Quantitative Proteomic Signature of Liver Cancer Cells: Tissue Transglutaminase 2 Could Be a Novel Protein Candidate of Human Hepatocellular Carcinoma Yulin Sun,†,¶ Wei Mi,§,¶ Jianqiang Cai,# Wantao Ying,§ Fang Liu,† Haizhen Lu,| Yuanyuan Qiao,‡ Wei Jia,§ Xinyu Bi,# Ning Lu,| Shangmei Liu,| Xiaohong Qian,*,§ and Xiaohang Zhao*,†,‡ State Key Laboratory of Molecular Oncology, Department of Abdominal Surgery, and Department of Pathology, Cancer Institute & Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, 100021 P. R. China, State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing, 102206 P. R. China, and Center of Basic Medical Sciences, Navy General Hospital, Beijing 100037, P. R. China Received February 27, 2008
Hepatocellular carcinoma (HCC) is one of the most common diseases worldwide, with extremely poor prognosis due to failure in diagnosing it early. Alpha-fetoprotein (AFP) is the only available biomarker for HCC diagnosis; however, its use in the early detection of HCC is limited, especially because about one-third of patients afflicted with HCC have normal levels of serum AFP. Thus, identifying additional biomarkers that may be used in combination with AFP to improve early detection of HCC is greatly needed. A quantitative proteomic analysis approach using stable isotope labeling with amino acids in cell culture (SILAC) combined with LTQ-FT-MS/MS identification was used to explore differentially expressed protein profiles between normal (HL-7702) and cancer (HepG2 and SK-HEP-1) cells. A total of 116 proteins were recognized as potential markers that could distinguish between HCC and normal liver cells. Certain proteins, such as AFP, intercellular adhesion molecule-1 (ICAM-1), IQ motif containing GTPase activating protein 2 (IQGAP2), claudin-1 (CLDN1) and tissue transglutaminase 2 (TGM2), were validated both in multiple cell lines and in 61 specimens of clinical HCC cases. TGM2 was overexpressed in some of the AFP-deficient HCC cells (SK-HEP-1 and Bel-7402) and in about half of the tumor tissues with low levels of serum AFP (17/32, AFP-negative HCC). Trace amounts of TGM2 were found to be expressed in the samples with high serum AFP (26/29, AFP-positive HCC). Moreover, TGM2 expression in liver tissues showed an inverse correlation with the level of serum AFP in HCC patients. Notably, TGM2 existed in the supernatant of the AFP-deficient SK-HEP-1, SMMC-7721 and HLE cells, and it was found to be induced in AFP-producing cells (HepG2) by specific siRNA silence assay. Serum TGM2 levels of 109 HCC patients and 42 healthy controls were further measured by an established ELISA assay; the levels were significantly higher in HCC patients, and they correlated with the histological grade and tumor size. These data suggest that TGM2 may serve as a novel histological/serologic candidate involved in HCC, especially for the individuals with normal serum AFP. These novel findings may provide important clues to identify new biomarkers of HCC and indirectly improve early detection of the disease. Keywords: hepatocellular carcinoma • quantitative proteomics • stable isotope labeling with amino acids in cell culture • transglutaminase 2 • biomarker
Introduction Liver cancer ranks sixth among common fatal cancers worldwide and it is usually diagnosed at the advanced stage * To whom correspondence should be addressed. For X.Z.: phone, +8610-67709015; fax, +86-10-87778360; e-mail,
[email protected]. For X.Q.: phone, +86-10-80705155; fax, +86-10-80705055; e-mail,
[email protected]. † State Key Laboratory of Molecular Oncology, CIH, CAMS & PUMC. § State Key Laboratory of Proteomics, BPRC, BIRM. ¶ These author contributed equally to this work. # Department of Abdominal Surgery, CIH, CAMS & PUMC. | Department of Pathology, CIH, CAMS & PUMC. ‡ Center of Basic Medical Sciences, Navy General Hospital. 10.1021/pr800153s CCC: $40.75
2008 American Chemical Society
with a low 5-year survival rate. Hepatocellular carcinoma (HCC) is the most common hepatic neoplasm and it represents a majority of all liver cancer cases with a dismal outcome.1 Alphafetoprotein (AFP), the only clinically available marker, has been widely used for serological diagnosis of human HCC. However, the measurement of AFP is not an ideal method for screening individuals at risk of developing HCC due to its poor sensitivity and specificity. Elevated levels of serum AFP are present in up to 20% of patients with chronic hepatitis, 20-60% of those with hepatitis cirrhosis, 60-70% of those with HCC, and in some patients with embryonic carcinomas.2–4 Importantly, about Journal of Proteome Research 2008, 7, 3847–3859 3847 Published on Web 07/23/2008
research articles one-third of HCC patients exhibit low levels of serum AFP, and some of these patients are in the early stage. Therefore, the discovery of additional biomarkers may improve the early diagnosis of HCC as well as provide better understanding of the mechanisms underlying tumorigenesis at the clinical level, especially for individuals who have no detectable serum AFP. Recently, genomics- and proteomics-based approaches have been used to identify more specific candidate markers of HCC. Jia et al. used microarray analysis to explore the expression profiles of 218 HCC specimens from patients with either high or low serum AFP levels and they identified 37 candidate genes.5 From the transcriptome levels, it was found that AFPproducing HCC cell lines such as HepG2 and Hep3B share common gene-expression profiles, thereby differentiating them from AFP-silent HCC cell lines.6,7 On the other hand, using 2-D DIGE, a proteomics approach, Yokoo et al. identified 11 proteins that distinguished 9 AFP-producing cell lines from 7 nonproducing ones.8 These studies suggest that AFP-positive and AFP-negative tumor cells may have different mechanisms underlying tumorigenesis and lead to different gene expression profiles in these two patient groups. For the identification of new biomarkers, it is necessary to analyze quantitatively the differential expressed protein profiles of liver cancer cells, those that either do or do not produce AFP, as compared with the profiles for normal liver cells in the same system. Stable isotope labeling with amino acids in cell culture (SILAC) assay combined with high accuracy MS identification has recently been proved to be a powerful approach for quantitative proteomic study in cell culture systems, and this approach allows the analysis of three different types of cells simultaneously.9,10 In the present study, we employed the SILAC method to characterize and quantitatively compare protein profiles among two liver cancer cells, HepG2 and SK-HEP-1, and a normal liver cell line, HL-7702. The aim of this study is to identify the protein signatures that distinguish HCC and normal liver cells and to find potential diagnostic biomarkers. From 116 differentially expressed proteins, several candidates, including AFP, tissue transglutaminase 2 (TGM2), IQ motif containing GTPase activating protein 2 (IQGAP2), claudin-1 (CLDN1), and intercellular adhesion molecule-1 (ICAM-1), were further validated in 7 HCC and one normal liver cell lines as well as in tumor tissue samples obtained from 61 clinical HCC individuals. Notably, TGM2 was found to be overexpressed in AFP-negative HCC cells and in about half of the tissue samples with low serum AFP levels. TGM2 expression in tumor tissue samples obtained from 61 clinical HCC individuals was inversely correlated with their serum AFP levels. In other words, patients who had low levels of AFP in serum showed higher TGM2 levels in their cancerous tissues. SiRNA-mediated AFP knockdown proved that increased TGM2 expression was accompanied by increased concentrations of siRNA duplex. Meanwhile, TGM2 could be secreted by AFP-deficient cells, SK-HEP-1, SMMC-7721 and HLE, but very little secretion was observed in HepG2 cells. Serum TGM2 was first found to be significantly elevated in HCC patients as compared with healthy donors. These data suggested TGM2 might serve as a potential histological/serologic biomarker of HCC, especially in individuals with low serum AFP.
Experimental Procedures Cell Culture, Measurement of AFP Expression, and Conditioned Media Collection. Human liver cancer cell linessHepG2 (ATCC HB-8065), Hep3B (ATCC HB-8064), and 3848
Journal of Proteome Research • Vol. 7, No. 9, 2008
Sun et al. SK-HEP-1 (ATCC HTB-52)swere purchased from American Type Culture Collection (Rockville, MD); Bel-7402, SMMC-7721, and normal liver cell line HL-7702 were purchased from the Institute of Biochemistry and Cell Biology of Chinese Academy of Sciences (Shanghai, China);11 and HLE and Huh-7 cell lines were obtained from the Human Science Research Resources Bank (Osaka, Japan).8 All cells were maintained in recommended media at 37 °C and 5% CO2. Approximately 4 × 104 liver cells of each kind mentioned above were cultured in a 10 cm dish with 10 mL of complete medium. Their supernatants were harvested after 5 days and the concentration of AFP was determined using RIA. The purified AFP at different concentrations was used as a standard. Approximately 1 × 107 HepG2, SK-HEP-1, SMMC-7721, or HLE cells were grown to 80% confluence and washed six times in PBS and incubated in serum-free MEM medium supplemented with 2.0 mM L-glutamine and 1.0 mM sodium pyruvate (Sigma-Aldrich, St. Louis, MO) for 24 h. The conditioned medium containing the secreted proteins was collected and filtered using 0.45 µm filter and subsequently concentrated using a 5000 Da MWCO Amicon Centrifugal Filter Device (Millipore, Bedford, MA). The samples were then concentrated again using a 3000 Da MWCO Microcon Spin Column (Millipore, Bedford, MA). The protein content was determined by Coomassie Plus Protein Assay (Pierce, Rockford, IL). Equal amounts of proteins from the supernatants of each kind of cell were run on a 10% SDS-PAGE gel and stained with Coomassie blue R250. Stable Isotope Labeling and Cells Lyses. The HL-7702 cells were cultured in regular RPMI-1640 medium (Gibco BRL, Grand Island, NY) with 10% dialyzed FBS (PAA Laboratories GmbH, Pasching, Austria). HepG2 and SK-HEP-1 cells were maintained in arginine-deficient MEM medium (PromoCell GmbH, Heidelberg, Germany) supplemented with 10% dialyzed FBS, 2.0 mM L-glutamine, and 1.0 mM sodium pyruvate and containing 0.30 mM medium (13C6-L-Arg hydrochloride) or heavy (13C6,15N4L-Arg hydrochloride) arginine for stable isotope labeling (Cambridge Isotope Labs, Andover, MA), respectively. The cells were grown for at least 5 passages to allow full incorporation of the labeled amino acids. The cells were then harvested using a protein lysis buffer (pH 7.4) containing 50 mM Tris-HCl, 150 mM NaCl, 1% NP-40, 0.1% SDS, and protease inhibitor cocktail (Roche, Mannheim, Germany). siRNA Preparation and Transfection. The AFP siRNA duplex was designed and synthesized by GeneChem Co., Ltd. (Shanghai, China). It corresponded to bases 560-579 from the open reading frame of the human AFP mRNA: 5′-CUU CCU GUA UGC ACC UAC AA-dTdT-3′. The RNA sequence used as a negative control for siRNA activity was 5′-UUC UCC GAA CGU GUC ACG U-dTdT-3′. HepG2 cells, maintained in MEM medium with 10% FBS at ∼50% confluence, were transfected with AFP siRNA (from 20 to 200 nM) using LipofectAMINE 2000 according to the manufacturer’s protocol (Invitrogen, Carslbad, CA). The cells were harvested at 72 h after transfection. RNA Extraction and Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR). When HepG2 was harvested after transfection, total cellular RNA (HepG2) was isolated using an RNeasy Mini Kit (Qiagen, Inc., Valencia, CA) and reversetranscribed using a reverse transcription system to generate cDNA (Promega Corporation, Madison, MI) by following the manufacturer’s protocol. Gene specific primers were then used for semiquantitative PCR amplification to detect the relative amounts of the transcript. The primer sequences were as
TGM2: A Novel Candidate of AFP-Negative Hepatocellular Carcinoma follows: AFP forward, 5′-AAG AGT TGC TAA AGG ATA CC-3′; AFP reverse, 5′-AAG CAA CGA GAA ACG CA-3′; TGM2 forward, 5′- CCT GAC TGA GGA GCA GAA GA-3′; AFP reverse, 5′- CCC TTC ACA GCC TTC AGC-3′; GAPDH forward, 5′-ACC ACA GTC CAT GCC ATC AC-3′; and GAPDH reverse, 5′-TCC ACC ACC CTG TTG CTG TA-3′. The amounts of AFP and TGM2 transcripts were normalized by that of GAPDH. Each primer set was optimized for PCR conditions that would detect the relative differences in samples with known amounts of a specific transcript. The PCR steps were programmed using a GeneAmp PCR System 9700 thermocycler (Applied Biosystems, Inc., Foster, CA). The reactions of AFP and GAPDH were performed at 95 °C for 2 min and followed by 30 cycles at 94 °C for 30 s, 52 °C for 30 s, 72 °C for 45 s and finally, 72 °C for 10 min. The reactions of TGM2 and GAPDH were performed at 95 °C for 2 min and followed by 30 cycles at 94 °C for 1 min, 59 °C for 45 s, 72 °C for 1 min, and finally, 72 °C for 10 min. RT-PCR products were analyzed by agarose gel electrophoresis. SDS-PAGE Separation and MS Identification. Equal amounts of proteins from cells labeled with light, medium, or heavy L-arginine were combined and separated by SDS-PAGE, followed by in-gel tryptic digestion, as described previously.10 LCMS/MS experiments were performed on a hybrid LTQ-FT mass spectrometer (Thermo, Waltham, MA) equipped with a nanospray source and an Agilent 1100 Series binary HPLC system (Agilent, Palo Alto, CA). Peptide mixtures were separated on a fused silica microcapillary column with an i.d. of 75 µm and an in-house prepared needle tip with an i.d. of approximately 15 µm. The columns were packed to a height of 10 cm with a C18 RP resin (SP-300-ODS-AP, 300 Å, 5 µm) (Jinouya Beijing). An electrospray voltage of 2.1 kV was applied. Sample solutions (20 µL, approximately 1 µg/µL or 0.01 µg/µL for dilution experiments) were loaded onto the precolumn and separated by using a mobile phase from 2% acetonitrile (ACN), 98% H2O, 0.1% FA (phase A) and 80% ACN, 20% H2O, 0.1% FA (phase B) and by applying a linear gradient ranging from 2 to 40% buffer B for 90 min at a flow rate of 300 nL/min. The LTQ-FT MS was operated in the data-dependent mode. A scan cycle was initiated with a full-scan survey MS experiment (m/z 400-2000) performed using an FT-ICR MS. The three most abundant ions detected in this scan were subjected to an FT-ICR SIM scan followed by an MS/MS experiment in the LTQ MS. The isolation widths were -10 to 10 Da for SIM scan and ( 2 m/z for the MS/MS experiments. Ions were selected for MS/MS when their intensity exceeded a minimum threshold of 1000 counts. Singly charged ions were not subjected to MS/MS and the exclusion time for a given parent ion was 30 s. Raw data from a completed LC-MS runs were converted into Mascot generic format flat files and then searched with Mascot (Matrix Science) with reference to the human International Protein Index protein sequence database (IPI, version 3.16) with variable modifications, that is, carboxyamidomethylation of cysteine; oxidation of methionine; 13C6-Arg; and 13C6,15N4-Arg. For the search, the peptide mass tolerance was set as 5 ppm and the fragment mass tolerances were set to 0.8 Da. Only proteins containing at least one unique peptide with a Mascot score of over 25 were considered in the list of quantified proteins. Quantification was first carried out using the open-source software MSQuant (http://msquant.sourceforge.net/) and then verified manually. For each protein, the ratio was calculated based on the average of all its quantified peptides. Clinical Specimen Collection and Preparation. Surgical tissue and serum specimens from HCC patients were collected
research articles
after obtaining informed consent and approval from the Institutional Review Board of the Cancer Institute and Hospital of Chinese Academy of Medical Sciences (Beijing, China). A total of 61 HCC tumor and paired adjacent nontumor liver tissue samples, 18 fresh and 47 formalin-fixed paraffin-embedded tissues (there were 4 overlapped cases with both fresh and fixed samples) were collected from HCC patients undergoing resection during the period of May 2004 to November 2006. Out of the 61 specimens, 32 were AFP-normal, while 29 were AFP-positive. The tissue samples were collected and washed right after surgical resection. They were then snap-frozen in liquid nitrogen immediately and stored at -80 °C. Fresh tissue samples were homogenized and the proteins were extracted using the protein lysis buffer described above. The supernatants were collected and the protein concentrations were measured (Coomassie Plus Protein Assay, Pierce, Rockford, IL). Serum samples were obtained after obtaining informed consent from 109 HCC patients (median age, 54 ( 11.7 SD; range 15-79 years) and 42 noncancer healthy controls (30 males and 12 females; mean age, 53 ( 13.3 SD; range 22-68 years). The healthy individuals showed no abnormalities in liver function tests, abdominal ultrasonic, blood routine test, and biochemistry examinations; infections of human immunodeficiency virus (HIV), hepatitis B virus (HBV), and hepatitis C virus (HCV) were excluded. Pathologic diagnosis was independently conducted by two senior pathologists and the histological grade of differentiation was judged by the criteria described previously.12–14 Measurements of hepatitis viral status and AFP level were standardized between studies. The clinical characteristics of these samples are included in Table 1. The clinically acceptable normal serum AFP was defined as
