Insulin release from pancreatic ␤-cells plays an essential role in blood glucose homeostasis. Several proteins controlling insulin exocytosis have been identified, but the factors determining the expression of the components of the secretory machinery of ␤-cells remain largely unknown. MicroRNAs are newly discovered small non-coding RNAs acting as repressors of gene expression. We found that overexpression of mir-9 in insulinsecreting cells causes a reduction in exocytosis elicited by glucose or potassium. We show that mir-9 acts by diminishing the expression of the transcription factor Onecut-2 and, in turn, by increasing the level of Granuphilin/Slp4, a Rab GTPase effector associated with ␤-cell secretory granules that exerts a negative control on insulin release. Indeed, electrophoretic mobility shift assays, chromatin immunoprecipitation, and transfection experiments demonstrated that Onecut-2 is able to bind to the granuphilin promoter and to repress its transcriptional activity. Moreover, we show that silencing of Onecut-2 by RNA interference increases Granuphilin expression and mimics the effect of mir-9 on stimulus-induced exocytosis. Our data provide evidence that in insulin-producing cells adequate levels of mir-9 are mandatory for maintaining appropriate Granuphilin levels and optimal secretory capacity.The insulin in the circulation of adult mammals is produced exclusively by pancreatic ␤-cells, making them the central regulators of glucose homeostasis. Alterations in ␤-cell secretory function can cause hyperglycemia and lead to diabetes mellitus. During the last few years a number of key components of the machinery controlling insulin exocytosis have been identified. These components include different SNARE 3 proteins, Rab3 and Rab27 GTPases, and their effectors Noc2 and Granuphilin/ Slp4 (1-3). Alteration in the function or in the expression level of these proteins results in insulin secretory defects both in vitro (4 -8) and in vivo (9 -11). Despite recent progress in understanding the molecular mechanisms regulating the final events in the secretory pathway of ␤-cells, the factors determining the expression of the proteins involved in insulin exocytosis are largely unknown.MicroRNAs (miRNAs) are newly discovered regulators of gene expression that act by targeting the 3Ј-untranslated region (3Ј-UTR) of mRNA sequences and by preventing the productive translation of the messengers (12-15). miRNAs have been implicated in many processes in invertebrates, including cell proliferation, apoptosis (16 -18), fat metabolism (17), and neuronal patterning (19). Because of their spatial and temporal expression pattern and their conservation across species (20 -22), miRNAs are believed to play similar roles in all animal cells. In mammals, only for a few miRNAs a specific function has been assigned. A subset of miRNAs has been shown to be involved in metabolic regulation: mir-143 participates in human adipocyte differentiation (23), and the levels of mir-375, a pancreatic islet-specific miRNA, influence insulin...

Islet-brain 1 (IB1) is the rat and human homologue of JIP-1, a murine inhibitor of the c-Jun amino-terminal kinase (JNK). It was termed IB1 since its expression is mostly detectable in pancreatic islets and in the brain (2). IB1 was identified by studying the transcriptional mechanisms responsible for the pancreatic ␤-cell-specific control of glucose transporter gene GLUT2 (2,4,24,36). Subsequently, the human MAPK8IP1 gene, encoding IB1, was established as a candidate gene for diabetes mellitus (35). Indeed, the direct sequencing of this gene in human type 2 diabetes patients revealed the presence of a missense mutation (resulting in protein mutation S59N) which cosegregated with a rare form of monogenic type 2 diabetes. Ex vivo, this mutation was shown to induce an accelerated apoptosis in pancreatic ␤ cells (35). These observations identified IB1 as a key regulator for ␤-cell survival since it modulates the activation of the JNK signaling pathway, a system which plays an essential role in maturation, differentiation, and/or apoptosis (8, 16). For example, when the IB1 protein content is decreased, ␤ cells are more sensitive to cytokineinduced apoptosis by increasing the JNK activity (3). Thus, the IB1 expression level is critical for ␤-cell function.The goal of the present study was to understand how MAPK8IP1 gene expression is controlled in a tissue-specific manner. Work by Atouf and coauthors has correlated the selective presence of several gene transcripts in pancreatic ␤ and neuronal cells with the absence in these cells of a transcription factor named REST (RE-1 silencing transcription factor, also termed as NRSF) (1). This protein is a zinc finger transcriptional repressor found to be widely expressed during embryogenesis in all tissues, except in endocrine pancreas and mature neuronal tissues (1, 5). The REST gene displays modular organization conserved across humans, rats, and mice within the protein-coding region and is regulated by alternative splicing of REST pre-mRNA. It was reported that the REST isoform with nine zinc fingers is the most predominant isoform found in various tissues (25). Alternatively spliced REST protein isoforms differ in their DNA-binding domains and transrepression domains (26), suggesting different functions of REST. For example, the REST4 isoform, which is a truncated protein, inhibits REST activity by acting as a dominant-negative form of REST (DNREST) (26,32). REST binds to a 21-bp cis element called the RE-1 silencer element (NRSE), also known as repressor element RE-1, to negatively regulate in nonneuronal tissues several genes preferentially expressed in neuronal cells such as the rat SCG10, the rat type II sodium channel, the human synapsin I, and the rat N-methyl-D-aspartate (NMDA) receptor 1 genes (6,18,20,33). The repression effect induced by REST required the interaction of REST with the corepressor mSin3 and histone deacetylase I (HDACI) to form a complex which induces hypoacetylation of histone (15). These authors proposed that a remodeling of the chromatin stru...

Aims/hypothesisPro-atherogenic and pro-oxidant, oxidised LDL trigger adverse effects on pancreatic beta cells, possibly contributing to diabetes progression. Because oxidised LDL diminish the expression of genes regulated by the inducible cAMP early repressor (ICER), we investigated the involvement of this transcription factor and of oxidative stress in beta cell failure elicited by oxidised LDL.MethodsIsolated human and rat islets, and insulin-secreting cells were cultured with human native or oxidised LDL or with hydrogen peroxide. The expression of genes was determined by quantitative real-time PCR and western blotting. Insulin secretion was monitored by EIA kit. Cell apoptosis was determined by scoring cells displaying pycnotic nuclei.ResultsExposure of beta cell lines and islets to oxidised LDL, but not to native LDL raised the abundance of ICER. Induction of this repressor by the modified LDL compromised the expression of important beta cell genes, including insulin and anti-apoptotic islet brain 1, as well as of genes coding for key components of the secretory machinery. This led to hampering of insulin production and secretion, and of cell survival. Silencing of this transcription factor by RNA interference restored the expression of its target genes and alleviated beta cell dysfunction and death triggered by oxidised LDL. Induction of ICER was stimulated by oxidative stress, whereas antioxidant treatment with N-acetylcysteine or HDL prevented the rise of ICER elicited by oxidised LDL and restored beta cell functions.Conclusions/interpretationInduction of ICER links oxidative stress to beta cell failure caused by oxidised LDL and can be effectively abrogated by antioxidant treatment.Electronic supplementary materialThe online version of this article (doi:10.1007/s00125-011-2165-x) contains supplementary material, which is available to authorised users.

In type 2 diabetes (T2D), hepatic insulin resistance is strongly associated with nonalcoholic fatty liver disease (NAFLD). In this study, we hypothesized that the DNA methylome of livers from patients with T2D compared with livers of individuals with normal plasma glucose levels can unveil some mechanism of hepatic insulin resistance that could link to NAFLD. Using DNA methylome and transcriptome analyses of livers from obese individuals, we found that hypomethylation at a CpG site in (encoding platelet-derived growth factor α) and overexpression are both associated with increased T2D risk, hyperinsulinemia, increased insulin resistance, and increased steatohepatitis risk. Genetic risk score studies and human cell modeling pointed to a causative effect of high insulin levels on CpG site hypomethylation, overexpression, and increased PDGF-AA secretion from the liver. We found that PDGF-AA secretion further stimulates its own expression through protein kinase C activity and contributes to insulin resistance through decreased expression of insulin receptor substrate 1 and of insulin receptor. Importantly, hepatocyte insulin sensitivity can be restored by PDGF-AA-blocking antibodies, PDGF receptor inhibitors, and by metformin, opening therapeutic avenues. Therefore, in the liver of obese patients with T2D, the increased PDGF-AA signaling contributes to insulin resistance, opening new therapeutic avenues against T2D and possibly NAFLD.