Characterisation of BHK-21 cells engineered to secrete human insulin

Autoimmune destruction of cells in the pancreas leads to type I, or insulin dependent diabetes mellitus (IDDM), through the loss of endogenous insulin production capacity. This paper describes an attempt to generate artificial cells using the fibroblast cell line BHK21. Stable transfectants expressing the human preproinsulin (PPI) gene were isolated and characterised. The resulting clone selected for further analysis (BHK-PPI-C16) was capable of secreting 0.12 pmol proinsulin/hr/10⁵ cells and maintained a steady cellular proinsulin content of 0.36 ± 0.04 pmol l^–1. There was no processing of the proinsulin to mature insulin. The cells were unresponsive to glucose but there was increased proinsulin secretion in the presence of agents that stimulated formation of intracellular cAMP. Transfection of cDNAs for the key elements of the glucose sensing apparatus (GLUT2 and glucokinase) led to a subphysiological stimulation of secretion when glucokinase was transfected alone while there was a complete loss of insulin secretion when both components were overexpressed. The deleterious effect on proinsulin secretion observed upon co-expression of the glucose sensing genes may have implications for applications requiring multigene expression in BHK21 cells.     

Identification of the Ubiquitin-like Domain of Midnolin as a New Glucokinase Interaction Partner

Glucokinase acts as a glucose sensor in pancreatic beta cells. Its posttranslational regulation is important but not yet fully understood. Therefore, a pancreatic islet yeast two-hybrid library was produced and searched for glucokinase-binding proteins. A protein sequence containing a full-length ubiquitin-like domain was identified to interact with glucokinase. Mammalian two-hybrid and fluorescence resonance energy transfer analyses confirmed the interaction between glucokinase and the ubiquitin-like domain in insulin-secreting MIN6 cells and revealed the highest binding affinity at low glucose. Overexpression of parkin, an ubiquitin E3 ligase exhibiting an ubiquitin-like domain with high homology to the identified, diminished insulin secretion in MIN6 cells but had only some effect on glucokinase activity. Overexpression of the elucidated ubiquitin-like domain or midnolin, containing exactly this ubiquitin-like domain, significantly reduced both intrinsic glucokinase activity and glucose-induced insulin secretion. Midnolin has been to date classified as a nucleolar protein regulating mouse development. However, we could not confirm localization of midnolin in nucleoli. Fluorescence microscopy analyses revealed localization of midnolin in nucleus and cytoplasm and co-localization with glucokinase in pancreatic beta cells. In addition we could show that midnolin gene expression in pancreatic islets is up-regulated at low glucose and that the midnolin protein is highly expressed in pancreatic beta cells and also in liver, muscle, and brain of the adult mouse and cell lines of human and rat origin. Thus, the results of our study suggest that midnolin plays a role in cellular signaling of adult tissues and regulates glucokinase enzyme activity in pancreatic beta cells.     

Glucose-regulated insulin production from genetically engineered human non-beta cells

In this report we describe development and characterization of four human cell lines that are able to secrete insulin and C-peptide in response to higher concentrations of glucose. These cell lines have been developed by stably and constitutively expressing human proinsulin with a furin-cleavable site, whereas expression of furin is regulated by glucose concentration. These cell lines have been cloned and, therefore, the transgene in each cell is located in an identical location of the genome leading to a uniform expression. Cloning has also allowed us to identify cell lines with more desirable properties such as higher basal insulin secretion and/or better glucose responsiveness. We have further shown that the insulin produced by these cells is biologically active and induces normoglycemia when injected in diabetic animals. Our objective in initiating these studies was to identify a cell line that could serve as a surrogate beta cell line for therapeutic intervention in type I diabetic patients.     

Apolipoprotein A-I primes beta cells to increase glucose stimulated insulin secretion

The increase of plasma levels of high-density lipoproteins and Apolipoprotein A-I (ApoA-I), its main protein component, has been shown to have a positive action on glucose disposal in type 2 diabetic patients. The current study investigates the unexplored function of ApoA-I to prime beta cells for improved insulin secretion.INS-1E rat clonal beta cells as well as isolated murine islets were used to study the effect of ApoA-I on responsiveness of the beta cells to high glucose challenge. Confocal and transmission electron microscopy were used to dissect ApoA-I mechanisms of action. Chemical endocytosis blockers were used to understand the role of ApoA-I internalization in mediating its positive effect.Pre-incubation of beta cells and isolated murine islets with ApoA-I augmented glucose stimulated insulin secretion. This effect appeared to be due to an increased reservoir of insulin granules at the cell membrane, as confirmed by confocal and transmission electron microscopy. Moreover, ApoA-I induced pancreatic and duodenal homeobox 1 (PDX1) shuttling from the cytoplasm to the nucleus, with the subsequent increase in the proinsulin processing enzyme protein convertase 1 (PC1/3). Finally, the blockade of ApoA-I endocytosis in beta cells resulted in a loss of ApoA-I positive action on insulin secretion.The proposed mechanisms of the phenomenon here described include ApoA-I internalization into beta cells, PDX1 nuclear translocation, and increased levels of proinsulin processing enzymes. Altogether, these events lead to an increased number of insulin granules.     

Real-time analysis of intracellular glucose and calcium in pancreatic beta cells by fluorescence microscopy

Glucose is the physiological stimulus for insulin secretion in pancreatic beta cells. The uptake and phosphorylation of glucose initiate and control downstream pathways, resulting in insulin secretion. However, the temporal coordination of these events in beta cells is not fully understood. The recent development of the FLII(12)Pglu-700μ-δ6 glucose nanosensor facilitates real-time analysis of intracellular glucose within a broad concentration range. Using this fluorescence-based technique, we show the shift in intracellular glucose concentration upon external supply and removal in primary mouse beta cells with high resolution. Glucose influx, efflux, and metabolism rates were calculated from the time-dependent plots. Comparison of insulin-producing cells with different expression levels of glucose transporters and phosphorylating enzymes showed that a high glucose influx rate correlated with GLUT2 expression, but was largely also sustainable by high GLUT1 expression. In contrast, in cells not expressing the glucose sensor enzyme glucokinase glucose metabolism was slow. We found no evidence of oscillations of the intracellular glucose concentration in beta cells. Concomitant real-time analysis of glucose and calcium dynamics using FLII(12)Pglu-700μ-δ6 and fura-2-acetoxymethyl-ester determined a glucose threshold of 4mM for the [Ca(2+)](i) increase in beta cells. Indeed, a glucose concentration of 7mM had to be reached to evoke large amplitude [Ca(2+)](i) oscillations. The K(ATP) channel closing agent glibenclamide was not able to induce large amplitude [Ca(2+)](i) oscillations in the absence of glucose. Our findings suggest that glucose has to reach a threshold to evoke the [Ca(2+)](i) increase and subsequently initiate [Ca(2+)](i) oscillations in a K(ATP) channel independent manner.     

Effects of acute or prolonged exposure to human leptin on isolated human islet function

The adipocyte-derived hormone leptin has been reported to inhibit, have no effect, or potentiate insulin secretion in-vitro; these effects mainly depend on the species considered, the concentrations used, and the length of exposure. We investigated the direct effects of recombinant human leptin (HL) on human pancreatic beta cell function by studying insulin secretion (IS), hexokinase and glucokinase activity and Km, and potassium channel permeability in purified human islets (HI). In acute experiments, no effect of 1, 5, 20, or 50 ng/ml HL on glucose or arginine stimulated insulin release was found, whereas 500 ng/ml HL caused a significant decrease of glucose induced IS. After 24h pre-culture with either 20 or 500 ng/ml HL, a significant reduction of glucose (but not arginine) stimulated IS was observed. Exposure to leptin caused a significant increase of potassium channel permeability, whereas hexokinase and glucokinase activity and Km remained unchanged. These results suggest that physiological human leptin concentration is able to importantly affect glucose (but not arginine) stimulated insulin release from human islets only after prolonged exposure. This effect is probably mediated by changes of potassium channel permeability, and is not accompanied by modifications of glucose phosphorylating enzymes properties.     

Impaired anaplerosis and insulin secretion in insulinoma cells caused by small interfering RNA-mediated suppression of pyruvate carboxylase

Anaplerosis, the synthesis of citric acid cycle intermediates, by pancreatic beta cell mitochondria has been proposed to be as important for insulin secretion as mitochondrial energy production. However, studies designed to lower the rate of anaplerosis in the beta cell have been inconclusive. To test the hypothesis that anaplerosis is important for insulin secretion, we lowered the activity of pyruvate carboxylase (PC), the major enzyme of anaplerosis in the beta cell. Stable transfection of short hairpin RNA was used to generate a number of INS-1 832/13-derived cell lines with various levels of PC enzyme activity that retained normal levels of control enzymes, insulin content, and glucose oxidation. Glucose-induced insulin release was decreased in proportion to the decrease in PC activity. Insulin release in response to pyruvate alone, 2-aminobicyclo[2,2,1]heptane-2-carboxylic acid (BCH) plus glutamine, or methyl succinate plus beta-hydroxybutyrate was also decreased in the PC knockdown cells. Consistent with a block at PC, the most PC-deficient cells showed a metabolic crossover point at PC with increased basal and/or glucose-stimulated pyruvate plus lactate and decreased malate and citrate. In addition, in BCH plus glutamine-stimulated PC knockdown cells, pyruvate plus lactate was increased, whereas citrate was severely decreased, and malate and aspartate were slightly decreased. The incorporation of 14C into lipid from [U-14C]glucose was decreased in the PC knockdown cells. The results confirm the central importance of PC and anaplerosis to generate metabolites from glucose that support insulin secretion and even suggest PC is important for insulin secretion stimulated by noncarbohydrate insulin secretagogues.     

Cdk5-dependent regulation of glucose-stimulated insulin secretion

Tight glycemic control in individuals with diabetes mellitus is essential to prevent or delay its complications. Present treatments to reduce hyperglycemia mainly target the ATP-sensitive K(+) (K(ATP)) channel of pancreatic beta cells to increase insulin secretion. These current approaches are often associated with the side effect of hypoglycemia. Here we show that inhibition of the activity of cyclin-dependent kinase 5 (Cdk5) enhanced insulin secretion under conditions of stimulation by high glucose but not low glucose in MIN6 cells and pancreatic islets. The role of Cdk5 in regulation of insulin secretion was confirmed in pancreatic beta cells deficient in p35, an activator of Cdk5. p35-knockout mice also showed enhanced insulin secretion in response to a glucose challenge. Cdk5 kinase inhibition enhanced the inward whole-cell Ca(2+) channel current and increased Ca(2+) influx across the L-type voltage-dependent Ca(2+) channel (L-VDCC) upon stimulation with high glucose in beta cells, but had no effect on Ca(2+) influx without glucose stimulation. The inhibitory regulation by Cdk5 on the L-VDCC was attributed to the phosphorylation of loop II-III of the alpha(1C) subunit of L-VDCC at Ser783, which prevented the binding to SNARE proteins and subsequently resulted in a decrease of the activity of L-VDCC. These results suggest that Cdk5/p35 may be a drug target for the regulation of glucose-stimulated insulin secretion.     

Differences between human and rodent pancreatic islets: low pyruvate carboxylase, atp citrate lyase, and pyruvate carboxylation and high glucose-stimulated acetoacetate in human pancreatic islets

Anaplerosis, the net synthesis in mitochondria of citric acid cycle intermediates, and cataplerosis, their export to the cytosol, have been shown to be important for insulin secretion in rodent beta cells. However, human islets may be different. We observed that the enzyme activity, protein level, and relative mRNA level of the key anaplerotic enzyme pyruvate carboxylase (PC) were 80-90% lower in human pancreatic islets compared with islets of rats and mice and the rat insulinoma cell line INS-1 832/13. Activity and protein of ATP citrate lyase, which uses anaplerotic products in the cytosol, were 60-75% lower in human islets than in rodent islets or the cell line. In line with the lower PC, the percentage of glucose-derived pyruvate that entered mitochondrial metabolism via carboxylation in human islets was only 20-30% that in rat islets. This suggests human islets depend less on pyruvate carboxylation than rodent models that were used to establish the role of PC in insulin secretion. Human islets possessed high levels of succinyl-CoA:3-ketoacid-CoA transferase, an enzyme that forms acetoacetate in the mitochondria, and acetoacetyl-CoA synthetase, which uses acetoacetate to form acyl-CoAs in the cytosol. Glucose-stimulated human islets released insulin similarly to rat islets but formed much more acetoacetate. β-Hydroxybutyrate augmented insulin secretion in human islets. This information supports previous data that indicate beta cells can use a pathway involving succinyl-CoA:3-ketoacid-CoA transferase and acetoacetyl-CoA synthetase to synthesize and use acetoacetate and suggests human islets may use this pathway more than PC and citrate to form cytosolic acyl-CoAs.     

Studies of insulin secretory responses and of arachidonic acid incorporation into phospholipids of stably transfected insulinoma cells that overexpress group VIA phospholipase A2 (iPLA2beta ) indicate a signaling rather than a housekeeping role for iPLA2beta

A cytosolic 84-kDa group VIA phospholipase A(2) (iPLA(2)beta) that does not require Ca(2+) for catalysis has been cloned from several sources, including rat and human pancreatic islet beta-cells and murine P388D1 cells. Many potential iPLA(2)beta functions have been proposed, including a signaling role in beta-cell insulin secretion and a role in generating lysophosphatidylcholine acceptors for arachidonic acid incorporation into P388D1 cell phosphatidylcholine (PC). Proposals for iPLA(2)beta function rest in part on effects of inhibiting iPLA(2)beta activity with a bromoenol lactone (BEL) suicide substrate, but BEL also inhibits phosphatidate phosphohydrolase-1 and a group VIB phospholipase A(2). Manipulation of iPLA(2)beta expression by molecular biologic means is an alternative approach to study iPLA(2)beta functions, and we have used a retroviral construct containing iPLA(2)beta cDNA to prepare two INS-1 insulinoma cell clonal lines that stably overexpress iPLA(2)beta. Compared with parental INS-1 cells or cells transfected with empty vector, both iPLA(2)beta-overexpressing lines exhibit amplified insulin secretory responses to glucose and cAMP-elevating agents, and BEL substantially attenuates stimulated secretion. Electrospray ionization mass spectrometric analyses of arachidonic acid incorporation into INS-1 cell PC indicate that neither overexpression nor inhibition of iPLA(2)beta affects the rate or extent of this process in INS-1 cells. Immunocytofluorescence studies with antibodies directed against iPLA(2)beta indicate that cAMP-elevating agents increase perinuclear fluorescence in INS-1 cells, suggesting that iPLA(2)beta associates with nuclei. These studies are more consistent with a signaling than with a housekeeping role for iPLA(2)beta in insulin-secreting beta-cells.     

A K ATP channel-dependent pathway within alpha cells regulates glucagon release from both rodent and human islets of Langerhans

Glucagon, secreted from pancreatic islet alpha cells, stimulates gluconeogenesis and liver glycogen breakdown. The mechanism regulating glucagon release is debated, and variously attributed to neuronal control, paracrine control by neighbouring beta cells, or to an intrinsic glucose sensing by the alpha cells themselves. We examined hormone secretion and Ca(2+) responses of alpha and beta cells within intact rodent and human islets. Glucose-dependent suppression of glucagon release persisted when paracrine GABA or Zn(2+) signalling was blocked, but was reversed by low concentrations (1-20 muM) of the ATP-sensitive K(+) (KATP) channel opener diazoxide, which had no effect on insulin release or beta cell responses. This effect was prevented by the KATP channel blocker tolbutamide (100 muM). Higher diazoxide concentrations (>/=30 muM) decreased glucagon and insulin secretion, and alpha- and beta-cell Ca(2+) responses, in parallel. In the absence of glucose, tolbutamide at low concentrations (<1 muM) stimulated glucagon secretion, whereas high concentrations (>10 muM) were inhibitory. In the presence of a maximally inhibitory concentration of tolbutamide (0.5 mM), glucose had no additional suppressive effect. Downstream of the KATP channel, inhibition of voltage-gated Na(+) (TTX) and N-type Ca(2+) channels (omega-conotoxin), but not L-type Ca(2+) channels (nifedipine), prevented glucagon secretion. Both the N-type Ca(2+) channels and alpha-cell exocytosis were inactivated at depolarised membrane potentials. Rodent and human glucagon secretion is regulated by an alpha-cell KATP channel-dependent mechanism. We propose that elevated glucose reduces electrical activity and exocytosis via depolarisation-induced inactivation of ion channels involved in action potential firing and secretion.     

The role of ion channels in insulin secretion.

Ion channels in beta cells regulate electrical and secretory activity in response to metabolic, pharmacologic, or neural signals by controlling the permeability to K+ and Ca2+. The ATP-sensitive K+ channels act as a switch that responds to fuel secretagogues or sulfonylureas to initiate depolarization. This depolarization opens voltage-dependent calcium channels (VDCC) to increase the amplitude of free cytosolic Ca2+ levels ([Ca2+]i), which triggers exocytosis. Acetyl choline and vasopressin (VP) both potentiate the acute effects of glucose on insulin secretion by generating inositol 1,4,5-trisphosphate to release intracellular Ca2+; VP also potentiates sustained insulin secretion by effects on depolarization. In contrast, inhibitors of insulin secretion decrease [Ca2+]i by either hyperpolarizing the beta cell or by receptor-mediated, G-protein-coupled effects to decrease VDCC activity. Repolarization is initiated by voltage- and Ca(2+)-activated K+ channels. A human insulinoma voltage-dependent K+ channel cDNA was recently cloned and two types of alpha 1 subunits of the VDCC have been identified in insulin-secreting cell lines. Determining how ion channels regulate insulin secretion in normal and diabetic beta cells should provide pathophysiologic insight into the beta cell signal transduction defect characteristic of non-insulin dependent diabetes (NIDDM).     

Homocysteine and other structurally-diverse amino thiols can alter pancreatic beta cell function without evoking cellular damage

Homocysteine and related amino thiols, homocysteic acid, cysteic acid, homocysteine sulphinic acid and cysteine sulphinic acid have been labelled as neurotoxins. Homocysteine thiolactone, a metabolic derivative of homocysteine, is cytotoxic to endothelial cells and other cell lineages. Since pancreatic beta cells share many phenotypic similarities with neuronal cells, the present study uses clonal pancreatic BRIN-BD11 cells to investigate possible detrimental effects of these amino thiols on insulin secretion and pancreatic beta cell function. Insulin secretion was concentration-dependently inhibited at both basal (1.1 mM) and stimulatory (16.7 mM) glucose by homocysteine, homocysteine thiolactone and homocysteine sulphinic acid. Cysteic acid concentration-dependently inhibited insulin secretion at 16.7 mM glucose. Cell viability was not compromised by any of the amino thiols. Insulin secretory responses to alanine were inhibited by homocysteine, homocysteine thiolactone, homocysteic acid and cysteic acid. Insulin secretion in the presence of elevated Ca(2+) and forskolin were lowered by all amino thiols, except homocysteic acid. The secretory responsiveness to PMA, GLP-1 and KCl were only impaired in the presence of homocysteine and homocysteine thiolactone. These findings indicate that homocysteine, homocysteine thiolactone and, to a lesser extent, other amino thiols cause dysfunctional insulin secretion from pancreatic beta cells.     

Ascorbic acid recycling by cultured beta cells: effects of increased glucose metabolism

Ascorbic acid is necessary for optimal insulin secretion from pancreatic islets. We evaluated ascorbate recycling and whether it is impaired by increased glucose metabolism in the rat beta-cell line INS-1. INS-1 cells, engineered with the potential for overexpression of glucokinase under the control of a tetracycline-inducible gene expression system, took up and reduced dehydroascorbic acid to ascorbate in a concentration-dependent manner that was optimal in the presence of physiologic D-glucose concentrations. Ascorbate uptake did not affect intracellular GSH concentrations. Whereas depletion of GSH in culture to levels about 25% of normal also did not affect the ability of the cells to reduce dehydroascorbic acid, more severe acute GSH depletion to less than 10% of normal levels did impair dehydroascorbic acid reduction. Culture of inducible cells in 11.8 mM D-glucose and doxycycline for 48 h enhanced glucokinase activity, increased glucose utilization, abolished D-glucose-dependent insulin secretion, and increased generation of reactive oxygen species. The latter may have contributed to subsequent decreases in the ability of the cells both to maintain intracellular ascorbate and to recycle it from dehydroascorbic acid. Cultured beta cells have a high capacity to recycle ascorbate, but this is sensitive to oxidant stress generated by increased glucose metabolism due to culture in high glucose concentrations and increased glucokinase expression. Impaired ascorbate recycling as a result of increased glucose metabolism may have implications for the role of ascorbate in insulin secretion in diabetes mellitus and may partially explain glucose toxicity in beta cells.     

BMP4-BMPR1A signaling in beta cells is required for and augments glucose-stimulated insulin secretion

Impaired glucose-stimulated insulin secretion (GSIS) and perturbed proinsulin processing are hallmarks of beta cell dysfunction in type 2 diabetes. Signals that can preserve and/or enhance beta cell function are therefore of great therapeutic interest. Here we show that bone morphogenetic protein 4 (Bmp4) and its high-affinity receptor, Bmpr1a, are expressed in beta cells. Mice with attenuated BMPR1A signaling in beta cells show decreased expression of key genes involved in insulin gene expression, proinsulin processing, glucose sensing, secretion stimulus coupling, incretin signaling, and insulin exocytosis and develop diabetes due to impaired insulin secretion. We also show that transgenic expression of Bmp4 in beta cells enhances GSIS and glucose clearance and that systemic administration of BMP4 protein to adult mice significantly stimulates GSIS and ameliorates glucose tolerance in a mouse model of glucose intolerance. Thus, BMP4-BMPR1A signaling in beta cells plays a key role in GSIS.     

Blockade of cannabinoid 1 receptor improves glucose responsiveness in pancreatic beta cells

Cannabinoid 1 receptors (CB1Rs) are expressed in peripheral tissues, including islets of Langerhans, where their function(s) is under scrutiny. Using mouse β‐cell lines, human islets and CB1R‐null (CB1R^?/?) mice, we have now investigated the role of CB1Rs in modulating β‐cell function and glucose responsiveness. Synthetic CB1R agonists diminished GLP‐1‐mediated cAMP accumulation and insulin secretion as well as glucose‐stimulated insulin secretion in mouse β‐cell lines and human islets. In addition, silencing CB1R in mouse β cells resulted in an increased expression of pro‐insulin, glucokinase (GCK) and glucose transporter 2 (GLUT2), but this increase was lost in β cells lacking insulin receptor. Furthermore, CB1R^?/? mice had increased pro‐insulin, GCK and GLUT2 expression in β cells. Our results suggest that CB1R signalling in pancreatic islets may be harnessed to improve β‐cell glucose responsiveness and preserve their function. Thus, our findings further support that blocking peripheral CB1Rs would be beneficial to β‐cell function in type 2 diabetes.