Comparative Profiling of Serum Glycoproteome by Sequential Purification of Glycoproteins and 2-Nitrobenzenesulfenyl (NBS) Stable Isotope Labeling: A New Approach for the Novel Biomarker Discovery for Cancer Koji Ueda,‡ Toyomasa Katagiri,‡ Takashi Shimada,† Shinji Irie,† Taka-Aki Sato,† Yusuke Nakamura,*,‡ and Yataro Daigo‡ Laboratory of Molecular Medicine, Human Genome Center, Institute of Medical Science, The University of Tokyo, Tokyo 108-8639, Japan, and Division of Advanced Clinical Proteomics, Institute of Medical Science, The University of Tokyo, Tokyo 108-8639, Japan Received February 25, 2007
The recent progress in various proteomic technologies allows us to screen serum biomarker including carbohydrate antigens. However, only a limited number of proteins could be detected by current conventional methods such as shotgun proteomics, primarily because of the enormous concentration distribution of serum proteins and peptides. To circumvent this difficulty and isolate potential cancerspecific biomarkers for diagnosis and treatment, we established a new screening system consisting of the sequential steps of (1) immunodepletion of 6 high-abundance proteins, (2) targeted enrichment of glycoproteins by lectin column chromatography, and (3) the quantitative proteome analysis using 12C6or 13C6-NBS (2-nitrobenzenesulfenyl) stable isotope labeling followed by MALDI-QIT-TOF mass spectrometric analysis. Through this systematic analysis for five serum samples derived from patients with lung adenocarcinoma, we identified as candidate biomarkers 34 serum glycoproteins that revealed significant difference in R1,6-fucosylation level between lung cancer and healthy control, clearly demonstrating that the carbohydrate-focused proteomics could allow for the detection of serum components with cancer-specific features. In addition, we developed a more simplified and practical technique, mass spectrometry-based glycan structure analysis and lectin blotting, in order to validate glycan structure of candidate biomarkers that could be applicable in clinical use. Our new glycoproteomic strategy will provide highly sensitive and quantitative profiling of specific glycan structures on multiple proteins, which should be useful for serum biomarker discovery. Keywords: biomarker • serum • lung cancer • glycosylation • glycoproteomics • stable isotope labeling • lectin • mass spectrometry
1. Introduction Lung cancer is the leading cause of cancer-related death in Japan and accounts for 19.0% of all cancer deaths in 2005 (Vital Statistics of Japan, Statistics and Information Dept., Minister’s Secretariat, Ministry of Health, Labour and Welfare). At the time of diagnosis, most of the lung cancer patients are at advanced stages; therefore, new diagnostic strategies detecting the early stage cancer have been eagerly expected to improve patients’ survival.1 Several tumor markers, such as ProGRP, NSE, cytokeratin 19-fragment (CYFRA 21-1), squamous-cell carcinoma antigen (SCC), and carcinoembryonic antigen (CEA) were found to be elevated in serum of patients with lung cancer, but no * Correspondence should be addressed to Yusuke Nakamura, M.D., Ph.D. Phone: +81-3-5449-5372. FAX: +81-3-5449-5433. E-mail:
[email protected]. ‡ Laboratory of Molecular Medicine, Human Genome Center, Institute of Medical Science, The University of Tokyo. † Division of Advanced Clinical Proteomics, Institute of Medical Science, The University of Tokyo. 10.1021/pr070103h CCC: $37.00
2007 American Chemical Society
tumor marker or their combination has been sufficiently useful for detection of lung cancer at potentially curative stage.2 In recent years, the use of proteomic technologies to discover cancer-specific proteins as tumor markers rapidly spread; however, only a few effective tumor markers were newly approved and introduced to cancer diagnosis in clinic.3 Protein glycosylation is one of the most abundant and structurally diverse post-translational modifications (PTMs).4,5 There is increasing evidence that glycosylation is highly sensitive to the biochemical environment and it plays a significant role in biological processes including transcription, differentiation, signaling, apoptosis, adhesion, and infiltration, as well as oncogenic transformation and metastasis.6-9 In fact, cancer cells produce various aberrant oligosaccharides that occur through genetic variations at glycosylation sites or structural change of glycan. Several reports suggested that the alterations in protein glycosylation on the cell surface and in body fluids might be associated with the development of certain types of Journal of Proteome Research 2007, 6, 3475-3483
3475
Published on Web 08/18/2007
research articles
Figure 1. Schema of high-throughput screening system to identify carbohydrate-targeting serum tumor biomarkers focusing on R1,6-fucosylation. This system was composed of 3 sequential steps: (1) removal of highly abundant proteins, (2) enrichment of glycoproteins by lectin column chromatography, and (3) quantitative analysis of glycosylation and protein identification by NBS stable isotope labeling and subsequent MALDIQIT-TOF mass spectrometric analysis. The MS intensity ratio of a 13C-NBS-labeled peptide derived from lung cancer versus the same fragment from control serum with 12C-NBS-label was designated as 13C/12C. This value obtained from the purified sample by LCA lectin column was described as LCA, while that from unpurified sample, Pre. Peptides satisfying the criteria in the box were selected as candidate biomarkers for lung cancer.
human cancer and other diseases.10-13 Highly tumor-specific alteration of glycan structures could be a potential target for cancer immunotherapy, such as epitopes for therapeutic monoclonal antibody.14 R1,6-Fucosylated N-linked glycan on R-fetoprotein (AFP-L3 fraction) is one of the well-established, carbohydrate-targeting tumor markers for clinical diagnosis. Although a level of total AFP protein in serum could be used as a tumor marker for hepatocellular carcinoma (HCC), positivity of serum AFP-L3 has much higher specificity (>95%) for HCC and relatively higher sensitivity for early stage HCC (
