Proteomics-Based Metabolic Modeling Reveals That Fatty Acid Oxidation (FAO) Controls Endothelial Cell (EC) Permeability
2015; Elsevier BV; Volume: 14; Issue: 3 Linguagem: Inglês
10.1074/mcp.m114.045575
ISSN1535-9484
AutoresFrancesca Patella, Zachary T. Schug, Erez Persi, Lisa J. Neilson, Zahra Erami, Daniele Avanzato, Federica Maione, Juan R. Hernández‐Fernaud, Gillian Mackay, Liang Zheng, Steven Reid, Christian Frezza, Enrico Giraudo, Alessandra Fiorio, Kurt I. Anderson, Eytan Ruppin, Eyal Gottlieb, Sara Zanivan,
Tópico(s)Advanced Proteomics Techniques and Applications
ResumoEndothelial cells (ECs) play a key role to maintain the functionality of blood vessels. Altered EC permeability causes severe impairment in vessel stability and is a hallmark of pathologies such as cancer and thrombosis. Integrating label-free quantitative proteomics data into genome-wide metabolic modeling, we built up a model that predicts the metabolic fluxes in ECs when cultured on a tridimensional matrix and organize into a vascular-like network. We discovered how fatty acid oxidation increases when ECs are assembled into a fully formed network that can be disrupted by inhibiting CPT1A, the fatty acid oxidation rate-limiting enzyme. Acute CPT1A inhibition reduces cellular ATP levels and oxygen consumption, which are restored by replenishing the tricarboxylic acid cycle. Remarkably, global phosphoproteomic changes measured upon acute CPT1A inhibition pinpointed altered calcium signaling. Indeed, CPT1A inhibition increases intracellular calcium oscillations. Finally, inhibiting CPT1A induces hyperpermeability in vitro and leakage of blood vessel in vivo, which were restored blocking calcium influx or replenishing the tricarboxylic acid cycle. Fatty acid oxidation emerges as central regulator of endothelial functions and blood vessel stability and druggable pathway to control pathological vascular permeability. Endothelial cells (ECs) play a key role to maintain the functionality of blood vessels. Altered EC permeability causes severe impairment in vessel stability and is a hallmark of pathologies such as cancer and thrombosis. Integrating label-free quantitative proteomics data into genome-wide metabolic modeling, we built up a model that predicts the metabolic fluxes in ECs when cultured on a tridimensional matrix and organize into a vascular-like network. We discovered how fatty acid oxidation increases when ECs are assembled into a fully formed network that can be disrupted by inhibiting CPT1A, the fatty acid oxidation rate-limiting enzyme. Acute CPT1A inhibition reduces cellular ATP levels and oxygen consumption, which are restored by replenishing the tricarboxylic acid cycle. Remarkably, global phosphoproteomic changes measured upon acute CPT1A inhibition pinpointed altered calcium signaling. Indeed, CPT1A inhibition increases intracellular calcium oscillations. Finally, inhibiting CPT1A induces hyperpermeability in vitro and leakage of blood vessel in vivo, which were restored blocking calcium influx or replenishing the tricarboxylic acid cycle. Fatty acid oxidation emerges as central regulator of endothelial functions and blood vessel stability and druggable pathway to control pathological vascular permeability. Endothelial cells (ECs) 1The abbreviations used are:ECendothelial cellFAOfatty acid oxidationTCActricarboxylic acid cycleFAfatty acidiMATintegrative metabolic analysis toolGSMMgenome-scale metabolic network modelSILACstable-isotope labeling with amino acids in cell cultureHUVEChuman umbilical vein endothelial cellsECMextracellular matrixTEERtrans-endothelial electrical resistanceDCAdichloroacetateVEGFvascular endothelial growth factor. line the inner layer of the blood vessel wall and constitute a barrier between blood and surrounding tissue. As such, a tight regulation of EC permeability is crucial to maintain vessel functionality and avoid excessive extravasation of fluid and plasma proteins (1Goddard L.M. Iruela-Arispe M.L. Cellular and molecular regulation of vascular permeability.Thromb. Haemost. 2013; 109: 407-415Crossref PubMed Scopus (109) Google Scholar). Increased endothelial permeability is typical in inflammatory states and a hallmark of diseases such thrombosis, atherosclerosis, and cancer (2Borissoff J.I. Spronk H.M. Heeneman S. ten Cate H. Is thrombin a key player in the "coagulation-atherogenesis" maze?.Cardiovasc. Res. 2009; 82: 392-403Crossref PubMed Scopus (176) Google Scholar, 3Jain R.K. Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy.Science. 2005; 307: 58-62Crossref PubMed Scopus (4344) Google Scholar). Because of their unique localization, ECs are constantly exposed to oxygen and nutrients that fuel cell metabolism and whose levels vary in physiological and pathological conditions. Yet, how cell metabolism regulates endothelial permeability remains incompletely understood. endothelial cell fatty acid oxidation tricarboxylic acid cycle fatty acid integrative metabolic analysis tool genome-scale metabolic network model stable-isotope labeling with amino acids in cell culture human umbilical vein endothelial cells extracellular matrix trans-endothelial electrical resistance dichloroacetate vascular endothelial growth factor. Previous studies have reported that EC cultures use glucose as predominant source of energy by producing lactate through glycolysis. However, also fatty acids and glutamine contribute to ATP and metabolic intermediate production (4De Bock K. Georgiadou M. Schoors S. Kuchnio A. Wong B.W. Cantelmo A.R. Quaegebeur A. Ghesquiere B. Cauwenberghs S. Eelen G. Phng L.K, Betz I. Tembuyser B. Brepoels K. Welti J. Geudens I. Segura I. Cruys B. Bifari F. Decimo I. Blanco R. Wyns S. Vangindertael J. Rocha S. Collins R.T. Munck S. Daelemans D. Imamura H. Devlieger R. Rider M. Van Veldhoven P.P. Schuit F. Bartrons R. Hofkens J. Fraisl P. Telang S. Deberardinis R.J. Schoonjans L. Vinckier S. Chesney J. Gerhardt H. Dewerchin M. Carmeliet P. Role of PFKFB3-driven glycolysis in vessel sprouting.Cell. 2013; 154: 651-663Abstract Full Text Full Text PDF PubMed Scopus (858) Google Scholar, 5Spolarics Z. Lang C.H. Bagby G.J. Spitzer J.J. Glutamine and fatty acid oxidation are the main sources of energy for Kupffer and endothelial cells.Am. J. Physiol. 1991; 261: G185-G190PubMed Google Scholar, 6Leighton B. Curi R. Hussein A. Newsholme E.A. Maximum activities of some key enzymes of glycolysis, glutaminolysis, Krebs cycle, and fatty acid utilization in bovine pulmonary endothelial cells.FEBS Lett. 1987; 225: 93-96Crossref PubMed Scopus (59) Google Scholar, 7Dagher Z. Ruderman N. Tornheim K. Ido Y. Acute regulation of fatty acid oxidation and amp-activated protein kinase in human umbilical vein endothelial cells.Circulation Res. 2001; 88: 1276-1282Crossref PubMed Scopus (166) Google Scholar). Recent in vivo studies have shown that glycolysis is necessary for EC proliferation and motility in physiological and pathological angiogenesis (4De Bock K. Georgiadou M. Schoors S. Kuchnio A. Wong B.W. Cantelmo A.R. Quaegebeur A. Ghesquiere B. Cauwenberghs S. Eelen G. Phng L.K, Betz I. Tembuyser B. Brepoels K. Welti J. Geudens I. Segura I. Cruys B. Bifari F. Decimo I. Blanco R. Wyns S. Vangindertael J. Rocha S. Collins R.T. Munck S. Daelemans D. Imamura H. Devlieger R. Rider M. Van Veldhoven P.P. Schuit F. Bartrons R. Hofkens J. Fraisl P. Telang S. Deberardinis R.J. Schoonjans L. Vinckier S. Chesney J. Gerhardt H. Dewerchin M. Carmeliet P. Role of PFKFB3-driven glycolysis in vessel sprouting.Cell. 2013; 154: 651-663Abstract Full Text Full Text PDF PubMed Scopus (858) Google Scholar, 8Schoors S. De Bock K. Cantelmo A.R. Georgiadou M. Ghesquiere B. Cauwenberghs S. Kuchnio A. Wong B.W. Quaegebeur A. Goveia J. Bifari F. Wang X. Blanco R. Tembuyser B. Cornelissen I. Bouche A. Vinckier S. Diaz-Moralli S. Gerhardt H. Telang S. Cascante M. Chesney J. Dewerchin M. Carmeliet P. Partial and transient reduction of glycolysis by PFKFB3 blockade reduces pathological angiogenesis.Cell. Metab. 2014; 19: 37-48Abstract Full Text Full Text PDF PubMed Scopus (349) Google Scholar). Moreover, the peroxisome proliferator-activated receptor gamma coactivator 1-α, which can activate oxidative phosphorylation, blocks EC sprouting in diabetes (9Sawada N. Jiang A. Takizawa F. Safdar A. Manika A. Tesmenitsky Y. Kang K.T. Bischoff J. Kalwa H. Sartoretto J.L. Kamei Y. Benjamin L.E. Watada H. Ogawa Y. Higashikuni Y. Kessinger C.W. Jaffer F.A. Michel T. Sata M. Croce K. Tanaka R. Arany Z. Endothelial PGC-1alpha mediates vascular dysfunction in diabetes.Cell. Metab. 2014; 19: 246-258Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar). The intriguing information emerging from these studies is that key metabolic pathways, such as glycolysis and oxidative phosphorylation in the mitochondria, play an important role in ECs and that they are actively involved in the regulation of key cell functions. Mitochondrial fatty acid oxidation (FAO) is the process that converts fatty acids (FAs) into acetyl-CoA, which fuels the tricarboxylic acid cycle (TCAc) and generates reducing factors for producing ATP via oxidative phosphorylation. Cells can incorporate FAs from the culture media or can generate FAs from the hydrolysis of triglycerides or through de novo synthesis. FAs, then, can access the mitochondria according to their length; whereas short and medium-chain FAs (up to 12 carbon atoms) diffuse through the mitochondrial membrane, long-chain FAs (with 13–21 carbon atoms) are actively transported by the carnitine O-palmitoyl transferase (CPT) proteins, which are rate-limiting enzymes for this pathway (10Carracedo A. Cantley L.C. Pandolfi P.P. Cancer metabolism: fatty acid oxidation in the limelight.Nat. Rev. Cancer. 2013; 13: 227-232Crossref PubMed Scopus (759) Google Scholar). Previous work suggested that FAO is poorly utilized by EC cultures (4De Bock K. Georgiadou M. Schoors S. Kuchnio A. Wong B.W. Cantelmo A.R. Quaegebeur A. Ghesquiere B. Cauwenberghs S. Eelen G. Phng L.K, Betz I. Tembuyser B. Brepoels K. Welti J. Geudens I. Segura I. Cruys B. Bifari F. Decimo I. Blanco R. Wyns S. Vangindertael J. Rocha S. Collins R.T. Munck S. Daelemans D. Imamura H. Devlieger R. Rider M. Van Veldhoven P.P. Schuit F. Bartrons R. Hofkens J. Fraisl P. Telang S. Deberardinis R.J. Schoonjans L. Vinckier S. Chesney J. Gerhardt H. Dewerchin M. Carmeliet P. Role of PFKFB3-driven glycolysis in vessel sprouting.Cell. 2013; 154: 651-663Abstract Full Text Full Text PDF PubMed Scopus (858) Google Scholar), however, under certain stress conditions such as glucose deprivation, FAO becomes a major source of energy (7Dagher Z. Ruderman N. Tornheim K. Ido Y. Acute regulation of fatty acid oxidation and amp-activated protein kinase in human umbilical vein endothelial cells.Circulation Res. 2001; 88: 1276-1282Crossref PubMed Scopus (166) Google Scholar). Although it is striking to note how cells can adapt and remodel their metabolism, the role of key FAO enzymes in the control of EC functions is still largely unclear. Because of the complexity of the cell metabolome, global-scale metabolomic studies for in depth and quantitative analysis of metabolic fluxes are still challenging and computational models have provided invaluable help to better understand cell metabolism. Among them, the integrative metabolic analysis tool (iMAT), which integrates gene expression data with genome-scale metabolic network model (GSMM), has been successfully used to predict enzyme metabolic flux in several model systems and diseases (11Shlomi T. Cabili M.N. Herrgard M.J. Palsson B.O. Ruppin E. Network-based prediction of human tissue-specific metabolism.Nat. Biotechnol. 2008; 26: 1003-1010Crossref PubMed Scopus (471) Google Scholar, 12Jerby L. Ruppin E. Predicting drug targets and biomarkers of cancer via genome-scale metabolic modeling.Clin. Cancer Res. 2012; 18: 5572-5584Crossref PubMed Scopus (72) Google Scholar). Because gene expression and protein levels do not always correlate, and because enzymes levels do not necessarily reflect their enzymatic activity or the flux of the reaction that they are involved in, iMAT uses expression data as cue for the likelihood, but not final determinant, of enzyme activity. Modern MS technology and robust approaches for protein quantification, such as stable-isotope labeling with amino acids in cell culture (SILAC) (13Ong S.E. Blagoev B. Kratchmarova I. Kristensen D.B. Steen H. Pandey A. Mann M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics.Mol. Cell. Proteomics. 2002; 1: 376-386Abstract Full Text Full Text PDF PubMed Scopus (4569) Google Scholar) and advanced label-free algorithms (14Cox J. Hein M.Y. Luber C.A. Paron I. Nagaraj N. Mann M. MaxLFQ allows accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction.Mol. Cell. Proteomics. 2014; Abstract Full Text Full Text PDF Scopus (2688) Google Scholar), allow global comparative proteomic analysis and accurate measurements of protein and post-translational modification levels (15Lamond A.I. Uhlen M. Horning S. Makarov A. Robinson C.V. Serrano L. Hartl F.U. Baumeister W. Werenskiold A.K. Andersen J.S. Vorm O. Linial M. Aebersold R. Mann M. Advancing cell biology through proteomics in space and time (PROSPECTS).Mol. Cell. Proteomics. 2012; 11Abstract Full Text Full Text PDF Scopus (56) Google Scholar). We reasoned that the integration of quantitative MS-proteomic data into GSMM could contribute to the study of cell metabolism. Moreover, metabolic changes trigger activation of protein kinases (16Sengupta S. Peterson T.R. Sabatini D.M. Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress.Mol. Cell. 2010; 40: 310-322Abstract Full Text Full Text PDF PubMed Scopus (984) Google Scholar, 17Hardie D.G. AMP-activated protein kinase: an energy sensor that regulates all aspects of cell function.Genes Dev. 2011; 25: 1895-1908Crossref PubMed Scopus (1169) Google Scholar) to rapidly remodel the intracellular signaling and enable cells to adapt to these sudden alterations. Protein phosphorylation therefore plays an important role in regulating cell response to metabolic alteration and may hide information on cellular pathways and functions controlled by specific metabolic activities. MS-based proteomic approaches therefore offer an additional opportunity to investigate in an unbiased manner the interplay between cell metabolism and cell function (18Reid S. Hernandez-Fernaud J.R. Zanivan S. In vivo quantitative proteomics for the study of oncometabolism.Methods Enzymol. 2014; 543: 235-259Crossref PubMed Scopus (2) Google Scholar). We have previously shown (19Zanivan S. Maione F. Hein M.Y. Hernandez-Fernaud J.R. Ostasiewicz P. Giraudo E. Mann M. SILAC-based proteomics of human primary endothelial cell morphogenesis unveils tumor angiogenic markers.Mol. Cell. Proteomics. 2013; 12: 3599-3611Abstract Full Text Full Text PDF PubMed Scopus (47) Google Scholar) that when human primary ECs are cultured for 1 day on the three-dimensional matrix matrigel and assemble into a complex network, a simplified model that recapitulates some aspects of vascular network assembly in vivo (20Arnaoutova I. George J. Kleinman H.K. Benton G. The endothelial cell tube formation assay on basement membrane turns 20: state of the science and the art.Angiogenesis. 2009; 12: 267-274Crossref PubMed Scopus (346) Google Scholar), the levels of metabolic enzymes are profoundly regulated. This result suggested an interplay between cell metabolism and EC behavior. Here we investigate further this aspect. Integrating label-free quantitative MS-proteomics, predictive metabolic modeling and metabolomics we discovered increased FAO when ECs are assembled into a fully formed network. Moreover, by inhibiting CPT1 pharmacologically, we elucidated that FAO is a central regulator of EC permeability in vitro and blood vessel stability in vivo. Thus, proteomics significantly contributes to the study of cell metabolism and here we identified FAO as a promising target for therapeutic intervention for the control of pathological vascular permeability. Human umbilical vein endothelial cells (HUVECs) isolated from 2–5 umbilical cords were pooled and cultured in EGM-2 (Lonza, Basel, Switzerland). For the SILAC labeling, cells were grown for three passages (P) in custom EGM-2 without arginine and lysine (Lonza) supplemented with l-arginine and l-lysine (SILAC light) (Sigma-Aldrich, St. Louis, MO), 13C6 l-arginine and D4 l-lysine (SILAC medium) or 13C615N4 l-arginine, and 13C615N2 l-lysine (SILAC heavy) (Cambridge Isotope Laboratories, Tewksbury, MA). BOECs were kindly provided by Dr. Maartje van den Biggelaar and cultured in EGM-2 medium 10% FBS. Cells were used between P2 and P6. If not otherwise stated, after 2 h etomoxir treatment (15 μg/ml) cells were treated or not with pyruvate (500 μm) for 30 min followed by dichloroacetate (5 mm) for 30 min before cells were used in experiments. Etomoxir, oxfenicine, dichloroacetate, pyruvate, thrombin, VEGF, and mouse anti-vinculin antibody were from Sigma-Aldrich; anti-CPT1A antibody (15184–1-AP) was from Protein Tech group (Chicago, IL); anti-β-tubulin was from Santa Cruz Biotechnology (Dallas, TX), anti-phospho ACACA was from Cell Signaling (Danvers, MA); anti-mouse IRDye 700CW, and anti-rabbit IRDye 800CW used for Western blot were from LI-COR Biosciences (Lincoln, NE). Matrigel and Cell recovery solution were from BD biosciences (Franklin Lakes, NJ). HUVECs were seeded and cultured on solidified Matrigel in EGM-2 medium with the indicated stimuli and harvested for MS analysis using Cell recovery solution according to manufacturer's instructions and as previously described (19Zanivan S. Maione F. Hein M.Y. Hernandez-Fernaud J.R. Ostasiewicz P. Giraudo E. Mann M. SILAC-based proteomics of human primary endothelial cell morphogenesis unveils tumor angiogenic markers.Mol. Cell. Proteomics. 2013; 12: 3599-3611Abstract Full Text Full Text PDF PubMed Scopus (47) Google Scholar). Pictures were taken with Axiovert microscope and the tubule length measured with ImageJ software. HUVECs were lysed in 2% SDS and 100 mm Tris-HCl pH 7.4 buffer. Proteins were precipitated and solubilized in 8 m urea, 75 mm NaCl and 50 mm TrisHCl. After reduction with dithiothreitol and alkylation with iodoacetamide, proteins were digested with trypsin. Proteome etomoxir: Light, medium, and heavy SILAC-labeled cell lysates (∼70 μg/sample) were mixed in equal amount, trypsin digested by filter-aided sample preparation (FASP) method (22Wisniewski J.R. Zougman A. Nagaraj N. Mann M. Universal sample preparation method for proteome analysis.Nat. Methods. 2009; 6: 359-362Crossref PubMed Scopus (5043) Google Scholar) and 50 μg of peptides fractionated into six fractions using on-tip strong anion exchange chromatography (21Wisniewski J.R. Zougman A. Mann M. Combination of FASP and StageTip-based fractionation allows in-depth analysis of the hippocampal membrane proteome.J. Proteome Res. 2009; 8: 5674-5678Crossref PubMed Scopus (437) Google Scholar). Light, medium, and heavy SILAC-labeled cell lysates (∼3 mg/sample) were mixed in equal amount, digested using FASP and peptides were fractionated using strong cation exchange chromatography followed by titanium dioxide enrichment (23Larsen M.R. Thingholm T.E. Jensen O.N. Roepstorff P. Jorgensen T.J. Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns.Mol. Cell. Proteomics. 2005; 4: 873-886Abstract Full Text Full Text PDF PubMed Scopus (1333) Google Scholar) for phosphorylated peptides as previously described (24van den Biggelaar M. Hernandez-Fernaud J.R. van den Eshof B.L. Neilson L.J. Meijer A.B. Mertens K. Zanivan S. Quantitative phosphoproteomics unveils temporal dynamics of thrombin signaling in human endothelial cells.Blood. 2014; 123: e22-e36Crossref PubMed Scopus (31) Google Scholar). Digested peptides were de-salted with Empore-C18 StageTips (25Rappsilber J. Ishihama Y. Mann M. Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics.Anal. Chem. 2003; 75: 663-670Crossref PubMed Scopus (1796) Google Scholar), eluted in 80% acetonitrile (ACN), 0.5% acetic acid, and stored at −80 °C until MS analysis. Tryptic peptides were separated on 20 cm fused silica emitter (New Objective, Woburn, MA) packed in-house with the reverse phase ReproSil-Pur C18-AQ, 1.9 μm resin (Dr. Maisch, GmbH, Ammerbuch-Entringen, Germany) and analyzed on a LTQ-Orbitrap Elite (Thermo Fisher Scientific) coupled on-line with a nano-HPLC (Easy nLC, Thermo Fisher Scientific). For each sample, ∼2 μg of digested peptides were eluted from reverse phase column with a flow of 200 nl/min in 190 min gradient, from 5% to 30% ACN in 0.5% acetic acid. For each time point three replicates were performed and each replicate was run at the MS twice. For each fraction, half of the peptides were loaded onto reverse phase column and eluted with a flow of 200 nl/min in 190 min gradient, from 5% to 30% ACN in 0.5% acetic acid. Triplicate experiments were performed swapping SILAC labeling conditions. For each experiment, 10 fractions enriched for phosphorylated peptides were analyzed at the MS. Two thirds of each sample was loaded onto reverse phase column and eluted with a flow of 200 nl/min in 90 min gradient, from 5% to 30% ACN in 0.5% acetic acid. The remaining 1/3 was pooled into two fractions that were analyzed at the MS. Triplicate experiments were performed swapping SILAC labeling conditions. MS spectra were acquired in the Orbitrap analyzer at a resolution of 120,000 at 400 m/z, and a target value of 106 charges. High collision dissociation fragmentation of the 10 most intense ions was performed using a target value of 40,000 charges and acquired in the Orbitrap at resolution 15,000 at 400 m/z. Data were acquired with Xcalibur software. MS data were processed using the MaxQuant software (26Cox J. Mann M. MaxQuant enables high peptide identification rates, individualized p. p. b.-range mass accuracies and proteome-wide protein quantification.Nat. Biotechnol. 2008; 26: 1367-1372Crossref PubMed Scopus (9154) Google Scholar) and searched with Andromeda search engine (27Cox J. Neuhauser N. Michalski A. Scheltema R.A. Olsen J.V. Mann M. Andromeda – a peptide search engine integrated into the MaxQuant environment.J Proteome Res. 2011; 10: 1794-1805Crossref PubMed Scopus (3450) Google Scholar) against the human UniProt database (release-2012 01, 81,213 entries). An initial maximal mass deviation of 7 ppm and 20 ppm was required to search for precursor and fragment ions, respectively. Trypsin with full enzyme specificity and peptides with a minimum length of seven amino acids were selected. Two missed cleavages were allowed. Oxidation (Met) and N-acetylation were set as variable modifications, as well as phospho(STY) for the phosphoproteome analysis, whereas Carbamidomethylation (Cys) as fixed modification. False discovery rate (FDR) of 1% was used for peptides, proteins and phosphopeptides identification. For the phosphosites, a minimum Andromeda phosphopeptides score of 40 was required, as previously described (28Sharma K. D'Souza R.C. Tyanova S. Schaab C. Wisniewski J.R. Cox J. Mann M. Ultradeep human phosphoproteome reveals a distinct regulatory nature of tyr and ser/thr-based signaling.Cell Rep. 2014; 8: 1583-1594Abstract Full Text Full Text PDF PubMed Scopus (636) Google Scholar). Peptides and proteins were quantified according to the MaxLFQ algorithm of MaxQuant (14Cox J. Hein M.Y. Luber C.A. Paron I. Nagaraj N. Mann M. MaxLFQ allows accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction.Mol. Cell. Proteomics. 2014; Abstract Full Text Full Text PDF Scopus (2688) Google Scholar) version 1.4.1.0. Only proteins uniquely identified with minimum one unique peptide and quantified in at least three MS runs were used for the analysis. The relative quantification of the proteins against their labeled counterpart was performed by MaxQuant (14Cox J. Hein M.Y. Luber C.A. Paron I. Nagaraj N. Mann M. MaxLFQ allows accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction.Mol. Cell. Proteomics. 2014; Abstract Full Text Full Text PDF Scopus (2688) Google Scholar) version 1.5.0.36. Only proteins identified with minimum one unique peptide and quantified with a minimum of two ratio counts were used for the analysis. Proteins were considered up-regulated if the SILAC ratio was higher than 0.3 (log2 scale), which was more than one standard deviation from the mean of the all calculated ratios, in a minimum of two replicates. The relative quantification of the phosphorylation sites against their labeled counterpart was performed by MaxQuant (14Cox J. Hein M.Y. Luber C.A. Paron I. Nagaraj N. Mann M. MaxLFQ allows accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction.Mol. Cell. Proteomics. 2014; Abstract Full Text Full Text PDF Scopus (2688) Google Scholar) version 1.4.1.6. Only class I sites (= sites accurately localized with localization probability > 0.75 and score difference > 5) were used for the analysis. Phosphorylation sites were considered up-regulated if the SILAC ratio was higher than 0.4 (log2 scale), which was more than one standard deviation from the mean of the all calculated ratios, in a minimum of two replicates. For the NetworKIN analysis (29Horn H. Schoof E.M. Kim J. Robin X. Miller M.L. Diella F. Palma A. Cesareni G. Jensen L.J. Linding R. KinomeXplorer: an integrated platform for kinome biology studies.Nat. Methods. 2014; 11: 603-604Crossref PubMed Scopus (209) Google Scholar), for each phosphorylation site only the predicted kinase with highest score was considered and we required a minimum NetworKIN score of 1.5. Motif-X analysis was performed using standard parameters, significance of 0.000001 and IPI Human Proteome as background (30Schwartz D. Gygi S.P. An iterative statistical approach to the identification of protein phosphorylation motifs from large-scale data sets.Nat. Biotechnol. 2005; 23: 1391-1398Crossref PubMed Scopus (716) Google Scholar). Predicted kinase activity was calculated by means of significantly overrepresented (Fisher test, with 2% FDR) kinase motifs (used "Motifs" column supplemental Table S3 which was generated with Perseus software, based on Human Protein Reference Database (31Keshava Prasad T.S. Goel R. Kandasamy K. Keerthikumar S. Kumar S. Mathivanan S. Telikicherla D. Raju R. Shafreen B. Venugopal A. Balakrishnan L. Marimuthu A. Banerjee S. Somanathan D.S. Sebastian A. Rani S. Ray S. Harrys Kishore C.J. Kanth S. Ahmed M. Kashyap M.K. Mohmood R. Ramachandra Y.L. Krishna V. Rahiman B.A. Mohan S. Ranganathan P. Ramabadran S. Chaerkady R. Pandey A. Human Protein Reference Database–2009 update.Nucleic Acids Res. 2009; 37: D767-D772Crossref PubMed Scopus (2518) Google Scholar)) within the 83 up-regulated sites upon etomoxir treatment. The 83 sites were queried against the entire phospho-dataset. Metabolic genes for which protein abundance levels (LFQ) were measured in experiments were mapped to the human genome-scale metabolic model (GSMM) (32Duarte N.C. Becker S.A. Jamshidi N. Thiele I. Mo M.L. Vo T.D. Srivas R. Palsson B.O. Global reconstruction of the human metabolic network based on genomic and bibliomic data.Proc. Natl. Acad. Sci. U.S.A. 2007; 104: 1777-1782Crossref PubMed Scopus (1064) Google Scholar). The mean (over three replicas) of protein abundance levels in each time point (i.e. 4 h, 22 h) were used to infer ternary presentation of the abundance levels using "quartile" partitioning. This allowed for integrating 50% of the measured data, such that proteins in the top 25% quartile were labeled "1" (highly abundant), proteins in the down 25% quartile were labeled "−1" (lowly abundant) and the rest were labeled "0" (moderately abundant), in each time point. Based on the GSMM gene-reaction rules, i.e. the logical dependence of each reaction on the activity of the genes associated with it, we infer the ternary state at the reaction level. This ternary representation was used as "cues" (soft constraints) to perform iMAT (11Shlomi T. Cabili M.N. Herrgard M.J. Palsson B.O. Ruppin E. Network-based prediction of human tissue-specific metabolism.Nat. Biotechnol. 2008; 26: 1003-1010Crossref PubMed Scopus (471) Google Scholar) in each time point. To assess the permissible flux range (i.e. minimal and maximal flux) of each reaction we performed flux-variability analysis (FVA) around the optimal solution that maximizes the agreement between the predicted fluxes and the proteomic measurements. Then, we sampled the solution space using ACHR algorithm and estimated the average flux of each reaction. Fold changes between 22 h and 4 h were derived based on the average fluxes. The pathway enrichment analysis based on fold change reaction flux between 4h and 22h matrigel (in supplemental Table S2) was performed using the one dimension (1D) annotation enrichment analysis available in the Perseus software (33Cox J. Mann M. 1D and 2D annotation enrichment: a statistical method integrating quantitative proteomics with complementary high-throughput data.BMC Bioinformatics. 2012; 16: S12Crossref Scopus (413) Google Scholar). To simulate the effect of CPT1A inhibition at 22h we simulated the metabolic state using iMAT twice: once when the reactions associated with CPT1A were active at their maximal flux, and once when they were inhibited, carrying no flux. FVA and sampling (ACHR) of the solution space were performed. Based on the average fluxes of the reactions we estimated the fold change following CPT1A inhibition as: fluxes when CPT1A was inactive/fluxes when CPT1A was active. HUVECs were seeded on a solidified matrigel (six well plate, 200 μl/9.6 cm2) in EGM-2 medium. After 3h, 22h, and 30h, cells were washed with PBS, and medium replaced with 1 ml EGM-2, 11 mm 13C6 glucose and 100 μm 12C16 palmitic acid or 11 mm 12C6 glucose and 100 μm 13C16 palmitic acid. After 6h incubation at 37 °C metabolites were extracted as follow from triplicate samples: Twenty microliters of supernatant were mixed with 980 μl of cold methanol/ACN/water (5:3:2 volumes) extraction
Referência(s)