European Journal of Cell Biology vol:36 issue:2 pages:269-76
Gap junction morphology was studied on freeze fracture replicas of pancreatic islet tissue, using morphometric techniques. In rat islets in situ, 60 percent of the connexions were polygonally packed in gap junctions, whereas the remaining part occurred in linear strands. After collagenase isolation, the islets presented similar numbers of gap junctions but contained virtually no linear strands. The distribution of connexions over polygonal or linear arrays also varied with the culture conditions: at 11.2 mM glucose, a higher percentage of particles occurred in gap junctions than at 5.6 mM glucose; this was also the case in other conditions with elevated cellular cyclic AMP levels. The total number of connexions increased when islets were cultured with dibutyryl cyclic AMP or with a phosphodiesterase inhibitor; conditions with an augmented number of gap junctions also displayed an elevated islet cyclic AMP content. A similar association was noted in newly formed aggregates of pancreatic B-cells purified by autofluorescence-activated. cell sorting. These results indicate that the number of classically defined gap junctions is not only dependent on the total number of connexions but also on their organization within the membrane. It is suggested that the distribution of connexions over polygonal and linear arrays follows a dynamic equilibrium varying with the extracellular conditions. Cyclic AMP appears to modulate the number of gap junctions between pancreatic B-cells both through an induction of new connexions and through an assembly of linearly organized particles into polygonal arrays.