Online

Fuchs, Jason R.

Psychology

2016

PhD

The neural mechanisms that support learning and memory are still poorly understood. Much work has focused on changes in neurotransmitter receptor expression, while changes in voltage-gated ion channel expression have been largely unexplored, despite the fact that voltage-gated ion channels govern neuronal excitability. Here we used eyeblink conditioning (EBC) in rats, a model of learning and memory with a well-understood neural circuit, to examine regulation of voltage-gated ion channels as a consequence of learning. EBC is a form of classical conditioning that involves pairings of a behaviorally neutral conditioned stimulus (CS) and an eyeblink eliciting unconditioned stimulus (US) over many trials to produce an eyeblink conditioned response (CR) to the CS in anticipation of the US. The acquisition and generation of the eyeblink CR is governed by plasticity at various sites in the cerebellum, both in the cerebellar cortex and the interpositus nucleus (IPN). Purkinje cells (PCs) are the primary neuron in the cerebellar cortex and these cells represent the sole output of the cerebellar cortex. PCs tonically inhibit the neurons of the IPN; the IPN is the start of the eyeblink pathway. In order for a CR to be generated, the inhibition of the IPN by PCs must be lifted. Basket cells (BCs) are small inhibitory interneurons that form synapses near the PC soma. These neurons are strategically located to strongly regulate PC output through inhibitory input near the axon hillock. BC axon terminals have the highest expression of Kv1.2, an alpha subunit of the Kv1 (Shaker) family of voltage-gated potassium channels, in the cerebellum. In addition, significant Kv1.2 expression is found on PC dendrites. Blocking Kv1.2 leads to increased GABAergic input to PCs and facilitates EBC. In the current work, we addressed the question of whether EBC itself regulates surface expression of Kv1.2 in cerebellar cortex. Rats received three days of either EBC, explicitly unpaired stimulus presentations, or no stimuli, and cerebellar tissue was harvested and analyzed via biotinylation/western blot (WB) and multiphoton microscopy (MP) techniques. In the first experiment, the Unpaired group showed significantly reduced surface Kv1.2 expression at BC axon terminals as measured by MP, but no changes observed with the WB measure, which measures expression at both BC axon terminals and PC dendrites. The second experiment used the same procedures but examined cerebellar tissue following a shorter training procedure. We hypothesized that the Paired and Unpaired groups would show similar Kv1.2 surface expression earlier in training. The Unpaired group showed increased surface Kv1.2 compared to the other two groups in the WB measures, but no differences were observed in the MP measure. Paired group rats that did not exhibit CRs showed the same pattern as the Unpaired group. Overall, we observed training and location specific changes in surface Kv1.2 expression, suggesting that learning does appear to regulate voltage-gated ion channel expression in the mammalian brain. Increased surface Kv1.2 early in training before CR expression emerges may set the stage for other mechanisms to govern the expression of the learned response. Prolonged stimulus input that is unmodulated by expression of a learned response, such as in the Unpaired group in the first experiment, leads to long-term changes in surface Kv1.2 expression exclusively at BC axon terminals.