SOMETIME in early May 1997, a three-year-old boy living on Hong Kong Island inhaled a flu virus. The virus was swept into his lungs, where it entered one of the cells and started replicating. Soon there were thousands of viruses invading neighbouring cells. Within a few days, the child developed a sore throat, dry cough and fever. Two weeks later he was dead.
At first, doctors thought he had succumbed to a common flu bug – the infection sometimes kills young children as well as the weak or elderly. But a closer look revealed something much more sinister: the culprit was a new type of human flu virus, previously only thought to infect birds. Soon there were more deaths. Scientists rushed to Hong Kong from around the world and their findings made them fear the worst: that the next great flu pandemic was upon us.
The virus was certainly new and lethal, but fortunately it didn’t yet meet the third criterion for a pandemic, the ability to spread between people – it could only be caught from birds. But how long would it be before the virus acquired this ability? In desperation, the Hong Kong authorities slaughtered the island’s entire chicken population, some 1.5 million birds. This drastic step seemed to work, as no more human infections were reported. The world breathed a collective sigh of relief.
Advertisement
But the threat of a pandemic has not lifted. There were three flu pandemics in the 20th century and experts believe another is bound to happen sooner or later. Worryingly, similar strains to the one that caused the 1997 outbreak are still circulating in wild bird populations in China, and have reappeared in Hong Kong chickens twice, necessitating further carnage at poultry markets. In the past five years, virologists have been trying to probe the secrets of the Hong Kong strain to find out what made it so deadly. It’s no exaggeration to say that their work – which finally seems to be bearing fruit – could be the only thing standing between us and the next flu pandemic.
The main reason flu causes us such problems is that the virus’s surface antigens – the proteins that are visible to the immune system – are constantly mutating. This “antigenic drift” explains why getting vaccinated or catching flu one year won’t necessarily stop you from catching it another year (although you’re likely to see off the second infection faster). But every so often, the surface proteins undergo a dramatic change, often as a result of human strains incorporating genes from animal strains, leaving people with no immunity at all. It’s such “antigenic shifts” that trigger pandemics.
Before 1997, all virologists could do was try to investigate the strains that had caused previous pandemics: the Spanish flu of 1918 that killed an estimated 40 million people, and the lesser outbreaks of 1957 and 1968. But in each case, it has been difficult to find any preserved bodies of victims. Researchers have dug up the well-preserved victims of the 1918 pandemic buried in the Alaskan permafrost (Â鶹´«Ă˝, 14 October 2000, p 28). But they failed to find any intact virus and have had to try piecing together the viral genome from fragments in different samples – a laborious process. One group of researchers has high hopes that a corpse from Twickenham cemetery in London will yield better samples (see “Help from beyond the grave”). Surprisingly, there seem to be even fewer remains from the later two epidemics.
The Hong Kong outbreak gave virologists a new direction for their research. The virus killed 6 of the 18 people it infected – an extremely high death rate – so they urgently wanted to identify the key genes that made it so lethal. Although we don’t yet know for sure, there have been intriguing hints, gleaned from differences found between the Hong Kong strain and related, benign viruses.
The genome of the flu virus consists of eight pieces of RNA that encode ten proteins. The virus’s two surface proteins, called haemagglutinin and neuraminidase, bind to specific receptor molecules on the surface of a host’s lung cells, allowing it to enter and exit through the cell membrane. The receptor molecules on human and bird lung cells are so different that researchers had assumed a single virus could not infect both species. Bird strains usually have to pick up haemagglutinin and neuraminidase genes from human strains before they can infect humans. The fact that the Hong Kong strain had broken this rule made it all the more sinister. People had never before been infected by a pure bird strain so they would have had no immunity to it, raising the likelihood of a further outbreak turning into a global pandemic.
The first lead came from a group led by Kanta Subbarao, then at the Centers for Disease Control and Prevention in Atlanta, Georgia, now at the National Institutes of Health in Bethesda, Maryland. It had been known for a while that bird flu viruses sometimes acquire a certain mutation in their haemagglutinin gene that allows them to infect a wider range of tissue types than just lung cells, turning a relatively benign localised infection into a deadly, systemic one. Could that be why the Hong Kong strain was so vicious?
Subbarao began investigating virus taken from the first victim, the three-year-old boy. Chickens infected with the strain had flu virus throughout their bodies – they died of massive internal bleeding. And when she sequenced the virus’s genome, she confirmed that the suspect mutation was present. This had never been observed before in human flu. This finding was seen as so important that Subbarao’s work was published in Science, despite it being based on only one patient (1998, vol 279, p 393).
The mutation, says Subbarao, adds several basic amino acids at a crucial site in the haemagglutinin protein that must normally be cleaved by a lung cell enzyme during its synthesis. Because of the mutation a wider range of enzymes can cleave the site, so the virus can replicate in tissues apart from lung cells. “It causes a multi-systemic infection,” she says. “That explains why chickens die.”
But the theory didn’t last long. As more people fell sick and died, their autopsies revealed no evidence of the virus anywhere other than the lungs, and unlike the chickens, no patients had experienced haemorrhages. The scientists were stumped.
It was another four years before the next lead came, from Yoshiro Kawaoka, an animal virologist who divides his time between Wisconsin and Tokyo Universities. The break came when Kawaoka developed a sophisticated technique for synthesising new flu viruses in the lab, called “reverse genetics” (Â鶹´«Ă˝, 7 August 1999, p 77). It allows researchers to mix and match the eight flu genes from different strains to create new ones. For each gene they create a small circle of DNA, called a plasmid, then they introduce several into cultured cells. From the plasmids, the cells manufacture the complementary RNA strands and the encoded proteins, which then all become incorporated into new virus particles.
Using this system, Kawaoka’s team found that one particular gene from the Hong Kong strain was enough to turn a harmless flu virus into one that was lethal in mice (Science, vol 293, p 1840, 2001). It was the PB2 gene, which encodes part of a polymerase enzyme, responsible for replicating viral genes. A single “point mutation” turned a non-lethal strain into a deadly one by switching the amino acid glutamic acid for lysine at position 627 in the protein. “This mutation has to play a role in human virulence,” Kawaoka says.
Frustratingly, Kawaoka hasn’t yet discovered the mutation’s biological effect, although he’s on the case. There are several possibilities – one is that the mutated enzyme might be more efficient at producing new copies of the viral genome. But the PB2 mutation can’t be the whole story. Some of the viruses that had a benign PB2 gene but other genes from the Hong Kong strain also killed mice, suggesting that at least one other gene mutation contributed to the strain’s deadliness.
Four months ago, researchers led by Robert Webster, a virologist at St Jude’s Hospital in Memphis, suggested which one that might be. Again using reverse genetics, his team pointed the finger at a gene called NS1 (non-structural 1), which encodes a protein that helps the virus evade the host’s immune response.
During viral infections, many cell types start making chemical messengers called cytokines (including interferon and tumour-necrosis factor alpha), which tell the immune system to gear up for battle. Somehow the NS1 protein stops these signals getting through.
What Webster’s group has found is that the NS1 gene of the Hong Kong strain has a point mutation at position 92, an amino acid change from aspartic acid to glutamic acid (Nature Medicine, vol 8, p 950). This seems to make the virus more than usually resistant to the host’s cytokine response. When Webster infected pigs with an artificial strain that had seven benign genes plus the Hong Kong NS1 gene, the animals became much sicker than normal.
And there’s direct evidence from the human victims of the Hong Kong strain that Webster is onto something. Two of the post-mortems found much higher than normal levels of cytokines in the lungs and airways, and this was a causative factor in the deaths. Webster says in his paper that it’s possible there was “a mounting host cytokine response to influenza viruses that escaped the cytokines’ antiviral effects”
There’s also an intriguing parallel here with the ongoing research into the 1918 pandemic strain by Jeffery Taubenberger of the Armed Forces Institute of Pathology in Washington DC and Adolfo Garcia-Sastre of the Mount Sinai School of Medicine in New York. They have discovered that the 1918 strain may also have had an unusual NS1 gene, suppressing the cytokine response.
Again using reverse genetics, the group engineered a virus that contained the 1918 NS1 gene and used it to infect a dish of human lung cells. With a device called a gene chip microarray, they compared the expression of 13,000 different genes in the lung cells with that of cells infected with a modern, benign flu strain. The 1918 strain was found to suppress many genes that are normally switched on by one of the chief antiviral cytokines, interferon.
Unlike the Hong Kong strain, however, the 1918 NS1 protein had no point mutation at amino acid 92. “The effect must reside in other changes,” says Taubenberger. There must be another explanation for the anti-cytokine action, which remains to be discovered.
So can these findings help us prevent the next flu pandemic? If the outbreak is triggered by a bird flu strain similar to the one that caused the Hong Kong outbreak, existing vaccine-manufacturing techniques won’t work. We make flu vaccines by introducing virus into fertilised chicken eggs and incubating them. The virus is then harvested, killed and used in vaccines. But the Hong Kong strain is lethal to chickens so the embryos would die before producing enough virus. Even if it worked, this system takes six to nine months to produce and safety-test sufficient amounts of vaccine – too long to do much good in the face of a fast-moving pandemic.
Fortunately, several research groups are now developing ways to grow human and bird flu virus in cultured monkey kidney cells. What’s more, this method can produce greater quantities of the vaccine much faster than chicken eggs. And using reverse genetics, scientists can mix and match the eight flu genes from different strains to create whatever combinations they like.
So the technology is there – if a little experimental – for us to stockpile vaccines against future outbreaks. Unfortunately there’s not yet any global consensus on which genes from which strains should go into the mix. The NS1 gene is now looking like it could play a crucial role. But there may well be other gene mutations that are equally important.
According to Subbarao at the NIH, who is one of those trying to develop vaccines to bird flu strains, the more we find out about the Hong Kong virus, the better. “What’s a bit alarming is that [the Hong Kong] viruses continue to circulate in ducks and geese in the region and those remain a potential source of further human infections,” she says.
All we can do, it seems, is cross our fingers that a pandemic holds off for just a few years longer – and hope that governments realise the importance of expediting these lines of research. Martin Myers, who until recently managed the “pandemic preparedness plan” of the CDC, has complained that some officials have “no sense of urgency”. Speaking in September at the Interscience Conference on Antimicrobial Agents and Chemotherapy in San Diego, he said more people needed to realise that another flu pandemic was bound to happen sooner or later – the only question was when. “The clock is running,” he said. “But we don’t know what time it is.”
Help from beyond the grave
When 20-year-old Phyllis Burn succumbed to flu during the great pandemic of 1918, she was just one victim out of an estimated 25 million across the world. But her wealthy family’s decision to bury her in an expensive, lead-lined coffin will help a present-day virologist in his quest to understand what turns an ordinary flu virus into a killer.
Phyllis had just returned to London from service as a Red Cross ambulance driver in France during the First World War. She had survived the horrors and hardships at the front, only to face another kind of war – with the virus that was spreading through her body.
It’s known that Phyllis spent her final days in a house a street away from the family home, perhaps because she was afraid of infecting her nearest and dearest. On 30 October she took her last breath. It was at the height of the flu epidemic, and some victims were being buried in cardboard coffins or simply wrapped in newspaper. But Phyllis’s family could afford the best that money could buy.
John Oxford, a virologist at Queen Mary’s School of Medicine in London, is glad they did. He’s been searching for preserved samples of the 1918 flu virus for several years. The hope is that sequencing the genome of the flu strain that caused the worst pandemic in recorded history will help us arm ourselves against future outbreaks. After poring over thousands of burial records in London from autumn 1918, Oxford came upon Phyllis, who is buried at Twickenham cemetery in south London.
A lining of lead makes a coffin airtight, and that can keep a corpse in near-perfect condition for 150 years, experience has shown. Oxford has identified nine other bodies that may do, but Phyllis is the best hope. “She died young and had been in good health before succumbing to the virus,” he says. He hopes to isolate the complete virus genome from Phyllis’s lungs and then sequence it to look for clues to its virulence.
Oxford has permission from Burn family descendants and is now awaiting the go-ahead from church and health authorities to exhume the body. He hopes to start work early next year. The team will have to take biohazard precautions, just in case there is still live virus around. “To have a complete body is unusual. It will be the first time,” says Oxford. “We don’t expect to find live virus – but then no one expected to find well-preserved bodies in lead coffins, either.”