Functional properties of the CNS (mostly with reference to learning and memory capabilities) should be studied in awake, freely moving animals. An attempt was made in our laboratory to study motor learning and memory capabilities of wild and transgenic mice with deficits in those functions (demential states, experimental models of Alzheimer’s disease, etc.). The classical conditioning of the eyelid response in mice was used for this aim. Reproducible surgical and implantation procedures were developed in order to standardize this preparation.

Experiments were carried out following EU and Spanish guidelines and, at the end of them, animals were humanely killed. A total of four 50 µm stainless-steel wires (wholly isolated except at the very tip) were implanted in the upper lid under deep anaesthesia (ketamine, 50 mg kg^-1, I.P.) and soldered to a 4-pin connector. Two of the electrodes were directed to the supraorbitary branch of the trigeminal nerve and the other two to the orbicularis oculi muscle. Conditioning sessions consisted of 60 paired presentation of a conditioned stimulus (CS, a 50 µs square, cathodic pulse, 1.5 X threshold) followed, 250 ms later, by an unconditioned stimulus (US, a 0.5 ms square, cathodic pulse, 3 X threshold). The CS-US pair of stimuli was presented every 30 ± 5 s. At the end of each session, five CS were presented alone. Also, CS was presented alone for whole habituation and extinction sessions. Until now, we have used C57 and Swiss mice. We did two habituation, ten conditioning and four extinction sessions. Also, we have carried out both fast (an extinction session per day) and slow (and extinction session every ten days) extinctions.

The results for control animals were as follows. (i) The EMG of the orbicularis oculi muscle proved to be a good parameter to identify the presence of true conditioned responses. (ii) Maximum percentage of responses (70-80 %) was reached for both strains by the 7th conditioning session. (iii) The learning curve (determined as mean percent of conditioned responses per session) for C57 rose faster than for Swiss mice, but both reached a similar learning asymptote after a few conditioning sessions. Differences between both learning curves were not statistically significant.

At present we are repeating the same experimental procedure on different transgenic mice that model Alzheimer’s disease. Thus C57BL76J adult male mice are being used to evaluate the effects of the overexpression of amyloid precursor protein (APP) and presenilin-1 (PS-1) on the acquisition of classically conditioned eyelid responses.

This work was supported by Aventis-Pharma, La Caixa, Junta Andalucía and DGICYT (BFI2002-00936) grants. We thank Mrs M. Sutil for her co-operation in mice conditioning.