ANN ARBOR—Researchers at the University of Michigan and the University of Freiberg in Germany have demonstrated that blocking a substance the body normally produces to help fight bacteria shows great promise in animal studies as a potential treatment for the highly lethal disease, sepsis.
The findings are published in the July 1999 issue of Nature Medicine.
Sepsis is a bacterial infection of the blood or body tissues that kills half of the hundreds of thousands of people who get the disease each year in the United States.
Investigators found that when they blocked C5a, a protein product that is normally activated by the immune system, they were able to provide a significant degree of protection in sepsis-induced rats. C5a is a product of the mammalian complement system—a complex system of proteins and plasma that has powerful immune functions. C5a normally binds to certain white blood cells, called neutrophils, stimulating them to kill bacteria.
Earlier studies had shown that in sepsis an excess of C5a is produced, causing the neutrophils to become overstimulated and paralyzing many of their defensive functions. In essence, sepsis causes the body’s own immune system to disarm its ability to fight the disease.
Researchers, led by Peter A. Ward, M.D., chair of the Department of Pathology in the University of Michigan Health System, found that septic rats injected with an immunoglobulin (IgG) antibody against C5a had powerful protection against sepsis. In the group of rats treated with the antibody against C5a, more than half survived. By contrast, in a second group treated with standard immunoglobulin, only 5 percent survived.
“This study reveals that in experimental sepsis, the system that produces protective mediators is overstimulated, leading to profoundly harmful consequences,” says Ward. “If this is also the case in humans with sepsis, there could be obvious implications for a new therapeutic approach for the treatment of sepsis.”
Researchers induced sepsis in rats by making a small perforation with a needle in the cecum (large intestine). The procedure—called cecal ligation and puncture, or CLP—causes a slow leak of the contents of the bowel, resulting in bacteria such as E. coli escaping into the body cavity. This model of sepsis and its physiological changes are very similar to those seen in human sepsis.
Three different groups of rats were studied. The first was treated with preimmune IgG immediately after CLP, while a second group received IgG antibody against C5a at the same time. In the group receiving preimmune IgG, only two of 21 rats survived more than a week, while half of the group treated with IgG antibody to C5a survived past 10 days. A third group had a different immune protein (C3) depleted and all animals died within five days.
Finally, Ward and his team tested the hypothesis that the reason bacteria counts in the antibody group were so low was because the therapy preserved the capacity of neutrophils to produce hydrogen peroxide—a vital product in the ability of white blood cells to kill bacteria. They took blood samples and stimulated the neutrophils to produce hydrogen peroxide. The neutrophils taken from rats treated with antibody to C5a produced high levels of hydrogen peroxide while those taken from the preimmune IgG group showed very little production.
“It appears, at least in this experimental model, that sepsis causes activation of the complement system,” says Ward. “You the get sustained, high levels of C5a generated, which bind to the circulating neutrophils and cause their de-activation. They then lose ability to kill bacteria because their ability to generate hydrogen peroxide is suppressed.”
Ward says the next steps are for these findings to be verified in other models of sepsis and in other non-human animals. Clinically, he says, the first thing that should be done is human trials that do not involve interventions. This would include taking blood samples, looking for C5a presence on neutrophils and at what point they peak, and trying to determine whether neutrophils in human sepsis patients are defective in their ability to produce hydrogen peroxide.
The study was made possible by a grant from the National Institutes of Health.
Further information about Ward’s Research Laboratory, along with other information on the University of Michigan’s Department of Pathology, can be found on the Internet at www.pathology.med.umich.edu
Or

Nature MedicinePeter A. WardNational Institutes of HealthResearch Laboratory