Comparative Proteomic Analysis of Matched Primary and Metastatic Melanoma Cell Lines Mohammad Al-Ghoul,†,‡ Thomas B. Bru ¨ ck,†,§ Janelle L. Lauer-Fields,† Victor S. Asirvatham,† † Claudia Zapata, Russell G. Kerr,| and Gregg B. Fields†,* Department of Chemistry and Biochemistry, Florida Atlantic University, 777 Glades Road, Boca Raton, Florida 33431, and Biomedical Sciences at Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3 Received March 4, 2008
Identification of the biochemical pathways involved in the transformation from primary to metastatic melanoma is an area under intense investigation. A 2DE proteomics approach has been applied herein to the matched patient primary and metastatic melanoma cell lines WM-115 and WM-266-4, respectively, to better understand the processes that underlie tumor progression. Image analysis between samples aligned 470 common gel spots. Quantitative gel analysis indicated 115 gel spots of greater intensity in the metastatic line compared with the primary one, leading to the identification of 131 proteins via database searching of nano-LC-ESI-Q-TOF-MS/MS data. This more than tripled the number of proteins previously shown to be of higher abundance during melanoma progression. Also observed were 22 gel spots to be of lesser intensity in the metastatic line with respect to the primary one. Of these gel spots 15 proteins could be identified. Numerous proteins from both groups had not been reported previously to participate in melanoma progression. Further analysis of one protein, cyclophilin A, confirmed that this protein is expressed at higher levels in metastatic melanoma compared with primary melanoma and normal fibroblasts. Overall, this study expands our knowledge of protein modulation during melanoma stages, and suggests new targets for inhibitor development. Keywords: melanoma • cyclophilin • melanoma progression • melanoma proteomics
Introduction Melanomas are a heterogeneous class of malignancies that may occur in skin, eye, or meninges, among other tissues of the body. Primary cutaneous melanomas are characterized by horizontal growth that is limited to the epidermis. These malignancies can become more aggressive, leading to rapid growth and invasion of the dermis. In contrast to primary melanoma, the later stages of the disease are always associated with rapid invasion and metastasis in tissue types other than the dermis.1 During the past four decades the incidence of melanoma has increased at an alarming rate. In 2007, the new cases of melanoma was estimated at 59 940, which is nearly double that of the previous estimates.2 If melanoma is detected prior to metastasis, the five year survival rate is ∼99%.3 However, metastatic melanoma has a very poor prognosis (∼12-15% five year survival).3,4 The low survival rate is partially due to inherent chemoresistance of the disease.5 Therefore, an im* To whom correspondence should be addressed. Dr. Gregg B. Fields; Phone, +1-210-567-1312; E-mail,
[email protected]. † Florida Atlantic University. ‡ Present address: University of Louisville School of Medicine, Department of Pathology and Laboratory Medicine, MDR Building, Room 221, 511 South Floyd Street, Louisville, KY 40292. § Present address: Corporate Research and Development, Su ¨ d-Chemie AG, Staffelsee Str. 6, 81477 Mu ¨ nchen, Germany. | University of Prince Edward Island. 10.1021/pr800174k CCC: $40.75
2008 American Chemical Society
portant goal to improve clinical diagnosis and treatment options for melanoma patients is the identification of specific biomarkers that correlate with the different stages of the disease. These include markers that distinguish benign melanocytic nevi from primary stage melanoma as well as those that differentiate noninvasive from metastatic melanoma. Dysregulation of cell signaling pathways during the process of melanoma cell transformation from the primary to the metastatic stage has been well documented.5 Silencing of tumor suppressor genes and the activation of oncogenes has also been implicated in the development and progression of melanoma.6,7 The detailed mechanisms of melanoma progression have been probed using a variety of genomic and proteomic tools.8,9 Potential molecular markers of melanoma have been discovered at the gene level using cDNA microarrays10-14 and at the protein level using proteomics.1,4,8,9,15-18 Although cancer is commonly regarded as a disease of genes, the protein products of genes are the functional effector molecules. Hepatoma-derived growth factor (HDGF), nucleophosmin B23, quinolinate phosphoribosyltransferase, cathepsin D, and 14-3-3 γ have been shown to be strongly correlated with melanoma progression in a prior proteomics study which utilized two-dimensional electrophoresis (2DE) for protein separation and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) and database searches for protein identification.1 The same methodolJournal of Proteome Research 2008, 7, 4107–4118 4107 Published on Web 08/13/2008
research articles ogy was also shown to be effective in the identification of proteins associated with the development of chemoresistance in melanoma cell lines, including aldolase, the GTP:AMP phosphotransferase, and nicotinamide N-methyltransferase.19 Prior studies of melanoma proteomics have utilized a rather restricted pI range (4-7) and SDS-PAGE resolving capability.1,8 In addition, matched patient samples for primary and metastatic melanoma were not examined. The aim of the present study was to extend the knowledge of melanoma tumor progression by comparing matched primary and metastatic melanoma cell lines WM-115 and WM-266-4 using a robust 2DE approach in combination with protein identification via nLC-Q-TOF-MS/MS and computer assisted database matching. In the matched melanoma cell lines, 15 proteins were less abundant in the metastatic cell line while 115 were of greater abundance compared to the primary cell line. Several proteins were found that are known to be associated with melanoma progression, and several novel protein classes were identified that have not previously been linked to melanoma. Specifically, the transport protein 14-3-3 ζ, the cell adhesion protein galectin-1, a molecular chaperone, and a nuclear chloride channel protein were present in greater abundance in primary melanoma compared to its metastatic match. Cyclophilin A (CypA) was found to be at a higher concentration in metastatic compared to primary melanoma. CypA is commonly linked to various inflammatory processes as well as pancreatic cancer. Western blot analysis revealed increased CypA in two metastatic cell lines compared to one primary and one fibroblast cell line. To the best of our knowledge CypA, as well as several other of the aforementioned proteins, have not previously been linked to melanoma progression.
Experimental Section Materials. All chemicals were of analytical grade and purchased from Sigma Chemical Co. (St. Louis, MO). All gels were cast in-house using gel casting stands provided with the Ettan Dalt12 PAGE system (GE Healthcare, Piscataway, NJ). To prevent exogenous protein contamination, all gels were cast and handled in a sterile class 2 cabinet. Gels were continuously stored in enclosed plastic containers during the fixing and staining processes. Subsequent protein spot cutting was accomplished automatically using the EXQuest spot cutter (BioRad) without manual interference. Finally, manual in-gel digestion of protein spots was carried out in a sterile class 2 cabinet. Therefore, gel exposure to outside air was kept to a minimum at all times. Cell Culture. Two melanoma cell lines were examined in this study, which were derived from the same 58 year old female patient. The cell line designated WM-115 was of epithelial-like origin and isolated from the primary skin tumor, while the matched cell line WM-266-4 was isolated from a metastatic site. All cell lines were purchased from the American type Cell Culture Collection (ATCC) (Manassas, VA). Melanoma cell lines were verified by RT-PCR for the melanoma-specific genes tyrosinase and MelanA/MART1. The cells were grown in OptiMEM I medium supplemented by 4% fetal bovine serum, 0.1 mg/mL gentamycin sulfate, 50 units/mL penicillin, and 0.05 mg/mL streptomycin sulfate in a humidified atmosphere of 5% CO2 at 37 °C. Sample Preparation. Cells were grown to 80% confluency and released from the plates using trypsin. Harvested cells were washed twice with 10 mL PBS and centrifuged for 10 min at 800 × g at room temperature. The supernatant was discarded 4108
Journal of Proteome Research • Vol. 7, No. 9, 2008
Al-Ghoul et al. and the pellet was suspended in 1.0 mL of protein extraction reagent type 4 [7 M urea, 2 M thiourea, 40 mM Tris base, 1% w/v 3-(4-heptyl)phenyl-3-hydroxypropyl)dimethylammoniopropanesulfonate (C7BzO), pH 10.4] also containing 1.0% v/v Biolyte 3-10 (BioRad, Hercules, CA), 5 mM tributylphosphine (TBP), 25 U/mL benzoase, 1 mM PMSF, 10 µL protease inhibitors cocktail (Sigma Aldrich, catalog # P1860), and 0.1% bromophenol blue. The solution was then disrupted by gentle up and down pipetting followed by a 5 s sonication cycle. The sample was left at room temperature for 1 h and centrifuged at 13 000 rpm for 1 h. The supernatant was transferred into an Ultrafree-4 centrifugal filter unit (10 kDa cut off; Millipore, Bedford, MA) for desalting and concentrating of proteins. The total protein content in each sample was determined with a BSA standard using the Bradford method.20 Two-Dimensional Gel Electrophoresis (2DE). Samples prepared from each cell line were subjected to 2DE as described elsewhere.17 First dimension IEF focusing was carried out with a Protean IEF cell (BioRad). Each protein sample (3 mg, 300 µL) was applied via cup loading at both the basic and acidic ends of preparative IPG ready 24 cm strips having a nonlinear pH 3-10 gradient (BioRad). IEF focusing started at 200 V and the voltage was gradually increased to 5000 V at rate of 3 V/min and then kept constant for a further 24 h (about 140,000 Vh total). After IEF focusing, IPG strips were initially equilibrated for 15 min while shaking in a reducing buffer containing 6 M urea, 20% glycerol, 2% SDS, and 5 mM TBP. Subsequent protein alkylation was achieved by incubating IPG strips for 15 min in 6 M urea, 20% glycerol, 2% SDS, and 2.5% w/v iodoacetamide. After equilibration, IPG strips were loaded on top of SDS-PAGE gels [10% (w/v) total acrylamide concentration; 25 × 20 × 1.5 cm] and sealed with a 0.5% (w/v) solution of low melting agarose in Laemmli buffer containing 0.1% w/v bromophenol blue. A gel plug containing a standard Precision Plus broad range protein molecular weight (Mr) ladder (Bio-Rad) was placed next to the IPG strip for Mr calibration. The second dimensional separation of cellular proteins was carried out using SDS-PAGE (24 × 20 × 1.5 cm). Employing the Ettan Dalt12 PAGE system, which runs 12 preparative gels simultaneously, gels were run at 15 mA per gel in Laemmli buffer (BioRad) until the bromophenol dye front reached the bottom of the gel. After the second dimension run, gels were fixed for 12 h in methanol-10% acetic acid (3:2, v/v), and the gels were stained with Coomassie Blue R250 (J.T Baker Chemical Co. Phillipsburg, NJ) using the “Blue Silver” methodology.21 The “Blue Silver” methodology is a modification of the classical colloidal Coomassie G250 staining protocol with significantly enhanced protein detection sensitivity. Sets of 12 2DE gels were stained simultaneously for 20 h under identical conditions using the Dodeca Gel Stainer (Bio-Rad) until dye saturation occurred. The gels were destained with acetic acid-methanoldouble deionized water (7:40:53, v/v) until blue spots appeared against a clear background. The gels are then washed with water and imaged using the EXQuest spot cutter. Image Analysis and Spot Identification. Image analysis, including spot detection, mapping, spot volume determination, and comparisons between replicate gel sets was performed using the PDQuest software package V.7.5 (BioRad) according to the manufacturer’s protocols. The Mr and pI values for the respective proteome patterns were determined by interpolation with the PDQuest calibration tools using the external molecular weight marker associated with each gel and the predetermined pI range of the IPG strip (pH 3-10). In addition, a second
Comparative Proteomics of Melanoma Progression method was applied to verify the obtained Mr/pI calibration grid using a commercially available 2D standard protein mixture. In this process a mixture of Mr and pI marker proteins obtained from Bio-Rad (catalog # 161-0320) was separated by 2DE. After gel processing, the resulting gel images of the cell lysates and the 2D marker proteins were overlaid using PDQuest V.7.4. The sample and marker protein spot patterns could be separated visibly by color-coding using the Multichannel viewer option in PDQuest. The external Mr, pI marker proteins were then used to construct an Mr and pI grid for each gel. By comparing the two methods for Mr and pI calibration, no significant differences in Mr or pI values of gel spots were observed. Prior to comparison of the WM-115 and WM-266-4 data sets, the relative spot volumes of each protein were normalized to the sum of the intensities (Intensity × Area in parts per million) of all valid spots in each gel. Triplicate gel images of each cell lysate were used to compile the respective master gel composites. The three replicate gels that were chosen to make up the respective master gel composite for the WM-115 and WM266-4 data sets provide significant clarity of protein spot patterns such that gel matching could be accomplished semiautomatically in PDQuest with minimal manual editing. Editing was restricted to deletion of noise and checking for falsely detected spots. Comparisons between master gel composites for WM-115 and WM-266-4 cell lines were made only on the basis of spots present in all gels of each cell type. At least 85% of all detected protein spots could be matched to the same location in the master gel composites of WM-115 and WM266-4 data sets. Dysregulation of protein expression was assessed by comparing normalized protein spot volumes between the master gel composites for the primary and metatstatic melanoma cell lines. Normalized protein spot volumes were considered differentially regulated if their relative volume deviated more than 3 fold.22 On this basis, protein dysregulation was defined as an increase or decrease of more than 3-fold, when the same spots in different cell lines were compared. In addition to the fold dysregulation in spot volumes, we also assessed the statistical significance between normalized spot intensities by using an unpaired two-tailed Student’s t test as previously described.23 For all the dysregulated spots reported in Tables 1 and 2, the resulting p-values were
