Cancer Cell Biology
Emmprin in epithelioid sarcoma: Expression in tumor cell membrane and stimulation of MMP-2 production in tumor-associated fibroblasts
Article first published online: 27 NOV 2006
Copyright © 2006 Wiley-Liss, Inc.
International Journal of Cancer
Volume 120, Issue 4, pages 761–768, 15 February 2007
How to Cite
Koga, K., Nabeshima, K., Aoki, M., Kawakami, T., Hamasaki, M., Toole, B. P., Nakayama, J. and Iwasaki, H. (2007), Emmprin in epithelioid sarcoma: Expression in tumor cell membrane and stimulation of MMP-2 production in tumor-associated fibroblasts. Int. J. Cancer, 120: 761–768. doi: 10.1002/ijc.22412
- Issue published online: 27 DEC 2006
- Article first published online: 27 NOV 2006
- Manuscript Accepted: 14 SEP 2006
- Manuscript Received: 24 MAR 2006
- Ministry of Education, Culture, Sports, Science and Technology of Japan. Grant Numbers: 16659109, 15390119
- Central Research Institute
- Fukuoka University School of Medicine. Grant Number: 041001
- epithelioid sarcoma;
- matrix metalloproteinase;
- stromal fibroblasts
Emmprin is a transmembrane glycoprotein on tumor cells that stimulates peritumoral fibroblasts to produce matrix metalloproteinases (MMPs). Emmprin and the induced MMPs play a crucial role in tumor progression, invasion and metastasis of human carcinomas (epithelial malignancies). However, only a few reports have addressed its role in soft tissue sarcomas. This study investigated the expression and role of emmprin in epithelioid sarcoma (ES). Immunoblot studies of 2 ES cell lines showed that they express emmprin, and co-culture of these ES cells with dermal fibroblasts resulted in upregulation of gelatinase A (MMP-2) in fibroblasts, as shown by zymography, immunoblotting and enzyme immunoassay. This stimulation was inhibited by an activity-blocking peptide against emmprin and by antiemmprin antibody. In addition, in vivo, immunohistochemical analysis of 5 ES patient cases demonstrated diffuse emmprin expression in ES cells and MMP-2 expression in both ES cells and peritumoral fibroblasts. The histopathological findings that peritumoral fibroblasts that were not in direct contact with emmprin-expressing ES cells exhibit upregulated MMP-2 prompted us to look for a soluble form of emmprin. Soluble full-length emmprin released from ES cells was detected in conditioned medium and shown to stimulate MMP-2 production by fibroblasts. In conclusion, emmprin is expressed in ES in both membrane and soluble forms and stimulates MMP-2 production via interactions with fibroblasts, which could play a role in ES cell stromal invasion and vascular involvement. © 2006 Wiley-Liss, Inc.
Degradation of extracellular matrix (ECM) components by matrix metalloproteinases (MMPs) is critical for tumor cell invasion and metastasis.1In situ hybridization studies using surgical specimens have demonstrated that most tumor-associated MMPs, such as interstitial collagenase (MMP-1), stromelysin-1 (MMP-3), stromelysin-3 (MMP-11) and gelatinase A (MMP-2), are mainly synthesized in peritumoral fibroblasts rather than tumor cells in breast, colon, lung, skin and head and neck cancers.2, 3, 4In vitro studies indicated that cell–cell interaction between tumor cells and fibroblasts is crucial for the upregulation of MMPs in the fibroblasts,5, 6, 7, 8, 9 and the upregulated fibroblast MMPs facilitate invasion of the ECM by carcinoma cells.10 These observations suggest the presence of an inducer molecule(s) that mediates the interaction between tumor cells and fibroblasts and thus its critical role in the regulation of MMPs during tumor progression.
Emmprin is a transmembrane glycoprotein that is frequently expressed on tumor cells and stimulates nearby fibroblasts to produce MMP-1, 2 and 3.7 It belongs to the immunoglobin (Ig) superfamily.7 Emmprin mediates MMP stimulation via direct cell–cell contact or as a soluble form(s) released from the cell surface.6, 11, 12 Two mechanisms of release are known: the first involves shed membrane vesicles that contain full-length emmprin,13, 14 and the second a metalloproteinase-dependent generation of a proteolytic cleavage product of emmprin lacking the carboxyl terminus.15 This tumor cell emmprin, in both membrane and soluble forms, appears to be responsible, at least in part, for the stimulation of MMP production by fibroblasts that is observed in vivo in a variety of tumors. This rationale for the function of emmprin is supported by several clinical studies. For example, the expression of emmprin is upregulated in urinary bladder, breast, lung, oral and esophageal carcinomas.16, 17, 18, 19 Furthermore, elevated emmprin expression correlated with tumor progression in gliomas, serous ovarian carcinoma and melanoma.9, 20, 21
Epithelioid sarcoma (ES) is a rare, aggressive soft-tissue tumor generally presenting as a subcutaneous or deep dermal mass in the upper extremities of adolescents and young adults, and has high risk for local recurrence and metastasis.22, 23, 24 Histologically, ES is characterized by nodular aggregates of cytologically malignant epithelioid and/or spindle cells, often with central necrosis.22, 23, 24 In an in vitro study, human ES cell lines secreted MMP-2 and MMP-9 and their low invasive potential was associated with overexpression of distinct tissue inhibitor of metalloproteinases (TIMPs) relative to the MMP-2 and MMP-9.25 However, the role of tumor cell–fibroblast interactions, specifically via emmprin, in the regulation of MMP levels has yet to be elucidated in ES. Thus, in this study, we investigated tumor cell–fibroblast interactions in ES with special reference to emmprin expression and MMP production both in vitro and in vivo. Additionally, we investigated emmprin release from ES cells.
Material and methods
Tissues and cells
This study included 5 patients with ESs (4 males and 1 female; age range 22–75 (mean = 52.6) years) diagnosed at the Department of Pathology, Fukuoka University, Japan. The anonymous use of redundant tissue is part of the standard treatment agreement with patients in our hospital when no objections have been expressed. Each specimen obtained at surgery was fixed in 20% formalin and embedded in paraffin. Histological diagnosis was made according to the WHO standards.26
FU-EPS-127 and SFT860628 cell lines were both established from patients with ESs who had not received any chemotherapy before surgical resection. The human dermal fibroblast ST353 was obtained from nonlesional dermis around nodular fascitis. Other human fibroblasts, CCD-25Sk and CCD-32Sk, were purchased from the American Type Culture Collection (Manassas, VA). These cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 U/ml penicillin G and 100 μg/ml streptomycin) in humidified atmosphere of 5% CO2 at 37°C.
Immunostaining of formalin-fixed, paraffin-embedded tissue sections was performed using a biotin-streptavidin method. Briefly, sections were deparaffinized, rehydrated in descending alcohol dilutions and washed in Tris-buffered saline (TBS), pH 7.6. After nonspecific sites were blocked with 3% bovine serum albumin and 1% nonfat dry milk in TBS for 10 min at room temperature, the sections were incubated with primary antibodies overnight at 4°C. Primary antibodies used included a monoclonal antibody (mAb) against MMP-2 (clone 75-7F7, Daiichi Fine Chemical Co., Takaoka, Japan) and a goat polyclonal antibody against emmprin (R&D System, Flanders, NJ). The sections were then washed in TBS, and incubated with biotinylated horse anti-mouse (Vector Laboratories, Burlingame, CA) or swine anti-goat IgG (DAKO, Carpinteria, CA) for 30 min at room temperature, followed by streptavidin conjugated to alkaline phosphatase (DAKO) for another 30 min. The reaction was detected with naphthol AS-BI phosphate (Sigma Chemical Co., St. Louis, MO) in 100 ml of 0.2 M TBS (pH 8.2) containing 4% hydrochloric acid and 4% nitric acid and counterstained with hematoxyline. The immunohistochemical specificity of the antibodies was confirmed by 2 types of negative controls: substituting mouse or goat nonimmune IgG for the primary antibodies and omitting the primary antibodies from the staining protocol.
Double immunostaining for emmprin and AE1/AE3 cytokeratins was performed using Envision labeled polymer reagent (DAKO) and a biotin-streptavidin method. First, sections were stained for emmprin. After nonspecific sites were blocked as described earlier, the sections were incubated with antiemmprin mAb (R&D System) for 1 hr at room temperature. The sections were then washed in TBS, and incubated with Envision reagent for 30 min at RT. The attached antibody was visualized by the 3,3′-diaminobenzidine (DAKO). The primary and secondary antibodies were removed from tissue by washing in 3 changes of 0.1 mol/l glycine-HCl (pH 2.2) for 1 hr. Then, the biotin-streptavidin method was repeated using AE1/AE3 anticytokeratin antibody (DAKO) as described earlier. The reaction was visualized with naphthol AS-BI phosphate or fast blue (Sigma).
Co-culture experiments were performed as described previously.6 Briefly, cultures containing either fibroblasts (ST353, CCD-25Sk, CCD-32Sk), ES cells (FU-EPS-1, SFT8606) or both were established in 20 mm diameter wells containing 1.0 ml DMEM with 10% FBS. A fixed number (4 × 104) of fibroblasts were incubated with increasing numbers (2, 4 and 8 × 104) of ES cells. The cells were allowed to attach for 24 hr at 37°C, after which their media were replaced with fresh serum-free (SF) DMEM containing 0.2% lactalbumin hydrolysate (0.5 ml/well) prior to beginning the experiment. Each experimental condition was carried out in duplicate or triplicate wells. Culture fluids were replaced with fresh SF DMEM at 3 days and harvested at 6 days. The harvested media were used for zymography, immunoblotting and enzyme immunoassay (EIA).
Experiments employing the activity-blocking peptide emp no. 2 were performed as described previously.29 Briefly, the FU-EPS-1 ES cells were preincubated with various concentrations of emp no. 2 at 37°C for 45 min before starting co-culture with fibroblasts. Thereafter, the experiment was performed in exactly the same way as described earlier except for the presence of emp no. 2. The emp no. 2 peptide consists of 20 amino acids (aa) (SLNDSATEVTGHRWLKGGVV) that correspond to the second 20-aa peptide from the N-terminus of the functionally active first Ig domain (ECI) of the emmprin molecule. This peptide specifically inhibits emmprin activity in co-cultures using transformed T-cells and fibroblasts.29 Relatively high doses (250 and 500 μg/ml) were needed for this inhibition, which was similar to a case of inhibition of lectin activity by a mouse neural cell adhesion molecule (NCAM)-derived 19-aa peptide.30 The fourth Ig-like domain of NCAM is a lectin domain, and the 19-aa peptide corresponds to a part of it. ECI of emmprin is homologous to this lectin domain of NCAM and exhibits lectin activity.31 Furthermore, inhibition experiments in co-cultures using antiemmprin blocking antibody (UM-8D6, Ancell Corporation, Bayport, MN) were performed as described previously.8
Gelatinolytic activities in conditioned media were demonstrated using gelatin as a substrate, as described previously.8 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed under nonreducing conditions using a 9% separating gel containing 1 mg/ml gelatin. After electrophoresis, the gel was shaken gently in detergent buffer (5 mM CaCl2, 2.5% Triton X-100 and 50 mM Tris-HCl, pH 7.6) at room temperature for 60 min to remove SDS, and then incubated in reaction buffer (0.15 M NaCl, 10 mM CaCl2, 0.02% NaN3 and 50 mM Tris-HCl, pH 7.6) at 37°C for 30 hr followed by staining with 2.5% Coomassie brilliant blue in 30% methanol and 10% acetate. Enzyme activity was detected as a clear band on the resulting blue background of undigested gelatin. Photos of the gels were analyzed digitally (Adobe Photoshop, Adobe System, Mountain View, CA) to quantitate relative activities.
Preparation and extraction of membranes
Extraction of emmprin from cell membranes was performed as described previously.6 Tumor cells and fibroblasts were harvested from confluent cultures in 2 culture bottles (25 cm2) by mechanical scraping and were then suspended and sonicated in 50 mM Tris-HCl (pH 7.4)/0.24 M sucrose. The sonicated cell suspension was centrifuged at 500g for 20 min and the supernatant was centrifuged at 100,000g for 1 hr. The membrane pellet was extracted with 0.15 M NaCl/0.5% Nonidet P-40/1 mM ethylenediaminetetra-acetic acid (EDTA)/2 mM phenylmethylsulfonyl fluoride/10 mM Tris-HCl (pH 8.2), at 4°C overnight. The extract was centrifuged at 100,000g for 1 hr and the supernatant was kept at 4°C until use. The presence of emmprin in the membrane extract was assayed using immunoblotting.
SDS-PAGE of membrane extracts and conditioned medium was performed under reducing conditions using a 5–15% gradient gel (Biomate, Tokyo, Japan). After electrophoresis, the proteins were transferred electrophoretically to Immobilon membrane (Millipore, Bedford, MA). Nonspecific sites were blocked with 5% nonfat dry milk in 0.05% Tween-20/TBS, pH 7.6 (TBS-T) at 37°C for 1 hr and the membrane was incubated overnight at 4°C with antibodies against MMP-2 (monoclonal 75-7F7, Daiichi Fine Chemical) or emmprin (monoclonal E11F46; goat polyclonal (R&D System); rabbit polyclonal against N- and C-terminal portions of emmprin (Santa Cruz Biotechnology, Santa Cruz, CA)) dissolved in TBS-T containing 1% BSA. After washing with TBS-T, the membrane was incubated for 1 hr with peroxidase-conjugated goat anti-mouse, rabbit anti-goat or goat anti-rabbit IgG. The color was developed with chemiluminescence reagents (DuPont NEN, Boston, MA) according to the instructions supplied by the manufacturer. The bands on the film were subjected to image analysis (Adobe Photoshop). Statistical analysis was performed using Student's t test.
Protein concentrations of pro-MMP-2 were measured using a commercially available EIA kit (Daiichi Fine Chemical) using the protocols supplied by the manufacturer. All assays were performed in triplicate and statistical analyses were performed using Student's t test.
Treatment of fibroblasts with ES-CM containing concentrated or reduced emmprin
ES-CM was obtained from confluent FU-EPS-1 cell culture in SF DMEM containing 0.2% LH for 3 days and concentrated using Centricon (Millipore). ES-CM protein concentrations were determined by the Bradford assay (Bio-Rad, Hercules, CA). To reduce emmprin levels, 60 μl of ES-CM were incubated with 1 μl antiemmprin monoclonal antibody (R&D System) at 4°C for 4 hr, followed by the addition of 20 μl agarose-conjugated protein A (Bio-Rad) (4°C for 1 hr) and centrifugation at 4,000 rpm for 1 min. The supernatant was added to fibroblasts. As control, ES-CM was incubated with non-immune IgG.
Emmprin expression in human ES cells and fibroblasts
First, we determined whether human ES cell lines FU-EPS-1 and SFT8606 and fibroblasts expressed emmprin. Immunoblots of membrane extracts prepared from FU-EPS-1 and SFT8606 showed a single band at 58 kDa. FU-EPS-1 also showed a broad band ranging from ∼35 to 50 kDa (Fig. 1). It is likely that this lower molecular weight broad band was generated by degradation because it appeared in both FU-EPS-1 and SFT8606 cells when more proteins were applied, and mAb from R&D System detected this band more efficiently than mAb E11F4. Fibroblasts (ST353, SK25 and SK32) also expressed emmprin as a single band at 58 kDa, although their expression levels were lower than those of ES cell lines; especially the band was barely visible for ST353.
Emmprin-dependent stimulation of fibroblast MMP-2 production by ES cells
Next, we investigated whether the emmprin expressed on ES cells could stimulate host fibroblasts to produce MMPs by using co-culture experiments. Figure 2a shows gelatin zymography of the conditioned media harvested on day 6 of co-cultures of FU-EPS-1 or SFT8606 ES cells with ST353 fibroblasts. Fibroblasts exhibited a weak gelatinolytic band at 68 kDa, while SFT8606 ES cells generated only a faint band and FU-EPS-1 ES cells did not produce any detectable gelatinolytic activities. In co-cultures, FU-EPS-1 ES cells enhanced the gelatinolytic activity at 68 kDa by ∼4.1, 4.8 and 5.8 times at the ratio of 1:0.5, 1:1 and 1:2 of fibroblasts to ES cells, respectively when compared with individual cultures of each cell type, and SFT8606 ES cells by ∼2.0, 2.4 and 2.7 times, respectively (Figs. 2a and 2b). This 68-kDa band corresponds to the molecular mass of the pro-form of MMP-2.32 The mechanisms involved in this difference between the 2 cell lines in inducing MMP-2 production in co-culture with fibroblasts are currently unknown. Moreover, FU-EPS-1 ES cells stimulated 3 fibroblast lines in co-culture, i.e. ST353, CCD-25Sk and CCD-32Sk, resulting in the enhancement of pro-MMP-2 gelatinolytic activity by approximately 5.8, 2.1 and 3.2 times, respectively (Fig. 2c). This upregulation of pro-MMP-2 in co-cultures was also confirmed by both an immunoblot and EIA analysis of conditioned medium with anti-MMP-2 MAb; only pro-MMP-2 reacted with this antibody (Figs. 3a and 3b). By immunoblotting, co-culture of FU-EPS-1 ES cells and ST353 fibroblasts stimulated the production of pro-MMP-2 by ∼6.0 times when compared with individual cultures (Fig. 3a). Consistent with zymography and immunoblot results, quantitative assessment by EIA also showed a significant increase in pro-MMP-2 production in co-cultures, i.e. 12.0 ± 1.1 ng/ ml (mean ± SEM) of pro-MMP-2 in individual cultures versus 110 ± 25 ng/ml in 1:1 co-cultures (Fig. 3b).
To determine the role of emmprin in the above stimulation of MMP-2 production, we performed co-culture experiments in the presence of various concentrations of emmprin activity-blocking emp no. 2 peptide and antibody (Figs. 3a–3c). Immunoblotting studies showed that emp no. 2 peptide inhibited the stimulation in a dose-dependent manner, i.e. ∼19.1% inhibition at 100 μg/ml, 55.1% at 250 μg/ml and 91.2% at 500 μg/ml, while a control 20-aa peptide did not cause any inhibition (Fig. 3a). On zymograms, emp no. 2 peptide showed (85.5 ± 4.4)% inhibition (n = 3) at 500 μg/ml (data not shown). This inhibition was also confirmed by EIA; 110 ± 25 ng/ml of MMP-2 produced in co-cultures was reduced to 38.8 ± 1.6 ng/ml (n = 3) by incubation of tumor cells with 500 μg/ml emp no. 2 peptide (∼73% inhibition) (Fig. 3b) while a control peptide caused 23% reduction. Emmprin activity-blocking antibody (UM-8D6) also demonstrated a dose-dependent inhibition of stimulation of MMP-2 production in co-cultures, i.e. ∼28.5% inhibition at 10 μg/ml and 59.6% inhibition at 40 μg/ml, while control non-immune IgG did not cause any inhibition (Fig. 3c). This was again confirmed by EIA; 48 ng/ml of MMP-2 produced in co-cultures was reduced to 14 ng/ml (mean, n = 2) by 40 μg/ml UM-8D6 (∼86% inhibition), while the control IgG caused no inhibition (0%). These data support the occurrence of emmprin-dependent MMP-2 upregulation in the co-cultures and suggest that the stimulation of MMP-2 production in co-culture is predominantly attributable to cell–cell interaction via emmprin.
Expression of emmprin and MMP-2 in co-cultures
Using immunocytochemistry, we also examined the expression of emmprin and MMP-2 in the co-cultures. ES cells were clearly distinguished from fibroblasts by their polygonal appearance and expression of cytokeratins (AE1/AE3, Fig. 4a). Both ES cells and fibroblasts showed cytoplasmic expression of MMP-2, but more prominent expression was observed in fibroblasts (Fig. 4b). Emmprin was predominantly demonstrated along the cell membrane of ES cells, while perinuclear staining was occasionally observed in fibroblasts (Fig. 4c). Double immunostaining for emmprin and AE1/AE3 cytokeratins was performed to confirm differentiation between ES cells and fibroblasts. Many of polygonal ES cells expressed both cytokeratin and emmprin (cytokeratin in the cytoplasm and emmprin along the cell membrane and also focally in the cytoplasm), whereas spindle-shaped fibroblasts showed only weak reactivity for emmprin in the cytoplasm (Fig. 4d).
Expression of emmprin and MMP-2 in human ES tissues
Since ES cells stimulated MMP-2 production by fibroblasts via emmprin in vitro, we examined the expression of emmprin and MMP-2 in human ES tissue specimens (4 males and 1 female; age range 24–75 (mean = 53) years) by immunohistochemistry. All cases evaluated in this study displayed typical histological features, including nodular collections of cytologically malignant epithelioid and spindle cells with eosinophilic cytoplasm separated by dense collagen and often revealing central necrosis (Fig. 5a). In all cases, MMP-2 was detected in the cytoplasm of ES cells and surrounding peritumoral fibroblasts (Fig. 5b), while emmprin was expressed predominantly on ES cells, along the cell membrane (Fig. 5c). Staining results are summarized in Table I. Double immunostaining for emmprin and AE1/AE3 cytokeratins was performed in 3 cases to differentiate ES cells from intratumoral fibroblasts. In general, ES cells frequently express AE1/AE3, while fibroblasts do not. As expected, many ES cells demonstrated co-expression of emmprin and cytokeratins (Fig. 5d, inset). Furthermore, emmprin expression was also occasionally detected in some fibroblasts present within the tumor, which stained negative for AE1/AE3 (Fig. 5e), whereas fibroblasts outside the tumor were negative (Fig. 5f), with the exception of 1 case (Table I). Although AE1/AE3-negative and emmprin-positive ES cells were occasionally seen, they could be differentiated from fibroblasts by their tumor cell features (such as large rounded nuclei, high N/C (nuclei/cytoplasm) ratio and occasional prominent nucleoli) and epithelioid features (polygonal cell shapes with cell–cell contact with one another) (Figs. 5e and 5f).
|Tumor cell||Intratumoral fibroblast||Peritumoral fibroblast||Tumor cell||Peritumoral fibroblast|
Release of full-length emmprin
Since peritumoral fibroblasts that were not in direct contact with emmprin-expressing ES cells showed up-regulated MMP-2 in tissue sections, we speculated that emmprin might be released from ES cells. Two mechanisms for emmprin release have been demonstrated. One is metalloproteinase-dependent generation of a proteolytic cleavage product of emmprin lacking the carboxyl terminus.15 The other is release of full-length emmprin from the cell surface via microvesicle shedding.14 To distinguish between these 2 possibilities, we examined the presence and the size of emmprin released into ES-conditioned medium. Immunoblotting with antiemmprin mAb E11F4 detected emmprin of 58 kDa, which also reacted with antibodies raised against N- and C-terminal portions of emmprin (Fig. 6a), indicative of the release of full-length emmprin. This emmprin-containing CM stimulated ST353 fibroblasts to produce pro-MMP-2 in a dose-dependent manner on immunoblots (Fig. 6b). EIA confirmed this stimulation; 242 ng/ml of pro-MMP-2 produced in nontreated fibroblasts (control) was stimulated to 754, 908, 958 and 1,193 ng/ml by 1×, 2×, 5× and 10×-concentrated ES-CM (mean, n = 2). This stimulation was partially inhibited on immunoblots when CM was pretreated with antiemmprin mAb before being added to cells, followed by addition of protein A agarose and centrifugation (Fig. 6c). This partial inhibition was also confirmed by EIA; 1,053 ng/ml of MMP-2 stimulated by ES-CM was reduced to 699 ng/ml (mean, n = 2) by the pretreatment. These results indicate that full-length emmprin released into CM can upregulate MMP-2 in fibroblasts.
To our knowledge, this is the first study of soft tissue sarcoma that has explored the role of emmprin in tumor cell–fibroblast interaction that leads to MMP-2 upregulation. Stimulation of MMP-2 production in co-cultures of 2 human ES cell lines and 3 human dermal fibroblasts was demonstrated quantitatively using gelatin zymography, immunoblotting and EIA. Emmprin activity-blocking emp no. 2 peptide and antiemmprin antibody both inhibited this stimulation in co-cultures, indicating involvement of emmprin in the stimulation. ES cells also released full-length emmprin into CM that enhanced MMP-2 production in fibroblasts. Furthermore, both in vitro and in vivo, ES cells predominantly expressed emmprin, while MMP-2 expression was shown in both ES cells and fibroblasts adjacent to the tumor cells. Emmprin-enhanced stromal MMP-2 production may be involved in ES tumor progression by facilitating tumor invasion and angiogenesis.
Experimental studies have demonstrated that emmprin induces several malignant properties associated with cancer, including invasiveness,18 angiogenesis,33 anchorage-independent growth34 and chemoresistance.35 Furthermore, emmprin expression in primary breast and ovarian cancer tissue correlates with tumor size and staging, and is predictive of poor prognosis.21, 36 Correlations between emmprin expression and malignancy have also been demonstrated in other cancers.9, 18, 20 Although the relationship between the expression of emmprin and malignancy of soft tissue sarcoma has not been investigated, there are reports demonstrating that MMP-2 expression correlates with malignant progression of soft tissue sarcoma in vivo.37, 38, 39, 40 ES tissues immunohistochemically revealed strong, diffuse MMP-2 expression,40 and in vitro study using ES cell lines demonstrated that low invasive potential was associated with overexpression of TIMPs relative to MMP-2 and -9.25 These observations suggest that ES cell-associated emmprin may be involved in ES invasion via interactions with stromal fibroblasts and consequent upregulation of MMP-2.
The possibility that emmprin is released from the cell surface, as a result of either proteolytic scission or shedding of membrane vesicles, was raised in early studies.5, 11 Recently, the presence of soluble emmprin in tumor cell culture medium has been shown by several groups.13, 14, 15 Full-length soluble emmprin that promoted MMP-2 production from fibroblasts was released from breast13 and lung14 cancer cells in culture. Microvesicle shedding was proposed as a mechanism for this type of emmprin release: emmprin is initially associated with microvesicles that are quickly degraded upon release from the cells to release full-length bioactive emmprin.14 On the other hand, MMP-dependent proteolytic shedding of the functionally active extracellular domain of emmprin was also described.15 Whatever the mechanism of emmprin release, soluble emmprin would be able to exert its effect on fibroblasts at distant sites as long as it is bioactive. In our study, ES cells released soluble full-length emmprin that retained both its C- and N-terminal domains. Furthermore, the soluble emmprin stimulated MMP-2 production in fibroblasts. These findings suggest that ES cells release emmprin via microvesicle shedding as described,14 resulting in MMP-2 production in fibroblasts apart from tumor nests. This is in accordance with the immunohistochemical findings in tissue sections that peritumoral fibroblasts, which were not in direct contact with emmprin-expressing ES cells, showed strong MMP-2 expression while those in nondiseased control soft tissue were negative.
Recent studies have shown possible factors that regulate emmprin expression.41, 42 The possibility of a positive feedback regulatory mechanism of emmprin expression in fibroblasts was also demonstrated.15 When fibroblasts were exposed to an emmprin stimulus, emmprin expression was upregulated in these cells at both mRNA and protein levels. In vivo, it was reported that emmprin expression was confined to tumor cells and not detected in the stroma in some tumors, including melanoma9 and lung and breast cancer.17 However, some other reports demonstrated emmprin mRNA expression in peritumoral fibroblasts in ovarian21 and breast43 carcinoma tissues. Moreover, in a nude mouse xenograft model, overexpression of emmprin in human tumor cells resulted in a profound increase in mouse emmprin expression in host cells, both at the periphery of the tumor and infiltrated into the tumor tissue.33 We also found emmprin expression in fibroblasts within ES tumors (intratumoral fibroblasts in close proximity to tumor cells), but not in the peritumoral fibroblasts. Additionally, in co-culture experiments, it is possible that a fibroblastic source of emmprin could auto-stimulate MMP-2 production by fibroblasts, although ES cell-expressed emmprin appears to play a major role since co-culture of fibroblasts with ES cells upregulated MMP-2 production more than 5 times when compared with culture of fibroblasts alone. The findings that immunocytochemical expression of emmprin in fibroblasts in co-culture appears to be increased mildly when compared with that in culture of fibroblasts alone (data not shown) also support this possibility. These observations may reflect a positive feedback regulatory mechanism of emmprin expression in fibroblasts.
The results of our study may suggest that the MMP-upregulation mechanism mediated by tumor-associated emmprin can be a potentially useful target in antitumor invasion therapy for soft tissue sarcomas.
- 25In vitro invasiveness of human epithelioid-sarcoma cell lines: association with cell motility and inverse correlation with the expression of tissue inhibitor of metalloproteinases. Int J Cancer 1999; 80: 406–12., , , , , , .
- 26Epithelioid sarcoma. In: MertensF, ed. WHO classification. Tumors of soft tissue and bone. Lyon: IARC Press, 2002. 205–7., .
- 32Concanavalin A produces a matrix-degradative phenotype in human fibroblasts. Induction and endogenous activation of collagenase, 72-kDa gelatinase, and Pump-1 is accompanied by the suppression of the tissue inhibitor of matrix metalloproteinases. J Biol Chem 1990; 265: 21141–51., .