Electrical coupling among heart cells in the absence of ultrastructurally defined gap junctions.

Cells from the ventricles of 7-day chick embryos were aggregated into spheroidal clusters by 48 hr of culture on a gyratory platform. All aggregates beat spontaneously and rhythmically. Microelectrode impalement of widely separated cells within aggregates indicated that they were coupled, as evidenced by a mean coupling ratio (delta V2/ delta V1) of 0.81 +/- 0.09, and by simultaneity of intrinsic electrical activity (action potentials and subthreshold voltage fluctuation). In freeze-fracture preparations, the cell surfaces contained numerous small groups of intramembrane protein (IMP) particles, arranged in macular clusters, and linear and circular arrays. Using the criterion of 4 clustered IMP particles to defined a minimal gap junction, 0.27% of the total P-face examined was devoted to gap junctional area. Within such clusters particles were packed at about 8200/micrometer2; in nonjunctional regions, particles were scattered at a density of about 2000/micrometer2. When exposed to cycloheximide (CHX: 50 micrograms/ml) for 24--48 hr, coupling ratio declined to 0.44. This decrease could be attributed largely to leakiness of the nonjunctional membrane. Aggregates continued to beat rhythmically and in a coordinated fashion even after 72 hr in inhibitor. However, between 3--21 hr in CHX gap junctional area declined to 0.10%, and all particle clusters disappeared from the P-faces of aggregates in CHX for 24 or 48 hr. Neither macular nor linear particle arrays were seen. We conclude that organized gap junctions are unnecessary for electrotonic coupling between embryonic heart cells. These findings support the idea that low-resistance cell-to-cell pathways may exist as isolated channels scattered throughout the area of closely apposed plasma membranes.