Figure 6.2 Kv1.2 channels prolong first spike latency and slow repetitive discharge. (a) The black line shows the voltage response of a medium spiny neuron recorded in a slice to a near-rheobase current injection. The gray line denotes the simulated Kv1.2 current evoked by the same voltage trajectory. Note that the Kv1.2 current deactivated slowly and remained activated during inter-spike period. (b) The same data as in (a) at higher gain to illustrate the Kv1.2 channel currents during the inter-spike period. (c) After bath application of 1 pM a-DTX, the first spike latency of the medium spiny neuron shown in (a) was shortened, the fast afterhyperpolarization reduced, and the repetitive discharge rate increased, as suggested by the Kv1.2 model. (d) Action potentials (APs) evoked by a symmetric 4 second long current ramp shown at the top of (d) in the same neuron before and after application of a-DTX (1 pM). Note that a-DTX shortened the latency to the first AP and increased the repetitive firing.

acetylcholine (Shen et al. 2004). Although they activate in this subthreshold voltage range, KCNQ channels differ in two important respects from Kv4 and Kv1.2 channels: (a) they do not inactivate, and (b) they activate (and deactivate) much more slowly than these other channels. Blockade of KCNQ channels depolarizes neurons held at UP state potentials and accelerates the discharge in medium spiny neurons in response to slow current ramps.

Do these channels contribute to regulation of UP state spiking in medium spiny neurons? As originally demonstrated by Wilson and Kawaguchi (1996), blockade of K+ channels with Cs+ has an enormous impact on the UP state potential. This led them to suggest that K+ channels were the key intrinsic factor governing the UP state membrane potential distribution. Subsequently, Wickens and Wilson (1998) have suggested that the difference between medium spiny neurons that spike and those that do not during UP state transitions in vivo (anesthetized rats) is that the UP state membrane potential of those that do is a few millivolts more positive than those that do not. Given the importance of Kv1.2 and KCNQ channels in this voltage range, it is easy to speculate that the modulation state of either channel (see below) could be responsible for the difference between spiking and not.

Do channels that give rise to inward currents play a role in the UP state? As the membrane potential moves above ca. -65 mV, voltage-dependent Na+ channels begin to activate providing an inward, depolarizing current that will speed the membrane potential movement toward the UP state plateau near -55 mV. These currents are attributable to a mix of Na+ channels with pore-forming sub-units of the Nav1.1, Nav1.2, andNav1.6 classes. Channels with these subunits give rise to both persistent and transient currents. However, the relative contribution of each channel to each component of the macroscopic current may vary from one cell type to another for as yet unknown reasons. For example, in cortical pyramidal neurons, roughly two thirds of the persistent Na+ current seen in somatic/proximal dendritic recordings is attributable to Nav1.6 channels. This may turn out to be significant for several reasons. One is that the subcellular distribution of these channel types does not appear to be uniform (Gong et al. 1999; Surmeier, unpublished). Nav1.1 channels are largely somatic, whereas Nav1.2 and Nav1.6 channels are more widely distributed. Another potentially important factor is phosphoregulation. Nav1.6 channels lack critical protein kinase A (PKA) phosphorylation sites, making them insensitive to modulation by this protein kinase, which is a major effector of D1 receptor signaling. D2 and M1 muscarinic receptor stimulation of protein kinase C (PKC), on the other hand, should modulate all forms of the Na+ channel alpha subunit.

There is no question that voltage-dependent Na+ channels shape the membrane potential trajectory in response to depolarizing current injection delivered to the soma, but do they contribute to the UP state transition and maintenance of the UP state? Wilson and Kawaguchi (1996) claim not. They found that UP states recorded in an urethane-anesthetized rat were unaffected by somatic loading of QX-314 through a sharp electrode. QX-314 blocks Na+ channels at sufficiently high concentrations, although not selectively. This result is a bit perplexing and is now being challenged by studies in in vitro preparations. In organotypic slice cultures, Na+ channels in dendritic regions of medium spiny neurons are capable of supporting backpropagating spikes well into distal regions (Kerr and Plenz 2002), suggesting that orthodromically initiated events should be capable of generating slow regenerative events in dendrites (as in cortical pyramidal neurons). In corticostriatal slices, state transitions resembling those seen in vivo can be produced under certain conditions (Vergara et al. 2003). An example of these transitions is shown in Figure 6.3. Surprisingly, the medium spiny neuron membrane potential distributions found in this preparation are bimodal and strongly resemble those found in anesthetized rodents. For comparison, a histogram taken from in vivo work by Wickens and Wilson (1998) is shown alongside that taken from the slice preparation.

In contrast to what has been found in vivo, plateau potentials (UP states) generated in the slice preparation are blocked by QX-314 delivered with a somatic patch pipette (unpublished observations). There are two obvious ways in which the differences between the in vitro and in vivo results can be explained. One is that because QX-314 was delivered with a sharp electrode, it was not present at high enough concentrations in the dendritic region to block Na+ channels effectively in the Wilson and Kawaguchi experiments. Another is that the mechanisms governing dendritic electrogenesis in the slice and in the anesthetized rat are different. This difference could stem from a GPCR-mediated suppression of dendritic Na+ channel opening in anesthetized animals, eliminating them as a significant factor in the response to synaptic stimulation and dendritic electro-genesis. Another possibility is that the degree of afferent synchrony achieved in the anesthetized preparation masks any contribution of intrinsic mechanisms to dendritic electrogenesis that might occur in other circumstances.

Another depolarizing influence at voltages near the UP state stems from voltage-activated Ca2+ channels. Mature medium spiny neurons express Cav1.2, Cav1.3, Cav2.1, Cav2.2, and Cav2.3 voltage-activated Ca2+ channels. The Cav2 family channels (N-, P/Q- and R-types) are activated only at membrane potentials achieved during spiking. Ca2+ entry through these channels during spiking shapes repetitive activity by opening nearby Ca2+-dependent K+ channels of the SK class (Vilchis et al. 2000). Another high voltage-activated Ca2+ channel opened only by spiking is the L-type Cav1.2 (C-type) channel. This channel is commonly a sensor of somatic spiking that helps control activity-dependent gene transcription (Bading et al. 1993). Medium spiny neurons express high levels of another L-type Ca2+ channel (Cav1.3 L-type Ca2+ channels) that activate at significantly more negative membrane potentials (ca. 15 mV) than do the other Ca2+ channels. Recent work suggests that these channels are preferentially located at excitatory synaptic sites in medium spiny neurons. Cav1.3 channels appear to interact with the synaptically targeted scaffolding protein Shank

(a) Slice recording

0 0

Post a comment