beta-Carbolines (including harmane and pinoline) stimulate insulin secretion by a mechanism that may involve interaction with imidazoline I(3)-receptors but which also appears to be mediated by actions that are additional to imidazoline receptor agonism. Using the MIN6 beta-cell line, we now show that both the imidazoline I(3)-receptor agonist, efaroxan, and the beta-carboline, harmane, directly elevate cytosolic Ca(2+) and increase insulin secretion but that these responses display different characteristics. In the case of efaroxan, the increase in cytosolic Ca(2+) was readily reversible, whereas, with harmane, the effect persisted beyond removal of the agonist and resulted in the development of a repetitive train of Ca(2+)-oscillations whose frequency, but not amplitude, was concentration-dependent. Initiation of the Ca(2+)-oscillations by harmane was independent of extracellular calcium but was sensitive to both dantrolene and high levels (20 mM) of caffeine, suggesting the involvement of ryanodine receptor-gated Ca(2+)-release. The expression of ryanodine receptor-1 and ryanodine receptor-2 mRNA in MIN6 cells was confirmed using reverse transcription-polymerase chain reaction (RT-PCR) and, since low concentrations of caffeine (1 mM) or thimerosal (10 microM) stimulated increases in [Ca(2+)](i), we conclude that ryanodine receptors are functional in these cells. Furthermore, the increase in insulin secretion induced by harmane was attenuated by dantrolene, consistent with the involvement of ryanodine receptors in mediating this response. By contrast, the smaller insulin secretory response to efaroxan was unaffected by dantrolene. Harmane-evoked changes in cytosolic Ca(2+) were maintained by nifedipine-sensitive Ca(2+)-influx, suggesting the involvement of L-type voltage-gated Ca(2+)-channels. Taken together, these data imply that harmane may interact with ryanodine receptors to generate sustained Ca(2+)-oscillations in pancreatic beta-cells and that this effect contributes to the insulin secretory response.