Proteome Analysis of Gastric Cancer Metastasis by Two-Dimensional Gel Electrophoresis and Matrix Assisted Laser Desorption/Ionization-Mass Spectrometry for Identification of Metastasis-Related Proteins Jie Chen,† Thilo Ka1 hne,‡ Christoph Ro1 cken,§ Tobias Go1 tze,† Jun Yu,| Joseph J. Y. Sung,| Minhu Chen,# Pinjin Hu,# Peter Malfertheiner,† and Matthias P. A. Ebert*,† Department of Gastroenterology, Hepatology and Infectious Diseases, Research Center Immunology/Institute of Experimental Internal Medicine, and Pathology, Otto-von-Guericke University, Magdeburg, Germany; Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong; and Department of Gastroenterology, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China Received April 27, 2004
A well-described animal model was used to understand the molecular mechanisms of carcinogenesis and metastasis of gastric cancer at the protein level. Gastric cancer was induced in 12 Wistar rats by oral administration of N-methyl-N′-Nitro-N-Nitrosoguanidine (MNNG). The protein expression patterns of normal gastric tissue, gastric cancer, and corresponding metastases were analyzed by proteomics in matched tissues of 3 rats. Proteins in the region of molecular masses of 15-75 kDa and an isoelectric point of 3-7 were separated by two-dimensional electrophoresis (2-DE) and identified by peptide fingerprinting with matrix-assisted laser desorption/ionization-time-of-flight-mass spectrometry (MALDITOF-MS). Twenty-seven spots corresponding to 25 different proteins served as landmarks for comparison between tissues. The identified proteins included cytoskeletal proteins, stress associated proteins, proteins involved in signal transduction, cell proliferation and differentiation, and metabolism. Eleven proteins were up-regulated and 2 proteins were down-regulated in tumor tissue when compared with normal tissue. Twelve proteins were up-regulated and 8 proteins were down-regulated in the metastases when compared with the primary tumor. The overexpression of HSP27 in gastric cancer was confirmed by immunohistochemical analysis of human gastric cancer specimens. Combining welldefined animal models with proteome analysis will improve our understanding of the fundamental changes that contribute to the process of carcinogenesis and the formation of metastases in gastric cancer. Keywords: gastric cancer • metastasis • proteome analysis • two-dimensional electrophoresis • matrix-assisted laser desorption/ionization-time-of-flight-mass spectrometry
Introduction Gastric cancer is one of the leading causes of cancer-related death throughout the world.1 Advanced gastric cancer is often accompanied by metastases to the peritoneum, lymph nodes, * To whom correspondence should be addressed. Matthias Ebert, MD, Department of Gastroenterology, Hepatology and Infectious Diseases, Ottovon-Guericke University, Leipziger Str. 44. D-39120 Magdeburg, Germany. Tel: +49-391-6713156. Fax: +49-391-67190054. E-mail: Matthias.Ebert@ medizin.uni-magdeburg.de. † Department of Gastroenterology, Hepatology and Infectious Diseases, Otto-von-Guericke University. ‡ Research Center Immunology/Institute of Experimental Internal Medicine, Otto-von-Guericke University. § Pathology, Otto-von-Guericke University. | Department of Medicine & Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong. # Department of Gastroenterology, The First Affiliated Hospital, Sun YatSen University. 10.1021/pr049916l CCC: $27.50
2004 American Chemical Society
or other organs, resulting in a high mortality rate. The molecular mechanisms underlying carcinogenesis and metastasis of gastric cancer have not been fully clarified. Many different genes and multiple interactions of aberrant expressed genes and proteins are involved in tumor development and progression. cDNA microarrays are able to analyze the expression of groups of genes and to recognize patterns of gene expression in different diseases including gastric cancer.2,3 However, this method has limitations. The level of mRNA expression frequently does not represent the amount of active protein in a cell4 and posttranslational modifications, which may be essential for protein function and activity, are undetectable. Recently, the global analysis of cellular proteins, termed proteomics, has become a key area of developing research in the post-genome era.5 Proteomics uses a combination of sophisticated techniques including two-dimensional gel elecJournal of Proteome Research 2004, 3, 1009-1016
1009
Published on Web 08/03/2004
research articles trophoresis, image analysis, mass spectrometry, amino acid sequencing, and bio-informatics to comprehensively resolve, quantify, and characterize proteins in cells, tissues and animal models. The application of proteomics provides new opportunities to elucidate disease mechanisms and to identify new diagnostic markers and therapeutic targets. The proteomic approach is also being actively applied in cancer research.6 Various human cancers, such as bladder,7 kidney,8 breast,9 lung,10 liver,11 head and neck,12 thyroid,13 ovarian,14 prostate15 and colorectal16,17 cancers, as well as melanoma18 and lymphoma,19 have been analyzed by proteomics analysis. With regard to gastric cancer, proteome analysis has been carried out primarily in gastric cancer cell lines;20,21 Recently, 2-DE maps have also been established for the human stomach22 which give an overview of the proteins expressed in the human stomach and form the basis for subsequent comparative proteome analysis in diseases of the stomach. More recently Ryu et al. analyzed eleven human gastric cancer samples by proteome analysis in order to find biomarkers of gastric cancer.23 To our knowledge, no proteome analysis has been carried out on gastric cancer metastasis until now. The MNNGinduced gastric cancer is a well-established animal model of gastric cancer. It has been shown that this mutagen, when given in drinking water, induces gastric atrophy, intestinal metaplasia, dysplasia, and adenocarcinoma in the pyloric mucosa of Wistar rats.24 The histology of these tumors closely resembles the human counterpart of a well-differentiated gastric adenocarcinoma.25 In this study, we induced gastric cancer in Wistar rats by oral administration of MNNG26 and then compared the protein expression pattern of normal gastric tissue, primary gastric tumor tissue and corresponding metastatic tumor tissues, respectively, by proteome analysis.
Material and Methods Animal Model Construction. Four week-old male Wistar rats with a weight of approximately 60 g were obtained from the Animal Experimental Center of Sun Yat-Sen University.The rats were randomized into two groups: untreated control group (n ) 5) and MNNG treated group (n ) 16). The rats were fed with food and water ad libitum and maintained on hardwood bedding under a 12-hour light/dark cycle. All rats were weighed weekly during the experimental period. Primary gastric adenocarcinoma was induced with MNNG as described previously.27 MNNG was prepared every other day with distilled water to a concentration of 100 µg/mL and was given to rats by drinking water. In addition, 1 mL of 10% sodium chloride was given weekly by oral gavage in the initial 6 weeks to enhance gastric cancer development.28 The Animal Experimentation and Ethics Committee of Sun Yat-Sen University approved all experiments. Rat Tissue Collection. The rats were sacrificed 48 weeks after tumor induction treatment. Rats were fasted for 12 h and were anaesthetized with ether. The stomach was cut along the greater curvature and carefully examined. Other organs including peritoneum, liver, and lung were also carefully inspected, and sliced, to find any metastasis. Tissue samples from the gastric tumor, corresponding nontumorous stomach and the corresponding hematogenous metastases were obtained, where present. Tissue samples were snap frozen in liquid nitrogen and subsequently stored at -80 °C for proteome analysis or fixed in 10% buffered formalin for histological processing. In all cases with tumor cells (either primary cancer or metastasis), histology confirmed that at least 75% of the tissue samples used for proteomics contained cancer cells. 1010
Journal of Proteome Research • Vol. 3, No. 5, 2004
Chen et al.
For our proteome analysis we selected cancers which exhibited similar histomorphological characteristics and which gave adequate protein for 2D analysis. From the available primary and metastatic cancers, we picked three cases of gastric cancers which were all well-differentiated and in which all three components, i.e., normal gastric mucosa, primary gastric cancer, and distant metastasis were preserved and frozen for further proteome analysis. In total, three animals were analyzed, from which enough frozen tissue from the normal nonmalignant gastric mucosa, the primary gastric cancer and the metastasis was available. Each of the 2D analyses was performed at least twice under identical conditions. Reproducibility in each case was excellent. Only proteins which were aberrantly expressed in 2 of the three animals were picked for further identification. Furthermore, the spots of interest were excised from the gel and pooled from the replicates before digestion and MS analysis. In addition, to further increase the accuracy of our analysis and to prevent experimental artifacts, we only analyzed spots that exhibited identical intensity within the replicates. Human Gastric Tissue Samples. Human tissues samples used in the present study were obtained from 12 patients (6 men, 6 women), diagnosed with intestinal-type (6 patients), or diffuse-type (6 patients) carcinomas. The age of patients ranged from 59 to 81 years (mean 67.2 ( 9.1 years). Tissue samples (gastric cancer, corresponding nonneoplastic tissue and lymph node metastases) had been obtained after gastrectomy and were processed as described below. Follow-up data on these gastric cancer patients was not available. Histology. All tissue specimens were fixed in 10% neutralized formalin, dehydrated in an increasing series of alcohols and xylol, and embedded in paraffin. After deparaffinisation, the 4-µm thick sections were stained with haematoxylin and eosin. The diagnosis and grading of gastric cancer was established on the criteria given in by the WHO Classification of gastrointestinal tumors29 Immunohistochemistry. Deparaffinized sections were stained using hematoxylin and eosin (H&E). For immunostaining of paraffin-embedded sections, the slides were deparaffinized and rehydrated in a graded alcohol series, and pretreated with 10 mM sodium citrate (3 × 10 min, 600 W microwave oven). Immunostaining was performed with a monoclonal antibody directed against human HSP27 (clone F-4; dilution 1:200). Incubation with the primary antibody was performed in a moist chamber at 37 °C for 1 h. Polyvalent anti-mouse IgG (30 min, room temperature; Immunotech, Marseilles, France) served as a secondary antibody. Slides were washed between steps with Tris-buffered saline. Immunoreactions were visualized via an avidin-biotin complex, using the Vectastain ABC alkaline phosphatase kit (distributed by CAMON, Wiesbaden, Germany). Fast red/Naphthol Mx (Immunotech, Marseille, France) served as chromogen. The specimens were counter-stained with hematoxylin. 2D-Gel-Electrophoresis. 25 mg of tissue sample was crushed under liquid nitrogen to a fine powder. The powder was dissolved in 750 µL lysis buffer (7 M urea, 2 M thiourea, 4% chaps, 50 mM DTT, 0.1% SDS, 2% Pharmalyte pH 3-10). After 30 min, at room temperature the samples were centrifuged at 100.000 × g for 20 min. The supernatants were diluted in equal amounts of rehydration buffer (7 M urea, 1% chaps, 50 mM DTT, and 1% Pharmalyte pH 3-10). 500 µL of each preparation were used to rehydrate an IPG-strip (24 cm, pH 3-7, Amersham Biosciences, Freiburg), respectively. After 12 h of rehydration,
Proteome Analysis for Gastric Cancer Metastasis
the strips were transferred to a dry-strip unit on a MultiphorElectrophoresis apparatus (Amersham Biosciences, Freiburg). Isoelectric focusing was performed at constant power (10 µA/ IPG-strip) at 500 V for 12 h and 3500 V for 90 h. Thereafter, the strips were equilibrated in 50 mM TRIS/HCl pH 8.8, 6 M urea, 30% glycerol, 2% SDS and 10 mM DTT for 10 min, washed with 50 mM TRIS/HCl pH 8.8, 6 M urea, 30% glycerol, 2% SDS and 200 mM iodoacetamide for another 10 min, transferred to the top of SDS-gradient gels (10%-16%), and finally embedded in low melting agarose. Gels for comparison were run simultaneously in a Hoefer IsoDalt electrophoresis unit at constant power (20 mA per Gel) overnight. Protein spots were visualized using a silver stain according to the protocols described by Blum et al.30 The amount of protein was originally quantified by the PlusOne 2D-Quant-Kit (Amersham Bioscience), and the sample load was adjusted accordingly. Tryptic Digestion and Mass Spectrometry (Peptide-MassFingerprinting). Protein spots of interest were excised from the gels after computer-aided comparison (Image Master Software, Amersham Biosciences, Freiburg). In-gel digestion was performed in an adapted manner according to Shevchenko et al..31 Gel pieces were washed by repeated addition and removing of 0.1 M NH4HCO3 and acetonitrile, respectively. Subsequently, the gel particles were dried in a vacuum centrifuge, rehydrated in a freshly prepared digestion buffer containing 50 mM NH4HCO3 and 12.5 ng/µL of trypsin (Boehringer Mannheim, modified, sequencing grade) and incubated at 37 °C overnight. The peptides were extracted from the gel by repeated addition of a sufficient volume of 25 mM NH4HCO3 and acetonitrile. The extraction was forced by sonification. All extracts were pooled and dried in a vacuum centrifuge. For mass spectrometric peptide mapping the peptides were redissolved in 5 µL 0.1% TFA and purified on a 200 nl reversedphase (C18)- nanocolumn. Peptides were eluted in 5 µL 70% (v/v) acetonitrile and subsequently co-crystallized with R-cyano-4-hydroxycinnamic acid (20 mg/mL) in 70% acetonitrile on a SCOUT 384-600 µm anchor-Target. The mass spectrometry was performed on a MALDI-TOF-MS (Reflex III, Bruker Daltonics, Germany) in reflector mode with external calibration. Annotation of the tryptic fragments was done using the XMASS and BioTools 2.0 software (Bruker Daltonics, Germany). The ProFound-Software (www.prowl.rockefeller.edu) was used for the database search. Spot intensity was assessed with the Image-Master-2-D Elite Software (Amersham Bioscience). Due to the low dynamic range of the silver staining, the differences in protein expression were divided into the following groups: “+” ) 5-30% difference in protein expression; “++” ) 3160%; “+++” ) 61-95%; “-“ ) no difference in protein expression detectable.
Results Animal Model. Twelve of 16 (75%) rats treated with MNNG had developed gastric cancer (Figure 1A), whereas 6 had additionally liver or lung metastases (Figure 1B). Histologically, 10 well-differentiated and 2 poorly differentiated adenocarcinomas arose in 12 stomach cancer-bearing animals in the MNNG treated group (Figure 1C). Gastric cancer was not observed in any of the rats from the untreated control group. For proteome analysis, only well-differentiated cancers were chosen in which all components, i.e., normal gastric mucosa, gastric cancer and metastasis were preserved and were frozen for proteome analysis. Proteome Analysis. 2-DE with immobilized pH gradients was used to study proteins extracted from gastric tumor tissue
research articles
Figure 1. Panel A, the stomach of one MNNG treated rat was cut along the large curvature, arrow indicated an elevated lesion with a central depression in the cardia of the stomach. Panel B, liver of the same rat, arrows indicated two metastatical tumors in the liver. Panel C, histology of the stomach tumor showed well differentiated adenocarcinoma.
(T), adjacent nontumor tissue (N) and the metastases (M) of the same rats. In total, 3 rats were analyzed in which all three components were readily available. To ensure reproducibility, every sample was studied at least twice and the same protein pattern was obtained. In our study, the analysis was carried out on soluble-fraction proteins with isoelectric point (pI) values in the range of 3-7 and molecular masses in the range of 15-75 kDa, because high quality 2-DE separations are more easily obtained for these proteins. When the protein pattern of primary tumor, corresponding nontumor tissue and metastasis were compared, we found multiple proteins to be differentially expressed. Figure 2A,B shows 2-DE patterns obtained from the tumor, corresponding nontumor and the metastasis. Twenty-seven spots representing 25 different proJournal of Proteome Research • Vol. 3, No. 5, 2004 1011
research articles
Chen et al.
Table 1. Protein Identified from 2-DE Gel by Peptide Mass Fingerprinting spot no.
proteins identified
1 similar to transient receptor protein 2 2 hypothetical protein 3 catecholamine-Omethyltransferase 4 calreticulin 5 ATP synthase β subunit 6 protein disulfide isomerase 7 R-1-antitrypsin precursor 8 desmin 9 tropomyosin β 2 10 tropomyosin β 2 11 tropomyosin β 2 12 tropomyosin R isoform 13 heat shock 27 kDa protein 14 haptoglobin precursor 15 phosphatase and tensin homolog 16 albumin 17 hypothetical protein 18 similar to ornithine decarboxylase 19 β-actin (aa 27-375) 20 similar to Sad 1/unc-84-like protein 21 glutathione S-transferase P (Chain 7) 22 olfactory receptor 23 hypothetical protein 24 similar to DJ-1 protein 25 hypothetical protein 26 peroxiredoxin 5 27 3-mercaptopyruvate sulfurtransferase
NCBI database theor. mol. accession no. mass(Da)/pI
XP•130617
19.89/4.8
XP•159481 NP•036663
17.88/6.8 29.81/5.4
NP•031617 AAB02288 NP•037130 AAA40788 NP•071976 A44131 A44131 P58775 AAK54242 P42930 P06866 NP•032986 NP•599153 XP•143995 XP•144851 CAA27396 XP•128279
48.15/4.3 51.18/4.9 57.25/4.8 45.99/5.7 53.50/5.2 32.90/4.6 32.90/4.6 32.93/4.7 32.77/4.7 22.93/6.1 39.04/6.1 47.70/5.9 70.70/6.1 28.35/6.8 26.53/6.0 39.45/5.8 35.89/5.6
P04906
23.65/6.9
AAC64584 XP•155778 AAH02187 XP•157422 NP•446028 NP•620198
24.72/6.8 31.10/7.0 20.23/6.3 21.70/7.0 24.86/5.6 33.21/5.9
teins were identified by MALDI-TOF-MS and served as landmarks for the comparison between the tissues, as shown in Table 1. Diverse groups of proteins were found, such as cytoskeletal proteins (actin, tropomyosin, and desmin), stress associated proteins (heat shock 27 kDa protein, calreticulin, and disulfide isomerase), serum proteins (albumin, haptoglobin precursor, and R-1-antitrypsin precursor), proteins involved in signal transduction, cell proliferation and differentiation (similar to transient receptor protein 2, phosphatase and tensin homolog, olfactory receptor, peroxiredoxin 5, similar to DJ-1 protein), and metabolism proteins (catecholamine-O-methyltransferase, similar to ornithine decarboxylase, glutathione S-transferase protein, 3-mercaptopyruvate sulfurtransferase and ATP synthase beta subunit). Several hypothetical proteins with unknown function were also identified and further studies will be necessary to clarify their role. Although 24 out of 25 proteins were present as single spots on the 2-DE gel, tropomyosin β 2 was present in 3 adjacent spots (spots 9, 10, and 11); this phenomenon is likely due to posttranslational modifications. The different expression levels of the proteins in gastric tumor, corresponding nontumor tissue and metastasis are listed in Table 2. Due to the low dynamic range and the variability of spot intensities of silver stained protein spots, we could not quantify the protein expression by intensity scan. Therefore, we divided the different spot intensities semiquantitatively as +, ++, and +++ (see Materials and Methods for further details). Of the 25 differentially expressed proteins, only 2, i.e., heat shock 27 kDa protein (Hsp27) and phosphatase and tensin homologue (PTEN), were previously reported in gastric cancer.23,32-34 So far, the other 23 proteins have not been linked to gastric cancer and its metastatic behavior. The protein 1012
Journal of Proteome Research • Vol. 3, No. 5, 2004
Table 2. Different Expression Level of Selected Proteins in Non-tumor Tissue (N), Tumor Tissue (T), and Metastasis Tumor Tissue (M)a spot no.
proteins
N
T
M
3 6 7 13 14 18 19 21 23 24 25 8 16 17 20 1 2 15 22 26 27 4 5 9 10 11 12
catecholamine-O-methyltransferase protein disulfide isomerase R-1-antitrypsin precursor heat shock 27 kDa protein haptoglobin precursor similar to ornithine decarboxylase β-actin (aa 27-375) glutathione S-transferase P (Chain 7) hypothetical protein similar to DJ-1 protein hypothetical protein desmin albumin hypothetical protein similar to Sad 1/unc-84-like protein similar to transient receptor protein 2 hypothetical protein phosphatase and tensin homolog olfactory receptor peroxiredoxin 5 3-mercaptopyruvate sulfurtransferase calreticulin ATP synthase β-subunit tropomyosin β 2 tropomyosin β 2 tropomyosin β 2 tropomyosin R isoform
+ + + + ++ ++ ++ ++ ++ ++ +++ +++ +++ +++ +++ +++
+ ++ ++ ++ ++ ++ + +++ ++ ++ + +++ ++ ++ ++ ++ + +++ +++ +++ +++
+++ ++ ++ +++ +++ +++ ++ ++ ++ ++ +++ ++ +++ +++ +++ -
a + for weak, ++ for medium, and +++ for high expression, - means not found.
expression patterns of nontumorous gastric tissue were compared with gastric cancer, and showed the up-regulation of 11 proteins and the down-regulation of 2 proteins in the tumor. The comparison of the primary tumor with the metastasis showed an up-regulation of 12 proteins and a down-regulation of 8 proteins in the metastasis. Expression of Human HSP27 in Gastric Cancer. Immunohistochemistry was performed on paraffin-embedded sections from 12 patients with metastasized gastric cancer. In the nontumorous gastric mucosa HSP27 was found in gastric glands of the corpus. It was not expressed in nonregenerative, nontumorous foveolar epithelium. Cytoplasmic expression of HSP27 was found in 12 of 12 (100%) gastric carcinomas and 12 of 12 (100%) lymph node metastases. In general, the intensity of immunostaining and the number of immunoreactive cells was similar, when lymph node metastases were compared with the corresponding primary tumor (Figure 3).
Discussion Carcinogenesis and metastasis of gastric cancer is a complex multistep process involving tumor initiation, promotion, uncontrolled proliferation, angiogenesis, invasion, and colony formation at other organs. In this study, we attempted to elucidate the molecular mechanisms underlying the carcinogenesis and metastasis of gastric cancer by a global proteome analysis. Although cell lines have been extensively used as experimental models to study the genetics, pharmacology and biochemistry of cancer cells, the loss of the interaction between the tumor cell and the host influences the study results. Direct analysis of clinical samples provides us with more reliable information, but the heterogeneous nature of clinical samples requires analysis of a large number of samples, to guarantee a statistical significance. Therefore, the study of an animal model
Proteome Analysis for Gastric Cancer Metastasis
research articles
Figure 2. A and B. Overview of 2-DE maps of rat gastric non-tumor tissue, gastric tumor tissue, and metastasis tumor tissue. Staining was done with silver nitrate. Arrows indicated proteins identified by MALDI-TOF-MS. The protein name and NCBI accession number are listed in Table 1. A/B, 26: peroxiredoxin 5, 28: not identified, 22: olfactory receptor, 20: similar to SAD1/unc-84, 55: not identified, 56: not identified, 17: hypothetical protein, 1: similar to transient receptor protein 2, 3: catecholamin-O-methyltransferase, 35: not identified, 31: not identified, 6: protein disulfide isomerase, 4: calreticulin.
became an optional strategy in our proteomics approach. The MNNG-rat gastric cancer model is a well-established animal model.24-26 Oral administration of MNNG to Wistar rats induces intestinal metaplasia and adenocarcinoma in the stomach.35,36 In this study, we successfully constructed this gastric cancer model. Tissue samples from gastric cancer, corresponding nonneoplastic gastric tissue and metastases were obtained from the same rat for proteome analyses. These samples enabled us to study the differential protein expression of metastatic
gastric cancer under a similar genomic background. To the best of our knowledge, this is the first proteome analysis of gastric cancer in an animal model until now. Twenty-five proteins in the molecular mass range of 15 to 75 kDa and with an isoelectric point between 3 and 7 were identified by 2-DE and MALDI-TOF-MS. These served as landmarks for comparison between tissues. Several groups of proteins including cytoskeletal proteins, stress associated proteins, proteins involved in signal transduction, cell proliferation and differentiation, and Journal of Proteome Research • Vol. 3, No. 5, 2004 1013
research articles
Figure 3. Immunohistochemical Expression of HSP27 in Human Gastric Cancer. The distribution and expression pattern of HSP27 in gastric carcinomas was investigated by immunohistochemistry. Nontumorous epithelium, tumor and lymph node metastases were stained with anti-HSP27 antibodies. HSP27 was found in the cytoplasm of tumor cells of diffuse type (A, B) and intestinal type (C, D) gastric cancer. HSP27 was not expressed in nontumorous and nonregenerative foveolar epithelium (A). Note expression of HSP27 in lymph node metastases (B, D). Hematoxylin counterstain; Original magnifications: ×400 (A, B) amd ×200 (C, D).
finally metabolism are likely to contribute to the complexity of gastric cancer carcinogenesis and metastasis. Hsp27, protein disulfide isomerase (PDI), desmin and catecholamine-O-methyltransferase (COMT) were up-regulated in the tumor. Hsp27 is a molecular chaperone with an ability to interact with a large number of proteins. It is involved in a wide variety of physiological and pathological processes. Under the influence of nonphysiological conditions, heat shock protein synthesis is accelerated to aid cell survival. Thus, over-production of heat shock proteins could protect malignantly transformed cells from apoptotic cell death and fosters resistance to chemotherapeutic agents and irradiation.37,38 Higher expression of Hsp27 protein had been previously reported in liver, breast, colon and gastric cancer.23,32,39-41 Its expression is associated with poor prognosis. The differential expression of HSP27 in rat gastric cancer was confirmed in the human counterpart. Similar to the rat, we found HSP27 in human gastric cancer, whereas nontumorous, nonregenerative human foveolar epithelium did not express HSP27. We then studied a series of lymph node metastases, and found HSP27 immunohistochemically also in the lymph node metastases of human gastric cancer. This result contrasts with the absence of HSP27 expression in the distant metastases of the animal model. The discrepancy may be either related to the different nature of 1014
Journal of Proteome Research • Vol. 3, No. 5, 2004
Chen et al.
metastases that were compared (lymph node metastasis versus distant hematogenic metastases) or by the higher sensitivity of the antibody-based detection of HSP27 in immunohistochemistry. PDI belongs to the protein disulfide isomerase family, which catalyzes the rearrangement of both intrachain and interchain disulfide bonds in proteins and is found to function in protein folding in vivo.42,43 PDI is a stress-associated protein and was reported to be up-regulated in hypoxia-insulted glial cells.44 Previous 2-DE studies revealed that PDI was also overexpressed in human breast, liver, and gastric cancer.9,11,23 In our study, PDI was found to be up-regulated in both primary and metastatic gastric cancer. Desmin is a muscle-specific intermediate filament protein and was reported as a histologic marker of tumors with muscle origin.45 Desmin was also found in some spontaneous histiocytic tumors of the rat.46 In this study, we found an up-regulation of desmin in the primary gastric cancer tissue, which than disappeared in the metastasis. As a cytoskeletal protein, the different expression levels of desmin in primary and metastatic tumor tissue implicate a role in carcinogenesis and metastasis of gastric cancer. The oxidative metabolism of 17-beta-estradiol (E2) and estrone (E1) to catechol estrogens (2-OHE2, 4-OHE2, 2-OHE1, and 4-OHE1) and estrogen quinones has been postulated to be a factor in mammary carcinogenesis. Catechol-O-methyltransferase (COMT) catalyzes the methylation of catechol estrogens to methoxy estrogens, which simultaneously lowers the potential for DNA damage and increases the concentration of 2-methoxyestradiol (2-MeOE2), an antiproliferative metabolite. COMT plays an important tumor-suppressive role in estrogen-induced cancers.47 COMT was also reported to be down-regulated in liver tumors.11 High expression of COMT in gastric cancer contrasts with the previous reports. This deserves further studies on a larger set of clinical specimens in order to obtain more detailed information about the role of COMT in gastric cancer pathogenesis. Calreticulin and ATP synthase beta subunit were downregulated in the primary tumor. Calreticulin is a Ca2+ binding protein with multiple functions, diverse cellular locations, and putative isoforms. Calreticulin and calreticulin fragments inhibit angiogenesis by directly targeting endothelial cells, and ultimately suppress tumor growth.48,49 Calreticulin-integrin surface complex functions as a symbiotic unit, transmitting information in both directions across the plasma membrane.50 Calreticulin is also a component of the nuclear matrix and was also found to be in the nuclear matrixes of various carcinoma cell lines.51 The formation and/or expansion of the calreticulinnuclear matrix may be related to the activated cell growth. Calreticulin was reported to be up-regulated in liver 39,51 and prostate cancer,52 whereas in our study the expression of this protein was decreased in both primary and metastatic cancer. These contradictory results may be explained by the diverse cellular locations and complex biological functions of calreticulin. Mitochondrial H+ATP synthase is required for cellular energy provision and for efficient execution of apoptosis. A previous study indicated that the expression of ATP synthase beta subunit (β-F1-ATPase) of mitochondria in carcinomas of the human liver, kidney, and colon was decreased significantly53 and demonstrated an impaired mitochondrial function of cancer cells. ATP synthase β subunit was also down-regulated in our model. The identification of proteins, such as the olfactory receptor, which have not been previously reported in studies using
Proteome Analysis for Gastric Cancer Metastasis
gastric cancer tissues, may reflect changes of gene expression in epithelial cells due to the transformation of these cells and the associated changes in epithelial differentiation. These changes are probably a result of the process of dedifferentiation that is a key feature of epithelial cell transformation. A further interesting finding of our study was that 3 proteins, phosphatase and tensin homologue (PTEN), tropomyosin β 2 and tropomyosin R isoform, were dramatically decreased in metastatic tumor tissues, while expressed at similar level in gastric cancer and corresponding nontumor tissue. PTEN is a dualspecificity phosphatase implicated in embryonic development, intestinal cell proliferation and differentiation, as well as tumor suppression,54 previously we reported reduced PTEN expression in gastric cancer and in the gastric mucosa of gastric cancer relatives.33 Tropomyosins (TMs) are a family of microfilament binding proteins, which are suppressed in the transformed cells. Decrease of tropomyosin β 2 and tropomyosin R isoform had been reported in human breast and prostate carcinoma.15,52,55 In the present study, we did not find a decrease of PTEN and TMs in rat gastric tumor tissue, but these proteins were dramatically decreased in metastatic tumor tissues. These results indicate that these proteins might contribute to the metastasis of gastric cancer rather than carcinogenesis. In conclusion, by using a well-established rat gastric cancer model, we analyzed the protein expression pattern of normal gastric tissue, primary gastric tumor tissue and corresponding metastatic tumor tissue and identified a series of proteins related to the carcinogenesis and metastasis of gastric cancer by proteome analysis. This is the first report of proteome analysis in gastric cancer metastasis. Further functional and clinical sample analysis of these cancer-associated proteins are necessary to elucidate their precise role in the process of gastric carcinogenesis and the formation of metastasis.
Acknowledgment. M.P.A.E. is supported by a grant from the Deutsche Forschungsgemeinschaft (Eb 187/4-1) and a Heisenberg-Stipend (Eb 187/5-1). This work was also supported by a grant awarded from the Land Sachsen-Anhalt to M.P.A.E. (3488A/0103M). References (1) Pisani, P.; Parkin, D. M.; Bray, F.; Ferlay, J. Int. J. Cancer 1999, 83, 18-29. (2) Boussioutas, A.; Li, H.; Liu, J.; Waring, P.; Lade, S.; Holloway, A. J.; Taupin, D.; Gorringe, K.; Haviv, I.; Desmond, P. V.; Bowtell, D. D. Cancer Res. 2003, 63, 2569-2577. (3) Hippo, Y.; Taniguchi, H.; Tsutsumi, S.; Machida, N.; Chong, J. M.; Fukayama, M.; Kodama, T.; Aburatani, H. Cancer Res. 2002, 62, 233-240. (4) Anderson, L., Seilhamer, J. Electrophoresis 1997, 18, 533-537. (5) Chambers, G.; Lawrie, L.; Cash, P.; Murray, G. I. J Pathol. 2000, 192, 280-288. (6) Alaiya, A. A.; Franzen, B.; Auer, G.; Linder, S. Electrophoresis 2000, 21, 1210-1217. (7) Celis, J. E.; Wolf, H.; Ostergaard, M. Electrophoresis 2000, 21, 2115-2121. (8) Sarto, C.; Marocchi, A.; Sanchez, J. C.; Giannone, D.; Frutiger, S.; Golaz, O.; Wilkins, M. R.; Doro, G.; Cappellano, F.; Hughes, G.; Hochstrasser, D. F.; Mocarelli, P. Electrophoresis 1997, 18, 599604. (9) Bini, L.; Magi, B.; Marzocchi, B.; Arcuri, F.; Tripodi, S.; Cintorino, M.; Sanchez, J. C.; Frutiger, S.; Hughes, G.; Pallini, V.; Hochstrasser, D. F.; Tosi, P. Electrophoresis 1997, 18, 2832-2841. (10) Okuzawa, K.; Franzen, B.; Lindholm, J.; Linder, S.; Hirano, T.; Bergman, T.; Ebihara, Y.; Kato, H.; Auer, G. Electrophoresis 1994, 15, 382-390.
research articles (11) Kim, J.; Kim, S. H.; Lee, S. U.; Ha, G. H.; Kang, D. G.; Ha, N. Y.; Ahn, J. S.; Cho, H. Y.; Kang, S. J.; Lee, Y. J.; Hong, S. C.; Ha, W. S.; Bae, J. M.; Lee, C. W.; Kim, J. W. Electrophoresis 2002, 23, 41424156. (12) Wu, W.; Tang, X.; Hu, W.; Lotan, R.; Hong, W. K.; Mao, L. Clin. Exp. Metastasis 2002, 19, 319-326. (13) Srisomsap, C.; Subhasitanont, P.; Otto, A.; Mueller, E. C.; Punyarit, P.; Wittmann-Liebold, B.; Svasti, J. Proteomics 2002, 2, 706-712. (14) Alaiya, A. A.; Franzen, B.; Fujioka, K.; Moberger, B.; Schedvins, K.; Silfversvard, C.; Linder, S.; Auer, G. Int. J. Cancer 1997, 73, 678-683. (15) Alaiya, A. A.; Oppermann, M.; Langridge, J.; Roblick, U.; Egevad, L.; Brindstedt, S.; Hellstrom, M.; Linder, S.; Bergman, T.; Jornvall, H.; Auer, G. Cell. Mol. Life Sci. 2001, 58, 307-311. (16) Cole, A. R.; Ji, H.; Simpson, R. J. Electrophoresis 2000, 21, 17721781. (17) Minowa, T.; Ohtsuka, S.; Sasai, H.; Kamada, M. Electrophoresis 2000, 21, 1782-1786. (18) Sinha, P.; Kohl, S.; Fischer, J.; Hutter, G.; Kern, M.; Kottgen, E.; Dietel, M.; Lage, H.; Schnolzer, M.; Schadendorf, D. Electrophoresis 2000, 21, 3048-3057. (19) Muller, E. C.; Schumann, M.; Rickers, A.; Bommert, K.; WittmannLiebold, B.; Otto, A. Electrophoresis 1999, 20, 320-330. (20) Sinha, P.; Poland, J.; Schnolzer, M.; Celis, J. E.; Lage, H. Electrophoresis 2001, 22, 2990-3000. (21) Tomlinson, A. J.; Hincapie, M.; Morris, G. E.; Chicz, R. M. Electrophoresis 2002, 23, 3233-3240. (22) Ha, G. H.; Lee, S. U.; Kang, D. G.; Ha, N. Y.; Kim, S. H.; Kim, J.; Bae, J. M.; Kim, J. W.; Lee, C. W. Electrophoresis 2002, 23, 25132524. (23) Ryu, J. W.; Kim, H. J.; Lee, Y. S.; Myong, N. H.; Hwang, C. H.; Lee, G. S.; Yom, H. C. J. Korean Med. Sci. 2003, 18, 505-509. (24) Sugimura, T.; Fujimura, S.; Baba, T. Cancer Res. 1970, 30, 455465. (25) Ushijima, T.; Yamamoto, M.; Suzui, M. Cancer Res. 2000, 60, 1092-1096. (26) Ohgaki, H.; Kleihues, P.; Sugimura, T. Ital. J. Gastroenterol. 1991, 23, 371-377. (27) Sugimura, T.; Fujimura, S. Nature 1967, 216, 943-944. (28) Tatematsu, M.; Ozaki, K.; Mutai, M.; Shichino, Y.; Furihata, C.; Ito, N. Carcinogenesis 1990, 11, 1975-1978. (29) Jass, J. R.; Sobin, L. H.; Watanabe, H. Cancer 1990, 66, 21622167. (30) Blum, H.; Beier, H.; Gross, H. J. Electrophoresis 1987, 8, 93-99. (31) Shevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Anal. Chem. 1996, 68, 850-858. (32) Kapranos, N.; Kominea, A.; Konstantinopoulos, P. A.; Savva, S.; Artelaris, S.; Vandoros, G.; Sotiropoulou-Bonikou, G.; Papavassiliou, A. G. J. Cancer Res. Clin. Oncol. 2002, 128, 426-432. (33) Fei, G.; Ebert, M. P.; Mawrin, C.; Leodolter, A.; Schmidt, N.; Dietzmann, K.; Malfertheiner P. Eur. J. Gastroenterol. Hepatol. 2002, 14, 297-303. (34) Kang, Y. H.; Lee, H. S.; Kim, W. H. Lab. Invest. 2002, 82, 285291. (35) Sasajima, K.; Kawachi, T.; Matsukura, N.; Sano, T.; Sugimura, T. J. Cancer Res. Clin. Oncol. 1979, 94, 201-206. (36) Matsukura, N.; Kawachi, T.; Sasajima, K.; Sano, T.; Sugimura, T.; Hirota, T. J. Natl. Cancer Inst. 1978, 61, 141-144. (37) Concannon, C. G.; Gorman, A. M.; Samali, A. Apoptosis 2003, 8, 61-70. (38) Witkin, S. S. Eur. J. Gynaecol. Oncol. 2001, 22, 249-256. (39) Yu, L. R.; Zeng, R.; Shao, X. X.; Wang, N.; Xu, Y. H.; Xia, Q. C. Electrophoresis 2000, 21, 3058-3068. (40) Garrido, C.; Fromentin, A.; Bonnotte, B.; Favre, N.; Moutet, M.; Arrigo, A. P.; Mehlen, P.; Solary, E. Cancer Res. 1998, 58, 54955499. (41) Lemieux, P.; Oesterreich, S.; Lawrence, J. A.; Steeg, P. S.; Hilsenbeck, S. G.; Harvey, J. M.; Fuqua, S. A. Invasion Metastasis 1997, 17, 113-123. (42) Wang, C. C. Methods Enzymol. 2002, 348, 66-75. (43) Gonzlez, R.; Andrews, B. A.; Asenjo, J. A. J Theor. Biol. 2002, 214, 529-537. (44) Tanaka, S.; Uehara, T.; Nomura, Y. J Biol Chem. 2000, 275, 10 388-10 393. (45) Osborn, M. J Invest. Dermatol. 1983, 81, 104s-109s. (46) Wright, J. A.; Goonetilleke, U. R.; Waghe, M.; Horne, M.; Stewart, M. G. J. Comp. Pathol. 1991, 104, 223-232.
Journal of Proteome Research • Vol. 3, No. 5, 2004 1015
research articles (47) Dawling, S.; Roodi, N.; Mernaugh, R. L.; Wang, X.; Parl, F. F. Cancer Res. 2001, 61, 6716-6722. (48) Pike, S. E.; Yao, L.; Setsuda, J.; Jones, K. D.; Cherney, B.; Appella, E.; Sakaguchi, K.; Nakhasi, H.; Atreya, C. D.; Teruya-Feldstein, J.; Wirth, P.; Gupta, G.; Tosato, G. Blood 1999, 94, 2461-2468. (49) Pike, S. E.; Yao, L.; Jones, K. D.; Cherney, B.; Appella, E.; Sakaguchi, K.; Nakhasi, H.; Teruya-Feldstein, J.; Wirth, P.; Gupta, G.; Tosato, G. J. Exp. Med. 1998, 188, 2349-2356. (50) Zhu, Q.; Zelinka, P.; White, T.; Tanzer, M. L. Biochem. Biophys. Res. Commun. 1997, 232, 354-358. (51) Yoon, G. S.; Lee, H.; Jung, Y.; Yu, E.; Moon, H. B.; Song, K.; Lee, I. Cancer Res. 2000, 60, 1117-1120.
1016
Journal of Proteome Research • Vol. 3, No. 5, 2004
Chen et al. (52) Alaiya, A.; Roblick, U.; Egevad, L.; Carlsson, A.; Franzen, B.; Volz, D.; Huwendiek, S.; Linder, S.; Auer, G. Anal. Cell. Pathol. 2000, 21, 1-9. (53) Cuezva, J. M.; Krajewska, M.; de Heredia, M. L.; Krajewski, S.; Santamaria, G.; Kim, H.; Zapata, J. M.; Marusawa, H.; Chamorro, M.; Reed, J. C. Cancer Res. 2002, 62, 6674-6681. (54) Kim, S.; Domon-Dell, C.; Wang, Q.; Chung, D. H.; Di Cristofano, A.; Pandolfi, P. P.; Freund, J. N.; Evers, B. M. Gastroenterology 2002, 123, 1163-1178. (55) Bharadwaj, S.; Prasad, G. L. Cancer Lett. 2002, 183, 205-213.
PR049916L
