humans are 37 degrees Celsius. The human nose and throat, where flu viruses usually enter, is usually around 33 degrees Celsius.
“So usually the bird flu doesn’t grow well in the nose or throat of humans,” Kawaoka said. This particular mutation allows H5N1 to live well in the cooler temperatures of the human upper respiratory tract.
H5N1 caused its first mass die-off among wild waterfowl in 2005 at Qinghai Lake in central China, where hundreds of thousands of migratory birds congregate.
That strain of the virus was carried across Asia to Africa and Europe by migrating birds. Its descendants carry the mutation, Kawaoka said.
“So the viruses circulating in Europe and Africa, they all have this mutation. So they are the ones that are closer to human-like flu,” Kawaoka said. (Australia Associated Press)
This is pretty scary sounding but it isn’t new scary sounding. Kawaoka confirmed and filled in the picture about a mutation we already knew about, which is why he looked at it more closely in this paper. As far as we know the suspicion that a mutation in the PB2 gene at position 627 that substitutes lysine for glutamic acid (the mutation is written E627K in shorthand) goes as far back as 1992 when Subbarao, London and Murphy showed it was needed for a bird virus to infect mammalian cells. The idea that temperature is important was reported by Massin and colleagues in 2001. The viral RNA genome is packaged inside ribonucleoprotein (RNP) particles comprised of nucleoprotein and three RNA-dependent RNA polymerase subunits, PA, PB1 and PB2. This package takes the virus’s RNA genetic material and uses it to make new viral protein and new viral genetic material. PB2 is part of this protein-making and gene replicating machine. It all gets packed up inside a viral outcoat that includes both the hemagglutinin and neuriaminidase proteins (the Hs and Ns of the various subtypes like H5N1) as well as some other proteins. This paper confirms that a switch from glutamine to lysine in PB2 makes a difference.
For some years we have been fixated on apparent differences between the cellular receptors preferred by bird flu viruses and human viruses. We have written quite a bit about it here (and here and here). But it was becoming ever clearer that the receptor story wasn’t the whole story or even the main story. Other things were involved. Early on, E627K came under suspicion. What Kawaoka and his colleagues have done is show that two H5N1 isolates taken from the same patient in Vietnam in 2004, one from the lower respiratory tract and the other from the upper respiratory tract, differed genetically in a few places, one being at position 627 of PB2. They then showed that this single change made a difference in the ability of the virus to infect cells from the upper and lower respiratory tract in the test tube and in inoculated mice. To show it wasn’t other genetic variations, they mutated that single position in each of the two isolates and showed that the biological properties of the virus switched with it: what before easily infected the lungs but not the upper tract now did the reverse, etc. It is quite elegant.
The relationship between transmissibility and infection of the upper tract is both on the giving and the receiving end. Infected cells in the upper tract are likely to shed more virus when the patient coughs and sneezes and an exposed person is more likely to expose susceptible cells. While we’ve known about this mutation for some time, this paper provides more direct evidence that the Glu-to-Lys mutation at PB2-627 affects growth in mammalian cells and the ability to infect cells in the upper respiratory tract. The effects are not completely clearcut. The avian version was also able to infect upper tract cells in some instances. The paper also shows it is not the only change involved. Other changes, as yet unknown, seem also to be important.
This work makes us more confident we are pursuing a reasonable line of inquiry, no small matter in a confusing question. But we haven’t as yet unlocked the key to what would make easy transmissibility in humans.