Figure 1.

Localization of the V_m determinants in the CT domain of rat Cx43. WT and the Cx43 channels truncated at positions S257stop and G242stop were expressed in Xenopus laevis oocytes pairs. (A) Sample records of nonjunctional and junctional currents (I₂) induced by the equal displacement of membrane potentials in the two cells (V₁ and V₂; steps of 40 s from −100 to 0 mV in increments of 20 mV, returning to −100 mV for 40 s between each step). The junctional currents (I_j) measured in oocyte 2 after application of small V_j pulses in oocyte 1 (+10 mV, 500 ms, 1 Hz). For WT (left) and S257stop (center) junctions, I_j decreased on depolarization, more rapidly and to progressively lower steady state values as more positive potentials were achieved. I_j recovered to its initial values on returning to −100 mV. The mutation G242stop largely abolished the dependence of junctional conductance (g_j) on V_m (right). (B) V_m dependence of the time constant of decrease in g_j. The time course of relaxations to G_jss was well fit by single exponentials (Inset) with long time constants that were shorter at more positive voltages. The smooth curve is derived from calculations assuming a single Boltzmann model. (C) Graphs of G_jss/V_m relations where G_jss is normalized to the value at −100 mV. The smooth curves are derived from fits to the square of a Boltzmann relation based on a model of two independent gates in series, one in each hemichannel (except for WT-G242stop, for which a single Boltzmann and gate were assumed). The parameters are given in Table 1. The G_jss of WT (○) and truncated S257stop (▴) junctions decreased by ≈50% when V_m increased from −100 to 0 mV. G_jss of truncated G242stop junctions (■) was quite insensitive to V_m. The G_jss/V_m curve of hybrid WT-G242stop junctions (●) was in between those of the homotypic junctions of the component combinations, suggesting that only the V_m gate of WT hemichannels contributed to their V_m dependence. Each point in B represents mean values (±SEM) of six pairs.