Single-cell transcriptomics of East-Asian pancreatic islets cells.

Single-cell RNA-seq (scRNA-seq) of pancreatic islets have reported on α- and ß-cell gene expression in mice and subjects of predominantly European ancestry. We aimed to assess these findings in East-Asian islet-cells. 448 islet-cells were captured from three East-Asian non-diabetic subjects for scRNA-seq. Hierarchical clustering using pancreatic cell lineage genes was used to assign cells into cell-types. Differentially expressed transcripts between α- and ß-cells were detected using ANOVA and in silico replications of mouse and human islet cell genes were performed. We identified 118 α, 105 ß, 6 δ endocrine cells and 47 exocrine cells. Besides INS and GCG, 26 genes showed differential expression between α- and ß-cells. 10 genes showed concordant expression as reported in rodents, while FAM46A was significantly discordant. Comparing our East-Asian data with data from primarily European subjects, we replicated several genes implicated in nuclear receptor activations, acute phase response pathway, glutaryl-CoA/tryptophan degradations and EIF2/AMPK/mTOR signaling. Additionally, we identified protein ubiquitination to be associated among East-Asian ß-cells. We report on East-Asian α- and ß-cell gene signatures and substantiate several genes/pathways. We identify expression signatures in East-Asian ß-cells that perhaps reflects increased susceptibility to cell-death and warrants future validations to fully appreciate their role in East-Asian diabetes pathogenesis.

Comparison of pancreatic beta cells and alpha cells under hyperglycemia: Inverse coupling in pAkt-FoxO1.

Type 2 diabetes manifests beta cell deficiencies and alpha cell expansion which is consistent with relative insulin deficiency and glucagon oversecretion. The effects of hyperglycemia on alpha cells are not as understood in comparison to beta cells. Hyperglycemia increases oxidative stress, which induces Akt activation or FoxO activation, depending on cell type. Several studies independently reported that FoxO1 translocations in alpha cells and beta cells were opposite. We compared the responses of pancreatic alpha cells and beta cells against hyperglycemia. Alpha TC-1 cells and Beta TC-6 cells were incubated with control (5mM Glucose) or high glucose (33mM Glucose) with or without PI3K inhibitor or FoxO1 inhibitor. We assessed PI3K, pAkt and phosphorylated FoxO1 (pFoxO1) in both cell lines. Immunostaining of BrdU and FoxO1 was detected by green fluorescence microscopy and confocal microscopy. Hyperglycemia and H2O2 decreased PI3K and pAKT in beta cells, but increased them in alpha cells. FoxO1 localizations and pFoxO1 expressions between alpha cells and beta cells were opposite. Proliferation of beta cells was decreased, but alpha cell proliferation was increased under hyperglycemia. Antioxidant enzymes including superoxide dismutase (SOD) and catalase were increased in beta cells and they were reversed with FoxO1 inhibitor treatment. Increased proliferation in alpha cells under hyperglycemia was attenuated with PI3K inhibitor. In conclusion, hyperglycemia increased alpha cell proliferation and glucagon contents which are opposite to beta cells. These differences may be related to contrasting PI3K/pAkt changes in both cells and subsequent FoxO1 modulation.

Enhanced transcellular penetration and drug delivery by crosslinked polymeric micelles into pancreatic multicellular tumor spheroids.

Many attempts have been made in the application of multicellular tumor spheroids (MCTS) as a 3D tumor model to investigate their biological responses upon introduction of polymeric micelles as nanocarriers for therapeutic applications. However, the micelle penetration pathways in MCTS are not yet known. In this study, micelles (uncrosslinked, UCM) were prepared by self-assembly of block copolymer poly(N-(2-hydroxypropyl) methacrylamide-co-methacrylic acid)-block-poly(methyl methacrylate) (P(HPMA-co-MAA)-b-PMMA). Subsequently, the shells were crosslinked to form relatively stable micelles (CKM). Both UCM and CKM penetrated deeper and delivered more doxorubicin (DOX) into MCTS than the diffusion of the free DOX. Additionally, CKM revealed higher delivery efficiency than UCM. The inhibition of caveolae-mediated endocytosis, by Filipin treatment, decreased the uptake and penetration of the micelles into MCTS. Treatment with Exo1, an exocytosis inhibitor, produced the same effect. Furthermore, movement of the micelles through the extracellular matrices (ECM), as modelled using collagen micro-spheroids, appeared to be limited to the peripheral layer of the collagen spheroids. Those results indicate that penetration of P(HPMA-co-MAA)-b-PMMA micelles depended more on transcellular transport than on diffusion through ECM between the cells. DOX-loaded CKM inhibited MCTS growth more than their UCM counterpart, due to possible cessation of endocytosis and exocytosis in the apoptotic peripheral cells, caused by faster release of DOX from UCM.

Pancreatic Endocrine Effects of Dopamine Receptors in Human Islet Cells.

OBJECTIVE: To date, there are no reports on the cellular localization of dopamine receptors in the human pancreas. In our study, we determined the localization and expression of 5 dopamine receptors (D(1), D(2), D(3), D(4), and D(5)) in normal human pancreas tissue. METHODS: Human nonpathological pancreas tissues were fixed with 4% paraformaldehyde, paraffin-embedded, and processed for immunohistochemical analysis to detect dopamine receptors in the human pancreas tissue by using double immunofluorescent labeling and confocal microscopy. RESULTS: We found that the D(1) receptor is present in ß cells; the D(2) receptor is expressed by α, δ, and pancreatic polypeptide cells; the D(4) receptor is expressed by ß and polypeptide cells; whereas the D(5) receptor is expressed only by δ cells. CONCLUSIONS: Our results identify the dopamine receptors (D(1)-D(5)) in normal pancreas tissue and provide a morphological basis for studying the pancreatic endocrine effects of dopamine and suggest a new target for the clinical treatment of diabetes.

Distribution Profile of Inositol 1,4,5-Trisphosphate Receptor/Ca2+ Channels in α and ß Cells of Pancreas: Dominant Localization in Secretory Granules and Common Error in Identification of Secretory Granule Membranes.

OBJECTIVES: The α and ß cells of pancreatic islet release important hormones in response to intracellular Ca increases that result from Ca releases through the inositol 1,4,5-trisphoshate receptor (IP3R)/Ca channels. Yet no systematic studies on distribution of IP3R/Ca channels have been done, prompting us to investigate the distribution of all 3 IP3R isoforms. METHODS: Immunogold electron microscopy was performed to determine the presence and the relative concentrations of all 3 IP3R isoforms in 2 major organelles secretory granules (SGs) and the endoplasmic reticulum of α and ß cells of rat pancreas. RESULTS: All 3 IP3R isoforms were present in SG membranes of both cells, and the IP3R concentrations in SGs were ∼2-fold higher than those in the endoplasmic reticulum. Moreover, large halos shown in the electron microscope images of insulin-containing SGs of ß cells were gap spaces that resulted from separation of granule membranes from the surrounding cytoplasm. CONCLUSIONS: These results strongly suggest the important roles of SGs in IP3-induced, Ca-dependent regulatory secretory pathway in pancreas. Moreover, the accurate location of SG membranes of ß cells was further confirmed by the location of another integral membrane protein synaptotagmin V and of membrane phospholipid PI(4,5)P2.

ATF3 expression is induced by low glucose in pancreatic α and ß cells and regulates glucagon but not insulin gene transcription.

The pancreas is critical for maintaining glucose homeostasis. Activating transcription factor 3 (ATF3) is an adaptive response transcription factor. There are major discrepancies in previous reports on pancreatic ATF3; therefore, its role in the pancreas is unclear. To better elucidate the role of ATF3 in the pancreas, we conducted in vitro studies using pancreatic α and ß cell lines, and also evaluated the use of ATF3 antibodies for immunohistochemistry. We determined ATF3 expression was increased by low glucose and decreased by high glucose in both αTC-1.6 and ßTC3 cells. We also showed that adenovirus-mediated ATF3 overexpression increased glucagon promoter activity and glucagon mRNA levels in αTC-1.6 cells; whereas, it had no effect on insulin promoter activity and insulin mRNA levels in ßTC3 cells. Although immunostaining with the C-19 ATF3 antibody demonstrated predominant expression in α cells rather than ß cells, ATF3 staining was still detected in ATF3 knockout mice as clearly as in control mice. On the other hand, another ATF3 antibody (H-90) detected ATF3 in both α cells and ß cells, and was clearly diminished in ATF3 knockout mice. These results indicate that previous discrepancies in ATF3 expression patterns in the pancreas were caused by the varying specificities of the ATF3 antibodies used, and that ATF3 is actually expressed in both α cells and ß cells.

Adhesion molecule CADM1 contributes to gap junctional communication among pancreatic islet α-cells and prevents their excessive secretion of glucagon.

Cell adhesion molecule-1 (CADM1) is a recently identified adhesion molecule of pancreatic islet α-cells that mediates nerve-α-cell interactions via trans-homophilic binding and serves anatomical units for the autonomic control of glucagon secretion. CADM1 also mediates attachment between adjacent α-cells. Since gap junctional intercellular communication (GJIC) among islet cells is essential for islet hormone secretion, we examined whether CADM1 promotes GJIC among α-cells and subsequently participates in glucagon secretion regulation. Dye transfer assays using αTC6 mouse α-cells, which endogenously express CADM1, supported this possibility; efficient cell-to-cell spread of gap junction-permeable dye was detected in clusters of αTC6 cells transfected with nonspecific, but not with CADM1-targeting, siRNA. Immunocytochemical analysis of connexin 36, a major component of the gap junction among αTC6 cells, revealed that it was localized exclusively to the cell membrane in CADM1-non-targeted αTC6 cells, but diffusely to the cytoplasm in CADM1-targeted cells. Next, we incubated CADM1-targeted and non-targeted αTC6 cells in a medium containing 1 mM glucose and 200 mM arginine for 30 min to induce glucagon secretion, and found that the targeted cells secreted three times more glucagon than did the non-targeted. We conducted similar experiments using pancreatic islets that were freshly isolated from wild-type and CADM1-knockout mice, and expressed glucagon secretion as ratios relative to baseline values. The increase in ratio was larger in CADM1-knockout islets than in wild-type islets. These results suggest that CADM1 may serve as a volume limiter of glucagon secretion by sustaining α-cell attachment necessary for efficient GJIC.

One-step purification of functional human and rat pancreatic alpha cells.

Pancreatic alpha cells contribute to glucose homeostasis by the regulated secretion of glucagon, which increases glycogenolysis and hepatic gluconeogenesis in response to hypoglycemia. Alterations of glucagon secretion are observed in diabetic patients and exacerbate the disease. The restricted availability of purified primary alpha cells has limited our understanding of their function in health and disease. This study was designed to establish convenient protocols for the purification of viable alpha cells from rat and human pancreatic islets by FACS, using intrinsic cellular properties. Islets were isolated from the pancreata of Wistar rats or deceased human organ donors. Dispersed islet cells were separated by FACS based on light scatter and autofluorescence. Purity of sorted cells was evaluated by immunocytochemistry using hormone specific antibodies. Relative hormone expression was further determined by quantitative RT-PCR. Viability was determined by Annexin V and propidium iodide staining and function was assessed by monitoring cytoplasmic free Ca(2+) concentration ([Ca(2+)](i)) using Fura-2/AM. We developed species-specific FACS gating strategies that resulted in populations consisting mainly of alpha cells (96.6 ± 1.4%, n = 3 for rat; 95.4 ± 1.7%, n = 4 for human, mean ± SEM). These cell fractions showed ~5-fold and ~4-fold enrichment (rat and human, respectively) of glucagon mRNA expression compared to total ungated islet cells. Most of the sorted cells were viable and functional, as they responded with an increase in [Ca(2+)](i) upon stimulation with L-arginine (10 mM). The majority of the sorted human alpha cells responded also to stimulation with kainate (100 µM), whereas this response was infrequent in rat alpha cells. Using the same sample preparation, but a different gating strategy, we were also able to sort rat and human populations enriched in beta cells. In conclusion, we have simplified and optimized a method for the purification of rat alpha cells, as well as established a novel approach to separate human alpha cells using neither antibodies nor dyes possibly interfering with cellular functions.

c-Kit and stem cell factor regulate PANC-1 cell differentiation into insulin- and glucagon-producing cells.

Recent evidence has shown that stem cell factor (SCF) and its receptor, c-Kit, have an important role in pancreatic islet development by promoting islet cell differentiation and proliferation. In this study, we examined the role of c-Kit and SCF in the differentiation and proliferation of insulin- and glucagon-producing cells using a human pancreatic duct cell line (PANC-1). Our study showed that increased expression of endocrine cell markers (such as insulin and glucagon) and transcription factors (such as PDX-1 and PAX-6) coincided with a decrease in CK19(+) and c-Kit(+) cells (P<0.001) during PANC-1 cell differentiation, determined by immunofluorescence and qRT-PCR. Cells cultured with exogenous SCF showed an increase in insulin(+) (26%) and glucagon(+) (35%) cell differentiation (P<0.01), an increase in cell proliferation (P<0.05) and a decrease in cell apoptosis (P<0.01). siRNA knockdown of c-Kit resulted in a decrease in endocrine cell differentiation with a reduction in PDX-1 and insulin mRNA, as well as the number of cells immunostaining for PDX-1 and insulin. Taken together, these results show that c-Kit/SCF interactions are involved in mediating islet-like cluster formation and islet-like cell differentiation in a human pancreatic duct cell line.

Characterization of Erg K+ channels in alpha- and beta-cells of mouse and human islets.

Voltage-gated eag-related gene (Erg) K(+) channels regulate the electrical activity of many cell types. Data regarding Erg channel expression and function in electrically excitable glucagon and insulin producing cells of the pancreas is limited. In the present study Erg1 mRNA and protein were shown to be highly expressed in human and mouse islets and in alpha-TC6 and Min6 cells alpha- and beta-cell lines, respectively. Whole cell patch clamp recordings demonstrated the functional expression of Erg1 in alpha- and beta-cells, with rBeKm1, an Erg1 antagonist, blocking inward tail currents elicited by a double pulse protocol. Additionally, a small interference RNA approach targeting the kcnh2 gene (Erg1) induced a significant decrease of Erg1 inward tail current in Min6 cells. To investigate further the role of Erg channels in mouse and human islets, ratiometric Fura-2 AM Ca(2+)-imaging experiments were performed on isolated alpha- and beta-cells. Blocking Erg channels with rBeKm1 induced a transient cytoplasmic Ca(2+) increase in both alpha- and beta-cells. This resulted in an increased glucose-dependent insulin secretion, but conversely impaired glucagon secretion under low glucose conditions. Together, these data present Erg1 channels as new mediators of alpha- and beta-cell repolarization. However, antagonism of Erg1 has divergent effects in these cells; to augment glucose-dependent insulin secretion and inhibit low glucose stimulated glucagon secretion.

DTA-mediated targeted ablation revealed differential interdependence of endocrine cell lineages in early development of zebrafish pancreas.

Cell lineage analysis is critical in understanding the relationship between progenitors and differentiated cells as well as the mechanism underlying the process of differentiation. In order to study the zebrafish endocrine pancreas cell lineage, transgenic expression of diphtheria toxin gene A chain (DTA) under two cell type-specific promoters derived from the insulin (ins) and somatostatin2 (sst2) genes was used to ablate the two types of endocrine cells: insulin-producing beta-cells and somatostatin-producing delta-cells, respectively. We found that ablation of beta-cells resulted in a reduction of not only beta-cells but also glucagon-producing alpha-cells; in contrast, delta-cells were largely unaffected. Ablation of delta-cells led to reduction of all three types of endocrine cells: alpha-, beta-, and delta. Interestingly, alpha-cells were more profoundly affected in both beta- and delta-cell ablations and were frequently reduced together with beta- and delta-cells. By taking advantage of Tg(ins:gfp) and Tg(sst2:gfp) lines, we also monitored the changes of different types of endocrine cells in vivo after ablation and found that both beta- and delta-cell populations significantly recovered by 3dpf after their ablation and it seemed that delta-cells had a better capability of recovery than beta-cells. Thus, our current observations indicated differential interdependence of these three cell lineages. The development of zebrafish alpha-cells, but not delta-cells, is dependent on beta-cells, while the development of both alpha- and beta-cells is dependent on delta-cells. In contrast, the development of delta-cells was independent of beta-cells.

Interleukin 22 receptor 1 expression in pancreas islets.

OBJECTIVES: Interleukin (IL) 22 is a recently identified T-cell-derived cytokine. IL-22 binds at the cell surface to a heterodimer receptor complex composed of IL-22 receptor (R) 1 and IL-10R2. In this study, we performed immunohistochemical analyses for IL-22R1 expression in human pancreatic tissue. METHODS: Normal human pancreatic tissue (n = 8) was immunostained with antihuman IL-22R1 antibodies following standard immunohistochemical procedures. RESULTS: In the normal human pancreas, IL-22R1 was expressed in the islets of Langerhans. IL-22R1 was not expressed by the acinar cells and ductal epithelium. Double-immunostaining experiments showed that the majority of insulin-expressing beta cells and glucagon-expressing alpha cells were immunopositive for IL-22R1. CONCLUSIONS: The islets of Langerhans are the local site for IL-22R1 expression in the human pancreas. It may be that the T-cell-mediated immune response modulates cell islet function through IL-22 signaling.

Direct visualisation of peptide hormones in cultured pancreatic islet alpha- and beta-cells by intact-cell mass spectrometry.

The application of intact-cell mass spectrometry (ICM) by matrix-assisted laser desorption/ionisation time-of-flight (MALDI-TOF) mass spectrometry to achieve direct protein-profiling of bacterial species is now well established. However, this methodology has not to our knowledge been applied to the analysis of mammalian cells in routine culture. Here, we describe a novel application of ICM by which we have identified proteins in intact cells from two lines representative of pancreatic islet alpha- and beta-cells. Adherent alphaTC1 clone 9 and betaTC6 F7 cells were harvested into phosphate-buffered saline (PBS) using enzyme-free dissociation buffer before 1 microL of cell suspension was spotted onto MALDI plates. Cells were overlaid with sinapinic acid then washed with pure water before application of a final coat of sinapinic acid. Data in the 2000-20,000 m/z range were acquired in linear mode on a Voyager DE-Pro mass spectrometer. The proteins which ionised were composed in large part of peptide hormones (e.g. insulin and glucagon) known to be packaged into the secretory granules of the beta- and alpha-cells respectively. However, in addition to visualising the peptides expected to be associated with these cells, a mass consistent with oxyntomodulin was identified in the cultured alpha-cells, a finding not previously reported to our knowledge. In summary, this paper describes, for the first time, a rapid and direct method useful for identifying secretory products in intact endocrine cells.

Intranuclear rodlets in human pancreatic islet cells.

OBJECTIVES: Intranuclear rodlets (INRs) are rod-shaped intranuclear inclusions that we have described in neurons of the human brain. We recently identified these structures in pancreatic islet cells. The objectives of this study are to describe the light microscopic features and cellular pattern of distribution of INRs in human pancreatic islet cells. METHODS: Double immunofluorescence staining was performed on 5 human pancreatic tissue samples for the detection of class III beta tubulin (C3T) to detect INRs and for promyelocytic leukemia (PML) protein to examine the relationship between PML and INRs. RESULTS: Intranuclear rodlets were detected in 22.99% of pancreatic B cells compared with only 3.11%, 1.80%, and 1.60% of A, D, and PP cells, respectively. Twenty-four percent of C3T-immunoreactive INRs showed partial or complete immunoreactivity for PML. Promyelocytic leukemia staining within the nuclei of B cells was confined to INRs and was not present in the typical PML bodies present in other cell types. Spatially, PML and C3T staining of islet cell INRs appeared to be mutually exclusive within individual INRs. CONCLUSIONS: Intranuclear rodlets are present within the nuclei of pancreatic islet cells, where they reside predominantly but not exclusively in B cells. Immunoreactivity of B-cell INRs for PML suggests that the functional significance of INRs may be related to that of PML and/or PML bodies. Conversely, the exclusive localization of PML staining to INRs in B cells indicates that PML's function in B cells is selectively associated with INRs. The mutually exclusive pattern of PML and C3T staining suggests dynamic interactions between these 2 proteins in B-cell INRs. In light of evidence for the involvement of INRs and of PML bodies in disease, it will be of interest to investigate these structures in animal models of diabetes and in human diabetes.

Noninsulinoma pancreatogenous hypoglycemia syndrome: quantitative and immunohistochemical analyses of islet cells for insulin, glucagon, somatostatin, and pancreatic and duodenal homeobox protein.

OBJECTIVE: To present a case of an elderly man with noninsulinoma pancreatogenous hypoglycemia syndrome (NIPHS) and to determine the pathogenesis of this syndrome. METHODS: The pancreas of our patient with NIPHS was immunocytochemically stained for insulin-, glucagon-, and somatostatin-secreting cells and pancreatic and duodenal homeobox protein (PDX-1). The clinical findings and morphologic and immunocytochemical analyses of the islets of our patient are described, along with a review of related published reports. RESULTS: A 78-year-old man presented with hyperinsulinemic hypoglycemia, with episodes unrelated to meals or fasting. An insulinoma could not be localized by preoperative imaging or by intraoperative ultrasonography or palpation. He underwent a 70% distal pancreatectomy. For assessment of the possibility that a nuclear transcription factor regulating islet beta-cell growth and development is overexpressed in this disease and is responsible for diffuse islet hyperfunction and proliferation of beta-cells, pancreatic sections from our patient were stained immunocytochemically for PDX-1, insulin, glucagon, and somatostatin. Morphologic findings were compared with pancreatic sections from normal control patients and normative data reported in the literature. Clinical findings and morphologic analyses were consistent with NIPHS. Islets were arranged in long clusters, both in the pancreatic tissue and in peripancreatic adipose tissue. Islets were small but increased in number, and insulin, glucagon, and somatostatin were present in the islets. The relative intensity of insulin staining was increased in our patient in comparison with that in the control patients, and PDX-1 was not overexpressed. CONCLUSION: The etiopathogenesis of NIPHS in this patient involved (1) an increased number of islets with development of ectopic islets in the peripancreatic adipose tissue; (2) alpha- and delta- as well as beta-cell proliferation; and (3) an early step in the development of the islet not involving overexpression of PDX-1.

Exophilin4/Slp2-a targets glucagon granules to the plasma membrane through unique Ca2+-inhibitory phospholipid-binding activity of the C2A domain.

Rab27a and Rab27b have recently been recognized to play versatile roles in regulating the exocytosis of secretory granules and lysosome-related organelles by using multiple effector proteins. However, the precise roles of these effector proteins in particular cell types largely remain uncharacterized, except for those in pancreatic beta cells and in melanocytes. Here, we showed that one of the Rab27a/b effectors, exophilin4/Slp2-a, is specifically expressed in pancreatic alpha cells, in contrast to another effector, granuphilin, in beta cells. Like granuphilin toward insulin granules, exophilin4 promotes the targeting of glucagon granules to the plasma membrane. Although the interaction of granuphilin with syntaxin-1a is critical for the targeting activity, exophilin4 does this primarily through the affinity of its C2A domain toward the plasma membrane phospholipids phosphatidylserine and phosphatidylinositol-4,5-bisphosphate. Notably, the binding activity to phosphatidylserine is inhibited by a physiological range of the Ca(2+) concentration attained after secretagogue stimulation, which presents a striking contrast to the Ca(2+)-stimulatory activity of the C2A domain of synaptotagmin I. Analyses of the mutant suggested that this novel Ca(2+)-inhibitory phospholipid-binding activity not only mediates docking but also modulates the subsequent fusion of the secretory granules.

Downstream regulatory element antagonistic modulator regulates islet prodynorphin expression.

Calcium-binding proteins regulate transcription and secretion of pancreatic islet hormones. Here, we demonstrate neuroendocrine expression of the calcium-binding downstream regulatory element antagonistic modulator (DREAM) and its role in glucose-dependent regulation of prodynorphin (PDN) expression. DREAM is distributed throughout beta- and alpha-cells in both the nucleus and cytoplasm. As DREAM regulates neuronal dynorphin expression, we determined whether this pathway is affected in DREAM(-/-) islets. Under low glucose conditions, with intracellular calcium concentrations of <100 nM, DREAM(-/-) islets had an 80% increase in PDN message compared with controls. Accordingly, DREAM interacts with the PDN promoter downstream regulatory element (DRE) under low calcium (<100 nM) conditions, inhibiting PDN transcription in beta-cells. Furthermore, beta-cells treated with high glucose (20 mM) show increased cytoplasmic calcium (approximately 200 nM), which eliminates DREAM's interaction with the DRE, causing increased PDN promoter activity. As PDN is cleaved into dynorphin peptides, which stimulate kappa-opioid receptors expressed predominantly in alpha-cells of the islet, we determined the role of dynorphin A-(1-17) in glucagon secretion from the alpha-cell. Stimulation with dynorphin A-(1-17) caused alpha-cell calcium fluctuations and a significant increase in glucagon release. DREAM(-/-) islets also show elevated glucagon secretion in low glucose compared with controls. These results demonstrate that PDN transcription is regulated by DREAM in a calcium-dependent manner and suggest a role for dynorphin regulation of alpha-cell glucagon secretion. The data provide a molecular basis for opiate stimulation of glucagon secretion first observed over 25 years ago.