Modeling and predicting selected immunological effects of a chemical stressor (3,4-dichloropropionanilide) using the area under the corticosterone concentration versus time curve

Many chemicals and drugs can induce a neuroendocrine stress response that c an be immunosuppressive. Mathematical models have been developed that allow prediction of the immunological impact of such stress responses in mice on the basis of exposure to the important stress-related mediator corticoster one. The area under the corticosterone concentration vs. time curve (AUC) h as been used as an indicator of cumulative corticosterone exposure in these modeling studies, In the present study, an immunotoxicant known to induce a stress response, 3,4-dichloropropionanilide (propanil), was evaluated to determine if corticosterone AUC values are related to suppression of immuno logical parameters in mice treated with this chemical. Linear relationships between corticosterone AUC values and suppression of the following paramet ers were noted in B6C3F1 female mice: thymus cellularity and thymus subpopu lation percentages, splenic subpopulation percentages, natural killer cell activity, MHC class II protein expression, and IgG1 and IgG2a antibody resp onses to antigen. Linear models derived in previous studies using mice trea ted with exogenous corticosterone or with restraint stress effectively pred icted the immunological effects of 3,4-dichloropropionanilide on the basis of corticosterone AUC values. The models derived using immobilization stres s were more effective (r(2) for observed vs. predicted = 0.90) than the mod els derived using mice treated with exogenous corticosterone (r(2) for obse rved vs, predicted = 0.65). This was expected, because most stressors induc e a variety of immunomodulatory mediators, not just corticosterone. These f indings have implications for risk assessment in immunotoxicology.