Gap junctional communication of primary human keratinocytes: characterization by dual voltage clamp and dye transfer

We have compared dye coupling in pairs of small (less than 10 microns in diameter) and large (greater than 20 microns in diameter) keratinocytes isolated from normal human epidermis, using Lucifer yellow microinjection. Under control conditions, dye coupling was found in only 1 out of the 25 small pairs tested, whereas it was evident in 75% of the large pairs (n = 52). After a 30-min incubation of the latter pairs in the presence of 10(-6) and 10(-4) M all-transretinoic acid (RA), the percentage of coupling was 53% (n = 15; NS) and 7% (n = 14; P less than 0.001), respectively. The almost complete uncoupling observed after 10(-4) M RA was not reversible even 30 min after return to control medium (n = 8). Dual whole-cell patch-clamp recordings from large keratinocyte pairs showed a macroscopic junctional conductance (gj) of 9 +/- 2 nS (n = 43), which was abolished by heptanol (3.5 mM) in a fully reversible way. Compared to heptanol, 10(-4) M RA abolished keratinocyte gj more slowly and irreversibly (n = 10). By contrast, 10(-6) M RA had no significant effect on gj (n = 8). Single-gap junctional channels were also identified between large keratinocytes. Events histograms of 152 transitions from three experiments revealed three main unitary conductances (gamma j) of 45 +/- 4, 78 +/- 4, and 106 +/- 7 pS. The dye coupling results indicate that junctional communication is markedly different in pairs of small and large cells, which showed the phenotype and keratin markers of basal and suprabasal keratinocytes, respectively. In the latter cell type, coupling is ensured by channels of three sizes and is blocked irreversibly by pharmacologic concentrations of RA.