Activity-dependent changes in extracellular Ca2+ and K+ reveal pacemakers in the spinal locomotor-related network

Abstract

Changes in the extracellular ionic concentrations occur as a natural consequence of firing activity in large populations of neurons. The extent to which these changes alter the properties of individual neurons and the operation of neuronal networks remains unknown. Here, we show that the locomotor-like activity in the isolated neonatal rodent spinal cord reduces the extracellular calcium ([Ca(2+)]o) to 0.9 mM and increases the extracellular potassium ([K(+)]o) to 6 mM. Such changes in [Ca(2+)]o and [K(+)]o trigger pacemaker activities in interneurons considered to be part of the locomotor network. Experimental data and a modeling study show that the emergence of pacemaker properties critically involves a [Ca(2+)]o-dependent activation of the persistent sodium current (INaP). These results support a concept for locomotor rhythm generation in which INaP-dependent pacemaker properties in spinal interneurons are switched on and tuned by activity-dependent changes in [Ca(2+)]o and [K(+)]o.

Figure 1. Variations of [Ca²⁺]_o and [K⁺]_o during Locomotor-like Activity

(A–D) Changes in [Ca²⁺]_o (A and C) and [K⁺]_o (B and D) in the locomotor CPG during (A and B) and at the onset (C and D) of NMA/5-HT induced locomotor-like activity in neonatal rat preparations. Raw (A and B) and rectified/integrated (C and D) traces of locomotor output recorded from opposite L5 ventral roots. Insets illustrate enlargements of the recordings (gray vertical line). (E–G) Time course of [Ca²⁺]_o and [K⁺]_o changes before and during locomotor-like activity in relation to amplitude (F) and frequency (G) of bursts. Error bars indicate SEM. Time 0: onset of locomotion. Dashed lines in (C)–(E) represent threshold of [Ca²⁺]_o and [K⁺]_o at the onset of locomotion. (H) Fictive locomotor episode induced by an aCSF composed of 0.9 mM [Ca²⁺]_o and 6 mM [K⁺]_o pressure applied over the CPG.

Figure 2. Changes in [Ca²⁺]_o and [K⁺]_o Precede and Trigger I_NaP-Dependent Bursting Properties

(A–E) Activity of a locomotor-related interneuron before (A), at the onset (B, enlargement of the gray area in A), and during (C) an NMA/5-HT-induced locomotorlike activity and 20 min (D) or 40 min (E) after the superfusion of riluzole (5 μM). (F) Voltage traces illustrate the effects of varying [Ca²⁺]_o or [K⁺]_o alone (within the range of fluctuations observed during locomotion) on the firing pattern of interneurons recorded in the locomotor CPG region of neonatal rats. (G) The threedimension histogram plots the proportion of interneurons exhibiting pacemaker properties (y axis) against [Ca²⁺]_o (x axis) and [K⁺]_o (z axis). Note that for each [K⁺]_o and [Ca²⁺]_o concentration in which bursts emerged, at least one bursting cell was tested for the sensitivity of their membrane oscillations to the blockade of I_NaP (by TTX or riluzole). (H) Activity of a pacemaker cell before, during, and after a brief (~3 min) application of TTX (0.1 μM). (I) Proportion of Hb9 interneurons exhibiting pacemaker properties (y axis) against [Ca²⁺]_o (x axis) and [K⁺]_o (z axis). (J and K) Firing patterns of two typical Hb9 cells when the [Ca²⁺]_o was decreased from 1.2 mM to 0.9 mM (J) or when [K⁺]_o was increased from 4 to 6 mM (K) while [K⁺]_o and [Ca²⁺]_o remained constant, respectively. For each cell, the leak-subtracted membrane current evoked by voltage ramp from −70 to −10 mV is plotted before (black trace) and after decreasing [Ca²⁺]_o or increasing [K⁺]_o(dark gray trace). A subsequent application of TTX (1 μM) abolished the current (light gray trace).

Figure 3. Involvement of I_NaP in Generation of the Pacemaker Bursting Activity: Insights from Computational Modeling

(A) Changes in I_NaP during slow voltage ramp for different values of V_mNaP1/2. (B) Firing pattern of neuron in the model at [K⁺]_o = 6 mM. At V_mNaP1/2 = −52 mV, the model exhibited tonic spiking activity (top); at V_mNaP1/2 = −53 mV, the neuron switched to bursting (middle), and further shifting V_mNaP1/2 to the left (−54 mV) increased spiking frequency within the bursts (bottom). (C) Dependence of the percentage of bursting cells on [Ca²⁺]_o and [K⁺]_o (in the modeled population of 50 uncoupled neurons with randomly distributed parameters).