Dendritic excitability microzones and occluded long-term depression after classical conditioning of the rabbit's...

We made intradendritic recordings in Purkinje cells (n = 164) from parasaggital slices of cerebellar lobule HVI obtained from rabbits given paired presentations of tone and periorbital electrical stimulation (classical conditioning, n = 27) or explicitly unpaired presentations of tone and periorbital stimulation (control, n = 16). Purkinje cell dendritic membrane excitability, assessed by the current required to elicit local dendritic calcium spikes, increased significantly in slices from animals that received classical conditioning. In contrast, membrane potential, input resistance, and amplitude of somatic and dendritic spikes were not different in slices from animals given paired or explicitly unpaired stimulus presentations. The location of cells with low thresholds for local dendritic calcium spikes suggested that there are specific sites for learning-related changes within lobule HVI. These areas may correspond to learning "microzones" and are consistent with locations of learning-related in vivo changes in Purkinje cell activity. Application of 4-aminopyridine, an antagonist of the rapidly inactivating potassium current IA, reduced the threshold for dendritic spikes in slices from naive animals to levels found in slices from trained animals. In cells where thresholds for eliciting parallel fiber-stimulated Purkinje cell excitatory postsynaptic potentials (EPSPs) were measured, levels of parallel fiber stimulation required to elicit a 6-mV EPSP as well as a 4-mV EPSP (n = 30) and a Purkinje cell spike (n = 56) were found to be significantly lower in slices from paired animals than unpaired controls. A classical conditioning procedure was simulated in slices of lobule HVI by pairing a brief, high-frequency train of parallel fiber stimulation (8 pulses, 100 Hz) with a brief, lower frequency train of climbing fiber stimulation (3 pulses, 20 Hz) to the same Purkinje cell. Following paired stimulation of the parallel and climbing fibers, Purkinje cell EPSPs underwent a long-term (> 20 min) reduction in peak amplitude (-24%) in cells (n = 12) from animals given unpaired stimulus presentations but to a far less extent (-9%) in cells (n = 20) from animals given in vivo paired training. Whereas 92% of cells from unpaired animals showed pairing-specific depression, 50% of cells from paired animals showed no depression and in several cases showed potentiation. Our data establish that there are localized learning-specific changes in membrane and synaptic excitability of Purkinje cells in rabbit lobule HVI that can be detected in slices 24 h after classical conditioning. Long-term changes within Purkinje cells that effect this enhanced excitability may occlude pairing-specific long-term depression.