Liraglutide protects ß-cells in novel human islet spheroid models of type 1 diabetes.

To enable accurate, high-throughput and longer-term studies of the immunopathogenesis of type 1 diabetes (T1D), we established three in-vitro islet-immune injury models by culturing spheroids derived from primary human islets with proinflammatory cytokines, activated peripheral blood mononuclear cells or HLA-A2-restricted preproinsulin-specific cytotoxic T lymphocytes. In all models, ß-cell function declined as manifested by increased basal and decreased glucose-stimulated insulin release (GSIS), and decreased intracellular insulin content. Additional hallmarks of T1D progression such as loss of the first-phase insulin response (FFIR), increased proinsulin-to-insulin ratios, HLA-class I expression, and inflammatory cytokine release were also observed. Using these models, we show that liraglutide, a glucagon-like peptide 1 receptor agonist, prevented loss of GSIS under T1D-relevant stress, by preserving the FFIR and decreasing immune cell infiltration and cytokine secretion. Our results corroborate that liraglutide mediates an anti-inflammatory effect that aids in protecting ß-cells from the immune-mediated attack that leads to T1D.

Human Islet Microtissues as an In Vitro and an In Vivo Model System for Diabetes.

Loss of pancreatic ß-cell function is a critical event in the pathophysiology of type 2 diabetes. However, studies of its underlying mechanisms as well as the discovery of novel targets and therapies have been hindered due to limitations in available experimental models. In this study we exploited the stable viability and function of standardized human islet microtissues to develop a disease-relevant, scalable, and reproducible model of ß-cell dysfunction by exposing them to long-term glucotoxicity and glucolipotoxicity. Moreover, by establishing a method for highly-efficient and homogeneous viral transduction, we were able to monitor the loss of functional ß-cell mass in vivo by transplanting reporter human islet microtissues into the anterior chamber of the eye of immune-deficient mice exposed to a diabetogenic diet for 12 weeks. This newly developed in vitro model as well as the described in vivo methodology represent a new set of tools that will facilitate the study of ß-cell failure in type 2 diabetes and would accelerate the discovery of novel therapeutic agents.

Genetic models rule out a major role of beta cell glycogen in the control of glucose homeostasis.

AIMS/HYPOTHESIS: Glycogen accumulation occurs in beta cells of diabetic patients and has been proposed to partly mediate glucotoxicity-induced beta cell dysfunction. However, the role of glycogen metabolism in beta cell function and its contribution to diabetes pathophysiology remain poorly understood. We investigated the function of beta cell glycogen by studying glucose homeostasis in mice with (1) defective glycogen synthesis in the pancreas; and (2) excessive glycogen accumulation in beta cells. METHODS: Conditional deletion of the Gys1 gene and overexpression of protein targeting to glycogen (PTG) was accomplished by Cre-lox recombination using pancreas-specific Cre lines. Glucose homeostasis was assessed by determining fasting glycaemia, insulinaemia and glucose tolerance. Beta cell mass was determined by morphometry. Glycogen was detected histologically by periodic acid-Schiff's reagent staining. Isolated islets were used for the determination of glycogen and insulin content, insulin secretion, immunoblots and gene expression assays. RESULTS: Gys1 knockout (Gys1 (KO)) mice did not exhibit differences in glucose tolerance or basal glycaemia and insulinaemia relative to controls. Insulin secretion and gene expression in isolated islets was also indistinguishable between Gys1 (KO) and controls. Conversely, despite effective glycogen overaccumulation in islets, mice with PTG overexpression (PTG(OE)) presented similar glucose tolerance to controls. However, under fasting conditions they exhibited lower glycaemia and higher insulinaemia. Importantly, neither young nor aged PTG(OE) mice showed differences in beta cell mass relative to age-matched controls. Finally, a high-fat diet did not reveal a beta cell-autonomous phenotype in either model. CONCLUSIONS/INTERPRETATION: Glycogen metabolism is not required for the maintenance of beta cell function. Glycogen accumulation in beta cells alone is not sufficient to trigger the dysfunction or loss of these cells, or progression to diabetes.

Evaluation of the Effects of Harmine on ß-cell Function and Proliferation in Standardized Human Islets Using 3D High-Content Confocal Imaging and Automated Analysis.

Restoration of ß-cell mass through the induction of proliferation represents an attractive therapeutic approach for the treatment of diabetes. However, intact and dispersed primary islets suffer from rapidly deteriorating viability and function ex vivo, posing a significant challenge for their experimental use in proliferation studies. Here, we describe a novel method for the assessment of compound effects on ß-cell proliferation and count using reaggregated primary human islets, or islet microtissues (MTs), which display homogeneous size and tissue architecture as well as robust and stable functionality and viability for 4 weeks in culture. We utilized this platform to evaluate the dose-dependent short- and long-term effects of harmine on ß-cell proliferation and function. Following compound treatment and EdU incorporation, islet MTs were stained and confocal-imaged for DAPI (nuclear marker), NKX6.1 (ß-cell marker), and EdU (proliferation marker), allowing automated 3D-analysis of number of total cells, ß-cells, and proliferating ß- and non-ß-cells per islet MT. In parallel, insulin secretion, intracellular insulin and ATP contents, and Caspase 3/7 activity were analyzed to obtain a comprehensive overview of islet MT function and viability. We observed that 4-day harmine treatment increased ß- and non-ß-cell proliferation, NKX6.1 expression, and basal and stimulated insulin secretion in a dose-dependent manner, while fold-stimulation of secretion peaked at intermediate harmine doses. Interestingly, 15-day harmine treatment led to a general reduction in harmine's proliferative effects as well as altered dose-dependent trends. The described methodology provides a unique tool for in vitro high-throughput evaluation of short- and long-term changes in human ß-cell proliferation, count and fraction along with a variety of functional parameters, in a representative 3D human islet model.

Gsα-dependent signaling is required for postnatal establishment of a functional ß-cell mass.

OBJECTIVE: Early postnatal life is a critical period for the establishment of the functional ß-cell mass that will sustain whole-body glucose homeostasis during the lifetime. ß cells are formed from progenitors during embryonic development but undergo significant expansion in quantity and attain functional maturity after birth. The signals and pathways involved in these processes are not fully elucidated. Cyclic adenosine monophosphate (cAMP) is an intracellular signaling molecule that is known to regulate insulin secretion, gene expression, proliferation, and survival of adult ß cells. The heterotrimeric G protein Gs stimulates the cAMP-dependent pathway by activating adenylyl cyclase. In this study, we sought to explore the role of Gs-dependent signaling in postnatal ß-cell development. METHODS: To study Gs-dependent signaling, we generated conditional knockout mice in which the α subunit of the Gs protein (Gsα) was ablated from ß-cells using the Cre deleter line Ins1Cre. Mice were characterized in terms of glucose homeostasis, including in vivo glucose tolerance, glucose-induced insulin secretion, and insulin sensitivity. ß-cell mass was studied using histomorphometric analysis and optical projection tomography. ß-cell proliferation was studied by ki67 and phospho-histone H3 immunostatining, and apoptosis was assessed by TUNEL assay. Gene expression was determined in isolated islets and sorted ß cells by qPCR. Intracellular cAMP was studied in isolated islets using HTRF-based technology. The activation status of the cAMP and insulin-signaling pathways was determined by immunoblot analysis of the relevant components of these pathways in isolated islets. In vitro proliferation of dissociated islet cells was assessed by BrdU incorporation. RESULTS: Elimination of Gsα in ß cells led to reduced ß-cell mass, deficient insulin secretion, and severe glucose intolerance. These defects were evident by weaning and were associated with decreased proliferation and inadequate expression of key ß-cell identity and maturation genes in postnatal ß-cells. Additionally, loss of Gsα caused a broad multilevel disruption of the insulin transduction pathway that resulted in the specific abrogation of the islet proliferative response to insulin. CONCLUSION: We conclude that Gsα is required for ß-cell growth and maturation in the early postnatal stage and propose that this is partly mediated via its crosstalk with insulin signaling. Our findings disclose a tight connection between these two pathways in postnatal ß cells, which may have implications for using cAMP-raising agents to promote ß-cell regeneration and maturation in diabetes.

Wisp1 is a circulating factor that stimulates proliferation of adult mouse and human beta cells.

Expanding the mass of pancreatic insulin-producing beta cells through re-activation of beta cell replication has been proposed as a therapy to prevent or delay the appearance of diabetes. Pancreatic beta cells exhibit an age-dependent decrease in their proliferative activity, partly related to changes in the systemic environment. Here we report the identification of CCN4/Wisp1 as a circulating factor more abundant in pre-weaning than in adult mice. We show that Wisp1 promotes endogenous and transplanted adult beta cell proliferation in vivo. We validate these findings using isolated mouse and human islets and find that the beta cell trophic effect of Wisp1 is dependent on Akt signaling. In summary, our study reveals the role of Wisp1 as an inducer of beta cell replication, supporting the idea that the use of young blood factors may be a useful strategy to expand adult beta cell mass.

Stress-Induced MicroRNA-708 Impairs ß-Cell Function and Growth.

The pancreatic ß-cell transcriptome is highly sensitive to external signals such as glucose oscillations and stress cues. MicroRNAs (miRNAs) have emerged as key factors in gene expression regulation. Here, we aimed to identify miRNAs that are modulated by glucose in mouse pancreatic islets. We identified miR-708 as the most upregulated miRNA in islets cultured at low glucose concentrations, a setting that triggers a strong stress response. miR-708 was also potently upregulated by triggering endoplasmic reticulum (ER) stress with thapsigargin and in islets of ob/ob mice. Low-glucose induction of miR-708 was blocked by treatment with the chemical chaperone 4-phenylbutyrate, uncovering the involvement of ER stress in this response. An integrative analysis identified neuronatin (Nnat) as a potential glucose-regulated target of miR-708. Indeed, Nnat expression was inversely correlated with miR-708 in islets cultured at different glucose concentrations and in ob/ob mouse islets and was reduced after miR-708 overexpression. Consistent with the role of Nnat in the secretory function of ß-cells, miR-708 overexpression impaired glucose-stimulated insulin secretion (GSIS), which was recovered by NNAT overexpression. Moreover, miR-708 inhibition recovered GSIS in islets cultured at low glucose. Finally, miR-708 overexpression suppressed ß-cell proliferation and induced ß-cell apoptosis. Collectively, our results provide a novel mechanism of glucose regulation of ß-cell function and growth by repressing stress-induced miR-708.

Generation of a Conditional Allele of the Transcription Factor Atonal Homolog 8 (Atoh8).

Atonal Homolog 8 (Atoh8) is a basic helix-loop-helix (bHLH) transcription factor that is highly conserved across species and expressed in multiple tissues during embryogenesis. In the developing pancreas, Atoh8 is expressed in endocrine progenitors but declines in hormone-positive cells, suggesting a role during early stages of the endocrine differentiation program. We previously generated a whole-body Atoh8 knockout but early lethality of null embryos precluded assessment of Atoh8 functions during organ development. Here we report the generation of a conditional Atoh8 knockout mouse strain by insertion of two loxP sites flanking exon 1 of the Atoh8 gene. Pancreas-specific Atoh8 knockout (Atoh8 Δpanc) mice were obtained by mating this strain with a Pdx1-Cre transgenic line. Atoh8 Δpanc mice were born at the expected mendelian ratio and showed normal appearance and fertility. Pancreas weight and gross pancreatic morphology were normal. All pancreatic cell lineages were present, although endocrine δ (somatostatin) cells were modestly augmented in Atoh8 Δpanc as compared to control neonates. This increase did not affect whole-body glucose tolerance in adult knockout animals. Gene expression analysis in embryonic pancreases at the time of the major endocrine differentiation wave revealed modest alterations in several early endocrine differentiation markers. Together, these data argue that Atoh8 modulates activation of the endocrine program but it is not essential for pancreas formation or endocrine differentiation in the mouse. Given the ubiquitous expression pattern of Atoh8, the availability of a mouse strain carrying a conditional allele for this gene warrants further studies using temporally regulated Cre transgenic lines to elucidate time or cell-autonomous functions of Atoh8 during development and in the adult.