ACUTE INFECTION OF MICE WITH HIGHLY VIRULENT GROUP-B STREPTOCOCCI AS A HOST-RESISTANCE MODEL FOR IMMUNOTOXICITY ASSESSMENT

This report describes a unique model for immunotoxicity evaluation in mice. The model is adapted from previously described mouse models for group B streptococcus (GBS) infections in human neonates. In this disease as well as a number of human diseases caused by highly virulent pathogens, the mechanisms of innate immunity are unable to protect the host, and survival is strictly dependent on acquired immunity. Unlike other host resistance models widely used in immunotoxicity studies, the GBS model utilizes bacteria that are highly virulent for mice (LD50 = 5-17 colony forming units). GBS is not virulent for adult humans and can be safely handled with typical precautions. Acquired immunity in the GBS model is induced during a 2 week period by two injections of heat-killed GBS. The immunizing doses are the minimum which will allow survival of 80-100% of mice in response to challenge with an otherwise lethal dose of live GBS (100 bacteria). Administration of the immunotoxic agents cyclophosphamide, carrageenan, or cobra venom factor during the immunization period and/or shortly before challenge significantly suppressed host resistance. For example, the composite mortality rate for unimmunized mice was 98% and the rate for immunized mice was 8.5%. For all groups treated with cyclophosphamide (one 75 mg/kg dose 48 h before each immunization) the mean mortality was 41 +/- 18%. The consistency of the model was evaluated by repetition of several treatments in independent experiments, and the model's consistency is comparable to that of other host resistance models. Finally, bacterial concentrations in the blood or peritoneum 16 h after challenge were evaluated as possible endpoints for the GBS model, and GBS concentration in the peritoneum seems to be a reliable indicator of immunotoxicity. Measurement of GBS concentration was comparable to survival time in discerning statistically significant differences between groups, and both of these parameters were superior to per cent mortality, particularly when small groups (six mice per group) were used.