Serum proteomic analysis reveals potential serum biomarkers for occupational medicamentosa-like dermatitis caused by trichloroethylene

Toxicology Letters 229 (2014) 101–110

Contents lists available at ScienceDirect

Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet

Serum proteomic analysis reveals potential serum biomarkers for occupational medicamentosa-like dermatitis caused by trichloroethylene Peiwu Huang a,d,1, Xiaohu Ren a,1, Zhijun Huang c,1, Xifei Yang a , Wenxu Hong a , Yanfang Zhang b , Hang Zhang a,d , Wei Liu a , Haiyan Huang a , Xinfeng Huang a , Desheng Wu a , Linqing Yang a , Haiyan Tang b , Li Zhou a , Xuan Li a , Jianjun Liu a, * a Key Laboratory of Modern Toxicology of Shenzhen, Medical Key Laboratory of Guangdong Province, Medical Key Laboratory of Health Toxicology of Shenzhen, Shenzhen Center for Disease Control and Prevention, No. 8, Longyuan Road, Nanshan District, Shenzhen 518055, China b Shenzhen Prevention and Treatment Center for Occupational Disease, Shenzhen 518001, China c The Emergency Department, Second Clinical Medical College (Shenzhen People’s Hospital), Jinan University, Shenzhen 518020, China d College of Life Science, Shenzhen University, Shenzhen 518060, China

H I G H L I G H T S

   

Identify 8 proteins differentially expressed between OMLDT patients and controls. The altered expressions were validated by Western blot and ELISA analysis. TTR and PBP4 were significantly down-regulated in OMLDT patients. Haptoglobin was up-regulated in OMLDT patients.

A R T I C L E I N F O

A B S T R A C T

Article history: Received 4 April 2014 Received in revised form 29 May 2014 Accepted 29 May 2014 Available online 21 June 2014

Trichloroethylene (TCE) is an industrial solvent with widespread occupational exposure and also a major environmental contaminant. Occupational medicamentosa-like dermatitis induced by trichloroethylene (OMLDT) is an autoimmune disease and it has become one major hazard in China. In this study, sera from 3 healthy controls and 3 OMLDT patients at different disease stages were used for a screening study by 2D-DIGE and MALDI-TOF-MS/MS. Eight proteins including transthyretin (TTR), retinol binding protein 4 (RBP4), haptoglobin, clusterin, serum amyloid A protein (SAA), apolipoprotein A-I, apolipoprotein C-III and apolipoprotein C-II were found to be significantly altered among the healthy, acute-stage, healingstage and healed-stage groups. Specifically, the altered expression of TTR, RBP4 and haptoglobin were further validated by Western blot analysis and ELISA. Our data not only suggested that TTR, RBP4 and haptoglobin could serve as potential serum biomarkers of OMLDT, but also indicated that measurement of TTR, RBP4 and haptoglobin or their combination could help aid in the diagnosis, monitoring the progression and therapy of the disease. ã 2014 Published by Elsevier Ireland Ltd.

Keywords: Occupational medicamentosa-like dermatitis induced by trichloroethylene (OMLDT) Serum biomarkers 2D-DIGE MALDI-TOF-MS/MS

Abbreviations: TCE, trichloroethylene; OMLDT, occupational medicamentosa-like dermatitis induced by trichloroethylene; TTR, transthyretin; RBP4, retinol binding protein 4; Hp, haptoglobin; SAA, serum amyloid A protein; CLU, clusterin; Apo, apolipoprotein; 2D-DIGE, two dimension difference gel electrophoresis; MALDI-TOF, matrixassisted laser desorption ionization time of flight tandem; MS/MS, mass spectrometry; ELISA, enzyme linked immunosorbent assay; DTT, DL-dithiothreitol; CHAPS, 3-[3cholamidopropyl]-dimethylammonio-1-propanesulfenate; Tris, trashydroxymethyl amino methane; IPG, immobilized pH gradient; IEF, isoelectric focusing; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; CHCA, a-cyano-4-hydroxycinnamic acid; PMF, peptide mass fingerprint; LIFT, laser-induced forward transfer; PFF, peptide fragment fingerprint; ROC, receiver operating characteristic; AUC, area under the curve; TP, total protein; APPs, acute phase proteins; TBIL, total bilirubin; ALT, alanine transaminase; AST, aspartate transaminase; GGT, gamma glutamyl transpeptidase; ALP, alkaline phosphatase. * Corresponding author. Fax: +86 755 25508584. E-mail address: [email protected] (J. Liu). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.toxlet.2014.05.024 0378-4274/ ã 2014 Published by Elsevier Ireland Ltd.

102

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

1. Introduction Trichloroethylene (TCE) is a volatile organic solvent that has been widely used since the early 1900s in industry as metal degreaser, industrial intermediates, dry cleaning and food processing agent, and even in medicine as an anesthetic drug. It is estimated that more than 400,000 workers are exposed to TCE annunally in the U.S., where approximately 50 million pounds of TCE are released annually into the environment (NIOSH, 1994; Goldman, 2010). Many studies have demonstrated that TCE exposure can cause severe toxic effect to the central nervous system, liver, kidney, immune system, reproductive system, etc. In 2011, the U.S. Environmental Protection Agency (EPA) released a final health assessment for TCE, characterizing it as a human carcinogen and non-cancer health hazard. TCE occupational exposure can lead to systemic skin disorders, named “occupational medicamentasa-like dermatitis by trichloroethylene” (OMLDT) accompanying with liver dysfunction (Xu et al., 2009; Chiu et al., 2013). Occurrences of OMLDT have been reported from USA, Japan, Spain, Singapore, China, Korea, Thailand, and the Philippines, while the number of OMLDT cases has increased since the mid-1990s in China (Kamijima et al., 2007), and 394 cases have been reported as of 2009 in Guangdong province, where an estimation of at least 20,000 new workers were exposed to TCE every year (Huang and Huang, 2010). Extensive researches have been carried out to study OMLDT in recent years. However, the studies mainly focus on retrospective epidemiological investigations (Kamijima et al., 2008), genetic polymorphisms associated susceptibility (Li et al., 2006), and immunological mechanisms (Dai et al., 2005). Very few studies have focused on valid population studies, the dose–response relationship, period of exposure before disease onset, etiology and recurrence of OMLDT still remains unclear (Nakajima et al., 2003). OMLDT is often misdiagnosed, and proper treatment is delayed due to unavailable biological diagnostic methods. OMLDT has become an ongoing health problem because of TCE exposure in China (Xu et al., 2009). Serum is an easily accessible body fluid that contains different types of proteins released by various diseased tissues. Proteome analysis of the serum provides insight into disease pathophysiology and mechanisms and helps to identify potential biomarkers with diagnostic and prognostic significance (Ray et al., 2011). In our previous study (Liu et al., 2009), we used serological proteomic analysis (SERPA) to screen and identify auto-antigens involved in TCE-induced autoimmune diseases. Six proteins were screened, and non-metastasis (NM23) protein was found to be a potential toxicological biomarker in TCE-induced autoimmune diseases. In this study, we compared the expression of proteins in the sera of OMLDT patients at three different stages using 2DDIGE and matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF-MS) strategy. Eight proteins were significantly altered among the healthy, acute-stage, healing-stage and healed-stage groups. The proteins TTR, RBP4 and haptoglobin were further validated by Western blot and ELISA. This study can provide potential serum biomarkers that may be used for the diagnosis and monitoring the progression of the disease. 2. Materials and methods 2.1. Subjects and serum collection This study was conducted in accordance with the principles of the Declaration of Helsinki (World Medical Association, 1997). Informed consent was obtained from all subjects, and the Medical Ethics Committee of Shenzhen Center for Disease Control and

Table 1 Age and TCE exposure duration of OMLDT patients. Patient no.

Age (year)

TCE exposure duration (day)

1a 2a 3a 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Average

18 16 19 26 18 20 39 27 21 31 22 39 19 17 17 43 14 25 23.94  8.70

14 21 36 51 46 25 32 30 32 20 37 44 25 29 42 13 31 34 31.22

a Patients whose serum specimens were included in the 2D-DIGE study and western validation experiments.

Prevention approved the study protocol. The skin manifestations of the OMLDT patients enrolled in the study and the details for serum collection were described in our previous study (Hong et al., 2013). For 2D-DIGE analysis and Western blot validation experiments, 3 male cases with typical symptoms were enrolled (Tables 1 and 2, patients 1–3). For the ELISA, the 3 cases used in the 2D-DIGE study and an additional 15 cases (10 males and 5 females; 23.9  8.7 years of age) were included (Tables 1 and 2, patients 4–18). The patients were hospitalized for treatment between March 2010 and November 2011. Eighteen age- and sex-matched normal subjects were also selected as the controls. 2.2. Depletion of high abundance serum proteins The serum samples were processed using the ProteoExtract Albumin/IgG removal kit according to the manufacturer’s instructions. Briefly, 40 mL of each serum was depleted by the albumin/IgG affinity resin columns, and the flow-through fraction was concentrated and desalted by ultrafiltration using a 3 kDa cut off centrifugal filter device (Millipore, USA). For the final centrifugation step, the binding buffer in the samples was displaced by 1 sample buffer, and the protein concentration of all depleted samples was determined with the 2-D Quant kit (GE Healthcare, USA). 2.3. CyDye minimal labeling The depleted serum samples were minimally labeled (25 mg protein per 200 pmol dyes) with Cy2, Cy3 or Cy5 fluorescent dyes according to the manufacturer’s instructions (GE healthcare). One half of the twelve test samples were labeled with Cy3 or Cy5, respectively. Cy2 was used to label an internal standard sample generated by pooling an aliquot of all twelve test samples. Equal portions (25 mg) of the internal standard were run on all six DIGE gels to assess the reproducibility and minimize the gel-to-gel variation. Three differently labeled samples were mixed in one gel, and an equal volume of 2 sample buffer (7 M urea, 2 M thiourea, 4% w/v CHAPS, 2% w/v DTT, and 2% v/v IPG buffer pH 3–10) was added to each mixed sample and incubated on ice for 10 min prior to bringing the total sample volume to 450 mL with rehydration buffer (8 M urea, 2% w/v CHAPS, 0.28% w/v DTT, 0.5% v/v IPG buffer pH 3–10, and 0.002% w/v bromophenol blue).

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

103

Table 2 Liver function test results of OMLDT patients. Patient no.

TP (g/L)

TBIL (mmol/L)

ALT (U/L)

AST (U/L)

GGT (U/L)

ALP (U/L)

1a 2a 3a 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Healthy range

52.5b /61.4c /62.9d 47.9/62.6/78.1 54.8/61.5/71.7 50b/60.1c/64.7d 48.9/62.4/74.2 45.8/61.0/57.5 43.2/66.2/72.1 43.4/56.4/55.7 49.9/56.5/59.3 51.3/57.9/66.1 46.1/56.8/70.0 54.2/58.7/67.7 49.6/60.2/67.5 52/53.1/61.6 54.6/66.4/64.8 48.8/55.7/70.7 40.6/56.3/69.4 47.2/58.0/65.7 62–87

84.2/29/8.7 176.2/66/7.1 23.3/6.8/5.5 253.1/36.3/10.6 182.0/49.7/11.6 71.8/19.1/3.7 268.2/73.8/8.2 126.4/30.6/8.5 49.0/17.4/9.1 37.3/18.7/5.9 251.5/76.9/7.3 68/24.5/8.1 141.2/49.1/12.3 61.9/30.4/15.8 88.2/29.6/7.9 40.4/20.7/9.5 218.6/57.3/6.3 217.8/78.1/7.3 1–26

5985/173/22 3067/678/11 1000/365/9 664/88/22 1480/347/18 2610/581/24 897/140/41 945/237/26 1504/415/27 849/277/11 678/213/15 619/189/19 521/176/37 1710/272/29 1396/413/8 1930/500/18 1206/257/22 738/219/41 0–40

5586/97/22 1678/328/19 748/251/19 291/43/15 1270/139/21 1530/133/17 221/71/24 353/47/16 286/112/21 437/119/17 228/83/14 424/101/18 237/70/18 1863/113/48 660/173/13 780/145/15 405/83/20 203/84/18 0–45

263/93/36 486/248/6 104/68/11 275/126/30 340/100/15 187/110/26 1565/569/33 297/131/27 173/163/22 450/127/22 279/78/39 280/107/10 276/158/17 166/123/39 178/112/13 737/217/49 623/202/86 1596/452/85 0–49

238/125/61 148/91/45 152/130/83 197/102/52 230/79/35 130/89/63 386/158/50 174/65/46 140/120/100 294/157/50 287/140/43 552/227/99 211/149/67 368/190/95 503/273/78 252/139/67 282/177/119 182/137/62 53–128

TP: total protein; TBIL: total bilirubin; ALT: alanine transaminase; AST: aspartate transaminase; GGT: gamma glutamyl transpeptidase; ALP: alkaline phosphatase. a Patients whose serum specimens were included in the 2D-DIGE study and western validation experiments. b Parameters in acute stage. c Parameters in healing stage. d Parameters in healed stage.

2.4. Two-dimensional gel electrophoresis For 2-DE, six pooled samples were subjected to isoelectric focusing (IEF) with individual immobilized pH gradient (IPG) strips (pH 3–10, 24 cm, GE Healthcare) on a 24 cm strip holder (GE Healthcare) using an Ettan IPGphor 3 isoelectric focusing system (GE Healthcare). The IEF was programmed in the following five steps: 30 V for 12 h, 500 V for 1 h, 1000 V for 1 h, 8000 V gradient for 5 h and then the unit was maintained at 8000 V for a total of 90,000 volt-hours (V h). After IEF, the strips went through two step equilibration before applying on 12.5% SDS-PAGE gels. Second dimension separation was conducted on an Ettan DALT six electrophoresis system (GE Healthcare) at 15  C at 1 W per gel for 1 h, followed by 11 W per gel for 6 h. 2.5. Image acquisition and analysis After the 2-DE was finished, the gels were immediately scanned using a Typhoon Trio Variable Mode Imager (GE Healthcare) at excitation/emission wavelengths of 488(blue)/520, 532(green)/ 580 and 633(red)/670 nm for Cy2, Cy3 and Cy5, respectively. The resulted 18 maps were processed by DeCyder 2D v6.5 (GE Healthcare) for differential analysis. Protein spots were detected and manually landmarked to the master gel to improve matching quality before the automatic matching. For protein expression comparison, the spot volume of the Cy3 and Cy5 samples was normalized to the corresponding spot volume of the Cy2 internal standard samples from the same gel. The Student’s t-test and oneway ANOVA were used to calculate the statistical significance. The spots considered statistically significant (p  0.05) with an average ratio  1.5 were marked for further evaluation. 2.6. Spot picking and in-gel enzymolysis The significantly different protein spots were manually excised from 2-DE gel stained by Coomassie Brilliant Blue G-250 staining solution. Gel plugs were destained with 50% acetonitrile in 25 mM ammonium bicarbonate followed by dehydration in 100% acetonitrile. Each gel piece was digested overnight at 37  C with 0.03 mg sequencing-grade trypsin (Promega, USA) in 15 mL 25 mM

ammonium bicarbonate buffer. The extracted peptides were stored at 80  C prior to protein identification. 2.7. MALDI-TOF-MS/MS analysis and database searching Protein identification was performed with an UltrafleXtreme MALDI-TOF/TOF mass spectrometer (Bruker Daltonics Inc., USA). Briefly, 2 mL extracted peptide mixtures were spotted onto a polished steel MTP 384 target plate (Bruker Daltonics Inc.), air dried and covered with 1 mL CHCA matrix (5 mg/mL in 50% CAN and 0.1% TFA). External calibration was performed using standard peptide calibration mixtures covering peaks between 700 and 4000 Da (Bruker Daltonics Inc.). The peptide mass fingerprint (PMF) spectra were acquired in the positive reflection mode and 7 strongest peptides per spot were selected automatically for MS/MS analysis in LIFT mode. Protein identification by PMF and MS/MS spectra was performed using the MASCOT search engine 2.2 (Matrix Science, USA) and BioTools software (Bruker Daltonics Inc.). Confident matches in SwissProt database were defined by the MASCOT score, the sequence coverage by matching peptides and statistical significance (p < 0.05). 2.8. Western blot analysis Three chosen proteins underwent Western blotting to validate the 2D-DIGE results. Equal volume of crude serum samples were separated on 10% polyacrylamide gels with 5% stacking gels and transferred onto PVDF membranes (Millipore, MA) in a semi-dry transfer unit (GE Healthcare, USA). Anti-RBP4 (diluted 1:5000, Abcam), anti-TTR/prealbumin and anti-haptoglobin (diluted 1:1000, Santa Cruz) antibodies were used as primary antibodies. Horseradish peroxidase–conjugated antibody (diluted 1:3000, Thermo Scientific, USA) was used as secondary antibody. Chemiluminescence signal was measured using ECL substrate reagents (Thermo Scientific) and Image Quant RT EC (GE Healthcare). 2.9. Enzyme-linked immunosorbent assays To confirm the differential expression of the identified proteins, the concentrations of the corresponding proteins in each serum

104

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

sample were quantified using commercially available ELISA kits for human haptoglobin (Abcam, England), human transthyretin (Uscn Life Science, China) and human RBP4 (EMD Millipore, USA) according to the manufacturer’s instructions. Briefly, equal amounts of crude serum samples of acute-stage, healing-stage and healed-stage from 18 OMLDT patients and 18 controls were used for the ELISAs. The standards and all samples were assayed in duplicate. The serum protein concentrations were calculated using the kit-specific standard curves generated by the CurveExpert 1.4 software. The statistical analysis was performed using the GraphPad Prism software. p < 0.05 was regarded as statistically significant. Receiver operating characteristic (ROC) curve analysis was generated in SPSS Statistics. 3. Results 3.1. Clinical characteristics of the OMLDT patients The mean age of the 18 patients was 23.9  8.7 years, and the duration of TCE exposure prior to disease occurrence was 31.2  10.3 days (Table 1). The serum levels of total protein (TP) were significantly decreased (p < 0.05) in the acute-stage (48.9  4.1 g/L) compared to the healed-stage (66.9  6.0 g/L), whereas the serum levels of TBIL, ALT, AST, GGT, ALP were significantly increased (p < 0.05) in the acute-stage (131.1 84.0 mmol/L, 1544  1309 U/L, 956  1275 U/L, 460  439 U/L and 263  121 U/L, respectively) compared to the healed-stage (8.5  2.8 U/L, 22  10 U/L, 20  8 U/L, 33  23 U/L and 68  23 U/L, respectively), see Table 2. 3.2. Efficiency evaluation for serum depletion One of the major challenges in serum proteomic analysis is the high abundance of proteins that obscure other proteins and cause a loss of resolution in the 2-DE gels. Albumin and IgG are the two most abundant serum proteins, comprising 50–70% and 10–20% of the total serum proteins, respectively. To obtain higher sensitivity and to visualize low abundance proteins, the ProteoExtract1 Albumin/IgG removal kit was used to specifically and simultaneously deplete the serum albumin and IgG from the serum samples. Densitometric analysis of the stained protein bands showed that more than 80% of the serum albumin and IgG (heavy chain and light chain) were removed from control serum sample and OMLDT serum sample. After depletion of these two highabundance proteins, the intensity of other protein bands was increased (Fig. 1), this procedure increased the resolution of 2-DE (Supplementary Fig. 1). 3.3. Quantitative analysis and identification of differential protein spots After image acquisition (Fig. 2), comparative analysis of the protein profiles was performed between the four test groups using the DeCyder software version 6.5. A total of 20 protein spots were statistically significant (p  0.05) among the four groups, 15 spots were down-regulated and 5 spots were up-regulated in the acutestage compared to the healing-stage, healed stage or control (Table 3, Fig. 3). The spots were trypsin-digested, MALDI-TOF/TOF mass spectrometer was applied to obtain high resolution peptide mass fingerprint and peptide fragment fingerprint (see Supplementary Figs. 2 and 3). Proteins were identified by searching the SwissProt database. Of the down-regulated spots, 2 spots were identified as clusterin (CLU), four were identified as apolipoprotein A-I (Apo-AI), three were retinol binding protein 4 (RBP4), four were transthyretin (TTR, Fig. 4), one was apolipoprotein C-III (Apo-CIII) and one was apolipoprotein C-II (Apo-CII), see Table 3. Of the

Fig. 1. The efficiency evaluation for removal of albumin and immunoglobulin G from serum samples. 30 mg of crude serum, albumin and IgG depleted serum fractions from healthy control and OMLDT patients were loaded and resolved by SDS-PAGE followed by stained with colloidal coomassie blue G-250. C: crude serum and D: depleted serum.

up-regulated spots, three were haptoglobin (Hp) and two were serum amyloid A protein (SAA). 3.4. Confirmation of differential protein expression Specific antibodies against TTR, RBP4 and Hp were used to detect their abundance in each test group using crude serum samples from the patients and controls enrolled in the initial 2DDIGE study. western blot analysis of TTR and RBP4 showed that the proteins were down-regulated in the three acute-stage samples compared to the healing-stage and healed-stage samples, the expression levels of which were similar to the control samples (Fig. 5A and B, left panel). The band intensity of Hp was increased in the acute-stage and decreased in the healing- and healed-stage, but remained higher than the control (Fig. 5C). These results were consistent with the 2D-DIGE data (Table 3, Fig. 4). The ELISA results also proved that the TTR level was significantly down-regulated (p < 0.05) in the serum samples from the acute-stage patients, and increased in the healing-stage to healed-stage patients (p < 0.05). No significant differences were observed between the healed-stage samples and the healthy control samples (Fig. 5A, right panel). The changes in the RBP4 level (Fig. 5B, right panel) were similar to TTR, whereas the Hp level had an opposite trend (Fig. 5C, right panel). These data further confirmed the results of the 2D-DIGE analysis. To determine the potential impact of TTR, RBP4 and Hp as markers to discriminate acute stage from healing stage of OMLDT, the ELISA results for these three proteins were used to generate the ROC curves. Fig. 6 reveals that area under the curve (AUC) values for TTR, RBP4 and Hp were 0.864, 0.830 and 0.821, respectively. 4. Discussion TCE occupational exposure can lead to OMLDT, which is usually accompanied with liver dysfunction. In our study, the liver function tests showed that all the 18 enrolled patients had high levels of TBIL, ALT, AST, GGT, and ALP. These proteins are sensitive

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

105

Fig. 2. Representative 2D-DIGE images comparing the serum proteome of healthy subjects and different stages of OMLDT patients. Acute-stage of OMLDT and control samples were labeled with Cy3 and Cy5, respectively, while the protein reference pool (internal standard) was labeled with Cy2.

Table 3 The twenty differential protein spots found to be statistically significant among different stages of OMLDT and normal controls by 2D-DIGE analysis and identified by MALDITOF-MS/MS. Master no. 902 922 1130 1133 1136 1168 1200 1201 1202 1277 1282 1284 1292 1314 1319 1325 1349 1351 1354 1370

Protein namea Clusterin Clusterin Apolipoprotein A-I Apolipoprotein A-I Apolipoprotein A-I Apolipoprotein A-I Retinol-binding protein 4 Retinol-binding protein 4 Retinol-binding protein 4 Transthyretin Transthyretin Transthyretin Transthyretin Haptoglobin Haptoglobin Haptoglobin Serum amyloid A protein Serum amyloid A protein Apolipoprotein C-III Apolipoprotein C-II

Accession IDb CLUS_HUMAN CLUS_HUMAN APOA1_HUMAN APOA1_HUMAN APOA1_HUMAN APOA1_HUMAN RET4_HUMAN RET4_HUMAN RET4_HUMAN TTHY_HUMAN TTHY_HUMAN TTHY_HUMAN TTHY_HUMAN HPT_HUMAN HPT_HUMAN HPT_HUMAN SAA_HUMAN SAA_HUMAN APOC3_HUMAN APOC2_HUMAN

MWc 53031 53031 30759 30759 30759 30759 23337 23337 23337 15991 15991 15991 15991 45861 45861 45861 13581 13581 10846 11277

PId 5.9 5.9 5.5 5.5 5.5 5.5 5.7 5.7 5.7 5.4 5.4 5.4 5.4 6.1 6.1 6.1 6.4 6.4 5.1 4.5

Cov.e 9 11 19 16 17 20 21 20 15 44 9 25 9 12 7 7 11 9 19 19

Mascot scoref 246 336 345 250 227 312 148 207 116 452 148 340 125 272 108 135 73 68 203 113

+: upregulation; –: downregulation. a Identified protein was named referring to each matched protein in the SwissProt database. b Accession ID was recorded as a reference for identification in the SwissProt database. c The theoretical molecular weight of matched protein in Da. d The theoretical isoelectric point of matched protein. e The percentage of identified sequence to the complete sequence of the known protein. f Spots identified by MS/MS analysis, the MASCOT score is indicated. g One-way analysis of variance among all test groups. h Independent tests between two groups. i Average ratio calculated between two groups after normalization. j Parameters of control against acute stage. k Parameters of healing stage against acute stage. l Parameters of healed stage against acute stage.

1-ANOVAg 0.0031 0.0058 0.009 0.04 0.01 0.043 0.00015 7.50E-06 0.0017 0.00011 0.00045 6.40E-05 1.30E-05 0.042 0.06 – 0.044 0.011 0.015 0.038

t-testh

Av. ratioi j

k

l

0.0043 /0.0033 /0.025 0.0052/0.018/0.0098 0.0058/0.06/0.015 0.0034/0.021/0.039 0.0024/0.037/0.018 0.017/–/0.043 0.00041/0.00048/0.0037 0.00014/0.00028/0.0006 0.0042/0.0063/0.0026 0.0003/0.00035/0.0029 0.0011/0.0015/0.004 0.002/0.00089/0.0023 0.0031/0.00014/0.0004 0.032/–/0.042 0.031/0.049/– 0.033/–/– 0.023/–/0.038 0.042/–/0.036 –/0.012/0.046 0.83/0.049/0.062

2.12j/3.09k/2.45l 2.12/2.48/2.73 2.55/1.93/2.26 2.56/2.13/2.49 2.32/2.06/2.07 2.91/2.34/2.53 3.07/3.51/3.47 2.54/2.91/3.02 1.62/1.53/1.72 3.85/3.93/3.94 2.34/3.19/3.62 3.63/4.07/4.17 3.9/5.16/5.22 2.38/ 1.16/ 1.13 2.5/ 1.78/ 1.87 2.29/ 1.2/ 1.26 44.75/ 3.25/ 3.07 18.06/ 3.73/ 3.62 1.31/5.36/4.92 1.07/2.42/2.51

106

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

Fig. 3. Differentially expressed protein spots identified by MALDI-TOF/TOF were marked on the DIGE gels.

indicators of liver injury and they are widely used in clinic to estimate liver function. However, their blood levels are elevated in nearly all kinds of liver injuries, thus, they are not good markers for specific disease, but rather serve as auxiliary indexes in diagnosis. Proteomic techniques to screen for disease-specific biomarkers are powerful tools for evaluating disease prognosis and for gaining deep insight into disease pathogenesis. Serum is an attractive clinical sample because it is easy to obtain and conveys information regarding the physiological and pathological state of the organism because blood circulates through every organ and tissue of the body (Anderson and Anderson, 2002). Previously, we performed a magnetic bead-based weak cation exchange chromatography (MB-WCX) for serum processing and MALDI-TOF-MS for peptide profiling to build a diagnostic model with selected peptide peaks that can discriminate the OMLDT patients from the healthy controls. Two differential peptide peaks were identified as fragments of ABCA12 and PRSS1, which may be potential biomarker of OMLDT (Hong et al., 2013). However, this strategy primarily focused on serum peptides with low molecular weight, which were difficult to identify and validate. Moreover, this technique requires expensive equipment and a well-trained technician to perform the procedures. To gain insight into the dynamic changes of large molecular weight serum proteins in OMLDT patients and to identify serum biomarkers that have more direct clinical applications, we quantified the protein levels of serum proteins in OMLDT patients at different disease stages and compared them to a control group using 2D-DIGE and MALDI-TOF-MS/MS. Twenty differentially expressed protein spots were screened and identified as 8 different proteins. These proteins were involved in a variety of biological processes, e.g., inflammatory response, transportation of thyroid hormones, retinol transport, hemoglobin transport, oxidative

stress, complement regulation, apoptosis and lipid transport. The relative downregulation of TTR, RBP4 and upregulation of Hp in the serum of the acute stage of OMLDT cases were validated by Western blot and ELISA. They were chosen for validation because of the following reasons: relatively low t-test, several protein spots were identified as the same protein, their representative functions in elucidating the pathogenesis of OMLDT and available antibodies and ELISA kits. TTR, also known as prealbumin, is a tetrameric unglycosylated plasma protein involved in the transport of both thyroid hormones and retinol through the mediation of RBP (Palha, 2002). The hepatic synthesis of TTR is exquisitely sensitive to nutritional level and energy intake, as well as inflammation (Myron Johnson et al., 2007). Wang and Burke showed that the mRNA levels of TTR were decreased after HepG2 cells were treated with IL-1b, IL-6 or TNF-a (Wang and Burke, 2007). Correspondingly, the serum levels of IL-1b, IL-6, IL-8 and TNF-a were significantly higher in the OMLDT patients than TCE exposed workers and non-exposed controls (Jia et al., 2012), suggesting that TCE may trigger an immune response, which results in OMLDT and the downregulation of TTR (Figs. 4 and 5A). During hospitalization, the TTR levels can alert the physician to estimating nutritional status, improve patient outcome, and shorten hospitalization (Beck and Rosenthal, 2002). In our study, the OMLDT patients showed a decreased level of total protein in the sera during the acute stage (Table 2), suggesting malnutrition due to liver damage, which may also contribute to the significantly down-regulated TTR levels. These findings indicate that serum TTR may be a potential biomarker to diagnose patients and evaluate the outcomes and recovery of OMLDT patients together with other clinical indices, thereby shortening hospitalization. The TTR concentration is also an important determinant of serum RBP4, which is an adipocytokine that is potentially

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

107

Fig. 4. Representative 2D-DIGE images of differentially expressed proteins by the Decyder software. The expression level of TTR among control sample and different stages of the same OMLDT patients were visualized by protein abundance maps (top panel), 2-DE images (bottom top panel) and three-dimensional spot images (bottom bottom panel).

associated with insulin resistance and fibrosis (Toyama et al., 2011). RBP4 interacts and exists as a complex with TTR, and the RBP4-TTR complex physically prevents glomerular filtration while maintaining RBP4 in the serum (Pullakhandam et al., 2012). In our study, both the TTR and the RBP4 levels behaved similarly, and their serum abundance was significantly decreased in acute-stage OMLDT patients, and rose to normal levels after healing. One possible explanation is that TCE induced high levels of cytokines (Jia et al., 2012), which down-regulated the expression of TTR. Low levels of TTR in the serum indicate that RBP4 is insufficiently bound and lost via filtration in the kidney (Naylor and Newcomer, 1999). However, this hypothesis must be tested with a rigorous experimental design. Hp, primarily synthesized in the liver, is responsible for transporting hemoglobin to the liver for metabolism. As a positive member of the acute phase proteins (APPs), Hp exerts antioxidant properties (Tseng et al., 2004). It also plays an important role in damaged tissue repair and maintaining homeostasis (Pullakhandam et al., 2012). The Hp levels were up-regulated in the acutestage OMLDT patients who suffered severe liver damage and returned to normal level after healing, indicating that Hp may play a role in repairing damaged liver cells. The expression of Hp is induced by the IL-6-like cytokine family (Moshage, 1997). As described above, the serum IL-6 levels were significantly higher in the OMLDT patients (Jia et al., 2012), which may explain the observed high levels of Hp in the acute-stage cohorts. Hp also has

anti-inflammatory effects in autoimmune diseases (Galicia et al., 2009), and may modulate the TCE induced autoimmune response (Cooper et al., 2010; Griffin et al., 2000; Liu et al., 2009). In conclusion, serum Hp may be a biomarker for OMLDT. SAA and CLU are also liver-originated APPs. SAA occurs at low levels in the serum of healthy individuals, and its expression levels are increased by cytokines under stress conditions (Son et al., 2004). Studies have shown that IL-6 increases the plasma concentration of SAA, which can reduce multifocal necrosis in the livers of mice (Kuzuhara et al., 2006). Because TCE can induce high levels of IL-6 and TNF-a (Jia et al., 2012), which are activators of SAA, we hypothesize that the elevated levels of SAA in acute-stage may protect against TCE induced liver injury. CLU is a highly conserved, secreted heterodimeric glycoprotein with a number of physiologic functions, including complement regulation, lipid transport, initiation of apoptosis and endocrine secretion (Rosenberg and Silkensen, 1995). Studies have shown that CLU is localized to immune deposits with complement components in a model of glomerulonephritis (Eddy and Fritz, 1991). Down-regulated CLU in acute phase OMLDT patients may be explained by the observations that TCE exposure may induce renal and immune toxicity (Mensing et al., 2002) and that OMLDT patients had relatively low levels of complement (Huang et al., 2012). Taken together, TCE exposure may triger an abnormal immune response, which finally results in the onset of OMLDT. Several apolipoproteins were down-regulated in acute-stage OMLDT patients. Apo-CIII and Apo-CII are regulators of lipid

108

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

Fig. 5. The validation results of candidate serum markers for OMLDT. The expression levels of these proteins in crude sera of controls and patients enrolled in the 2D-DIGE analysis were validated by Western blot (left panel), and 15 more controls and patients were used for ELISA (right panel) analysis. Paired Student’s t-test has been used for the statistical analysis of the ELISA results. *p < 0.01; (A) the expression trend of TTR among different test groups; (B) the expression trend of RBP4; and (C) the expression trend of Hp.

metabolism. Apo-CIII inhibits lipoprotein lipase (LPL), and hepatic lipase plays an important role in the catabolism of very low density lipoprotein (VLDL) and chylomicrons (CM) and regulates triglyceride levels in the plasma (Mendivil et al., 2010). Low levels of Apo-CIII indicate increased activity of LPL, a lipid detoxification

enzyme, which may contribute to the lipid peroxidation induced by TCE in the liver (Ogino et al., 1991). Apo-CII activates lipoprotein lipase in the capillaries, and has an important biological role in the catabolism of triglyceride-rich lipoproteins (Kim et al., 2006). Another down-regulated apolipoprotein,

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

109

Transparency document The Transparency document associated with this article can be found in the online version. Acknowledgments This work was supported by the NSFC (National Natural Science Foundation of China) (81273126), the Key Project of the Guangdong Natural Science Foundation (S2012020010903), the Project of Shenzhen Basic Research Plan (JCYJ20130329103949656, JCYJ20130329161137543), and the Upgrade Scheme of the Shenzhen Municipal Key Laboratory (JCYJ20130401102255980, CXB201104220031A). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.toxlet.2014.05.024. Fig. 6. The ELISA data based receiver operating characteristic (ROC) curves for TTR, RBP4 and Hp in discriminating acute stage from healing stage. The area under the curve (AUC) values are 0.864, 0.830 and 0.821 for TTR, RBP4 and Hp, respectively.

Apo-AI, promotes anti-inflammatory effects and plays a role in the non-specific immune system (Frank and Marcel, 2000). The down-regulation of Apo-AI mRNA in the liver appears to be mediated by IL-6 and TNF-a (Navarro et al., 2005). We postulate that abnormal lipid metabolism may be involved in the pathogenesis of OMLDT. In addition, some serum proteins have highly variable modifications, which can signal the onset of diseases and may be used as biological markers for diagnostics (Nedelkov et al., 2005). In this era of advanced mass spectrometry, it is no longer sufficient to speak simply of increased or decreased expression of proteins without carefully examining the splice variants of these proteins (Omenn, 2013). In our study, 8 differentially expressed proteins were identified, and six of these proteins appeared in more than 1 protein spot on the DIGE maps. This change in the PI and MW may be explained by post-translational modifications and differential cleavage of the proteins. The post-translational modifications and cleavage of these proteins may be of more clinically significant for OMLDT. 5. Conclusions In conclusion, we applied albumin/IgG depletion and ultrafiltration to optimize serum protein extraction procedure. 8 proteins were identified as differentially expressed in the serum of OMLDT patients in different disease stages by a proteomic-based study in combination of 2D-DIGE analysis and MALDI-TOF-MS. Among these, the altered proteins levels of TTR, RBP4 and haptoglobin were validated by Western blot analysis and ELISA assay. Out data not only suggested that TTR, RBP4 and haptoglobin could serve as potential serum biomarkers of OMLDT, but also indicated that the measurement of TTR, RBP4 and haptoglobin, or their combination could help aid in the diagnosis, monitoring the progression and therapy of the disease. Conflict of interest None of the authors has any potential conflict of interest or financial interests to disclose.

References Anderson, N.L., Anderson, N.G., 2002. The human plasma proteome: history, character, and diagnostic prospects. Mol. Cell. Proteom. 1, 845–867. Beck, F.K., Rosenthal, T.C., 2002. Prealbumin: a marker for nutritional evaluation. Am. Fam. Physician 65, 1575–1578. Chiu, W.A., Jinot, J., Scott, C.S., Makris, S.L., Cooper, G.S., Dzubow, R.C., Bale, A.S., Evans, M.V., Guyton, K.Z., Keshava, N., Lipscomb, J.C., Barone Jr., S., Fox, J.F., Gwinn, M.R., Schaum, J., Caldwell, J.C., 2013. Human health effects of trichloroethylene: key findings and scientific issues. Environ. Health Perspect. 121, 303–311. Cooper, G.S., Wither, J., Bernatsky, S., Claudio, J.O., Clarke, A., Rioux, J.D., Fortin, P.R., 2010. Occupational and environmental exposures and risk of systemic lupus erythematosus: silica, sunlight, solvents. Rheumatology (Oxford) 49, 2172–2180. Dai, Y.F., Niu, Y., Cheng, J., Leng, S.G., Zheng, Y.X., 2005. Immunological mechanism in development of allergic dermatitis in guinea pig induced by trichloroethylene in vitro. Zhonghua lao dong wei sheng zhi ye bing za zhi = Zhonghua laodong weisheng zhiyebing zazhi = Chin. J. Ind. Hyg. Occup. Dis. 23, 129–131. Eddy, A.A., Fritz, I.B., 1991. Localization of clusterin in the epimembranous deposits of passive Heymann nephritis. Kidney Int. 39, 247–252. Frank, P.G., Marcel, Y.L., 2000. Apolipoprotein A-I: structure–function relationships. J. Lipid Res. 41, 853–872. Galicia, G., Maes, W., Verbinnen, B., Kasran, A., Bullens, D., Arredouani, M., Ceuppens, J.L., 2009. Haptoglobin deficiency facilitates the development of autoimmune inflammation. Eur. J. Immunol. 39, 3404–3412. Goldman, S.M., 2010. Trichloroethylene and Parkinson’s disease: dissolving the puzzle. Expert Rev. Neurother. 10, 835–837. Griffin, J.M., Blossom, S.J., Jackson, S.K., Gilbert, K.M., Pumford, N.R., 2000. Trichloroethylene accelerates an autoimmune response by Th1 T cell activation in MRL +/+ mice. Immunopharmacology 46, 123–137. Hong, W.X., Liu, W., Zhang, Y., Huang, P., Yang, X., Ren, X., Ye, J., Huang, H., Tang, H., Zhou, G., Huang, X., Zhuang, Z., Liu, J., 2013. Identification of serum biomarkers for occupational medicamentosa-like dermatitis induced by trichloroethylene using mass spectrometry. Toxicol. Appl. Pharmacol. 273, 121–129. Huang, Y.S., Huang, H.L., 2010. Research progress on immune injury resulted from occupational medicamentose-like dermatitis induced by trichloroethylene. China Occup. Med. 522, 157–159. Huang, Z., Yue, F., Yang, X., Xia, L., Chen, C., Qiu, X., Huang, J., Li, L., Kamijima, M., Nakajima, T., Huang, H., 2012. Upregulation of calprotectin and downregulation of retinol binding protein in the serum of workers with trichloroethyleneinduced hypersensitivity dermatitis. J. Occup. Health 54, 299–309. Jia, Q., Zang, D., Yi, J., Dong, H., Niu, Y., Zhai, Q., Teng, Y., Bin, P., Zhou, W., Huang, X., Li, H., Zheng, Y., Dai, Y., 2012. Cytokine expression in trichloroethylene-induced hypersensitivity dermatitis: an in vivo and in vitro study. Toxicol. Lett. 215, 31–39. Kamijima, M., Hisanaga, N., Wang, H., Nakajima, T., 2007. Occupational trichloroethylene exposure as a cause of idiosyncratic generalized skin disorders and accompanying hepatitis similar to drug hypersensitivities. Int. Arch. Occup. Environ. Health 80, 357–370. Kamijima, M., Wang, H., Huang, H., Li, L., Shibata, E., Lin, B., Sakai, K., Liu, H., Tsuchiyama, F., Chen, J., Okamura, A., Huang, X., Hisanaga, N., Huang, Z., Ito, Y., Takeuchi, Y., Nakajima, T., 2008. Trichloroethylene causes generalized hypersensitivity skin disorders complicated by hepatitis. J. Occup. Health 50, 328–338. Kim, S.Y., Park, S.M., Lee, S.T., 2006. Apolipoprotein C-II is a novel substrate for matrix metalloproteinases. Biochem. Biophys. Res. Commun. 339, 47–54. Kuzuhara, H., Nakano, Y., Yamashita, N., Imai, M., Kawamura, Y., Kurosawa, T., Nishiyama, S., 2006. Protective effects of alpha1-acid glycoprotein and serum

110

P. Huang et al. / Toxicology Letters 229 (2014) 101–110

amyloid A on concanavalin A-induced liver failure via interleukin-6 induction by ME3738. Eur. J. Pharmacol. 541, 205–210. Li, H.S., Dai, Y.F., Huang, H.L., Feng, X.L., Sun, Y.F., Niu, Y., Cheng, J., Zhang, L.Y., Li, L.Y., Chen, B., Zheng, Y.X., 2006. Possible association between polymorphisms of human leukocyte antigen-DQ genes and susceptibility to trichloroethyleneinduced severe generalized dermatitis. Zhonghua yu fang yi xue za zhi = Chin. J. Prev. Med. 40, 173–176. Liu, J., Xing, X., Huang, H., Jiang, Y., He, H., Xu, X., Yuan, J., Zhou, L., Yang, L., Zhuang, Z., 2009. Identification of antigenic proteins associated with trichloroethyleneinduced autoimmune disease by serological proteome analysis. Toxicol. Appl. Pharmacol. 240, 393–400. Mendivil, C.O., Zheng, C., Furtado, J., Lel, J., Sacks, F.M., 2010. Metabolism of very-lowdensity lipoprotein and low-density lipoprotein containing apolipoprotein C-III and not other small apolipoproteins. Arterioscler. Thromb. Vasc. Biol. 30, 239–245. Mensing, T., Welge, P., Voss, B., Fels, L.M., Fricke, H.H., Bruning, T., Wilhelm, M., 2002. Renal toxicity after chronic inhalation exposure of rats to trichloroethylene. Toxicol. Lett. 128, 243–247. Moshage, H., 1997. Cytokines and the hepatic acute phase response. J. Pathol. 181, 257–266. Myron Johnson, A., Merlini, G., Sheldon, J., Ichihara, K., 2007. Clinical indications for plasma protein assays: transthyretin (prealbumin) in inflammation and malnutrition. Clin. Chem. Lab. Med. CCLM/FESCC 45, 419–426. Nakajima, T., Yamanoshita, O., Kamijima, M., Kishi, R., Ichihara, G., 2003. Generalized skin reactions in relation to trichloroethylene exposure: a review from the viewpoint of drug-metabolizing enzymes. J. Occup. Health 45, 8–14. Navarro, M.A., Carpintero, R., Acin, S., Arbones-Mainar, J.M., Calleja, L., Carnicer, R., Surra, J.C., Guzman-Garcia, M.A., Gonzalez-Ramon, N., Iturralde, M., Lampreave, F., Pineiro, A., Osada, J., 2005. Immune-regulation of the apolipoprotein A-I/CIII/A-IV gene cluster in experimental inflammation. Cytokine 31, 52–63. National Institute for Occupational Safety and Health (NIOSH), 1994. National Occupational Exposure Survey, 1981–1983. NIOSH, Cincinnati, OH. Naylor, H.M., Newcomer, M.E., 1999. The structure of human retinol-binding protein (RBP) with its carrier protein transthyretin reveals an interaction with the carboxy terminus of RBP. Biochemistry 38, 2647–2653.

Nedelkov, D., Kiernan, U.A., Niederkofler, E.E., Tubbs, K.A., Nelson, R.W., 2005. Investigating diversity in human plasma proteins. Proc. Natl. Acad. Sci. U. S. A. 102, 10852–10857. Ogino, K., Hobara, T., Kobayashi, H., Ishiyama, H., Gotoh, M., Imamura, A., Egami, N., 1991. Lipid peroxidation induced by trichloroethylene in rat liver. Bull. Environ. Contam. Toxicol. 46, 417–421. Omenn, G.S., 2013. Plasma proteomics, the human proteome project, and cancerassociated alternative splice variant proteins. Biochim. Biophys. Acta . Palha, J.A., 2002. Transthyretin as a thyroid hormone carrier: function revisited. Clin. Chem. Lab. Med. CCLM/FESCC 40, 1292–1300. Pullakhandam, R., Palika, R., Ghosh, S., Reddy, G.B., 2012. Contrasting effects of type 2 and type 1 diabetes on plasma RBP4 levels: the significance of transthyretin. IUBMB Life 64, 975–982. Ray, S., Reddy, P.J., Jain, R., Gollapalli, K., Moiyadi, A., Srivastava, S., 2011. Proteomic technologies for the identification of disease biomarkers in serum: advances and challenges ahead. Proteomics 11, 2139–2161. Rosenberg, M.E., Silkensen, J., 1995. Clusterin: physiologic and pathophysiologic considerations. Int. J. Biochem. Cell Biol. 27, 633–645. Son, D.S., Roby, K.F., Terranova, P.F., 2004. Tumor necrosis factor-alpha induces serum amyloid A3 in mouse granulosa cells. Endocrinology 145, 2245–2252. Toyama, T., Asano, Y., Takahashi, T., Aozasa, N., Akamata, K., Noda, S., Taniguchi, T., Ichimura, Y., Sumida, H., Tamaki, Z., Masui, Y., Tada, Y., Sugaya, M., Sato, S., Kadono, T., 2011. Clinical significance of serum retinol binding protein-4 levels in patients with systemic sclerosis. J. Eur. Acad. Dermatol. Venereol. . Tseng, C.F., Lin, C.C., Huang, H.Y., Liu, H.C., Mao, S.J., 2004. Antioxidant role of human haptoglobin. Proteomics 4, 2221–2228. Wang, Z., Burke, P.A., 2007. Effects of hepatocyte nuclear factor-4alpha on the regulation of the hepatic acute phase response. J. Mol. Biol. 371, 323–335. World Medical Association Declaration of Helsinki, 1997. Recommendations guiding physicians in biomedical research involving human subjects. JAMA 277, 925–926. Xu, X., Yang, R., Wu, N., Zhong, P., Ke, Y., Zhou, L., Yuan, J., Li, G., Huang, H., Wu, B., 2009. Severe hypersensitivity dermatitis and liver dysfunction induced by occupational exposure to trichloroethylene. Ind. Health 47, 107–112.