Simple Method for Quantitative Analysis of N-Linked Glycoproteins in Hepatocellular Carcinoma Specimens Hyoung-Joo Lee,† Keun Na,† Eun-Young Choi,† Kyung Sik Kim,‡ Hoguen Kim,§ and Young-Ki Paik*,† Department of Biochemistry and Biomedical Sciences, College of Life Sciences and Biotechnology, World Class University Program, Yonsei Proteome Research Center and Biomedical Proteome Research Center, Department of Surgery, and Department of Pathology, College of Medicine, Yonsei University, Seoul 120-749, Korea Received July 22, 2009
Changes in N-linked glycan structures are related to the initiation and progression of hepatocellular carcinoma (HCC), one of the most common fatal cancers worldwide. In this study, we describe a simple and an efficient strategy for the selective identification and quantitation of N-linked glycoproteins that does not require extensive enrichment steps prior to MS/MS analysis. With this approach, N-linked glycoprotein differences between the plasma of healthy and HCC patients were selectively quantified after iTRAQ labeling. We identified a total of 14 N-linked glycopeptides with higher expression in HCC patient plasma than in healthy controls (g1.5 fold). Additionally, we characterized alterations in the glycan structures of vitronectin (Asn-169, 242) and antithrombin III (Asn-225) that were identified in HCC patient plasma. The intact glycopeptides with native glycan structures were also elevated in HCC tumor tissue. Taken together, these data support the utility of our approach for high throughput global profiling and quantification of the N-linked glycopeptides to identify disease biomarkers. Keywords: N-linked glycopeptides • iTRAQ • HCC • quantitative analysis • exclude peak list
Introduction Protein glycosylation is a major post-translational modification (PTM) that controls protein folding, conformational distribution, stability, and activity.1,2 Most secreted proteins are glycosylated, including important tumor biomarkers that are found at low levels in patient plasma.3,4 A critical goal in the field of biomarker identification with mass spectrometry (MS) is to develop the ability to accurately quantify small changes in glycosylated protein levels between samples from healthy subjects and subjects with disease.5 Therefore, glycoprotein isolation and enrichment prior to MS analysis are standard steps when screening for disease biomarkers. Many attempts have been made to develop an efficient enrichment method for glycoproteins present in complex mixtures. In particular, lectin-based affinity chromatography enriches for glycoproteins by exposing a complex mixture to an immobilized lectin bed and washing away unbound nonglycoproteins with a binding buffer.6-9 The interacting glycoproteins are subsequently recovered by eluting with a hapten sugar-based elution buffer. However, sometimes the extensive binding, washing, and elution results in sample loss, which reduces the efficiency and reproducibility for quantitative analysis. The low efficiency of * To whom correspondence should be addressed: Yonsei University, 134 Shinchon-dong, Sudaemoon-ku, Seoul, 120-749, Korea, Email: paikyk@ yonsei.ac.kr, Tel: 82-2-2123-4242, Fax: 82-2-393-6589. † Department of Biochemistry and Biomedical Sciences, College of Life Sciences and Biotechnology, World Class University Program, Yonsei Proteome Research Center. ‡ Department of Surgery. § Department of Pathology.
308 Journal of Proteome Research 2010, 9, 308–318 Published on Web 11/10/2009
recovery also increases contamination from dust, solvent, and other reagent impurities. It is therefore crucial to minimize the number of handling and transfer steps during the glycopeptides enrichment and analysis. N-linked glycans are conjugated to asparagine residues within the consensus sequence Asn-X-Ser/Thr.10 Changes in N-linked glycan modification occur during tumorigenesis and the progression of hepatocellular carcinoma (HCC),11-14 one of the most common fatal cancers worldwide.15,16 The analysis and structural characterization of glycan modifications in glycopeptides is performed using MS.17,18 However, MS analysis of glycoproteins is challenged by the inherently poor ionization efficiency of glycoproteins due to the suppression of their signals by nonglycosylated peptides during MS analysis. In addition, most glycosylation sites have various glycoforms, which further reduces the detection limit of the glycopeptide.19 In this study, we describe a simple and efficient strategy for the selective identification of N-linked glycoproteins in human plasma without extensive enrichment steps prior to MS analysis. With our approach, the differences in the structural repertoire of N-linked glycoproteins between healthy and HCC patient plasma were selectively quantified using iTRAQ labeling. This strategy has been validated for the characterization of intact N-linked glycopeptides attached to oligosaccharides in complex mixtures such as human plasma and human liver tissue. 10.1021/pr900649b
2010 American Chemical Society
N-Linked Glycoproteins in Hepatocellular Carcinoma
Materials and Methods Sample Preparation. Standard proteins (bovine serum albumin, ovalbumin, fetuin, alpha-1-acid-glycoprotein, fibrinogen, IgG, beta-casein, transferrin, vitronectin and antithrombim III) and PNGase-F were purchased from Sigma Aldrich (St. Louis, MO). Plasma was collected from two healthy and two HCC patients. Tumor and adjacent nontumor liver tissue samples were collected from two HCC patients. The samples were obtained with informed consent and in accordance with IRB guidelines from the Yonsei University College of Medicine (Seoul, Korea). Plasma was prepared from patient blood by incubation with K2-EDTA as previously described20,21 and stored at -80 °C until use. Tissue samples were frozen in liquid nitrogen, ground to fine powder form, and homogenized in lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 10 mM Tris, 5 mM magnesium acetate, and one complete proteinase inhibitor cocktail tablet (Roche, Switzerland)). The homogenized lysate was sonicated on ice for 1 min with 2 s pulses every 2 s, and then for 30 min at room temperature with repeated agitations. Insoluble cellular debris was removed by centrifugation at 98 235g for 60 min at 4 °C. The supernatants were filtered through a Nanosep MF filter (0.45 µm, PALL, Port Washington, NY) and stored in aliquots at 4 °C. Protein concentrations were determined with the Bradford assay kit (BioRad, Hercules, CA) using bovine serum albumin as a standard. Depletion of Abundant Proteins. The top six (albumin, transferrin, IgG, IgA, haptoglobin, and antitrypsin) or top 14 (albumin, IgG, antitrypsin, IgA, transferrin, haptoglobin, fibrinogen, alpha2-macroglobulin, alpha1-acid glycoprotein, IgM, apolipoprotein A1, apolipoprotein A2, complement C3, and transthyretin) high abundance plasma proteins were depleted from human plasma with a Multiple Affinity Removal Column (MARC) (Agilent, Wilmington, DE) as previously described.20,22 Immunoblotting. Samples were separated by SDS-PAGE on 10% polyacrylamide gels and transferred onto nitrocellulose (NC) membranes using the iBLOT dry blotting system (Invitrogen, Carlsbad, CA). NC membranes were blocked with TBS-T buffer (20 mM Tris, 137 mM NaCl, 0.1% Tween-20, pH 7.6) containing 5% nonfat dry milk for 1 h at room temperature, and incubated for 2 h with a 1:200 dilution of monoclonal antivitronectin antibody (Santa Cruz Biotechnology, La Jolla, CA) in TBS-T at 4 °C. After washing for 30 min with TBS-T buffer, membranes were incubated for 1 h with a 1:5,000 dilution of horseradish peroxidase-conjugated donkey antigoat antibody in TBS-T (Santa Cruz Biotechnology), and then for 3 min with ECL Plus Immunoblotting detection reagents (GE Healthcare, Piscataway, NJ). In-Solution Tryptic Digestion and iTRAQ Labeling. Samples were reduced, alkylated, and subjected to tryptic digestion as previously described.20 Each 100 µg sample in 20 µL of 500 mM TEAB was reduced, alkylated, digested, and labeled with iTRAQ reagents (114:116) according to the manufacturer protocol (Applied Biosystems, Foster City, CA). Deglycosylation. The samples were enzymatically deglycosylated using PNGase F (Q-A Bio, Palm Desert, CA) according to the manufacturer protocol. Briefly, the sample (100 µg) was dissolved in 25 mL NH4HCO3 (pH 8.0), deglycosylated by adding 1 µL of PNGase F, and incubated 2 h at 37 °C. Glycoprotein Enrichment with Multi-Lectin Columns. Enrichment of human plasma glycoproteins was performed as previously described with modifications.8 The multilectin af-
research articles finity column (MLAC) was prepared by mixing equal amounts of agarose-bound concanavalin A, agarose-bound wheat germ agglutinin, agarose-bound Sambucus nigra agglutinin (SNA), agarose-bound Aleuria aurantia agglutinin (AAL) in an empty disposable column (Pierce Biotech, Rockford, IL). The one milligram serum sample was diluted with MLAC equilibrium buffer (20 mM Tris, 0.15 M NaCl, 1 mM Mn2+, and 1 mM Ca2+, pH 7.4) to a volume of 1 mL and loaded onto a newly packed MLAC. After 30 min incubation, the unbound proteins were washed away with 10 mL of equilibrium buffer, and the captured proteins were released with 15 mL of a displacer solution (20 mM Tris, 0.5 M NaCl, 0.17 M methyl-R-Dmannopyranoside, 0.17 M N-acetylglucosamine, 0.2 M Dlactose, 0.2 ML R-L-fucose, pH 7.4). Fractions were collected from the MLAC and concentrated with 15-mL, 5-kDa membrane filters (PALL, Port Washington, NY). LC-MS/MS for Peptides Analysis. NanoLC-MS/MS analysis was performed with a Tempo nano-LC system (Applied Biosystems, Forster City, CA). The capillary column used for LC-MS/MS analysis (150 mm × 0.075 mm) was obtained from Proxeon (Odense, Denmark) and the slurry packed in-house with a 5 µm, 100 Å pore size Magic C18 stationary phase resin (Michrom BioResources, Auburn, CA). The mobile phase A for LC separation was 0.1% formic acid in deionized water and the mobile phase B was 0.1% formic acid in acetonitrile. The chromatography gradient was designed for a linear increase from 5% B to 35% B in 100 min, 40% B to 60% B in 10 min, 90% B in 15 min, and 5% B in 20 min. The flow rate was maintained at 300 nL/min. Q-STAR ELITE mass spectrometry (Applied Biosystems, Forster City, CA) was used for the identification and quantification of peptides. The Analyst QS software (version 2.0, Applied Biosystems, Forster City, CA) was used to generate peak lists Product ion spectra were collected in the information-dependent acquisition (IDA) mode using continuous cycles of one full scan TOF MS from 400-1500 m/z (1.0 s) plus three product ion scans from 50-1500 m/z (1.5 s each). Precursor m/z values were selected starting with the most intense ion using a selection quadrupole resolution of 3 Da. The rolling collision energy feature was used, which determines collision energy based on the precursor value and charge state. The dynamic exclusion time for precursor ion m/z values was 120 s. An LTQ mass spectrometer was used for data collection as previously described.20 Identification of Peptides and Quantification of iTRAQ Labeled Peptides. ProteinPilot (version 2.0, Applied Biosystems, Forster City, CA) software was used for peak generation using Paragon algorithms (Applied Biosystems, Forster City, CA). The peptides were identified using IPI protein sequence database (IPI human_03.26. 2007) which is nonredundant database. Database search criteria were as described: taxonomy Homo sapiens, 52015 sequences for taxonomy, no fixed modification, oxidized at methionine (+16) residues and carboxyamidomethylated (+57) at cysteine residues for variable modifications, one maximum allowed missed cleavage, 0.5 Da MS tolerance, and a 0.2 Da MS/MS tolerance. Only peptides resulting from trypsin digests were considered. Proteins with at least one unique “identity” scored peptide were considered unambiguously identified with Pro Group algorithm. The minimum threshold was set to 30 ion counts. The peptides were filtered with a significance threshold of probability of >95%. Quantification was carried out by calculating the ratio between the peak areas of the 114/116 iTRAQ reporter groups. To eliminate any masking of changes in expression due to peptides that are Journal of Proteome Research • Vol. 9, No. 1, 2010 309
research articles
Lee et al.
Scheme 1. Strategy for Selective Detection of N-Glycopeptides in Human Plasma
shared between proteins, we calculated protein ratio using only ratios from the spectra that are distinct to each protein. All quantitative results were calculated based on default biascorrections. Reversed sequences were used for the evaluation of the false discovery rate (FDR). We set the FDR to
