Caveats noted, but CDC can't make H5N1 transmit

Hybrid of avian, human strain created in lab

The world is on pandemic watch, with the prevailing public health fear that a currently circulating avian influenza A (H5N1) will mutate with a human strain of influenza and become transmissible from person to person. While that certainly remains possible, researchers at the Centers for Disease Control and Prevention have been unable to create such a hybrid strain in the lab.

Certainly good news, but Julie Gerberding, MD, MPH, director of the CDC, hesitates to find the findings reassuring.

"I'm cautious about using the word "reassuring,'" Gerberding said at a recent press conference on the CDC research. "I think this is very important work, and it's reassuring to know that we have a laboratory tool that will help us understand transmission dynamics with much more precision than we've ever been able to do before. But I'm not reassured from the public health perspective because this virus is still out there, it's still evolving. Flu is always unpredictable. Though we weren't able to [induce transmissibility] through some simple gene exchanges, there are many other combinations and subtle changes that the virus itself could make."

As of Aug. 14, H5N1 had infected 238 people throughout the world and 139 have died. There are cases suggesting some limited human-to-human transmission has occurred, but for all practical purposes avian flu remains a zoonotic infection transmitted by birds. The concern is that avian flu will become transmissible in humans by reassorting with a circulating human virus, a scenario that unfolded in the pandemics of 1957 and 1968.

"In general, these viruses evolve by two main mechanisms," Gerberding said. "In one case, they might make a set of genetic changes over time, becoming progressively more transmissible from one person to person. That's probably what happened in 1918 with the pandemic virus that caused influenza on a global scale back then. But the other way that viruses can become transmissible to people is by exchanging genes with strands that are already transmissible."

The CDC created a lab protocol using ferrets — a good surrogate for human infection — to assess potential transmissibility with a reassorted virus. CDC researchers took an H5N1 strain of virus and used reverse genetic techniques to combine it with a strain of H3N2, one of the most common human influenza viruses.

"Simple combinations of genes from both parent viruses have not led to enhanced transmissibility in the ferrets," Gerberding said. "These data do not mean that H5N1 cannot convert to be transmissible from one person to another person. They mean that it is probably not a simple process and more than simple genetic exchanges are necessary. That complexity is something that we will be working on in subsequent experiments at CDC. [These results] do not mean that H5N1 cannot develop into a pandemic strain. It means that the genetics of that transformation are more complicated than a simple one-to-one exchange."

The researchers used a specialized caging system designed to study the transmission through respiratory droplets. "We found that the human H3N2 influenza viruses transmitted efficiently using the special caging system whereby the animals were placed in close proximity, but such that only transmission through respiratory droplets could occur," explained Jacqueline Katz, PhD, an influenza researcher at the CDC and co-author of the published study.1 "When we looked at a number of H5N1 viruses we found that these viruses were not able to transmit efficiently. And so, this told us that the model was working and reflected the transmission that we see of these two viruses in humans at present."

Creating the hybrid

The next step was to generate reassorted viruses, so researchers created two viruses that contained the surface-like protein gene from the avian H5N1 virus and the internal genes from a human H3N2 virus. "But when we assessed the ability for transmission of these reassortant viruses, we found that they were not able to transmit efficiently," Katz said. "And, in fact, they were also not as able to cause severe disease as the original H5N1 virus."

Another aspect of the experiment was to assess whether genetic changes might occur if hybrid viruses were "passaged" from one ferret to another, she explained. "Would this give the virus a chance to acquire additional genetic changes that could increase its ability to transmit efficiently?" wondered Katz. "We conducted this experiment with one of the hybrid viruses and passed it from ferret to ferret through taking nasal secretions from one ferret and transferring it to the next. We found the virus did not acquire any additional capacity to transmit efficiently from an infected ferret to a healthy ferret."

Caveats abound — including that the published paper deals only with experiments using a 1997 strain of H5N1. However, at the press conference, Katz said subsequent experiments using more recent strains produced similar results. "We only did use a relatively small number of reassortants and there are many more combinations that are possible," Katz added. "Over 50 combinations are out there, so that would take years of work to go through all of those reassortants. "We chose to use some that had the greatest likelihood of being viable, a good virus that grew well and was viable, and therefore had the potential for transmission."

The research was done at the CDC in a biosafety Level 3 laboratory since the experiment certainly had the potential to create the very pandemic flu strain that may eventually emerge in nature. "We always use these precautions when we do any work with H5N1 virus," Katz said.


  1. Maines TR, Chen LM, Matsuka Y, et al. Lack of transmission of H5N1 avian-human reassortant influenza viruses in a ferret model. Proc Natl Acad Sci 2006; Aug 8;103(32):12,121-12,126.