TITLE

IMPROVING RELEASE OF INSULIN

DESCRIPTION Technical field

The present invention relates to the use of pyruvate dehydrogenase kinase inhibitors to improve insulin response on serum glucose in type 2 diabetes mammals.

Background of the invention Modern life style often leads to overweight which in turn causes a metabolic syndrome, which includes type 2 diabetes, which is a result of impaired release of insulin in response to glucose. Glucose stimulus-secretion coupling in pancreatic beta-cells involves at least two pathways, the well-defined triggering pathway, and the less well-characterized amplifying pathway. The amplifying pathway has been shown to be dependent on glucose metabolism (1 ), where the fate of pyruvate is a key event. Pyruvate can enter the TCA cycle either via carboxylation by pyruvate carboxylase (PC) to oxaloacetate or via oxidative decarboxylation to acetyl-CoA by the pyruvate dehydrogenase complex (PDC). Flux via lactate dehydrogenase is limited by the normally low activity of this enzyme in pancreatic beta-cells (2). PC is highly expressed in pancreatic beta-cells, as compared to islet non-beta cells (3), and ~ 40% of pyruvate entering the TCA cycle during glucose stimulation is carboxylated by PC (4). Hence, PC is a key enzyme in glucose metabolism, and its importance for glucose-stimulated insulin secretion (GSIS) has been established by us and others (5-8). The role of PDC is more unclear. On the one hand, it has been shown that activation of PDC by dichloroacetate or adenoviral expression of the catalytic subunit of pyruvate dehydrogenase phosphatase (9), or inactivation of PDC by expression of PDK3, a PDK isoform of high activity not normally expressed in islets (10), has no effect on glucose metabolism or on insulin secretion. On the other hand, Cline at al (1 1 ) recently showed that fluxes through PDH and PC were equally important for insulin secretion by INS-1 cells, and increases in PDK4 protein expression in islets after fasting is associated with suppression of GSIS (12). The activity of PDH is tightly regulated via phosphorylation by pyruvate dehydrogenase kinases (PDKs) and dephosphorylation by pyruvate dehydrogenase phosphatases (PDPs) (13). There are four different isoforms of PDK (PDK1 -4) (14; 15), and their activation has been suggested to have implications for insulin secretion (12).

The mechanisms underlying the impairment of insulin secretion in Type 2 diabetes are not known; however, it is well established that chronic exposure to high concentrations of glucose, a main characteristic of diabetes, is detrimental to the beta-cells (16). This state is known as glucotoxicity, where the insulin response to glucose is blunted. It is clear that glucose modifies gene expression in pancreatic beta-cells (17; 18) and since glucose metabolism is crucial for stimulation of GSIS, changes in the expression of metabolic genes might be causing some of these disturbances.

To identify genes important for insulin secretion, we established a model to mimic the elevated blood glucose concentration in diabetes. We cultured robustly glucose-responsive INS-1

832/13 cells for 48 h in the presence of either low (2.8 mM) or high (16.7 mM) concentrations of glucose: the latter markedly impairing GSIS. We then proceeded to screen the 832/13 cells for gene changes caused by the high concentrations of glucose, which presumably might be involved in regulation of insulin secretion. For that purpose, we designed an array of assays for quantitative real-time PCR, allowing robust and simultaneous quantification of multiple genes involved in glucose metabolism. With this approach, we revealed that differential regulation of expression of PDKs may underlie glucose responsiveness in 832/13 beta-cells. The significance of PDK1 in beta-cell glucose stimulus-secretion coupling to GSIS was validated using knock-down by RNA interference (RNAi).

MATERIALS AND METHODS

Reagents and siRNA

All chemicals were from Sigma (St Louis, MO, USA) if not stated otherwise. Small interfering

RNAs (siRNAs) were 21 nt long duplexes designed, synthesized and supplied by Ambion (Austin, TX, USA). Three candidate sequences against PDK1 were tested in preliminary experiments. The optimal sequence, used in all following experiments was: sense strand 5'- GCAUAAAUCCAAACUGUGAtt -3' and antisense strand 5'-UCACAGUUUGGAUUUAUGCtt- 3'. As negative control Silencer® Negative Control #2 from Ambion was used.

Cell culture

Cells were cultured in RPMI-1640 containing 2.8, 16.7 or 1 1 .1 mM D-glucose and supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100μg/ml streptomycin, 10 mM HEPES, 2 mM glutamine, 1 mM sodium pyruvate, and 50 μM β-mercaptoethanol, at 37 ^Q C in a humidified atmosphere containing 95% air and 5% CO2.

Gene expression analysis Q-PCR RNA was extracted using RNAeasy (Qiagen, Hilden, Germany). 0.5 μg RNA was used for cDNA synthesis with Superscript (Invitrogen, Carlsbad, CA, USA). 20 μl reactions mixture with 20 ng cDNA, 10 μl TaqMan mastermix (Applied Biosystems, Foster City, CA USA), and 900 nM TaqMan gene expression assay were run in a 7900HT Fast Real-Time System (Applied Biosystems). The Q-PCR was carried out as follows: 50 ^Q C for 2 minutes, 95 ^Q C for 10 minutes, 40 cycles of 95 ^Q C for 15 seconds, and 60 ^Q C for 1 minute. The amount of mRNA was calculated relative to the amount of HPRT mRNA in the same sample by the formula X0/R0 = 2 ^{ctR Ctx} , where XO is the original amount of mRNA for the gene of interest, RO is the original amount of HPRT mRNA, CtR the Ct value for HPRT, and CtX the Ct value for the gene of interest.

Low density array (LDA)

Four different house keeping genes were evaluated for stable expression at high and low glucose concentrations with Q-PCR; 18S, 29S, hypoxanthine-guanine phospho-ribosyl transferase (HPRT) and β-actin. Expression of HPRT varied the least at different glucose concentrations, and HPRT was therefore included in the low density array. TaqMan Low

Density Arrays (Applied Biosystems,) are 384-well microfluidic cards, pre-loaded with TaqMan gene expression assays of your own choice. We designed the card to comprise assays for 46 of the essential genes in glycolysis, the TCA cycle and the respiratory chain; the assays were pre-designed and validated by Applied Biosystems. INS-1 832/13 cells were cultured at 2.8 or 16.7 mM glucose for 48 h, and afterwards RNA and cDNA were prepared as above. 50 ng cDNA were then loaded to each port on the micro fluidic card, and the card was run in a 7900HT Fast Real-Time System.

Insulin secretion INS-1 832/13 cells were cultured in 24-well dishes for 48 h at 2.8 or 16.7 mM glucose or transfected with siRNA at 72 h prior to assay. When assayed, the cells were washed in HEPES balanced salt solution (HBSS; 1 14 mM NaCI; 4.7 mM KCI; 1.2 mM KH ₂ PO ₄ ; 1 .16 mM MgSO4; 20 mM HEPES; 2.5 mM CaCI ₂ ; 25.5 mM NaHCO ₃ ; 0.2% BSA, pH 7.2) supplemented with 2.8 mM glucose for 2 h at 37 C. Insulin secretion was then measured by static incubation

of the cells for 1 h in 1 ml HBSS containing 2.8 or 16.7 mM glucose or 2.8mM glucose combined with 10 mM leucine or 10 mM methyl-succinate. Insulin was measured by the Coat- a-Count kit (DPC, Los Angeles, CA), which recognizes human insulin and cross reacts approximately 20% with rat insulin.

Insulin content measurement

INS-1 832/13 cells were cultured in 24-well dishes for 48 h at 2.8 or 16.7 mM glucose. Cells were washed in PBS, 100 μl H ₂ O were added per well, the cells were scraped off and sonicated. Afterwards the cells were centrifuged and the supernatant was diluted 1 OX in acidified ethanol. The samples were stored at -20 ^Q C until assay with RIA.

Transfection

832/13 cells were nucleofected using the Amaxa nucleofector (Amaxa, Cologne, Germany). The cells were trypsinised, counted and centrifuged. After this, 3X106 cells were resuspended in 100 μl nucleofector solution V mixed with 500 nM siRNA, and nucleofected with program T- 27. Afterwards the cells were seeded in 24- or 12-well plates and cultured for three days at 1 1 .1 mM glucose before experiments were performed.

Glucose oxidation Cells cultured in 12-well plates were trypsinated and resuspended in 1 ml HBSS. The cells were then centrifuged for 2 min at 4000xg, resuspended in 100 μl HBSS containing 2.8 mM glucose and transferred to a cup (Kimble-Kontes, Vineland, NJ, USA) suspended from a rubber sleeve stopper (Fisher, Pittsburgh, PA, USA), which was inserted into a glass scintillation vial. A 100 μl reaction mixture containing 13 kBq D-[ ¹⁴ C(U)]glucose (NEN Life Science Products, Boston, MA, USA) and glucose at a final concentration of 2.8 or 16.7 mmol/l was added to the cups and the vials were sealed. After 1 h at 37 ^Q C the reaction was terminated by injection of 100 μl 10% trichloroacetic acid into the suspended cup. The oxidation rate was measured as 14CO ₂ trapped in 300μl benzethonium hydroxide added to the bottom of the sealed vials, followed by an additional incubation overnight at room temperature. Released 14CO ₂ was determined by scintillation counting.

Western blotting

Cells were transfected with siRNA against PDK1 or a negative control, and cultured for 72h before harvesting. They were then homogenized in a homogenization buffer containing 0.25 M

sucrose, 1 mM EDTA, 1 mM dithiothreitol (DTT), 20 μg/ml leupeptin, 10 μg/ml antipain, and 1 μg/ml pepstatin. Determination of protein concentration was performed using a BCA protein kit (Pierce, Rockford, IL) and 200 μg of protein were loaded onto the gel. Samples were subjected to SDS-PAGE on 8% acrylamide gels and then electroblotted onto nitrocellulose membranes (HyBond-C Extra, Amersham Biosciences, Uppsala, Sweden). As primary antibody, a polyclonal rabbit anti-rat PDK1 antibody (1 :1000) was used, and a horseradish peroxidase- linked anti-rabbit IgG antibody (Amersham Biosciences) was used as secondary antibody (1 :5000). The blots were developed with enhanced chemiluminescence and detection was performed using a CCD camera (Fuji Photo Film Co., Ltd, Tokyo, Japan).

ATP measurements

Cells were washed in HBSS, followed by preincubation for 1 h in HBSS at 2.8 mM glucose; the buffer was then replaced by HBSS containing 2.8 or 16.7 mM glucose. After an additional 5 min, trichloroacetic acid was added at a final concentration of 2% to the cells, which were kept on ice for 20 min. After centrifugation (13,000 g), the supernatant was assayed for ATP content, using bioluminescence (BioThema, Haninge, Sweden). ADP was measured after ATP had been depleted by ATP sulphurylase followed by conversion of ADP to ATP by pyruvate kinase, as described in detail (19)

Statistical analysis

Statistical analysis was performed using an unpaired Student's t test.

RESULTS

Insulin secretion and content To establish conditions in 832/13 cells where metabolic regulation of insulin secretion was affected we cultured cells for a prolonged time period (48 h) at either low (2.8 mM) or high (16.7 mM) glucose. It was predicted that exposure to the high glucose concentrations would induce a glucotoxic condition with impaired GSIS. Results are shown in figure 1. After 48h culture at 2.8 mM glucose, insulin secretion increased 6.8 fold upon acute (1 h) stimulation with 16.7 mM glucose. Under conditions where the KATP-channels were bypassed with 35 mM KCI and 250 mM diazoxide, insulin secretion at 2.8 mM glucose increased by 6 fold; a further 3-fold increase in insulin secretion occurred when the glucose was increased to16.7 mM. Insulin secretion at basal glucose was lower (P<0.01 ) when the cells had been pre- cultured at high glucose. Furthermore, cells cultured for 48 h at 16.7 mM glucose increased

their insulin secretion by only 3.5-fold when stimulated with 16.7 mM glucose compared to 2.8 mM. Similarly, under KATP-channel-independent conditions, insulin secretion rose only twofold upon stimulation with 16.7 mM glucose. Culture at high glucose decreased total cellular insulin content by approximately 80% (594 ± 144 vs. 105 ±50 ng insulin/mg protein).

Gene expression analysis

In the next step, expression of the candidate metabolic genes in the LDA designed by us was assessed under the established conditions where insulin secretion was enhanced (pre-culture at 2.8 mM glucose) or diminished (pre-culture at 16.7 mM glucose). Out of the 46 genes, six were significantly differentially expressed following culture at 16.7 mM glucose versus 2.8 mM glucose (figure 2). Of these, five were upregulated at high glucose and one was down- regulated. The upregulated genes were glyceraldehyde 3-phosphate dehydrogenase (2.6 fold increase; P<0.05), the glycolytic enzymes triosephosphate isomerase 1 (2.5 fold increase; P<0.05) and phosphoglycerate kinase 1 (2.4 fold increase; P<0.05), the lipogenic enzyme acetyl-CoA carboxylase (2.3 fold increased; P<0.001 ), and the regulatory kinase PDK1 (2.4 fold increase; P<0.01 ). PDK2, a second PDK isoform expressed in these Knock-down of PDK1

Since we found that the expression of PDK1 increased after culture in 16.7 mM glucose where insulin secretion was decreased, we hypothesized that specifically-enhanced expression of this regulatory kinase for PDC phosphorylation might be involved in the metabolic disturbance underlying impaired GSIS. We therefore investigated the effect of knock-down of PDK1 . Treatment with siRNA against PDK1 decreased its mRNA level by 80%, as quantified with Q- PCR figure 3A); Western blot showed a clear decrease in PDK1 protein expression (figure 3B). Insulin secretion after knock-down of PDK1

In the next step, we examined the effect of knock-down of PDK1 on insulin secretion in response to glucose and two additional nutrient secretagogues, leucine and succinate (Fig. 4). 832/13 beta-cells transfected with the negative control increased their insulin secretion by 2.5 fold over basal (2.8 mM glucose) when stimulated with 16.7 mM glucose. Knockdown of PDK1 did not affect basal insulin secretion but, remarkably, cells treated with PDK1 siRNA increased their insulin secretion by four-fold (P<0.01 ) when stimulated with 16.7 mM glucose. In contrast, insulin secretion provoked by the two other metabolic fuels, leucine and succinate, was not influenced by knock-down of PDK1 .

Glucose oxidation

We hypothesized that knockdown of PDK1 might allow relief of inhibition of PDC, allowing increased decarboxylation of pyruvate to acetyl-CoA and, potentially, increased TCA cycle flux. We therefore analyzed the effect of PDK1 knockdown on glucose oxidation to CO ₂ using [U- ¹⁴ C] glucose. At 2.8 mM glucose, PDK1 knockdown caused a non-significant trend towards increased glucose oxidation. With control cells, increasing glucose from 2.8 M to 16.7 mM increased glucose oxidation from 101 ± 24 to 237± 27 pmol/h/1 x 10 ^'6 cells (2.3-fold, P<0.05). When the glucose concentration was increased from 2.8 to 16.7 mM in siRNA-treated cells, glucose oxidation increased from 146 ± 30 to 243 ± 46 pmol/h/1 x10 ^'6 cells. Thus, at high (16.7 mM) glucose, glucose oxidation did not differ between control and siRNA-treated cells.

ATP/ADP measurements We examined whether the knock-down of PDK had any effect on ATP/ADP ratios. ATP/ADP ratios at basal glucose were similar in control and siRNA-treated cells. In control cells, the ATP/ADP ratio doubled from 6.5 to 13.5 in response to high concentrations of glucose; an increase of similar magnitude was observed in cells treated with PDK1 siRNA.

DISCUSSION

Previous studies have reported discrepant data regarding the potential roles of the PDKs in modulating insulin secretion, depending on the system used. We have now found that the expression of PDK1 is upregulated in glucose-responsive clonal beta-cells after prolonged (48 h) exposure to high glucose concentrations in culture, a condition which leads to impaired GSIS upon subsequent acute challenge with secretagogues. Under the same circumstances the other PDK expressed in clonal beta-cells, PDK2, was downregulated. Interestingly, PDK4 mRNA and protein was contrary to previous findings in islets (12), not detectable in 832/13 cells. This might indicate that the protein is not expressed in beta-cells but rather in non-beta- cells. Alternatively, beta-cells cultured in the absence of fatty acids may express PDK4 only at very low levels.

The PDK isoform shift observed in response to sustained changes in glucose concentration points to a specific role for the individual PDKs. PDKs phosphorylate three serine-residues on PDC E1 alpha: serine-264 (site 1 ), serine-271 (site 2) and serine-203 (site 3) (20). All PDKs

can phosphorylate site 1 and site 2, whereas PDK1 uniquely phosphorylates site 3 (21 ). This suggests that site 3 is of special importance in tissues where PDK1 is expressed: pancreatic islets (12), heart (22) and, to a lesser extent, skeletal muscle (23). Since reactivation of E1 requires removal of all three phosphate groups, PDK1 phosphorylation can be particularly effective in maintaining PDC inactive. Upregulation of PDK expression has previously been observed following starvation, where PDK4 is upregulated in liver, kidney, skeletal muscle, heart and pancreatic islets (12; 22; 23). Increased mRNA expression of all three PDK isoforms has been reported to occur in mouse islet cultures with palmitate, whereas both PDK1 and PDK2 mRNA expression were increased, PDK4 mRNA expression was decreased in mouse islets cultured with high glucose (10). In this study, protein expression was not quantified and since islets were used, which cell types (beta versus non-beta) expressing each of these specific isoforms were not addressed. However, our data showing a specific PDK isoform shift towards PDK1 in glucose-responsive beta-cells imply that PDK isoform expression is differentially regulated by high glucose in alpha and beta-cells

Hypoxia causes cancer cells to induce glycolytic enzymes: this induction, mediated by the transcription factor HIF-1 alpha, increases glycolytic flux to pyruvate and increased glycolytic production of ATP. Recent studies have also shown that HIF-1 also influences mitochondrial function in cancer cells, inducing over-expression of PDK1 , an effect observed in conjunction with suppression of respiration (24; 25). Selectively blocking HIF-induced expression of PDK1 induces apoptosis (9) this increased apoptosis following selective PDC activation was attributed to either oxygen depletion (25) or enhanced production of reactive oxygen species (ROS) (24). Interestingly, a mutation in the HIF-1 alpha gene, causing a higher activity of the transcription factor, has been identified as a risk factor for type 2 diabetes (26). The present study has not examined whether HIF-1 alpha mediates upregulation of PDK1 expression in pancreatic beta-cells by exposure to high glucose in culture. However, it is interesting to note that normoxic activation of HIF1 alpha has been observed in other cell systems. Mechanisms that have been identified to date include oncogenic activation (27) and activation by depressed mitochondrial ROS production, due to mitochondrial abnormalities (28). High glucose stimulates c-Myc gene expression in rat beta-cells and pancreatic islets (29) and it has been reported, although this remains controversial, that rates of superoxide formation are lower in beta-cells following prolonged exposure to high glucose (30).

The upregulation of PDK1 brought us to question whether the expression changes were involved in the impairment of insulin secretion evoked by antecedent chronic exposure to high glucose levels. Therefore we investigated the effects on insulin secretion when PDK1 was knocked down with RNAi. This intervention caused a clear decrease in PDK1 mRNA and protein levels, together with a significant increase in GSIS, supporting our hypothesis.

Knockdown of PDK1 was without effect on the cellular ATP/ADP ratios. Although our finding that glucose oxidation was unchanged when PDK was knocked down might seem contradictory to our conclusion that PDC activation is a central event, we speculate that increased flux via PDC results in increased cataplerosis rather than increased TCA cycle flux, which in turn stimulates GSIS. Efflux and cycling of citrate has previously been suggested to be of functional importance for glucose-stimulated insulin secretion (5; 31 ). In fact, exit of citrate into a cycle to pyruvate would explain that glucose oxidation was not affected. Such a cycle would neither give rise to CO ₂ production, whereby oxidation of glucose is monitored, nor to NADH or FADH ₂ in the TCA cycle, which drive the respiratory chain and ultimately ATP production, which we also found was unaffected. Instead, other signaling pathways could be in operation. First, a cycle from citrate to pyruvate would involve malic enzyme (5; 32). The novel finding that the mitochondrial citrate/isocitrate carrier plays a regulatory role in insulin secretion (33), further supports the cycling hypothesis. The formation of pyruvate would thus result in generation of the reducing equivalent NADPH, raising the NADPH/NADP+ ratio. An increased NADPH/NADP+ ratio has recently has received strong support as a metabolic coupling factor in GSIS (34; 35), which further supports our findings. Second, acetyl-CoA, formed in the reaction catalyzed by citrate lyase, may be carboxylated to malonyl-CoA by acetyl-CoA carboxylase. This metabolite is a strong allosteric inhibitor of carnitine palmitoyl transferase 1 , thus preventing flux of long-chain acyl-CoAs into the mitochondrion for subsequent oxidation. Malonyl-CoA has been proposed to be an important metabolic coupling factor, via a presumed elevation of cytosolic long chain acyl-CoA, which stimulates exocytosis by unresolved mechanisms (36). However, the support for the role of malonyl-CoA in stimulus-secretion coupling in 832/13 cells has not been unambiguous (37; 38). Moreover, while we found that acetyl-CoA carboxylase was strongly upregulated, we found that fatty acid oxidation was unchanged in PDK1 knock-down cells (data not shown). On balance, these observations imply that this pathway plays no significant role under the conditions that we have examined. The fact that insulin secretion stimulated by leucine or succinate was not affected by knock-down of PDK1 further supports our hypothesis, since these two fuels would enter the TCA cycle distal to the presumed exit of citrate, and hence not contribute to enhanced cycling.

In conclusion, we have identified a potential role for the PDKs in regulation of glucose responsiveness in beta-cells. Our data so far, suggest, but do not prove, stimulation of a cycling pathway from citrate to pyruvate, via malic enzyme, a pathway previously associated with glucose responsiveness in clonal beta-cells (5-8).

The above results show that Inhibition of PDK1 will be a therapy for impaired insulin secretion in Type 2 Diabetes, and thus there are a number of rational reasons for developing inhibition of PDK1 as new treatment modality in Type 2 Diabetes. 1. Inhibition of PDK1 will correct a pathogenetic abnormality in β-cells in Type 2 Diabetes.

Thus, using this modality we are attacking a cause and not a consequence of the disease. This may be important for the long-term success of this potential therapy in Type 2 Diabetes.

2. The effect of inhibition of PDK1 on insulin secretion is glucose-dependent. Thus, there is no risk for hypoglycemia, the most-feared and potentially serious adverse treatment effect in Type 2 Diabetes.

3. An inhibitor of PDK1 aimed at β-cells and insulin secretion would not be envisaged to exert negative effects in other tissues. This isoforms of the kinase is expressed predominantly in β-cells and the heart. 4. If there were to be an effect in other tissues as well, we would expect to see increases in whole body glucose metabolism, which is beneficial. In fact, this effect has been observed in animal models where PDK2-antagonsists have been used (3).

The term "PDK-inhibitors" as used herein means inhibitors of pyruvate dehydrogenase kinase and include, but are not limited to, those compounds disclosed by Aicher et al in J. Med. Chem. 42 (1999) 2741 -2746.

Abbreviations used herein

HBSS: HEPES balanced salt solution, HPRT: hypoxanthine-guanine phosphoribosyl transferase, K _A τp-channels: ATP-sensitive K+ channels, LDA: low density array, PC: pyruvate carboxylase, PDC: pyruvate dehydrogenase complex, PDK: pyruvate dehydrogenase kinase, Q-PCR: quantitative PCR, SDS-PAGE: sodium dodecyl sulphate polyacrylamid gelelectrophoresis, siRNA: small interfering RNA, TCA: tricarboxylic acid.

FIGURE LEGENDS

FIG. 1 Insulin secretion after culture in 2.8 or 16.7 mM glucose for 48 h.

Insulin secretion by INS 832/213 cells was assayed following acute (1 h) exposure to 2.8 mM glucose (2.8G) and 16.7 mM glucose (16.7G), and under K _AT p-independent conditions (with 2.8 mM glucose or 16.7mM glucose) in combination with 35 mM KCI plus 25QDM diazoxide. Grey bars = pre-culture at 2.8 mM glucose and black bars = pre-culture at 16.7 mM glucose. Values are means ± S. E. M. for 6 independent experiments. ^** P< 0.01 , ^*** P< 0.001 FIG. 2 Alterations in gene expression after culture in 2.8 or 16.7 mM glucose for 48 h. Cells were cultured for 48h in 2.8 or 16.7 mM glucose, and gene expression was then determined with Q-PCR. ACC= acetyl-CoA carboxylase, GAPDH= glyceraldehyde-3- phosphate dehydrogenase, PGK= phosphoglycerate kinase 1 , PDKI = pyruvate dehydrogenase kinase 1 , PDK2= pyruvate dehydrogenase kinase 2, TPI= triosephosphate isomerase 1. Values are means ± S. E. M. for 5 independent experiments. ^* P<0.05, ^** P< 0.01 , ^*** P< 0.001. FIG. 3 Figure 3. Knock-down of PDK1.

PDK1 mRNA levels relative to HPRT in control and siRNA-treated cells (A) and protein levels (B) 72 h after transfection. Values are means ± S. E. M. for 4 independent experiments in A, ^*** P< 0.001. B shows a representative blot. FIG. 4 Figure 4. Insulin secretion is stimulated by knock-down of PDK1. 72h after siRNA transfection, cells were stimulated with 2.8 mM glucose, 16.7mM glucose or 2.8 mM glucose combined with 10 mM leucine or succinate for 1 h. Black bars = siRNA-treated cells, grey bars = control. Values are means ± S. E. M. for 6 independent experiments for GSIS and 3 independent experiments for leucine and succinate. ^** P<0.01

REFERENCES

1. Gembal M, Gilon P, Henquin JC: Evidence that glucose can control insulin release independently from its action on ATP-sensitive K+ channels in mouse B cells. J Clin Invest 89:1288-1295, 1992

2. Sekine N, Cirulli V, Regazzi R, Brown LJ, Gine E, Tamarit-Rodriguez J, Girotti M, Marie S, MacDonald MJ, Wollheim CB, et al.: Low lactate dehydrogenase and high mitochondrial glycerol phosphate dehydrogenase in pancreatic beta-cells. Potential role in nutrient sensing. J Biol Chem 269:4895-4902, 1994 3. Schuit F, De Vos A, Farfari S, Moens K, Pipeleers D, Brun T, Prentki M: Metabolic fate of glucose in purified islet cells. Glucose-regulated anaplerosis in beta cells. J Biol Chem

272:18572-18579, 1997

4. Khan A, Ling ZC, Landau BR: Quantifying the carboxylation of pyruvate in pancreatic islets.

J Biol Chem 271 :2539-2542, 1996 5. Farfari S, Schulz V, Corkey B, Prentki M: Glucose-regulated anaplerosis and cataplerosis in pancreatic beta-cells: possible implication of a pyruvate/citrate shuttle in insulin secretion.

Diabetes 49:718-726, 2000

6. Fransson U, Rosengren AH, Schuit FC, Renstrom E, Mulder H: Anaplerosis via pyruvate carboxylase is required for the fuel-induced rise in the ATP:ADP ratio in rat pancreatic islets. Diabetologia 49:1578-1586, 2006

7. Liu YQ, Jetton TL, Leahy JL: beta-Cell adaptation to insulin resistance. Increased pyruvate carboxylase and malate-pyruvate shuttle activity in islets of nondiabetic Zucker fatty rats. J Biol Chem 277:39163-39168, 2002

8. Lu D, Mulder H, Zhao P, Burgess SC, Jensen MV, Kamzolova S, Newgard CB, Sherry AD: 13C NMR isotopomer analysis reveals a connection between pyruvate cycling and glucose- stimulated insulin secretion (GSIS). Proc Natl Acad Sci U S A 99:2708-2713, 2002

9. Nicholls Ll, Ainscow EK, Rutter GA: Glucose-stimulated insulin secretion does not require activation of pyruvate dehydrogenase: impact of adenovirus-mediated overexpression of PDH kinase and PDH phosphate phosphatase in pancreatic islets. Biochem Biophys Res Commun 291 :1081 -1088, 2002

10. Xu J, Han J, Epstein PN, Liu YQ: Regulation of PDK mRNA by high fatty acid and glucose in pancreatic islets. Biochem Biophys Res Commun 344:827-833, 2006

1 1 . Cline GW, Lepine RL, Papas KK, Kibbey RG, Shulman Gl: 13C NMR isotopomer analysis of anaplerotic pathways in INS-1 cells. J Biol Chem 279:44370-44375, 2004

12. Sugden MC, Bulmer K, Augustine D, Holness MJ: Selective modification of pyruvate dehydrogenase kinase isoform expression in rat pancreatic islets elicited by starvation and activation of peroxisome proliferator-activated receptor-alpha: implications for glucose- stimulated insulin secretion. Diabetes 50:2729-2736, 2001 13. Linn TC, Pettit FH, Reed LJ: Alpha-keto acid dehydrogenase complexes. X. Regulation of the activity of the pyruvate dehydrogenase complex from beef kidney mitochondria by phosphorylation and dephosphorylation. Proc Natl Acad Sci U S A 62:234-241 , 1969 14. Gudi R, Bowker-Kinley MM, Kedishvili NY, Zhao Y, Popov KM: Diversity of the pyruvate dehydrogenase kinase gene family in humans. J Biol Chem 270:28989-28994, 1995 15. Rowles J, Scherer SW, Xi T, Majer M, Nickle DC, Rommens JM, Popov KM, Harris RA, Riebow NL, Xia J, Tsui LC, Bogardus C, Prochazka M: Cloning and characterization of PDK4 on 7q21 .3 encoding a fourth pyruvate dehydrogenase kinase isoenzyme in human. J Biol Chem 271 :22376-22382, 1996

16. Marshak S, Leibowitz G, Bertuzzi F, Socci C, Kaiser N, Gross DJ, Cerasi E, Melloul D: Impaired beta-cell functions induced by chronic exposure of cultured human pancreatic islets to high glucose. Diabetes 48:1230-1236, 1999

17. Laybutt R, Hasenkamp W, Groff A, Grey S, Jonas JC, Kaneto H, Sharma A, Bonner-Weir S, Weir G: beta-cell adaptation to hyperglycemia. Diabetes 50 Suppl 1 :S180-181 , 2001

18. Roche E, Farfari S, Witters LA, Assimacopoulos-Jeannet F, Thumelin S, Brun T, Corkey BE, Saha AK, Prentki M: Long-term exposure of beta-INS cells to high glucose concentrations increases anaplerosis, lipogenesis, and lipogenic gene expression. Diabetes 47:1086-1094, 1998

19. Schultz V, Sussman I, Bokvist K, Tornheim K: Bioluminometric assay of ADP and ATP at high ATP/ADP ratios: assay of ADP after enzymatic removal of ATP. Anal Biochem 215:302- 304, 1993

20. Teague WM, Pettit FH, Yeaman SJ, Reed LJ: Function of phosphorylation sites on pyruvate dehydrogenase. Biochem Biophys Res Commun 87:244-252, 1979

21 . Korotchkina LG, Patel MS: Site specificity of four pyruvate dehydrogenase kinase isoenzymes toward the three phosphorylation sites of human pyruvate dehydrogenase. J Biol Chem 276:37223-37229, 2001

22. Wu P, Sato J, Zhao Y, Jaskiewicz J, Popov KM, Harris RA: Starvation and diabetes increase the amount of pyruvate dehydrogenase kinase isoenzyme 4 in rat heart. Biochem J 329 ( Pt 1 ):197-201 , 1998

23. Spriet LL, Tunstall RJ, Watt MJ, Mehan KA, Hargreaves M, Cameron-Smith D: Pyruvate dehydrogenase activation and kinase expression in human skeletal muscle during fasting. J Appl Physiol 96:2082-2087, 2004

24. Kim JW, Tchernyshyov I, Semenza GL, Dang CV: H I F- 1 -mediated expression of pyruvate dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell

Metab 3:177-185, 2006

25. Papandreou I, Cairns RA, Fontana L, Lim AL, Denko NC: HIF-1 mediates adaptation to hypoxia by actively downregulating mitochondrial oxygen consumption. Cell Metab 3:187-197, 2006 26. Yamada N, Horikawa Y, Oda N, lizuka K, Shihara N, Kishi S, Takeda J: Genetic variation in the hypoxia-inducible factor-1 alpha gene is associated with type 2 diabetes in Japanese. J Clin Endocrinol Metab 90:5841 -5847, 2005

27. Semenza GL: Targeting HIF-1 for cancer therapy. Nat Rev Cancer 3:721 -732, 2003

28. Bonnet S, Michelakis ED, Porter CJ, Andrade-Navarro MA, Thebaud B, Haromy A, Harry G, Moudgil R, McMurtry MS, Weir EK, Archer SL: An abnormal mitochondrial-hypoxia inducible factor-1 alpha-Kv channel pathway disrupts oxygen sensing and triggers pulmonary arterial hypertension in fawn hooded rats: similarities to human pulmonary arterial hypertension. Circulation 1 13:2630-2641 , 2006

29. Jonas JC, Laybutt DR, Steil GM, Trivedi N, Pertusa JG, Van de Casteele M, Weir GC, Henquin JC: High glucose stimulates early response gene c-Myc expression in rat pancreatic beta cells. J Biol Chem 276:35375-35381 , 2001

30. Martens GA, Cai Y, Hinke S, Stange G, Van de Casteele M, Pipeleers D: Glucose suppresses superoxide generation in metabolically responsive pancreatic beta cells. J Biol Chem 280:20389-20396, 2005 31 . Flamez D, Berger V, Kruhoffer M, Orntoft T, Pipeleers D, Schuit FC: Critical role for cataplerosis via citrate in glucose-regulated insulin release. Diabetes 51 :2018-2024, 2002 32. MacDonald MJ: Feasibility of a mitochondrial pyruvate malate shuttle in pancreatic islets. Further implication of cytosolic NADPH in insulin secretion. J Biol Chem 270:20051 -20058, 1995 33. Joseph JW, Jensen MV, llkayeva O, Palmieri F, Alarcon C, Rhodes CJ, Newgard CB: The mitochondrial citrate/isocitrate carrier plays a regulatory role in glucose-stimulated insulin secretion. J Biol Chem, 2006

34. Ivarsson R, Quintens R, Dejonghe S, Tsukamoto K, in 't Veld P, Renstrom E, Schuit FC: Redox control of exocytosis: regulatory role of NADPH, thioredoxin, and glutaredoxin. Diabetes 54:2132-2142, 2005

35. Ronnebaum SM, llkayeva O, Burgess SC, Joseph JW, Lu D, Stevens RD, Becker TC, Sherry AD, Newgard CB, Jensen MV: A pyruvate cycling pathway involving cytosolic NADP- dependent isocitrate dehydrogenase regulates glucose-stimulated insulin secretion. J Biol Chem, 2006

36. Prentki M, Corkey BE: Are the beta-cell signaling molecules malonyl-CoA and cystolic long-chain acyl-CoA implicated in multiple tissue defects of obesity and NIDDM? Diabetes 45:273-283, 1996

37. Mulder H, Lu D, Finley Jt, An J, Cohen J, Antinozzi PA, McGarry JD, Newgard CB: Overexpression of a modified human malonyl-CoA decarboxylase blocks the glucose-induced increase in malonyl-CoA level but has no impact on insulin secretion in INS-1 -derived (832/13) beta-cells. J Biol Chem 276:6479-6484, 2001 38. Peyot ML, Nolan CJ, Soni K, JoIy E, Lussier R, Corkey BE, Wang SP, Mitchell GA, Prentki M: Hormone-sensitive lipase has a role in lipid signaling for insulin secretion but is nonessential for the incretin action of glucagon-like peptide 1. Diabetes 53:1733-1742, 2004