Read has also worked on the evolution of bacteria that resist antibiotics. In a provocative paper published last year, he suggested that typically prescribed doses of antibiotics can sometimes favor resistant strains. Smaller doses, which allow the immune system to kill resistant bacteria, could prove a better long-term approach.
In infectious disease, individual bugs compete within a host, as do individuals in any population of organisms. If a drug or vaccine doesn't completely wipe out or prevent infection, the intervention can tip the evolutionary playing field and possibly favor not only bugs that resist that particular drug, but ones that spread faster, or cause more severe symptoms, Read said.
The malaria findings, published in last week's issue of the journal PLoS Biology, are being taken seriously by vaccine researchers. "If there is an effect like this we need to be mindful of it and take steps to minimize this kind of impact," said Patrick Duffy, chief of the laboratory of malaria immunology and vaccination at the National Institutes of Health.
The findings don't apply to most existing vaccines, such as those used against smallpox, measles, and mumps. Those prevent people from ever getting infected, so there's no opportunity for the viruses to evolve in immunized people, Read said. But the malaria parasite is a more complex organism and no vaccine being tested now can wipe it out.
So far, the best scientists can do with vaccines is to stave off infection and ameliorate the symptoms, so that infected children are less likely to die. But vaccinated people still harbor the parasite, and, through mosquitoes, can spread it to others. Such imperfect vaccines are referred to as "leaky" because the pathogens can still multiply and evolve inside infected people. Other "leaky" vaccines may eventually be used to treat HIV.
Evolution may also play a role in how other diseases react to vaccines. Whooping cough is one of them, Read said, since the vaccine can wear off.
Researchers should also be vigilant about evolution of human papillomavirus (HPV), which causes cervical cancer, he said. The current vaccine doesn't protect against all strains of the virus and may allow nontargeted strains to become more virulent.
"We're moving into situations where parasites are evolving in the presence of immunized populations," he said. "We have to be sure we don't create situations where we're allowing hotter strains to spread."
While the malaria parasite that infects humans can't be given to laboratory mice or rats, these animals can be infected with a different version of the parasite that in nature infects African rodents called thicket rats.
For a vaccine in this mouse experiment, Read and collaborator Victoria Barclay chose a protein called AMA-1, which is the key component in several vaccines now in human trials in Africa. They allowed the parasite to spread through 10 different immunized mouse hosts. Then they let it infect unvaccinated mice.
This new malaria parasite made unvaccinated mice much sicker than malaria that hadn't evolved in vaccinated hosts. Read said they are still trying to figure out the specific mechanism by which the malaria parasites became more aggressive and virulent.
It's possible, he said, that the immune response prompted by the vaccine knocked out the weakest parasites, leaving the more aggressive, fast-spreading ones to grow without competition.
NIH's Duffy said the results will influence how they conduct human vaccine trials now going on in several African countries. In Read's experiment, it was the unvaccinated mice that got the worst disease down the road. If things play out the way they did for Read's experimental mice, he said, the vaccines might benefit some people but make the situation more dangerous for those who remain unvaccinated.
"We not only need to follow the people who we've given the vaccine to, but other people in the community," Duffy said.
Researchers are exploring a number of different tactics for vaccine development, Duffy said. They're targeting the parasite in different stages of its complicated life cycle, including the phase in which it lives inside the mosquito. Some of these stages may be less likely than others to allow the parasite to evolve.
He and Read agree that this new paper should increase awareness of evolution but not fuel irrational fears of vaccines. "Vaccines are still our best public health tool," Duffy said. "They are our most cost-effective way to improve public health."
Contact Faye Flam at 215-854-4977, firstname.lastname@example.org, or follow on Twitter @fayeflam. Read her blog at philly.com/evolution.