"We found the virus is doing what viruses do. It's mutating," said Pardis Sabeti of Harvard University and the Broad Institute, who led the massive study of samples from 78 people in Sierra Leone, all of whose infections could be traced to a faith healer whose claims of a cure attracted Ebola patients from Guinea, where the virus first took hold.
The findings, published in Science, suggest the virus is mutating quickly and in ways that could affect current diagnostics and future vaccines and treatments, such as GlaxoSmithKline's Ebola vaccine, which was just fast-tracked to begin clinical trials, or the antibody drug ZMapp, being developed by California biotech Mapp Biopharmaceutical.
The findings come as the World Health Organization said that the epidemic could infect more than 20,000 people and spread to more countries. A WHO representative could not immediately be reached for comment on the latest genetic study.
Study coauthor Robert Garry of Tulane University said the virus is mutating at twice the rate in people as it was in animal hosts, such as fruit bats.
Garry said the study has shown changes in the glycoprotein, the surface protein that binds the virus to human cells, allowing it to start replicating in its human host. "It's also what your immune system will recognize," he said.
In an unusual step, the researchers posted the sequences online as soon as they became available, giving other researchers early access to the data.
Erica Ollmann Saphire of the Scripps Research Institute in La Jolla, California, has already checked the data to see if it impacts the three antibodies in ZMapp, a drug in short supply that has been tried on several individuals, including the two U.S. missionaries who contracted Ebola in Sierra Leone and who have since recovered.
"It appears that they do not (affect ZMapp)," said Saphire, who directs a consortium to develop antibody treatments for Ebola and related viruses. But she said the data "will be critical to seeing if any of the other antibodies in our pool could be affected."
Saphire said the speed with which Sabeti and colleagues mapped genetic changes in the virus gives researchers information that "will also be critical" to companies developing RNA-based therapeutics.
That could impact treatments under way from Vancouver-based Tekmira Pharmaceuticals Corp and privately held Profectus BioSciences of Tarrytown, New York.
Part of what makes the data useful is the precise picture it paints as the epidemic unfolded. Sabeti credits years of work by her lab, colleagues at Tulane and the Sierra Leone Ministry of Health and Sanitation in developing a response network for Lassa fever, a virus similar to Ebola that is endemic in West Africa.
Several of the study authors gave their lives to the work, including Dr Sheik Humarr Khan, the beloved "hero" doctor from the Kenema Government Hospital, who died from Ebola.
The team had been doing surveillance for two months when the first case of Ebola arrived from Guinea on May 25. That case involved a "sowei" or tribal healer, whose claim of a cure lured sick Ebola victims from nearby Guinea.
"When she contracted Ebola and died, there were a lot of people who came to her funeral," Garry said. One of these was a young pregnant woman who became infected and traveled to Kenema Government Hospital, where she was diagnosed with Ebola.
With the Lassa surveillance team in place, they quickly began testing samples.
"We've been able to capture the initial spread from that one person and to follow all of these contacts and everything with sequencing," Garry said.
The team used a technique called deep sequencing in which sequences are done repeatedly to generate highly specific results, allowing them to see not only how the virus is mutating from person to person, but how it is mutating in cells within the same person.
What is not clear from the study is whether the mutations are fueling the epidemic by allowing the virus to grow better in people and become easier to spread. That will require further tests in the lab, Garry said.