Expression of peroxisome proliferator-activated receptor g, E-cadherin and matrix metalloproteinases-2 in gastric carcinoma and lymph node metastases

Gastric cancer remains the second most common cause of cancer death in the world although its incidence is decreasing.¹ The major cause of death from gastric cancer is metastasis that is usually resistant to the conventional treatment. Mechanisms by which primary tumors invade local tissues and spread to distant sites are beginning to be understood. The relevant factors include cell-to-cell and cell-to-extracellular matrix interactions, motility factors and genetic controls.² The investigation about regulatory and migratory factors associated with this process might help to elucidate the mechanisms of gastric carcinogenesis and metastases and to prevent these processes.

Peroxisome proliferator-activated receptors (PPARs) are members of the nuclear hormone receptor family. The PPARs function as ligand-activated transcription factors forming heterodimer with the retinoid X receptor (RXR). Upon ligand binding, the complex of PPAR and RXR binds to specific recognition sites on DNA, the peroxisome proliferator response elements (PPREs) and regulates transcription of specific genes. Three major subtypes of PPARs have been described, PPARα, PPARb/d and PPARg, which are involved in lipid metabolism, cellular homeostasis, inflammation control, and wound healing. It has been demonstrated that PPARg also participates in the processes of cellular proliferation, differentiation and apoptosis. PPARg is predominantly expressed in adipose tissue and functions as a key trigger of adipocyte differentiation. PPARγ is expressed in various organs including adipose tissue, breast epithelium, small intestine, lungs, liver and some other tumor tissues. However the expression levels of PPARγ in different tumor cells are different compared with corresponding normal tissues. Activation of PPARg has recently been demonstrated to inhibit various tumor cells growth, progression and metastasis.³

E-cadherin is a 120 kDa glycoprotein taking part in calcium-mediated cellular adhesion. It is localized at the epithelial junction complex (α-, β- and γ-catenin) and is responsible for the organization, maintenance and morphogenesis of epithelial tissues. Loss or reduction of E-cadherin mediated cellular adhesion is an important step in the development of invasion and metastasis in multiple carcinomas, including gastric carcinoma.⁴ Recent studies have shown that E-cadherin actually plays an early and important role in the control of carcinogenesis and acts as a tumor suppressor gene, particularly in gastric cancer.⁵

Matrix metalloproteinases (MMPs) are zinc proteinases responsible for the degradation of extracellular matrix macromolecules in pathophysiological conditions such as tissue remodeling and tumor invasion. MMPs may be related to the invasiveness of tumor cells and patients' survival from carcinomas. MMP-2 (gelatinase A; 72-kDa gelatinase; type IV collagenase) is an important enzyme of the MMP family, which is able to degrade collagen IV, a basic component of constitutive basement membranes. MMP-2 is a prerequisite for metastasizing tumor cells and the higher expression of MMP-2 associate with poor prognosis of cancer patients.⁶ Specific inhibitors of MMPs have been shown to inhibit tumor cell invasion.⁷

At present, little information is available on the comparison of PPARg expression in human metastatic gastric cancer with their matched lymph node metastases. We intended to investigate human gastric carcinomas for the expression of PPARg and E-cadherin, which is involved in cell-cell adhesion and thus in cell motility, and MMP-2, which is involved in matrix degradation and thus in invasiveness. Further to explore their correlation in order to evaluate whether PPARg affected both factors and to elucidate the role of PPARg in gastric carcinogenesis and metastases.

Subjects
Fifty-four surgically resected advanced gastric carcinomas with their regional lymph node metastases and the adjacent non-tumor tissues were retrieved from the Department of Pathology, the First Affiliated Hospital of Sun Yat-sen University. All patients underwent total gastrectomy or subtotal gastrectomy with extended lymphadenectomy between May 2000 and July 2003 at the Department of Surgery. No previous treatments, including chemotherapy, surgery and radiotherapy were given to these patients. The mean age of the patients was 56 years (range 30 to 81 years). Among them, 33 patients were male and 21 were female (male: female = 1.57:1). For each specimen, slide stained with hematoxylin and eosin (HE) was reviewed to verify the tumor morphotype, degree of differentiation (well: 1, moderate: 11, poor: 42) and other relevant histopatho- logical features.

Tissue preparation
Immediately after the surgical resection, segments of the gastric carcinomas and lymph node metastases and the adjacent non-tumor tissues were collected in phosphate-buffered saline (PBS; 0.9% NaCl in 0.01 mol/L sodium phosphate buffer, pH 7.4). After being washed in PBS three times to remove gastric juice and blood, the specimens were fixed in 10% formalin for 24 hours at room temperature. Then they were washed with PBS three times for 10 minutes to remove the fixative. Finally, the specimens were embedded in paraffin and sections of 4 mm thickness were cut by a microtome.

Immunohistochemistry
The paraffin-sections were dewaxed in xylene and were brought to water through graded alcohol solutions. Microwave treatment was performed in 10 mmol/L citrate buffer (pH 6.0) for 10 minutes for antigen unmasking for MMP-2 and E-cadherin. For PPARg retrieval, 1 mmol/L Tris–EDTA buffer (pH 9.0) was added and incubated for 20 minutes. Then, the slides were cooled down to room temperature. The sections were treated with freshly prepared 3% hydrogen peroxide in the dark for 15 minutes at room temperature to block the endogenous peroxides activity. The identical specimen was incubated with primary antibodies (mouse monoclonal antibodies) of PPARg (Santa Cruz, USA), MMP-2 (Antibody Diagnostic, USA) and E-cadherin (Dako, USA), in a humid chamber at 4℃ overnight, respectively. The concentrations of the primary antibody were 1:50 for PPARg, 1:100 for MMP-2, 1:50 for E-cadherin in PBS. Color development was accomplished using the Envision nonbiotin detection system with DAB Chromogen (Envision+Peroxidase, mouse/rabbit, Dako, USA) and the specimens were counterstained with hematoxylin. Then, the sections were dehydrated through graded alcohols. Finally, the slides were mounted in a mounting medium, covered with covers lips. Primary antibodies were omitted for negative controls.

Evaluation of immunoreactivity
The immunoreactivities of PPARg, E-cadherin and MMP-2 were examined under a light microscope. PPARg immunoreactivity was observed in cytoplasm and/or nucleus. The staining patterns of E-cadherin, as reported previously,⁸ were classified into the following four groups: membrane pattern (A) in which E-cadherin immunoreactivity was present on the basolateral surfaces of the epithelial cells; absent pattern (B) in which E-cadherin immunoreactivity was not observed; cytoplasmic pattern (C) in which E-cadherin immunoreactivity was present in the cytoplasm but not on the membrane; and heterogeneous pattern (D) in which E-cadherin immunoreactivity was present both in the cytoplasm and on the membrane. Only the membrane pattern (A) was considered as the normal pattern while the other three patterns (B, C and D) were considered as abnormal. In the case of mixed patterns in some sections, the classification was determined by the dominant pattern. Scoring of MMP-2 was exclusively restricted to cell staining whereas stromal and fibroblasts staining were not considered.

A numerical scoring system⁹ with two categories was used to assess the intensity and the extent of PPARg, E-cadherin and MMP-2 immunoreactivity in tumor cells and normal gastric mucosa. Category A documented the number of immunoreactive cells as 0 (no immunoreactive cells), 1 (immunoreactive cells were <25%), 2 (25%–75%) and 3 (>75%). A positive case was defined as having a category A value of 1 or over. Category B documented the intensity of the immunostaining as 0 (no staining), 1 (weak), 2 (moderate), and 3 (strong). Finally, the values of categories A and B were added to give the “immunoreactivity scores” (IRSs), which were ranged from 0 to 6. It should be noticed that in the IRS calculating system, the score of an individual patient was the sum of the values of categories A and B and it could not be 1. Immunostaining was assessed by a single experienced pathologist who was blinded to the clinical data of the patients.

Statistical analysis
All statistical analyses were done using the SPSS 11.0 software. The numbers of positive expression were compared using the Chi-square test. Expression levels of PPARg, E-cadherin and MMP-2 were analyzed using nonparametric K related tests. The correlations for PPARg, E-cadherin and MMP-2 expression were analyzed with Pearson tests. Statistical significances were defined as P < 0.05.

Expression of PPARg in normal gastric tissues, gastric carcinomas and matched lymph node metastases
PPARg immunoreactivity was present abundantly and intensively in the epithelial cell nuclei in 90.7% of normal gastric tissues (49 of 54, Fig. 1B). Occasionally, PPARg immunoreactivity was located in the cytoplasm of some glandular epithelium. The nuclei expression of PPARg was significantly weaker in primary tumors (12 of 54, 22.2%, Fig.1D) and in corresponding metastases (23 of 54, 42.6%, Fig.1F) than in normal gastric tissues (P<0.001, Table). But the expression of PPARg was relatively stronger in metastases than in primary tumors (P<0.05). Similarly the expression levels of PPARg in primary tumors were higher than that in metastases, but lower than that in the normal tissues. There was no significant difference in the nuclear expression of PPARg in well-, moderately-, and poorly-differentiated cancers (data not shown). In addition PPARg immunoreactivity was observed in the cytoplasm of 30.0% of primary tumors (17 of 54) and in 14.8% of metastases (8 of 54).

Expression of E-cadherin in normal gastric tissues, gastric carcinomas and matched lymph node metastases
E-cadherin was expressed in the all normal gastric tissues with the normal pattern (A): membrane pattern (Fig. 2B). However, in all neoplastic cells, E-cadherin was expressed with abnormal patterns: absent pattern (B), cytoplasm pattern (C), and/or cytoplasm and membrane pattern (D) (Fig. 2D and F). E-cadherin immunoreactivity was observed in 36 primary tumors (66.7%) and 24 lymph node metastases (44.4%) (P<0.05, Table) and decreased compared with normal tissues (P<0.001). Moreover, the expression level of E-cadherin significantly decreased in gastric carcinoma and further decreased in lymph node metastases (P<0.001). The express levels of E-cadherin decreased as the differentiation stage turned from well-, moderately-, and poorly-differentiated cancers (data not shown).

Expression of MMP-2 in normal gastric tissues, gastric carcinomas and matched lymph node metastases
In normal gastric specimens, MMP-2 immunoreactivity was found in the membranes of foveolar epithelia of the mucosa, with different intensity, whereas the deep glandular epithelium of mucosa was not stained (Fig. 3B). The expression of MMP-2 was predominantly located in the cytoplasm of primary tumor cells (Fig. 3D) and metastatic cells (Fig. 3F), and occasionally presented in the membranes of malignant cells. The expression of MMP-2 was observed in 30 primary tumors (55.6%) and 41 lymph node metastases (75.9%, P<0.05, Table). The expression level of MMP-2 significantly decreased in primary tumors compared with lymph node metastases (P<0.001, Table). The express levels of MMP-2 increased as the differentiation stage turned from well-, moderately-, and poorly-differentiated cancers (data not shown).

view in a new window

Fig. 3. Representative photographs of MMP-2, immunohisto- chemical staining in normal gastric mucosa (A and B), primary gastric tumors (C and D) and matched lymph node metastases (E and F). Sequential section for HE staining and immunohistochemical analysis (Original magnification×400).

Correlation among PPARg, E-cadherin and MMP-2 expression
In neoplastic foci, there was significant relationship between PPARg and MMP-2 expression (P<0.05). However, no significant relationship between E-cadherin expression and MMP-2 (P>0.05) and PPARg expression (P>0.05) was revealed.

In the present study, we found that the expression levels of PPARg were much lower in gastric carcinomas and matched lymph node metastases compared with normal gastric tissues. There was no significant difference in the expression level of PPARg among well-, moderately-, and poorly-differentiated cancers. Sato et al¹⁰found that PPARg protein shown by immunohistochemistry was expressed not only in well differentiated, moderately differentiated and poorly differentiated gastric adenocarcinoma, but also in normal mucosa with intestinal metaplasia adjacent to cancer, moreover the staining result was consistent in all cases.

Another study showed that the frequency of samples with positive cytosolic PPARg immunohistochemical staining decreased as the differentiation stage turned from intestinal metaplasia to adenoma, well-, moderately-, and poorly-differentiated cancers. Simultaneously, there was a tendency toward an increased frequency of samples with positive nuclear PPARg staining as the differentiation stage transformed from intestinal metaplasia to poorly-differentiated cancer.¹¹ These results were controversial. The different genes and number of samples studied may help to explain different results.

Previously published data showed that PPARg expression decreased in esophageal cancer, lung cancer, follicular thyroid cancer and human cervical carcinoma.^12-16 Decreased PPARg expression was correlated with poor prognosis in patients with esophageal cancer and lung cancer.^{14, 15} Badawi et al¹⁶ reported that down-regulation of PPARg mRNA expression was characterized as predictors of breast cancer metastases. Thus down-regulation of PPARg expression in cancer cells and tissues may be associated with a progression of gastric cancer. In contrast, other studies demonstrated that PPARg expression level was higher in testicular, pancreatic and breast carcinoma than that in corresponding normal tissues.^17-19 PPARg is not consistently expressed among the various tissues, suggesting that this protein is tissue-specific. In our study, the expression of PPARg was increased in metastases compared with primary tumors. The cause of PPARg re-expression after metastases is unknown and further study on the mechanism is needed.

The pattern of PPARg staining was nuclear and cytoplasmic in the present study. In normal gastric tissue, dominant nuclear staining was observed. However, primary tumor staining was mainly cytoplasmic, while metastases were dominantly nuclear. It has been reported that PPARg immunoreactivity was mainly located in cytoplasma in lung carcinoma and human neuroblastoma cells.^{20, 21} In addition, some studies demonstrated that the intracellular distribution-translocation of nuclear receptors as PPARs and RXRs is achieved by their phosphorylation.^{22, 23} Alterations of this phosphorylation process, as it usually occurs in cancer, results in the loss of receptor-mediated effect,²² by blocking the protein in the cytoplasm. Further studies are still needed to elucidate the reasons and mechanism about intracellular distribution-translocation as mutation of PPARg gene is a very rare event in human malignancies.²⁴

In the present study, abnormal expressions of E-cadherin (cytoplasm pattern, cytoplasm and membrane pattern, or absent pattern) were 100% in primary tumors and matched metastases. This rate is markedly higher than that in other studies.^{8, 25, 26}This difference might be due to the fact that only advanced gastric carcinomas with metastases were evaluated in our study. Normal expression (membrane pattern) was observed in all normal epithelium in both normal gastric mucosa and sections containing primary tumor or metastases. It has been shown that abnormal E-cadherin expression is closely associated with disease stages and lymph node metastasis, and thus E-cadherin expression is viewed as a differentiation marker in gastric carcinoma.^{8, 27} Abnormal E-cadherin expressions may be caused by the inactivation of E-cadherin gene in tumor tissues.²⁸

Our study revealed that E-cadherin expression decreased in primary tumors than in the normal gastric tissues and further decreased in metastases. Moreover the E-cadherin expression reduced as the differentiation stage turned from well-, moderately-, and poorly-differentiated cancers. It has been demonstrated that down-regulation of E-cadherin expression might lead to the reduction of the inhibition on proliferation, then allow the tumor cells to escape from growth control signals, and be also indispensable for cancer invasion and metastasis.⁴ Transfection of E-cadherin negative cell lines with E-cadherin complimentary DNA is able to suppress the invasive behavior.²⁹

The present study showed that MMP-2 was expressed in normal foveolar epithelium membrane in almost all cases, which was consistent with the flow cytometry results reported by Koyama et al.³⁰ However, our results are different from the data of earlier studies, which showed that there was no MMP-2 expression in normal gastric epithelium.^31-33 In malignant cells of primary tumors and metastases, the expression of MMP-2 was dominantly detected in the cytoplasm and occasionally on the membrane. The expression of MMP-2 was stronger in metastases than in primary tumors. Moreover the express levels of MMP-2 increased as the differentiation stage turned from well-, moderately-, and poorly-differentiated cancers. Metastasis formation is a complex multi-step process. Degradation of basement membrane components, especially type IV collagen, is a necessary step in metastasis.The higher MMP-2 expression in metastases is beneficial for neoplastic cells to migrate from the primary site and thrive in the microenvironment of the secondary site.³⁴

In this study, in primary tumors and metastases, PPARg expression showed a significant correlation with MMP-2 expression. Some studies have shown that PPARg ligands inhibit cancer cells' invasiveness and MMP-2 secretion in human pancreatic cancer, adrenocortical cancer, and myeloid leukemia cells.^35-37 Sasaki et al³⁸ found that linoleic acid with activation of PPARg inhibits invasion and metastasis of human gastric cells. According to these data, we postulated that PPARg might modulate MMP-2 expression and affect gastric cancer metastases.

In addition, Annicotte et al³⁹ found that PPARg regulated E-cadherin expression and inhibited growth and invasion of prostate cancer. Conjugated linoleic acid (CLA) regulates PPARg expression by selectively acting as an agonist and may influence E-cadherin/beta-catenin pathway and invasiveness of MCF-7 breast cancer cells.⁴⁰ Thiazolidindiones (TZDs), synthetic activators of PPARg, markedly increased differentiation markers including E-cadherin in a human pancreatic cancer cell line, BxPC-3.⁴¹ However in our study, there was no correlation between E-cadherin and PPARg or MMP-2 expression. Because abnormal expression of E-cadherin gene was potentially inactivated by mutations in a significant number of gastric caner,^28,29 we think that this may help to explain the lack of correlation between E-cadherin, PPARg and MMP-2 expression in gastric carcinoma.

In conclusion, our study showed that the expression of PPARg and E-cadherin was down-regulated in human gastric carcinoma and metastases compared with normal gastric tissues. Moreover, E-cadherin expression was further down-regulated and MMP-2 expression was further up-regulated in metastases compared with the primary tumors. These results suggest that down-regulation of E-cadherin and PPARg and up-regulation of MMP-2 may be helpful to gastric carcinogenesis and metastases. An inverse relationship between PPARg and MMP-2 in human gastric carcinoma suggests that PPARg might modulate MMP-2 expression and affect gastric cancer metastases. Further experimental in vitro and in vivo studies are necessary to explore the PPARg participation and the potential of PPARg in gastric cancer metastases.

1. Pisani P, Parkin DM, Bray F, Ferly J. Erratum: estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 1999; 83: 870-873.

2. Liotta LA, Kohn E. Cancer invasion and metastases. JAMA 1990; 263: 1123-1126.

3. Grommes C, Landreth GE, Heneka MT. Antineoplastic effects of peroxisome proliferator-activated receptor γ agonists. Lancet Oncol 2004; 5: 419-429.

4. Hazan RB, Qiao R, Keren R, Badano I, Suyama K. Cadherin switch in tumor progression. Ann N Y Acad Sci 2004; 1014: 155-163.

5. Kyrlagkitsis I, Karamanolis DG. Genes and gastric cancer. Hepatogastroenterology 2004; 51: 320-327.

6. Stetler-Stevenson WG, Yu AE. Proteinases in invasion: matrix metalloproteinases. Semin Cancer Biol 2001; 11: 143-152.

7. DeClerck YA, Yean TD, Chan D, Shimada H, Langley KE. Inhibition of tumor invasion of smooth muscle cell layers by recombinant human metalloproteinase inhibitor. Cancer Res 1991; 51: 2151-2157.

8. Tanaka M, Kitajima Y, Edakuni G, Sato S, Miyazaki K. Abnormal expression of E-cadherin and beta-catenin may be a molecular marker of submucosal invasion and lymph node metastasis in early gastric cancer. Br J Surg 2002; 89: 236-244.

9. Chen J, Rocken C, Hoffmann J, Kruger S, Lendeckel U, Rocco A, et al. Expression of carbonic anhydrase 9 at the invasion front of gastric cancers. Gut 2005; 54: 920-927.

10. Sato H, Ishihara S, Kawashima K, Moriyama N, Suetsugu H, Kazumori H, et al. Expression of peroxisome proliferator- activated receptor (PPAR) γ in gastric cancer and inhibitory effects of PPAR γ agonists. Br J Cancer 2000; 83: 1394-1400.

11. Nomura S, Nakajima A, Ishimine S, Matsuhashi N, Kadowaki T, Kaminishi M. Differential expression of peroxisome proliferator-activated receptor in histologically different human gastric cancer tissues. J Exp Clin Cancer Res 2006; 25: 443-448.

12. Aldred MA, Morrison C, Gimm O, Hoang-Vu C, Krause U, Dralle H, et al. Peroxisome proliferator-activated receptor γ is frequently downregulated in a diversity of sporadic nonmedullary thyroid carcinomas. Oncogene 2003; 22: 3412-3416.

13. Jung TI, Baek WK, Suh SI, Jang BC, Song DK, Bae JH, et al. Down-regulation of peroxisome proliferator-activated receptor γ in human cervical carcinoma. Gynecol Oncol 2005; 97: 365-373.

14. Terashita Y, Sasaki H, Haruki N, Nishiwaki T, Ishiguro H, Shibata Y, et al. Decreased peroxisome proliferator-activated receptor γ gene expression is correlated with poor prognosis in patients with esophageal cancer. Jpn J Clin Oncol 2002; 32: 238-243.

15. Sasaki H, Tanahashi M, Yukiue H, Moiriyama S, Kobayashi Y, Nakashima Y, et al. Decreased perioxisome proliferator- activated receptor γ gene expression was correlated with poor prognosis in patients with lung cancer. Lung Cancer 2002; 36: 71-76.

16. Badawi AF, Badr MZ. Expression of cyclooxygenase-2 and peroxisome proliferator-activated receptor-γ and levels of prostaglandin E [2] and 15-deoxy-Δ[12,14]-prostaglandin J[2] in human breast cancer and metastasis. Int J Cancer 2003; 103:84-90.

17. Hase T, Yoshimura R, Mitsuhashi M, Segawa Y, Kawahito Y, Wada S, et al. Expression of peroxisome proliferator- activated receptors in human testicular cancer and growth inhibition by its agonists. Urology 2002; 60: 542-547.

18. Kubota T, Koshizuka K, Williamson EA, Asou H, Said JW, Holden S, et al. Ligand for peroxisome proliferator-activated receptor γ (troglitazone) has potent antitumor effect against human prostate cancer both in vitro and in vivo. Cancer Res 1998; 58: 3344-3352.

19. Takahashi N, Okumura T, Motomura W, Fujimoto Y, Kawabata I, Kohgo Y. Activation of PPAR γ inhibits cell growth and induces apoptosis in human gastric cancer cells. FEBS Lett 1999; 455: 135-139.

20. Theocharis S, Kanelli H, Politi E, Margeli A, Karkandaris C, Philippides T, et al. Expression of peroxisome proliferator activated receptor γ in non-small cell lung carcinoma: correlation with histological type and grade. Lung Cancer 2002; 36: 249-255.

21. Han SW, Greene ME, Pitts J, Wada RK, Sidell N. Novel expression and function of peroxisome proliferator-activated receptor γ (PPARγ) in human neuroblastoma cells. Clin Cancer Res 2001; 7: 98-104.

22. Jiang WG, Redfern A, Bryce RP, Mansel RE. Peroxisome proliferator activated receptor-γ (PPAR-γ) mediates the action of γ linolenic acid in breast cancer cells. Prostaglandins Leukot Essent Fatty Acids 2000; 62: 119-127.

23. Katagiri Y, Takeda K, Yu ZX. Modulation of retinoid signalling through NGF-induced nuclear export of NGFI-B. Nat Cell Biol 2000; 2: 435-440.

24. Ikezoe T, Miller CW, Kawano S, Heaney A, Williamson EA, Hisatake J, et al. Mutational analysis of the peroxisome proliferator-activated receptor γ gene in human malignancies. Cancer Res 2001; 61: 5307-5310.

25. Ramesh S, Nash J, McCulloch PG. Reduction in membranous expression of beta-catenin and increased cytoplasmic E-cadherin expression predict poor survival in gastric cancer. Br J Cancer 1999; 81: 1392-1397.

26. Ohno T, Aihara R, Kamiyama Y, Mochiki E, Asao T, Kuwano H. Prognostic significance of combined expression of MUC1 and adhesion molecules in advanced gastric cancer. Eur J Cancer 2006; 42: 256-263.

27. Zheng Z, Pan J, Chu B, Wong YC, Cheung AL, Tsao SW. Downregulation and abnormal expression of E-cadherin and beta-catenin in nasopharyngeal carcinoma: close association with advanced disease stage and lymph node metastasis. Hum Pathol 1999; 30: 458-466.

28. Hirohashi S, Kanai Y. Cell adhesion system and human cancer morphogenesis. Cancer Sci 2003; 94: 575-581.

29. Miyaki M, Tanaka K, Kikuchi-Yanoshita R, Muraoka M, Konishi M, Takeichi M. Increased cell-substratum adhesion, and decreased gelatinase secretion and cell growth, induced by E-cadherin transfection of human colon carcinoma cells. Oncogene 1995; 11: 2547-2552.

30. Koyama S. Enhanced cell surface expression of matrix metalloproteinases and their inhibitors, and tumor-induced host response in progression of human gastric carcinoma. Dig Dis Sci 2004; 49: 1621-1630.

31. Monig SP, Baldus SE, Hennecken JK, Spiecker DB, Spiecker DB, Grass G, et al. Expression of MMP-2 is associated with progression and lymph node metastasis of gastric carcinoma. Histopathology 2001; 39: 597-602.

32. Kabashima A, Maehara Y, Kakeji Y, Baba H, Koga T, et al. Clinical pathological features and overexpression of matrix metalloproteinases in intramucosal gastric carcinoma with lymph node metastasis. Clin Cancer Res 2000; 6: 3581-3584.

33. Mori M, Mimori K, Yoshikawa Y, Shibata K, Utsunomiya T, Sadanaga N, et al. Analysis of the gene-expression profile regarding the progression of human gastric carcinoma. Surgery 2002; 131: S39-S47.

34. Rhee JS, Coussens LM. RECKing MMP function: implications for cancer development. Trends Cell Biol 2002; 12: 209-211.

35. Liu J, Lu H, Huang R, Lin D, Wu X, Lin Q, et al. Peroxisome proliferator activated receptor-γ ligands induced cell growth inhibition and its influence on matrix metallo-proteinase activity in human myeloid leukemia cells. Cancer Chemother Pharmacol 2005; 56: 400-408.

36. Ferruzzi P, Ceni E, Tarocchi M, Grappone C, Milani S, Galli A, et al. Thiazolidinediones inhibit growth and invasiveness of the human adrenocortical cancer cell line H295R. J Clin Endocrinol Metab 2005; 90: 1332-1339.

37. Galli A, Ceni E, Crabb DW, Mello T, Salzano R, Grappone C, et al. Antidiabetic thiazolidinediones inhibit invasiveness of pancreatic cancer cells via PPARγ independent mechanisms. Gut 2004; 53: 1688-1697.

38. Sasaki T, Fujii K, Yoshida K, Shimura H, Sasahira T, Ohmori H, et al. Peritoneal metastasis inhibition by linoleic acid with activation of PPARγ in human gastrointestinal cancer cells. Virchows Arch 2006; 448: 422-427.

39. Annicotte JS, Iankova I, Miard S, Fritz V, Sarruf D, Abella A, et al. Peroxisome proliferator-activated receptor γ regulates E-cadherin expression and inhibits growth and invasion of prostate cancer. Mol Cell Biol 2006; 26: 7561-7574.

40. Bocca C, Bozzo F, Francica S, Colombatto S, Miglietta A. Involvement of PPAR γ and E-cadherin/beta-catenin pathway in the anti-proliferative effect of conjugated linoleic acid in MCF-7 cells. Int J Cancer 2007; 121: 248-256.

41. Ohta T, Elnemr A, Yamamoto M, Ninomiya I, Fushida S, Nishimura G, et al. Thiazolidinedione, a peroxisome proliferator-activated receptor γ ligand, modulates the E-cadherin/beta-catenin system in a human pancreatic cancer cell line, BxPC-3. Int J Oncol 2002; 21: 37-42.