By Victoria Gill
Science reporter, BBC News
How flu strains adapt to be able to infect a new species is still a mystery
The flu virus is a survivor. It must continually evolve in order to evade its biggest threat - the immune system.
Mammals, including humans, make antibodies, which recognise and target the virus. "So it has to keep mutating to escape being destroyed," explains David Morens from the US National Institute of Allergy and Infectious Diseases.
Despite these tactics, most of the strains that make people ill during the so-called "flu season" are sufficiently similar to infections most of us have been exposed to before.
Our immune systems recognise common parts that these new strains share with their ancestors, and can launch an effective defence.
Every so often, however, a different strain emerges and infects people - one that contains new genes from an animal virus.
Its novelty is its most effective weapon against our immune defences. And if it is infectious enough to find its way easily into a new host - perhaps via an innocent sneeze - it can spread rapidly and cause a global epidemic - or pandemic.
"These events seem to be cyclical - they've occurred about every 20-40 years," says Mark Honigsbaum a researcher at the Wellcome Trust Centre for the History of Medicine at University College London.
"And all of these pandemics have been associated with just a few strains," he says.
Viral strains take their name from the various different types of two important proteins on their surface. Their initials are H for haemaglutanin, and N for neuraminidase.
The haemaglutanin molecule protrudes from the surface of the virus as spikes, and enables it to lock on to receptors on the surface of cells. Neuraminidase allows the virus to be released from infected cells and infect new ones.
"There are 16 Hs and 9 Ns, but as far as we know, only H1,2,3 and only N1 and N2 have ever made it into a human virus," says Dr Morens.
But beyond what is currently speculation that only a few strains may be transmissible among humans, the structure of flu gives us almost no clues about how it might behave and change.
"History teaches us to take these pandemics very seriously," says Howard Markel, director of the Center for the History of Medicine at the University of Michigan. "But it also teaches us that they are predictably unpredictable."
The first of the four pandemics was Russian flu - each takes its name from the country where the first case was reported - which emerged in 1889.
This was well before the science of virology had even been conceived.
It rapidly spread through Europe, and reached North America and Latin America, lasting until 1892 and eventually killing an estimated one million people.
Medical facilities were overwhelmed by the 1918 flu pandemic
"We have very little information about this strain," says Dr Morens. "But in the mid 1900s, scientists started collecting, and more importantly preserving, serum samples from people who had been alive during the flu pandemic. So there is now some evidence that it could have been an H3 virus.
"If it turned out to be true that only limited types of influenza genes have the ability to work in a human virus, that would be important, because you could vaccinate the world's population against the known possibilities of a pandemic virus," he says.
Although at its early stages, work is already progressing on a universal vaccine - one that would stimulate human immune systems to recognise, and raise antibodies against parts of the influenza virus shared by all strains.
'Mother of pandemics'
But, during the last century, the virus has shown a deadly ability to change beyond recognition.
In 1918, an influenza pandemic started that became a global disaster - eventually killing more people than the Great War.
Estimates of the death toll from the 1918 outbreak of Spanish flu range from 20 million to 40 million. Some historians argue it could have been as high as 100 million.
"There was a mild wave in the spring, but the very serious, lethal wave was in the autumn to the winter," says Professor Markel. "Then a third wave in January to April 1919, and a fourth wave in the winter of 1920."
This tendency for "waves" of infection and re-infection makes the virus yet more unpredictable.
New vaccines for circulating strains are now designed every year
At the time the medical consensus was that the disease was caused not by a virus, but by a bacterium called Haemophilus influenzae.
So some countries, including the US and UK, distributed vaccinations against the wrong disease-causing agent.
"Another problem was that the authorities stuck their heads in the sand," says Mr Honigsbaum.
"Their priority was the war and they didn't have the resources to deal with the health crisis.
"There's an argument that if they'd been more proactive, and diverted doctors and nurses from the front to civilian needs, they could have saved more lives."
Dr Morens refers to the 1918 H1N1 strain as "the mother of all pandemics".
"In the category of Influenza A, which is the category of virus that has caused all human epidemics and pandemics, every virus circulated since 1918 has been a descendent of this virus in one way or another," he says.
"Descendants of the 1918 pandemic are still infecting human beings, but they have mutated again and again and again to be able to survive."
Hybrids and mutants
The 1918 influenza pandemic gripped a vulnerable, unprepared human population, but its ability to "reassort" - or exchange its genes with other viruses - was what made it "dangerously novel".
"Every species has its own flu - when those species live together, and they can transmit their flu to different species, the virus itself changes its structure," says Professor Markel.
Since our immune systems recognise and respond mainly to the H and N part of the molecule, scientists suspect that pandemics arise when a strain emerges with a big change in the structure of one of these proteins.
But how exactly the virus adapts to enable it to attach to receptors on the cells of a different species, is an "unanswerable" question, says Dr Morens.
"With only four pandemics in over 100 years, our sample size is too small to say, but it seems that a new H [on the surface of the virus] has been the major factor," says Dr Morens.
"We don't know where the 1918 virus came from, but the evidence is that it was a new virus.
"At the same time that it infected humans, it also infected pigs. And at that point, we began to have two lineages of that disease - the human virus, and the pig virus, which persisted too," he adds.
The progeny of the 1918 influenza strain evolved and mutated as they were transmitted from one host to another.
And on two further occasions, these strains incorporated completely new genes and spread globally once again.
"That's what happened in 1957 and 1968 - a hybrid formed of the 1918 virus with genes that were never part of it before," says Dr Morens.
In the case of the 1957 Asian flu outbreak, a human H2N2 virus combined with the genes of a strain found in wild ducks.
The pandemic killed an estimated one million people worldwide.
An outbreak of H3N2 Hong Kong flu in 1968, when avian and human virus genes combined once again, claimed another million lives.
In both cases, the impact was minimised by health authorities, who identified the virus, and made vaccines available.
"Now, every year, around summer time, a group of flu experts get together and see what strains are circulating so they can design an appropriate vaccine," says Dr Markel.
And in the last few years, principally because of the global concern about avian flu, anti-viral drugs that target influenza, such as Tamiflu, have been introduced.
The emergence of a virus that crosses a species barrier is extremely rare
But Professor Markel points out that, despite having reached new levels of medical preparation, "we live in a world of emerging infectious diseases".
"We have learned to take avian flu very seriously, and we have learned to take the animal kingdom very seriously," he says.
But in the rare event that a virus does develop that is able to cross the species barrier, he points out that the close proximity of domestic farm animals to humans provides an opportunity for human infection.
"Human beings travel farther and faster than ever before. All of this means that we are set up for a potential epidemic or pandemic," concludes Professor Markel.
"We learn more every time, but the story of flu pandemics is still very much a story in progress."