Web edition: February 11, 2013
Print edition: March 23, 2013; Vol.183 #6 (p. 10)
Mice are poor stand-ins for people in experiments on some types of inflammation, a new study concludes. But some scientists say that critique discounts the value of mouse studies, and that many biomedical experiments simply couldn’t be done without the animals.
More attention — and money — should go toward studying disease in people than on mouse research, a consortium of scientists contends online February 11 in the Proceedings of the National Academy of Sciences. Too often, researchers make a discovery in mice and assume that humans will react in the same way, says study coauthor Ronald Tompkins, chief of the Massachusetts General Hospital burn service. “The presumption is not justifiable,” he says. As a result, drug trials — often based heavily on data gleaned from studies with mice — can fail.
But other scientists say that critique isn’t new and is overstated. Clinical trials are unsuccessful for many reasons, says Derry Roopenian, an immunologist and mouse geneticist at the Jackson Laboratory in Bar Harbor, Maine. “There’s frailty all along the process. That’s not a failure of the mouse.”
He and other critics worry that the study, conducted with a generic strain of laboratory mouse called Black6, unfairly tarnishes the reputation of all mice, even ones engineered to be as much like humans as possible. The group’s conclusions, were they accepted by policy makers, could set back biomedical research by jeopardizing funding for mouse studies, critics warn. “Without the mouse, progress is going to be slowed to a standstill,” Roopenian says.
Most of the researchers agree that creating mice with biologic responses that more closely mirror humans is important to understand diseases and develop new drugs. The sticking point appears to be how to balance mouse-based research with research involving humans.
In the new study, a group of Canadian and American researchers compared how human and mouse genes respond to certain types of trauma. They looked at gene activity in the blood of 167 people severely injured either by burns or by blunt trauma, such as in a car accident. They compared those results with gene activity in the blood of mice with similar injuries. The team also examined the reactions to an injection of endotoxin, a bacterial toxin, of four healthy people and 16 mice.
People’s immune systems reacted in predictable ways, cranking up activity in genes that cause inflammation and quieting other immune system genes, Tompkins says. When the researchers compared the gene activity changes in people with burns versus people with trauma, they found 97 percent similarity. For injuries and endotoxin exposure, 88 percent of genes responses were similar.
But mouse responses were more varied, says study coauthor Wenzhong Xiao, a genome scientist at Harvard University and Stanford University. Compared with what goes on in humans, 47 to 63 percent of mouse genes changed activity in the same way. The result was not far from what researchers would expect from random chance.
The results indicate that humans and mice react differently to traumas that often land people in intensive care units. “We need to take those differences into serious consideration, which people currently do not,” Xiao says.
No one expects mouse and human physiology to match exactly, Roopenian counters. And the differences between the species might even inform research. For instance, the study authors cite as a shortcoming that mice can withstand doses of endotoxin 1 million times as high as the dose that sends people into shock. But Roopenian sees that difference as an opportunity to learn how to boost human tolerance to the substance.
Researchers should do more of the kinds of genetic analyses performed in this study when choosing an appropriate mouse strain, says Klaus Schughart, a geneticist at the Helmholtz Center for Infection Research in Braunschweig, Germany. Schughart studies infectious diseases including influenza and uses mice as human substitutes. He sees benefits in the fact that mice don’t perfectly mimick complex human diseases.
“We purposely oversimplify,” he says, to strip away confounding factors and get at the basic mechanisms of disease. “We should value the mouse for what it’s worth and not condemn it for its shortcomings.”
Tompkins agrees that basic research should continue, but adds that “we need to recognize that its relevance to human disease is a leap of faith.”
J. Seok et al. Genomic responses in mouse models poorly mimic human inflammatory diseases. Proceedings of the National Academy of Sciences. Published online February 11, 2013. [Go to]
N. Seppa. Novel fused protein quells inflammation. Science News Vol. 172, December 22, 2007, p. 402. [Go to]
T. H. Saey. Inflammation feeds E. coli. Science News Online. February 7, 2013. [Go to]
N. Seppa. Experimental drug fends off emphysema in mice. Science News Vol. 175, January 17, 2009, p. 8. [Go to]