Isles within islets: The lattice origin of small-world networks in pancreatic tissues

Abstract

The traditional computational model of the pancreatic islets of Langerhans is a lattice of -cells connected with gap junctions. Numerous studies have investigated the behavior of networks of coupled -cells and have shown that gap junctions synchronize bursting strongly. This simplistic architecture of islets, however, seems increasingly untenable at the face of recent experimental advances. In a microfluidics experiment on isolated islets, Rocheleau et al. (2004) showed a failure of penetration of excitation when one end received high glucose and other end was not excited sufficiently; this suggested that gap junctions may not be efficient at inducing synchrony throughout the islet. Recently, Stozer et al. (2013) have argued that the functional networks of -cells in an islet are small world. Their results implicate the existence of a few long-range connections among cells in the network. The physiological reason underlying this claim is not well understood. These studies cast doubt on the original lattice model that largely predict an all-or-none synchrony among the cells. Here we have attempted to reconcile these observations in a unified framework. We assume that cells in the islet are coupled randomly to their nearest neighbors with some probability, . We simulated detailed -cell bursting in such islets. By varying systematically we were led to network parameters similar to those obtained by Stozer et al. (2013). We find that the networks within islets break up into components giving rise to smaller isles within the super structure—isles-within-islets, as it were. This structure can also account for the partial excitation seen by Rocheleau et al. (2004). Our updated view of islet architecture thus explains the paradox how islets can have strongly synchronizing gap junctions, and be weakly coordinated at the same time.