VALIDATION OF AUTOMATED BEHAVIORAL-TEST SYSTEMS

A positive control study was conducted as parr of the ongoing validati on program for developmental neurotoxicity testing in our laboratory u sing a standard battery of automated systems, consisting of rotorod, m otor activity, acoustic startle, and two-way active avoidance. Female Sprague-Dawley rats were given 10 mg/kg diazepam (DZ) by SC injection or 20 mg/kg methimazole (MET) by gavage from gestation day 15 (DZ) or 17 (MET) through postpartum day 10; a group of control animals remaine d untreated. Offspring were assessed for growth, survival, development al landmarks, and behavior. Although this study was considered useful for obtaining historical data, it offered few advantages in terms of v alidation of automated behavior test systems. Perinatal treatment with DZ resulted in no maternal toxicity and no adverse effects on growth or development of F-1 offspring; a deficit in acoustic startle respond ing was the only behavioral effect observed. Treatment with MET result ed in maternal toxicity, reduced neonatal body weights, and developmen tal delays. Behavioral effects included impaired rotorod performance a nd acoustic startle responding (neonates), and enhanced motor activity and acoustic startle responding (young adults). However, effects on s huttle avoidance were not observed for either drug, and only one direc tion of behavioral effect occurred for the rotorod and motor activity systems. These results, as well as those from subsequent studies in ou r laboratory, suggest that it may be preferable to validate automated behavior systems using short-term studies in which young adult animals are treated directly with positive control agents.