Proteomic Analysis of Colorectal Cancer Metastasis: Stathmin-1

Dec 19, 2011 - Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, 8 Medical Drive, Singapore 117597, ...
0 downloads 0 Views 7MB Size
Article pubs.acs.org/jpr

Proteomic Analysis of Colorectal Cancer Metastasis: Stathmin-1 Revealed as a Player in Cancer Cell Migration and Prognostic Marker Hwee Tong Tan,† Wei Wu,† Yi Zhen Ng,† Xuxiao Zhang,† Benedict Yan,‡ Chee Wee Ong,‡ Sandra Tan,§ Manuel Salto-Tellez,‡ Shing Chuan Hooi,∥ and Maxey C. M. Chung*,†,§ †

Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, 8 Medical Drive, Singapore 117597, Singapore ‡ Department of Pathology, National University Hospital, 5 Lower Kent Ridge Road, Singapore 119074, Singapore § Department of Biological Sciences, Faculty of Science, National University of Singapore, 14 Science Drive 4, Singapore 117543, Singapore ∥ Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, 4 Medical Drive, Singapore 117597, Singapore S Supporting Information *

ABSTRACT: Metastasis accounts largely for the high mortality rate of colorectal cancer (CRC) patients. In this study, we performed comparative proteome analysis of primary CRC cell lines HCT-116 and its metastatic derivative E1 using 2-D DIGE. We identified 74 differentially expressed proteins, many of which function in transcription, translation, angiogenesis signal transduction, or cytoskeletal remodeling pathways, which are indispensable cellular processes involved in the metastatic cascade. Among these proteins, stathmin-1 (STMN1) was found to be highly up-regulated in E1 as compared to HCT-116 and was thus selected for further functional studies. Our results showed that perturbations in STMN1 levels resulted in significant changes in cell migration, invasion, adhesion, and colony formation. We further showed that the differential expression of STMN1 correlated with the cells’ metastatic potential in other paradigms of CRC models. Using immunohistochemistry, we also showed that STMN1 was highly expressed in colorectal primary tumors and metastatic tissues as compared to the adjacent normal colorectal tissues. Furthermore, we also showed via tissue microarray analyses of 324 CRC tissues and Kaplan−Meier survival plot that CRC patients with higher expression of STMN1 have poorer prognosis. KEYWORDS: colorectal cancer, metastasis, proteomics, Stathmin-1, prognosis

1. INTRODUCTION Colorectal cancer (CRC) is one of the top malignant diseases in the world, affecting both men and women. Cancer metastases is the main cause of its high mortality rate, as 20− 50% of CRC patients die within 5 years after diagnosis due to extensive metastatic diseases.1 Metastasis is a complex multistep process describing the spread of tumor cells from the tissue of origin to other sites in the body.2 It consists of a series of events such as self-renewal to initiate tumor growth, resistance to extracellular death signals, alteration of cellular adhesion, and attainment of migratory and invasive capability. The metastasis cascade has been described extensively in a review by Gupta and Massague.2 Although several proteins have been identified to be associated with metastasis, including those in cell growth, motility and apoptosis, the underlying molecular mechanisms of CRC metastasis remain to be elucidated.3,4 There are currently no effective biomarkers for detecting early CRC metastasis. Hence, the identification of proteins associated with CRC metastasis would help to gain insights into the key players © 2011 American Chemical Society

involved in CRC malignancy, and thus aid in the development of novel biomarkers and therapeutic candidates for CRC. In recent years, cancer proteomics has been used to study proteome alterations in cancer metastasis to discover novel diagnostic markers, find therapeutic interventions and gain insights into the mechanism of metastasis.3−8 However, the reliability of the CRC metastasis markers identified by these studies has been limited by the choice of samples. These studies are not ideal as the proteins may be differentially expressed due to the intrinsic differences between two unassociated cell lines or between human tissues and in vitro cell lines. For example, Brunagel et al. (2002) compared patient samples of liver metastasis to primary colorectal cancer cell lines.5 In a study by Zhang et al. (2005), the proteomes of metastatic (LS174T) and nonmetastatic (SW480) CRC cells derived from different patients were analyzed, which identified proteins involved in cell growth, motility, invasion, adhesion, apoptosis and tumor Received: November 3, 2011 Published: December 19, 2011 1433

dx.doi.org/10.1021/pr2010956 | J. Proteome Res. 2012, 11, 1433−1445

Journal of Proteome Research

Article

immunity.3 Comparing two cell lines with the same genetic background would overcome these constraints. Katayama et al. (2006) compared SW480 and SW620 cells that are derived from primary colon tumor and lymph node metastasis of the same patient respectively, and showed that there was upregulation of alpha-enolase and triosephosphate isomerase.4 These two cell lines have also been used in other studies on CRC metastasis and have led to the identification of the overexpression of trefoil factor 3, growth/differentiation factor 15,7 and heat shock protein-27 (HSP27).8 Pei et al. (2007) has performed 2-DE of lymph node metastasis of CRC using clinical tissues and detected increased levels of HSP-27, glutathione S-transferase (GST), and Annexin II but decreased expression of liver-fatty acid binding protein (L-FABP).6 However, lymph node metastasis is only the intermediate step in CRC metastasis. To identify proteins that are truly associated with CRC metastasis, primary CRC cell line HCT116 and its metastatic derivative E1 were chosen as a CRC metastasis model for functional proteomics study in this work. The metastatic E1 cell line was generated via repeated intrasplenic injections of HCT-116 in nude mice.9 These two CRC cell lines with a similar genetic background were studied as differences in protein expressions observed in E1 as compared to HCT-116 are more likely attributed to the acquisition of metastatic potential in E1. Recently, gene expression profiling of E1 cells had identified phalladin’s role in the acquisition of epithelial-mesenchymal transition (EMT) features by E1 cells.9 In this study, comparative proteomics of HCT-116 versus E1 was performed using 2-dimensional difference gel electrophoresis (2-D DIGE) analysis of protein fractions enriched from heparin affinity chromatography of these two cell lines. In total, 74 unique proteins are identified from the 90 differentially expressed protein spots. This is the first report that showed the involvement of a comprehensive coverage of proteins, which included transcription, translation, cytoskeletal remodeling and signaling pathway, in CRC metastasis. STMN1 was found to be the most up-regulated protein in E1 as compared to HCT-116. The identification of STMN1, a microtubule destabilizing protein, is of particular significance as cytoskeletal structure plays a crucial role in cancer metastasis. In this study, we showed that STMN1 may play a putative role in CRC metastasis, and its expression correlates with poor prognosis of CRC patients.

thiourea (Fluka, Buchs, Switzerland), 4% CHAPS (USB), 10 mM Tris (J. T. Baker, Phillipsburg, NJ), 1× Halt protease inhibitor cocktail (Pierce, Rockford, IL), 50 μg/mL DNase I and 50 μg/mL RNase A (Roche Diagnostic, Mannheim, Germany). Proteins were extracted from SW480 and SW620 cells using the RIPA buffer containing 1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, 140 mM sodium chloride and 10 mM Tris HCl, pH 8. The cell lysates were centrifuged at 18800× g for 60 min at 15 °C to obtain the supernatant. For heparin affinity chromatography, harvested HCT-116 and E1 cells were resuspended in 10 mM sodium phosphate buffer (pH 7.0) (Merck, Darmstadt, Germany), 1x HaltTM protease inhibitor cocktail, 50 μg/mL DNase I and 50 μg/mL RNase A. The cells were then lysed by sonication, and cell lysates were then centrifuged at 75000× g for 1 h at 4 °C as described previously.10 Heparin Affinity Chromatography

3 mg of proteins from HCT-116 and E1 cell lysates were each fractionated by a HiTrap Heparin HP column connected to the Ettan Liquid Chromatography system (GE Healthcare, Uppsala, Sweden) as described previously10 with some modifications. The heparin column was equilibrated with 10 mM sodium phosphate (pH 7.0) at a flow-rate of 0.5 mL/min prior to sample loading. Flow-through fractions which contained the unbound proteins were collected until the optical density at 280 nm returned to the baseline level. Heparin-bound proteins were then fractionated via stepwise elution using 10 mM sodium phosphate (pH 7.0) containing 0.8 M sodium chloride. The eluates were desalted and concentrated using the 2-D Clean-Up kit (GE Healthcare) and the Ultracel YM-10 Microcon centrifugal filter device (Millipore, Billerica, MA) according to the manufacturer’s instructions. Protein Assay

Protein concentration of each sample was determined using the Coomassie Plus Protein Assay Reagent kit (Pierce) with bovine serum albumin (BSA) as the calibrating standard. The samples that are lysed in the cocktail solution were subjected to at least 4× dilution so that the concentrations of urea and CHAPS are compatible with the assay. The samples that are lysed in the RIPA buffer are subjected to at least 40× dilution as recommended by the manufacturer’s instructions. 2-D DIGE

2. MATERIALS AND METHODS

Minimal labeling of the protein fractions from the heparin affinity chromatography for 2-D DIGE was performed according to the protocol stated in the Ettan DIGE User Manual (GE Healthcare) and as described previously.10 Three biological batches of HCT-116 and E1 cells were used. Cy3 and Cy5 dyes were used to label proteins from HCT-116 and E1 cells respectively. An equal amount of Cy2 was added to a mixture of the six protein fractions (heparin bound or unbound protein fractions) obtained from the three batches of HCT-116 and E1 cells to act as the internal standard. Duplicate 2-D DIGE gels were run. Isoelectric focusing (IEF) was performed on Immobilized pH gradient (IPG) strips (GE Healthcare) (18 cm, pH 3−10, nonlinear) using the following focusing parameters: (i) 200 V, 200 Vhr; (ii) 500 V, 500 Vhr; (iii) 1000 V, 500 Vhr; (iv) 1000−8000 V, 4500 Vhr; and (v) 8000 V, 32000 Vhr. After IEF was completed, SDS-PAGE was carried out as described previously.10

Cell Culture

E1 cells were generated from HCT-116 cells as described previously.9 Human colorectal carcinoma HCT-116, HT-29 (American Type Culture Collection, Rockville, MD) and E1 cell lines were cultured in modified McCoy’s 5A media (Sigma, St. Louis, MO) whereas SW620 and SW480 (American Type Culture Collection) were cultured in DMEM-HG (Sigma) in a humidified incubator at 37 °C and 5% CO2. Both media were supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA). HCT-116 and HT-29 cells were treated with 5 mM butyrate (Sigma) for 24 h as described before.10 The cells were harvested via trypsinization upon reaching 80−90% confluency. Sample Preparation

HT-29, HCT-116 and E1 cells were disrupted using a cocktail solution containing 7 M urea (USB, Cleveland, OH), 2 M 1434

dx.doi.org/10.1021/pr2010956 | J. Proteome Res. 2012, 11, 1433−1445

Journal of Proteome Research

Article

Fluorescent images of the 2-D DIGE gels were obtained and processed as described previously.10 The 2-D DIGE patterns were evaluated using DeCyder v6.5 software (GE Healthcare) as described in the Ettan DIGE user manual. A threshold limit of 1.5-fold difference was set as the statistically significant quantitative change. Student’s t-test at 95% statistical confidence (p < 0.05) was used. In addition, the protein spots must be present in all the six replicate 2-D DIGE gels for heparin bound or unbound protein fractions. The 2-D DIGE gels were silver stained, and protein spots that satisfied the above criteria were excised for in-gel tryptic digestion as described previously.10 The experimental Mr and pI of proteins spots in the 2-D gels were estimated using an in-house program.

(1:1000 in 1% milk overnight at 4 °C) from OriGene Technologies Inc. (Rockville, MD) were used as primary antibodies in detection of endogenous and overexpressed STMN1 respectively. Rabbit antiglyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:2500) from Santa Cruz, and α-actin (1:5000) from Sigma were used as loading controls. HRP-conjugated antirabbit IgG (1:5000) from Pierce and antimouse IgG (1:5000) from GE Healthcare were used as secondary antibodies. Immuno-reactive protein spots or bands were visualized using the Enhanced Chemiluminescence (ECL) detection system (GE Healthcare), SuperSignal West Dura Extended Duration Chemiluminescent Substrate (Pierce) or Lumigen TMA-6 (GE Healthcare), and Biomax film (Eastman Kodak Company, New York, NY). Relative quantitation of protein expression was determined using the Quantity One software (Bio-Rad).

Mass Spectrometry Analysis and Database Searching

STMN1 shRNA and Overexpression Contructs

Peptide digests from each gel spot were reconstituted with 1.2 μL CHCA matrix solution (5 mg/mL α-cyano-4-hydroxycinnamic acid in 0.1% TFA, 50% ACN) and spotted onto a MALDI target plate. The samples were analyzed using a 4800 Proteomics Analyzer MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA) for MS and MS/MS analyses as described previously.11 For MS analysis, typically 500 shots were accumulated for each well of sample. MS data were automatically obtained with the five most intense ions selected for MS/MS. The peptides were subsequently subjected to MS/MS analyses using air with a collision energy of 2 kV and a collision gas pressure ∼2 × 10−6 Torr. Stop conditions were implemented so that 2000 to 3000 shots were accumulated depending on the quality of the spectra. MASCOT search engine (version 2.1; Matrix Science) was used to search all of the tandem mass spectra. GPS Explorer software Version 3.6 (Applied Biosystems) was used to create and search files with MASCOT search engine for peptide and protein identification. The search parameters allowed N-terminal acetylation, cysteine C-terminal carbamidomethylation and methionine oxidation as variable modification. One missed cleavage was allowed. Keratin (m/z 804.4097, 973.5313, 1234.68, 1475.785, 1320.583, 1707.773, 1791.728, 1993.977 and 2705.161) and tryptic (m/z 842.51, 1045.56, 1940.95, 2211.10, 2225.12, 2239.13, 2283.18 and 2807.30) peptides were excluded from MS/MS acquisition. Peptide mass tolerance and fragment mass tolerance were set to 150 ppm and ±0.4 Da respectively. The precursor ions formed by the MALDI-TOF/TOF MS has a charge of 1+. The International Protein Index (IPI) human database (Version 3.45, 71983 sequences)12 was used for the search, which was restricted to tryptic peptides. The threshold expectation value used for protein identification in this study was based on the significance level recommended by MASCOT which is an expectation value less than 0.05 (p value