European Review for Medical and Pharmacological Sciences 2014; 18: 2215-2227 Metabolic signaling of insulin secretion by pancreatic β-cell and its derangement in type 2 diabetes

C.-Y. ZOU, Y. GONG, J. LIANG

Department of Endocrinology, Xuzhou Central Hospital, Xuzhou Clinical School of Xuzhou Medical College; The Affiliated Xuzhou Hospital of Medical College of Southeast University, Jiangsu, China; Xuzhou Institute of Medical Sciences, Xuzhou Institute of Diabetes, Xuzhou, Jiangsu, China

Introduction Abstract. Pancreatic β-c ell is responsible for insulin secretion in response to the avail- ability of nutrients. Type 2 diabetes mellitus Type 2 diabetes mellitus (T2D) is the manife- (T2D) is the result of pancreatic β-c ell failure station of pancreatic islet β-cell failure to supply to supply sufficient amount of insulin accom- sufficient amount of insulin and decreased sensiti- panied with decreased sensitivity of the body vity of the body tissues to insulin. It results from tissues to respond to insulin. The insulin se- genetic susceptibility and epigenetic changes in cretion apparatus of β -cell is uniquely equipped with multiple metabolic and signal- the context of toxic environmental factors such as 1-3 ing steps that are under rigorous control. The malnutrition and reduced physical activity . The metabolic machinery of β -cell is designed to islet of Langerhans is a key fuel sensing micro-or- sense the fluctuations in blood glucose level gan that constantly adjusts the release of insulin in and supply insulin accordingly to the needs of response to the levels and stimulus strength of nu- body. Besides glucose, amino acids including trients and hormonal factors. The insulin secretion glutamine and leucine and also fatty acids are apparatus of -cells is uniquely equipped with known to either stimulate the β -cells directly β or potentiate the glucose stimulated insulin multiple metabolic and signaling steps that are un- secretion (GSIS) response. Glucose metabo- der rigorous control. The glucose response of β- lism dependent GSIS is linked with the pro- cell to secrete insulin is considered to be highest in + duction of ATP that is needed for K ATP chan- comparison to other calorigenic nutrient secreto- nel inhibition and influx of calcium, necessary 4,5 gogues . Thus, the metabolic machinery of β-cell for insulin granule exocytosis. Besides glu- is designed to sense the fluctuations in blood glu- cose metabolism, amino acid metabolism and lipid metabolism derived metabolites mediate cose level and supplies insulin accordingly to the needs of the body5,6. Besides glucose, some amino the optimal glucose response of β -cells to se- crete insulin. Metabolites derived from nutri- acids including glutamine and leucine and also ent secretagogues that directly or indirectly fatty acids are known to either stimulate the β-cel- participate in the enhancement of GSIS are ls directly or potentiate the glucose stimulated in- considered as metabolic coupling factors. In sulin secretion (GSIS) response5,7. The early pre- this review, we will discuss the regulation of absorptive phase of insulin release, seen within insulin secretion by β -cell keeping the recent developments in metabolic signaling in focus. few minutes after food ingestion, is due to the pa- The relevant metabolic pathways in pancreat- rasympathetic nerves supplying the islets8. ic β -cell and their role in the control of fuel- Unlike in most cell types where activation of an stimulated insulin secretion will be reviewed energy consuming biological process (e.g., con- to arrive at a consensus picture with respect traction) lowers the ATP/ADP ratio, which in turn to the metabolic signaling of insulin secre- promotes cellular metabolism to produce ATP, in tion. β-cell metabolic activation is primarily driven by (fuel) availability9,10, rather than as a se- condary effect to enhanced insulin release11. Glu- Key Words: cose metabolism in β-cell is linked to the produc- tion of ATP and a rise in the cytoplasmic ATP/ADP Pancreatic -cell, Insulin secretion, Type 2 diabetes + β ratio that is needed for K ATP channel inhibition and mellitus.RETRACTED depolarization of the plasma membrane. The im-

Corresponding Author: Jun Liang, MD, Ph.D; e-mail: [email protected] 2215 C.-y. Zou, Y. Gong, J. Liang

+ portance of K ATP channel closure for opening up peak initially, followed by a slowly developing se- the voltage dependent L-type calcium channels in cond phase. The first phase secretion is reduced in the plasma membrane with the resultant Ca2+ in- prediabetes, whereas it is almost completely aboli- flux, as one of the primary events for insulin granu- shed in T2D along with significantly reduced se- le exocytosis, is well established2,12,13 (Figure 1). cond-phase secretion1. However, the biphasic pat- + Glucose metabolism driven K ATP channel inhibi- tern of insulin secretion probably does not exist in tion has also been implicated in the regulation of - vivo under physiological conditions, where gluco- 14,15 β cell mass . The ability of the β-cell to respond to se concentration does not rise in a step-wise man- the alterations in the blood glucose levels in the ner. However, this feature observed in vitro, helps (patho)physiological range (3 to 16 mM) is accom- in understanding the biochemical basis of fuel-in- plished because of the affinities of two key regula- duced insulin secretion and in identifying predia- tory proteins for glucose. These are Glut-1 and betes condition. The first phase corresponds to the Glut-2 glucose transporters that have high Km exocytosis of a small number of the secretory gra- (~17 mM), in human and in rodent β-cells, respec- nules already docked on the plasma membrane tively, which rapidly equilibrate external and inter- that release insulin, whereas the second phase in- nal glucose and ( IV), the volves mobilization of granules from the storage rate limiting that catalyzes the first step of pool. There are 10,000-12,000 insulin granules in glycolysis, which has a Km of ~8 mM for gluco- a β-cell, with a small pool of them (~500) docked se16. The Glut-1/2-Glucokinase tandem ensures a to the plasma membrane, of which 50-100 granu- steady increase in glycolysis and ATP production les, considered as the ready-releasable pool, are in the β-cell with increasing blood glucose levels tethered to the membrane in close association with and, thus, the glucose concentration dependent in- Ca2+ channels and contribute to the rapid first pha- 17 25 sulin secretion response . GSIS in β-cells is achie- se secretion . The ready-releasable pool is reple- ved by a tight link between glycolysis and mito- nished with fresh granules from the storage pool, chondrial metabolism for the quantitative direction which eventually contribute to the second sustai- of glucose carbons into mitochondria, due to the ned phase of secretion. very low expression of lactate dehydrogenase18,19. The molecular basis of how granules pools are The exocytotic process in the β-cell is orchestra- linked to phasic secretion is currently being ted by several components, in particular Ca2+ ions worked out. Recent studies showed that in human and exocytotic effector proteins located at the surfa- islets, nascent insulin granules contribute to first ces of secretory granules and plasma membranes20, phase and the mature granules to the second phase which facilitate the fusion of the insulin containing insulin secretion26. Different pools of insulin gra- large dense-core vesicles with the plasma membra- nules that are functionally distinct have been de- ne. Besides, synaptic-like micro vesicles (SLMV) scribed in β-cells. Regions of β-cells with preas- are also present in β-cells and these vesicles contain sembled soluble NSF-attachment protein receptor small molecules like γ-aminobutyric acid (GABA) (SNARE) proteins showed fast exocytosis in re- and may release their contents – in a manner similar sponse to rise in Ca2+, but in regions without to that of synaptic vesicles of neurons21,22. However, preassembled SNAREs, the exocytosis is slower21. not much is known about how fuel stimuli influen- Whether these granule pools related to biphasic ce the release of SLMV. secretion is uncertain. Recently, the model of pha- In this review, we will specifically discuss the sic secretion with respect to granule pools has regulation of insulin secretion by fuel stimuli in been questioned24. The current model proposes the β-cell keeping the recent developments in me- that first phase insulin secretion results from a rea- tabolic signaling in focus. The relevant metabolic dily-releasable pool composed of granules docked pathways and their role in the control of nutrient to the plasma membrane, whereas the second pha- stimulated insulin secretion will be reviewed to se results from a reserve pool of granules located arrive at a consensus picture with respect to the farther away that are recruited upon stimulation, biochemical basis of insulin secretion promoted docked, and followed by fusion with the plasma by glucose, amino acids and fatty acids. membrane. In a new currently accepted model, in- sulin granules are recruited upon β-cell stimula- Biphasic modes of insulin secretion tion and immediately fused to the plasma mem- A step rise in glucose concentration induces the brane, in both the phases. This model promotes release of insulin in a biphasic pattern, both in vi- the idea that the second phase secretion actually tro andRETRACTED in vivo23,24, consisting of a rapid 3-10 min consists of iteration of the first phase.

2216 Metabolic signaling of insulin secretion by pancreatic β-cell

Figure 1. Fuel metabolism and production of metabolic coupling factors in the β-Cell. Glucose, fatty acids and glutamine are metabolized through the Krebs cycle, pyruvate cycle and triglyceride/ fatty acid (TG/FA) cycle in the β-cell. Glutamine is con- verted by glutaminase in cytosol to glutamate, which is oxidized by glutamate dehydrogenase (GDH) in the mitochondria to produce α-ketoglutarate that enters Krebs cycle. Further oxidation of α-ketoglutarate in Krebs cycle leads to production of GTP, which inhibits GDH. Generation of α-ketoglutarate by GDH functions as anaplerosis as it provides substrate for Krebs cycle. Reducing equivalents (NADH) produced by Krebs cycle are oxidized by electron transport chain (ETC) to generate ATP and as a side , reactive oxygen species (ROS). ROS can directly stimulate insulin exocytosis. ATP exits mitochondria 2+ and in cytosol, it inhibits K+ATP channel and triggers Ca influx and insulin granule exocytosis. ATP is also used by adenylate cyclase to produce cyclic AMP (cAMP), when activated in response to stimuli like binding of GLP1 to its Gs-coupled receptor. cAMP activates exchange protein directly activated by cyclic AMP (EPAC2) and also protein -A (PKA), both of these promote exocytosis. Fatty acids are activated to fatty acyl-CoA by acyl-CoA synthase, long chain present on the sur- face of different subcellular membranes. Under conditions of low glucose, fatty acyl-CoAs are oxidized to generate acetyl- CoA, which enters Krebs cycle at citrate synthase step. Fatty acyl-CoAs are channeled to the TG/FA cycle under conditions of increased glucose availability, when β-oxidation is reduced. Glucose enters β-cell via Glut1 or Glut 2 transporters (in human or rodent, respectively) and enters glycolysis after its conversion to glucose-6-phosphate by glucokinase (GK); glycolysis gives ri- se to pyruvate and acetyl-CoA, which enters Krebs cycle. Approximately 50% of pyruvate is also converted to oxaloacetate via the anaplerotic enzyme pyruvate carboxylase inside mitochondria. Pyruvate participates in pyruvate cycles exchanging substra- tes, viz., citrate, isocitrate and malate, with Krebs cycle. Transport of these metabolites from mitochondria to the cytoplam (ca- taplerosis) facilitates pyruvate cycles. NADPH produced by pyruvate cycles either directly or in combination with glutare- doxin, or via ROS formation through the action of NADPH oxidase, promotes insulin granule exocytosis. Dihydroxyacetone phosphate (DHAP), produced during glycolysis, is reduced to glycerol 3-phosphate using NADH, byglycerol-3-phosphate dehydrogenase. Glycerol-3-phosphate and fatty acyl-CoA react together to form lysophosphatidic acid and to enter TG/FA cy- cle. Malonyl-CoA, which is formed from acetyl-CoA, as part of pyruvate/citrate cycle, inhibits fatty acid oxidation and positi- vely influences TG/FA cycle by diverting fatty acyl-CoA into TG/FA cycle, with participating enzymes distributed in endopla- smic reticulum (ER), mitochondria, cytosol and plasma membrane. TG/FA cycle produces several lipid signaling molecules in- cluding monoacylglycerol (MAG), which stimulate insulin secretion. MAG activates Munc13-1, an exocytosis facilitating pro- tein. MAG can also be produced by the hydrolysis of diacylglycerol (DAG), produced at plasma membrane from phosphoino- sitides (PIP2) during the activation of Gq-coupled receptors like GPR40, the fatty acid receptor. Nutrient metabolism in β-cell generatesRETRACTED metabolic coupling factors (MCF) that positively influence insulin granule exocytosis at different steps.

2217 C.-y. Zou, Y. Gong, J. Liang

Metabolic coupling factors Signals from anaplerosis and cataplerosis Insulin secretion is a multi-component process Anaplerosis is the process that contributes to and different steps involved are influenced by meta- the replenishment of Krebs cycle intermediates. bolites produced during glucose, amino acid and li- Once elevated, certain Krebs cycle intermediates, pid metabolism. It has recently been suggested that like citrate, not only enhance the cycle activity the metabolic signals can be ‘early or late effec- catalytically, but also participate in additional tors’, on the basis of whether the corresponding af- metabolic pathways that lead to the production of fected step is an early event or late step in the pro- different MCF (e.g., malonyl-CoA, Glutamate, cess of insulin exocytosis27. A metabolic coupling NADPH) in the cytoplasm. This process is com- factor (MCF) can be a signal that contributes to the plemented by cataplerosis, which refers to the regulation of nutrient-stimulated insulin secretion exit of Krebs cycle intermediates from the mito- process, either by modulating the nutrient metaboli- chondrial matrix to cytoplasm. Various transpor- sm (early regulators) or by directly influencing the ters located on the mitochondrial inner membra- component(s) of exocytotic machinery (late regula- ne facilitate the transmitochondrial movement of tors). Table I gives examples of various MCF and the Krebs cycle intermediates, such as the di- and their proposed targets and roles in insulin secretion. tri-carboxylate carriers. In the cytoplam, some of Considering the importance of insulin secretion, β- the Krebs cycle intermediates (citrate, isocitrate, cell harbors metabolic pathways that generate these α-ketoglutarate and malate) participate in pyru- multiple MCFs to ensure proper insulin secretion. vate cycling processes that generate cytoplasmic As mentioned above, the relatively low affini- NADPH (Figure 1), an important MCF27. ties of Glut1 and 2 and glucokinase for glucose Pyruvate metabolism via pyruvate carboxylase 28,29 control the flux of glucose metabolism in the β- (PC), which is highly expressed in β-cells , is cell, which in turn dictates the rate and magnitude central to anaplerosis. Many studies have shown of insulin secretion in response to blood glucose that as compared to the rates of decarboxylation level. Glucose metabolism produces ATP, which and oxidation of pyruvate30, the rate of pyruvate 2+ closes the KATP channel with the resultant Ca in- carboxylation correlates well with the glucose flux that promotes insulin granule exocytosis. Pro- dose dependence of GSIS31,32. Studies using PC duction of ATP in mitochondria is facilitated by inhibitor phenylacetic acid32, RNAi knockdown the efficient transfer of glucose-derived NADH and overproduction of PC in INS-1 cells and from cytosol to mitochondria by the glycerol-3 islets clearly demonstrated the significance of PC phosphate and malate/aspartate shuttles. Activated in GSIS33-35. Besides formation of oxaloacetate mitochondrial metabolism, consisting of anaplero- by PC, the oxidative deamination of glutamate to sis and cataplerosis, is central for enhanced Krebs α-ketoglutarate by mitochondrial glutamate cycle activity and ATP production and also for the dehydrogenase (GDH) is also a significant con- generation of additional MCF that take part in the tributor to anaplerosis by amino acids as discus- amplification of GSIS (Figure 1). sed below36. Mitochondrial GDH is important for

Table I. Metabolic coupling factors implicated in insulin secretion.

MCF Target of action Mode and site of action AMP AMPK Negative regulation of GSIS 2+ ATP KATP Ca signaling cAMP PKA, Epac2 Ca2+, channels, exocytosis proteins Citrate ACL, ACC Pyruvate cycling Diacylglycerol PKC, Munc13-1 Channels, exocytosis Fatty acyl-CoA Lipogenic enzymes TG/FA cycling Glutamate GDH Anaplerosis GTP GDH, GTP-SCS Anaplerosis, Ca2+ signaling 2+ 2+ Inositol poly-PO4 IP3 receptor, L-type Ca channel. Ca influx, exocytosis α-Ketoglutarate α-KG dehydrogenase, HIF1α-hydroxylase Exocytosis Malonyl-CoA CPT-1, FAS Fatty acyl group partitioning MgADP decrease SUR1 Ca2+ signaling Monoacylglycerol Munc13-1 Vesicle fusion, Exocytosis NADPH Glutaredoxin, Kv Exocytosis redox control, Ca2+ ROSRETRACTEDExocytosis proteins Exocytosis complex redox control

2218 Metabolic signaling of insulin secretion by pancreatic β-cell amino acid (glutamine plus leucine) induced in- ring a glucose tolerance test57, suggesting a role sulin secretion and gain of function mutation of for NNT and NADPH in GSIS. NNT-generated GDH is associated with a hyperinsulinemic hy- NADPH can be used by mitochondrial NADP-de- 37 poglycemic syndrome . pendent isocitrate dehydrogenase, which using α- ketoglutarate, can produce isocitrate, followed by Pyruvate cycles and cytosolic NADPH the export of isocitrate (or citrate) to the cytopla- There are four pyruvate cycling processes, viz., sm, for regenerating NADPH via pyruvate cycling pyruvate/citrate, pyruvate/malate, pyruvate/isocita- processes (Figure 1). There is strong evidence im- te and pyruvate/phosphoenol-pyruvate cycles, plicating a role for cytosolic NADPH in GSIS. which are critical for anaplerosis/cataplerosis-deri- Thus, NAPDH was shown to directly promote 36,38-40 ved signaling and MCF production and for exocytosis in patch-clamped β-cells (Ivarsson et producing NADPH in the cytosol using mitochon- al, 2005) and RNAi silencing of cytosolic NADPH drial NADH. Inasmuch as PC converts approxi- generating malic enzyme and isocitrate dehydro- 48-50 mately 50% of the pyruvate to OAA in β-cell mi- genase enzymes reduces GSIS . NADPH likely tochondria41 and citrate levels both in cytosol and targets glutaredoxin58 and voltage-dependent K+ mitochondria are elevated in proportion to glucose (Kv) channels59. Glutaredoxin is important for the concentration, and because pharmacological inter- post-translational modifications of exocytotic pro- vening at different steps of the pyruvate/citrate cy- teins and the Kv channel β-subunit, which can cle causes reduced GSIS in β-cells, it has been bind NADPH (Figure 1), is thought to be the sen- suggested earlier that pyruvate/citrate cycling is sor of intracellular redox potential that in turn re- quantitatively important in the production of MCF gulates the channel and, thus, controls glucose-sti- 32,38,42 59 and in GSIS . This cycle may also be linked mulated action potentials in β-cells . to metabolic oscillations and, thus, contribute to the pulsatile insulin release from -cells that paral- Pyruvate cycling and malonyl-CoA 2+ β lels oscillations in [Ca ]i, ATP, NAD(P)H and ci- Acetyl-CoA formed by ATP-citrate (ACL) trate levels43-45. In the pyruvate/isocitrate cycle, cy- is carboxylated by acetyl-CoA carboxylase (ACC) tosolic isocitrate dehydrogenase (cICDH) converts to malonyl-CoA, which has signaling role in the 10 isocitrate to α-ketoglutarate and NADP to control of GSIS by inhibiting carnitine palmitoyl- NADPH. The precise role of this cycle in GSIS is -1 (CPT-1). Inhibition of CPT-1 diverts uncertain as cICDH RNAi-knockdown studies fatty acids from β-oxidation to lipid synthesis, and were shown to both decrease GSIS46, as well as in- some of these lipids play important role in the am- crease GSIS47. It has been shown that knockdown plification of GSIS10. Build up of fatty acyl-CoA of cytosolic malic enzyme, which takes part in the due to CPT-1 inhibition, can lead to the activation 48-50 60 + 61 pyruvate/malate cycle, reduces GSIS . The of -C enzymes and K ATP channel fourth pyruvate cycle, the “pyruvate/phosphoenol- and also stimulate GSIS62. The view that malonyl- pyruvate” cycle51 is linked to ATP formation in the CoA/CPT-1 interaction is needed for optimal GSIS cytoplasm and GTP hydrolysis in the mitochon- was supported by the studies showing impaired dria. The relative importance of these cycles is GSIS in INS cells overexpressing a mutant CPT-1 currently controversial38,48,52-54. that is insensitive to malonyl-CoA63. However, ove- Glucose stimulation of -cell causes a rapid rise rexpression of cytosol-directed malonyl-CoA de- 55 β in cytosolic NADPH that occurs even prior to β- carboxylase appears to lower GSIS only in the pre- cell depolarization, increased intracellular Ca2+ sence of fatty acids64,65. Fatty acids also directly and insulin granule fusion to plasma membrane56. bind to cell surface GPCRs, GPR40 and GPR120 The pentose phosphate shunt pathway, which can and stimulate GSIS66. also generate NADPH is unlikely to be important for GSIS, as flux through this pathway in β-cell is Glutamate, GDH and GTP much lower, thus suggesting pyruvate cycling Inasmuch as glucose stimulation of islets is ac- pathways to be the major source of cytosolic companied by augmented glutamate levels sup- NADPH in these cells. NADPH is also formed in ports the view that glutamate is an MCF67,68. In ad- + mitochondrial matrix by the reduction of NADP dition, reduction of β-cell glutamate levels by glu- by nicotinamide nucleotide transhydrogenase tamate decarboxylase overexpression reduces in- (NNT) and by mitochondrial NADP-dependent sulin secretion. Islets from β-cell-specific GDH isocitrate dehydrogenase. NNT mutant mice show KO mice display reduced GSIS, indicating gluta- glucoseRETRACTED intolerance and reduced insulinemia du- mate metabolism via GDH reaction is necessary

2219 C.-y. Zou, Y. Gong, J. Liang

68 for the stimulation of GSIS . In mitochondria, al- enhanced FFA β-oxidation and reduces lipolysis, losteric inhibition of glutamate dehydrogenase by thereby reducing the production of lipid signals GTP inhibits oxidative deamination of glutamate, for the amplification of GSIS77. Recent work indi- thereby negatively affecting insulin secretion (Fi- cated that AMPK activation causes metabolic de- gure 1; Table I). Gain of function mutations of celeration in the β-cell by slowing down the glu- GDH, which render GDH to be less susceptible to cose metabolism and, thus, insulin secretion at GTP inhibition are associated with hyperinsuline- glucose concentrations < 10 mM, whereas at hi- mia36,69. Both cytosolic and mitochondrial GTP gher glucose levels above 16 mM, this metabolic has an effect on insulin secretion. Mitochondrial deceleration effect is absent78, suggesting that GTP is predominantly produced by the GTP-spe- AMPK activation offers protection to β-cells from cific succinyl-CoA synthase (GTP-SCS), whereas the toxicity of fuel surplus and exhaustive oversti- in cytosol nucleoside diphosphate kinase is re- mulation78. Recent studies suggested that AMPK sponsible for GTP formation. On the other hand, also likely controls the activity of SIRT1, which mitochondrial GTP has also been shown to modu- also is known to regulate insulin secretion in β- late mitochondrial metabolism and Ca2+ and to po- cells76,78. LKB1/AMPK enzymes have been impli- 70 76 sitively influence GSIS in β-cells . Thus, RNAi- cated as negative regulators of insulin secretion . knockdown of GTP-SCS lowers β-cell ATP levels and reduces GSIS70. GTPase enzymes associated Reactive Oxygen Species (ROS) with insulin exocytosis utilize cytosolic GTP to ROS include superoxide and hydroxyl radicals promote secretion and it has been shown that GTP and hydrogen peroxide (H2O2) and these are pro- levels in cytosol rise at high glucose concentra- duced physiologically in many cells, during nu- tion. Besides, recent evidence strongly implicated trient oxidation. In β-cells, it has been proposed a role for cGMP in stimulating insulin secretion. that ROS acts as MCF for promoting GSIS79,80. Thus, activation of AMPA receptors by glutamate However, chronic production of elevated levels of + can lead to elevated cGMP, which inhibits K ATP ROS can be detrimental for β-cell function. Mito- channel and stimulate secretion71 and the presence chondrial electron transport chain components of guanylate cyclases, which produce cGMP from Complex-I and Complex-III are the major site for 72 36,81 GTP in β-cells has been confirmed . ROS formation . Besides mitochondria, ROS can also be produced by plasma membrane elec- ATP, ADP, AMP and AMPK tron transporting NADPH oxidase complex79 and + Inasmuch as K ATP channel inhibition by ATP in peroxisomes. It has been suggested that in β- is a necessary step for insulin secretion, altera- cells peroxisomal fatty acid oxidation is the major + tions in ATP/ADP ratio in the vicinity of K ATP source for H2O2, which leads to β-cell dysfunc- channel are relevant and are regulated by adeny- tion and death whereas mitochondrial β-oxidation late kinase-1, which likely associates with Kir6.2 does not contribute significantly to ROS82. In mi- + 73 subunit of K ATP channel . Adenine nucleotides, tochondria, nicotinamide nucleotide transhydro- whose cellular levels are dependent on the meta- genase, which produces NADPH, contributes to 57 bolic state of the β-cell, directly influence the in- free radical detoxification and altered activity of sulin exocytotic machinery at the plasma mem- this enzyme is associated with proportional chan- brane. Overall regulation of K+ channels by the ges in insulin secretion57,83. adenine nucleotides depends on the net inhibitory The rise in influx of Ca2+, while necessary for effect of ATP and the net activating effect of insulin granule exocytosis, can also cause MgADP on SUR1 component of the channel74. NADPH oxidase activation resulting in increased 79 Besides ATP/ADP ratio, ATP/AMP ratio is also production of H2O2 . Attenuation of ROS signal important in the overall regulation of insulin secre- is mainly accomplished by superoxide dismutase, tion in β-cells as this influences the activity of glutathione peroxidase, thioredoxin and peroxire- AMP activated protein kinase (AMPK), which is a doxins in human β-cell whereas these enzymes master controller of cellular energy metabolism75. are expressed at low levels in rodent islets79. The

There have been several excellent reviews on positive effects of ROS, in particular H2O2 and 76 – 2+ AMPK and we focus here exclusively on the re- O2 on GSIS include enhancement of Ca in- cent developments. Cellular AMP levels are regu- flux80 and activation of volume regulated anion lated via its utilization by and th- channels84 even though the precise targets are not rough its formation during fatty acid and amino identified. The detrimental effects of ROS are ac- acidRETRACTED activation. Activation of AMPK triggers tivation of mitochondrial UCP2, oxidative modi-

2220 Metabolic signaling of insulin secretion by pancreatic β-cell

77 fication and inhibition of aconitase, adenine nu- insulin secretion in β-cells . Thus, agents that cleotide and glyceraldehyde-3- block TG/FA cycle at different steps are known phosphate dehydrogenase, and oxidation of mito- to reduce GSIS78,89-91. It is important to note that chondrial cardiolipin resulting in reduced ATP le- lipid intermediates of lipogenic arm have various vels, decreased insulin secretion and apoptosis 36, signaling functions and disruption of the any of 79, 81. Production of ROS is likely to be elevated in the involved enzymes has varying effects on the pancreatic islets from T2D patients than from GSIS (Figure 2). Lysophosphatidic acid is known nondiabetic subjects85. Thus, ROS can have dual to affect Ca2+ influx while phosphatidic acid was function in the regulation of GSIS (Figure 1) – thought to directly influence exocytosis77. Howe- both in stimulating secretion as well as in cau- ver, agents that block lipolysis arm of the TG/FA 27 sing β-cell dysfunction, when chronically produ- cycle have been shown to reduce GSIS . ced in high amounts. Recent studies demonstrated that ATGL produ- Glucose stimulation of β-cell has been shown ces 1,3- and 2,3-DAG species, which have no to cause a rapid turnover of inositol containing li- known signaling function, rather than the signaling pids, particularly the plasma membrane associa- competent sn1,2-DAG that activates C-kinase ted phosphatidylinositol-4,5-bisphosphate (PIP2) enzymes. In addition, intracellularly, HSL prima- and -3,4,5-trisphosphate (PIP3) and it has been rily hydrolyzes DAG at position 1 or 3, thus produ- suggested that these polyphosphoinositides likely cing 1- or, 2-MAG (Figure 2). It has been proposed play a facilitating role in GSIS, probably by con- that sn1,2-DAG is the lipid signal for insulin secre- trolling intracellular Ca2+ levels and also DAG le- tion as this lipid can activate certain protein kinase- vels86. Recent studies implicated inositol-tripho- C isoenzymes and also activate the exocytotic pro- sphate and other inositol polyphosphates in the tein Munc13-1 in the β-cells. However, conside- regulation of GSIS87, probably by their direct ac- ring that conditions that lead to accumulation of tion on L-type Ca2+ channels. DAG generally cause a decrease in GSIS rather than increase suggests that MAG is the likely lipid Triglyceride/fatty acid cycling signal derived from lipolysis. Recent studies con- and lipid MCF signals firmed this as suppression of the MAG The triglyceride/ fatty acid (TG/FA) cycle con- ABHD6 in β-cells led to enhanced GSIS, both in sists of two segments – lipid synthesis and lipid vitro and in vivo and also added MAG stimulated breakdown and produces many lipid intermedia- GSIS in islets by activating Munc13-188. tes (Figure 2). Lipogenic arm of TG/FA cycle is Glucose stimulation of β-cells increases li- initiated by the fatty acid esterification of glyce- polysis92 and the release of FFA, which can acti- rol-3-phosphate, arising from the action of cyto- vate GPR40, the Gq-coupled FFA receptor lea- solic glycerol-3-phosphate dehydrogenase on ding to sn1,2-DAG production and subsequent glycolytic intermediate, dihydroxyacetone pho- activation of protein kinase C enzymes or protein sphate (Figures 1 and 2). This first step of esteri- kinase D93, that have been implicated in GSIS fication, catalyzed by glycerol-3-phosphate acyl- (Figure 2)94. The role of released FFA acting as transferase isoenzymes forms lysophosphatidic autocrine /paracrine signals in GSIS is yet to be acid (LPA), that is further converted to phospha- established. tidic acid by LPA acyltransferase, followed by Cyclic AMP and hormonal modulators: Seve- removal of phosphate by lipin to sn1,2-diacylgly- ral hormonal and neurotransmitter stimuli to β- cerol (DAG) and then conversion to triglyceride cells lead to elevated cyclic AMP levels via Gs- (TG) by DAG acyltransferase. TG, then, enters coupled GPCR activation, without influencing lipolytic arm, by sequential hydrolysis to produce intracellular Ca2+. Incretins like glucagon-like either sn2,3-DAG or sn1,3-DAG (by adipose tri- peptide-1 (GLP-1), which are released from inte- glyceride lipase), followed by the formation of stinal L-cells in response to high blood glucose either 2-monoacylglycerol (MAG) or 1-MAG (by potentiate insulin secretion by β-cell by stimula- hormone sensitive lipase) and finally glycerol ting production of cAMP, which activates protein and FA by the recently discovered / -hydrolase kinase A (PKA)-dependent and -independent me- 88 α β containing-6 (Figure 2) . TG in β-cells is stored chanisms of exocytosis, and KATP-channel as micro lipid droplets, distributed beneath the closure95. cAMP also acts via PKA-independent cell membrane (Pinnick et al, 2010). Many of the mechanism mediated by the cAMP-sensing pro- intermediates of TG/FA cycle show signaling tein Epac2 (Figure 1), which has been shown to functionsRETRACTED and some are known to participate in be a target of sulphonylureas96.

2221 C.-y. Zou, Y. Gong, J. Liang

Figure 2. Lipid signals and insulin secretion. Continuous synthesis and breakdown of triglycerides and other glycerolipids is termed as triglyceride/fatty acid (TG/FA) cycle. This metabolic pathway starts with the condensation of glucose derived glyce- rol-3-phosphate with fatty acyl-CoA (FA-CoA), produced from fatty acid by acyl-CoA synthase, long chain (ACSL). This first step of TG/FA cycle is catalyzed by glycerol-3-phosphate acyltransferase isoenzymes located on mitochondria and endopla- smic reticulum, to form lysophosphatidic acid. Lysophosphatidic acid is further acylated to phosphatidic acid on ER by ly- sophosphatidic acid acyltransferase (LPAAT) isoenzymes. Formation of 1,2-diacylglycerol (1,2-DAG) from phosphatidic acid is catalyzed by lipin enzymes, which are distributed in lipid droplets, ER, nucleus and cytosol. Reversal of this reaction, i.e., formation of phosphatidic acid from 1,2-DAG is conducted by several DAG kinase (DAGK) isoenzymes, distributed throu- ghout the cell. TG is formed by the acylation of 1,2-DAG, the final step of lipogenic segment, by diacylglycerol acyltransfera- ses-1 & 2 (DGAT) present on ER and lipid droplets. Lipolysis of TG is initiated by adipose triglyceride lipase (ATGL), which hydrolyzes TG to DAG. Depending on its activation by Comparative Gene Identification 58 protein (CGI-58) on the surface of lipid droplets, ATGL generates either 2,3-DAG (with CGI58) or 1,3-DAG (in the absence of CGI58) from TG. DGAT can use 1,3 DAG also to form TG. Hormone sensitive lipase (HSL) hydrolyzes 1,3-DAG and 2,3-DAG to 1-monoacylglycerol (MAG) and 2-MAG, respectively. MAG can be converted to lysophosphatidic acid by acylglycerol kinase (AGK). Plasma membrane associated α/β-domain containing hydrolase-6 (ABHD6) catalyzes the hydrolysis of MAG to glycerol and FFA in β-cells. Gly- cerol is not further metabolized in the β-cells, due to the lack of , and leaves the cell via aquaglyceroporins. TG/FA cycle generates several lipid signals that promote insulin secretion. These include (1), fatty acyl-CoA, which targets + 2+ K ATP channels; (2), lysophosphatidic acid, which influences Ca levels; (3), phosphatidic acid, which affects exocytosis; (4), 1,2-DAG, which likely binds and activates exocytosis promoting protein Munc13-1, protein kinase-C (PKC) and protein kina- se-D (PKD); (5), 1-MAG, which activates Munc13-1 and (6), FA, which activates GPR40.

Insulin secretory defects in diabetes: Under tion ultimately culminates in T2D. Thus, it has conditions of excess fuel supply and insulin resi- been shown that in subjects who potentially deve- stance due to obesity, pancreatic β-cell from nor- lop T2D, there is an increase in blood insulin le- mal non-diabetic individuals responds by com- vels during the prediabetic stage, where normo- pensatory hypersecretion of insulin in order to glycemia is maintained and this is followed with maintain normoglycemia. Loss of this ability of time by a steady decline in circulating insulin le- β-cellsRETRACTED for compensatory elevated insulin secre- vels due to β-cell failure, associated with elevated

2222 Metabolic signaling of insulin secretion by pancreatic β-cell

–––––––––––––––––––-–– fasting glycemia above 5.5 mM. The mechanisms Acknowledgements involved in -cell compensation are not clear but β Dr. Jun Liang’s research was sponsored by Jiangsu Provin- animal studies implicated both -cell mass expan- cial Bureau of Health Foundation (H201356) & Internatio- β 17 sion as well as enhanced β-cell function . The nal Exchange Program and Jiangsu Six Talent Peaks Pro- elevated compensatory insulin secretion by pan- gram (2013-WSN-013), It was also supported by the creatic islets can be due to increased fuel (glucose Xuzhou Outstanding Medical Academic Leader project and and fatty acids) supply, increased growth factor a Xuzhou Science and Technology Grant (XM13B066, XZZD1242). and incretin signaling. Several monogenic forms of obesity and diabetes, including maturity-onset –––––––––––––––––––-–– diabetes of the young (MODY) have been descri- Conflict of interest bed97. However, T2D is a polygenic disease and The Authors declare that they have no conflict of interests. shows more complex genetics, in which varia- tions within multiple genes, each independently contributing some risk for disease development98. References The insulin secretory defect in T2D is multifacto- rial and likely involves reduced β-cell mass, im- 1) NOLAN CJ, DAMM P, PRENTKI M. Type 2 diabetes paired -cell glucose sensing, and defective -cell across generations: From pathophysiology to pre- β β vention and management. Lancet 2011; 378: 169- secretory machinery and MCF production. These 181. defects are not readily recognized in in vivo stu- 2) ASHCROFT FM, RORSMAN P. Diabetes mellitus and the dies, making it difficult to understand the specific beta cell: The last ten years. Cell 2012; 148: 1160- disease mechanisms coupled to T2D risk loci. A 1171. recent study in T2D patients revealed that such 3) MARTIN-GRONERT MS, OZANNE SE. Metabolic pro- T2D risk imposing genetic variants can affect gramming of insulin action and secretion. Dia- either glucose sensing, exocytosis or structural betes Obes Metab 2012; 14(Suppl 3): 29-39. elements of secretory machinery99. 4) NEWSHOLME P, KRAUSE M. Nutritional regulation of insulin secretion: implications for diabetes. Clin Biochem Rev 2012; 33: 35-47. 5) NOLAN CJ, PRENTKI M. The islet beta-cell: fuel re- Conclusions sponsive and vulnerable. Trends Endocrinol Metab 2008; 19: 285-291. Despite much knowledge and technological 6) MEGLASSON MD, MATSCHINSKY FM. Pancreatic islet advancement the precise mechanisms involved in glucose metabolism and regulation of insulin se- cretion. Diabetes Metab Rev 1986; 2: 163-214. GSIS are not entirely clear. This lack of under- 7) NOLAN CJ, MADIRAJU MS, DELGHINGARO-AUGUSTO V, standing stems from the complexity and degene- PEYOT ML, PRENTKI M. Fatty acid signaling in the racy of multiple metabolic pathways, that are beta-cell and insulin secretion. Diabetes 2006; linked to signaling processes to control GSIS. 55(Suppl 2): S16-23. Such high level of organization consisting of 8) RUIZ DE AZUA I, GAUTAM D, GUETTIER JM, WESS J. Novel structural proteins, metabolic enzymes, metaboli- insights into the function of beta-cell M3 muscarinic acetylcholine receptors: Therapeutic implications. tes, ion channels and metal ions, ensures the deli- Trends Endocrinol Metab 2011; 22: 74-80. very of the needed amounts of insulin into circu- 9) MATSCHINSKY FM. A lesson in metabolic regulation lation, as excess or low insulin can be detrimen- inspired by the glucokinase glucose sensor para- tal. Considering the multitude of signaling pro- digm. Diabetes 1996; 45: 223-241. cesses involved in GSIS, defects at one or more 10) PRENTKI M. New insights into pancreatic beta-cell steps of these pathways can contribute to the de- metabolic signaling in insulin secretion. Eur J En- docrinol 1996; 134: 272-286. velopment of T2D. Failure of β-cell can occur when any of these metabolic signaling pathways 11) PEYOT ML, GRAY JP, LAMONTAGNE J, SMITH PJ, HOLZ GG, MADIRAJU SR, PRENTKI M, HEART E. Glucagon- are compromised either because of genetic, envi- like peptide-1 induced signaling and insulin secre- ronmental or epigenetic factors. Further research tion do not drive fuel and energy metabolism in should focus on β-cell metabolic signaling me- primary rodent pancreatic beta-cells. PLoS One chanisms that are altered in T2D and to prevent 2009; 4: e6221. or reverse such pathological alterations in meta- 12) BENNETT K, JAMES C, HUSSAIN K. Pancreatic beta-cell katp channels: hypoglycaemia and hyperglycaemia. bolism. Identification of these pathways will help Rev Endocr Metab Disord 2010; 11: 157-163. in understanding the molecular basis of -cell β 13) NICHOLS CG, REMEDI MS. The diabetic beta-cell: hy- failure in diabetes and to discover new targets to perstimulated vs. hyperexcited. Diabetes Obes developRETRACTED antidiabetic drugs. Metab 2012; 14(Suppl 3): 129-135.

2223 C.-y. Zou, Y. Gong, J. Liang

14) PORAT S, WEINBERG-COREM N, TORNOVSKY-BABAEY S, 29) SCHUIT F, DE VOS A, FARFARI S, MOENS K, PIPELEERS D, SCHYR-BEN-HAROUSH R, HIJA A, STOLOVICH-RAIN M, BRUN T, PRENTKI M. Metabolic fate of glucose in pu- DADON D, GRANOT Z, BEN-HUR V, WHITE P, GIRARD rified islet cells. Glucose-regulated anaplerosis in CA, KARNI R, KAESTNER KH, ASHCROFT FM, MAGNUSON ß-cells. J Biol Chem 1997; 272: 18572-18579. MA, SAADA A, GRIMSBY J, GLASER B, DOR Y. Control of 30) SRINIVASAN M, CHOI CS, GHOSHAL P, PLISS L, PANDYA pancreatic beta cell regeneration by glucose me- JD, HILL D, CLINE G, PATEL MS. {beta}-cell-specific tabolism. Cell Metab 2011; 13: 440-449. pyruvate dehydrogenase deficiency impairs glu- 15) LEVITT HE, CYPHERT TJ, PASCOE JL, HOLLERN DA, ABRA- cose-stimulated insulin secretion. Am J Physiol HAM N, LUNDELL RJ, ROSA T, ROMANO LC, ZOU B, Endocrinol Metab 2010; 299: E910-917. O’DONNELL CP, STEWART AF, GARCIA-OCAÑA A, ALONSO 31) CLINE GW, LEPINE RL, PAPAS KK, KIBBEY RG, SHULMAN LC. Glucose stimulates human beta cell replica- GI. 13C NMR isotopomer analysis of anaplerotic tion in vivo in islets transplanted into NOD-severe pathways in INS-1 cells. J Biol Chem 2004; 279: combined immunodeficiency (SCID) mice. Dia- 44370-44375. betologia 2011; 54: 572-582. 32) FARFARI S, SCHULZ V, CORKEY B, PRENTKI M. Glucose- 16) MEGLASSON MD, MATSCHINSKY FM. New perspectives regulated anaplerosis and cataplerosis in pancre- on pancreatic islet glucokinase. Am J Physiol atic beta-cells: possible implication of a pyru- 1984; 246: E1-13. vate/citrate shuttle in insulin secretion. Diabetes 17) PRENTKI M, NOLAN CJ. Islet beta cell failure in type 2000; 49: 718-726. 2 diabetes. J Clin Invest 2006; 116: 1802-1812. 33) HASAN NM, LONGACRE MJ, STOKER SW, BOONSAEN T, 18) SEKINE N, CIRULLI V, REGAZZI R, BROWN LJ, GINE E, JITRAPAKDEE S, KENDRICK MA, WALLACE JC, MACDON- TAMARIT-RODRIGUEZ J, GIROTTI M, MARIE S, MACDONALD ALD MJ. Impaired anaplerosis and insulin secre- MJ, WOLLHEIM CB, RUTTER GA. Low lactate dehy- tion in insulinoma cells caused by small interfer- drogenase and high mitochondrial glycerol phos- ing RNA-mediated suppression of pyruvate car- phate dehydrogenase in pancreatic beta-cells. boxylase. J Biol Chem 2008; 283: 28048-28059. Potential role in nutrient sensing. J Biol Chem 34) XU J, HAN J, LONG YS, EPSTEIN PN, LIU YQ. The role 1994; 269: 4895-4902. of pyruvate carboxylase in insulin secretion and 19) SCHUIT F, VAN LOMMEL L, GRANVIK M, GOYVAERTS L, DE proliferation in rat pancreatic beta cells. Dia- FAUDEUR G, SCHRAENEN A, LEMAIRE K. Beta-cell- betologia 2008; 51: 2022-2030. specific gene repression: a mechanism to protect 35) MACDONALD MJ, LONGACRE MJ, LANGBERG EC, TIBELL against inappropriate or maladjusted insulin se- A, KENDRICK MA, FUKAO T, OSTENSON CG. De- cretion? Diabetes 2012; 61: 969-975. creased levels of metabolic enzymes in pancreat- 20) KWAN EP, GAISANO HY. Rescuing the subprime ic islets of patients with type 2 diabetes. Dia- meltdown in insulin exocytosis in diabetes. Ann N betologia 2009; 52: 1087-1091. Y Acad Sci 2009; 1152: 154-164. 36) MAECHLER P, LI N, CASIMIR M, VETTERLI L, FRIGERIO F, BRUN T. Role of mitochondria in beta-cell function 21) KASAI H, HATAKEYAMA H, OHNO M, TAKAHASHI N. Exo- cytosis in islet beta-cells. Adv Exp Med Biol 2010; and dysfunction. Adv Exp Med Biol 2010; 654: 654: 305-338. 193-216. 37) STANLEY CA. Two genetic forms of hyperinsulinemic 22) SUCKOW AT, CRAIGE B, FAUNDEZ V, CAIN WJ, CHESSLER SD. An AP-3-dependent mechanism drives hypoglycemia caused by dysregulation of gluta- synaptic-like microvesicle biogenesis in pancreat- mate dehydrogenase. Neurochem Int 2011; 59: ic islet beta-cells. Am J Physiol Endocrinol Metab 465-472. 2010; 299: E23-32. 38) SUGDEN MC, HOLNESS MJ. The pyruvate carboxy- lase-pyruvate dehydrogenase axis in islet pyru- 23) CURRY DL, BENNETT LL, GRODSKY GM. Dynamics of vate metabolism: going round in circles? Islets insulin secretion by the perfused rat pancreas. 2011; 3: 302-319. Endocrinology 1968; 83: 572-584. 39) MACDONALD MJ, FAHIEN LA, BROWN LJ, HASAN NM, 24) SEINO S, SHIBASAKI T, MINAMI K. Dynamics of insulin BUSS JD, KENDRICK MA. Perspective: emerging evi- secretion and the clinical implications for obesity dence for signaling roles of mitochondrial and diabetes. J Clin Invest 2011; 121: 2118-2125. anaplerotic products in insulin secretion. Am J 25) WANG Z, THURMOND DC. Mechanisms of biphasic Physiol Endocrinol Metab 2005; 288: E1-15. insulin-granule exocytosis—roles of the cy- 40) JENSEN MV, JOSEPH JW, RONNEBAUM SM, BURGESS SC, toskeleton, small GTPases and SNARE proteins. SHERRY AD, NEWGARD CB. Metabolic cycling in con- J Cell Sci 2009; 122: 893-903. trol of glucose-stimulated insulin secretion. Am J 26) MICHAEL DJ, XIONG W, GENG X, DRAIN P, CHOW RH. Physiol Endocrinol Metab 2008; 295: E1287-1297. Human insulin vesicle dynamics during pulsatile 41) FRANSSON U, ROSENGREN AH, SCHUIT FC, RENSTROM E, secretion. Diabetes 2007; 56: 1277-1288. MULDER H. Anaplerosis via pyruvate carboxylase 27) PRENTKI M, MATSCHINSKY FM, MADIRAJU SR. Metabolic is required for the fuel-induced rise in the signaling in fuel-induced insulin secretion. Cell ATP:ADP ratio in rat pancreatic islets. Diabetolo- Metab 2013; 18: 162-185. gia 2006; 49: 1578-1586. 28) MACDONALD MJ. Feasibility of a mitochondrial 42) FLAMEZ D, BERGER V, KRUHOFFER M, ORNTOFT T, pyruvate malate shuttle in pancreatic islets. Fur- PIPELEERS D, SCHUIT FC. Critical role for cataplerosis ther implication of cytosolic NADPH in insulin se- via citrate in glucose-regulated insulin release. Di- RETRACTEDcretion. J Biol Chem 1995; 270: 20051-20058. abetes 2002; 51: 2018-2024.

2224 Metabolic signaling of insulin secretion by pancreatic β-cell

43) CORKEY BE, TORNHEIM K, DEENEY JT, GLENNON MC, 55) PRALONG WF, BARTLEY C, WOLLHEIM CB. Single islet PARKER JC, MATSCHINSKY FM, RUDERMAN NB, PRENTKI beta-cell stimulation by nutrients: relationship be- M. Linked oscillations of free Ca2+ and the tween pyridine nucleotides, cytosolic ca2+ and ATP/ADP ratio in permeabilized RINm5F insulino- secretion. EMBO J 1990; 9: 53-60. ma cells supplemented with a glycolyzing cell- 56) HEART E, YANEY GC, CORKEY RF, SCHULTZ V, LUC E, LIU free muscle extract. J Biol Chem 1988; 263: 4254- L, DEENEY JT, SHIRIHAI O, TORNHEIM K, SMITH PJ, 4258. CORKEY BE. Ca2+, NAD(P)H and membrane po- 44) LONGO EA, TORNHEIM K, DEENEY JT, VARNUM BA, tential changes in pancreatic beta-cells by methyl TILLOTSON D, PRENTKI M, CORKEY BE. Oscillations in succinate: comparison with glucose. Biochem J cytosolic free Ca2+, oxygen consumption, and in- 2007; 403: 197-205. sulin secretion in glucose-stimulated rat pancreat- 57) FREEMAN H, SHIMOMURA K, COX RD, ASHCROFT FM. ic islets. J Biol Chem 1991; 266: 9314-9319. Nicotinamide nucleotide transhydrogenase: a link 45) MACDONALD MJ, FAHIEN LA, BUSS JD, HASAN NM, between insulin secretion, glucose metabolism FALLON MJ, KENDRICK MA. Citrate oscillates in liver and oxidative stress. Biochem Soc Trans 2006; and pancreatic beta cell mitochondria and in INS- 34: 806-810. 1 insulinoma cells. J Biol Chem 2003; 278: 51894- 58) IVARSSON R, QUINTENS R, DEJONGHE S, TSUKAMOTO K, 51900. IN ‘T VELD P, RENSTROM E, SCHUIT FC. Redox control 46) RONNEBAUM SM, ILKAYEVA O, BURGESS SC, JOSEPH JW, of exocytosis: regulatory role of NADPH, thiore- LU D, STEVENS RD BECKER TC, SHERRY AD, NEWGARD doxin, and glutaredoxin. Diabetes 2005; 54: 2132- CB, JENSEN MV. A pyruvate cycling pathway involv- 2142. ing cytosolic NADP-dependent isocitrate dehydro- 59) JITRAPAKDEE S, WUTTHISATHAPORNCHAI A, WALLACE JC, genase regulates glucose-stimulated insulin se- MACDONALD MJ. Regulation of insulin secretion: cretion. J Biol Chem 2006; 281: 30593-30602. role of mitochondrial signalling. Diabetologia 47) GUAY C, JOLY E, PEPIN E, BARBEAU A, HENTSCH L, PINE- 2010; 53: 1019-1032. DA M, MADIRAJU SR, BRUNENGRABER H, PRENTKI M. A 60) PRENTKI M, MADIRAJU SR. Glycerolipid metabolism role for cytosolic isocitrate dehydrogenase as a and signaling in health and disease. Endocr Rev negative regulator of glucose signaling for insulin 2008; 29: 647-676. secretion in pancreatic b-cells. PLoS One 2013; 8: 61) GRIBBLE FM, PROKS P, CORKEY BE, ASHCROFT FM. e77097. Mechanism of cloned ATP-sensitive potassium 48) GUAY C, MADIRAJU SR, AUMAIS A, JOLY E, PRENTKI M. A channel activation by oleoyl-CoA. J Biol Chem role for atp-citrate lyase, malic enzyme, and pyru- 1998; 273: 26383-26387. vate/citrate cycling in glucose-induced insulin se- 62) YANEY GC, CORKEY BE. Fatty acid metabolism and cretion. J Biol Chem 2007; 282: 35657-35665. insulin secretion in pancreatic beta cells. Dia- 49) PONGRATZ RL, KIBBEY RG, SHULMAN GI, CLINE GW. Cy- betologia 2003; 46: 1297-1312. tosolic and mitochondrial malic enzyme isoforms 63) HERRERO L, RUBI B, SEBASTIAN D, SERRA D, ASINS G, differentially control insulin secretion. J Biol Chem MAECHLER P, PRENTKI M, HEGARDT FG. Alteration of 2007; 282: 200-207. the malonyl-CoA/carnitine palmitoyltransferase I 50) XU J, HAN J, LONG YS, LOCK J, WEIR GC, EPSTEIN PN, interaction in the beta-cell impairs glucose-in- LIU YQ. Malic enzyme is present in mouse islets duced insulin secretion. Diabetes 2005; 54: 462- and modulates insulin secretion. Diabetologia 471. 2008; 51: 2281-2289. 64) RODUIT R, NOLAN C, ALARCON C, MOORE P, BARBEAU 51) STARK R, PASQUEL F, TURCU A, PONGRATZ RL, RODEN M, A, DELGHINGARO-AUGUSTO V, PRZYBYKOWSKI E, MORIN CLINE GW, SHULMAN GI, KIBBEY RG. Phospho- J, MASSÉ F, MASSIE B, RUDERMAN N, RHODES C, enolpyruvate cycling via mitochondrial phospho- POITOUT V, PRENTKI M. A role for the malonyl- enolpyruvate carboxykinase links anaplerosis and CoA/long-chain acyl-CoA pathway of lipid signal- mitochondrial GTP with insulin secretion. J Biol ing in the regulation of insulin secretion in re- Chem 2009; 284: 26578-26590. sponse to both fuel and nonfuel stimuli. Diabetes 2004; 53: 1007-1019. 52) JOSEPH JW, JENSEN MV, ILKAYEVA O, PALMIERI F, ALAR- CON C, RHODES CJ, NEWGARD CB. The mitochondrial 65) ANTINOZZI PA, SEGALL L, PRENTKI M, MCGARRY JD, citrate/isocitrate carrier plays a regulatory role in NEWGARD CB. Molecular or pharmacologic pertur- glucose-stimulated insulin secretion. J Biol Chem bation of the link between glucose and lipid me- 2006; 281: 35624-35632. tabolism is without effect on glucose-stimulated insulin secretion. J Biol Chem 1998; 273: 16146- 53) HEART E, CLINE GW, COLLIS LP, PONGRATZ RL, GRAY JP, 16154. SMITH PJ. Role for malic enzyme, pyruvate car- boxylation, and mitochondrial malate import in 66) KEBEDE MA, ALQUIER T, LATOUR MG, POITOUT V. Lipid glucose-stimulated insulin secretion. Am J Physiol receptors and islet function: therapeutic implica- Endocrinol Metab 2009; 296: E1354-1362. tions? Diabetes Obes Metab 2009; 11(Suppl 4): 10-20. 54) JOSEPH JW, ODEGAARD ML, RONNEBAUM SM, BURGESS SC, MUEHLBAUER J, SHERRY AD, NEWGARD CB. Normal 67) MAECHLER P, WOLLHEIM CB. Mitochondrial glutamate flux through ATP-citrate lyase or fatty acid syn- acts as a messenger in glucose-induced insulin thase is not required for glucose-stimulated in- exocytosis. Nature 1999; 402: 685-689. sulin secretion. J Biol Chem 2007; 282: 31592- 68) VETTERLI L, CAROBBIO S, POURNOURMOHAMMADI S, MAR- RETRACTED31600. TIN-DEL-RIO R, SKYTT DM, WAAGEPETERSEN HS, TAMARIT-

2225 C.-y. Zou, Y. Gong, J. Liang

RODRIGUEZ J, MAECHLER P. Delineation of glutamate 84) CRUTZEN R, SHLYONSKY V, LOUCHAMI K, VIRREIRA M, pathways and secretory responses in pancreatic HUPKENS E, BOOM A, SENER A, MALAISSE WJ, islets with beta-cell specific abrogation of the glu- BEAUWENS R. Does NAD(P)H oxidase-derived tamate dehydrogenase. Mol Biol Cell 2012; 23: H2O2 participate in hypotonicity-induced insulin 3851-3862. release by activating VRAC in -cells? Pflugers 69) LI M, LI C, ALLEN A, STANLEY CA, SMITH TJ. The struc- Arch 2012; 463: 377-390. ture and of glutamate dehy- 85) ROBERTSON RP, HARMON JS. Pancreatic islet beta-cell drogenase. Neurochem Int 2011; 59: 445-455. and oxidative stress: The importance of glu- 70) KIBBEY RG, PONGRATZ RL, ROMANELLI AJ, WOLLHEIM tathione peroxidase. FEBS Lett 2007; 581: 3743- CB, CLINE GW, SHULMAN GI. Mitochondrial GTP 3748. regulates glucose-stimulated insulin secretion. 86) BARKER CJ, BERGGREN PO. The pancreatic beta cell Cell Metab 2007; 5: 253-264. as a paradigm for advances in inositide research. 71) WU ZY, ZHU LJ, ZOU N, BOMBEK LK, SHAO CY, WANG Adv Biol Regul 2012; 52: 361-368. N, ET AL. AMPA receptors regulate exocytosis and 87) ILLIES C, GROMADA J, FIUME R, LEIBIGER B, YU J, JUHL insulin release in pancreatic beta cells. Traffic K, YANG SN, BARMA DK, FALCK JR, SAIARDI A, BARKER 2012; 13: 1124-1139. CJ, BERGGREN PO. Requirement of inositol py- 72) RUSSELL MA, MORGAN NG. Expression and func- rophosphates for full exocytotic capacity in pan- tional roles of guanylate cyclase isoforms in creatic beta cells. Science 2007; 318: 1299-1302. BRIN-BD11 -cells. Islets 2010; 2: 374-382. 88) ZHAO S, IGLESIAS J, MUGABO Y, PEYOT ML, JOLY E, ADIRAJU RENTKI 73) SCHULZE DU, DUFER M, WIERINGA B, KRIPPEIT-DREWS P, M SR, P M. Monoacylglycerol as a DREWS G. An adenylate kinase is involved in KATP metabolic coupling factor in glucose-stimulated in- channel regulation of mouse pancreatic beta sulin secretion. Can J Diabetes 2013; 37: 62. cells. Diabetologia 2007; 50: 2126-2134. 89) PEYOT ML, GUAY C, LATOUR MG, LAMONTAGNE J, LUSSIER R, PINEDA M, RUDERMAN NB, HAEMMERLE G, 74) DREWS G, KRIPPEIT-DREWS P, DUFER M. Electrophysiol- ogy of islet cells. Adv Exp Med Biol 2010; 654: ZECHNER R, JOLY E, MADIRAJU SR, POITOUT V, PRENTKI 115-163. M. Adipose triglyceride lipase is implicated in fu- el- and non-fuel-stimulated insulin secretion. J Bi- 75) RUDERMAN N, PRENTKI M. AMP kinase and malonyl- ol Chem 2009; 284: 16848-16859. CoA: targets for therapy of the metabolic syn- drome. Nat Rev Drug Discov 2004; 3: 340-351. 90) SORHEDE WINZELL M, AHREN B. Glucagon-like pep- tide-1 and islet lipolysis. Horm Metab Res 2004; 76) FU A, EBERHARD CE, SCREATON RA. Role of AMPK in 36: 795-803. pancreatic beta cell function. Mol Cell Endocrinol 2013; 366: 127-134. 91) KONRAD RJ, MAJOR CD, WOLF BA. Diacylglycerol hy- drolysis to arachidonic acid is necessary for in- 77) PRENTKI M, MADIRAJU SR. Glycerolipid/free fatty acid cy- sulin secretion from isolated pancreatic islets: se- cle and islet beta-cell function in health, obesity and quential actions of diacylglycerol and monoacyl- diabetes. Mol Cell Endocrinol 2012; 353: 88-100. glycerol lipases. Biochemistry 1994; 33: 13284- 78) LAMONTAGNE J, PEPIN E, PEYOT ML, JOLY E, RUDERMAN 13294. NB, POITOUT V, MADIRAJU SR, NOLAN CJ, PRENTKI M. 92) MARTINS EF, MIYASAKA CK, NEWSHOLME P, CURI R, Pioglitazone acutely reduces insulin secretion CARPINELLI AR. Changes of fatty acid composition in and causes metabolic deceleration of the pancre- incubated rat pancreatic islets. Diabetes Metab atic beta-cell at submaximal glucose concentra- 2004; 30: 21-27. tions. Endocrinology 2009; 150: 3465-3474. 93) SUMARA G, FORMENTINI I, COLLINS S, SUMARA I, WINDAK 79) NEWSHOLME P, MORGAN D, REBELATO E, OLIVEIRA- R, BODENMILLER B, RAMRACHEYA R, CAILLE D, JIANG H, EMILIO HC, PROCOPIO J, CURI R, CARPINELLI A. Insights PLATT KA, MEDA P, AEBERSOLD R, RORSMAN P, RICCI R. into the critical role of NADPH oxidase(s) in the Regulation of PKD by the MAPK p38delta in in- normal and dysregulated pancreatic beta cell. Di- sulin secretion and glucose homeostasis. Cell abetologia 2009; 52: 2489-2498. 2009; 136: 235-248. 80) PI J, BAI Y, ZHANG Q, WONG V, FLOERING LM, DANIEL 94) FERDAOUSSI M, BERGERON V, ZARROUKI B, KOLIC J, K, REECE JM, DEENEY JT, ANDERSEN ME, CORKEY BE, CANTLEY J, FIELITZ J, OLSON EN, PRENTKI M, BIDEN T, COLLINS S. Reactive oxygen species as a signal in MACDONALD PE, POITOUT V. G protein-coupled re- glucose-stimulated insulin secretion. Diabetes ceptor (GPR)40-dependent potentiation of insulin 2007; 56: 1783-1791. secretion in mouse islets is mediated by protein 81) MA ZA, ZHAO Z, TURK J. Mitochondrial dysfunction kinase D1. Diabetologia 2012; 55: 2682-2692. and beta-cell failure in type 2 diabetes mellitus. 95) DZHURA I, CHEPURNY OG, LEECH CA, ROE MW, DZHU- Exp Diabetes Res 2012; 2012: 703538. RA E, XU X, LU Y, SCHWEDE F, GENIESER HG, SMRCKA 82) Gehrmann W, Elsner M, Lenzen S. Role of meta- AV, HOLZ GG. Phospholipase C- links Epac2 acti- bolically generated reactive oxygen species for vation to the potentiation of glucose-stimulated in- lipotoxicity in pancreatic beta-cells. Diabetes sulin secretion from mouse islets of Langerhans. Obes Metab 2010; 12(Suppl 2): 149-158. Islets 2011; 3: 121-128. 83) WONG N, BLAIR AR, MORAHAN G, ANDRIKOPOULOS S. 96) ZHANG CL, KATOH M, SHIBASAKI T, MINAMI K, SUNAGA The deletion variant of nicotinamide nucleotide Y, TAKAHASHI H, YOKOI N, IWASAKI M, MIKI T, SEINO S. transhydrogenase (NNT) does not affect insulin The cAMP sensor Epac2 is a direct target of an- secretion or glucose tolerance. Endocrinology tidiabetic sulfonylurea drugs. Science 2009; 325: RETRACTED2010; 151: 96-102. 607-610.

2226 Metabolic signaling of insulin secretion by pancreatic β-cell

97) GIUFFRIDA FM, REIS AF. Genetic and clinical charac- 99) ROSENGREN AH, BRAUN M, MAHDI T, ANDERSSON SA, teristics of maturity-onset diabetes of the young. TRAVERS ME, SHIGETO M ZHANG E, ALMGREN P, LADEN- Diabetes Obes Metab 2005; 7: 318-326. VALL C, AXELSSON AS, EDLUND A, PEDERSEN MG, JONS- 98) PARIKH H, GROOP L. Candidate genes for type 2 di- SON A, RAMRACHEYA R, TANG Y, WALKER JN, BARRETT A, abetes. Rev Endocr Metab Disord 2004; 5: 151- JOHNSON PR, LYSSENKO V, MCCARTHY MI, GROOP L, 176. SALEHI A, GLOYN AL, RENSTRÖM E, RORSMAN P, ELIAS- SON L. Reduced insulin exocytosis in human pan- creatic beta-cells with gene variants linked to type 2 diabetes. Diabetes 2012; 61: 1726-1733.

RETRACTED

2227