Inhibition of Foxo1 Protects Pancreatic Islet Β-Cells Against Fatty Acid And

Diabetes Publish Ahead of Print, published online January 9, 2008

Inhibition of Foxo1 Protects Pancreatic Islet ®-Cells Against Fatty Acid and ER-Stress Induced Apoptosis

Sara C. Martinez1, Katsuya Tanabe1, Corentin Cras-Méneur1, Nada A. Abumrad2, Ernesto Bernal-Mizrachi1, M. Alan Permutt1

1 Division of Endocrinology, Metabolism, and Lipid Research, 2 Department of Medicine, Division of Nutritional Science Washington University School of Medicine, St. Louis, MO

Running title: Foxo1 Inhibition Protects Islet ®-Cells

Corresponding Author: M. Alan Permutt, MD Division of Endocrinology, Metabolism, and Lipid Research Washington University School of Medicine 660 S. Euclid Avenue, Campus Box 8127 St. Louis, MO 63110 [email protected]

Received for publication 30 April 2007 and accepted in revised form 27 December 2007

Additional information for this article can be found in an online appendix at http://diabetes.diabetesjournals.org.

ABSTRACT

Objective: β-cells are particularly susceptible to fatty acid (FA) induced apoptosis associated with decreased insulin receptor/PI3-kinase/Akt signaling and the activation of stress kinases. We examined the mechanism of FA induced apoptosis of mouse - cells especially as related to the role played by ER stress-induced Foxo1 activation and whether decreasing Foxo1 activity could enhance cell survival.

Research Design and Methods: Mouse insulinoma (MIN6) cells were treated with FA and the role of Foxo1 in mediating effects on signaling pathways and apoptosis was examined by measuring Foxo1 activity and using dominant negative Foxo1.

Results: Increasing FA concentrations (100-400 μM palmitate or oleate) led to early Jun-N-terminal kinase (JNK) activation that preceded induction of ER stress markers and apoptosis. Foxo1 activity was increased with FA treatment and by pharmacologic inducers of ER-stress and this increase was prevented by JNK inhibition. FA induced nuclear localization of Foxo1 at 4 hours when Akt activity was increased indicating that FoxO1 activation was not mediated by JNK inhibition of Akt. In contrast, FA treatment for 24 hours was associated with decreased insulin signaling. A dominant-negative Foxo1 adenovirus (Adv-DNFoxo) conferred cells with protection from ER stress and FA mediated apoptosis. Microarray analysis revealed that FA induction of gene expression was in most cases reversed by Adv-DNFoxo, including the pro-apoptotic transcription factor, CHOP.

Conclusions: Early induction of JNK and Foxo1 activation play an important role in FA- induced apoptosis. Expressing a dominant-negative allele of Foxo1 reduces expression of apoptotic and ER stress markers and promotes β-cell survival from FA and ER- stress, identifying a potential therapeutic target for preserving β-cells in type 2 diabetes.

2 Foxo1 Inhibition Protects Islet ®-Cells

nsulin-resistant states are regulation of insulin signaling (12). JNK characterized by an initial adaptive also activates Foxo1 by phosphorylation Iexpansion of β-cell mass to increase at sites independent of those insulin production for maintaining phosphorylated by Akt (threonine 447 and euglycemia. However, a significant threonine 451) (13; 14). Deletion of JNK number of insulin resistant individuals in mice reduces the obesity-associated develop impaired insulin secretion insulin-resistance (15). In summary, FA associated with a reduction in β-cell mass treatment of β-cells has been shown in that subsequently results in separate studies to activate JNK, to hyperglycemia and overt diabetes (1). A promote ER stress, and to decrease common feature of insulin resistant states insulin signaling. How these pathways is the high serum fatty acids (FA) (2). In interact and the sequence of events addition to reducing insulin sensitivity in involved in FA-induced apoptosis of β- peripheral tissues, FAs also impact islet cells remain unknown. A prime candidate β-cell function. While short term FA for mediating FA induced toxicity exposure (hours) augments insulin downstream of decreased insulin secretion, chronic exposure (days) results signaling is the pro-apoptotic anti- in decreased secretion in rodent and proliferative transcription factor Foxo1 human islets and in insulinoma cells in that is negatively regulated by insulin/PI3- culture (3; 4; 5, respectively). Chronic kinase/Akt activation. Phosphorylation of elevation of FAs has also been shown to Foxo1 by Akt promotes its nuclear be associated with increased β-cell exclusion and inhibits its function (16). apoptosis in rodents, and has been Glucose stimulation of β-cells resulted in implicated in etiology of the reduced β- Akt phosphorylation with inhibition of cell mass of type 2 diabetes(6; 7). Foxo1 nuclear activity by an Oleate treatment of insulinoma cells autocrine/paracrine effect of released induced apoptosis that was dependent on insulin on its receptor (17). In contrast, FA decreased protein kinase B (PKB)/Akt treatment of insulinoma β-cells decreased phosphorylation (8). Palmitate and oleate phosphorylation of Akt and Foxo1 (8). induced apoptosis was associated with Oxidative stress resulted in nuclear to expression of endoplasmic-reticulum (ER) cytoplasmic translocation of Pdx1 and stress markers such as Bip and CHOP this was blocked by dominant negative (7). Pharmacologic induction of ER-stress Foxo (18). However, none of the previous within the β-cell was associated with studies evaluated the role of Foxo1 decreased Akt phosphorylation, and with activation in FA-induced β-cell death. We activation of the c-Jun NH2-terminal here demonstrate an essential role of kinase, JNK (9). In the liver, high FAs Foxo1 activation in β--cell apoptosis. We promote an ER-stress response, show that FAs as well as pharmacologic associated with JNK activation (10). In ER-stress inducing agents rapidly induce adipocytes, FA reduction of insulin JNK, the ER stress response, and signaling is associated with JNK increase Foxo1 nuclear localization and activation, and is reversed by JNK activity at a time when Akt inhibition (11). JNK phosphorylates an phosphorylation is increased. Foxo1 inhibitory serine residue (serine 307) on activation is blocked by JNK inhibition. the IRS-1 protein, which results in down Expressing a dominant-negative allele of

3 Foxo1 Inhibition Protects Islet ®-Cells

Foxo1 reduces expression of apoptotic Animal Studies Committee. Animals were and ER stress markers and promotes β- killed in a carbon dioxide chamber. Islets cell survival from FA and ER-stress. A of 8-10 week-old wildtype male C57BL/6 potential mechanism for the pro-apototic mice were isolated by ductal collagenase action of Foxo1 is provided by distension/digestion of the pancreas (20), documenting its transcriptional regulation followed by filtering and washing through of CHOP. a 70 μM Nylon cell strainer (BD Biosciences, San Jose, CA). Isolated RESEARCH DESIGN AND METHODS islets were then maintained in RPMI Materials. The thapsigargin, tunicamycin, medium containing 11 mM glucose, 10% oleic acid, palmitic acid, formalin, DAPI, FBS, 200 units/mL Penicillin, and Hoechst dye and propidium iodide were 200 g/mL streptomycin in humidified 5% purchased from Sigma (Saint Louis, MO). CO2, 95% air at 37 C. The SP610025 was from Tocris Fatty Acid Treatment of MIN6 Cells and Bioscience (Ellisville, MO). Antibodies Islets. MIN6 cells were incubated in used were anti- -Tubulin (monoclonal; modified DMEM media with 0.5% (w/v) Sigma), anti-Bip, anti-Gadd153/CHOP BSA alone or 0.2, 0.3, or 0.4 mM (Santa Cruz Biotechnology, Santa Cruz, palmitate or 0.4 mM oleate complexed to CA), anti-phosphoEIF2 , anti-phospho- 0.5% (w/v) BSA for 24 hours. Preparation JNK1/2, anti-total-JNK1/2, anti-phospho- of the 0.4 mM FA media was carried out cJun, anti-Cleaved Caspase 3, anti- as described (8). Briefly a 20 mM solution Foxo1, anti-phospho-Ser473-Akt, anti- of the FA in 0.01 M NaOH was incubated phospho-Thr308-Akt (Cell Signaling at 70°C for 30 minutes. Then, 330 μL of Technology, Beverley, MA). The anti-HA 30% BSA and 200, 300, or 400 μL of the antibody is a mouse monoclonal from free FA/NaOH mixture was mixed Covance Research Products together and filter sterilized with 20 mL of (Cumberland, VA). All primary antibodies either the DMEM or RPMI media. These were used at a 1:1000 dilution in 5% conditions were chosen as we originally BSA/TBS-T/0.1% NaN3, with the showed that 25mM vs. 5mM glucose exception of the CHOP antibody, which inhibited Foxo1 activation (Martinez, was used at a 1:200 dilution in 5% 2006) and was protective of apoptosis milk/TBS-T. (Srinivasan, 2002). In the current studies, Cell Culture of Mouse Insulinoma we chose to use standard culture media Cells. MIN6 cells were maintained in (25 mM) to begin the experiments with Dulbecco’s modified Eagle’s medium Foxo1 activity and apoptosis at a (DMEM) containing 25 mM glucose, with minimum in order to optimize the effects 15% fetal bovine serum (FBS), 100 of FFA. The approximate molar ratio of units/mL Penicillin, 100 g/mL FA:BSA is 6:1 with 0.4 mM palmitate. The streptomycin, 100 g/mL L-glutamine, addition of BSA or a FA:BSA mixture has and 5 μL/L -mercaptoethanol in not shown to affect the pH of the media. humidified 5% CO2, 95% air at 37 C For TUNEL staining, cells were kept in (19). MIN6 cells used were between culture in Lab-Tek II, CC2-treated passages 21-31. chamber slides (Nunc, Rochester, NY). Islet Isolation and Culture. All Propidium Iodide/DAPI Cell Death procedures were performed in Assay. MIN6 cells were grown on glass accordance with Washington University’s coverslips within the wells of a 6-well

4 Foxo1 Inhibition Protects Islet ®-Cells

plate and incubated with, either normal nitrocellulose membranes, followed by media, thapsigargin, BSA alone or 0.2, immunoblotting using all primary 0.3, or 0.4 mM palmitate complexed with antibodies according to manufacturer’s BSA for 24 hours. For the last hour of instructions. Immunodetection was incubation, 10 g/mL of propidium iodide developed with ECL Advance (Amersham (PI) and 20 g/mL DAPI were added Biosciences, Buckinghamshire, U.K.) and directly to the media. After this incubation, imaged with a charge-coupled device the MIN6 cells were washed 3X with PBS camera (Alpha Innotech, San Leandro, and fixed with 3.7% formalin for 15 CA). minutes at 4 C. After fixation, the MIN6 Plasmid Constructs. HA-tagged 256- cells were washed again 3X with PBS Foxo1 in the pCMV5 vector (DN-Foxo) and then mounted with anti-fading gel was a gift from D. Accili (Columbia mounting medium (Biomeda Corporation, University, New York, NY). The IGFBP-1 Foster City, CA) onto glass slides. Each promoter/luciferase gene construct condition reported represents over 600 (p925GL3) was a gift from M. Rechler cells counted by randomized field (NIDDK, NIH, Bethesda, MD). The selection. The percentage of cell-death is CHOP-promoter/luciferase construct reported as the number of PI stained (pGL2/3) was a gift from the lab of D. Ron nuclei over the total number of nuclei (Skirball Institute, New York University, stained by DAPI as quantitated by Image New York, NY). The pRL-TK control J v1.3.8s (21). TUNEL staining was vector contains the thymidine kinase performed using the Rhodamin A promoter of the herpes simplex virus Rhodamie ApopTAG kit (Amersham upstream of the Renilla luciferase BioSciences, Piscataway, NJ) according (Promega, Madison, WI). to the manufacturer’s instructions. Luciferase Assay. MIN6 cells were Nuclear and Cytoplasmic plated in 12-well plates 2 days before Fractionation. MIN6 cells were plated on transfection. At ~60-70% confluence, 10cm dishes and allowed to grow to 60- each well was transfected with 100 ng 70% confluence. Media was then IGFBP-1/luciferase or CHOP/luciferase exchanged for 0.5% BSA or 0.4 mM plasmid, 20 ng pRL-TK control vector, palmitate/0.5% BSA or 0.4 mM and wither 100 ng of the pCMV5- 256- oleate/0.5% BSA media for the times Foxo1 or an empty pCMV5 vector in 2 μL indicated. Cells were trypsinized and Lipofectamine 2000 (Invitrogen, Carlsbad, fractionated according to a previously CA) in 100 μL of OptiMem. For cell lysis, published protocol evaluating Foxo1 200 μL of passive lysis buffer (Promega) localization (18). was used. The firefly and Renilla Western Blotting. Protein was extracted luciferase activities were measured after with a cell lysis buffer (diluted from 10X the indicated hours of incubation with cell lysis buffer from Cell Signaling BSA or 0.4 mM palmitate in a Monolight Technology and an additional protease 3010 luminometer (BD Biosciences, San cocktail tablet from Roche at 1 Jose, CA) using the dual-luciferase tablet/10 mL final buffer volume). Protein reporter assay system (Promega). samples (30 g) were separated by SDS- Adenovirus Infection. The HA-tagged electrophoresis through either 4-15% dominant-negative Foxo1 adenovirus gradient or 15% polyacrylamide gels (AdV- 256-Foxo1) and the HA-tagged (BioRad, Hercules, CA) and transferred to wildtype Foxo1 adenovirus were

5 Foxo1 Inhibition Protects Islet ®-Cells generous gifts from D. Accili (22). The palmitate complexed to 0.5% (w/v) BSA GFP-adenovirus was a gift from D. Kelly for 24 hours. Preparation of the 0.4 mM (Washington University, St. Louis, MO). free fatty acid media was carried out as Infection of the MIN6 cells was carried out described (8). Briefly a 20 mM solution of at the indicated m.o.i. for one hour in palmitic acid (Sigma) in 0.01 M NaOH serum-free media. The MIN6 were then was incubated at 70°C for 30 minutes. washed in PBS, maintained in the Then, 330 μL of 30% BSA and 400 μL of DMEM/15% FBS media, and then the palmitic acid/NaOH mixture was experiments were carried out 24 hours mixed together and filter sterilized with 20 after infection. mL of the RPMI media. The approximate Generation of β-cell specific dominant molar ratio of FFA:BSA is 6:1 with negative Foxo transgenic mice. The 0.4 mM palmitate. MIN6 cells were truncated and hemaglutinin (HA) tagged incubated in modified DMEM media with dominant-negative Foxo1 allele (DNFoxo) 0.5% (w/v) BSA alone or 0.4 mM (23) was a generous gift from D. Accili palmitate with 0.5% (w/v) BSA for 24 (Columbia University, New York, NY). It hours. was inserted into a RIP-I/β-globin Thapsigargin treatment of islets. After expression vector (24; 25), sequenced, isolation and a 6 hour recovery period and microinjected into fertilized eggs of after isolation in RPMI/10% FBS media, C57BL/6 x CBA mice according to primary islets from wild type or RIP- standard protocol of the Mouse Genetics DNFoxo mice were pooled and divided Core of the Washington Univerisity into 12 well plates with 1 mL of School of Medicine. Nine founders were RPMI/10%FBS media to be treated with obtained, of which five passed the either 10 µM thapsigargin (Sigma) (10 µL transgene through germline transmission, of a 1 mM stock) or the vehicle (10 µL of evidenced by genotyping and ethanol) alone for 48 hours. backcrossing to C57BL/6J mice (The Statistical analysis. Western blot Jackson Laboratory). Of the five which quantitation consisted of acquiring a transmitted, one line maintained per/lane ratio of the chemiluminescent consistent and strong HA-staining within signal intensities from the phospho- pancreatic islets, designated as RIP- Ser256-Foxo1 or Gadd153 to its respective DNFoxo. Transgenic and non-transgenic α-Tubulin using NIH-ImageJ (21), then littermate male (designated as wild type normalizing each signal to the first or WT) mice from C57BL/6J-backcrossed condition of the starved state in each F4-F7 generations of this one line, known respective blot. Significance in the as RIP-DNFoxo, were used in all western blot figures noted with an experiments in accordance with asterisk (*) represents a p-value <0.05 in Washington University’s Animal Studies significance different from the starved Committee. These mice are currently state. A double asterisk (**) indicates being phenotyped. significance (p < 0.05) from the intensity Fatty acid treatment of primary islets of the glucose phosphorylation of Foxo1. and MIN6 cells. Primary islets from wild For the luciferase assays, ratios of type or RIP-DNFoxo mice were pooled (3 luciferase to renilla were generated with per genotype) and divided to be standard deviations and error. Error was incubated in modified RPMI media with propagated for fold calculations between 0.5% (w/v) BSA alone or with 0.4 mM different conditions. Student’s two-tailed t-

6 Foxo1 Inhibition Protects Islet ®-Cells test for independent samples was used in arrays on the chip. For each gene, the significance calculations. The p-value for Illumina software calculated the p-value significance is mentioned in the figure of the reliability of the spot. Only spots legends. that had a significant p-value were kept RNA isolation from MIN6 cells. MIN6 for further analysis. For each gene, the cells infected with either a GFP or per-chip variance was integrated across DNFoxo adenovirus were maintained in the 30 repeats per gene on each array triplicate samples within 6-well plates in and the array variance. For each ratio DMEM with either 0.5% BSA or 0.5% (GFP Palm : BSA; DNFoxo Palm : BSA, BSA and 0.4 mM palmitate for 24 hours. 95% confidence intervals were calculated Each well was then aspirated and rinsed taking in account internal repeats (approx once with PBS. Then 1 ml of ice cold 30) on multiple gene arrays. The cutoff Trizol (Invitrogen, Carlsbad, CA) was fold change was arbitrarily chosen at 2.0: added to each well. The protocol was Genes up- or down-regulated over two followed according to manufacturer until fold and with a 95% confidence interval the aqueous phase was isolated. RNAse not overlapping 0 (in Log10 scale) were free 70% ethanol was added (twice the considered for further analysis. aqueous volume), and the mixture was Annotation for the microarray platform run through a purification column within was collected through the SOURCE the RNeasy mini-kit (Qiagen Valencia, repository (26). All transcripts significantly CA). Purification of total RNA was regulated by FA were then clustered completed using the kit and eluted with according to their Gene Ontology (27) RNase free water. RNA quality was functions using Genesis 1.7.2 (28). assessed with gel electrophoresis and the The normalized, unprocessed data is Agilent bioanalyzer (Agilent currently available on Technologies, Santa Clara, CA). http://drtc.im.wustl.edu/files/apermutt/micr Microarray experiment. We utilized the oarray/FA/normalized_data.csv and will MouseRef-8 Expression BeadChip be made available in the MIAME standard (Illumina, Inc., San Diego, CA), which on the NCBI Gene Expression Omnibus allows for the processing of 8 samples to http://www.ncbi.nlm.nih.gov/geo/. be analyzed simultaneously for the The microarray data was validated by quantitation of absolute expression of up real-time PCR on selected genes. cDNA to 16,435 genes and controls with a 30- was generated (Superscript III, fold redundancy of 50-mer oligos per Invitrogen) and amplified by real-time gene. For control samples, four PCR (Power SYBR Green Master Mix, independent RNA preparations from the Applied Biosystems) on a ABI 7500 Real- AdV-GFP BSA were pooled together, as Time PCR system (Applied Biosystems). were four independent RNA samples Primers are shown in Table S2. Results from the AdV-DNFoxo infected samples were normalized to the acidic ribosomal treated with BSA and without palmitate. protein 36B4 and quantified using the These two control groups occupied two comparative Ct method (29). arrays. For the experimental condition Promoter analysis. For each gene treated with 0.4 mM palmitate, three analyzed, 2500 bp were collected from independent RNA samples for each of the the NCBI ENTREZ databases GFP and DNFoxo virally infected samples (http://www.ncbi.nlm.nih.gov/) using EZ- occupied the other six gene expression Retreive

7 Foxo1 Inhibition Protects Islet ®-Cells

(http://siriusb.umdnj.edu:18080/EZRetriev of the stress-activated kinases also e/). Perl Global Regular Expression known as c-Jun N-terminal kinases, Patterns were then used in BBEdit 8.6.1 JNK1/2, in conjunction with decreased (Bare Bones Software, Inc. Bedford, MA, insulin signaling (10; 11; 15). We USA) to find all consensus sequences or confirmed these observations using a 24h their reverse complements in the treatment of MIN6 cells with palmitate sequences. and oleate. As FA concentration was increased, an apparent dose-response RESULTS increase in JNK1/2 activation was Fatty acids induce dose-dependent observed as evidenced by the two bands apoptosis in MIN6 cells. Glucose- representing increased JNK1/2 responsive MIN6 insulinoma cells were phosphorylation and concomitant assessed for apoptosis at 24-hours after phosphorylation of its substrate target, c- treatment with a saturated FA, palmitate Jun (Figure 2A). This treatment did not (16:0), or a monounsaturated FA oleate affect the total amount of the JNK (18:1). Effects of palmitate vs. oleate proteins. The effects on JNK activation were compared because of the known appeared to be greater for palmitate than different potencies of saturated versus for oleate. A decrease in Akt monounsaturated FA in inducing cell phosphorylation, most evident with 0.4 apoptosis (6; 30; 31), and in contributing mM palmitate or oleate was noted (Figure to symptoms of diabetes and 2B), consistent with previous findings (8). cardiovascular disease (32). Cells were Pdx-1 is a transcription factor known incubated at a constant 0.5% BSA (w/v) to be involved in the regulation of insulin and at concentrations of palmitate or production and protection of pancreatic oleate of 0.2, 0.3, and 0.4 mM. Apoptosis, islet β-cells from apoptosis (33). A recent characterized by propidium iodide report observed that palmitate treatment staining, normalized with DAPI staining, of rat islets resulted in decreased insulin was visibly enhanced with increasing gene expression that was associated with concentrations of FAs, illustrated with a nuclear to cytoplasmic translocation of palmitate in Figure 1A. A dose response Pdx-1 without a change in total Pdx-1 was observed in cell death rates reaching protein (34). In our studies a dose- 11.6% and 3.4%, respectively, for dependent decrease in Pdx-1 protein treatment with 0.4 mM palmitate and expression was observed with FA oleate (Figure 1B). Evidence that treatment (Figure 2C). This was palmitate-induced cell death in MIN6 associated with dose-dependent insulinoma cells is contributed at least in increases in expression of the pro- part by apoptosis, was demonstrated by apoptotic transcription factor CHOP. increased TUNEL staining (Figure S1). Cleaved and activated Caspase 3 also Fatty acids induce JNK activation, a increased, consistent with its role as a dose-dependent reduction in insulin major participant in the apoptotic cell signaling, and reciprocal increased death pathway. expression of apoptotic markers and JNK activation by fatty acids precedes decreased Pdx-1 expression. Several markers of ER-stress and Caspase 3 recent studies in peripheral insulin target activation. To define the sequence of tissues have documented the relationship molecular events that occur in β-cells between FA treatment and the activation upon FA exposure, MIN6 cells were

8 Foxo1 Inhibition Protects Islet ®-Cells

incubated with either 0.4 mM palmitate or translocation and mitochondrial stress. oleate and analyzed over an 8-hour Fatty acid treatment and activators of period (Figure 3A). JNK activation, ER stress in MIN6 cells result in an evidenced by phosphorylation of what increase in Foxo1 nuclear localization appeared to be 3 isoforms, was noted as and transcriptional activation. To early as one hour after treatment with examine the relationship between FA either FA. The lowest phosphorylated treatment, ER stress and Foxo1 activity, JNK band, possibly JNK3, which has MIN6 cells were treated with palmitate or been shown to be expressed primarily in oleate for 4 hours, and the intracellular neurons (35), demonstrated transient localization of Foxo1 was examined. As expression. Activated JNK1/2 persisted shown, FA induced a shift of Foxo1 from over the 8-hour incubation, further the cytoplasm to the nucleus (Figure 4A). evidenced by phosphorylation of the JNK Consistent with the increased effect of substrate, c-Jun. palmitate on apoptosis and intracellular The temporal relationship between FA signaling, palmitate treatment induced induced ER-stress and JNK activation more Foxo1 translocation than oleate was evaluated next. While FA treatment treatment. Using an IGFBP-1 promoter- resulted in insignificant changes in BiP luciferase construct containing a Foxo1 (HSPA5, GPR78) protein, a major binding site (39), palmitate treatment chaperone of the ER lumen (36), there resulted in a sustained increase in was a progressive increase in endogenous Foxo1 transcriptional activity phosphorylation of eIF2 , a protein that from 4 to 24 hours of incubation (Figure attenuates translation and initiates a 4B). Next the relationship between ER program of apoptosis with continued stress and Foxo1 transcriptional activity in stress exposure (37). This was most insulinoma cells was assessed. We apparent at 4 hours (Figure 3B) and was previously demonstrated that the ER followed by increased expression of stress agent thapsigargin reduced Akt CHOP and cleaved Caspase 3. These phosphorylation (9). As shown in Figure data showed that JNK activation 4C, the addition of thapsigargin (1 μM) to preceded induction of ER stress markers MIN6 cells resulted in significant cytosolic that in turn preceded induction of to nuclear translocation of a GFP-tagged apoptotic markers. Foxo1 within one hour of treatment that To confirm the role of ER stress in FA was sustained up to 18 hours. Using an induced apoptosis, cells were examined IGFBP-1 promoter-luciferase construct 24 hours after palmitate treatment in the containing a Foxo1 binding site (39), we presence and absence of the ER stress assessed Foxo1 activity of MIN6 cells inhibitor TUDCA, a chemical chaperone after 12 hours of treatment with two ER- (38). Addition of TUDCA reduced the stress inducing agents, thapsigargin levels of cleaved Caspase 3, phospho- (1 μM) and tunicamycin (2 μg/mL). eIF2a, and CHOP protein expression by Consistent with an increase in Foxo1 Western analysis (Figure 3C), although nuclear translocation, a significant incomplete inhibition suggests the increase in Foxo1 activity was observed possibility of other mechanisms (data not shown). The results of these contributing to FA induced cell death. studies indicated that the effect of FA TUDCA is not a selective inhibitor of ER treatment to induce ER stress could stress, and also inhibits BAX contribute to activation of Foxo1.

9 Foxo1 Inhibition Protects Islet ®-Cells

JNK activity contributes to Foxo1 expression of a constitutively-active Akt localization and apoptosis in prevented oleate- induced apoptosis (8). palmitate-treated MIN6 cells. The data We confirmed a decreased Akt in Fig 3 indicated that JNK activation is phosphorylation 24 hours after FA observed early after FA treatment. To test treatment (Figure 2B). However, the time JNK’s contribution to the action of FA to course of Akt phosphorylation after FA induce Foxo1 and apoptotic markers we shown in Figure 5E was complex. An used SP600125, a JNK-specific inhibitor early decrease at 1 hour after FA (40). The addition of SP600125 resulted treatment was consistently followed by in a dose dependent decrease of increased phosphorylation at 4 hours palmitate-induced phosphorylation of c- when Foxo1 activity was increased. Thus, Jun along with expression of cleaved the JNK-dependent Foxo1 activation Caspase 3 (Figure 5A). Treatment with (Figure 5B) appeared to reflect a direct 0.4 mM palmitate resulted in an increase action of JNK rather than JNK effects to of nuclear Foxo1 at four hours (Figure decrease insulin signaling. We have 5B). A one-hour pre-incubation and observed a time course for GSK3 continued incubation with 300nM of phosphorylation that is very similar to that SP600125 resulted in maintenance of of Akt phosphorylation (data not shown). Foxo1 in the cytoplasm and decreased This observation supports our Foxo1 in the nucleus despite the interpretation that Akt is presence of 0.4 mM palmitate. A repeat phosphorylated/activated following its of this experiment for a 12-hour early transient dephosphorylation and incubation yielded similar findings. As cell that this does not coincide with FoxO fractionation can result in cross retention in the cytoplasm. The contamination of nuclear and cytoplasmic progressive decline in Akt proteins, to confirm these findings the phosphorylation that followed after 4 hour experiment was repeated by direct (Figure 5E) suggested that decreased visualization with a GFP-tagged Foxo1 insulin signaling likely contributes showing significant cytosolic to nuclear additionally to JNK induced Foxo1 translocation following 4 hours treatment activation at later time points. with palmitate, and a reduction following Inhibiting Foxo1 activity reduces ER- JNK inhibition with SP600125 (Figures stress and palmitate-induced 5C &5D). apoptosis in MIN6 cells. To directly These results demonstrated that JNK examine the contribution of Foxo1 to FA- activation contributes an important part of induced ER stress and cell death we FA-induced cytoplasmic to nuclear used an adenoviral construct of an HA- translocation of Foxo1. JNK activation tagged dominant-negative Foxo (AdV- enhances transcriptional activity of Foxo1 DNFoxo) and evaluated its effects on ER- via direct binding and active stress and FA induced apoptosis in MIN6 phosphorylation of Foxo1 (14). However, cells. This DNFoxo is a truncated allele of JNK could also activate Foxo1 indirectly Foxo1 that retains the DNA-binding through inhibition of insulin signaling (41). domain (amino acid residues 1-256) but The latter mechanism was previously lacks the transactivation domain (23). suggested since decreased Akt activity This construct was created to examine was measured 24 hours after oleate the contribution of the 3 Foxo proteins treatment of insulinoma cells, and over- expressed in β-cells under normal

10 Foxo1 Inhibition Protects Islet ®-Cells conditions, and with loss of β-cell mass in µM thapsigargin for 48 hours to induce insulin resistant models. This allele has ER-stress.This resulted in cleaved been demonstrated to function as a Caspase3 expression only in wild type dominant-negative inhibitor of islets (Figure 7A). The resistance to endogenous Foxo activity in the liver (42), apoptosis in RIP-DNFoxo islets also in myoblasts (43), and in the β-cell ((17; applied to FA treatment. Isolated islets 18)). MIN6 cells transfected with 200 from wild type and RIP-DNFoxo mice m.o.i. of the DNFoxo adenovirus cultured in the presence of 0.4 mM exhibited a nearly 60% reduction in palmitate for 24 hours exhibited similar propidium iodide staining and cell death JNK phosphorylation without changes in when exposed to 100nM thapsigargin, total JNK protein. (Figure 7B). However, compared to that in control GFP protein expression of the pro-apoptotic adenovirus (AdV-GFP) transfected cells transcription factor Chop was only (Figure 6A and 6B). Similar protection observed in wild type islets. RIP-DNFoxo against cell death was also observed with islets also resisted Caspase3 activation palmitate treatment of MIN6 cells (Figure by palmitate, which was reproduced with 6C and 6D). The effects of adenoviral wild type primary islets. Additionally, the treatment after 24 hours of palmitate were abundance of phosphorylated Akt and the further examined by western blot analysis transcription factor, Pdx1, were elevated of protein extracts (Figure 6E). Increasing in RIP-DNFoxo islets when compared to the m.o.i. of the adenovirus resulted in similarly treated wild type islets. augmented expression of the HA-tagged Microarray analysis reveals that most DNFoxo protein, indicated by western blot of the genes regulated by palmitate using an HA-antibody. Increasing titers of treatment are Foxo dependent. To the AdV-DNFoxo in cells exposed to define the contribution of Foxo activation 0.4 mM palmitate for 24 hours resulted in to FA regulation of gene expression decreased expression of CHOP, along relevant to apoptosis, we performed two with a significant reduction in cleaved parallel microarray experiments. A first Caspase 3 as compared to cells similarly experiment compared expression profiles treated with palmitate but transfected with in cells treated or not treated with FA and AdV-GFP. These results correlate with infected with Adv-GFP. The second the observation of DNFoxo reducing the experiment further evaluated the gene percentage of propidium iodide positive expression profiles of MIN6 cells in FA cells (Figure 6D). They support the treated and untreated cells but infected interpretation that Foxo1 activation is with AdV-DN-Foxo. This latter experiment important in mediating cell death after ER permitted comparison of the effects of stress and FA treatment. FAs in the presence and absence of Foxo RIP-DNFoxo islets resist ex vivo signaling. Expression profiles of stimuli of apoptosis. The results approximately 16000 genes were obtained with MIN6 cells expressing assessed. After 24 hours treatment with DNFoxo were confirmed with primary 0.4 mM palmitate and AdV-GFP, 124 islets as shown in Figure 7. Islets were genes were significantly regulated (see isolated from transgenic RIP-DNFoxo Methods) over two-fold (91 genes up- mice expressing the dominant-negative regulated and 33 down-regulated). The Foxo driven by the rat insulin promoter Foxo-dependent FA regulated genes, and were cultured in the presence of 10 were clustered according to their

11 Foxo1 Inhibition Protects Islet ®-Cells

Ontology physiological processes (27; transcription of CHOP with nutrient 44). The degree of activation of the withdrawal that was inhibited by co- individual regulated genes and the effects expression with a plasmid expressing of treatment with AdV-DNFoxo are DNFoxo (17). This established that illustrated in Table 1. There was a CHOP-induced expression with nutrient pronounced participation of genes withdrawal was Foxo dependent. In the involved primarily in cellular metabolism, current studies, we sought to determine cellular physiology, and transport. These how FAs and ER stress-induced CHOP included a number of genes that could expression and Foxo activation were influence FA induced apoptosis. related. MIN6 cells were transfected with Immediately apparent was that the a luciferase construct driven by 10kb of overwhelming majority of genes that were the CHOP promoter (45), either an empty regulated by palmitate in the control pCMV vector or pCMV-DNFoxo, and a condition were not regulated to the same thymidine kinase promoter driving renilla extent when Foxo was inhibited, expression as a control vector. CHOP suggesting that the FA induced promoter activity was induced in cells, transcriptional regulation of these genes treated with 0.4 mM palmitate for 24 is predominantly Foxo-dependent (Figure hours, as shown in Figure 8. Co- S2). The magnitude of expression of the transfection of the DNFoxo plasmid genes up- and down-regulated by significantly reduced CHOP promoter palmitate in control condition (AdV-GFP) activity in a dose-dependent manner. and in the presence of AdV-DNFoxo are Non-specific effects of pCMV-DNFoxo on further illustrated in Figures S2-A and -B. CHOP-luciferase were corrected by These results were validated by an normalization with inclusion of pTK- independent assessment of changes in renilla. These results clearly mRNA by quantitative RT-PCR on demonstrated that ER-stress induced selected transcripts (Figure S3). CHOP expression is regulated at least in Foxo1 regulates ER stress and fatty part by Foxo activation. In the absence of acid induced CHOP (Ddit3) expression. mutation of the Foxo binding sites in the One of the genes that contained multiple CHOP promoter however, these results putative conserved Foxo binding sites are consistent with an indirect effect of whose transcription was activated by FA Foxo activation on CHOP expression. treatment was CHOP (Ddit3) (Table 1). Previous studies had noted an DISCUSSION association between FA-induced ER This study was designed to examine stress and CHOP expression in the sequence of molecular signaling insulinoma cells (7; 30). We previously events in FA induced apoptosis in studied genes activated in MIN6 insulinoma cells and islets and to test the insulinoma cells on nutrient withdrawal by hypothesis that Foxo1 activation plays a microarray analysis and noted that the significant role in this process. Fatty acids most highly expressed gene was DNA have been shown to contribute to β-cell damage inducible transcript 3 (Ddit3, apoptosis by decreasing insulin/Akt CHOP, Gadd153). Observing a signaling (7; 8). Only one study examined conserved canonical Foxo binding site in Foxo1 involvement in this process, the CHOP promoter, we demonstrated measuring changes in phosphorylation of with a CHOP-luciferase assay increased Foxo1 as an indicator of protein activity,

12 Foxo1 Inhibition Protects Islet ®-Cells

but not definitively or directly implicating effects (31; 46; 48; 49). The relatively Foxo in the process (8). In the current greater effects of the saturated FA on study Foxo involvement is documented in Foxo1 activation and apoptosis may have a definitive way, showing changes in its relevance to nutritional treatment of nuclear localization and its role in patients with T2DM. transcriptional effects of fatty acids. In A time course of treatment with either addition, these studies determined the palmitate or oleate in MIN6 cells earliest events in FA-induced apoptosis suggested that JNK phosphorylation and documented the involvement of JNK precedes markers of ER stress and (as opposed to Akt) in the early action of apoptosis, including phosphorylation of FA to induce nuclear localization of eIF2 , expression of CHOP, and the Foxo1. Compelling evidence is presented eventual increase of cleaved Caspase 3. to indicate that ER stress via activation of In addition, JNK inhibition decreased JNK and Foxo1 contributes to the nuclear Foxo1 along with the reduction in apoptotic outcome after FA treatment. FA induced apoptosis. These findings Finally, novel evidence shows that FA position JNK as an important, early alter expression of a large number of participant upstream of Foxo1 in FA proapototic genes in islets and that many activation of the pro-apoptotic pathway. of these effects are mediated by Foxo1. Previous studies on islets and insulinoma Treatment of MIN6 cells with palmitate cells demonstrated JNK activation via or oleate at varying concentrations oxidative stress, cytokines, and ER-stress established a pattern of intracellular inducing agents (9; 18; 50) but the signaling that was common to both FAs. underlying mechanisms remain unclear. The data indicated that both FAs regulate A recent study demonstrated palmitate the same signaling events with palmitate binding to the pro-inflammatory Toll-Like being more potent possibly as a result of Receptor 2 (TLR2), leading to enhanced its less effective processing by JNK activation and insulin resistance in esterification (6; 30; 31). In various murine myocytes (51). Another report mammalian cell lines including β-cells, documented dependence of JNK lower FA toxicity was associated with activation in macrophages on signaling increased cellular capacity for triglyceride via the membrane FA translocase CD36 (TG) synthesis and with up-regulation of [Rahaman SO et 2006]. stearoyl coenzyme A desaturase 1 The important role of decreased (SCD1) (31; 46). SCD1 functions in FA insulin signaling and Akt phosphorylation desaturation prior to incorporation into in FA-induced β-cell apoptosis has been TG, and its expression was shown to be reported previously (8). However, our down-regulated by Foxo1 (47). This effect data would indicate that early activation of of Foxo1 could contribute to cell death Foxo1, which would set in motion the indirectly by inhibiting FA incorporation apoptotic program after FA, is not into the neutral TG pool, consequently dependent on Akt activity. The transient increasing available FA and the potential apparent increase in Akt activity at 4 for accumulation of toxic metabolites such hours after FA treatment is unexplained, as ceramides (48; 49). Possibly, FA although it could be secondary to the activation of Foxo1 with subsequent acute stimulatory effect of FA on insulin inhibition of SCD1 would create a self- secretion (5). Alternatively, a recent sustained cycle with deleterious cellular report has described a complex

13 Foxo1 Inhibition Protects Islet ®-Cells

interaction between the activities of Foxo codon for all of the FA regulated genes and Akt (52). Treatment of primary were assessed in the mouse, rat, and hepatocytes with either wild type, human genomes for those that contained constitutively nuclear, or dominant an evolutionarily-conserved seven base negative Foxo adenovirus resulted in pair sequence (T[A/G]TT[T/C][A/C]C) activation of PKB pathways. In the current encoding Forkhead binding sites (53; 54). report, at 4 hours after FA treatment, a By chance alone, the consensus binding time when Foxo1 was observed to be site can be found once every 1024 base nuclear and transcriptionally active, we pairs (44x23x0.5). The number of putative also observed increased phosphorylation Forkhead binding sites for each gene is of Akt (see Figure 3B). Possibly JNK represented in Supplemental Table 1 activated Foxo1 may have contributed to (available at this transient increase in Akt http://diabetes.diabetesjournals.org) by phosphorylation. We repeatedly observed asterix appended to the symbol of the decreased Akt phosphorylation after 24 gene. Notable is that very few genes in hours of FA treatment, a finding this group had fewer than 2 potential consistent with that in another insulinoma Foxo binding sites. However, whether cell line (8). Taken together these results these genes are directly regulated by suggest that reduced Akt phosphorylation Foxo binding can only be determined by chronically contributes to maintenance of direct experimentation. the early, direct activation of Foxo1 by The FA induced genes that could be JNK. mediating apoptosis in -cells were While the current results show that FA examined by a whole genome expression treatment results in increased Foxo study. Over 100 genes exhibited more activity, previous studies (8) have shown than a two-fold up- or down-regulation in that FA treatment results in reduced expression after 24 hours of treatment phosphorylation of another Akt substrate, with palmitate (Table 1). Validation of glycogen synthase kinase 3β (Gsk3β), these expression results is suggested by that could also contribute to apoptosis. In the concomitant observation of FA preliminary studies we have confirmed reduced expression of Pdx-1, and that palmitate and oleate treatment of increased expression of CHOP proteins MIN6 cells results in reduced by Western analysis (Figures 2C & 3B) phosphorylation of Gsk3β, and extended and by RT-PCR of selected transcripts these studies to show that inhibitors of (Figure S3). A number of FA regulated Gsk3β activity reduced FA induced genes that may contribute to reduce apoptosis (ADA Annual Meeting 2007, insulin signaling or secretion were noted. http://scientificsessions.diabetes.org/inde Trib3 (Trb3) is a known inhibitor of Akt x.cfm?fuseaction=Locator.SearchAbstract activation and potentiator of apoptosis s&CalledByID=1006). These combined that has been incriminated in the insulin results thus indicate that reduced insulin resistance of type 2 diabetes (55). Its 3.6- signaling leads to activation of at least fold induction by FA suggests the two pro-apoptotic proteins, Foxo and involvement of Trib3 in FA impairment of Gsk3β, that contribute to the deleterious insulin signaling. Trib3 was also identified effects of FA on β-cells. as a target of ER stress induced Twenty-five hundred base pairs of the apoptosis in human embryonic kidney promoter regions upstream of the start (293) cells (56). Further it was shown that

14 Foxo1 Inhibition Protects Islet ®-Cells

Trib3 was down-stream of CHOP reduction in Pdx-1 gene expression activation, and that up-regulation of Trib3 following palmitate treatment which by ER stress was required for apoptosis. correlated with the decrease in Pdx-1 The hypothesis that Trib3 is downstream protein (Fig. 2C). FA reduction of Pdx-1 of FA induced ER-stress and CHOP expression was blunted by 69% with mediated apoptosis in insulinoma cells ADv-DN-Foxo treatment documenting can now be tested. Foxo-dependence of the effect. Previous Other FA regulated genes included work had shown that IRS2 deficiency was Eif4ebp1, a protein that binds EIF4E to associated with Pdx deficiency, and in inhibit translation. Phosphorylation of this model -cell destruction was Eif4ebp1, leading to its dissociation from corrected by either Foxo EIF4E has been implicated at least in part haploinsufficiency or overexpression of in insulin-mediated enhancement of Pdx-1 (61). Together these results translation and cell growth (57). Similarly emphasize the link between FA/ER stress Cdkn1a (or p21) is a protein that binds to reduction of insulin signaling, Foxo cyclin complexes and inhibits cell cycle activation and subsequent Pdx regulation progression (58), and its activation is and apoptosis. associated with ER stress and apoptosis We examined potential mechanisms of β-cells cells (59). PPAR is a for the Foxo1-dependent induction or transcription factor that senses and repression of genes by FA. β-cells have a regulates lipid metabolism. Its induction in highly developed and active ER to β-cells has been associated with maintain insulin secretion which increased FA uptake, triglyceride translates to a high susceptibility to stress synthesis, and with inhibited insulin (66). Foxo1 has been shown to contribute secretion (58; 60-62). to apoptosis through transcription of Remarkably, regulation of more than several pro-apoptotic genes such as Bim, 95% of the FA sensitive genes was FasL, and TRAIL (54). Our study blunted or completely reversed with identifies a novel mechanism by which concomitant over-expression of Foxo1 could promote β-cell apoptosis dominant-negative Foxo1. For example, a through its activation of CHOP (66). 6 fold induction by FA of Cxcl12, a pro- Conditions which increase CHOP inflammatory cytokine responsive to expression have consistently been lipopolysaccharide, TNF, or IL1 (63), was associated with ER-stress, decreased blunted by 85%. The 3 fold upregulation insulin signaling, and apoptosis in β-cells of Eif4ebp1 was reduced by 70%. Genes (7; 9). Disruption of the CHOP gene up-regulated by palmitate treatment that delayed although did not prevent the exhibited blunted response in the onset of hyperglycemia and β-cell death presence of dominant-negative Foxo1 in an in vivo model of pancreatic β-cell also included Dgat2, responsible for ER stress, the Akita mouse, conversion of diglycerides to triglycerides heterozygous for a point mutation in the (64), and Acaa2, which catalyzes the last insulin 2 gene (Ins2C96Y) (67). Thus, step of mitochondrial β-oxidation (62; 65). these results suggest redundancy in ER- Among the genes down-regulated by stress mediated pathways leading to FA treatment that may contribute to β- apoptosis, potentially exclusive of, or in cells apoptosis, Pdx1 is especially addition to, Foxo1 and Chop. interesting. There was a two-fold This study documented the role of

15 Foxo1 Inhibition Protects Islet ®-Cells

FoxO activation by FA or ER stress in pharmacologic ER-stress inducers. We mediating β-cell death. The data do not were able to significantly decrease β-cell rule out effects of FA and oxidative stress apoptosis from ER-stress or FA by over- in inducing mitochondrial dysfunction expressing a dominant-negative Foxo1. which triggers programmed cell death, as Inhibiting Foxo1 could be clinically reviewed recently (68; 69). There is relevant for promoting the survival of β- limited information related to Foxo cells in obesity and diabetes, conditions proteins and mitochondrial stress but a associated with insulin resistance, recent study suggested differential roles increased serum FA levels and ER- of Foxo1 and Foxo3a. Silencing of stress. These results thus identify Foxo1 FOXO1 did not alter susceptibility of as a potential new therapeutic target for endometrial cells to oxidative cell death preservation of β-cells mass and function. while silencing of FOXO3a prevented apoptosis induced by hydrogen peroxide ACKNOWLEDGEMENTS (70). The authors would like to thank H. Our findings suggest that further in Harding and D. Ron (New York vivo characterization of the role of Foxo1 University) for reagents, D. Accili within β-cells is merited especially in the (Columbia) for insightful comments and context of the β-cell failure associated reagents, K. Polonsky for helpful with diabetes. Prolonged treatment of comments on the experiments, Cris M. islets or insulinoma cells with FAs results Welling for lab assistance, and other in decreased glucose-stimulated insulin members of the Permutt lab for helpful secretion (5; 71). Foxo1 may mediate discussions. This work was supported in some of this effect as it contributes to the part by an NIH grant (R37 DK16746, to phenotype of insulin resistance by M.A. Permutt), a Ruth L. Kirschstein Pre- increasing FA utilization and decreasing Doctoral National Research Service glucose utilization in skeletal muscle and Award (F32 DK06537, to S.C. Martinez), by increasing gluconeogenesis in the liver and a grant from the Philip Morris USA (22; 72; 73). Foxo1 activation in β-cells External Research Program. The could likewise down-regulate glucose Washington University DRTC (NIH P60 metabolism and thus insulin secretion DK20579) is acknowledged for and synthesis. Nutrient withdrawal and assistance from the Immunoassay, decreased insulin signaling have Morphology, and Transgenic Mouse previously been shown to activate Foxo1 Cores, and the Adipocyte Biology Core, in β-cells (17), and now we extend the Clinical Nutrition Research Unit (NIH agonists of Foxo1 to include FA and DK56351).

16 Foxo1 Inhibition Protects Islet ®-Cells

REFERENCES

1. Marchetti P, Del Prato S, Lupi R, Del Guerra S: The pancreatic beta-cell in human Type 2 diabetes. Nutr Metab Cardiovasc Dis 16 Suppl 1:S3-6, 2006 2. Randle PJ, Garland PB, Hales CN, Newsholme EA: The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1:785-789, 1963 3. Newsholme P, Keane D, Welters HJ, Morgan NG: Life and death decisions of the pancreatic beta-cell: the role of fatty acids. Clin Sci (Lond) 112:27-42, 2007 4. Zhou YP, Grill VE: Long-term exposure of rat pancreatic islets to fatty acids inhibits glucose-induced insulin secretion and biosynthesis through a glucose fatty acid cycle. J Clin Invest 93:870-876, 1994 5. Noushmehr H, D'Amico E, Farilla L, Hui H, Wawrowsky KA, Mlynarski W, Doria A, Abumrad NA, Perfetti R: Fatty acid translocase (FAT/CD36) is localized on insulin- containing granules in human pancreatic beta-cells and mediates fatty acid effects on insulin secretion. Diabetes 54:472-481, 2005 6. Maedler K, Spinas GA, Dyntar D, Moritz W, Kaiser N, Donath MY: Distinct effects of saturated and monounsaturated fatty acids on beta-cell turnover and function. Diabetes 50:69-76, 2001 7. Kharroubi I, Ladriere L, Cardozo AK, Dogusan Z, Cnop M, Eizirik DL: Free fatty acids and cytokines induce pancreatic beta-cell apoptosis by different mechanisms: role of nuclear factor-kappaB and endoplasmic reticulum stress. Endocrinology 145:5087- 5096, 2004 8. Wrede CE, Dickson LM, Lingohr MK, Briaud I, Rhodes CJ: Protein kinase B/Akt prevents fatty acid-induced apoptosis in pancreatic beta-cells (INS-1). J Biol Chem 277:49676-49684, 2002 9. Srinivasan S, Ohsugi M, Liu Z, Fatrai S, Bernal-Mizrachi E, Permutt MA: Endoplasmic Reticulum Stress-Induced Apoptosis Is Partly Mediated by Reduced Insulin Signaling Through Phosphatidylinositol 3-Kinase/Akt and Increased Glycogen Synthase Kinase- 3{beta} in Mouse Insulinoma Cells. Diabetes 54:968-975, 2005 10. Ozcan U, Cao Q, Yilmaz E, Lee AH, Iwakoshi NN, Ozdelen E, Tuncman G, Gorgun C, Glimcher LH, Hotamisligil GS: Endoplasmic reticulum stress links obesity, insulin action, and type 2 diabetes. Science 306:457-461, 2004 11. Nguyen MT, Satoh H, Favelyukis S, Babendure JL, Imamura T, Sbodio JI, Zalevsky J, Dahiyat BI, Chi NW, Olefsky JM: JNK and tumor necrosis factor-alpha mediate free fatty acid-induced insulin resistance in 3T3-L1 adipocytes. J Biol Chem 280:35361- 35371, 2005 12. Aguirre V, Uchida T, Yenush L, Davis R, White MF: The c-Jun NH(2)-terminal kinase promotes insulin resistance during association with insulin receptor substrate-1 and phosphorylation of Ser(307). J Biol Chem 275:9047-9054, 2000 13. Matsumoto M, Accili D: All roads lead to FoxO. Cell Metab 1:215-216, 2005 14. Essers MA, Weijzen S, de Vries-Smits AM, Saarloos I, de Ruiter ND, Bos JL, Burgering BM: FOXO transcription factor activation by oxidative stress mediated by the small GTPase Ral and JNK. Embo J 23:4802-4812, 2004

17 Foxo1 Inhibition Protects Islet ®-Cells

15. Hirosumi J, Tuncman G, Chang L, Gorgun CZ, Uysal KT, Maeda K, Karin M, Hotamisligil GS: A central role for JNK in obesity and insulin resistance. Nature 420:333-336, 2002 16. Biggs WH, 3rd, Meisenhelder J, Hunter T, Cavenee WK, Arden KC: Protein kinase B/Akt-mediated phosphorylation promotes nuclear exclusion of the winged helix transcription factor FKHR1. Proc Natl Acad Sci U S A 96:7421-7426, 1999 17. Martinez SC, Cras-Meneur C, Bernal-Mizrachi E, Permutt MA: Glucose Regulates Foxo1 Through Insulin Receptor Signaling in the Pancreatic Islet {beta}-cell. Diabetes 55:1581-1591, 2006 18. Kawamori D, Kaneto H, Nakatani Y, Matsuoka TA, Matsuhisa M, Hori M, Yamasaki Y: The forkhead transcription factor Foxo1 bridges the JNK pathway and the transcription factor PDX-1 through its intracellular translocation. J Biol Chem 281:1091- 1098, 2006 19. Miyazaki J, Araki K, Yamato E, Ikegami H, Asano T, Shibasaki Y, Oka Y, Yamamura K: Establishment of a pancreatic beta cell line that retains glucose-inducible insulin secretion: special reference to expression of glucose transporter isoforms. Endocrinology 127:126-132, 1990 20. Bernal-Mizrachi E, Wen W, Stahlhut S, Welling CM, Permutt MA: Islet beta cell expression of constitutively active Akt1/PKB alpha induces striking hypertrophy, hyperplasia, and hyperinsulinemia. J Clin Invest 108:1631-1638, 2001 21. Girish V, Vijayalakshmi A: Affordable image analysis using NIH Image/ImageJ. Indian J Cancer 41:47, 2004 22. Nakae J, Kitamura T, Silver DL, Accili D: The forkhead transcription factor Foxo1 (Fkhr) confers insulin sensitivity onto glucose-6-phosphatase expression. J Clin Invest 108:1359-1367, 2001 23. Nakae J, Barr V, Accili D: Differential regulation of gene expression by insulin and IGF-1 receptors correlates with phosphorylation of a single amino acid residue in the forkhead transcription factor FKHR. Embo J 19:989-996, 2000 24. Dandoy-Dron F, Itier JM, Monthioux E, Bucchini D, Jami J: Tissue-specific expression of the rat insulin 1 gene in vivo requires both the enhancer and promoter regions. Differentiation 58:291-295, 1995 25. Dandoy-Dron F, Monthioux E, Jami J, Bucchini D: Regulatory regions of rat insulin I gene necessary for expression in transgenic mice. Nucleic Acids Res 19:4925-4930, 1991 26. Diehn M, Sherlock G, Binkley G, Jin H, Matese JC, Hernandez-Boussard T, Rees CA, Cherry JM, Botstein D, Brown PO, Alizadeh AA: SOURCE: a unified genomic resource of functional annotations, ontologies, and gene expression data. Nucleic Acids Res 31:219-223, 2003 27. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G: Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25:25-29, 2000 28. Sturn A, Quackenbush J, Trajanoski Z: Genesis: cluster analysis of microarray data. Bioinformatics 18:207-208, 2002

18 Foxo1 Inhibition Protects Islet ®-Cells

29. Mancuso DJ, Sims HF, Han X, Jenkins CM, Guan SP, Yang K, Moon SH, Pietka T, Abumrad NA, Schlesinger PH, Gross RW: Genetic Ablation of Calcium-independent Phospholipase A2{gamma} Leads to Alterations in Mitochondrial Lipid Metabolism and Function Resulting in a Deficient Mitochondrial Bioenergetic Phenotype. J Biol Chem 282:34611-34622, 2007 30. Karaskov E, Scott C, Zhang L, Teodoro T, Ravazzola M, Volchuk A: Chronic Palmitate but not Oleate Exposure Induces Endoplasmic Reticulum Stress Which May Contribute to INS-1 Pancreatic {beta}-cell Apoptosis. Endocrinology, 2006 31. Listenberger LL, Han X, Lewis SE, Cases S, Farese RV, Jr., Ory DS, Schaffer JE: Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proc Natl Acad Sci U S A 100:3077-3082, 2003 32. Corcoran MP, Lamon-Fava S, Fielding RA: Skeletal muscle lipid deposition and insulin resistance: effect of dietary fatty acids and exercise. Am J Clin Nutr 85:662-677, 2007 33. Johnson JD, Ahmed NT, Luciani DS, Han Z, Tran H, Fujita J, Misler S, Edlund H, Polonsky KS: Increased islet apoptosis in Pdx1+/- mice. J Clin Invest 111:1147-1160, 2003 34. Hagman DK, Hays LB, Parazzoli SD, Poitout V: Palmitate inhibits insulin gene expression by altering PDX-1 nuclear localization and reducing MafA expression in isolated rat islets of Langerhans. J Biol Chem 280:32413-32418, 2005 35. Keramaris E, Vanderluit JL, Bahadori M, Mousavi K, Davis RJ, Flavell R, Slack RS, Park DS: c-Jun N-terminal kinase 3 deficiency protects neurons from axotomy-induced death in vivo through mechanisms independent of c-Jun phosphorylation. J Biol Chem 280:1132-1141, 2005 36. Urano F, Bertolotti A, Ron D: IRE1 and efferent signaling from the endoplasmic reticulum. J Cell Sci 113 Pt 21:3697-3702, 2000 37. Holcik M, Sonenberg N: Translational control in stress and apoptosis. Nat Rev Mol Cell Biol 6:318-327, 2005 38. Ozcan U, Yilmaz E, Ozcan L, Furuhashi M, Vaillancourt E, Smith RO, Gorgun CZ, Hotamisligil GS: Chemical chaperones reduce ER stress and restore glucose homeostasis in a mouse model of type 2 diabetes. Science 313:1137-1140, 2006 39. Ooi GT, Tseng LY, Tran MQ, Rechler MM: Insulin rapidly decreases insulin-like growth factor-binding protein-1 gene transcription in streptozotocin-diabetic rats. Mol Endocrinol 6:2219-2228, 1992 40. Bogoyevitch MA, Boehm I, Oakley A, Ketterman AJ, Barr RK: Targeting the JNK MAPK cascade for inhibition: basic science and therapeutic potential. Biochim Biophys Acta 1697:89-101, 2004 41. Lee YH, Giraud J, Davis RJ, White MF: c-Jun N-terminal kinase (JNK) mediates feedback inhibition of the insulin signaling cascade. J Biol Chem 278:2896-2902, 2003 42. Altomonte J, Richter A, Harbaran S, Suriawinata J, Nakae J, Thung SN, Meseck M, Accili D, Dong H: Inhibition of Foxo1 function is associated with improved fasting glycemia in diabetic mice. Am J Physiol Endocrinol Metab 285:E718-728, 2003 43. Hribal ML, Nakae J, Kitamura T, Shutter JR, Accili D: Regulation of insulin-like growth factor-dependent myoblast differentiation by Foxo forkhead transcription factors. J Cell Biol 162:535-541, 2003

19 Foxo1 Inhibition Protects Islet ®-Cells

44. Lambrix P, Habbouche M, Perez M: Evaluation of ontology development tools for bioinformatics. Bioinformatics 19:1564-1571, 2003 45. Boyce M, Bryant KF, Jousse C, Long K, Harding HP, Scheuner D, Kaufman RJ, Ma D, Coen DM, Ron D, Yuan J: A selective inhibitor of eIF2alpha dephosphorylation protects cells from ER stress. Science 307:935-939, 2005 46. Busch AK, Gurisik E, Cordery DV, Sudlow M, Denyer GS, Laybutt DR, Hughes WE, Biden TJ: Increased fatty acid desaturation and enhanced expression of stearoyl coenzyme A desaturase protects pancreatic beta-cells from lipoapoptosis. Diabetes 54:2917-2924, 2005 47. Zhang W, Patil S, Chauhan B, Guo S, Powell DR, Le J, Klotsas A, Matika R, Xiao X, Franks R, Heidenreich KA, Sajan MP, Farese RV, Stolz DB, Tso P, Koo SH, Montminy M, Unterman TG: FoxO1 regulates multiple metabolic pathways in the liver: effects on gluconeogenic, glycolytic, and lipogenic gene expression. J Biol Chem 281:10105- 10117, 2006 48. Kelpe CL, Moore PC, Parazzoli SD, Wicksteed B, Rhodes CJ, Poitout V: Palmitate inhibition of insulin gene expression is mediated at the transcriptional level via ceramide synthesis. J Biol Chem 278:30015-30021, 2003 49. Maedler K, Oberholzer J, Bucher P, Spinas GA, Donath MY: Monounsaturated fatty acids prevent the deleterious effects of palmitate and high glucose on human pancreatic beta-cell turnover and function. Diabetes 52:726-733, 2003 50. Storling J, Binzer J, Andersson AK, Zullig RA, Tonnesen M, Lehmann R, Spinas GA, Sandler S, Billestrup N, Mandrup-Poulsen T: Nitric oxide contributes to cytokine- induced apoptosis in pancreatic beta cells via potentiation of JNK activity and inhibition of Akt. Diabetologia 48:2039-2050, 2005 51. Senn JJ: Toll-like receptor-2 is essential for the development of palmitate-induced insulin resistance in myotubes. J Biol Chem 281:26865-26875, 2006 52. Naimi M, Gautier N, Chaussade C, Valverde AM, Accili D, Van Obberghen E: Nuclear Foxo1 Controls And Integrates Key Signaling Pathways In Hepatocytes. Endocrinology, 2007 53. O'Brien R M, Streeper RS, Ayala JE, Stadelmaier BT, Hornbuckle LA: Insulin- regulated gene expression. Biochem Soc Trans 29:552-558, 2001 54. Greer EL, Brunet A: FOXO transcription factors at the interface between longevity and tumor suppression. Oncogene 24:7410-7425, 2005 55. Du K, Herzig S, Kulkarni RN, Montminy M: TRB3: a tribbles homolog that inhibits Akt/PKB activation by insulin in liver. Science 300:1574-1577, 2003 56. Ohoka N, Yoshii S, Hattori T, Onozaki K, Hayashi H: TRB3, a novel ER stress- inducible gene, is induced via ATF4-CHOP pathway and is involved in cell death. Embo J 24:1243-1255, 2005 57. Shen WH, Boyle DW, Wisniowski P, Bade A, Liechty EA: Insulin and IGF-I stimulate the formation of the eukaryotic initiation factor 4F complex and protein synthesis in C2C12 myotubes independent of availability of external amino acids. J Endocrinol 185:275-289, 2005 58. Rowland BD, Peeper DS: KLF4, p21 and context-dependent opposing forces in cancer. Nat Rev Cancer 6:11-23, 2006

20 Foxo1 Inhibition Protects Islet ®-Cells

59. Yamada T, Ishihara H, Tamura A, Takahashi R, Yamaguchi S, Takei D, Tokita A, Satake C, Tashiro F, Katagiri H, Aburatani H, Miyazaki J, Oka Y: WFS1-deficiency increases endoplasmic reticulum stress, impairs cell cycle progression and triggers the apoptotic pathway specifically in pancreatic beta-cells. Hum Mol Genet 15:1600-1609, 2006 60. Parton LE, Diraison F, Neill SE, Ghosh SK, Rubino MA, Bisi JE, Briscoe CP, Rutter GA: Impact of PPARgamma overexpression and activation on pancreatic islet gene expression profile analyzed with oligonucleotide microarrays. Am J Physiol Endocrinol Metab 287:E390-404, 2004 61. Kitamura T, Nakae J, Kitamura Y, Kido Y, Biggs WH, 3rd, Wright CV, White MF, Arden KC, Accili D: The forkhead transcription factor Foxo1 links insulin signaling to Pdx1 regulation of pancreatic beta cell growth. J Clin Invest 110:1839-1847, 2002 62. Drucker DJ, Nauck MA: The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 368:1696-1705, 2006 63. Zlotnik A: Chemokines and cancer. Int J Cancer 119:2026-2029, 2006 64. Millar JS, Stone SJ, Tietge UJ, Tow B, Billheimer JT, Wong JS, Hamilton RL, Farese RV, Jr., Rader DJ: Short-term overexpression of DGAT1 or DGAT2 increases hepatic triglyceride but not VLDL triglyceride or apoB production. J Lipid Res 47:2297- 2305, 2006 65. Kim JJ, Battaile KP: Burning fat: the structural basis of fatty acid beta-oxidation. Curr Opin Struct Biol 12:721-728, 2002 66. Zhang P, McGrath B, Li S, Frank A, Zambito F, Reinert J, Gannon M, Ma K, McNaughton K, Cavener DR: The PERK eukaryotic initiation factor 2 alpha kinase is required for the development of the skeletal system, postnatal growth, and the function and viability of the pancreas. Mol Cell Biol 22:3864-3874, 2002 67. Oyadomari S, Koizumi A, Takeda K, Gotoh T, Akira S, Araki E, Mori M: Targeted disruption of the Chop gene delays endoplasmic reticulum stress-mediated diabetes. J Clin Invest 109:525-532, 2002 68. Ott M, Gogvadze V, Orrenius S, Zhivotovsky B: Mitochondria, oxidative stress and cell death. Apoptosis 12:913-922, 2007 69. Riedl SJ, Salvesen GS: The apoptosome: signalling platform of cell death. Nat Rev Mol Cell Biol 8:405-413, 2007 70. Kajihara T, Jones M, Fusi L, Takano M, Feroze-Zaidi F, Pirianov G, Mehmet H, Ishihara O, Higham JM, Lam EW, Brosens JJ: Differential expression of FOXO1 and FOXO3a confers resistance to oxidative cell death upon endometrial decidualization. Mol Endocrinol 20:2444-2455, 2006 71. Milburn JL, Jr., Hirose H, Lee YH, Nagasawa Y, Ogawa A, Ohneda M, BeltrandelRio H, Newgard CB, Johnson JH, Unger RH: Pancreatic beta-cells in obesity. Evidence for induction of functional, morphologic, and metabolic abnormalities by increased long chain fatty acids. J Biol Chem 270:1295-1299, 1995 72. Bastie CC, Nahle Z, McLoughlin T, Esser K, Zhang W, Unterman T, Abumrad NA: FoxO1 stimulates fatty acid uptake and oxidation in muscle cells through CD36- dependent and -independent mechanisms. J Biol Chem 280:14222-14229, 2005 73. Nakae J, Kitamura T, Kitamura Y, Biggs WH, 3rd, Arden KC, Accili D: The forkhead transcription factor Foxo1 regulates adipocyte differentiation. Dev Cell 4:119-129, 2003

21 Foxo1 Inhibition Protects Islet ®-Cells

TABLE 1. Expression profiles of all genes regulated by FA in Control condition or with the Adv-DNF

22 Foxo1 Inhibition Protects Islet ®-Cells

For each gene significantly regulated by FA in the Control (Adv-GFP) condition, the fold change is indicated in the FA-Ctrl column and the corresponding regulation in the presence of the Adv-DNFoxo is indicated in the FA-DNFoxo column. A heat-map provides a visual representation of the regulation in both conditions. Genes are clustered according to their Gene Ontology (Cellular physiological processes) functions. Some of the genes belong to more than one category. A scale bar indicates the extent of the fold changes in the heat-map (in Log10 scale).

23 Foxo1 Inhibition Protects Islet ®-Cells

FIGURE LEGENDS

Figure 1. Fatty acids result in a dose-dependent increase in MIN6 cell death. Propidium iodide and DAPI staining of MIN6 cells grown on coverslips and treated with the indicated concentrations of palmitate and 0.5% BSA for 24 hours. B. Quantitation of the percentage of propidium iodide stained nuclei of MIN6 cells treated with the indicated concentrations of palmitate or oleate and 0.5% BSA for 24 hours. A single asterisk (*) and double asterisk (**) indicate a p-value of significance < 0.05 and < 0.01, respectively, for the FA exposure compared to the treatment with 0.5% BSA alone.

Figure 2. Intracellular signaling with varying concentrations of palmitate and oleate. MIN6 cells were grown at the indicated concentrations of 0.5% BSA alone or with additional palmitate or oleate, and upon 24 hours of treatment, whole cell lysates were evaluated for protein expression. A. Western blot of increasing phosphorylation of JNK and its substrate c-Jun with increasing concentrations of FAs. Total JNK was used as a loading control. B. Western blot of Akt phosphorylation and total Akt as a loading control. C. Western blot of decreasing Pdx-1 expression with increasing concentrations of palmitate or oleate, correlating with increased CHOP and cleaved Caspase 3 (CC3) expression. Alpha-Tubulin was used as a loading control.

Figure 3. Time course of JNK and ER-stress activation with palmitate and oleate treatment of MIN6 cells. Cells were treated with palmitate (0.4 mM) or oleate (0.4 mM) for the indicated time points, and harvested for whole cell lysates to be used in western blots. A. A time course of JNK activation with FA treatment, indicated by western blotting for JNK phosphorylation and its target, c-Jun. Total JNK was probed as well. Thapsigargin treatment for 8 hours (300 nM) was used as a positive control for JNK activation. B. Bip protein, phosphorylation of EIF2-alpha, CHOP protein, Pdx1, and cleaved Caspase 3 (CC3) were assessed by western blot during the time course of treatment indicated. Alpha-Tubulin was probed as a loading control. C. the effects of the addition of the ER stress inhibitor TUDCA on expression of cleaved Caspase 3, phospho eIF2a, and CHOP 24 hours after the addition of palmitate.

Figure 4. Fatty acid treatment increased Foxo1 nuclear localization and transcriptional activity at a time when Akt phosphorylation was increased. A. Western blot for total Foxo1 in cytoplasmic and nuclear protein fractions of MIN6 cells incubated for four hours in 0.5% BSA alone or BSA with 0.4 mM palmitate or oleate. B. MIN6 cells co-transfected with the IGFBP1 promoter-luciferase and Renilla vectors were exposed to 0.4 mM palmitate/0.5% BSA for the indicated times. The double asterisks (**) indicate significance with a p-value < 0.01 compared to incubation in 0.5% BSA treatment alone. C. Intracellular localization of GFP-tagged Foxo1 expressed in MIN6 cells is represented in a graph of the percentage of nuclear localization with treatment of 1 μM thapsigargin at the indicated hours of exposure.

Figure 5. JNK inhibition decreased nuclear Foxo1 localization and apoptosis with palmitate treatment. A. Western blot of cleaved Caspase 3 and phospho c-Jun from MIN6 cells incubated in BSA, BSA and palmitate, and the indicated concentrations of

24 Foxo1 Inhibition Protects Islet ®-Cells

SP600125 for 24 hours. B. Western blot of total Foxo1 of nuclear and cytoplasmic protein extractions from MIN6 cells incubated for 4 and 12 hours in either 0.5% BSA alone or BSA and 0.4 mM palmitate with or without the addition of 300 nM SP610025, a concentration which provided JNK inhibition. C. Intracellular quantitation of GFP-tagged Foxo1 in MIN6 cells incubated for 4 hours with BSA, palmitate, or palmitate and 300 nM of SP600125 D. Cellular localization of GFP-tagged Foxo1 and DAPI staining by immunofluorescence. E. Akt phosphorylation was assessed by Western blot during the time course of treatment with palmitate (0.4 mM) or oleate (0.4 mM).

Figure 6. Dominant-Negative Foxo1 expression decreased thapsigargin and palmitate-induced cell death. A. Propidium iodide and DAPI staining of MIN6 cells infected with either GFP or DNFoxo adenovirus (Adv-GFP and Adv-DNF respectively) and exposed for 18 hours in either normal culture media or media plus the addition of 100 nM thapsigargin. B. Cell death from thapsigargin as quantitated by the percentage of nuclei staining for propidium iodide and DAPI in three independent experiments. C. Propidium iodide and DAPI staining of cells infected with either GFP or DNFoxo adenovirus and incubation for 24 hours in either 0.5% BSA alone or BSA and 0.4 mM palmitate. D. Quantitation of cell death by 0.4 mM palmitate in GFP or DNFoxo virally infected MIN6 cells from three independent experiments. Significance is indicated within the figures of quantitation. E. Western blot of cleaved Caspase 3, CHOP, and HA expression of MIN6 cells infected with the GFP adenovirus and increasing titers of DNFoxo adenovirus and incubated in 0.4 mM palmitate and 0.5% BSA for 24 hours.

Figure 7. Ex vivo treatment of wild type and RIP-DNFoxo islets with thapsigargin and palmitate. A. Western blot for cleaved Caspase 3 expression and α-Tubulin expression as a loading control of pooled isolated islets from each genotype which were divided into two culture conditions with and without 10 µM thapsigargin for 48 hours. B. Immunoblotting for JNK activation, Caspase activation, CHOP, phosphorylated-Akt, and Pdx-1 in pooled islets from wild type and RIP-DNFoxo mice cultured for 24 hours in RPMI containing 0.5% BSA with or without 0.4 mM palmitate

Figure 8. Foxo1 activity regulates CHOP promoter activity. CHOP promoter activity at 24 hours of thapsigargin (100 nM) or palmitate (0.4 mM with 0.5% BSA) treatment with co-transfection of empty pCMV5 plasmid or pCMV5-DNFoxo (as indicated), measured by a luciferase expression and normalized to TK renilla expression.

25 Foxo1 Inhibition Protects Islet ®-Cells

26 Foxo1 Inhibition Protects Islet ®-Cells

27 Foxo1 Inhibition Protects Islet ®-Cells

28 Foxo1 Inhibition Protects Islet ®-Cells

29 Foxo1 Inhibition Protects Islet ®-Cells

30 Foxo1 Inhibition Protects Islet ®-Cells

31 Foxo1 Inhibition Protects Islet ®-Cells

32 Foxo1 Inhibition Protects Islet ®-Cells