Pancreatic islet B cell function was studied in vitro using three structurally different preparations of islet tissues: isolated, intact islets, dispersed islet cells attached singly to microcarrier beads, and reaggregated islet cells. Mechanisms of intercellular communication are eliminated with single cell preparations, whereas in aggregates cell to cell communications are reestablished and a defined microenvironment restored. Perifusion studies measured nonstimulated and glucose- and arginine-stimulated insulin release from the three islet tissues. Insulin secretion rates were expressed as a function of cellular DNA content, permitting direct comparison between tissues. During perifusion with low (2.8 or 5.5 mM) glucose concentrations, secretion rates of single islet cells were up to 6-fold greater (P less than 0.001) than those of intact islets. Perifusion of islet cells with 2.8 mM glucose and 100 or 500 pg glucagon/ml had no effect whereas GH-release-inhibiting factor (330 and 1000 pg/ml) decreased nonstimulated insulin secretion rates by 15% (P less than 0.05). After reaggregation, basal insulin secretion rates were restored toward those of intact islets. Glucose (5.5-30 mM) and L-arginine (5-20 mM) elicited first phase insulin responses from single islet cells that were not significantly different from those observed with intact islets; in contrast, second phase responses of single islets to glucose were approximately 50% those seen with intact islets, and their second phase responses to arginine were absent. Single islet cell first and second phase insulin responses to 5.5 mM glucose were enhanced 2.2-fold (P less than 0.01) and 2.8-fold (P less than 0.05), respectively, in the presence of exogenous glucagon, resulting in secretory profiles characteristic of intact islets. Reaggregation of single islet cells was associated with markedly increased first and second phase insulin responses to both glucose and arginine stimulation. These data show that disruption of the islet microanatomy results in alteration of insulin secretory responses and that these effects can be reversed, in part by exogenous glucagon and GH-release-inhibiting factor, and by reaggregation. Although different mechanisms appear important for nonstimulated, first and second phase insulin release, the findings support a role for both direct intercellular communication and hormonal secretion by islet A and D cells in the modulation of B cell function.

Isolated adult rat pancreatic islets were dispersed into single cells and cultured free-floating for 3 to 4 d, during which time islet cells reaggregated spontaneously into spherical clusters or pseudoislets. The gross morphology of these tissues resembled nondissociated islets. Electron microscopy revealed well-preserved cell ultrastructure and intercellular membrane connections. Immunofluorescent localization of islet cell types showed that A cells tended to be peripherally distributed around a B cell core, with D cells scattered throughout the aggregate mass. The dynamics of insulin release from pseudoislets were evaluated in vitro by perifusion techniques. Pseudoislets exhibited clear biphasic dose-dependent insulin responses to 30 min glucose stimulation over the range 5.5 to 30 mM. Repeated 2-min pulses with 22 mM glucose elicited brief monophasic spikes of insulin release of consistent magnitude. L-Arginine (5 to 20 mM) evoked biphasic insulin release but these responses were not dose-dependent. These data indicate that islet cells reaggregate into structures with close morphologic similarities to intact islets, and that pseudoislet B cells continue to secrete insulin in response to nutrient secretagogues, comparable to that seen with islets in vitro and in situ.

The use of islet DNA content to standardize insulin secretion rates from pancreatic islets of different sizes has been studied. Isolated intact islets were sorted into 4 size categories and perifused with 22 mM glucose, collecting effluent in 5 min fractions for insulin RIA. DNA content of perifused islets was measured by fluorometric assay, and insulin secretion expressed as pmoles/ug DNA/unit time. For islets with diameters less than 300 u (1) insulin secretion was proportional to islet size; (2) insulin release per islet and islet DNA content were strongly correlated; (3) when expressed as a function of DNA content, insulin secretion from different sized islets was not significantly different. These relationships did not continue for very large islets (above 300 u) suggesting a limiting islet size for insulin secretion in vitro. The data demonstrates that expression of insulin secretion from pancreatic islets with diameters less than 300 u, as a function of their DNA content standardizes secretion irrespective of islet size and number, and should allow direct comparison of secretory responses between different islet tissue preparations.

Dispersed adult rat pancreatic islet cells were incubated with Cytodex-3 microcarrier beads for 72 h, during which time single cells adhered firmly to bead surfaces. Electron microscopy revealed well-preserved ultrastructure of attached A, B, and D cells. Perifusion of these cultures showed stable basal insulin release, brisk, biphasic insulin responses to 30-min glucose stimulation, and consistent, monophasic spikes of insulin release in response to repeated, brief pulses of glucose. These results indicate that adult rat islet cells attach to microcarriers and remain viable in culture. This preparation offers advantages for studies of hormone secretory dynamics of differentiated single islet cells, free from cell-to-cell interactions.