May 28, 2008 (CIDRAP News) – Scientists have found evidence that North American avian influenza viruses of the H7 subtype are becoming more like human flu viruses in their ability to attach to host cells, which suggests they may be improving their capacity to infect humans.
The investigators determined that several recent North American H7 viruses have an increased ability to bind to a type of receptor molecule that is abundant on human tracheal cells and is less common in birds. Their results were published this week by the Proceedings of the National Academies of Science.
The finding—which comes as the deadly Eurasian H5N1 virus continues to be seen as the likeliest candidate to spark a pandemic—"underscores the necessity for continued surveillance and study of these [North American H7] viruses as they continue to resemble viruses with pandemic potential," says the report. The study was done by scientists from the US Centers for disease Control and Prevention (CDC), Emory University in Atlanta, and the Scripps Research Institute in La Jolla, Calif.
H7 viruses have caused a number of disease outbreaks in poultry in Europe and North America in recent years, though far fewer than the widespread outbreaks caused by the H5N1 virus. H7 viruses also have occasionally infected humans, typically causing only mild conjunctivitis. But a veterinarian died of an H7N7 infection during the devastating poultry outbreaks in the Netherlands in 2003.
Previous research has established that avian flu viruses prefer to link up with cell receptor molecules known as alpha 2-3 glycans, whereas human flu viruses, such as H1N1 and H3N2, prefer to attach to receptors called alpha 2-6. The terms refer to the nature of the link between sialic acid (SA), which forms the tip of the receptor molecule, and galactose, an adjoining sugar unit.
Differences in receptor binding, transmissibility
In the new study, with Jessica A. Belser of the CDC and Emory as first author, investigators examined the binding preferences of H7 viruses by exposing glycan microarrays—receptor molecules laid out in grids on glass slides—to solutions containing the viruses.
In addition, they dosed ferrets with H7 viruses and tested whether they became infected and whether the infection spread to other ferrets housed with them. Ferrets are considered the best animal model for studying flu transmission, because the distribution of cell receptor types resembles that in humans.
The viruses tested included two highly pathogenic H7N7 isolates from humans infected during the Dutch outbreaks of 2003, three H7N2 isolates from poultry outbreaks and a human case in the eastern United States in 2002 and 2003, and two H7N3 viruses from human conjunctivitis cases linked to a poultry outbreak in British Columbia in 2004.
The scientists found that the two Dutch viruses, one of which (called NL/219) came from the fatal human case, showed the typical avian preference for alpha 2-3 glycans. In contrast, the three eastern US isolates showed significantly increased binding to alpha 2-6 receptors. One of these three, which came from a New York state man who was infected in 2003, showed both a sharply increased preference for alpha 2-6 glycans and reduced binding to alpha 2-3 receptors—a characteristic that was also observed in H1, H2, and H3 viruses when they were first introduced into humans, according to the report.
The two Canadian isolates also showed an increased preference for alpha 2-6 receptors, compared with the Dutch strains, but it was less marked than that of the eastern US isolates.
To study transmission, the researchers inoculated groups of three ferrets with one of six viruses and then caged them with three other ferrets to see if the viruses spread by direct contact. They found that the viruses multiplied in all of the inoculated ferrets, and some showed respiratory signs such as sneezing.
However, most of the isolates did not spread efficiently to the other ferrets. The one clear exception was the New York virus, which was transmitted to all three of the previously unexposed ferrets and multiplied to high levels. None of the viruses spread via airborne droplets, as shown by the absence of infection in ferrets that were housed in cages next to the inoculated ferrets.
The scientists write that the virus from the New York man showed both "the most dramatic shift in receptor specificity" and the greatest transmissibility in ferrets. However, they also note that an H7N2 chicken virus from Connecticut showed similar receptor preferences but spread by contact to only one of three ferrets.
The authors say previous work shows that the preference for alpha 2-6 glycans seems to be essential for transmissibility in existing human flu viruses such as H1N1. But their results, they add, suggest that this characteristic alone is not sufficient to render avian flu viruses transmissible in ferrets, and that reduced binding to alpha 2-3 receptors may be one of the additional requirements.
"These results indicate that H7 influenza viruses from the North American lineage have acquired sialic acid–binding properties that more closely resemble those of human influenza viruses and have the potential to spread to naïve animals," the researchers conclude.
A reminder of unpredictability
Infectious disease expert Michael T. Osterholm, PhD, MPH, called the study scientifically "outstanding" and said it is a reminder that it's impossible to predict which breed of flu virus will evolve into the next pandemic strain. Osterholm is director of the University of Minnesota Center for Infectious Disease Research and Policy, publisher of CIDRAP News.
"It goes to the point that we don't know what the next pandemic strain will be," he said. "However, we're reminded with the very disturbing data in this paper that it very well could be an H7. The most important message we can take from this is that there will be another pandemic strain that will emerge—tomorrow, next week, next year, whenever, but it's going to occur."
Dr. Terence Tumpey, senior author of the CDC study, said North American H7 viruses studied since about 2002 generally seem to have shown an increasing affinity for human-type receptors, according to a May 26 Canadian Press (CP) story.
"These viruses are partially adapted to recognize the receptors preferred by human influenza viruses, but not completely," he told CP. "It needs to be adapted further. But I think it shows that potentially these viruses are changing."
David A. Halvorson, DVM, a veterinarian and avian flu expert at the University of Minnesota in St. Paul, called the CDC study very interesting but cautioned against concluding that North American H7 viruses in general are becoming more like human-adapted strains, since the study did not include viruses from wild birds.
He noted that the three eastern US isolates studied—the New York human virus, the Connecticut chicken isolate, and one from a turkey in Virginia—all showed increased binding to alpha 2-6 glycans, compared with a 1993 H7N1 isolate from North Carolina. The two Canadian isolates showed a lesser increase in alpha 2-6 binding.
"The H7 viruses associated with the live poultry markets in New York were first detected in 1994 and remained there until 2006, so when the 2002 and 2003 isolates were obtained they had been circulating in chickens (mostly) for 8 to 10 years," Havorson said via e-mail. "If we allow (conservatively) one infection cycle per week, that would mean the viruses had been passed 52 times per year or 416 to 520 passages in chickens.
"As such they do not generally represent 'North American isolates of H7,' which would include hundreds if not thousands of wild bird H7s perhaps more typical of the rhea [North Carolina] isolate. It would be more correct to confine the title, discussion, and conclusions to human isolates of H7 and high-chicken-passage isolates of H7. Those findings are important to our understanding of how avian influenza viruses might become infectious for humans, but the findings do not necessarily apply to North American isolates in general."
Research on H7 vaccines
In view of the possible threat posed by H7 viruses, the National Institute of Allergy and Infectious Diseases (NIAID) is sponsoring several studies of human H7 vaccines, according to NIAID Director Dr. Anthony S. Fauci.
A phase 1-2 clinical trial of a vaccine based on a US H7N7 virus was launched in March, Fauci told CIDRAP News today. The trial, based at Baylor School of Medicine in Houston, involves 125 volunteers who received doses ranging from 7.5 to 90 micrograms of antigen to study the vaccine's safety and immunogenicity. The egg-based vaccine was made by Sanofi, he said.
In addition, the NIAID recently conducted an intramural phase 1 clinical trial of a cold-adapted H7N3 vaccine made from the British Columbian strain, Fauci reported. He said the results show that the vaccine is safe, but the immunogenicity findings are still being analyzed.
Fauci said some additional research on H7 vaccines is under way in NIAID labs in Bethesda, Md. "The bottom line is there is stuff going on," he said.
Belser JA, Blixt O, Chen L, et al. Contemporary North American influenza H7 virses possess human receptor specificity: implications virus transmissibility. Proc Nat Acad Sci 2008 May 27;105(21)7558-63 [Full text]
Jan 14 CIDRAP News story "Study refines view of H5N1 virus's binding preferences"