An in vitro system for studying insulin release: effects of glucose and glucose-6-phosphate.

1. Investigation of the ionic requirements of the in vitro insulin release system, which consists of cod islet plasma membrane and rabbit islet granules incubated at pH 6.5, showed that the presence of Ca(2+) was obligatory for the system to operate.2. Glucose-initiated insulin release was as effective in the presence of beta-gamma-methylene ATP, as it was in the presence of ATP. This analogue of ATP is a substrate neither for adenylate cyclase nor for any known animal membrane proteases. The effect of ATP on glucose mediated release is allosteric.3. Glucose (16 mM)-initiated insulin release was slower than that induced by glucose-6-phosphate (4 mM); 150 and 120 sec, respectively.4. The lag found with glucose-mediated insulin release was dependent upon glucose concentration. The lower the glucose concentration, the longer the lag. With 1 mM glucose the lag extended to 30 min.5. Once insulin release was initiated, the rate and amount of insulin release was independent of the glucose concentration.6. Pre-incubation of membranes with Ca(2+), glucose and ATP prior to the addition of granules, abolished the extended lag that had been obtained with 1 mM glucose. Events in the plasma membrane are the major contributor to the generation of the extended lag.7. The glucose analogue 5'thio-D-glucose, although not able to release insulin, was shown to compete with glucose for the glucoreceptor. By increasing the ratio of analogue to glucose the lag time increased. Thus, the lag time is dependent upon the ;effective' external glucose concentration.8. The max. amount of insulin released by 4 ng of membrane in the presence of glucose (16 mM) was 300 ng. The fact that membranes became refractory to glucose after this max. amount of insulin was released showed that recycling of release sites was not taking place in vitro and that granule: granule interactions were not occurring.9. The 120 sec lag before glucose-6-phosphate-initiated release was independent of glucose-6-phosphate concentration. The rate of insulin release with glucose-6-phosphate was concentration dependent.10. Glucose-6-phosphate did not cause further insulin release from a membrane that had released the max. amount of insulin it was capable of in the presence of glucose. The addition of tolbutamide (10 mM) to such a membrane did cause insulin release. This suggests that glucose and glucose-6-phosphate share a final common pathway.11. Adrenaline and somatostatin did not inhibit glucose-mediated insulin release.