The susceptibility of mice to bacterial endotoxins.

Albino mice (Rockefeller NCS strain) raised and maintained free of ordinary bacterial pathogens, as well as of intestinal Escherichia coli and of Proteus bacilli, were found to be highly resistant to the lethal effect of bacterial endotoxins. When newborn mice of this NCS colony were nursed by foster mothers from another colony raised under ordinary conditions (SS colony from which the NCS colony was derived), they acquired the intestinal flora of the latter animals and became susceptible to the lethal effects of endotoxins. NCS adult mice could be rendered susceptible to the lethal effect of endotoxins by vaccination with heat killed Gram-negative bacilli. The susceptibility thus induced exhibited a certain degree of specificity for the bacterial strain used in vaccination. Although untreated NCS mice were resistant to the lethal effect of endotoxins, they proved exquisitively susceptible to the infection-enhancing effect of these materials. For example, 1 microg. or less of endotoxin was found sufficient to help establish a rapidly fatal septicemia with Staphylococcus aureus. Small amounts of endotoxin (1 microg. or less), administered alone, caused a marked but transient loss of weight. Vaccination with heat-killed Gram-negative bacilli or with killed BCG increased the resistance of NCS mice to the infection-enhancing effect of small amounts of endotoxin. This protective effect exhibited a certain degree of specificity for the bacterial strain from which the toxin used in the infection-enhancing test was derived. These various findings can be explained by assuming that the pathological effects of endotoxins involve at least two unrelated mechanisms; (a) a primary toxicity illustrated in this study by the loss of weight and enhancement of infection resulting from the injection of small doses of toxin; (b) an immunological reaction with lethal consequences which became manifest only in animals sensitized to the endotoxin by prior exposure to Gram-negative bacilli.