Participation of a chloride conductance in the subthreshold behavior of the rat sympathetic neuron

The presence of a novel voltage-dependent chloride current, active in the s ubthreshold range of membrane potential, was detected in the mature and int act rat sympathetic neuron in vitro by using the two-microelectrode voltage -clamp technique. Hyperpolarizing voltage steps applied to a neuron held at -40/-50 mV elicited inward currents,, whose initial magnitude displayed a linear instantaneous current-voltage (I-V) relationship; afterward, the cur rents decayed exponentially with a single voltage-dependent time constant ( 63.5 s at -40 mV: 10.8 s at -130 mV). The cell input conductance decreased during the command step with the same time course as the current. On return ing to the holding potential, the ensuing outward currents were accompanied by a slow increase in input conductance toward the initial values; the inw ard charge movement during the transient ON response (a mean of 76 nC in 8 neurons stepped from -50 to -90 mV) was completely balanced by outward char ge displacement during the OFF response. The chloride movements accompanyin g voltage modifications were studied by estimating the chloride equilibrium potential (E-Cl) at different holding potentials from the reversal of GABA evoked currents. [Cl-](i) was strongly affected by membrane potential, and at steady state it was systematically higher than expected from passive io n distribution. The transient current was blocked by substitution of isethi onate for chloride and by Cl- channel blockers (9AC and DIDS). It proved in sensitive to K+ channel blockers, external Cd2+, intracellular Ca2+ chelato rs [bis-(o-aminophenoxy)-N,N,N',N'-tetraacetic acid (BAPTA)] and reduction of [Na+](e). It is concluded that membrane potential shifts elicit a chlori de current that reflects readjustment of [Cl-](i). The cell input conductan ce was measured over the -40/-120-mV voltage range, in control medium, and under conditions in which either the chloride or the potassium current was blocked. A mix of chloride, potassium, and leakage conductances was detecte d at all potentials. The leakage component was voltage independent and cons tant at similar to 14 nS. Conversely, gCl decreased with hyperpolarization (80 nS at -40 mV, undetectable below -110 mV), whereas gK displayed a maxim um at -80 mV (55.3 nS). Thus the ratio gCl/gK continuously varied with memb rane polarization (2.72 at -50 mV; 0.33 at -110 mV). These data were forced in a model of the three current components here described, which accuratel y simulates the behavior observed in the "resting" neuron during membrane m igrations in the subthreshold potential range, thereby confirming that acti ve K and Cl conductances contribute to the genesis of membrane potential an d possibly to the control of neuronal excitability.