Lipid Metabolism and Peroxisome Proliferator-Activated Receptor

Dec 30, 2010 - Signaling Pathways Participate in Late-Phase Liver Regeneration. Xing Yuan,. †,‡,§. Shikai Yan,. †,|| Jing Zhao,. §,^. Duo Shi,...
0 downloads 0 Views 5MB Size
ARTICLE pubs.acs.org/jpr

Lipid Metabolism and Peroxisome Proliferator-Activated Receptor Signaling Pathways Participate in Late-Phase Liver Regeneration Xing Yuan,†,‡,§ Shikai Yan,†,|| Jing Zhao,§,^ Duo Shi,‡ Bin Yuan,‡ Weixing Dai,§ Binghua Jiao,‡ Weidong Zhang,*,§,|| and Mingyong Miao*,‡ ‡

Department of Biochemistry and Molecular Biology, Second Military Medical University, Shanghai 200433, People's Republic of China School of Pharmacy, Second Military Medical University, Shanghai 200433, People's Republic of China School of Pharmacy, Shanghai Jiao Tong University, Shanghai 200030, People's Republic of China ^ Department of Mathematics, Logistical Engineering University, Chongqing 400016, People's Republic of China

)

§

bS Supporting Information ABSTRACT: Liver regeneration (LR) is of great clinical significance in various liver-associated diseases. LR proceeds along a sequence of three distinct phases: priming/initiation, proliferation, and termination. Compared with the recognition of the first two phases, little is known about LR termination and structure/function reorganization. A combination of “omics” techniques, along with bioinformatics, may provide new insights into the molecular mechanism of the late-phase LR. Gene, protein, and metabolite profiles of the rat liver were determined by cDNA microarray, two-dimensional electrophoresis, and HPLC-MS analysis. Pathway enrichment analysis was performed to identify the pathways: 427 differentially expressed genes extracted from the microarray experiment revealed two expression patterns representing the early and late phase of LR. Functionally, the genes expressing at a higher level at the early phase than at the late phase were mainly involved in the response to stress, proliferation, and resistance to apoptosis, while those expressing at a lower level at the early phase than at the late phase were mainly engaged in lipid metabolism. Compared with the sham-operation control (SH) group, 5 proteins in the 70% partial hepatectomy (70%PHx) group were upregulated at the protein level, and 3 proteins were downregulated at 168 h after the 70%PHx. E-FABP, an upregulated fatty acid binding protein, was found to be involved in the peroxisome proliferatoractivated receptor (PPAR) signaling pathway. The metabolomic data confirmed the enhancement of lipid metabolism by the detection of the intermediate and final metabolites. We've concluded that increased lipid metabolism and activated PPAR signaling pathways play important roles in late-phase LR. KEYWORDS: liver regeneration, genomics, proteomics, metabolomics, lipid metabolism pathway, peroxisome proliferatoractivated receptor signaling pathway

’ INTRODUCTION The liver is one of the few glands with the capacity to regenerate after surgical, microbiological, or chemical trauma,1-4 and it seems that the liver “knows” when to start and when to stop growing.5,6 A 70% partial hepatectomy (70%PHx) induces cell proliferation until the original mass of the liver is restored.7 In parallel with the experimental work, human partial liver transplantation from cadavers and living donors continues to increase. The new method of cell transplantation is also particularly useful in patients with acute liver failure and widespread cell necrosis.8 Furthermore, tumors possess the characteristic of high proliferation similar to the priming phase of liver regeneration (LR). Unlike cells whose reproduction is precisely regulated during LR, tumor cells have a r 2010 American Chemical Society

very poor ability to apoptose. Thus, the knowledge generated from laboratory work with rodents in this field has great promise to solve clinical problems. Previous studies usually adopted conventional molecular biology methods to study the signal transmission pathways and regulatory mechanisms that control the replication of hepatocytes,9-12 while the complexity of the regeneration mechanisms including priming, proliferation and termination is still under exploration. With the development of systems biology, high throughput “omics” technology has been applied in the research field of LR, and a comprehensive analysis of LR at the gene, protein and Received: September 19, 2010 Published: December 30, 2010 1179

dx.doi.org/10.1021/pr100960h | J. Proteome Res. 2011, 10, 1179–1190

Journal of Proteome Research metabolite levels can be achieved by using genomics, proteomics and metabolomics.13-19 In our opinion, an integration of “omics” information at different levels is promising for a better explanation of the complexity of the LR course. The LR course includes priming, proliferation and termination phases. Although elements that trigger LR have been the object of intensive studies,20-22 the mechanism of LR inactivation by the termination factor system after the proliferation phase has not been well investigated. At present, most known factors involved in the termination response comprise the TGF family, including the negative regulatory element TGFβ and Activins.23-26 However, recent studies showed that some factors of the TGF family may play positive regulatory roles during the termination phase of LR.27-29 It is therefore worthy to further explore unknown mechanisms of termination. In the present study, we utilized genomics, proteomics and metabolomics methods to examine biological variations of LR. In addition, we applied bioinformatics approaches to integrate these high throughput “omics” data and analyze significant variations of the biological function of regulatory or signaling pathways. The use of systems biology approaches to explore the overall perspective on complicated systematic variations in 70%PHx hepatocytes and especially to study biological function regulation in response to late-phase LR resulted in an overall view of the 70%PHx-induced late-phase responses of rat hepatocytes. The proposed pathways and biomarkers obtained through this approach might enable us to develop new ideas about the termination and structure/function reorganization of LR.

’ METHODS Animals and Surgery

Forty adult male Sprague-Dawley rats (220-250 g) obtained from the Experimental Animal House at the Second Military Medical University (Shanghai, China) were equally randomized into 4 groups at 6 or 168 h after either 70%PHx or sham surgery. All groups were housed with a 12 h light/12 h dark cycle and were given a standard diet until they were euthanized. A 70% partial hepatectomy (70%PHx) was performed in the test (PH) group as previously described;1 the median and left lateral lobes of the liver were removed without injuring the remaining tissue. The control (SH) group was subjected to a sham operation with the same procedure as the test group but without removing the liver. Sample Collection

Necropsy was carried out immediately after euthanasia. The removed liver lobe was immediately weighed, flash-frozen in liquid nitrogen and stored at -80 C for subsequent genomic, proteomic and metabolomic analysis. Microarray Analysis

cDNA Microarray Analysis. A rat 10k cDNA microarray (Hujing Biotech Ltd., Shanghai, China; NCBI registration No. GPL4294) was used in the experiments. The microarray contained 9234 cDNA probe sets, with 6455 unigenes and 2125 homologous genes. Finally, 6753 genes were identified after data analysis (Supporting Information, Additional File 1). To ensure homogeneity of the microarray, multiple replicates of the microarray were set up for quality control of the samples. Every small spotting matrix contained several positive controls and negative controls. To ensure the accuracy of the results, clones of all the genes underwent strict sequencing verification.

ARTICLE

RNA Extraction and Preparation of Probe cDNA. Total liver RNA was extracted using Trizol (Takara, Japan) and purified using the QIAGEN quick PCR purification kit and QIAGEN quick Nucleotide Removal Kit (QIAGEN Inc., German). Probe cDNA was generated using 100 μg total RNA. The total RNA from the test group was labeled with red-fluorescent Cy3-dUTP during reverse transcription, and the total RNA from the control group was labeled with blue-fluorescent Cy5-dUTP (both from Amersham Pharmacia Inc., Sweden). Reverse transcription was performed by Oligo-dT primer (Operon Inc., USA) and SuperScript II RNase H-Reverse Transcriptase (Invitrogen Inc., Carlsbad, CA) for 1 h at 42 C. After reverse transcription, probe cDNA was prepared as described previously.30 Hybridization and Data Scanning. After filtration, separate images from each fluorescent signal were scanned by a ScanArray 5000 confocal laser scanner and analyzed using Scannalyze2 software. The color images of the hybridization results were generated by representing the Cy3 fluorescent image as red and the Cy5 fluorescent image as green, and then the two color images were merged. Each signal was evaluated using Scannalyze2 for data analysis. Data Analysis. To increase the accuracy of the data, the same experiment was performed three times, and the data that were consistent were extracted. The images were converted to numerical information by ImaGene (Scan resolution 10 μm, PMT 100%), the data were standardized with GeneSpring software and the ratio was expressed as Cy3/Cy5 (representative of PH/ SH). The discrepant genes were selected according to the following criteria: a ratio g2 for upregulated genes and a ratio e0.5 for downregulated genes. The final data were represented as a log-ratio (base 2): log2 (Cy3/Cy5). Real-Time PCR. The reliability of the cDNA microarray results was detected by conventional reverse transcriptional polymerase chain reaction (RT-PCR) on both the PH group and SH group liver tissues for the selected genes. The total RNA was from liver tissues at 6 and 168 h after 70%PHx, and cDNAs were prepared according to the manufacturer’s instructions (Invitrogen, Carlsbad, USA), and quantitative PCR was performed on the TP800 real-time PCR system (Takara Bio) using a standard SYBR Green PCR kit (Takara Bio). The relative expression of the measured genes in each sample was compared to β-actin, and the significance was calculated by using the average of the β-actin-normalized 2-ΔΔCt value. PCR amplification was performed using the following primers for the indicated genes: rat Cyp7a1 (50 -GCTTTACAGAGTGCTGGCCAA-30 and 50 -CTGTCTAGTACC GGCAGGTCATT-30 ), Cyp8b1 (50 -GTACACATGGACCCCGACATC-30 and 50 -GGG TGCCATCAGGG TTGAG-30 ). Proteomic Analysis and Result Validation

Sample Preparation. Equal portions of frozen liver tissues (0.1 g) from either group of PH or SH at 168 h after 70%PHx (10 per group) were mixed and thawed at room temperature and were homogenized in 10 volumes of tissue lysis buffer containing 9 M urea, 4% CHAPS, 65 mM DTT, 0.5% carrier ampholytes Bio-Lyte 3-10 (Bio-Rad, Richmond, CA) and protease inhibitors. The suspensions were sonicated for approximately 40 s and centrifuged for 1 h at 15 000 g to remove DNA, RNA and any particulate materials. The supernatants contained total liver proteins solubilized in sample buffer. The protein concentrations in the supernatants were determined using a Bradford Assay Kit (Bio-Rad). All protein samples were aliquoted and stored at 80 C until use. 1180

dx.doi.org/10.1021/pr100960h |J. Proteome Res. 2011, 10, 1179–1190

Journal of Proteome Research Two-Dimensional Electrophoresis (2-DE) and Image Analysis. 2-DE was performed essentially according to the manufacturer’s (Bio-Rad) protocol. For the first dimension, 1 mg protein was applied to immobilized 17-cm, pH 3-10 nonlinear IPG strips (Bio-Rad). Rehydration was achieved at 50 V for 14 h. Isoelectric focusing (IEF) was started at 250 V using a Protean IEF Cell apparatus (Bio-Rad), and the voltage was increased gradually to 10 000 V and kept constant until 80 000 Vh. Separation in the second dimension was performed on 10% SDS-PAGE gels (230 mm  200 mm  1.5 mm) at a constant current of 30 mA/gel in a Criterion gel apparatus (Bio-Rad). 2D gels were stained with Coomassie blue G-250 for 13 h,31 and after electrophoresis, the gels were scanned with an Image Scanner (UMAX). Spot detection, matching and quantitative intensity analysis were performed using PDQuest7.4 software (Bio-Rad). Briefly, protein spots were first automatically detected and then manually refined, and the gels were normalized and matched. Each intergroup comparison of the samples between PH and SH was carried out on three separate paired gels. Similar qualitative changes had to be detected in all three paired repeats of a given comparison and the expression change had to be at least 2-fold to ensure that changes in the protein spots reflected genuine differences in actual protein expression. MALDI-TOF-MS and Database Searching. Each sample was suspended in 0.7 μL matrix solution (R-cyano-4-hydroxycinnamic acid (CHCA) in acetonitrile/water (1:1, v/v) acidified with 0.1% (v/v) TFA), and the mixture was immediately spotted onto the MALDI target and allowed to dry and crystallize. Analyses were performed on a 4700 Proteomics Analyzer (TOF/TOF) (Applied Biosystems, Foster City, CA) equipped with a 355-nm Nd:YAG laser. The proteins were identified by peptide mass fingerprinting (PMF) and tandem mass spectrometry (MS/MS) using the MASCOT program (version 1.9, Matrix Science, London, U.K.) against the Swiss-Prot database with GPS Explorer software (Applied Biosystems), and MASCOT protein scores (based on the combined MS and MS/MS spectra) greater than 56 were considered statistically significant (P e 0.05). Validation: Western Blot Analysis. Western blot analysis was performed according to the method previously described.32 Antibodies including E-FABP (epidermal-type fatty acid-binding protein) (Santa Cruz, CA) and β-actin (Thermo Scientific) were used at the appropriate dilution as recommended by the manufacturer. Validation: Immunohistochemical (IHC) Analysis. IHC staining was performed according to the method previously described.33 Fresh liver tissues of PH and SH at 168 h after surgery were fixed in 10% formalin and embedded in paraffin. On paraffin sections, a two-step indirect EnVision (Dako, Carpentaria, CA) technique was used for the E-FABP antibody. At the final stage, the sections were developed using 3,30-diaminobenzidine-tetrahydrochloride (DAB), and the slides were counterstained by Mayer’s hematoxylin and mounted. E-FABP-positive cells exhibited the deposition of brown DAB precipitate. Positive reactions were classified using the following criteria: (-) negative, (þ) 10% or more to less than 20%, (þþ) 20% or more to less than 40%, and (þþþ) 40% or more. Tissues with more than 20% of the hepatocytes stained were classified as E-FABP protein-positive. Identification of Pathways Enriched with LR-Associated Proteins. Differentially expressed genes identified by our microarray experiments and the genes corresponding to

ARTICLE

differentially expressed proteins in the proteomic experiments were considered LR-associated genes. These genes were mapped onto the rat pathways in the KEGG database, and then P-values were used to determine whether a pathway was enriched with LR-associated genes instead of by chance. Assuming that the total LR-associated genes (K) were mapped onto the rat pathways in KEGG, which included N distinct genes, a hypergeometric cumulative distribution function could model the probability of identifying at least k genes from a pathway of size n by chance, such that the P-value is given by: ! ! K N -K kX -1 kX -1 i n -i ! f ðiÞ ¼ 1 P ¼ 1N i¼0 i¼0 n Given the significance level R = 0.01, a P-value smaller than R demonstrated a low probability that the LR-associated genes appeared in the pathway by chance, that is, the pathway can be regarded as being significantly regulated by these gene-encoded proteins. Metabolomic Analysis

Sample Preparation. Liver tissues of PH and SH were homogenized in 2 mL physiological saline solution on ice, and the suspensions were then centrifuged for 5 min at 3000 g. Four volumes of acetonitrile (J. T. Baker, Phillipsburg, NJ) was added to the supernatant and shaken vigorously (60 s), and then the mixture was placed in a refrigerator at 4 C for 10 min and centrifuged at 12 000 g for 15 min. The supernatant was then stored in a refrigerator at -20 C prior to HPLC-MS analysis. HPLC-MS/MS Analysis. LC-MS analysis was performed on an Agilent 1200 rapid resolution liquid chromatograph (RRLC) coupled with a quadrupole time-of-flight mass spectrometer (Q-TOF/MS, Agilent, Andover, MA) using a C18 RPODS column (2.1 mm 100 mm; 1.8 μm; Agilent). The mobile phases were composed of water (A) and acetonitrile (B). The gradient was as follows: 0 min, 95% A, 5% B; 2 min, 95% A, 5% B; 13-22 min, 5% A, 95% B. Elution was performed at a solvent flow rate of 0.2 mL/min. The column compartment was kept at 25 C, and the injection volume was 2 μL. ESI-MS was performed using the following conditions: drying gas N2, 10 L/min; temperature, 350 C; pressure of nebulizer, 30 psi; capillary voltage, 4000 and 3500 V in positive ionization mode and negative ionization mode, respectively; scan range, 50-1000 m/z. Data Preprocessing and Partial Least-Squares (PLS). LCMS data were deconvoluted and aligned with mass and retention time tolerances using GeneSpring software (Version 1.1, Agilent, Andover, MA) to generate a matrix containing information on mass, retention time and intensities for all peaks. Prior to multivariate statistical analysis, the data of each chromatogram were normalized to a constant integrated intensity of the number of peaks to partially compensate for the concentration bias of each sample, and then the original variables were pareto-scaled before multivariate analysis.34 The data set produced by HPLC-MS profiling involved a large degree of redundant information, and PLS was performed to extract the most important information. PLS allowed the visualization of an overview of all samples in a score plot, and significant metabolites were statistically recognized in a loading plot. 1181

dx.doi.org/10.1021/pr100960h |J. Proteome Res. 2011, 10, 1179–1190

Journal of Proteome Research

ARTICLE

Table 1. Top 10 Significantly Enriched GO Terms for Genes Differentially Expressed between the Early Phase and Late Phase GO ID

GO term

differentially expressed genes

total genes in GO term

P-value

Pattern 1 genes GO:0006950

response to stress

57

1504

5.81  10-15

GO:0010033

response to organic substance

46

1117

5.46  10-13

GO:0001666

response to hypoxia

16

205

4.93  10-8

GO:0070482

response to oxygen levels

16

220

1.39  10-7

GO:0009719

response to endogenous stimulus

28

675

1.54  10-7

GO:0051716

cellular response to stimulus

30

829

1.02  10-6

GO:0009725

response to hormone stimulus

25

603

1.16  10-6

GO:0006807 GO:0032502

nitrogen compound metabolic process developmental process

45 63

1648 2794

1.35  10-6 1.67  10-6

GO:0060548

negative regulation of cell death

19

377

2.60  10-6

GO:0006629

lipid metabolic process

36

688

1.71  10-12

GO:0032787

monocarboxylic acid metabolic process

23

304

4.80  10-11

GO:0019752

carboxylic acid metabolic process

28

517

6.96  10-10

GO:0043436

oxoacid metabolic process

28

517

6.96  10-10

GO:0006082 GO:0042180

organic acid metabolic process cellular ketone metabolic process

28 28

522 529

8.71  10-10 1.19  10-9

GO:0044281

small molecule metabolic process

44

1334

1.81  10-8

GO:0008202

steroid metabolic process

15

176

7.03  10-8

GO:0008610

lipid biosynthetic process

18

288

3.25  10-7

GO:0008152

metabolic process

100

5478

1.05  10-5

Pattern 2 genes

’ RESULTS Differentially Expressed Genes Associated with Early/Late Phases during LR

Differentially expressed genes associated with early/late phases during LR were identified according to the following three principles: Cy3 and Cy5 fluorescent images were acquired, the genes from 6 and 168 h possessed expression differences and the genes could be found in the Swiss-Prot database. The Cy3/Cy5 ratio classified the expression levels of genes into upregulated (ratio g 2), downregulated (ratio e 0.5), and normally expressed (0.5 < ratio