Because this vaccine contains a live virus, it may be more immune-activating than avian flu vaccines prepared by traditional methods, say the researchers.

Furthermore, because it is grown in cells, it can be produced much more quickly than traditional vaccines, making it an extremely attractive candidate for preventing the spread of the virus in domestic livestock populations and, potentially, in humans, according to the study, published in the Feb 15 issue of the Journal of Virology and made available early online.

"The results of this animal trial are very promising, not only because our vaccine completely protected animals that otherwise would have died, but also because we found that one form of the vaccine stimulates several lines of immunity against H5N1," said Andrea Gambotto, M.D., assistant professor in the departments of surgery and molecular genetics and biochemistry, University of Pittsburgh School of Medicine, and lead author of the study.

Dr. Gambotto and his colleagues constructed the vaccine by genetically engineering a common cold virus, called adenovirus, to express either all or parts of an avian influenza protein called hemagglutinin (HA) on its surface. Found on the surface of all influenza viruses, HA allows the virus to attach to the cell that is being infected and is, therefore, critical to the influenza virus' ability to cause illness and death.

Since the late 1990s, a number of outbreaks of the avian influenza H5N1 in poultry have occurred in Cambodia, China, Indonesia, Japan, Laos, South Korea, Thailand and Vietnam. Outbreaks recently have been reported in Turkey and Romania. To date, H5N1 has caused the most large-scale and widespread bird deaths in known history--an estimated 150 to 200 million birds have either died in the outbreaks or been killed as part of infection control actions in the last eight years.

The H5N1 virus does not usually infect humans. However, in 1997, the first case of spread from a bird to a human occurred in Hong Kong during an outbreak of bird flu in poultry. The virus caused severe respiratory illness in 18 people, six of whom died. Since that time, more than 170 cases of known H5N1 infection have occurred among humans worldwide, approximately half of whom died.

Based on the published sequence of the Vietnam strain of the H5N1 avian influenza virus, members of the University of Pittsburgh Vector Core Facility, led by Wentao Gao, Ph.D., research instructor in the School of Medicine's department of surgery, constructed several adenovirus "vectors"--viruses that have been modified to serve as a vector, or delivery vehicle, for foreign genes or DNA--containing either the full genetic sequence of the HA protein or sequences for only parts, or subunits, of HA.

They also constructed a vector containing sequences for a portion of the HA protein from the H5N1 Hong Kong strain.

Collaborating with investigators Xiuhua Lu, Ph.D., Doan C. Nguyen, M.D., Yumi Matsuoka, Ph.D., Ruben O. Donis, Ph.D., and Jaquelin M. Katz, Ph.D., of the Influenza Branch of the Centers for Disease Control and Prevention, Dr. Gambotto's team tested the ability of their slightly different vaccines to protect mice from infection by wild-type H5N1 by comparing its performance to an adenovirus vector containing no H5N1 genes, or an "empty vector."

The investigators then observed the H5NI-exposed mice for any signs of illness, including weight loss and death, and also checked their blood for anti-viral antibodies and other markers of H5N1-specific immunity.

All of the mice immunized with the empty vector vaccine experienced substantial weight loss beginning about three days after exposure to wild-type H5N1, and all were dead within six to nine days of avian flu exposure. In sharp contrast, most of the mice immunized with the adenovirus containing either the whole or part of the HA protein showed only mild and short-lived weight loss and survived H5N1 infection.

When the investigators looked for evidence of a specific immune response to H5N1, they found similar results. Although they were able to isolate high levels of infectious H5N1 from multiple organs in the mice vaccinated with the empty vector, and to various degrees in animals vaccinated with the vectors containing the HA subunits, they isolated only very small amounts of H5N1 from the mice immunized with the full-length HA vaccine three days after infection. Six days after infection, they could not detect any infectious H5N1 in the organs of mice immunized with the full-length HA vaccine.

Moreover, when they looked at the cellular immune response to vaccination, they found that all of the animals immunized with full-length HA or the subunit vaccines developed strong cellular immune responses. However, only the full-length HA-immunized mice developed strong T-cell responses to both of the HA subunits.

According to Simon Barratt-Boyes, B.V.Sc., Ph.D., associate professor, department of infectious diseases and microbiology, University of Pittsburgh Graduate School of Public Health, and one of the co-authors of the study, the ability of this particular recombinant vaccine--a vaccine carrying only the important immune-stimulating proteins--to induce both antibody- and T cell-directed immunity is extremely encouraging.

"This means that this recombinant vaccine can stimulate several lines of defense against the H5N1 virus, giving it greater therapeutic value. More importantly, it suggests that even if H5N1 mutates, the vaccine is still likely to be effective against it. How effective, we are not sure," Dr. Barratt-Boyes cautioned. "We won't know until that occurs."

Based on the superior degree of protection that they found in mice vaccinated with full-length HA vaccine, Dr. Gambotto's group, working with David E. Swayne, D.V.M., Ph.D., at the U.S. Department of Agriculture, tested its effectiveness in chickens, which have almost a 100 percent mortality rate to H5N1 exposure.

In all, the researchers inoculated four groups of chickens either through their noses (intranasally) or with subcutaneous injections of either the HA-containing vaccine or the empty vector vaccine. The chickens were then challenged with a dose of whole H5N1 virus 10,000 times greater than the dose given to the mice and significantly greater than the dose farm chickens are likely to be exposed to during a natural outbreak.

Interestingly, all of the chickens that were immunized subcutaneously survived exposure to H5N1, developed strong HA-specific antibody responses and showed no clinical signs of disease. In contrast, half of the chickens immunized intranasally died and half survived.

All of the chickens immunized with the empty vector (intranasally and subcutaneously) died within two days of H5N1 exposure. The researchers are still not yet sure why the subcutaneous delivery is more effective than the intranasal delivery of the vaccine, but they suggested it may be because the adenovirus vector they used has limited infectivity via the nose and respiratory tract.

Dr. Gambotto and his colleagues suggest that rather than replacing traditional inactivated influenza vaccines, their adenovirus-based vaccine could be a critically important complement to them. Because it appears to be so successful in immunizing chickens against H5N1, widespread inoculation of susceptible poultry populations could provide a significant barrier to the spread of the virus via that route in this country and other countries that have so far been spared from avian flu.

Also, if there were a disruption in the traditional vaccine production pipeline, a recombinant vaccine could be an attractive alternative for human immunization as well, they said.

Indeed, according to Dr. Gambotto, there are several major advantages to this type of vaccine development approach over traditional approaches. Flu vaccines currently are prepared in fertilized chicken eggs, a process developed more than 50 years ago that requires millions of fertilized eggs that would be in short supply if a pandemic--a widespread, global outbreak--were to occur. The recombinant vaccine approach grows the vaccine in cell cultures, which are unlimited in supply. Another major advantage of this approach is its speed.

"It takes a little over a month for us to develop a recombinant vector vaccine compared to a minimum of several months via traditional methods," he explained. "This capacity will be particularly invaluable if the virus begins to mutate rapidly, a phenomenon that often limits the ability of traditional vaccines to contain outbreaks of mutant strains." Dr. Gambotto added that his group is planning a small clinical trial of the vaccine in humans in the very near future.

The research was supported by internal University of Pittsburgh funds. Others involved in this study include Paul D. Robins, Ph.D. and Angela Montecalvo, Ph.D., University of Pittsburgh School of Medicine; and Adam C. Soloff, B.S., University of Pittsburgh Graduate School of Public Health.

Related Links
University of Pittsburgh Medical Center