Background and aims : The scaffold protein beta-arrestin2 (ARRB2) is known to uncouple G protein coupled receptors (GPCRs) from the G protein and to recruit new signaling pathways (such as ERK1/2). It has been reported in non beta cells a direct interaction of ARRB2 with the GLP-1 receptor (GLP-1R), while its interaction with the GIP receptor (GIPR) is unclear. Our aim was to determine if ARRB2 is involved in both incretin receptor signaling in mouse beta cells.Materials and methods : The experiments were carried out in beta cells from five-month-old Arrb2+/+ and Arrb2-/- male mice. cAMP production (CAMPS-epac), endogeneous PKA (AKAR3) and ERK1/2 (EKAR) activations were measured by live microscopy after adenoviral infection of cells with FRET-based sensors of interest. Epac2 (Epac2-GFP) recruitment to the plasma membrane was assessed by TIRF microscopy. Cytosolic Ca2+ concentration ([Ca2+]c) was measured by Fura2-LR.Results: The genetic deletion of Arrb2 in mice was associated with a better oral glucose tolerance and a concomitant increase in plasma insulin concentration (p<0.05) despite an impaired i.p. glucose tolerance and a decrease in beta cell mass, suggesting a greater incretin effect. However, opposite effects in response to both incretin hormones were observed in Arrb2-/- islets for insulin secretion. Whereas insulin secretion from Arrb2-/- islets was larger with physiological concentrations of GLP-1 (1-10pM GLP-1; p<0.01), it was reduced by 50% in response to GIP (100pM-10nM, p<0.01). When ARRB2 (ARRB2-GFP) was re-expressed in Arrb2-/- beta cells, insulin secretion in response to GLP-1 or GIP was restored to a similar level than in Arrb2+/+ islets. The larger insulin release induced by 10pM GLP-1 in Arrb2-/- beta cells was associated with an increased cAMP production, PKA activation, and [Ca2+]c (oscillations frequencies and mean level), while the Epac2 recruitment remained unchanged. In contrast, GLP-1-induced activation of ERK1/2 was strongly decreased (~50%, p<0.05) in Arrb2-/- beta cells. Unexpectedly, under GIP stimulation, the cAMP (cAMP, PKA, Epac2), [Ca2+]c and ERK1/2 signaling pathways were similar in Arrb2+/+ and Arrb2-/- beta cells. Additionally, GLP-1 compared to GIP induced PKA long lasting activation after removal of the stimulation (>25 min vs 5min) suggesting slow versus fast recycling receptors, respectively, that were independent of ARRB2 expression. Finally, glucolipotoxic conditions induced a 20% decrease of ARRB2 expression in human islets (p<0.05).Conclusion: Our study revealed a differential role of ARRB2 in GLP-1R and GIPR signalling, and therefore a specificity of action for two GPCR positively coupled to cAMP. ARRB2 contributes to a partial uncoupling of cAMP/PKA signalling in the pM range of GLP-1, and consequently reduced insulin secretion. By contrast, ARRB2 plays a crucial role in GIP potentiation of insulin secretion that is independent and/or distal to cAMP, [Ca2+]c and ERK1/2 changes. Therefore, any variation in the expression of ARRB2, as observed in diabetic states, should functionally impact the incretin effect.