RABBIT CEREBELLAR SLICE ANALYSIS OF LONG-TERM DEPRESSION AND ITS ROLEIN CLASSICAL-CONDITIONING

Cerebellar long-term depression (LTD) has been proposed as a mechanism underlying classical conditioning of the rabbit nictitating membrane/ eyelid response (NMR). However, LTD has only been obtained reliably wh en (1) cerebellar slices are bathed in GABA antagonists which abolish disynaptic inhibitory post synaptic potentials, and (2) the temporal s equence of stimulation used in slice or intact preparations is the opp osite of that used in classical conditioning. Based on intradendritic Purkinje cell recordings obtained from rabbit cerebellar slices, we re port that stimulation of climbing fibers and then parallel fibers in t he presence of the GABA antagonist, bicuculline, produced significant depression of parallel fiber excitatory post synaptic potential (epsp) amplitude that continued to increase for at least 20 min after stimul ation. However, application of the same stimulation protocol without G ABA antagonists produced a brief depression of parallel fiber epsps th at disappeared within minutes. Activation of parallel fibers and then climbing fibers in an order opposite to the LTD-producing sequence (i. e. a classical conditioning-like order) produced a brief depression th at dissipated quickly. Stimulation of parallel fibers alone produced a small, slowly developing potentiation, but stimulation of parallel fi bers during depolarization-induced local dendritic calcium spikes prod uced significant depression almost immediately which then declined slo wly to more modest levels. Finally, stimulation of parallel fibers at frequencies used in in vivo parallel fiber-climbing fiber stimulation experiments (e.g. 100 Hz) produced an immediate and profound long-last ing epsp depression. The depression occurred, however, whether paralle l and climbing fibers were stimulated separately (unpaired) or in a cl assical conditioning-like protocol (paired) where parallel fiber stimu lation coterminated with climbing fiber stimulation (10 Hz). The depre ssion observed in both cases was reminiscent of transmitter depletion. Thus, despite obvious differences between in vitro and in vivo prepar ations, the present intradendritic Purkinje cell recordings in a rabbi t cerebellar slice do caste some doubt on the hypothesis that LTD func tions as a mechanism underlying classical conditioning of the rabbit N MR.