A Cornell University virologist has isolated a highly contagious equine flu virus that is spreading a sometimes-fatal respiratory flu among dogs, and is responsible for a major dog-flu outbreak in New York state. There is no evidence that the virus could infect people.

According to a paper published in the Sept. 26 issue of Science Express (Vol. 309, No. 5743), the online version of Science magazine, this is the first time an equine flu virus has been found to jump species.

The equine influenza virus, H3N8, was isolated at Cornell's College of Veterinary Medicine after University of Florida researchers sent fluid and tissue samples from greyhound race dogs that had died from a respiratory illness at a Florida racetrack in January 2004.

"Of all animals, dogs have the most intimate contact with humans on a daily basis, so the potential for human infection has to be in the back of our minds," said Ed Dubovi, director of the virology center at Cornell's Animal Health Diagnostic Laboratory, who isolated the virus from the University of Florida samples.

Still, he added, there is no evidence of the virus jumping to humans, and there is no expectation of it doing so. It is possible the equine virus has been infecting dogs for some time, although the symptoms are very similar and could be mistaken for common "kennel cough," a bacterial disease related to pertussis (whooping cough) in children. Nevertheless, the paper cautions that the newly discovered flu virus must be closely monitored.

With close to 100 percent of dogs exposed to the virus becoming infected and about 80 percent of infected dogs showing symptoms, the flu could be spreading throughout the country. It was originally documented in greyhounds at tracks and kennels but now is infecting all breeds of dogs. Ongoing testing is being done to track the spread of the virus to different regions of the country.

"Right now, we have a major outbreak of this disease in all breeds of dogs in New York state," said Dubovi, noting that symptoms can include high fever and a respiratory infection that lasts a few weeks, although 1 to 5 percent die from related hemorrhagic pneumonia. From January to May 2005, outbreaks occurred at 20 racetracks in 10 states (Florida, Texas, Arkansas, Arizona, West Virginia, Kansas, Iowa, Colorado, Rhode Island and Massachusetts), according to the paper.

"This infection will become a major concern for all dog owners, since 100 percent of dogs are susceptible to infection by this virus ," said Dubovi. "With 50 million pet dogs in this country, even if you have 1 percent mortality, this is going to result in a number of dogs dying from it."

The Centers for Disease Control and Prevention (CDC) in Atlanta sequenced the virus' entire genome and found all the segments were from the equine virus. This is unusual, because flu viruses will often swap genetic material with other flu viruses when they jump species. For a new virus to enter another species it must overcome a number of barriers, such as finding a cell receptor to bind to in order to enter the cell and to reproduce sufficiently in the new host.

"There are probably many examples of viruses jumping species, but then it becomes a dead-end issue," said Dubovi. Researchers have long known that equine flu was capable of growing in dogs, since scientists experimenting with equine influenza use cell lines from canine kidney cells.

When Dubovi first received the University of Florida samples, he and his colleagues isolated a virus and determined that it was an influenza not typically found in dogs. The next step was to test to see if it was an avian flu virus, like the virulent H5N1 that has jumped from birds to humans over the last few years. A highly sensitive test -- called a PCR (polymerase chain reaction) that amplifies and detects small amounts of DNA or RNA in a blood or tissue sample -- ruled out avian flu strains H5 and H7.

The isolate was then sent to the CDC where Ruben Donis, Cornell Ph.D. '87, one of Dubovi's former graduate students and currently chief of molecular genetics for the influenza branch, found that it tested positive for equine influenza virus. Donis also ruled out the possibility that the sample had been contaminated with equine virus from another source.

The study by lead author, Cynda Crawford, an immunologist at the University of Florida's College of Veterinary Medicine, was funded by the State of Florida's Division of Parimutuel Wagering in response to ongoing respiratory problems in racing greyhounds.