Skip to main content
NCBI home page
Journal of Cancer Research and Clinical Oncology logoLink to Journal of Cancer Research and Clinical Oncology
. 2005 Sep 23;131(12):809–814. doi: 10.1007/s00432-005-0030-3

Coordinate cell-surface expression of matrix metalloproteinases and their inhibitors on cancer-associated myofibroblasts from malignant ascites in patients with gastric carcinoma

Shohei Koyama 1,✉
PMCID: PMC12161200  PMID: 16180025

Abstract

Purpose:The crucial role of tumor stroma in cancer cell invasion has been described in human carcinoma tissues. However, myofibroblastic invasion remains largely unexplored in malignant ascites. Purpose of this study is to investigate the spatial localization or regulation of matrix metalloproteinases (MMP-2, -7 -9, MT1-MMP) and their inhibitors (TIMP-2 and -4) on myofibroblasts from malignant ascites in 20 patients with gastric carcinoma. Methods: The quantitative flow cytometric analysis of MMPs or TIMPs on myofibroblasts was based on the percentage of double positive cells defined by anti MMPs or anti TIMPs, and anti α-smooth muscle actin (α-SMA) antibodies. Result: The results clearly showed that the coordination of the high level of cell-surface expression of secreted MMPs and TIMPs was noted on the α-SMA+ myofibroblasts. The finding suggests the possible formation of ternary complex, MT1-MMP/TIMPs/MMPs on the cells. The events might be a cause and result of activation processing of MMPs on the cells. Conclusion: This study provides the presence of invasive myofibroblasts with activated MMPs in close association with MMPs+ and TIMPs+ cancer cells and tumor-infiltrating lymphocytes from malignant ascites, emphasizing the importance of molecular cross-talk in tumor-host microenvironment for cancer invasion, metastasis and progression.

Keywords: Myofibroblasts, MMP-2, MT1-MMP, TIMP-2, Malignant acites, Gastric carcinoma

Introduction

The majority of human neoplasm is carcinoma that arises from epithelial cells by definition. Consequently, the focus of research for cancer metastasis has been on epithelial cells, or more specifically genetic changes that occur in epithelial cells as they progress from normal to malignant. In previous study, we showed the increased cell-surface expression of matrix metalloproteinases (MMPs) and their inhibitors of MMP (TIMPs) on cancer cells, and tumor-infiltrating lymphocytes (TIL) during cancer invasion and metastasis (Koyama 2004a). However, it has been shown that concomitant changes also occur in stromal cells surrounding the malignant tumor (Liotta and Kohn 2001; DeWeber and Mareel 2003). The tumor stroma is now considered to be an integral part of a neoplasm with a major role in several steps of the metastatic cascade. The stromal cells are histopathologically composed of fibroblasts, inflammatory cells, immune cells, smooth muscle cells, and endothelial cells. The fibroblasts associated with tumor cells are seen significantly differentiated into so-called “myofibroblasts” (Desmouliere et al. 1993), which are also found in the process of wound healing. The trait of myofibroblasts containing both fibroblasts and smooth muscle cell lineages has been defined by positive expression of smooth muscle cell markers such as α-smooth muscle actin (α-SMA) (Skalli et al. 1986; Sappino et al. 1990).

On the other hand, the migration of myofibroblasts into carcinoma tissues has been shown in human cancers including prostate, endometrium, pancreas, stomach, head and neck, colon, breast and lung (Olumi et al. 1999; Orimo et al. 2001; Yen et al. 2002; Nakayama et al. 2002; Tuxhorn et al. 2002; Barth et al. 2004; Ishikawa et al. 2004; Poulsom et al. 1992; Bisson et al. 2003; Heppner et al. 1996). Although the myofibroblasts have shown to express several cytokines and protease, MMPs and TIMPs (Liotta and Kohn 2001; DeWeber and Mareel 2003), there are a few reports about the expression of MMPs and TIMPs in myofibroblasts within carcinoma tissues (Poulsom et al. 1992; Heppner et al. 1996; Bisson et al. 2003; Barth et al. 2004; Ishikawa et al. 2004). Moreover, the expressive analysis of MMPs and/or TIMPs in the conventional studies (Poulsom et al. 1992; Bisson et al. 2003; Ishikawa et al. 2004) has been immunohistochemically assessed. Such an approach by particularly using formalin-fixed, paraffin-embedded specimen is no longer useful to identify whether the MMPs and/or TIMPs are localized at the cell membrane or cytoplasma or the both. Therefore, the precise role of MMPs and TIMPs in myofibroblasts within carcinoma-tissues has not been well understood. In this regard, the analysis has been made by a flow cytometry on viable cells to address the difficult problem (Koyama 2000, 2003a, b, 2004a, b; Koyama et al. 1998, 1999, 2001, 2002). Furthermore, the expression of MMPs and TIMPs on cancer-associated myofibroblasts has not been examined in malignant ascites of human cancers, nor have studies been performed in exact localization patterns of these molecules. Such a detailed spatial localization or regulation (cancer-stromal cell relation) concerning these proteolytic enzymes is invariably needed to evaluate the actual relevance of these cellular or/and molecular mechanisms in cancer-associated myofibroblasts. The cancer-associated myofibrobalsts in the present study were obtained from the malignant ascites in patients with gastric carcinoma, because the dissemination of cancer cells in the serous cavities is pivotal one of the models in cancer invasion or/and metastasis. The myofibroblasts were floating in a protein-rich fluid containing large amounts of cancer cells and TIL.

This study was designed to quantitatively analyze the cell-surface expression of MMP-2, -7, -9, MT1-MMP, TIMP-2, and -4 in cancer-associated myofibroblasts by the method of two-color flow cytometry. In this report, MMPs and their inhibitors, TIMPs are coordinately expressed at the cell-surface by a vast majority of cancer-associated α-SMA+ myofibroblasts. The biological roles and significances of the results are discussed with special reference in molecular cross-talk between cancer cells and myofibroblasts, followed by cancer invasion and metastasis.

Materials and methods

Specimens

Myofibroblasts mixed with metastatic cancer cells and TIL were obtained from malignant peritoneal effusions of 20 patients (age range, 50–79 years; average 64.7 years) with advanced gastric carcinoma. The cancer cells were judged by morphological examination of Papanicolaou’s staining and/or by Wright-Giemsa-stained smears. None of the patients studied had received anticancer drugs before paracentasis. This study was performed in accordance with declaration of Helsinki. Informed consent was obtained from all patients.

Monoclonal antibodies (mAbs) and two-color flow cytometric analysis

mAbs for MMP-2 (mouse IgG1/κ, clone 42-5D11), MMP-7 (mouse IgG1/κ, clone 141-7B2), MMP-9 (mouse IgG1/κ, clone 56-2A4), MT1-MMP (mouse IgG1/κ, clone 114-1F2), TIMP-2 (mouse IgG1/κ, clone 67-4H11) and TIMP-4 (IgG2a/κ, clone 216-1G2) were obtained from Daiichi Fine Chemical Ltd. (Takaoka City, Toyama, Japan) to screen for their usefulness in flow cytometry. To determine the cell-surface molecules of MMPs and TIMPs, the prepared cells mixed with cancer cells, TIL and myofibroblasts were incubated with heat-treated human aggregate IgG (GII 2388, Cohn fraction II and III, Sigma, St Louis, MO, USA) for 15 min in PBS containing 0.1% BSA and 0.1% sodium azide to prevent antibody binding to FcR. The cells were then stained with a saturating concentration of each primary mAb for 30 min on ice with occasional shaking, washed twice ice-cold PBS, and then incubated with [phycoerythrine (PE)-conjugated goat anti-mouse IgG antibody (Biosource International, Camarillo, CA 93012, USA)] for an additional 30 min on ice and washed twice. Mouse IgG (Inter-cell Technologies, Hopewell, NJ 08525, USA) was added to block free IgG-binding sites on anti-mouse IgG, and 15 min later the cells were then stained with a saturating concentration of fluorescein-isothiocyanate (FITC)-conjugated anti-α-SMA mAb (mouse IgG2a, clone 1A4, Sigma, Saint Louis, MI 63103, USA) for 30 min on ice and washed twice, as described previously (Koyama 2000, 2003a, b, 2004a, b; Koyama et al. 1998, 1999, 2001, 2002). As a negative control, an aliquot of the cells from each case was stained with an irrelevant mAb of the same phenotype and a secondary antibody as described previously (Koyama 2000; Koyama et al. 1998, 1999, 2001). Flow cytometric analysis was performed on FACScan (Becton Dickinson, Mountain View, CA, USA) and data were processed by using the CellQuest program (Becton Dickinson).

To detect myofibroblasts at first, the prepared cells stained with FITC-conjugated anti-α-SMA alone were applied to FACScan. As displayed in FITC-dot plots of FL1 and FL2 (Fig. 1a) at ungated condition, α-SMA+ -area was determined by comparison with the isotype matched control cells. In order to address the position of α-SMA+-area (R1) in Fig. 1a, light scatter plots based on forward and side light scatter (FSC vs. SSC) with viable α-SMA+ cells were drawn, and then the main α-SMA+ events (R2) were gated on FSC vs SSC dot plots as shown in Fig. 1b. Next, the ungated dot-plots of FSC vs SSC were created to include representative cell populations in malignant ascites, such as tumor cells, TIL and α-SMA+ cells (Fig. 1c), and the gated area (R2) of Fig. 1c was subsequently copied from the gated area of Fig. 1b. It has clearly been shown that right lower dotted area belongs to tumor cells and left lower dotted area is mainly TIL (Fig. 1c). Namely, almost tumor cells, TIL and debris were excluded by these gating method. In addition, we have confirmed that the tumor cells and TIL gated are not stained by anti-α-SMA mAb. The R2-region containing α-SMA+ cells was assessed for expression of MMPs and TIMPs on myofibroblasts. In general, cell number of R2 area in Fig. 1c was approximately 5–15% of the total cell number exhibited, although it depended on the sample of the cases. The analysis was based on the collection of 10,000 events in gate of the R2-region. FACScan machine was calibrated before use to optimize instrument detection of fluorescence and scatter signals. The percentage of fluorescence-positive events within the area was determined after setting gate to contain 99% or more of the negative control samples. α-SMA-negative cells were eliminated from the collected 10,000 cells to revise a population of MMPs+ and TIMPs+ myofibroblasts from among α-SMA+ cells, since cell-populations containing α-SMA+ cells at R2 region in Fig. 1c were intermingled with other unknown α-SMA-negative cells except cancer cells and TIL to some extent.

Fig. 1.

Fig. 1

Open in a new tab

Gating strategy to determine MMPs and TIMPs on α-SMA+ cell populations from malignant ascites. a A gated area (solid box) shows α-SMA+ cell populations in FL1 versus FL2-dot plots. Vertical cursor distinguishes positive cells, defined by comparison with the isotype matched control cells. b A distribution of FSC versus SSC-dot plots corresponding to an area of α-SMA+ cell populations in Fig. 1a. A gated region of R2 (solid box) reveals main populations of α-SMA+ cells. c FACS-dotted plots of FSC versus SSC were shown. A gated R2-area (solid box) containing α-SMA+ cell populations was copied from Fig. 1b. Note that almost tumor cells, TIL and cell debris were capable to exclude for determination of MMPs and TIMPs expression on α-SMA+ myofibroblasts

Statistical analysis

Results are expressed as the mean ± SEM percentages of the cells having a given antigen expression. The data were compared by the Student’s t test and considered significant if a P value was <0.05.

Results

Optimal gating strategy confirmed the precision with two-color FACS analysis for determination of MMPs and TIMPs on α-SMA+ cells from malignant peritoneal effusion in 20 patients with gastric carcinoma. The representative FACS dot-plots illustrating the surface expression of MMP-2, -7, -9, MT1-MMP and their inhibitors, TIMP-2 and -4 on α-SMA+ myofibroblasts are shown in Fig. 2. Significant levels of these molecules were clearly demonstrated as double positive cells by α-SMA+ cells. Fig. 3 showed summary results of the cell-surface expression of MMPs and TIMPs on α-SMA+ myofibroblasts from malignant ascites. α-SMA+ cells disclosed an enhanced cell surface expression of MMPs paralleled by the expression of TIMPs with up-regulation. Namely, the mean percentages of MMP-2+, -7+, -9+, MTI-MMP+, TIMP-2+ and -4+ myofibroblasts within metastatic carcinoma were 58.9%±3.9, 59.3%±3.3, 62.9%±3.4, 61.7%±3.7, 61.4%±3.6 and 68.1%±3.5, respectively. There were no significant differences among the mean percentages of MMP-2+, -7+, -9+, MTI-MMP+, TIMP-2+ and -4+ cells.

Fig. 2.

Fig. 2

Open in a new tab

Representative cell-surface expression of MMPs or TIMPs molecules on α-SMA+ myofibroblasts from malignant ascites. X-parameter: FL1 (FITC-α-SMA); Y-parameter: FL2 (PE-MMPs or -TIMPs). Horizontal and vertical cursors distinguish positive cells, defined by as those exhibiting fluorescence. The exact percentages of α-SMA+ MMPs+ or α-SMA+ TIMPs+ cells were correctly calculated by omitting of α-SMA-negative cells from collected 10,000 cells

Fig. 3.

Fig. 3

Open in a new tab

Summary results on the cell-surface expression of MMP-2, -7, -9, MT1-MMP, and TIMP-2 and -4 on α-SMA+ myofibroblasts from malignant ascites. Values are the mean percentage positive cells ± SEM

Discussion

Two-color FACS analysis has clearly shown that MMP-2, -7, -9, MT1-MMP and their inhibitors, TIMP-2 and -4 were coordinately expressed on α-SMA+ myofibroblasts from malignant ascites. It is characteristic that even secreted MMPs and TIMPs are localized preferentially on the cell-surface of the myofibroblasts at least as shown in membrane type 1-MMP.

To my knowledge, this is the first study to evaluate the precision of flow cytometry for determining the localization patterns of the MMPs and TIMPs expression on myofibroblasts within carcinoma tissues. Most studies assessed by conventional immunohistochemistry have been shown to remain obscure or to be mainly expressed in cytoplasm of myofibroblasts (Poulsom et al. 1992; Bisson et al. 2003; Ishikawa et al. 2004). The mean percentages of MMPs+ and TIMPs+ myofibroblats presented here were comparable with mean percentages of those of tumor cells from malignant ascites (Koyama 2004a). The localization patterns of MMPs and TIMPs on myofibroblasts revealed basically the same with previous findings in normal epithelial or tumor cells, and TIL (Koyama 2004a, b). As for the expression of the MMPs and TIMPs on cell-surface of myofibroblasts, these molecules are presumably secreted from myofibrobrasts themselves, but not from cancer cells, nor from TIL. Actually, it has been reported that increased mRNA or protein level of MMPs or/and TIMPs is mainly expressed in stromal cells (Ishikawa et al. 2004; Poulsom et al. 1992; Bisson et al. 2003; Heppner et al. 1996).

The organ tissues consist of epithelial and stromal cells, so-called tissue-cell society which gives influence each other. Indeed, the expression of MMPs has been confined to myofibroblasts in close contact with tumor cells (Poulsom et al. 1992; Bisson et al. 2003; Ishikawa et al. 2004). It has also been shown that the MMPs are intracellularly expressed at certain levels in resting-state cells, but these intracellular MMPs rapidly appear on the cell-surface by stimuli, such as antigens (e.g., tumor-associated antigens, bacterial antigens) with numerous cytokines and growth factors or physical cellular interactions within local host microenvironment (Koyama 2004a, b). In the current study, therefore, intracellular expression of MMPs and TIMPs in myofibroblasts was not shown since the presence of these molecules in the cells do not always indicate the actual function of the proteolytic enzymes (Koyama 2004a).

The MMPs-activities in ovarian cancer ascites have been determined by Graves et al (2004). Their results showed that the majority of active MMP-2 and -9 were from membrane-bound vesicle, but not from cell-free and vesicle-free supernatant of malignant ascites, although cellular source of MMPs-activities remained unknown. Their findings are essentially accorded with the cell-surface expression of MMPs and TIMPs in the previous (Koyama 2004a, b) and current results. Thus, cell-surface detection of soluble MMPs and TIMPs secreted presumably from myofibroblasts implies the possible formation of a ternary complex, MT1-MMP/TIMPs/MMPs to become fully active as described previously (Koyama 2004a, b). In addition to the trimolecular complex, MMP-2 also interacts with α versus β3-integrin via its haemopexin-like domain (Brooks et al. 1996). MMP-7 binds cell-surface heparin sulphate proteoglycans (HSPG), including HSPG isoforms of hyaluronan receptor CD44 (Yu et al. 2002), and even MMP-9 binds several cell surface receptors including CD44 (Yu and Stamenkovic 1999; Bourguignon et al. 1998) and ICAM-1 (Fiore et al. 2002). These MMPs concentrated in cell-membrane would have a critical role to degrade components of the ECM, indicating an essential process in the migration and invasion of tumor cells or/and myofibroblasts.

The current study shows coordinately increased expression of MMPs and TIMPs on the cells. Although the role of the TIMPs essentially involves the inhibition of activated forms of MMPs on the cell-surface, significance of higher expression of TIMPs in cancer progression has recently been explained (Koyama 2004a). Briefly, as a major inhibitor of MT1-MMP, TIMPs bind to the catalytic site of MT1-MMP. However, MT1-MMP inhibited by TIMPs acts as a receptor of MMPs, such as MMP-2. Thus, TIMPs have a critical role eventually to activate MMPs on cells, forming a trimolecular complex, MT1-MMP/TIMPs/MMPs. In order to regulate the dangerous MMPs, the prodomain of the trimolecular complex is then generally clipped by another mature MT1-MMP located near the trimolecular complex. Whereas the clipping by MT1-MMP is subject to inhibition by excess TIMPs in tumor-host microenvironment, thereby enabling effector MMPs to act long on the cells. Such myofibroblasts activation might be reflected in modifications of the adjacent ECM that are favorable to the microinvasion of cancer cells. Therefore, the activated MMPs on myofibroblasts within tumor-host microenvironment are an integral part of the invasion machinery of cancer cells.

Taken together with previous results (Koyama 2004a), the present study suggests that the MMPs/TIMPs acting on the cell membrane of myofibroblasts as well as tumor cells and TIL are axis of the proteolysis involved in pericellular ECM-degrading, and intraperitoneal cancer dissemination and/or metastasis. As this occurs, increased pericellular proteolysis presumably cleaves cell adhesion systems to offset normal contact inhibition, with subsequent forming single or small clumps of floating tumor cells and myofibroblasts in ascitic fluid. Thus, stromal myofibroblasts might be a possible new target for cancer therapy.

In conclusion, tumor-host cross-talk between myofibroblasts and cancer cells might stimulate the cell-surface expression of both MMPs and TIMPs on these cells, and then the MMPs and TIMPs would undergo ternary complex, MT1-MMP/TIMPs/MMPs to become fully active. This study in malignant ascites would contribute to a better understanding of stromal contribution to cancer invasion, metastasis and progression.

Acknowledgements

This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. Ms. Y. Jinsenji is greatly acknowledged for technical help for flow cytometric analysis.

References

  1. Barth PJ, zu Schweinsberg TS, Ramaswamy A, Moll R (2004) CD34+ fibrocytes, α-smooth muscle antigen-positive myofibroblasts, and CD117 expression in the stroma of invasive squamous cell carcinomas of the oral cavity, pharynx, and larynx. Virchows Arch 444:231–234 [DOI] [PubMed] [Google Scholar]
  2. Bisson C, Blacher S, Polette M, Blanc J-F, Kebers F, Desreux J, Tetu B, Rosenbaum J, Foidart J-M, Birembaut P, Noel A (2003) Restricted expression of membrane type 1-matrix metalloproteinase by myofibroblasts adjacent to human breast cancer cells. Int J Cancer 105:7–13 [DOI] [PubMed] [Google Scholar]
  3. Bourguignon LY, Gunja-Smith Z, Iida N, Zhu HB, Young LJT, Muller WJ, Cardiff RD (1998) CD44v (3, 8–10) is involved in cytoskeleton-mediated tumor cell migration and matrix metalloproteinase 9 (MMP-9) association in metastatic breast cancer cells. J Cell Physiol 176:206–215 [DOI] [PubMed] [Google Scholar]
  4. Brooks PC, Stromblad S, Saunders LC, Sanders LC, von Schalscha TL, Aimes RT, Stetler-Stevenson WG, Quigley JP, Cheresh DA (1996) Localization of matrix metalloproteinase MMP-2 to the surface of invasive cells by interaction with integrin αvβ3. Cell 85:683–693 [DOI] [PubMed] [Google Scholar]
  5. De Weber O, Mareel M (2003) Role of tissue stroma in cancer cell invasion. J Pathol 200:429–447 [DOI] [PubMed] [Google Scholar]
  6. Desmouliere A, Geinoz A, Gabbiani F, Gabbiani G (1993) Transforming growth factor-β1 induces α-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts. J Cell Biol 122:103–111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fiore E, Fusco C, Romero P, Stamenkovic I (2002) Matrix metalloproteinase 9 (MMP-9/gelatinase B) proteolytically cleaves ICAM-1 and participates in tumor cell resistance to natural killer cell-mediated cytotoxicity. Oncogene 21:5213–5223 [DOI] [PubMed] [Google Scholar]
  8. Graves LE, Ariztia EV, Navari JR, Matzel HJ, Stack MS, Fishman DA. (2004) Proinvasive properties of ovarian cancer ascites-derived membrane vesicles. Cancer Res 64:7045–7049 [DOI] [PubMed] [Google Scholar]
  9. Heppner KJ, Matrsian LM, Jensen R A, Rodgers WH (1996) Expression of most matrix metalloproteinase family members in breast cancer represents a tumor-induced host response. Am J Pathol 149:279–282 [PMC free article] [PubMed] [Google Scholar]
  10. Ishikawa S. Takenaka K, Yanagihara K, Miyahara R, Kawano Y, Otake Y, Hasegawa S, Wada H, Tanaka F (2004) Matrix metalloproteinase-2 status in stromal fibroblasts, not in tumor cells, is a significant prognostic factor in non-small-cell lung cancer. Clin Cancer Res 10:6579–6585 [DOI] [PubMed] [Google Scholar]
  11. Koyama S (2000) Apoptotic depletion of infiltrating mucosal lymphocytes associated with Fas ligand expression by Helicobacter pylori-infected gastric mucosal epithelium: Human glandular stomach as a site of immune privilege. Dig Dis Sci 45:773–780 [DOI] [PubMed] [Google Scholar]
  12. Koyama S (2003a) Flow cytometric measurement of tumor necrosis factor-related apoptosis inducing ligand and its receptors in gastric epithelium and infiltrating mucosal lymphocytes in Helicobacter pylori-associated gastritis. J Gastroenterol Hepatol 18:763–770 [DOI] [PubMed] [Google Scholar]
  13. Koyama S (2003b) Differential expression of intracellular apoptotic signaling molecules in tumor and tumor-infiltrating lymphocytes during the development of invasion and/or metastasis of gastric carcinoma. Dig Dis Sci 48:2290–2300 [DOI] [PubMed] [Google Scholar]
  14. Koyama S (2004a) Enhanced cell-surface expression of matrix metalloproteinases and their inhibitors, and tumor-induced host response in progression of human gastric carcinoma. Dig Dis Sci 49:1621–1630 [DOI] [PubMed] [Google Scholar]
  15. Koyama S (2004b) Significance of cell-surface expression of matrix metalloproteinases and their inhibitors on gastric epithelium and infiltrating mucosal lymphocytes in progression of Helicobacter pylori-associated gastritis. Scand J Gastroenterol 39:1046–1053 [DOI] [PubMed] [Google Scholar]
  16. Koyama S, Maruyama T, Adachi S, Nozue M (1998) Expression of costimulatory molecules, B7–1 and B7–2 on human gastric carcinoma. J Cancer Res Clin Oncol 124:383–388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Koyama S, Maruyama T, Adachi S (1999) Expression of epidermal growth factor receptor and CD44 splicing variants sharing exsons 6 and 9 on gastric and esophageal carcinomas: a two-color flow-cytometric analysis. J Cancer Res Clin Oncol 125:47–54 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Koyama S, Koike N, Adachi S (2001) Fas receptor counterattack against tumor-infiltrating lymphocytes in vivo as a mechanism of immune escape by gastric carcinoma. J Cancer Res Clin Oncol 127:20–26 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Koyama S, Koike N, Adachi S (2002) Expression of TNF-related apoptosis inducing ligand (TRAIL) and its receptors in gastric carcinoma and tumor-infiltrating lymphocytes: a possible mechanism of immune evasion of the tumor. J Cancer Res Clin Oncol 128:73–79 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liotta LA, Kohn EC (2001) The microenvironment of the tumor-host interfaces. Nature 411:375–379 [DOI] [PubMed] [Google Scholar]
  21. Nakayama H, Enzan H, Miyazaki E, Toi M (2002) α Smooth muscle actin positive stromal cells in gastric carcinoma. J Clin Pathol 55:741–744 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Olumi AF, Grossfeld GD, Hayward SW, Carroll PR, Tlsty TD, Cunha GR (1999) Carcinoma-associated fibroblasts direct progression of initiated human prostatic epithelium. Cancer Res 59:5002–5011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Orimo A, Tomioka Y, Shimizu Y, Sato M, Oigawa S, Kamata K, Nogi YT, Inoue S, Takahashi M, Hata T, Muramatsu M (2001) Cancer-associated myofibroblast possess various factors to promote endometrial tumor progression. Clin Cancer Res 7:3097–3105 [PubMed] [Google Scholar]
  24. Poulsom R, Pignatelli M, Stetler-Stevenson WG, Liotta LA, Wright PA, Jeffery RE, Longcroft JM, Rogers L, Stamp GWH (1992) Stromal expression of 72 kda type IV collagenase (MMP-2) and TIMP-2 mRNAs in colorectal neoplasia. Am J Pathol 141:389–396 [PMC free article] [PubMed] [Google Scholar]
  25. Sappino AP, Schurch W, Gabbiani G (1990) Differentiation repertoire of fibroblastic cell: expression of cytoskeletal proteins as marker of phenotypic modulations. Lab Invest 63:144–161 [PubMed] [Google Scholar]
  26. Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen D, Gabbiani G (1986) A monoclonal antibody against α-smooth muscle actin: A new probe for smooth muscle differentiation. J Cell Biol 103:2787–2796 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Tuxhorn JA, Ayala GE, Smith MJ, Smith VC, Dang TD, Rowley DR (2002) Reactive stroma in human prostate cancer: induction of myofibrblast phenotype and extracellular matrix remodeling. Clin Cancer Res 8:2912–2923 [PubMed] [Google Scholar]
  28. Yen TWF, Aardal NP, Bronner MP, Thorning DR, Savard CE, Lee SP, Bell RH Jr (2002) Myofibroblasts are responsible for the desmoplastic reaction surrounding human pancreatic carcinomas. Surgery 131:129–134 [DOI] [PubMed] [Google Scholar]
  29. Yu Q, Stamenkovic I (1999) Localization of matrix metalloproteinase MMP-9 to the cell surface provides a mechanism for CD44-mediated tumor invasion. Genes Dev 13:35–48 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Yu WH, Woessner JF Jr, McNeish JD, Stamenkovic I (2002) CD44 anchors the assembly of matrilysin/MMP-7 with heparin-binding epidermal growth factor receptor and ErbB4 and regulates female reproductive organ remodeling. Genes Dev 16:307–323 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Cancer Research and Clinical Oncology are provided here courtesy of Springer

RESOURCES