Comparison of effects of global cerebral ischaemia on spatial learning in the standard and radial water maze: relationship of hippocampal damage to performance

Abstract

Groups of rats which had undergone global ischaemia for 10, 15 or 20 min using the four-vessel occlusion technique were compared with sham-operated controls on learning to locate a submerged platform in both acquisition and working memory tasks in a standard Morris water maze, and in a working memory task in an eight-channel water radial maze. Ischaemic rats showed duration-related impairments in all three tasks. The water radial maze task was learned more slowly than standard water maze tasks, but deficits were long-lasting. In the first phase of training in the radial water maze controls were more reluctant than ischaemic rats to visit all arms of the maze, and were subsequently found to spend less time on the open arms of an elevated plusmaze. However, differences in anxiety are not likely to account for differences in working memory performance in the radial water maze, as groups showed similar error rates before and after habituation to the maze. Histological examination showed that cell loss occurred chiefly in the CAI field of the hippocampus and was linearly related to duration of occlusion. Cell loss was significantly correlated with the extent of impairment, but the pattern of relationships varied across the different tasks. For water maze acquisition, deficits in latency, heading angle and time spent in the training quadrant related more strongly to CAI than CA3 cell loss, but radial water maze impairments showed the reverse tendency. In all cases correlations were substantially reduced following exclusion of rats with maximal CAI cell loss, although a modest relationship with CAI damage remained for latency in acquisition and working memory tasks, and heading angle on the probe trial. These results suggested that relationships between water maze impairments and cell loss are robust only after near total destruction of the dorsal CAI field.