A new measure for the strength of electrical synapses

Abstract

Electrical synapses, like chemical synapses, mediate intraneuronal communication. Electrical synapses are typically quantified by subthreshold measurements of coupling, which fall short in describing their impact on spiking activity in coupled neighbors. Here, we describe a novel measurement for electrical synapse strength that directly evaluates the effect of synaptically transmitted activity on spike timing. This method, also applicable to neurotransmitter-based synapses, communicates the considerable strength of electrical synapses. For electrical synapses measured in rodent slices of the thalamic reticular nucleus and in simple model neurons, spike timing is modulated by tens of ms by activity in a coupled neighbor.

Keywords: connexin36; efficacy; electrical synapses; gap junction; thalamic reticular nucleus.

Figure 1

Subthreshold and suprathreshold measurements for the strength of an electrical synapse. (A) Coupling measured with current pulses. Coupling coefficients are a ratio of voltage deflections. Here, voltage deflections were initiated by a step in current delivered in one neuron (gray) that echoed in the coupled neuron (black). cc = 0.17 for the pair shown. Scale bar 1 mV, 50 ms. (B) Coupling coefficient measured by spike and spikelet amplitudes. Spikes were elicited in one neuron (gray) and spikelets in the coupled neighbor (black). Scale bar 2.5 mV (black,) 25 mV (gray), 25 ms. (C) Coupling coefficient measured by burst and burstlet amplitudes, for a longer burst event in one neuron (gray) and a burstlet in the coupled neighbor (black). Scale bar 2.5 mV (black), 25 mV (gray), 25 ms. (D) Coupling measured by latency changes. Modulation of spike latency δL was measured by comparing timing of spikes elicited in one cell alone (black; gray cell quiet), and with the coupled neighbor also driven to spike (gray). Scale bar 2 mV (gray), 20 mV (black), 25 ms. All data presented are from the same pair.

Figure 2

Measuring δL, latency modulation. (A) Spiking in one cell of a coupled pair (blue) in response to current pulses of increasing amplitude (lower, shown in gray). Scale bar 25 ms, 20 mV. The coupled cell was quiet and is not shown. (B) Spiking in the same cell (light blue) for the same current pulses as in (A) (lower, shown in gray), with the coupled neighbor also spiking (lower, shown in green). Responses from (A) are repeated, vertically offset for clarity (darker blue). (C) Latency of spiking in (A) (pulse alone) and (B) (pulse + GJ input) plotted against input amplitude. For peri-threshold inputs (100 pA) in this cell, δL, the percentage change in perithreshold spike latency, was 50%.

Figure 3

Comparison of δL to other measures of electrical synapse strength. (A) δL plotted against coupling coefficient cc in each direction for a set of n = 18 pairs. R² = 0.38. (B) δL plotted against coupling conductance G_C; R² = 0.56. (C) δL plotted against absolute change in latency for each cell in 18 pairs.

Figure 4

(A) Coupling demonstrated by a hyperpolarizing current pulse in a pair of simple Hodgkin-Huxley neurons; cc = 0.15. Scale bar 2 mV, 25 ms. (B) Spiking one of the model cells (blue) for a minimal input (lower, gray); the coupled neuron was quiet. (C) Spiking in the same cell (light blue) for the same current pulses as in (A) (lower, shown in gray), with the coupled neighbor also spiking (lower, shown in green). Responses from (A) are repeated, vertically offset, for clarity (dark blue). Scale bar 10 mV, 25 ms. (D) δL plotted against coupling coefficient in the modeled pair, for three values of excitability [sodium conductances of 60 (pink), 75 (maroon) and 90 (red) μS/cm²].