Article pubs.acs.org/jpr
PPAR Alpha: A Novel Radiation Target in Locally Exposed Mus musculus Heart Revealed by Quantitative Proteomics Omid Azimzadeh,*,†,‡ Wolfgang Sievert,†,§ Hakan Sarioglu,∥ Ramesh Yentrapalli,‡,⊥ Zarko Barjaktarovic,‡ Arundhathi Sriharshan,‡ Marius Ueffing,∥,# Dirk Janik,¶ Michaela Aichler,○ Michael J. Atkinson,‡,§ Gabriele Multhoff,§ and Soile Tapio‡ ‡
Helmholtz Zentrum München - German Research Center for Environmental Health, Institute of Radiation Biology, Neuherberg, Germany § Department of Radiation Oncology, Klinikum Rechts der Isar, Technical University of Munich, Munich, Germany ∥ Helmholtz Zentrum München - German Research Center for Environmental Health, Research Unit Protein Science, Neuherberg, Germany ⊥ Centre for Radiation Protection Research, Department of Genetics, Microbiology and Toxicology, Stockholm University, Stockholm, Sweden # Centre of Ophthalmology, University Medical Centre, Tuebingen, Germany ¶ Helmholtz Zentrum München, German Research Center for Environmental Health, Institute of Pathology, Neuherberg, Germany ○ Helmholtz Zentrum München, German Research Center for Environmental Health, Research Unit Analytical Pathology, Neuherberg, Germany S Supporting Information *
ABSTRACT: Radiation exposure of the thorax is associated with a markedly increased risk of cardiac morbidity and mortality with a latency period of decades. Although many studies have confirmed the damaging effect of ionizing radiation on the myocardium and cardiac endothelial structure and function, the molecular mechanism behind this damage is not yet elucidated. Peroxisome proliferator-activated receptor alpha (PPAR alpha), a transcriptional regulator of lipid metabolism in heart tissue, has recently received great attention in the development of cardiovascular disease. The goal of this study was to investigate radiation-induced cardiac damage in general and the role of PPAR alpha in this process in particular. C57BL/6 mice received local heart irradiation with X-ray doses of 8 and 16 gray (Gy) at the age of 8 weeks. The mice were sacrificed 16 weeks later. Radiation-induced changes in the cardiac proteome were quantified using the Isotope Coded Protein Label (ICPL) method followed by mass spectrometry and software analysis. Significant alterations were observed in proteins involved in lipid metabolism and oxidative phosphorylation. Ionizing radiation markedly changed the phosphorylation and ubiquitination status of PPAR alpha. This was reflected as decreased expression of its target genes involved in energy metabolism and mitochondrial respiratory chain confirming the proteomics data. This study suggests that persistent alteration of cardiac metabolism due to impaired PPAR alpha activity contributes to the heart pathology after radiation. KEYWORDS: ionizing radiation, proteomics, ICPL, PPAR alpha, endothelial cell, heart, cardiovascular disease
■
left-sided breast cancer patients.9 A broad range of cardiac disorders such as cardiomyopathy, myocardial fibrosis, valvular disorders, and coronary artery disease are known to be causally associated with exposure to high-dose ionizing radiation.10,11 Furthermore, animal studies have provided accruing data for cardiac injury after radiation.1 Although different studies clearly indicate the damaging effect of radiation to myocardial and endothelial structure and function, the mechanism leading to radiation-induced heart disease is not fully understood. It is known that radiation induces several cardiac pathologies in C57BL/6 mice, resulting from enhanced inflammatory and
INTRODUCTION Increased risk for cardiovascular disease (CVD) induced by radiation remains one of the important health concerns.1 Epidemiological studies of radiation-exposed populations show an increased risk of CVD associated with high local doses of ionizing radiation to the heart, as observed in patients after thoracic radiotherapy treatment for breast cancer, Hodgkińs disease or childhood cancers.2−5 A significant difference in the mortality from CVD has been reported for patients treated by radiotherapy for left-sided breast cancer comparing to those with right-sided cancer.2,6−8 Although the recent development of radiation therapy practice and equipment has decreased the heart dose from lefttangential radiotherapy considerably over the past 40 years, parts of the heart still receive more than 20 Gy in approximately half of © 2013 American Chemical Society
Received: January 23, 2013 Published: April 8, 2013 2700
dx.doi.org/10.1021/pr400071g | J. Proteome Res. 2013, 12, 2700−2714
Journal of Proteome Research
Article
before suspending in ICPL lysis buffer (SERVA). Protein concentration was determined by Bradford assay following the manufacturer’s instructions (Thermo Fisher). Protein ICPL Labeling and Separation. The labeling was done as previously reported.12 Briefly, triplicate aliquots of 50 μg of tissue lysate proteins obtained from either sham or irradiated mice were labeled with ICPL reagents (SERVA) following the manufacturer’s instructions. ICPL0 was used for control tissue and ICPL6 for irradiated tissue according to the manufacturer’s protocol. The heavy and light labeled samples were combined, and separated by 12% SDS gel electrophoresis before staining with colloidal Coomassie solution. SDS-PAGE lanes were cut into 5 slices. Prior to digestion, proteins were destained with 50 mM NH4HCO3 in 30% acetonitrile (ACN). In-gel digestion was performed overnight with trypsin of sequencing grade (SERVA Electrophoresis GmbH, Germany) using a total protein to enzyme ratio of 50:1 in 10 mM NH4HCO3. Peptides were extracted and acidified with 1% formic acid for subsequent mass spectrometry analysis. A protein mixture with known ratios of heavy and light label containing bovine serum albumin (1:1), chicken ovalbumin (4:1) and bovine carbonic anhydrase II (1:2) was used as an internal standard for labeling efficiency and data acquisitions. The labeling was done using three biological replicates. LC−ESI−MS/MS Analysis. The digested peptides were separated by reversed phase chromatography (PepMap, 15 cm × 75 μm ID, 3 μm/100 Å pore size, LC Packings) operated on a nano-HPLC (Ultimate 3000, Dionex) with a nonlinear 170 min gradient using 2% acetonitrile in 0.1% formic acid in water (A) and 0.1% formic acid in 98% acetonitrile (B) as eluted with a flow rate of 250 nL/min. The gradient settings were subsequently: 0− 140 min: 2−30% B, 140−150 min: 31−99% B, 151−160 min: Stay at 99% B and equilibrate for 10 min at starting conditions. The nano-LC was connected to a linear quadrupole ion trapOrbitrap (LTQ Orbitrap XL) mass spectrometer (Thermo Fisher, Bremen, Germany) equipped with a nano-ESI source. The mass spectrometer was operated in the data-dependent mode to automatically switch between Orbitrap-MS and LTQMS/MS acquisition. Survey full scan MS spectra (from m/z 300 to 1500) were acquired in the Orbitrap with resolution R = 60 000 at m/z 400 (after accumulation to a target of 1 000 000 charges in the LTQ). The method used allowed sequential isolation of the most intense ions, up to 10, depending on signal intensity, for fragmentation on the linear ion trap using collisioninduced dissociation at a target value of 100 000 ions. High resolution MS scans in the Orbitrap and MS/MS scans in the linear ion trap were performed in parallel. Target peptides already selected for MS/MS were dynamically excluded for 30 s. General mass spectrometry conditions were: electrospray voltage, 1.25−1.4 kV; no sheath and auxiliary gas flow. Ion selection threshold was 500 counts for MS/MS, and an activation Q-value of 0.25 and activation time of 30 ms were also applied for MS/MS. The acquired MS/MS spectra were searched against the Ensembl Mus musculus database using an in-house version of Mascot (Matrix Science, version 2.3.02; 20121023, Number of residues, 26 203 053; Number of sequences, 56416) with the following parameters: MS/MS spectra were searched with a precursor mass tolerance of 10 ppm and a fragment tolerance of 0.8 Da; Arg-C was selected as enzyme. One missed cleavage was allowed and carbamidomethylation was set as a fixed modification. Oxidised methionine and the heavy and light
antioxidative responses coupled to increased production of mitochondrial reactive oxygen species, disturbed lipid and pyruvate metabolism and remodelling of cardiac cytoskeleton.12−15 Seemann et al. observed marked structural and microvascular damage after a clinically relevant local cardiac high-dose (16 Gy) radiation in this mouse model.16 Vascular leakage, inflammation, diffuse amyloidosis and even sudden death between 30 and 40 weeks in 38% of mice was reported.16 Peroxisome proliferator-activated receptor alpha (PPAR alpha), a key regulator of lipid metabolism in the heart tissue, has recently received great attention in the development of cardiovascular disease.17,18 Impairment of lipid metabolism has been described as a consequence of altered transcriptional activity of PPAR alpha.19 PPAR alpha null mice show a characteristic low expression of proteins involved in fatty acid metabolism and transport,20 suggesting that PPAR alpha plays a critical role in regulation of myocardial energy homeostasis.19 Further, PPAR alpha is a regulator of anti-inflammatory and antifibrotic responses.17,21 The goal of this study was to investigate mechanisms leading to radiation-induced heart injury and the role of PPAR alpha in this process. C57BL/6 mice were exposed locally on the heart with X-ray doses of 8 and 16 Gy. To avoid losing animals to sudden death, cardiac tissue was investigated 16 weeks after the irradiation by quantitative proteomics. The proteomics data were further analyzed using transcriptomics, immunoblotting, bioinformatics, immunohistochemistry, electron microscopy and serum lipid profiling.
■
EXPERIMENTAL SECTION
Materials
Beta-octylglucoside, SDS, and ammonium bicarbonate were obtained from Sigma (St. Louis, MO); RapiGest from Waters (Milford, MA); acetone, acetonitrile, formic acid, and trifluoroacetic acid (TFA) from Roth (Karlsuhe, Germany); dithiothreitol (DTT), iodoacetamide, tris-(hydroxymethyl) aminomethane (Tris) and sequencing grade trypsin were obtained from Promega (Madison, WI); and cyano-4-hydroxycinnamic acid was obtained from Bruker Daltonik (Bremen, Germany). All solutions were prepared using HPLC grade water from Roth (Karlsuhe, Germany). Animals
All animal experiments were approved and licensed under federal law (Certificate No. 211-2531-54/01). 9-week-old male C57BL/ 6 mice (Charles River) received local cardiac irradiation with a single X-ray dose of 8 or 16 Gy (200 kV, 10 mA); age-related control mice received sham irradiation. Each mouse was immobilized without anesthetic in a specially designed jig. Preceding the exposure the position of the heart inside the jig was localized by digital radiographs. The heart irradiation field consisted of a 9 × 13 mm2 window in a lead plate of 2 mm of thickness. To avoid losing animals to sudden death, the mice were sacrificed 16 weeks after irradiation. Heart tissue was prepared as described before.12 Briefly, after sacrificing mice by cervical dislocation, the hearts were rapidly removed and rinsed with phosphate buffered saline. The heart tissue was snap-frozen in liquid nitrogen for further usage. A total number of 12 animals were used for this study. Proteomics
Protein Extraction. To lyse the tissue, frozen heart was ground to a fine powder with a cold (−20 °C) mortar and pestle 2701
dx.doi.org/10.1021/pr400071g | J. Proteome Res. 2013, 12, 2700−2714
Journal of Proteome Research
Article
tool STRING version 8.2 (http://string-db.org) coupled to the Reactome database (http://www.reactome.org).28 Pathway-focused Gene Expression Profiling with qRT-PCR. Total RNA was extracted from controls and the 8 and 16 Gy irradiated heart using the mirVana miRNA isolation kit (Applied Biosystems; Foster City, CA) following the protocol for total RNA isolation. The quantity and quality of the total RNA and miRNA was measured with the Nanodrop spectrophotometer (PeqLab Biotechnology; Germany). The mouse PPAR signaling pathway RT2 Profiler PCR arrays (Qiagen) were used to profile the expression of 84 genes related to PPAR signaling. Single stranded cDNA was synthesized from 100 ng of total RNA using the SuperArray reaction ready first strand cDNA synthesis kit. The cDNAs were mixed with SuperArray RT2 Real time SYBR Green/ROX PCR master mix and real time PCR performed in accordance with the manufacturer’s instructions. Thermal cycling and fluorescence detection were performed using a StepOne Sequence Detection System (Applied Biosystems, Foster City, CA), according to the manufacturer’s instructions., and expression of PPAR regulated transcripts were compared between the groups using student’s t test; p ≤ 0.05 was considered significant.
ICPL labels of lysines as well as heavy and light labels of the protein N-terminus were set as variable modifications. Quantification with ICPL. Data processing for protein identification and quantification of ICPL pairs was performed using Proteome Discoverer version 1.3 (Thermo Fisher) as described before.12 Briefly, the software provides automated strict statistical analysis of the protein quantification using only unique peptides. To minimize experimental bias the software was set to normalize on the protein median (minimum protein count: 20). The complete peptide and protein profiles were filtered using high peptide confidence and top one peptide rank filters. The false discovery rate (FDR) was calculated at the peptide level for all experimental runs using the Decoy option in Mascot; this rate was estimated to be lower than 5% using the identity threshold as the scoring threshold system. Proteins with a lower score were manually scrutinized and regarded as unequivocally identified if they fulfilled the following four criteria: (a) they had fragmentation spectra with a long, nearly complete y- and/or bseries; (b) all lysines were modified; (c) the numbers of lysines predicted from the mass difference of the labeled pair had to match the number of lysines in the given sequence from the search query and (d) at least one mass of a modified lysine was included in the detected partial fragment series.22 Due to these criteria, the ICPL method noticeably lowered the significance level of a protein score and increased the probability of a significant protein hit. Differentially labeled isotopic pairs were detected with a mass precision of 2 ppm and a retention time window of 0.3 min. The calculated peptide ratio variability in the Proteome Discoverer software is an alternative of coefficient of variation (CV) used to calculate a particular protein ratio. For the quantification data as replicates, software calculates the protein ratios for single searches as CV for log-normal distributed data and then calculates the classical coefficient variation for these ratios. The average heavy/light ratio and ratio variability were applied for protein quantification wherever multiple peptides were identified for a protein. The proteins identified by at least 2 unique peptides in two out of three replicates and quantified by the H/L variability less than 30% were considered for further evaluation. In the present study, proteins with ratios of heavy/ light (H/L) label greater than 1.3-fold or less than 0.7-fold (One sample t test; p ≤ 0.05; Perseus statistical tool) were defined as significantly differentially expressed. The biological significance of this fold change cut off is in good agreement with the previously published data.22−24
Immunoblotting Analysis
Proteins separated by SDS-PAGE were transferred to nitrocellulose membranes (GE Healthcare) using a TE 77 semidry blotting system (GE Healthcare) at 1 mA/cm for 1h. The membranes were blocked using 3% BSA in PBS, pH 7.4, for 1 h at room temperature, washed three times in 10 mM Tris-HCl, pH 7.4, 150 mM NaCl for 5 min and incubated overnight at 4 °C with primary antibodies using dilutions recommended by the manufacturer. After washing three times, the blots were incubated with either horseradish peroxidase-conjugated or alkaline phosphatase-conjugated antimouse, antirabbit or antigoat secondary antibody (Santa Cruz Biotechnology) for 2 h at room temperature and developed using the ECL system (GE Healthcare) or 1-step NBT/BCIP method (ThermoFisher) following standard procedures. GAPDH was unchanged in proteomics profiles after 8 and 16 Gy and was therefore used as a loading control. Quantification of digitized images of immunoblot bands from three biological replicates was done using ImageJ (http://rsbweb.nih.gov/ij/). Protein Carbonylation Detection
To detect the protein targets of oxidative stress reflected by protein carbonylation, the carbonyl groups were detected following derivatization with dinitrophenylhydrazine (Sigma) as previously described.12 Samples of sham-irradiated and irradiated hearts were loaded on the same gel in similar amounts using same transfer conditions. The amount of the total protein was confirmed by Ponceau S staining for accurate comparison between the two groups. Samples of nonderivatized proteins were run in parallel to correct for nonspecific antibody binding. Total intensity of protein bands was quantified using ImageJ software by integration of all the pixel values in the band area after background correction.
Pathway and Functional Correlation Analysis
Gene Ontology (GO) Analysis. The differentially expressed proteins in all samples were categorized using the PANTHER (Protein Analysis Through Evolutionary Relationships) bioinformatics tool (http://www.pantherdb.org)25 and Database for Annotation, Visualization and Integrated Discovery (DAVID; http://www.david.niaid.nih.gov).26 Gene ontology (GO) categories “molecular function” and “cellular component” were analyzed. Protein−Protein Interaction and Signaling Network. The analyses of protein−protein interaction and signaling networks were performed by the software tool INGENUITY Pathway Analysis (IPA) (INGENUITY System, http://www. INGENUITY.com). IPA is a knowledge database generated from peer-reviewed scientific publications that enables discovery of highly represented functions and pathways (p < 0.001) from large quantitative data sets.27 The analysis of protein−protein interaction and signaling networks was performed by the search
Histology and Immunohistochemistry
The hearts were excised and snap-frozen in liquid nitrogen. Light microscopic evaluation of the hearts was performed on longitudinal sections encompassing left and right atria, ventricles and the interventricular septum. Images were taken with the Hamamatsu NanoZoomer 2.HT Slide scanning system (Hamamatsu Photonics K.K., Japan). Formalin fixed and paraffin embedded (FFPE) samples of murine hearts were stained with 2702
dx.doi.org/10.1021/pr400071g | J. Proteome Res. 2013, 12, 2700−2714
Journal of Proteome Research
Article
GmbH, Oberkochen, Germany). Pictures were acquired using a Slow Scan CCD-camera and iTEM software (Olympus Soft Imaging Solutions, Münster, Germany). Mitochondria were counted in five fields of view at 5000× magnification of two hearts of each experimental group. p-values were determined by the t test.
hematoxylin and eosin (HE), Periodic Acid Schiff (PAS), Prussian blue, Masson trichrome and Congo red. Longitudinal sections were stained with HE to detect morphological alterations and inflammatory infiltrations. PAS staining was performed to identify glycogen and proteoglycans. To investigate the presence of iron-laden macrophages indicating previous hemorrhage, Prussian blue staining was performed. The number of iron-positive cells was counted separately in the epicardium and myocardium. To compensate for variable tissue size, the numbers were normalized to heart circumference and heart area, respectively. Masson trichrome staining was used to detect an increase of interstitial collagen or fibrous thickening of the epicardium or endocardium. To detect amyloid deposits in the myocardium or vessel walls, a Congo red staining with polarizing microscopy was performed. In addition adipophilin immunohistochemistry was carried out to visualize an increased lipid droplet accumulation. Myocardial micro vessel density (MVD) was measured by immunohistochemical staining for CD31. Three randomized areas of the left ventricular myocardium (0.025 mm2) were chosen for quantification using the criteria of Weidner et al.:29 any brown-staining of endothelial cells or cell clusters, clearly separate from adjacent microvessels or other tissue elements, was considered as a single, countable micro vessel. The number of vessels counted in any of the areas was recorded and the mean and standard deviation were calculated. Inflammatory endothelial activation was measured by immunohistochemical staining of the adhesion molecule E-selectin. The expression of E-selectin was graded by evaluating the incidence and intensity of positively stained endothelium of myocardial vessels using the criteria of Tsokos.30 The intensity of the immunostaining was graded as absent (0), weak (1), moderate (2) and strong (3) and the percentage of blood vessels involved per visual field was graded as absent (0), < 10% (1), 10−50% (2) and >50% (3). The total score was achieved by adding both variables. An antibody against von Willebrand Factor (vWF) was used to identify thrombotic changes. CD45 immunohistochemistry was performed to detect inflammatory leukocytic infiltrates. Evaluation was performed as described for iron-positive cells. Irradiated and control groups were compared using the nonparametric Kruskal−Wallis rank sum test. Group differences were considered statistically significant if p < 0.05. To determine functional changes in the microvasculature, immunohistochemical staining with antibody against alkaline phosphatase was performed. Immunohistochemical staining of laminin was conducted to investigate alterations of the basal lamina. Each immunohistochemical staining was performed simultaneously with identical incubation times and concentrations for the primary and secondary antibody and diamino benzidine (DAB) solution.
Serum Triglyceride and Free Fatty Acids Assay
Blood was collected directly from heart after sacrificing. Serum was and kept at −80 °C for further experiments. For the quantitative determination of nonesterified (free) fatty acids (FFA), the serum concentrations of triglyceride (TG) and high-, low-, and very lowdensity lipoprotein (HDL, LDL and VLDL) cholesterol were measured by an in vitro enzymatic colorimetric assay (BioVision). Data Availability
Access to raw MS data is provided in RBstore (http://www. rbstore.eu), the archive of radiobiology studies where all MS/MS raw files for this publication can be found.
■
RESULTS
Radiation Response in the Heart Proteome
We used the ICPL method to investigate radiation-induced changes in the cardiac proteome. The Heavy/Light (H/L) ratios associated with significant variation of protein expression were determined as described before.24,22 The average values of the coefficient of variation (CV) obtained for H/L ratios of all quantified proteins were 11% (8 Gy) and 12.7% (16 Gy). Consequently, the variability of 30% (
