Article pubs.acs.org/jpr
Role of the Saturated Nonesterified Fatty Acid Palmitate in Beta Cell Dysfunction Michael Maris,† Sofie Robert,† Etienne Waelkens,‡,§ Rita Derua,‡ Miriam H. Hernangomez,∥ Wannes D’Hertog,† Miriam Cnop,∥,⊥ Chantal Mathieu,†,§,# and Lut Overbergh*,†,§,# †
Laboratory of Clinical and Experimental Endocrinology, Herestraat 49, KU Leuven, Leuven, Belgium Laboratory of Protein Phosphorylation and Proteomics, KU Leuven, Leuven, Belgium § SyBioMa, Herestraat 49, KU Leuven, Leuven, Belgium ∥ Laboratory of Experimental Medicine, Université Libre de Bruxelles (ULB), Route de Lennik, 808, 1070 Brussels, Belgium ⊥ Division of Endocrinology, Erasmus Hospital, Université Libre de Bruxelles (ULB), Route de Lennik, 1070 Brussels, Belgium ‡
S Supporting Information *
ABSTRACT: Sustained elevated levels of saturated free fatty acids, such as palmitate, contribute to beta cell dysfunction, a phenomenon aggravated by high glucose levels. The aim of this study was to investigate the mechanisms of palmitate-induced beta cell dysfunction and death, combined or not with high glucose. Protein profiling of INS-1E cells, exposed to 0.5 mmol/L palmitate and combined or not with 25 mmol/L glucose, for 24 h was done by 2D-DIGE, both on full cell lysate and on an enriched endoplasmic reticulum (ER) fraction. Eighty-three differentially expressed proteins (P < 0.05) were identified by MALDI-TOF/TOF mass spectrometry and proteomic results were confirmed by functional assays. 2D-DIGE analysis of whole cell lysates and ER enriched samples revealed a high number of proteins compared to previous reports. Palmitate induced beta cell dysfunction and death via ER stress, hampered insulin maturation, generation of harmful metabolites during triglycerides synthesis and altered intracellular trafficking. In combination with high glucose, palmitate induced increased shunting of excess glucose, increased mitochondrial reactive oxygen species production and an elevation in many transcription-related proteins. This study contributes to a better understanding and revealed novel mechanisms of palmitate-induced beta cell dysfunction and death and may provide new targets for drug discovery. KEYWORDS: lipotoxicity, glucolipotoxicity, beta cells, palmitate, type 2 diabetes, proteomics, endoplasmic reticulum, free fatty acids, 2D-DIGE
■
dysfunction and death.7 Palmitate is a strong activator of ER stress, inducing an up-regulation of chaperone levels, an inhibition of translation and an induction of activating transcription factor 3 (ATF3) and C/EBP homologous protein (CHOP), thereby triggering activation of pro-apoptotic signals.8 Another mechanism implicated in lipotoxicity is ceramide formation, a lipid molecule synthesized out of palmitate. Ceramide impairs pancreatic and duodenal homeobox-1 (PDX-1) subcellular localization and MafA expression with reduced insulin gene expression as a consequence9 and contributes to palmitate-induced beta cell death by mechanisms that have not yet been unraveled. Different proteomic approaches have been applied to investigate the effect of palmitate in beta cells. Using a SELDI-TOF MS and 2D-PAGE approach on palmitateexposed INS-1E cells, Sol et al. revealed the up-regulation of CPT-I and down-regulation of calmodulin.10 A recent 2DPAGE-based study identified 9 palmitate-regulated proteins,
INTRODUCTION Insulin resistance and dysfunction of pancreatic beta cells leads to hyperglycemia and the onset of type 2 diabetes (T2D).1 The causes for beta cell failure are a genetic predisposition and environmental factors, such as inflammation and metabolic stressors such as glucose and nonesterified fatty acids (NEFA).2,3 Saturated NEFA are detrimental for proper beta cell function and survival and this harmful outcome is aggravated in the presence of high glucose, an effect called glucolipotoxicity.4 A number of in vitro studies on isolated islets and beta cell lines showed hampered insulin gene expression, impeded glucose-stimulated insulin secretion (GSIS) and increased apoptosis after exposure to NEFA, effects that were more pronounced in the presence of high glucose.5 These observations were corroborated in vivo: rats receiving an infusion of NEFA and glucose showed lowered insulin gene expression and insulin content, as well as impeded GSIS.6 As the most common saturated NEFA in human plasma is palmitate, this NEFA has often been used in mechanistic studies. The induction of endoplasmic reticulum (ER) stress is an important mechanism in NEFA-induced beta cell © 2012 American Chemical Society
Received: July 2, 2012 Published: November 21, 2012 347
dx.doi.org/10.1021/pr300596g | J. Proteome Res. 2013, 12, 347−362
Journal of Proteome Research
Article
Two-dimensional Difference Gel Electrophoresis (2D-DIGE)
including glycolysis-, ER stress- and proteasome-related proteins.11 Jeffrey et al. applied a 2D-DIGE approach on palmitate-exposed human islets and identified carboxypeptidase E (CPE) as a key protein linking hyperlipidemia, insulin processing and beta cell apoptosis.12 Overall, these proteomic studies of palmitate-exposed beta cells detected a low number of differentially expressed proteins. A possible explanation could be that palmitate mainly affects low abundant proteins that are difficult to detect by gel-based approaches. To overcome this problem, subcellular prefractionation can reduce the complexity of the sample and/or enrich certain proteins. Fountoulakis et al. applied preparative electrophoresis to enrich low-abundance rat liver cytosolic proteins prior to proteomic analysis.13 Thus, subcellular fractionation is a powerful tool to enrich samples in order to study low abundant and organelle specific proteins and to elucidate important molecular events in more detail. The aim of this study was to analyze palmitate-induced protein changes and to gain more insights into the harmful effects of palmitate and the aggravating role of high glucose in beta cell dysfunction and death. To our knowledge, no proteomic studies focusing on glucolipotoxicity are performed yet, while this phenomenon is a major cause of beta cell dysfunction in the setting of type 2 diabetes. In addition, the limited studies available on palmitate-induced beta cell changes could only detect a very limited amount of proteins as being differentially expressed. To overcome this, we used the sensitive 2D-DIGE technique on full protein lysates of INS-1E cells and on an ER enriched subcellular fraction. This fraction was chosen based on the knowledge that palmitate acts, at least in part, through activation of the ER stress response.7,8,12
INS-1E cells (5 × 106) were exposed to a control condition of 11 mmol/L glucose (standard INS-1E culturing conditions), the palmitate condition consisting of 0.5 mmol/L palmitate and 11 mmol/L glucose and the combination of palmitate and glucose consisting of 0.5 mmol/L palmitate and 25 mmol/L glucose. For each condition, quadruplicate experiments were performed, originating from four independent experiments. After the exposure time, medium was removed and cells were washed twice with warm PBS to remove potential necrotic cells. Cells were detached using accutase, washed twice in PBS and resuspended in 100 μL of lysis buffer (7 mol/L urea, 2 mol/L thiourea, 4% (w/v) CHAPS, 40 mmol/L Tris base, 1% (w/v) DTT and a mixture of protease inhibitors (Complete protease inhibitor, Roche Diagnostics, Vilvoorde, Belgium)). Samples were sonicated and the supernatant was desalted by dialysis (PlusOne Mini Dialysis kit, GE Healthcare, Diegem, Belgium). Samples for 2D-DIGE on ER enriched fractions were prepared as described above. For both full and ER enriched lysate, 50 μg of protein was labeled with Cy3 or Cy5, and a pooled internal standard was labeled with Cy2. Dye swapping made sure that each condition was equally labeled with Cy3 or Cy5, eliminating potential bias from the labeling reaction. For the first dimension, pooled samples, containing sample loading buffer, were loaded onto the rehydrated IPG strips (24 cm) (GE Healthcare) using anodic cup loading and separated according to their isoelectric point on an Ettan IPGphorII manifold (GE Healthcare). The first dimension was ended when the current reached a stable phase (at 60 kV.h) (see also Supporting Information). Prior to the second dimension, the strips were equilibrated during two intervals of 15 min each in an equilibration buffer (7 mol/L urea, 2 mol/L thiourea, 4% (w/v) CHAPS, 0.5% (v/v) IPG buffer, 0.05% (w/v) OrangeG), and either 1% (w/v) DTT (for IPG strips pH 4−7) or 1.2% (v/v) Destreak (for IPG strips pH 6−9) and the equilibrated strips were placed on top of 12.5% SDS-polyacrylamide gel (24 × 20 cm) and separated on an Ettan DaltSix system (GE Healthcare) at 8 mA/gel, 9 W and 600 V for 1 h and then at 16 mA/gel, 15 W and 600 V until the bromophenol blue dye front reached the bottom of the gel. Scanning of the gels was performed using a Typhoon9400 (GE Healthcare). Initial scan parameters were press sample, depth plus 3 mm, 500−550 V photomultiplier tube setting and 1000-μm pixel size, before repeating at 100-μm pixel size for a high-resolution scan. Spot detection and matching was performed automatically using the BVA module of DeCyder Version 7.0 software followed by careful manual rematching of wrongly matched spots or unmatched spots. Spot detection parameters were set as described by the manufacturer: spot detection algorithm 6.0, estimated number of spots: 10000 and spot volumes below 30000 were excluded. The 12 spot maps corresponding to the 4 gels from each pH range were used to calculate average abundance. Protein spot expression levels which showed a statistically significant (P < 0.05) increase or decrease in 6 out of 8 (Cy3 or Cy5 labeled) gels were accepted as being differentially expressed.
■
METHODS INS-1E cell culture, cell death analysis, Western blotting and quantitative RT-PCR are available as Supporting Information. Subcellular ER Fractionation
The protocol for subcellular ER fractionation was based on Stone et al.14 and optimized for INS-1E cells (Supporting Information). INS-1E cells (225 × 106) were exposed to palmitate, combined or not with high glucose, for 24 h. After the exposure time, medium was removed and cells were washed twice with warm PBS to remove potential necrotic cells. Cells were detached using accutase, washed twice in PBS, pelleted by centrifugation (1500 rpm for 5 min at 25 °C) and resuspended in isolation buffer (210 mmol/L mannitol, 1 mmol/L EGTA, 70 mmol/L sucrose and 10 mmol/L MOPS (pH 7.2)). The cells were gently disrupted and homogenized using a Douncer with tight pestle B. Proper cell disruption was verified by microscopy. The homogenate was centrifuged at 625 g for 5 min (4 °C) to remove cellular debris and nuclei, and the supernatant was centrifuged at 12000× g for 10 min at 4 °C. Thereafter, the supernatant was centrifuged at 21000× g for 10 min (4 °C) to remove mitochondria and finally centrifuged at 100000× g for 1 h (4 °C) in a Beckman Coulter OptimaTM MAX to pellet the ER. The ER pellet was resuspended in lysis buffer (7 mol/L urea, 2 mol/L thiourea, 4% (w/v) CHAPS, 40 mmol/L tris base, 1% (w/v) DTT and protease inhibitors (Complete protease inhibitor, Roche Diagnostics). ER pellets were purified using the 2D Clean-Up Kit (GE Healthcare), the pH was adjusted to 8.5 and protein content quantified by the Bradford method.
Spot Digestion and Protein Identification by MALDI-TOF/TOF Analysis
For spot picking, two preparative gels for each pH range were run (350 μg protein lysate each, both for full cell lysates and for ER enriched fractions). Similar electrophoretic parameters for first and second dimension runs were used as for analytical gels, 348
dx.doi.org/10.1021/pr300596g | J. Proteome Res. 2013, 12, 347−362
Journal of Proteome Research
Article
except that Cy dye-labeling was omitted. Glass plates (27 × 21 cm) were pretreated with BindSilane and two reference markers were applied to enable automatic picking. The gels (24 × 20 cm) were poststained using the fluorescent Krypton dye (Pierce). Matching with the analytical gels was performed using the BVA module of the DeCyder V7.0 software. A pick list was generated and exported into the Spot Picker V1.20 software which controls the Ettan Spot Picker (GE Healthcare). Spot digestion and peptide purification was performed as previously reported.15 MS/MS analyses were performed on a 4800 MALDI TOF/TOF (Applied Biosystems, CA, USA). The instrument was calibrated with the Applied Biosystems Calibration Mixture 1. The mass accuracy (external calibration) was
