Trace eyeblink conditioning is hippocampally dependent in mice

The goal of this study was to determine whether trace eyeblink conditioning is a hippocampally dependent associative learning task in the mouse. First, we examined trace intervals of 0, 250, and 500 ms to determine a relatively long trace interval that would support eyeblink conditioning in young adult C57BL/6 mice. Mice rapidly acquired conditioned responses (CRs) with a 0-ms trace interval, acquired CRs with a 250-ms trace interval in approximately 2 days (2 sessions per day), and showed little acquisition with a 500-ms trace interval. Control mice were presented randomly unpaired stimuli and failed to show conditioning. We then determined the effect of lesioning dorsal hippocampal neurons on trace eyeblink conditioning. The hippocampus was injected bilaterally with vehicle (phosphate-buffered saline), 0.1% ibotenic acid, or 1% ibotenic acid. The vehicle group showed >60% CRs. The 0.1% group showed significantly fewer CRs (35-45%). The 1% group showed a level of CRs similar to that of the control mice. All the lesioned mice exhibited >60% CRs when subsequently trained with a 0-ms trace interval. A regression analysis indicated that the volume of area CA1 lesioned was more predictive of the behavioral impairment than the lesion volume of either CA3 or dentate gyrus, or even the total lesion volume. We conclude that dorsal hippocampal neurons play a critical role in eyeblink conditioning when a 250-ms trace interval is used with the C57BL/6 mouse, and that this paradigm will be useful for studying behavior and the in vivo and in vitro electrophysiology of hippocampal neurons in normal and transgenic or knockout mice.