Comparison of four spatial maze learning tests with methylnitrosourea-induced microcephaly rats

The usefulness of four spatial maze learning test methods, single T-maze, Biel water maze, Morris water maze, and radial eight-arm maze, was compared using rats with different degrees of microcephaly, which were induced by single intraperitoneal injection with 3 and 5 mg/kg of methylnitrosourea (MNU) on day 13 of gestation. The single T-maze test did not detect a learning defect in either MNU group. In the Biel water maze test, the swimming time and number of errors were increased in the 5 mg/kg group, but these parameters were comparable to the control values in the retention and path B (reverse course) tests. On the Morris water maze task, the 5 mg/kg group required longer swimming time and distance to reach the goal. In the retest with the goal set on an opposite side, the swimming distance in the 5 mg/kg group was as long as the value in the original test. The radial eight-arm maze test detected fewer correct choices in both MNU groups, which showed different responses on this task. The brain weights in the 3 and 5 mg/kg groups were approximately 80 and 60% of the control value, respectively. The cerebral cortex weights were 77 and 50% of the control value, and the hippocampus weights were 78 and 51%. Among the four maze tests, only the radial eight-arm maze test could detect the effect of both doses of MNU. This shows that the method has the highest sensitivity to spatial learning defect in MNU-induced microcephalic rats.