May 3, 2012 (CIDRAP News) – The first of two controversial H5N1 avian influenza studies to see print suggests that just four mutations in one of the virus's surface proteins may be enough to equip it to spread among mammals, but the findings are freighted with qualifiers.
After months of debate and discussion, Nature yesterday published the report by Yoshihiro Kawaoka, DVM, PhD, and colleagues describing a lab-derived hybrid virus, with elements of H5N1 and pandemic H1N1, that spread among ferrets via respiratory droplets. Kawaoka works at the University of Wisconsin.
The authors say their findings answer the "fundamental question" of whether flu viruses wearing the H5 hemagglutinin (HA) surface protein found in H5N1 can spread in mammals.
However, the virus did not kill any of the ferrets or even make them severely ill. Moreover, the authors caution that they don't know if the four mutations that supported transmission in the hybrid would accomplish the same thing in an H5N1 virus containing no foreign genes. Their reassortant virus contained seven genes from 2009 H1N1 and just one—the HA—from H5N1.
The results "indicate that H5 HA can convert to an HA that supports efficient viral transmission in mammals; however, we do not know whether the four mutations in the H5 HA identified here would render a wholly avian H5N1 virus transmissible," the report states. "The genetic origin of the remaining seven viral gene segments may also critically contribute to transmissibility in mammals."
At the same time, the researchers say one of the four key mutations they identified has already been seen in circulating H5N1 viruses on three continents, so it would take only three more to create an HA that supports mammalian transmission. They assert that their findings, by alerting scientists to potentially dangerous H5N1 mutations, will help surveillance.
Publication caps long process
The Kawaoka study is one of two H5N1 transmissibility studies that sparked concern and controversy when they were first described in general terms last fall. The other was led by Ron Fouchier, PhD, of Erasmus Medical Center in the Netherlands, whose report is expected to be published soon in Science.
Out of fear that someone could exploit the findings to create and release a highly dangerous virus, the US National Science Advisory board for Biosecurity (NSABB) recommended in December that both studies be stripped of key details before publication. But after the authors—especially Fouchier—provided some additional information and clarifications, the board recommended on Mar 30 that the full versions of both studies be published, and federal health officials endorsed the recommendation.
Kawaoka told CIDRAP News that he and his team revised their report twice. In response to the initial NSABB recommendation, they expanded their discussion of the benefits of the research and of the biosafety and biosecurity measures used, he said.
"After my presentation at the WHO H5N1 consultation in Geneva in February, the consensus was that including even more detail in these areas would be worthwhile," Kawaoka said.
Meanwhile, a companion editorial published in Nature today said the "essential scientific elements in the Kawaoka paper were unchanged between the first and second NSABB deliberations."
Introducing HA mutations
As the Kawaoka report explains, HA is the protein that enables flu viruses to attach to host cells to start the infection process. The H5 HA prefers to bind to sialic acids found on avian cell receptors, called alpha2,3 sialic acids, whereas the HA in human-adapted viruses such as H1N1 prefers to attach to alpha2,6 sialic acids.
Kawaoka's team started with a 2004 strain of H5N1 from Vietnam. The article describes a complex process whereby the team induced random mutations in the HA of this virus and then identified those that caused increased binding to alpha2,6 (human type) receptors. The researchers then used reverse genetics techniques to create reassortants consisting of the mutant H5 HAs and seven other genes from the pandemic H1N1 virus.
The reasons the team used the 2009 H1N1 virus included the "high genetic compatibility" of that virus with H5N1 and the fact that both viruses can infect pigs, which can serve as viral mixing vessels, the report explains. "The coexistence of H5N1 and 2009 pandemic H1N1 viruses could provide an opportunity for the generation of transmissible H5 avian-human reassortants in mammals," it says.
The team then infected ferrets intranasally with several different hybrid viruses and placed them in cages next to uninfected ferrets to identify mutations that were associated with respiratory-droplet transmission. In an initial round, a reassortant with three mutations (N158D, N224K, and Q226L) was found to be transmissible, with five of six contact ferrets showing serologic evidence of infection, and with actual virus recovered from two of those.
'A fine balance'
In further experiments, viruses with these three mutations spontaneously added a few more mutations, including one (T318I) that was found to make transmission more efficient.
The experiments also showed that viruses that combined a nonmutant H5 with H1N1 did not spread, nor did a hybrid virus that contained an H5 with only two mutations.
As for severity of illness, none of the ferrets infected with the reassortant viruses died, and most showed only modest weight loss (less than 10%), the researchers say. Ferrets infected with a 2009 H1N1 virus (with no H5N1 elements) lost more weight, but the difference was not significant. (Natural H5N1 viruses usually cause fatal illness in ferrets.)
The team determined that three of the four key mutations (N158D, N224K, and Q226L) contributed to the ability to bind to human receptors, while the fourth (T318I) lowered the pH at which the protein could release genetic material into an infected cell, according to a "News and Views" article accompanying the report. In addition, the N158D and T318I mutations made the HA more heat-stable, which in turn contributed to the virus's ability to spread by airborne droplets, the article says.
The authors concluded that "a fine balance of mutations affecting different functions in HA (such as receptor-binding specificity and HA stability) may be critical to confer transmissibility in ferrets."
Overall, the team said, "the transmissible H5 reassortant virus preferentially recognized human-type receptors, replicated efficiently in ferrets, caused lung lesions and weight loss, but was not highly pathogenic and did not cause mortality."
The scientists also sought to determine if the US government's H5N1 vaccine and the antiviral oseltamivir (Tamiflu) would be effective against the transmissible virus.
They found that serum from humans immunized with the H5N1 vaccine reacted more strongly to the lab-derived virus than to a wild-type virus, suggesting that the vaccine could provide some protection. In addition, the virus turned out to be "highly susceptible" to oseltamivir, suggesting that the drug would be useful.
Big question marks
The authors say they don't know if the four key mutations that conferred transmissibility in ferrets would also "support sustained human-to-human transmission." Components of the seven H1N1-derived gene segments in the virus, such as the neuraminidase gene, may have "critically contributed" to its airborne droplet transmissibility, they caution.
However, they observe that one of the four mutations, N158D, which causes loss of a glycosylation site, has been found in many H5N1 viruses in the Middle East, Africa, Asia, and Europe. "Therefore, only three nucleotide changes are needed for the HA of these viruses to support efficient transmission in ferrets."
Knowledge of the four key mutations will help those who conduct surveillance in H5N1-endemic countries to recognize "key residues that predict the pandemic potential of isolates," the authors predict. (In the debate over publication of Kawaoka's and Fouchier's findings, however, some experts have said that surveillance is too sparse and spotty to take advantage of such findings.)
"Although a pandemic H5N1 virus may not possess the amino acid changes identified in our study," the authors conclude, "the findings described here will advance our understanding of the mechanisms and evolutionary pathways that contribute to avian influenza virus transmission in mammals."
In the accompanying Nature News & Views piece, Hui-Ling Yen, PhD, and J. S. Malik Peiris, PhD, of the University of Hong Kong's Centre of Influenza Research write that the findings "demonstrate that H5N1 viruses do have the potential to cause a human pandemic."
Although the transmissible virus is a laboratory creation, "it should not be considered an experimental artifact," according to Yen and Peiris. "Natural emergence of an H5N1-H1N1 hybrid virus is plausible," because some H5N1 and H1N1 viruses trade genes in lab experiments, and the viruses could mix in pigs in many parts of the world.
Some other experts offered different views on what the findings imply about the risk of natural emergence of an H5N1 pandemic strain.
Philip K. Russell, MD, a veteran vaccinologist, infectious disease researcher, and former director of the Walter Reed Army Institute of Research, said the findings do not greatly change his views on the pandemic risk.
Russell told CIDRAP News he was very concerned about the risk of an H5N1 pandemic when the virus emerged and spread widely, but his concern has ebbed as the virus has failed to gain human transmissibility even as it has continued to cause rare human cases.
"I think it's still a remote possibility," he said. "I don't think these finding change my views that much, because the experiment has gone on in nature, and that keeps on saying it's probably not going to jump. But those [Kawaoka] experiments do indicate that yes, it's possible."
Russell added that continued surveillance of H5N1 viruses is important. "And I think this genetic information would be useful in improving our surveillance," he said. "If we see some viruses popping up with the right constellation of mutations, we could increase our level of concern and maybe increase our vaccine preparedness."
Masato Tashiro, MD, PhD, director of the WHO Collaborating Center for Reference and Research on Influenza at Japan's National Institute of Infectious Diseases in Tokyo, offered a different view of the study's implications for pandemic risk.
He said many policymakers have underestimated the risk of an H5N1 pandemic. The findings by Kawaoka and Fouchier, he said, suggest that an "H5N1 pandemic will occur, although several experts said it will not." Also, the findings show that "only a few mutations will result in a pandemic virus."
These mutations have not been observed together in nature as yet, but they have been seen separately "in many human isolates of H5N1 as well as avian isolates, especially of clade 2.2, recently prevalent in Egypt," Tashiro said.
He also said the study implies that an H5N1 virus would not lose its pathogenicity in gaining mammalian transmissibility. "The highly pathogenic properties may be essentially retained by the pandemic virus, although the change in the receptor binding specificity may reduce the pathogenicity to some extent," he commented.
He noted that the airborne-transmissible virus in the study did not kill the ferrets, but it did cause some lung pathology. Other reassortant viruses containing more H5N1 genetic material could be more pathogenic than Kawaoka's hybrid, he said.
Tashiro agreed with Russell that the findings will be an aid to surveillance and risk assessment: "The pandemic risk assessment of the virus isolates can be done more efficiently focusing on these mutations."
Nature assessed risk of publishing
At Nature's behest, a biodefense agency outside the United States conducted an assessment and concluded that the benefits of publishing Kawaoka's study outweighed the risks. The assessment was posted online by the journal.
The assessment states, "There is no doubt that this information could be used by an exceptionally competent laboratory to provide the foundation for a programme to develop a pandemic strain of this virus. There is no evidence that this reassortant virus would be fully pathogenic in humans."
On the other hand, the information in the study offers several benefits, according to the assessment. It will help scientists understand the pandemic potential of flu viruses, the risks posed by the deliberate manipulation of pathogens, and the characteristics needed for an effective flu vaccine, the anonymous author wrote.
Russell agreed that the risks of publishing the full study details outweigh the benefits. He said he sees little risk of bioterrorists exploiting the information, given the scientific sophistication they would need. "If there is a danger, it's a danger of escape from the lab," he added, but he said Kawaoka's team "did it right."
Tashiro commented that the risk of someone using the findings to create and unleash a very dangerous virus "cannot be excluded—but the risk of a naturally occurring pandemic virus is more realistic." And even without Kawaoka's study, information about potentially dangerous mutations could be gathered from many already published papers, he said.
Lessons about publishing
In an editorial, Nature's editors commented that the long debate over whether to publish only a redacted version of Kawaoka's (and Fouchier's) studies was worthwhile. The NSABB had initially recommended finding a way to provide the full details of the studies to selected scientists with a need to know.
"Having now considered these matters in depth, the editors of this journal have decided that we will not consider either alternative for papers in Nature in the foreseeable future," the editorial says.
"A paper that omits key results or methods disables subsequent research and peer review. Furthermore, after much internal and external deliberation, we cannot imagine any mechanism or criterion by which to sensibly judge who should or should not be allowed to see the work. Nor do we believe that any restricted information distributed to university laboratories would stay confidential for long."
Imai M, Watanabe T, Hatta M, et al. Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets. (Letter) Nature 2012 (published online May 2) [Full text]
May 2 Nature editorial
May 2 Nature News & Views extract
Study risk assessment posted May 2 by Nature
Jan 25 CIDRAP News story "Virus in one controversial H5N1 study wasn't lethal"