Impact of AMPK on Glucosylceramide Metabolism

The membrane glycolipid glucosylceramide (GlcCer) plays a critical role in cellular homeostasis. Its intracellular levels are thought to be tightly regulated. How cells regulate GlcCer levels remains to be clarified. AMP-activated protein kinase (AMPK), which is a crucial cellular energy sensor, regulates glucose and lipid metabolism to maintain energy homeostasis. Here, we investigated whether AMPK affects GlcCer metabolism. AMPK activators (AICAR and metformin) decreased intracellular GlcCer levels and synthase activity in mouse fibroblasts. AMPK inhibitors or AMPK siRNA reversed these effects, suggesting that GlcCer synthesis is negatively regulated by an AMPK-dependent mechanism. Although AMPK did not affect the phosphorylation or expression of GlcCer synthase protein, the amount of UDP-glucose, an activated form of glucose required for GlcCer synthesis, decreased under AMPK-activating conditions. Importantly, the UDP-glucose pyrophosphatase Nudt14, which degrades UDP-glucose generating UMP and glucose-1-phosphate, was phosphorylated and activated by AMPK. On the other hand, http://www.jbc.org/cgi/doi/10.1074/jbc.M115.658948 The latest version is at JBC Papers in Press. Published on June 5, 2015 as Manuscript M115.658948 Copyright 2015 by The American Society for Biochemistry and Molecular Biology, Inc. by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 2 suppression of Nudt14 by siRNA had little effect on UDP-glucose levels, indicating that mammalian cells have an alternative UDP-glucose pyrophosphatase that mainly contributes to the reduction of UDP-glucose under AMPK-activating conditions. Since AMPK activators are capable of reducing GlcCer levels in cells from Gaucher disease patients, our findings suggest that reducing GlcCer through AMPK activation may lead to a new strategy for treating diseases caused by abnormal accumulation of GlcCer. Glycosphingolipids (GSLs) are amphipathic compounds consisting of oligosaccharides and ceramide moieties. They are ubiquitous in the outer leaflet of the plasma membrane and believed to be involved in a large number of cellular processes, including signal transduction, membrane trafficking, cytoskeletal organization, and pathogen entry (1,2). Since most mammalian GSLs are generated from glucosylceramide (GlcCer), GlcCer synthesis is an important step for determining cellular GSL levels. GlcCer is synthesized by GlcCer synthase (UDP-glucose:ceremide glucosyltransferase; UGCG, EC 2.4.1.80) from ceramide and uridine diphosphate-glucose (UDP-Glc) at the cytosolic surface of the Golgi apparatus (3). GlcCer is a fundamental GSL found in organisms ranging from mammals to fungi. In vivo studies demonstrate that UGCG plays critical roles in development, differentiation, and energy homeostasis (4-7). For instance, ugcg knockout mice die in utero, because gastrulation is blocked by ectodermal apoptosis (4,8). The forebrain-neuron-specific deletion of UGCG in mice results in obesity, hypothermia, and lower sympathetic activity (9). This study demonstrated that UGCG expression in neurons of the adult central nervous system (CNS) regulates the leptin signaling pathway and controls energy homeostasis at the whole-animal level. On the other hand, abnormal accumulation of GlcCer and GSLs is known to be closely related to several diseases or disorders such as Gaucher disease, Parkinson disease, insulin resistance, breast cancer, and polycystic kidney disease (PKD) (10-16). In ob/ob mice, an animal model of type II diabetes, GlcCer levels are increased in several tissues, such as liver and muscle. Interestingly, insulin sensitivity, glucose homeostasis, and adipocyte function are improved by treating the mice with an UGCG inhibitor (17,18). The expression of ugcg mRNA is significantly increased in tumors of the breast, small intestine, cervix, and rectum compared to that in normal human tissues (11). Ugcg overexpression is associated with drug resistance in several cancer cells and the maintenance of pluripotency in breast cancer stem cells (19,20). Suppression of UGCG expression sensitizes cancer cells to anticancer agents (21). Previous studies indicated that GlcCer levels should be strictly regulated to appropriate levels to maintain biological activities. How cells regulate the activity of UGCG is yet to be clarified. Elucidating the mechanisms underlying the regulation of UGCG may lead to new and more effective treatment options for diseases caused by the abnormal accumulation of GlcCer/GSLs. The hydrolysis of ATP drives all energy-requiring processes in living cells. To maintain ATP at a sufficient level, eukaryotic by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 3 cells have an important nutrient and energy sensor, AMP-activated protein kinase (AMPK) (22). AMPK is a heterotrimeric serine/threonine kinase that enhances signaling in ATP-generating pathways, such as glycolysis or fatty acid oxidation, while inhibiting anabolic processes, such as biosynthesis of fatty acids, cholesterol, glycogen, and triacylglycerol under energy reducing conditions (i.e., increasing AMP/ATP or ADP/ATP ratios). Since AMPK is closely involved in the metabolism of glucose and fatty acids, which are components of GlcCer, we expected that AMPK could also control GlcCer metabolism in mammalian cells. To test this possibility, we assessed the intracellular UGCG activity and cellular GlcCer levels under AMPK-activating conditions. In the present study, we found that AMPK affects the GlcCer biosynthesis pathway. Intracellular GlcCer levels and UGCG activity were reduced by AMPK-activating drugs, such as 5-aminoimidazole-4-carboxamide 1--D-ribofuranoside (AICAR) and the anti-diabetic drug metformin. On the other hand, an AMPK inhibitor and AMPK siRNA overrode the reduced GlcCer synthase activity or cellular GlcCer levels under AMPK-activating conditions, indicating that AMPK is a negative regulator of GlcCer synthesis. The expression or phosphorylation levels of UGCG were unchanged under AMPK-activating and AMPK-inhibiting conditions. Instead, we found that cellular sugar nucleotides including UDP-Glc, a precursor of GlcCer synthesis, were decreased by AMPK-activating compounds. In addition, we found that UDP-Glc degrading enzyme, UDP-Glc pyrophosphatase Nudt14, is phosphorylated and activated by AMPK, which is partly, but significantly, involved in the reduction of UDP-Glc. The present study provides mechanistic insights into the regulation of GlcCer synthesis by AMPK, by which UDP-Glc levels were decreased through the activation of UDP-Glc pyrophosphatase. Importantly, AMPK activators were capable of reducing cellular GlcCer levels of cells derived from patients with Gaucher disease. Our findings suggest that reduction of GlcCer via AMPK activation may serve as the basis for new treatment options for diseases caused by the accumulation of GlcCer. EXPERIMENTAL PROCEDURES Materials — Ammonium formate, AICAR, 6-[4-(2-Piperidin-1-ylethoxy)phenyl]-3-pyridin4-ylpyrazolo[1,5-a]pyrimidine (Compound C), and anti-α-tubulin antibody were purchased from Sigma Aldrich (Germany). Pre-coated Silica gel 60 TLC plates were purchased from Merck (Germany). Metformin HCl was purchased from LKT Laboratories (USA). Tetrabutylammonium hydrogen sulfate was purchased from Tokyo Chemical Industry Co., LTD. (Japan). Formic acid, methanol, chloroform, and acetonitrile were purchased from Nacalai Tesque (Japan). Pro-Q Diamond phosphoprotein gel stain, SYPRO Ruby protein gel stain, 7-nitro-2,1,3-benzoxadiazole (NBD) C6-ceramide, NBD C6-ceramide conjugated to bovine serum albumin (BSA), anti--actin mouse antibody, Alexa Fluor 568 anti-mouse IgG, and Alexa Fluor 488 anti-rabbit IgG were purchased from Life Technologies (USA). NBD C6-GlcCer was purchased from Matreya (USA). Antibodies against DYKDDDDK (FLAG) Tag, AMPK1/2, phospho-AMPK1/2 (Thr172), by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 4 ACC, phospho-ACC (Ser79), phospho-(Ser/Thr) AMPK substrate, and LKB1 were purchased from Cell Signaling Technology (USA). Nudt14 antibody and donkey anti-goat IgG HRP were obtained from Santa Cruz Biotechnology (USA). GM130 antibody was purchased from BD Biosciences (USA). Construction of expression vectors — PCR was carried out by Phusion polymerase (Thermo Scientific) using mouse kidney cDNA (GenoStaff, Japan) as a template and primers listed in Table 1 for amplification of Nudt14 and UGP2. Phosphorylation-site-deleted mutants, Nudt14 T141A and UGP2 S448A, were generated using mutation primers listed in Table 1 (Nudt14 T141A-S, Nudt14 T141A-A, UGP2 S448A-S, and S448A-A). The amplified products were inserted into HindIII-digested p3xFLAG-CMV-10 (Sigma Aldrich) to generate N-terminal 3xFLAG-tagged proteins by using an In-Fusion HD Cloning kit (Clontech). The expression vectors of constitutively active and dominant negative forms of AMPK 1 were prepared as described in (23). C-terminal FLAG-tagged UGCG expression vector was a gift from Dr. Shun Watanabe, RIKEN, Japan. Cell cultures and gene transfection — NIH3T3 (3T3), human embryonic kidney (HEK)293, HeLa, Madin-Darby canine kidney (MDCK), mouse embryonic fibroblast (MEF), and Chinese hamster ovary (CHO) cells were cultured in Dulbecco’s modified Eagle Medium (DMEM), containing 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 g/ml streptomycin at 37C in 5% CO2. The cell lines ugcg (lox/lox) MEF and ugcg (-/-) MEF, which were established from ugcg (lox/lox) MEF by expressing Cre recombinase, were kindly donated by Dr. Shun Watanabe, RIKEN (24). Type 1, 2, and 3 fibroblasts derived from Gaucher disease patients (GM00372, GM00877, and GM00852, respectively) were purchased from Coriell Institute (USA). These fibroblast cells were cultured in MEM with Earle’s salts and non-essential amino acids, containing 15% FBS, 100 U/ml penicillin, and 100 g/ml streptomycin at 37C in 5% CO2. Cells were transfected with expression vectors by using TurboFect Transfection Reagent (Thermo Scientific) according to the manufacturer’s instructions. Immunoblot analysis — Cells were lysed in ice-cold lysis buffer (20 mM Tris HCl buffer [pH 7.5], containing 150 mM NaCl, 1 mM EDTA, 1% NP-40, and phosphatase and protease inhibitor cocktails [Roche diagnostics]) by sonication for 45 s. Cell debris was removed by centrifugation (18,000 x g for 10 min at 4C). The amount of protein in the supernatant was determined by bicinchoninic acid protein assay (Pierce), with BSA as a standard. For immunoprecipitation (IP) of FLAG-tagged protein, the protein extract prepared in lysis buffer was subjected to IP with ANTI-FLAG M2 Agarose Affinity Gel (Sigma-Aldrich). Then the gel was washed with PBS three times, and the FLAG-tagged protein was eluted with lysis buffer containing 100 g/ml of 3 x FLAG peptide (Sigma-Aldrich). Equal amounts of protein were separated by SDS-PAGE using Mini-PROTEAN TGX Gel (BioRad) and then blotted onto PVDF membranes by Trans Blot Turbo (BioRad). The membranes were blocked by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 5 in Blocking One-P (Nacalai Tesque) and incubated with Signal Enhancer HIKARI (Nacali tesque) containing antibodies against AMPK1/2 (1:2000); phospho-AMPK1/2 (Thr172) (1:2000); ACC (1:2000); phospho-ACC (Ser79) (1:2000); LKB1 (1:3000); DYKDDDDK tag (1:6000); phospho-(Ser/Thr) AMPK substrate (1:1000); -tubulin (1:20000); Nudt14 (1:250); or -actin (1:20000). The blots were washed with TBS-T, and then incubated with HRP-conjugated anti-mouse or anti-rabbit IgG antibody (1:10000) (Cell Signaling Technology) or anti-goat IgG antibody. The blots were washed again with TBS-T. The protein was detected by chemiluminescence using Luminate Forte Western HRP substrate (Millipore) and then analyzed with a LAS-3000 Luminescence Image Analyzer (Fujifilm). To avoid dephosphorylation of the AMPK  subunit and ACC, cells treated with or without AMPK-activating compounds were immediately frozen with liquid nitrogen after incubation. Immunostaining — To determine the subcellular localization of Nudt14 or Nudt14 T141, 3T3 cells were transfected with plasmids encoding FLAG-Nudt14 or FLAG-Nudt14 T141A. After transfection, cells were treated with AICAR (1 mM) or metformin (8 mM), fixed in 3% paraformaldehyde for 15 min, and then washed with PBS. The fixed cells were blocked in Blocking One Histo (Nacalai Tesque) and incubated with antibodies against FLAG tag (1:500) and GM130 (1:500), a Golgi marker, for 12 h at 4C. After incubation, cells were washed with PBS and incubated with Alexa Fluor 568 anti-mouse IgG (1:500) and Alexa Fluor 488 anti-rabbit IgG (1:500) for 1 h. Nuclei were counterstained with 4',6-diamidino-2-phenylindole, dihydrochloride (DAPI) for 5 min. Cells were mounted in Immu-Mount reagent (Thermo Scientific) and observed under a confocal laser-scanning microscope (FV1000, Olympus). Measurement of intracellular UGCG and glucosylceramidase activity — The assays were performed as described previously with modification (25). Cells grown in 24-well plates containing DMEM/10% FBS were exposed to AICAR, metformin, CC, or AMPK 1/2 siRNA for an appropriate period. The culture medium was switched to 250 l of DMEM/10% FBS containing 0.5 M NBD C6-ceramide conjugated to BSA (Invitrogen). After 90 min of incubation at 37C, cells were rinsed with ice-cold D-PBS (Nacalai Tesque), scraped from the wells using 50 l of ice-cold water, and transferred to 1.5 ml tubes. Lipids were extracted by the addition of 190 l of chloroform/methanol (1/2, v/v). After incubation at room temperature for 10 min, 62.5 l of water and chloroform were added and then the tubes were centrifuged at 12,000 x g for 5 min. The lower phase was collected and dried with a speed vac concentrator. The dried sample was dissolved in 20 l of chloroform/methanol (1/2, v/v), and 3 l of the sample was applied onto a TLC plate, which was developed with chloroform/methanol/water (65/25/4, v/v/v). The reaction products were visualized with a LAS-3000 equipped with blue-light-emitting diode (460 nm EPI) and a Y515-Di filter, and then quantified with Image J 1.43u software (NIH). The extent of synthesis of NBD-labeled by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 6 GlcCer was calculated as follows: synthesis (%) = (peak area for NBD C6-GlcCer) x 100 / (peak area for NBD C6-ceramide + peak area for NBD C6-GlcCer + peak area for NBD C6-sphingomyelin). Under this experimental condition, NBD fatty acid, which is generated from ceramide by ceramidase (26), was not detected. To measure intracellular glucosylceramidase (GCase) activity, we used BSA-conjugated NBD C6-GlcCer, which was prepared as described in (27), instead of NBD C6-ceramide. The extent of degradation of NBD-labeled GlcCer was calculated as follows: GCase activity (%) = (peak area for NBD C6-ceramide) x 100 / (peak area for NBD C6-ceramide+ peak area for NBD C6-GlcCer). Lipid extraction and quantification of sphingolipids by LC-ESI MS/MS — Cells were washed once with cold PBS, collected in 100 l of cold PBS, and sonicated with a Handy Sonic Disruptor (Tomy Seiko Co., Japan) for 10 s. Part of the sample (5 l) was used in the bicinchoninic acid protein assay to determine the amount of protein. Total lipids were extracted by adding 375 l of chloroform/methanol (1/2, v/v) containing 62.5 pmol of each component of Ceramide/Sphingoid Internal Standard Mixture II (Avanti Polar Lipids, Inc.) from the rest of sample. The single-phase mixture was incubated at 48°C overnight. After cooling, 125 l of water and chloroform were added. The lower phase was collected and dried with a speed vac concentrator. The dried sample was suspended in chloroform/methanol (2/1, v/v) containing 0.1 N KOH, incubated for 2 h at 37C and neutralized with an equal amount of acetic acid, and then water was added. The lower phase was withdrawn and dried, then resuspended in 200 l of methanol, sonicated for 10 s, and centrifuged at 14,000 x g for 5 min. The supernatant was transferred to autoinjector vials. The amount of GlcCer, LacCer, Cer, and SM were analyzed by LC-ESI-MS using a triple quadruple mass spectrometer 4000Q TRAP (AB SCIEX) coupled to an Agilent 1100 series HPLC system (Agilent). A binary solvent gradient with a flow rate of 0.2 ml/min was used to separate sphingolipids by normal-phase chromatography using an InertSustain NH2 (2.1 x 150 mm, 5 m bead size, GL Science, Japan). The gradient was started at 20% buffer B (methanol/water/formic acid, 89/9/1, v/v/v, with 20 mM ammonium formate) in buffer A (acetonitrile/methanol/formic acid, 97/2/1, v/v/v, with 5 mM ammonium formate). The gradient reached 100% B in 4 min and was maintained at 100% B for 2 min. Finally, the gradient was returned to the starting conditions, and the column was equilibrated for 5 min before the next run. Sphingolipids containing C16:0, C18:0, C20:0, C22:0, C24:0, and C24:1 fatty acids were detected using a multiple reaction monitoring (MRM) method, as described in (28). RNA interference — Stealth RNAi duplexes (Life technologies) were designed to mouse AMPK 1 (5’CCAGGUCAUCAGUACACCAUCUGAU-3’) and 2 (5’AAGUGAAGAUUGGAGAACACCAAUU) using the BLOCK-iT RNAi Designer. Predesigned Stealth RNAi siRNA (MSS287847, Life technologies) was used to knockdown Nudt14. Stealth RNAi siRNA Negative Control Med GC or Low GC were used as control by gest on Sptem er 7, 2016 hp://w w w .jb.org/ D ow nladed from Impact of AMPK on Glucosylceramide Metabolism 7 siRNAs for knockdown of AMPK  subunits and Nudt14. The reverse transfection method was used to transfect Stealth RNAi into 3T3 cells. Stealth RNAi duplexes (30 pmol) were mixed with 5 l of Lipofectamine RNAiMAX Reagent (Life technologies) in 500 l of Opti-MEM I medium (Gibco) in the wells of 6-well plates, and incubated for 20 min at room temperature. Cells in complete growth medium were seeded into 6-well plates at 150,000 cells per well, incubated for 48 h, and then the medium was changed to flesh complete growth