The metabolic amplification of insulin secretion is strongly affected by cell culture

Background and aims: Metabolic amplification of insulin secretion is still poorly understood in spite of considerable research efforts. Here, we have tested the hypothesis that at least part of the conflicting results is due to different cellular models.

Methods: Freshly isolated or cultured (24 to 48h in RPMI1640) mouse islets were used to measure insulin secretion by perifusion and ELISA of the fractionated efflux. The mitochondrial redox potential was assessed by measuring islet NAD(P)H- and FAD-autofluorescence.

Results: A 60 min pre-perifusion with 500µM tolbutamide in absence of nutrients virtually abolished the insulinotropic effect of 30 mM glucose on freshly isolated islets. Under control condition (no tolbutamide in the pre-perifusion) glucose produced a modest response during the initial 10 min, which continuously increased during the next 30 min. When cultured islets were used, the effect of 30 mM glucose under control condition consisted in a marked first phase, followed by a moderately elevated plateau after 20 min. Surprisingly the same pattern was observed when 500µM tolbutamide was present. The secretory response was even augmented both during the first phase and the following plateau. In cultured islets glucose caused an increase of NAD(P)H and a decrease of FAD-fluorescence with the same kinetics as in freshly isolated islets. However, the resulting increase of NAD(P)H/FAD ratio was twice as high and tolbutamide had a much less diminishing effect.

Conclusion: Cell culture may not simply relieve the stress of collagenase isolation but may profoundly alter the secretory response of primary beta cells.