To study the neural mechanisms of interhippocampal transfer of lateralized place navigation engrams in rats, lidocaine was injected via chronically implanted cannulae to reversibly inactivate the hippocampal formation and the visual cortex on one side. The eye opposite the blocked side was occluded. Under these conditions, rats learned the location of an invisible platform in a water maze [mean escape latencies per four-trial block(t) = 5-6 s at the performance asymptote]. Monocular intact brain retrieval with the trained eye (t = 7) was better than with the untrained eye (t = 13). However, analysis of each retrieval trial indicated untrained eye performance was only poor on the first trial (t = 30). To test whether trans-commissural read-out alone or write-in (i.e. interhippocampal transfer) of the lateralized engram explains the above results, rats acquired a new platform location (t = 5). Two groups were then given a 30-s ''free swim'' (the platform was removed) with intact brain and either the trained or untrained eye occluded. A third group did not have this ''transfer'' trial. Retrieval was tested with the trained hippocampus and visual cortex blocked. With the trained eye occluded, retrieval in the rats that had the transfer trial (t = 11) was better than in those that did not (t = 25), but slightly worse than in rats tested with the untrained eye, hippocampus and visual cortex blocked (t = 7). Additionally, retrieval was similar, independent of whether the trained (t = 12) or untrained (t = 11) eye was open on the transfer swim. The 30-s swim alone did not induce comparable savings. We conclude that interhippocampal transfer of lateralized place learning is easily induced, is equal if the transfer is facultative or imperative, and involves both trans-commissural read-out and write-in processes.
