"Just three mutations are needed to make bird flu a potential pandemic strain that could kill millions," is the alarming headline from the Mail Online. However, the chance of all three mutations occurring has been described as "relatively low".
Bird flu hit the headlines in 1997 when it was found that a strain of flu virus was spreading from poultry to humans in Hong Kong. The good news is that this strain didn't spread quickly between humans and therefore didn't spark a global pandemic in the same way as swine flu in 2009-10.
In a new study researchers analysed a strain of bird flu (H7N9) to see whether a particular surface protein on the virus could bind to human tissue. If it could, this would make human-to-human transmission of the H7N9 flu virus more likely.
The researchers found that three mutations of amino acids helped the virus bind specifically to human tissue and, in theory, could allow a human-to-human type transmission.
It is important to note however that the researchers did not actually engineer a virus capable of passing between people.
Although the media coverage has suggested that a "pandemic" of bird flu in humans could happen very soon, we have not seen the combination of mutations investigated in this study occur naturally. For now, these findings are useful for monitoring flu strains and to further our understanding of transmission but they are not a cause for alarm.
The study was carried out by researchers from several institutions in the US and the Netherlands, including The Scripps Research Institute and Utrecht University. It was funded by National Institutes of Health Grants, the Scripps Microarray Core Facility, the Centers for Disease Control (CDC), and the Kwang Hua Educational Foundation (JCP).
Some of the UK media's headlines were over-dramatic and needlessly alarming; suggesting that a pandemic of "human" bird flu could be on its way and millions could die.
This is absolutely not what this research suggests. Human-to-human transmission of these mutations was not tested, nor have we seen this combination of mutations occurring naturally.
The body of the Mail and The Independent's reporting was more measured and accurate.
This was a laboratory study on the H7N9 strain of bird flu. Researchers wanted to explore whether it could mutate into a strain capable of human-to-human transmission, similar to other infectious flu strains.
During an avian flu outbreak in 2013, the H7N9 influenza strain was transmitted from poultry to humans.
This occurred as a result of mutations in what are known as receptors – specialised cells that respond to external stimuli such as another flu virus. In the case of the H7N9 influenza strain, mutations caused a switch in what were once purely avian-type receptors to human-type receptors.
Once this happens there is also a chance subsequent mutations could enable transmission of the virus between humans, such as that seen in previous strains of H2N2 and H3N2.
Until now, H7N9 has only mutated to allow transmission of the virus from poultry to humans. The researchers involved in this study wanted to investigate how many further mutations would be needed for human-to-human transmission to become possible.
Laboratory studies such as this are useful as early stage research to get an indication of how biological processes may work. However, it isn't necessarily possible to accurately predict how these mutations would occur in nature, or how long they would take.
The researchers analysed both chicken and human "receptors" so that they could investigate which mutations are needed for the surface proteins on the H7N9 virus to attach to human tissues.
They introduced mutations previously seen in human pandemic strains of flu which had caused a switch between avian-type receptors to human-type receptors. For example, they looked at one called G228S which was involved in the H2N2 and H3N2 viruses.
The model was then validated using artificially constructed H7 proteins found to be able to bind to human tissue in the same way as the human influenza virus.
The researchers found they had to alter three amino acids before the flu virus could target human cells.
If all three mutations took place naturally then the virus could potentially spread from human to human (or mammal to mammal).
The researchers were unable to test this possibility using mammals such as ferrets as that type of experiment is currently legally prohibited under US law.
The researchers concluded: "In this study, we show that recombinant H7 proteins need three amino acid mutations to change specificity to human-type receptors. Although we are not allowed to assess if these mutations would lead to efficient transmission in the ferret model, this knowledge will aid in surveillance. If these amino acid mutations are observed to arise during natural selection in humans, timely actions could be taken."
This laboratory study analysed an H7N9 strain of bird flu. Researchers wanted to explore whether a particular change to the surface proteins of a virus was capable of allowing the strain to bind to human tissue. This would theoretically lead to human-to-human transmission of the flu virus.
It is worth noting that this ability to attach to human cells does not necessarily mean a mutated bird flu virus will be able to infect, replicate and transmit between humans. Other changes would also be required.
However, they were unable to further investigate whether this surface change could lead to human-to-human transmission of the virus because this type of experiment is not permitted under US law.
This research attracted the attention of several experts in the field.
Dr Fiona Culley, spokesperson for the British Society for Immunology, & Senior Lecturer in Respiratory Immunology at Imperial College London commented:
"This is a good, thorough study which specifically aimed to identify which changes in bird flu would allow the virus to attach to human cells."
"The authors found that certain combinations of three mutations were needed for the bird flu to be able to attach to human lung cells. They could potentially happen, but there is currently no evidence that they have ever occurred and the chances of all three occurring together is relatively low."