Autocrine/paracrine interactions moendocrine pancreasdulating hormone release in the endocrine pancreas

Abstract

Human islet transplantation is emerging as an alternative to pancreas transplantation or insulin therapy in the treatment of type I diabetes. It has been possible to achieve euglycemia by transplanting isolated human islets in patients who suffer from the long term side effects of diabetes and insulin therapy. However, the islet transplantation procedure is still experimental. To obtain FDA approval, each batch of islets will have to be labeled with its potency to meet the transplantation criteria. Consequently, finding an in vitro test that predicts the quality of the islets prior to transplantation will require a better understanding of the human islet physiology.

In this thesis we report on the generation of an in vitro perifusion machine with high throughput capabilities which integrates with other commercial high content screening systems. These technologies were successfully applied to measure [Ca2+]i and hormone release from batches of damaged islets with decreased viability, and it was possible to differentiate those islets from their undamaged counterparts. Pharmacological profiling of individual batches of islet cells proved feasible, measuring both [Ca2+]i and hormone release. Moreover, we showed that the hormone release assay can be used to distinguish batches of human islets from healthy donors and donors with type II diabetes.

Our data on human islets revealed a cytoarchitecture that differs from that of other animal models used to study islet physiology and diabetes. In the human islet, all endocrine cells are intermingled throughout the islets without the mantle-core segregation observed in the rodent islet. The endocrine cells are facing blood vessels and they appear without a specific pattern. Additionally, [Ca2+]i handling in human islets is different from that in mouse islets. While the entire human islet does not show oscillations in [Ca2+]i as described for the mouse islet, single human beta cells show oscillations in [Ca2+]i that resemble those found in mouse.

Our data showed that glutamate is a potent stimulus for glucagon secretion but not for insulin secretion. While glutamate induced increases in [Ca2+]i in alpha cells through activation of the voltage gated Ca2+ channels, it does not cause any change in [Ca2+]i in beta cells. We provided evidences that alpha cells express the machinery needed for glutematergic signaling. We propose that glutamate released from the alpha cell activates the glutamate receptors in the alpha cell plasma membrane to allow more Ca2+ into the alpha cells and further increase glucagon release.

Finally, we demonstrated that ATP exerts different effects in human and in rodent islets. In human islets, ATP potentiated insulin release, at basal and at high glucose concentration, but it did not do so in mouse, rat, or pig islets. This potentiation most likely occurs by activation of purinergic receptors of the P2X3 type located in the beta cell plasma membrane. Upon activation, these receptors become permeable to Ca2+, allowing an influx of this ion into the beta cell cytoplasm, which stimulates further insulin release. Hence, ATP serves as an autocrine signal that forms a positive feedback loop stimulating insulin release in a glucose independent manner.

The suggestion that both alpha and beta cells utilize positive feedback loops to potentiate the secretion of their respective hormones might indicate that glucose alone is insufficient to achieve adequate glucagon and insulin release from the alpha and beta cells and that other additional autocrine/paracrine signals are required to achieve fine-tuned exocytosis.