Glucose‐induced insulin secretion: is the small G‐protein Rab27A the mediator of the KATP channel‐independent effect?

Abstract

From yeast to neurones, living cells communicate with their environment by secretion of substances through regulated exocytosis. The substances are stored in secretory granules within the cytosol. As the cells, their needs and function are extremely divergent, synthesis, storage and secretion of cellular products are adapted. Although the regulation of exocytosis is cell specific the secretory machineries are amazingly similar. Whether in yeast, neurone or insulin-secreting cells SNARE proteins enable the fusion of the two membranes, the vesicular and plasma membrane and the release of cell-made material. SNARE protein interaction, i.e. fusion, is largely regulated by Ca2+. Small G proteins comprise another protein family that is evolutionary highly conserved. Distinct members of this family regulate the formation, the directed trafficking and the fusion of vesicles. Rab3A and Rab27A are two small G proteins identified to be involved in insulin release (Regazzi et al. 1992; Olszewski et al. 1994; Yi et al. 2002). Another vesicular protein, granuphilin that mediates granule docking to the plasma membrane is an effector of Rab27A. Kasai et al. (2005) demonstrated that rab27A-deficient ashen mice exhibit glucose intolerance due to insufficient glucose-induced insulin secretion while granuphilin-deficient insulin secreting cells release a higher amount of insulin. These observations are puzzling and suggest that granuphilin inhibits while Rab27A promotes granule fusion. Measurement of membrane capacitance using the patch clamp method allows the on-line observation of exo- and endocytotic events in cells under voltage clamp and the analysis of kinetic changes of plasma membrane surface area. Brief depolarizing voltage pulses are applied to mimic glucose-induced action potentials that result in Ca2+ influx through voltage-dependent Ca2+ channels. The immediate quantal increase in capacitance has been proposed to represent the readily releasable pool (RRP) of already docked insulin-containing vesicles (Olofsson et al. 2002). With a rise in glucose concentration, the increase in capacitance is enlarged due to a recruitment of additional granules to the RRP. This effect of glucose is mimicked by larger and repetitive depolarizing pulses. Shorter and smaller depolarization results in smaller increases in capacitance from the immediate releasable pool (IRP) of granules. It seems logical that the RRP represents the first, rapid phase of insulin secretion after glucose stimulation while the second slow phase of secretion when glucose remains high may mainly result from granules recruited to the plasma membrane from the reserve pool. The study published in this issue of The Journal of Physiology by Merrins & Stuenkel (2008) used the approach of capacitance measurements to examine the effect of rab27A deficiency in ashen islet cells on IRP and RRP mobilization and refilling. Via a pulse protocol of five short (50 ms) and eight long (500 ms) depolarizing pulses, the size of IRP and RRP, respectively, was analysed. The repetition of the pulse protocol allowed the analysis of the refilling of the pools. The capacitance changes to the first pulse trail were not different between wild type and ashen mouse islet cells while the answer to the second pulse trail was reduced in ashen islet cells. This observation confirms convincingly that docking, i.e. IRP and RRP, is not changed by Rab27A deficiency but refilling the pools. This refilling deficiency was antagonized by cAMP, an effect blocked by the inhibition of protein kinase A, indicating that cAMP facilitates exocytosis independent of Rab27A expression. Increasing the glucose concentration incompletely restored refilling of RRP in ashen islet cells. Interestingly, the deficiency of refilling is thus specific for glucose. Total internal reflection microscopy (TRIFM), a method that allows the analysis of granule movement and fusion in living cells, was used by Nagamatsu and colleagues (Nagamatsu, 2006). They confirmed that after glucose stimulation, mainly docked granules are released immediately followed by recruited granules. Kasai et al. (2008) present a detailed analysis of the provenance of granules during K+- and glucose-induced insulin release. They found that K+ initiated secretion within 5 s and mainly docked granules (80%) underwent fusion. In contrast, glucose-induced exocytosis started after 20 s and only 30% of granules originated from the docked pool but 70% from the reserve pool. In ashen mice, granule fusion in islet cells induced by glucose but not by K+ was reduced mainly due to reduced recruitment from the reserve pool. These observations support a new concept that IRP and RRP are granule pools with variable Ca2+ sensitivities and are docked to and blocked at the plasma membrane via granuphilin. These granules become more sensitive to Ca2+-mediated fusion when granuphilin-dependent blocking is ameliorated by Rab27A. Glucose exerts a dual effect: it stimulates docked granules through Ca2+ influx as does K+ and via an unknown mechanism it stimulates Rab27A that activates docked granules as well as recruits granules to the plasma membrane. In this concept Rab27A may mediate the KATP channel-independent effect of glucose on insulin secretion.