Article pubs.acs.org/est
Global Liver Proteome Analysis Using iTRAQ Labeling Quantitative Proteomic Technology to Reveal Biomarkers in Mice Exposed to Perfluorooctane Sulfonate (PFOS) Feng Tan,*,† Yihe Jin,† Wei Liu,† Xie Quan,† Jingwen Chen,† and Zhen Liang‡ †
Key Laboratory of Industrial Ecology and Environmental Engineering (MOE), School of Environmental Science and Technology, Dalian University of Technology, Dalian 116024, China ‡ Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China S Supporting Information *
ABSTRACT: Proteomic analysis allows detection of changes of proteins expression in organisms exposed to environmental pollutants, leading to the discovery of biomarkers of exposure and understanding of the action mechanism of toxicity. In the present study, we applied iTRAQ labeling quantitative proteomic technology for global characterization of the liver proteome in mice exposed to perfluorooctane sulfonate (PFOS). This successfully identified and quantified 1038 unique proteins. Seventy-one proteins showed a significant expression change in the treated groups (1.0, 2.5, 5.0 mg/kg of body weight) compared with the control group, and 16 proteins displayed strong dose-dependent changes. Gene ontology analysis showed that these differential proteins were significantly enriched and mainly involved in lipid metabolism, transport, biosynthetic processes, and response to stimulus. We detected significantly increased expression levels of enzymes regulating peroxisomal β-oxidationincluding long-chain acyl-CoA synthetase, acyl-CoA oxidase 1, bifunctional enzyme, and 3-ketoacyl-CoA thiolase A. PFOS also significantly induced cytochrome P450s and glutathione S-transferases that are responsible for the metabolism of xenobiotic compounds. The expressions of several proteins with important biological functions−such as cysteine sulfinic acid decarboxylase, aldehyde dehydrogenase, and apolipoprotein A-I, also correlated with PFOS exposure. Together, the present results provide insight into the molecular mechanism and biomarkers for PFOS-induced effects.
■
INTRODUCTION Perfluorooctane sulfonate (PFOS) is an emerging persistent organic pollutant. Due to its extreme chemical and thermal stabilities and high surface activity, PFOS has been widely used in industry and consumptive products in the past decades, and direct or indirect PFOS emissions during its manufacture, usage, and disposal have resulted in its widespread distribution in the environment.1 It has also been widely detected in wildlife and in the human body, even in cord blood and breast milk.2,3 Recent toxicological research has indicated that PFOS is correlated with multiple toxicities, including hepatotoxicity, carcinogenicity, immunotoxicity, and developmental effects.4 These findings have resulted in increasing worldwide public concern regarding the environmental health risks associated with PFOS. It is a great challenge to evaluate the potential risks of lowdose exposure to environmental pollutants, since the exposure usually does not rapidly result in obvious clinical symptoms, while chemical analysis on its own is not enough to provide useful insight into the health risk posed by the pollutants. On the other hand, it is possible to detect some molecular signatures associated with the exposures in the body before the onset of clinically apparent disease,5 which provides a great © 2012 American Chemical Society
advantage for toxicity assessment and early clinical diagnosis. Proteins are important functional molecules in cellular processes; investigating the protein changes induced by xenobiotics can provide insight into the action mechanism, and enable development of biomarkers for both exposure effect and susceptibility that can be used in health risk assessment.6,7 This can be achieved by analyzing the proteome of exposed animals by proteomic technologies.8 Conventional proteomic analysis heavily relies on twodimensional gel electrophoresis (2-DE), followed by mass spectrometry identification.9−11 Protein profiles in zebrafish embryos and gills of sentinel fish exposed to PFOS have been recently obtained with 2-DE technology by two different research groups.12,13 These studies have revealed a few differential proteins related to PFOS, providing new clues to further understanding the molecular mechanism underlying PFOS-induced effects and leading to the discovery of potential biomarkers associated with the exposure to PFOS. However, 2Received: Revised: Accepted: Published: 12170
August 7, 2012 October 8, 2012 October 9, 2012 October 9, 2012 dx.doi.org/10.1021/es3027715 | Environ. Sci. Technol. 2012, 46, 12170−12177
Environmental Science & Technology
Article
piperazineethanesulfonic acid (HEPES) and 8 M urea at pH 7.4, respectively, and was homogenized with a Dounce homogenizer for 1 min, then sonicated for 30 s. The homogenate was centrifuged under 32 000g at 4 °C for 1 h. The supernatant was removed and transferred into Eppendorf tubes and stored at −80 °C for use. Final protein concentrations measured by BCA kit according to the manufacturer’s protocol were 19−23 mg/mL. A 100-μg portion of the proteins from each group was transferred into one Eppendorf tube, respectively, and denatured, reduced, and blocked at room temperature according to the iTRAQ reagent reference guide. Ten μL of trypsin solution (1 μg/μL) was added to each tube and incubated at 37 °C overnight, respectively. The obtained tryptic peptides from the control and treated groups (1.0, 2.5, and 5.0 mg/kg) were labeled with iTRAQ reagents with 114.1, 115.1, 116.1, 117.1 tags for 1 h at room temperature, respectively. Then the labeled peptides were pooled together. High-pH Reverse-Phase Liquid Chromatography Fractionation. The pooled peptides were cleaned up with SCX cartridges and desalted by C18 SPE cartridges, then reconstituted in 10 mM ammonium formate solution at pH 9.5 for high-pH reverse-phase liquid chromatography (RPLC) separation, using Yilite P2300 HPLC system with an C18 analytical column (250 × 4.6 mm, 5 μm particles), with 0.8 mL/min of flow rate. The mobile phase consisted of the following: A phase, 10 mM ammonium formate solution at pH 9.5; B phase, 10 mM ammonium formate and 90% ACN solution at pH 9.5. The following linear gradient was used: 0−2 min, 2% B; 2−2.5 min, 2 to 5% B; 2.5−58 min, 5 to 40% B; 58.1−62 min, 80% B. A total of twenty fractions were collected from 2 to 60 min at 3-min intervals and combined into 10 fractions by combining fractions of 1 and 11; 2 and 12; 3 and 13, and so on. The combined fractions were dried, lyophilized, and reconstituted in 100 μL of 2% ACN-0.1% formic acid (FA) solution for nanoLC-MS/MS analysis. NanoLC-MS/MS Analysis. NanoLC-MS/MS analysis was carried out on an Agilent 1100 series system coupled with a LTQ-Orbitrap XL mass spectrometer (Thermo-Fisher Scientific). Twenty μL of the samples were loaded onto a C18 analytical column (15 cm × 75 μm × 5 μm, 200 Å) for reverse phase separation. The mobile phase consisted of the following: A phase, 2% ACN-0.1% FA; B phase, 98% ACN-0.1% FA. The following linear gradient was used: 0−10 min, 0% B; 10−12 min, 0 to 5% B; 12−90 min, 5 to 40% B; 90−100 min, 80% B; 100.1−115 min, 0% B, with a 300 nL/min of flow rate. The LTQ-Orbitrap mass spectrometer settings were as follows: spray voltage, 2.1 kV; temperature of ion transfer capillary, 250 °C. Full MS scans were acquired over a mass range of m/z 400−1800 with the Orbitrap mass analyzer at a resolution of 60 000 (m/z 400). Fragment ion spectra obtained by high energy collision dissociation (HCD) were acquired in the Orbitrap mass analyzer with a resolution of 7500 over a mass range of m/z 100−2000. Dynamic exclusion function was as follows: repeat count, 2; repeat duration, 30 s; exclusion duration, 60 s; normalized collision energy, 45%. A full MS scan was followed by ten data-dependent MS/MS scans for the top intense ions with signal threshold above 5000. All MS/MS data were acquired with Xcalibur 2.1 software (Thermo-Fisher Scientific). Experiments were performed in triplicate analyses. Data Analysis. The acquired peak-lists of all MS/MS spectra were searched with Mascot Daemon (version 2.3, Marix Science, London, U.K.) against the International Protein Index
DE has poor separation efficiency for proteins with extreme molecular mass, isoelectric point, and hydrophobic character. Furthermore, the relatively low sensitivity restricts the identification of low abundance proteins.14 iTRAQ labeling is an efficient quantitative proteomic technology that uses a family of isobaric isotope tags to label tryptic peptides from differential protein samples, followed by liquid chromatography-tandem mass spectrometry (LC-MS/ MS) analysis to identify and quantify the proteins. Because this process can simultaneously identify and quantify the proteins from multiple differential samples associated with various exposures,15 it is advantageous for revealing the dose−effect relationship between expression levels of the differential proteins and exposure level of the pollutants, which is an important reference for biomarkers discovery. Previous studies showed that PFOS could activate peroxisome proliferator activated receptor-alpha (PPARα) based on gene expression profiling.16,17 Since PPARα has important roles in many biological processes, the aim of the present study is investigate the proteins (more toxicologically relevant parameters) in liver associated with PFOS by a highthroughput iTRAQ labeling quantitative proteomic technology. The results of this proteomic analysis indicate that PFOS significantly regulated the expression levels of liver proteins in mice. The differentially expressed proteins were mainly involved in lipid metabolism, transport, biosynthetic processes, and response to stimulus.
■
MATERIALS AND METHODS Chemicals. PFOS (Potassium salt, purity >98%) and bicinchonic assay (BCA) kit were purchased from SigmaAldrich (MO, USA). An iTRAQ 4-plex application kit was purchased from Applied Biosystems (CA, USA). HPLC-grade acetonitrile (ACN) was purchased from Fisher (MA, USA). Strong cation exchange (SCX) and C18 solid-phase extraction (SPE) cartridges were purchased from Supelco (PA, USA). Modified trypsin was purchased from Promega (WI, USA). All other chemicals of analytical grade or biochemical grade were purchased from Sigma-Aldrich (MO, USA). Animal Administration. Male Kunming mice at 3 weeks of age (mean body weight (BW) of ∼32 g) were purchased from the experimental animal care center of Dalian Medical University. They were housed in plastic cages for natural diet at controlled conditions: room temperature of 23 ± 2 °C, relative humidity of 45 ± 10%, and a 12 h light-dark cycle. After acclamation for one week, 20 mice were randomly distributed into one control group (control) and three treated groups (treated 1, treated 2, and treated 3), each group containing five mice. The three treated groups received an intraperitoneal injection of PFOS water solution (0.2 mL/10 g of BW) with the doses of 1.0, 2.5, and 5.0 mg/kg of BW at every day with the help of a bulb-tipped gastric gavage needle at consecutive seven days, respectively. The control group received deionized water by the same way. Animals were sacrificed by neck dislocation 24 h after the last treatment. Liver tissues were quickly harvested and weighted, then stored in liquid nitrogen. BWs of all mice at the treatment before and after were measured. Protein Extraction, Digestion, and iTRAQ-Labeling. To obtain a representative proteome, 0.2 g of pieces were dissected from every liver in each group and pooled. One g of the pooled liver pieces from each group were blended in 5 mL of the extraction buffer consisting of 20 mM 4-(2-hydroxyethyl)-112171
dx.doi.org/10.1021/es3027715 | Environ. Sci. Technol. 2012, 46, 12170−12177
Environmental Science & Technology
Article
Table 1. Alteration of BW, Coefficient (coef.) of Liver Weight to BW, and Concentration of PFOS in Serum (Mean Value ± SE) body weight (g, n = 5) group control treated 1 treated 2 treated 3 a
treatment before 32.18 32.34 32.00 31.36
± ± ± ±
1.02 1.10 0.63 0.81
treatment after 34.28 34.94 35.20 32.08
± ± ± ±
coef. of liver weight to BW (n = 5)
0.82 1.69 1.11 2.75
0.055 0.060 0.064 0.071
± ± ± ±
0.0008 0.0020 0.0028a 0.0017a
concn of PFOS in serum (mg/L, n = 3−5) 19 1450 6244 16018
± ± ± ±
1.9 218.1a 478.1a 2116.8a
p < 0.05 relative to the control group.
identified by more than five peptides. The use of a highresolution Orbitrap mass analyzer and low FDR enabled confidence in the identification of proteins with one peptide match. Out of the 1502 unique proteins identified, 1038 were simultaneously quantified at all four labeling channels corresponding to the control and three treated groups (SI Table S1). With a few exceptions, fold changes of the quantified proteins in the treated groups compared to the control group were tightly clustered near a one-fold change (Figure 1), representing no obvious change in protein expression levels between the control and treated groups for most of the liver proteins identified.
(IPI) mouse database (v.3.87), which contains 59 534 sequences, and reversed sequences were appended to the database for estimating false discovery rate (FDR). Search parameters included the following: two missed cleavages, fixed modifications of cysteines carbamidomethylation and iTRAQ 4-plex, variable modification of methionine oxidation, mass tolerance of 20 ppm for precursors and 0.5 Da for fragment ions, and a 95% confidence interval (p < 0.05), the Mascot score for each identification to be at least 42. Protein quantification settings were as follows: protein ratio type of “weighted”; normalization of “average ratio”, outlier removal of “automatic”, peptide threshold of “at least homology”, a 95% confidence interval (p < 0.05) to assess accuracy of protein ratio, and a minimum of 3 spectra for quantifying a protein. Automatic isotope correction was carried out using the values supplied with the iTRAQ reagent certification. iTRAQ ratios were normalized to the control group. Final ratios of the differential proteins were reported with the mean values of multiple quantifications. Analysis of PFOS Concentrations in Serum. The blood samples were analyzed as serum after centrifugation and the serum was extracted by IPE method described by Hansen et al.18 Details are presented in the Supporting Information (SI).
■
RESULTS AND DISCUSSION There were no significant differences in BW among the control and treated groups at the beginning and ending of the experiment (p > 0.05). However, the treated 3 with the high dose of PFOS (5.0 mg/kg of BW) showed a slower increase of BW compared with other groups (Table 1). We observed that there was slight liver hypertrophy for the treated 3, which were proved by the significant increase of coefficient of liver weight to BW in the treated 2 and 3 compared with the control (p < 0.05). Concentration of PFOS in serum of all mice was determined by HPLC-MS/MS. There was a small background value for the control, and a significant dose-dependent increase for the three treated groups. The detected PFOS concentrations in serum could be comparable with those reported by Austin et al. in a previous publication.19 Identification and Quantification Results of Liver Proteome. We used high-throughput iTRAQ labeling quantitative proteomic technology to analyze protein samples from mice of the control and three treated groups, and to characterize the expression changes of liver proteins associated with PFOS exposure in these samples. The acquired MS/MS spectra were searched against the IPI mouse reference database using the Mascot search engine. Totals of 1268, 1252, and 1219 proteins were identified in triplicate analyses, respectively. After the redundancies were removed, 1502 unique proteins were identified at more than a 95% confidence level, with a FDR less than 0.4%. More than 70% of the proteins were identified by at least two distinct peptides, and ∼22% of the proteins were
Figure 1. Fold changes of the protein expression levels in the treated groups with 1.0 mg/kg (blue), 2.5 mg/kg (black), and 5.0 mg/kg (red) doses compared with the control group.
Using a general criterion for differentially expressed proteins in iTRAQ labeling quantification consider fold changes of >1.5 or
