Multiple-channel conductance states and voltage regulation of embryonic chick cardiac gap junctions

Summary

We used the double whole-cell voltage-clamp technique on ventricle cell pairs isolated from 7-day chick heart to measure the conductance of their gap junctions (G _j) and junctional channels (γ _j) with a steady-state voltage difference (V _j) applied across the junction. Currents were recorded from single gap junction channels (i _j) as symmetrical rectangular signals of equal size and opposite sign in the two cells, and γ _j was measured from i _j/V _j. We observed channel openings at six reproducible conductance levels with means of 42.6, 80.7, 119.6, 157.7, 200.4 and 240.3 pS. More than half of all openings were to the 80-and 160-pS conductance levels. The probability that a high conductance event (e.g., 160 or 240 pS) results from the random simultaneous opening of several 40-pS channels is small, based on their frequency of occurrence and on the prevalence of shifts between small and large conductance states with no intervening 40-pS steps. Our results are consistent with three models of embryonic cardiac gap junction channel configuration: a homogeneous population of 40-pS channels that can open cooperatively in groups of up to six; a single population of large channels with a maximal conductance near 240 pS and five smaller substates; or several different channel types, each with its own conductance.

G _j was determined from the junctional current (I _j) elicited by rectangular pulses of applied transjunctional voltage as I _j/V _j. It was highest near 0 V _j and was progressively reduced by application of V _j between 20 and 80 mV or −20 and −80 mV. In response to increases in V _j, G _j decayed in a voltage-and timedependent fashion. After a 6-sec holding period at 0 V _j, the initial conductance (G _init) measured immediately after the onset of an 80-mV step in V _j was nearly the same as that measured by a 10-mV prepulse. However, during 6-sec pulses of V _j>±20 mV, G _j declined over several seconds from G _init to a steady-state value (G _ss). At potentials greater than ±20 mV the current decay could be fit with biexponential curves with the slow decay time constant (τ ₂) 5–20 times longer than τ ₁. For the response to a step to 80 mV V _j, for example, τ ₁=127 msec and τ ₂=2.6 sec. The rate of current decay in response to smaller positive or negative steps in V _j was slower, the magnitude of the decline was smaller, and the ratio τ ₂/τ ₁ decreased. The relationship between G _init and V _j was approximately linear between 0 and 80 mV or −80 mV. whereas the relationship between G _ss and V _j was nonlinear beyond ±20 mV. Upon returning to 0 V _j, G _j recovered with a biexponential time course, reaching its maximal value after several seconds; recovery time constants after a step in V _j from 80 to 0 mV were 225 msec and 1.9 sec. In the resting state, at low junctional voltage, high conductance channel activity (160–240 pS) is favored. Voltage-dependent decline of G _j results in part from a shift from high to lower conductance states.