It’s good to talk: cell–cell communication by gap junctions

Gap junctions facilitate the exchange of solutes, metabolic precursors and electrical currents between neighbouring cells. They appear as clusters (or plaques) of tightly packed particles, in which each particle is a single channel. Although it was known that the appearance of such plaques is associated with the electrical coupling of two cells, previous methods measured only the average conductance over an entire cell population and not the properties of a single junction.

In a recent study, Bukauskas et al. (Proc. Natl Acad. Sci. USA 97, 2556–2561; 2000) used GFP-tagged connexin 43 (Cx43–EGFP) and dual whole-cell patch clamps to investigate the relationship between clusters and junctional conductance (g_j) in cell–cell pairs. When this construct was transfected into cells that were defective in communication, Cx43–EGFP fluorescence was observed throughout, except at areas of cell-cell contact, where punctate staining was present at the cell membrane. Most pairs of cells with large plaques (>0.2 µm in diameter) showed electrical coupling, with g_j values ranging from 11–60 nS, whereas those joined by smaller plaques were frequently uncoupled.

As the intensity of fluorescence in a large plaque is essentially constant (see picture), Bukauskas et al. were able to correlate the activity measured across a pair of cells with the activity of a single channel. The fluorescence per unit area within the plaque, together with previous calculations of channel density in plaques, was used to estimate the fluorescence intensity of a single channel. Next, pairs of cells with a single gap junction were identified, and the total fluorescence of each plaque was determined. Using their estimate of single-channel fluorescence intensity, the authors could then estimate the number of channels within the plaque. They also measured the conductance between the pairs of cells. Armed with this information, they calculated the g_j of a single channel within a plaque, and identified three categories. Small plaques (90–330 channels) showed no electrical coupling and had no active channels. Slightly larger plaques (200–400 channels) exhibited weak conductance (g_j values of 0.05–0.7 nS) and contained only 1–2 active channels. Large plaques (≥500 channels), however, possessed 35 or more active channels and had g_j values of ≥4 nS.

From Bakushkas et al. Proc. Natl. Acad. Sci. USA

These findings were unexpected in several ways. First, it seems that a minimum cluster size, in terms of the number of channels, is required to open a gap junction. Second, only a fraction of channels within a gap junction are active at any given time. Third, in gap junctions above a certain critical size, the proportion of channels that are active does not seem to increase significantly with an increasing number of channels. Thus, gating by gap junctions seems to be an all-or-nothing phenomenon that occurs only when a certain channel concentration is attained, but in which an overlying regulatory step limits the number of channels that are active. These results offer new insights and also raise several questions — how is clustering initiated? What senses when the threshold number of channels has been reached? And how is channel activity regulated so that only a certain proportion of the channels within a junction is active at any one time?