FASEB J. Avanti Polar Lipids
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.07-101386.
 (The FASEB Journal. 2008;22:3059-3067.)
© 2008 FASEB
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
fj.07-101386v1
22/8/3059    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Becker, R.
Right arrow Articles by Christian, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Becker, R.
Right arrow Articles by Christian, S.

Tumor stroma marker endosialin (Tem1) is a binding partner of metastasis-related protein Mac-2 BP/90K

Renate Becker*, Martin C. Lenter{dagger}, Tobias Vollkommer*, Anja M. Boos*, Dennis Pfaff*, Hellmut G. Augustin*,1 and Sven Christian*,1

* Joint Research Division Vascular Biology of the Medical Faculty Mannheim, University of Heidelberg and the German Cancer Research Center, Heidelberg, Germany; and

{dagger} Genomics Group—Proteomics, Boehringer Ingelheim Pharma, Biberach, Germany

1Correspondence: H.G.A., Vascular Biology, Medical Faculty Mannheim, University of Heidelberg, and German Cancer Research Center Heidelberg, INF 581; D-69221, Heidelberg, Germany. E-mail: augustin{at}angiogenese.de; S.C., BHC-GDD-TD-TIV, P.O. Box 10709; D-42096, Wuppertal, Germany. E-mail: sven.christian{at}bayerhealthcare.com


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Tumor development involves complex bidirectional interactions between tumor cells and host stromal cells. Endosialin (Tem1) has been identified as a highly O-glycosylated transmembrane glycoprotein, which is specifically expressed by tumor vessel-associated pericytes and stromal fibroblasts of a wide range of human tumors. Recent experiments in endosialin-deficient mice have unraveled a critical role of endosialin in site-specific tumor progression and metastasis. To molecularly understand the mechanisms of endosialin function, we aimed to identify extracellular endosialin ligands and identified Mac-2 BP/90K as a specific interaction partner. Detailed biochemical analyses identified a C-terminal fragment of Mac-2 BP/90K, which was shown to contain binding sites for galectin-3, and collagens as the structures responsible for endosialin binding. Subsequent expression analysis of Mac-2 BP/90K in vivo revealed weak or no expression in most normal tissues and strong up-regulation in tumor cells of human neoplastic tissues. Intriguingly, the expression patterns of Mac-2 BP/90K and endosialin were mutually exclusive in all human tissues. Correspondingly, loss-of-function adhesion experiments of Mac-2 BP/90K-expressing tumor cells on endosialin-expressing fibroblasts revealed a repulsive outcome of the Mac-2 BP/90K interaction. Taken together, the experiments identify a novel repulsive interaction between endosialin on stromal fibroblasts and Mac-2 BP/90K on tumor cells.—Becker, R., Lenter, M. C., Vollkommer, T., Boos, A. M., Pfaff, D., Augustin, H. G., Christian, S. Tumor stroma marker endosialin (Tem1) is a binding partner of metastasis-related protein Mac-2 BP/90K.


Key Words: activated fibroblasts • mural cells


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
TUMOR DEVELOPMENT, PROGRESSION, and metastasis are highly dependent on the tumor stroma, consisting of new blood vessels, activated fibroblasts, infiltrating inflammatory cells, and extracellular matrix. Tumor blood vessels differ structurally and functionally from normal blood vessels, a characteristic that is extensively explored as a novel antitumor therapy (1 , 2) . Endosialin was originally described as the antigen recognized by mAb FB5 as a highly specific marker of tumor blood vessels. Immunohistochemical methods showed that the FB5 antigen is detectable on blood vessels of a majority of human tumors, but not in normal human tissues (3) . The antigen recognized by mAb FB5 was identified in parallel by two independent lines of investigations. First, endosialin was isolated by an affinity purification approach that identified the molecule as a transmembrane protein of the C-type lectin family. Biochemical analysis identified endosialin as 90-kDa core protein that is highly modified by O-glycosylation (4) . Second, endosialin was identified as "Tumor endothelial marker 1 (Tem1)" by a SAGE approach comparing transcripts from tumor endothelium to blood vessels from matching normal tissue (5) . Recently, we and others have shown that endosialin is not expressed by tumor endothelial cells but rather by tumor stroma fibroblasts and tumor vessel-associated pericytes (6 , 7) . The function of the protein during development and progression is unknown. Endosialin-deficient mice show a reduction in gastrointestinal tumor growth, vascularization, and metastasis, but no apparent alterations in early development or wound healing, suggesting a site-specific role for endosialin during tumor angiogenesis and metastatic tumor cell dissemination (8) . To further elucidate the role of endosialin in these processes, it is crucial to identify extracellular ligands or intracellular signaling mechanisms. Here, we report the identification of Mac-2 BP/90K as an extracellular ligand for endosialin. Expression of Mac-2 BP/90K is described to be elevated in human plasma of cancer patients, and its expression correlates with distant metastasis (9 , 10) . In the present study, we analyzed Mac-2 BP/90K expression in a wide range of human carcinomas and its corresponding normal tissue. We identified weak or no expression in most of the normal tissues tested and strong up-regulation in different carcinomas. In all cases, expression of Mac-2 BP/90K is of epithelial origin and is mutually exclusive compared to the endosialin expression in the mesenchymal compartment, suggesting a repulsive positioning effect of the endosialin-Mac-2 BP/90K interaction. In conclusion, the identification of Mac-2 BP/90K as a specific endosialin ligand may provide an important functional link between the expression of endosialin in the tumor stroma compartment and its role in metastatic spread.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Primary cells and cell lines
LA1–5s neuroblastoma cells were cultured in RPMI (PAA Laboratories, Pasching, Austria) with 10% fetal calf serum and 1% penicillin/streptavidin (PAA). HeLa cells were cultured in Dulbecco modified Eagle medium (DMEM; PAA) with 10% fetal calf serum and 1% penicillin/streptavidin (PAA).

Cloning of endosialin-IgG
The endosialin-IgG fusion protein was cloned into the baculoviral expression vector pVL1393 as described previously (11) . The extracellular part of endosialin (aa 1–360), consisting of the N-terminal signal sequence, the C-type lectin domain, the Sushi domain, and the three EGF-like domains, was generated by polymerase chain reaction (PCR) for subsequent cloning. Endosialin-IgG was expressed in SF9 cells (Spodoptera frugiperda) and purified via Protein-A Sepharose column (Amersham Pharmacia, Uppsala, Sweden).

Western blot analysis
Cell extracts or affinity-purified samples were separated by SDS-PAGE and transferred onto nitrocellulose membranes for 1 h at 100 V. The membranes were blocked overnight at 4°C with 5% milk powder in TBS-T (140 mM NaCl; 10 mM Tris-HCl, pH 7.4; and 0.05% Tween), and incubated with the primary antibody as described (0.5–2 µg/ml in TBS-T) for 1 h at RT. After extensive washing, the membranes were incubated with peroxidase-coupled secondary antibody, and bound antibody was detected with ECL (Amersham Biosciences, Uppsala, Sweden) and exposure to Biomax MR film (Eastman Kodak, Rochester, NY, USA).

Cell surface biotinylation
For cell surface expression analysis, cells (5x106 cells) were washed 3 times with cold PBS on a cell culture plate and incubated with biotinylation buffer (20 mM HEPES, pH 7.45; 5 mM KCl; 130 mM NaCl; 0.8 mM MgCl2; 1 mM CaCl2; and 0.5 mg/ml EZ link sulfo-NHS-biotin; Pierce Chemical, Rockford, IL, USA) for 1 h at 4°C. After removal of the reagent, cells were washed 3 times with wash buffer (50 mM Tris, pH 7.5; 150 mM NaCl; 1 mM MgCl2; and 1 mM CaCl2) and lysed in 1% Triton X-100 lysis buffer for 1 h. Lysates were centrifuged for 15 min at 15,000 g. Precipitated proteins were separated by SDS-PAGE and transferred to nitrocellulose. After blocking, nitrocellulose membranes were incubated with ExtrAvidin–peroxidase conjugates diluted at 1:5000 (Sigma-Aldrich, St. Louis, MO, USA) for 1 h at RT. Bands were detected after incubation with ECL reagent and exposure to Biomax MR film.

Enzymatic digest
For enzymatic digests, precipitates were washed, split into equal aliquots, and incubated for 16 h at 37°C with either 1 U of PNGase F (N-glycanase F) from Flavobacterium meningosepticum; 10 mU sialidase from Arthrobacter ureafaciens; a mixture of 10 mU sialidase and 0.5 mU O-glycosidase from Diplococcus pneumoniae (Roche Diagnostics, Indianapolis, IN, USA); a mixture of all three enzymes; or 25 µl of reaction buffer alone (50 mM sodium phosphate, pH 7.0, and 0.5% Triton X-100). Precipitates were washed and eluted in Laemmli buffer. Signals were detected after SDS-PAGE by exposure to Biomax MR film.

Protein purification
Affinity purification of Mac-2 BP/90K from LA1–5s detergent extracts was performed as described previously (12) . In brief, 2 x 108 cells were pelleted and lysed in 20 ml of RIPA buffer (1% Triton X-100; 0.5% deoxycholic acid; 0.1% SDS; 10 mM Tris-HCl, pH 7.6; 150 mM NaCl; and 1% protease inhibitor cocktail for mammalian cells; Sigma-Aldrich). Lysates were incubated with endosialin-IgG affinity matrix (2 µg of construct covalently coupled to 20 µl of Affi-Gel 10). Matrix beads were washed 3 times with 0.025% Triton X-100, 50 mM Tris-HCl (pH 8.4), 150 mM NaCl, 1 mM CaCl2, and 0.02% azide and washed twice with 25 mM Tris-HCl (pH 8.4) and 250 mM NaCl. The bound proteins were eluted with 30 µl SDS gel sample buffer. Affinity-purified proteins were separated on a 10% polyacrylamide gel and visualized by colloidal blue staining (Invitrogen, Carlsbad, CA, USA). Bands that appeared in the endosialin-IgG eluate, but not in the control, were cut out, digested with trypsin, and analyzed by mass spectroscopy using a matrix-assisted laser desorption ionization, time-of-flight instrument.

Immunoprecipitation assays
For reprecipitation studies, LA1–5s cells were metabolically labeled with 500 µCi [35S] methionine and 500 µCi [35S] cysteine in 1.2 ml methionine- and cysteine-free DMEM (Invitrogen) for 16 h. Cell lysates were precleared with Protein A-Sepharose (Amersham Pharmacia) and precipitated with FB5 or mouse IgG coupled to CNBr-Sepharose (Amersham Pharmacia) at a ratio of 2.0 mg/ml of gel or anti-Mac2 BP/90K mAb SP-2 (Axxora, San Diego, CA, USA; 1 mg/ml of gel) coupled via rabbit anti-mouse IgG (1 mg/ml of gel) to Protein A-sepharose. FB5 precipitates were eluted from beads using 0.1 M glycin and HCl (pH 2.5) containing 1% TX-100. The precipitate was split: 85% was subjected to a second precipitation for 16 h using anti-Mac-2 BP/90K mAb SP-2 coupled to protein A-sepharose, as described above, while 15% was retained for SDS-PAGE. Finally, precipitates were washed, eluted in Laemmli buffer, and reduced with 50 mM dithiothreitol. Signals were detected after SDS-PAGE by exposure to Biomax MR film.

Immunohistochemistry
Cryosections of normal and malignant human tissues (NCT, Heidelberg, Germany, and BioCat, Heidelberg, Germany) were acetone fixed and blocked with methanol and 3% H2O2 for 30 min, avidin and biotin for 10 min each, and goat serum for 20 min before incubating with either 10 µg/ml monoclonal mouse anti-Mac-2 BP/90K antibody SP-2 or monoclonal mouse anti-endosialin antibody FB5 for 1 h. ChromPure mouse IgG (Dianova, Hamburg, Germany) was used as a control. Bound antibody was detected by light microscopy after treatment with EnVision + Dual Link HRP secondary antibody (Dako, Glostrup, Denmark) and DAB solution (Dako). Nuclei were stained with Mayers hemalaun.

Cell adhesion assay
Endosialin-siRNA transfected and control-siRNA transfected fibroblasts were grown to confluence. After 48 h, Mac-2 BP/90K-siRNAs or control-siRNA transfected HeLa cells were prelabeled using the PKH26 red fluorescent cell linker kit (Sigma), and 2.5 x 105 HeLa cells were added on top of the fibroblast monolayers. Tumor cells were allowed to adhere for 30 min under rigid agitation. Cells were washed twice with PBS and fixed in 4% PFA. Quantification was performed with the particle analysis Cell-F software (Olympus, Tokyo, Japan) by counting the labeled cells under the x4 objective. Counting was performed in triplicate, and 9 spots were counted per well. The assay was performed twice, and the mean of two individual experiments was analyzed.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Identification of Mac-2 BP/90K as extracellular endosialin ligand
To characterize the function of endosialin in the stromal compartment of human tumors, we aimed to identify potential secreted or cell surface proteins that interact with endosialin. Therefore, we used an endosialin-IgG fusion protein (comprising the extracellular globular part—aa 1–360—of human endosialin) to pull down cell surface biotinylated proteins from an LA1–5s neuroblastoma cell extract. A major protein band at 75 kDa and a minor band at 90 kDa were specifically detected with streptavidin-HRP in endosialin-IgG (En-Fc) but not in control precipitation experiments (Fig. 1 A). Large-scale affinity purification from lysates of the endogenously endosialin-expressing cell line LA1–5s revealed, in Coomassie-stained gels, four specific bands interacting with endosialin-IgG, including the previously detected major band migrating at 75 kDa (Fig. 1B ). Mass spectrometry analysis identified endosialin- and IgG-derived peptide sequences (accession nos. gi 13277301 and gi 226787), resulting from a bleeding of the endosialin-IgG fusion protein from the matrix, as well as the beta subunit of F1-ATPase (accession no.: gi 28940) and Mac-2 BP/90K (accession no.: gi 483474). In contrast to Mac-2 BP/90K, the beta subunit of F1-ATPase failed to coimmunoprecipitate with endosialin from LA1–5s cell lysates (data not shown). To further confirm the specificity of the Mac-2 BP/90K-endosialin interaction, we compared precipitation experiments of protein lysates generated from cell surface biotinylated LA1–5s using endosialin-IgG or a monoclonal antibody against Mac-2 BP/90K as bait. Confirming previous literature data, the monoclonal anti-Mac-2 BP/90K antibody bound to cell surface-expressed proteins with a molecular mass of 75 and 90 kDa (Fig. 2 A). Endosialin-IgG bound the same cell surface-expressed proteins with a preferential binding of the 75-kDa form of Mac-2 BP/90K (Fig. 2A , asterisk). The specific interaction of the 75-kDa Mac-2 BP/90K species with endosialin was confirmed by copurification with endogenous endosialin expressed by LA1–5s in immunoprecipitation experiments using the endosialin-specific mAb FB5 (Fig. 2C , open arrow). Again, although the smaller 75-kDa form and the larger 90-kDa form of Mac-2 BP/90K were expressed by LA1–5s in almost equal amounts, there was a preference of the 75-kDa species to interact with endosialin. Western blot analysis showed that the 90-kDa form of Mac-2 BP/90K was preferentially expressed as a monomer or multimers secreted into the cell culture medium, whereas the 75-kDa was mainly cell surface associated (Fig. 2B ).


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		Figure 1. Identification of Mac-2 BP/90K as a endosialin ligand. A) Cell surface biotinylated LA1–5s neuroblastoma cells were used for precipitation experiments, identifying a major 75-kDa band and a minor 90-kDa band that precipitated with the endosialin-IgG matrix (EN-Fc) but not with the control matrix (ctrl). B) Large-scale purification of endosialin-interacting proteins identified several protein bands in a colloidal Coomassie stain that specifically bind to endosialin-IgG but not to the control matrix, including the previously identified 75-kDa band (asterisk).


Figure 2
View larger version (16K):
[in this window]
[in a new window]

  		Figure 2. Preferential binding of the cell surface-associated 75-kDa form of Mac-2 BP/90K by endosialin. A) Cell surface biotinylated LA1–5s neuroblastoma cells were used for precipitation experiments using control beads (ctrl), endosialin-IgG (EN-Fc), or a monoclonal anti-Mac-2 BP/90K antibody (SP-2) as bait. Endosialin-IgG specifically bound the cell surface exposed 75-kDa form of Mac-2 BP/90K (asterisk), whereas the monoclonal anti-Mac-2 BP/90K antibody bound both the 75-kDa and the 90-kDa form. Precipitated cell surface proteins were detected using streptavidin-peroxidase B) Western blot analyses of the localization of Mac-2 BP/90K in the cell culture supernatant (SN) vs. cell surface (cells) using monoclonal antibody SP-2 showed that the 90-kDa form was mainly secreted, whereas the 75-kDa form was mainly cell surface associated. C) Endosialin immunoprecipitation from metabolically labeled LA1–5s cells. A small aliquot (15%) of eluted proteins (50 mM glycin and HCl, pH 2.5) was subjected directly to SDS-PAGE (FB5), revealing the endosialin antigen at an apparent molecular mass of 165 kDa (solid arrow), whereas the majority of the eluate (85%) was subjected to anti-Mac-2 BP/90K reprecipitation (1. FB5; 2. SP-2), indicating the presence of the 75-kDa Mac-2 BP/90K species (open arrow) in the FB5 eluate. Mouse IgG bound to CNBr sepharose served as the control.

Binding of endosialin is mediated by the Mac-2 BP/90K core protein domains 3 and 4
Mac-2 BP/90K is known to interact with galectin-3 through its sugar moieties (13) . To test whether the interaction of endosialin and Mac-2 BP/90K is mediated by the Mac-2 BP/90K core protein or by its carbohydrate moieties, the binding of Mac-2 BP/90K to endosialin-IgG was competed with free glucose or lactose. None of the carbohydrates inhibited binding of Mac-2 BP/90K to endosialin, suggesting a protein-protein binding mode of the two proteins (Fig. 3 A). These findings were further substantiated by enzymatic removal of N- and O-linked carbohydrate moieties. Incubation of Mac-2 BP/90K bound to endosialin-Fc with N-Glycosidase F led to a small but distinct shift of the band recognized by monoclonal anti-Mac-2 BP/90K antibody SP-2, whereas incubation with enzymes removing O-linked sugars did not lead to a molecular mass shift in Western blot analysis. This finding is in line with the established preferential N-glycosylation of Mac-2 BP/90K. However, neither removal of N-linked nor of O-linked sugars inhibited the endosialin-Mac-2 BP/90K interaction (Fig. 3B ), strengthening the previous observation of a protein-protein interaction mode between the two proteins. Finally, binding of Mac-2 BP/90K to endosialin was found to be independent of divalent cations, since Mac-2 BP/90K captured by endosialin-IgG could not be released by 5 mM EDTA as indicated in the immunoprecipitation/Western blot analyses (Fig. 3C ).


Figure 3
View larger version (10K):
[in this window]
[in a new window]

  		Figure 3. Identification of the Mac-2 BP/90K C-terminal core protein as endosialin ligand. A) Lysates from LA1–5s cells were used for precipitation experiments using endosialin-IgG (EN-Fc) or anti-Mac-2 BP/90K. Endosialin-IgG specifically bound the 75-kDa band of Mac-2 BP/90K. The 75-kDa band was not eluated (eluate) after addition of 250 mM glucose or lactose and was instead still binding to the endosialin-IgG beads (beads). B) Endosialin-IgG bound the 75-kDa form of Mac-2 BP/90K with or without enzymatic removal of N- or O-bound sugars. C) Binding of Mac-2 BP/90K to endosialin was not dependent on divalent cations. Mac-2 BP/90K was released by sample buffer (sb-), but not in the presence of PBS containing 5 mM EDTA (EDTA). Released antigen was detected by Western blot analyses using the monoclonal anti-Mac-2 BP/90K antibody SP-2. D) HeLa cells were control transfected (C) or transfected with either a Myc-tagged N-terminal version (M1,2) or a C-terminal version (M3,4) of Mac-2 BP/90K. Expression of the recombinant protein was validated by an anti-Myc Western blot analysis (L.). Transfected cells were used for precipitation experiments using control (–) or endosialin-IgG beads (+). Anti-Myc Western blot analysis revealed that M2BP-3,4 (asterisk) but not M2BP-1,2 interacted with endosialin-IgG. E) Myc tagged-M2BP-3,4 transfected or control transfected HeLa cells were tested for the expression of M2BP-3,4 by anti-Myc Western blot analysis. Removal of N- and O-bound sugars (N+O) did not affect the binding to endosialin-IgG. M2BP-3,4 was still bound to the beads but was not detectable in the supernatant (eluate).

Mac-2 BP/90K has previously been shown to consist of two autonomous folding units, an N-terminal part (M2BP-1,2), mediating protein dimerization and a C-terminal part (M2BP-3,4) containing binding sites for galectin-3, nidogen, and collagen V and VI (14) . As shown in Fig. 3D , Myc-tagged M2BP-1,2 and M2BP-3,4 were expressed in HeLa cells with an apparent molecular mass of the proteins as described previously (37 kDa for M2BP-1,2 and 50 kDa for M2BP-3,4). Western blot analyses showed that M2BP-3,4 but not M2BP-1,2 was able to interact with endosialin-IgG (Fig. 3D , asterisk), indicating that the same domains that bind extracellular matrix components were also responsible for the binding to endosialin. Complete deglycosylation of M2BP-3,4 had no effect on its ability to interact with endosialin (Fig. 3E ).

Mac-2 BP/90K is weakly expressed by epithelial cells in normal tissues and strongly up-regulated in malignant tissues
For expression analysis of Mac-2 BP/90K, we performed immunohistochemistry in serial sections of human tissues. Weak expression of Mac-2 BP/90K was detectable in the stomach, small intestine, colon, kidney, and ovary (Fig. 4 ). The protein was not detectable in sections of human brain, uterus, and skin. The only normal tissue with strong Mac-2 BP/90K expression was the normal prostate. Notably, expression of Mac-2 BP/90K was in all tissues strictly restricted epithelial cells. Because endosialin is a marker for mesenchyme-derived activated tumor stroma with weak expression in normal tissues, we tested the expression of the endosialin ligand Mac-2 BP/90K in human malignant tissues. The monoclonal anti-Mac-2 BP/90K antibody SP-2 was used to analyze a panel of tumor tissues and observed a strong expression of Mac-2 BP/90K expression in almost all tested carcinomas, like carcinomas of the small intestine, renal cell carcinomas, and adenocarcinoma of the colon and the uterus. For all of these cases, Mac-2 BP/90K was exclusively expressed by the neoplastic cells and not by the surrounding stroma (Fig. 5 ). In contrast, endosialin expression was mutually exclusive in the mesenchymal tumor stroma of serial sections (Fig. 5) .


Figure 4
View larger version (126K):
[in this window]
[in a new window]

  		Figure 4. Weak expression of Mac-2 BP/90K in normal human tissues. A broad immunohistochemical analysis of Mac-2 BP/90K expression on a number of normal human tissues was performed. Tissues stained were small intestine, kidney, brain, colon, uterus, ovary, stomach, skin, and prostate. Mac-2 BP/90K was absent or weakly expressed in normal tissues with the exception of prostate and epithelial cells of the colon. Scale bars = 500 µm.


Figure 5
View larger version (178K):
[in this window]
[in a new window]

  		Figure 5. Up-regulated expression of Mac-2 BP/90K in tumors. Immunohistochemical analysis of Mac-2 BP/90K expression on a number of human tumor tissues was performed. Tissues stained were adenocarcinoma of the intestine, renal cell carcinoma (RCC), and adenocarcinoma of the uterus of the colon. Mac-2 BP/90K was found strongly up-regulated in the tumor tissue. Interestingly, Mac-2 BP/90K expression was restricted to the tumor parenchyma and not detectable in the surrounding stroma. Staining for endosialin on serial sections of adenocarcinoma of the uterus and the colon revealed mutually exclusive expression pattern of endosialin in the mesenchymal tumor stroma and Mac2 BP/90K in tumor cells. Scale bars = 500 µm (top panels), 100 µm (bottom panels).

Mac-BP/90K-expressing tumor cells display decreased binding to endosialin-expressing fibroblasts
To study the functional implications of the Mac-2 BP/90K–endosialin interaction, we used siRNA-mediated knock-down approaches to down-regulate Mac-2 BP/90K in human tumor cells and/or endosialin in primary human fibroblasts. As shown in Fig. 6 A, all siRNAs led to a strong and specific down-regulation of the corresponding protein. Adhesion of tumor cells to human primary fibroblasts was significantly increased after down-regulation of either Mac-2 BP/90K in tumor cells or endosialin in fibroblasts with an additive effect after simultaneous down-regulation of both proteins. These results were confirmed by a second independent set of siRNAs against endosialin and Mac-2 BP/90K. Thus, the observed gain-of-function phenotype following a loss-of-molecule experimental approach validates the specificity of the observed findings and argues strongly for a repulsive role of the Mac-2 BP/90K–endosialin interaction.


Figure 6
View larger version (13K):
[in this window]
[in a new window]

  		Figure 6. Repulsive cellular effects of the Mac-2 BP/90K–endosialin interaction. A) Western blot analysis confirms the specific down-regulation of Mac-2 BP/90K in tumor cells and of endosialin in human primary fibroblasts, respectively, after transfection with two independent siRNAs against endosialin and Mac-2 BP/90K. Control siRNA had no effect on the expression of either Mac-2 BP/90K or endosialin. B) Cell adhesion assays of either Mac-2 BP/90K-siRNAs or control-siRNA-transfected tumor cells onto endosialin-siRNA or control-siRNA transfected fibroblasts reveal an increased adhesion after down-regulation of endosialin in fibroblasts, Mac-2 BP/90K in tumor cells, or both proteins simultaneously. Results were confirmed by two independent siRNAs against endosialin (EN siRNA 1 and 2) and two independent siRNAs against Mac-2 BP/90K (M2 siRNA 1 and 2). *P < 0.05; **P < 0.005.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
We identified Mac-2 BP/90K as a secreted ligand for endosialin. Although endosialin possesses an N-terminal C-type lectin domain, a domain involved in carbohydrate binding, this interaction was found to be independent of carbohydrates being mediated by the C-terminal part of Mac-2 BP/90K (15) . In addition, we performed a detailed expression analysis of Mac-2 BP/90K and show that the protein is weakly expressed in epithelial layers of normal tissues and strongly up-regulated in the tumor cell compartment of most analyzed carcinomas. We observed a mutually exclusive expression pattern of endosialin and Mac-2 BP/90K in the tumor stroma and in the tumor cells, respectively, indicating that the interaction between endosialin and Mac-2 BP/90K may mediate repulsive positioning signals between epithelial and mesenchymal compartments. This concept was supported by siRNA-mediated loss-of-function experiments that revealed a decreased adhesion of Mac-2 BP/90K-expressing tumor cells to endosialin-expressing human primary fibroblasts.

Mac-2BP/90K was first identified as a ligand for galectin-3 (Mac-2) by protein purification using the human breast carcinoma cell line SK-BR-3 and human breast milk as a source. In addition, Western blot analysis identified the protein in human serum, semen, saliva, urine, and tear fluid (16) . It was shown to self-assemble into high-molecular-weight complexes supporting cell adhesion by binding to β1 integrins, collagens, and fibronectin (17) . The binding of Mac-2 BP/90K to extracellular matrix components was shown to be mediated by the C-terminal part of the protein independently of galectin-3 or carbohydrates (15 , 18) . Accordingly, we found Mac-2 BP/90K interacting with endosialin through its C-terminal domains in a carbohydrate-independent manner. Therefore, endosialin is a candidate for Mac-2 BP/90K-mediated cell adhesion mechanisms. Interestingly, endosialin displays a preferential binding to a 75-kDa cell surface-associated form Mac-2 BP/90K. Although the interaction partner of Mac-2 BP/90K on the surface of tumor cells is unknown, galectin-3 was described to bind Mac-2 BP/90K on the surface of the melanoma cell line A375 in a carbohydrate-dependent manner (18) .

The closest mouse homolog of Mac-2 BP/90K was identified as cyclophilin C-associated protein (CyCAP) as a ligand for cyclophilin C (19) . This protein was shown to be up-regulated in rat fibroblasts during wound healing and in penumbral neurons and microglia after focal cerebral ischemia (20 , 21) down-modulating endotoxin and proinflammatory responses in vivo (22) . Similar regulatory mechanisms for Mac-2 BP/90K have not yet been described. However, the endosialin protein structure shows a striking similarity with selectins that play critical roles in leukocyte recruitment. It may, therefore, be possible that the endosialin-Mac-2 BP/90K interaction mediates communication between the tumor cell compartment, the stromal fibroblasts, and inflammatory cells (23) . Cellular adhesion assays revealed a repulsive function of tumor cell expressed Mac-2 BP/90K on endosialin-expressing primary fibroblasts. Repulsive signaling cues have been described for several protein families, including semaphorins, ephrins, slits, and netrins. These play important roles in multiple biological processes, including axon guidance and the outgrowth of newly formed blood vessels (24 , 25) . Interestingly, repulsive signaling cues are also critical during tissue boundary formation (26) .

The expression of Mac-2 BP/90K has been shown to correlate with distant metastasis and poor survival of cancer patients (10) . This is in line with tumor studies performed in endosialin-deficient mice (8) , which showed reduced metastatic potential after implantation of abdominal tumors. Therefore, endosialin and Mac-2 BP/90K seem to be also functionally linked during tumor development and progression. As there is no difference in tumor development and progression of subcutaneously growing tumors in endosialin-deficient mice vs. wild-type mice, it is very likely due to a different expression pattern of Mac-2 BP/90K in subcutaneously vs. abdominal growing tumors. The exact functional mechanisms of Mac-2 BP/90K-endosialin enhanced metastasis are subject of ongoing studies. However, as Mac-2 BP/90K mediates adhesion of tumor cells, the interaction of endosialin and Mac-2 BP/90K possibly regulates tumor cell adhesion and/or migration of endosialin-expressing stroma cells on Mac-2 BP/90K in the extracellular matrix or tumor cells (27) . In that regard, our functional in vitro results suggest that endosialin expression leads to an increase in migration and invasion by binding to ligands located in the extracellular matrix. Indeed, Tomkowicz and colleagues recently described endosialin binding to fibronectin, as well as collagen I and IV, leading to increased migration of endosialin-expressing cells (28) . Binding of Mac-2 BP/90K acts likely as a negative regulator of this interaction to control reduced infiltration and therefore stimulate boundary formation between the mesenchymal and the epithelial compartment.

In conclusion, the present study has 1) identified Mac-2 BP/90K as extracellular ligand of endosialin, 2) determined the preferential tumor-associated expression of both molecules with mutually exclusive expression pattern in the tumor stroma and in the tumor cells, and 3) provided evidence for a repulsive role of the Mac-2 BP/90K–endosialin interaction in controlling tumor cell–stromal cell crosstalk. Together with the well-established functional roles of both molecules during tumor growth, these data strongly suggest important roles of the Mac-2 BP/90K–endosialin interaction during tumor progression and metastasis. Further functional analysis of this interaction promises not just to shed light into an important novel mechanism of tumor progression mechanisms, but to hold also strong potential to validate the Mac-2 BP/90K–endosialin axis as a novel antitumor target.


   ACKNOWLEDGMENTS
 
This work was supported by grants from the Deutsche Forschungsgemeinschaft SPP1190, "The tumor-vessel interface" (Au83/9-1), and the European Union (LSHG-CT-2004-503573). H.G.A. is supported by an endowed chair from the Aventis Foundation.

Received for publication December 12, 2007. Accepted for publication April 17, 2008.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ferrara, N., Kerbel, R. S. (2005) Angiogenesis as a therapeutic target. Nature 438,967-974[CrossRef][Medline]
  2. Teicher, B. A. (2007) Newer vascular targets: endosialin (review). Int. J. Oncol. 30,305-312[Medline]
  3. Rettig, W. J., Garin-Chesa, P., Healey, J. H., Su, S. L., Jaffe, E. A., Old, L. J. (1992) Identification of endosialin, a cell surface glycoprotein of vascular endothelial cells in human cancer. Proc. Natl. Acad. Sci. U. S. A. 89,10832-10836[Abstract/Free Full Text]
  4. Christian, S., Ahorn, H., Koehler, A., Eisenhaber, F., Rodi, H. P., Garin-Chesa, P., Park, J. E., Rettig, W. J., Lenter, M. C. (2001) Molecular cloning and characterization of endosialin, a C-type lectin-like cell surface receptor of tumor endothelium. J. Biol. Chem. 276,7408-7414[Abstract/Free Full Text]
  5. St Croix, B., Rago, C., Velculescu, V., Traverso, G., Romans, K. E., Montgomery, E., Lal, A., Riggins, G. J., Lengauer, C., Vogelstein, B., Kinzler, K. W. (2000) Genes expressed in human tumor endothelium. Science 289,1197-1202[Abstract/Free Full Text]
  6. MacFadyen, J. R., Haworth, O., Roberston, D., Hardie, D., Webster, M. T., Morris, H. R., Panico, M., Sutton-Smith, M., Dell, A., van der, G. P., Wienke, D., Buckley, C. D., Isacke, C. M. (2005) Endosialin (TEM1, CD248) is a marker of stromal fibroblasts and is not selectively expressed on tumour endothelium. FEBS Lett. 579,2569-2575[CrossRef][Medline]
  7. Christian, S., Winkler, R., Helfrich, I., Boos, A. M., Besemfelder, E., Schadendorf, D., Augustin, H. G. (2008) Endosialin (Tem1) is a marker of tumor-associated myofibroblasts and tumor vessel-associated mural cells. Am. J. Pathol. 172,486-494[Abstract/Free Full Text]
  8. Nanda, A., Karim, B., Peng, Z., Liu, G., Qiu, W., Gan, C., Vogelstein, B., St Croix, B., Kinzler, K. W., Huso, D. L. (2006) Tumor endothelial marker 1 (Tem1) functions in the growth and progression of abdominal tumors. Proc. Natl. Acad. Sci. U. S. A. 103,3351-3356[Abstract/Free Full Text]
  9. Iacobelli, S., Sismondi, P., Giai, M., D'Egidio, M., Tinari, N., Amatetti, C., Di Stefano, P., Natoli, C. (1994) Prognostic value of a novel circulating serum 90K antigen in breast cancer. Br. J. Cancer 69,172-176[Medline]
  10. Marchetti, A., Tinari, N., Buttitta, F., Chella, A., Angeletti, C. A., Sacco, R., Mucilli, F., Ullrich, A., Iacobelli, S. (2002) Expression of 90K (Mac-2 BP) correlates with distant metastasis and predicts survival in stage I non-small cell lung cancer patients. Cancer Res. 62,2535-2539[Abstract/Free Full Text]
  11. Fiedler, U., Krissl, T., Koidl, S., Weiss, C., Koblizek, T., Deutsch, U., Martiny-Baron, G., Marme, D., Augustin, H. G. (2003) Angiopoietin-1 and angiopoietin-2 share the same binding domains in the Tie-2 receptor involving the first Ig-like loop and the epidermal growth factor-like repeats. J. Biol. Chem. 278,1721-1727[Abstract/Free Full Text]
  12. Christian, S., Ahorn, H., Novatchkova, M., Garin-Chesa, P., Park, J. E., Weber, G., Eisenhaber, F., Rettig, W. J., Lenter, M. C. (2001) Molecular cloning and characterization of EndoGlyx-1, an EMILIN-like multisubunit glycoprotein of vascular endothelium. J. Biol. Chem. 276,48588-48595[Abstract/Free Full Text]
  13. Inohara, H., Raz, A. (1994) Identification of human melanoma cellular and secreted ligands for galectin-3. Biochem. Biophys. Res. Commun. 201,1366-1375[CrossRef][Medline]
  14. Hellstern, S., Sasaki, T., Fauser, C., Lustig, A., Timpl, R., Engel, J. (2002) Functional studies on recombinant domains of Mac-2-binding protein. J. Biol. Chem. 277,15690-15696[Abstract/Free Full Text]
  15. Weis, W. I., Drickamer, K. (1996) Structural basis of lectin-carbohydrate recognition. Annu. Rev. Biochem. 65,441-473[CrossRef][Medline]
  16. Koths, K., Taylor, E., Halenbeck, R., Casipit, C., Wang, A. (1993) Cloning and characterization of a human Mac-2-binding protein, a new member of the superfamily defined by the macrophage scavenger receptor cysteine-rich domain. J. Biol. Chem. 268,14245-14249[Abstract/Free Full Text]
  17. Sasaki, T., Brakebusch, C., Engel, J., Timpl, R. (1998) Mac-2 binding protein is a cell-adhesive protein of the extracellular matrix which self-assembles into ring-like structures and binds beta1 integrins, collagens and fibronectin. EMBO J. 17,1606-1613[CrossRef][Medline]
  18. Inohara, H., Akahani, S., Koths, K., Raz, A. (1996) Interactions between galectin-3 and Mac-2-binding protein mediate cell-cell adhesion. Cancer Res. 56,4530-4534[Abstract/Free Full Text]
  19. Friedman, J., Trahey, M., Weissman, I. (1993) Cloning and characterization of cyclophilin C-associated protein: a candidate natural cellular ligand for cyclophilin C. Proc. Natl. Acad. Sci. U. S. A. 90,6815-6819[Abstract/Free Full Text]
  20. Kong, W., Li, S., Longaker, M. T., Lorenz, H. P. (2007) Cyclophilin C-associated protein is up-regulated during wound healing. J. Cell. Physiol. 210,153-160[CrossRef][Medline]
  21. Shimizu, T., Imai, H., Seki, K., Tomizawa, S., Nakamura, M., Honda, F., Kawahara, N., Saito, N. (2005) Cyclophilin C-associated protein and cyclophilin C mRNA are upregulated in penumbral neurons and microglia after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 25,325-337[CrossRef][Medline]
  22. Trahey, M., Weissman, I. L. (1999) Cyclophilin C-associated protein: a normal secreted glycoprotein that down-modulates endotoxin and proinflammatory responses in vivo. Proc. Natl. Acad. Sci. U. S. A. 96,3006-3011[Abstract/Free Full Text]
  23. Bevilacqua, M. P., Nelson, R. M. (1993) Selectins. J. Clin. Invest. 91,379-387[Medline]
  24. Halloran, M. C., Wolman, M. A. (2006) Repulsion or adhesion: receptors make the call. Curr. Opin. Cell Biol. 18,533-540[CrossRef][Medline]
  25. Pfaff, D., Fiedler, U., Augustin, H. G. (2006) Emerging roles of the angiopoietin-tie and the ephrin-Eph systems as regulators of cell trafficking. J. Leukoc. Biol. 80,719-726[Abstract/Free Full Text]
  26. Pasquale, E. B. (2005) Eph receptor signalling casts a wide net on cell behaviour. Nat. Rev. Mol. Cell. Biol. 6,462-475[CrossRef][Medline]
  27. Ulmer, T. A., Keeler, V., Loh, L., Chibbar, R., Torlakovic, E., Andre, S., Gabius, H. J., Laferte, S. (2006) Tumor-associated antigen 90K/Mac-2-binding protein: possible role in colon cancer. J. Cell. Biochem. 98,1351-1366[CrossRef][Medline]
  28. Tomkowicz, B., Rybinski, K., Foley, B., Ebel, W., Kline, B., Routhier, E., Sass, P., Nicolaides, N. C., Grasso, L., Zhou, Y. (2007) Interaction of endosialin/TEM1 with extracellular matrix proteins mediates cell adhesion and migration. Proc. Natl. Acad. Sci. U. S. A. 104,17965-17970[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Pathol.Home page
F. Sotgia, F. Del Galdo, M. C. Casimiro, G. Bonuccelli, I. Mercier, D. Whitaker-Menezes, K. M. Daumer, J. Zhou, C. Wang, S. Katiyar, et al.
Caveolin-1-/- Null Mammary Stromal Fibroblasts Share Characteristics with Human Breast Cancer-Associated Fibroblasts
Am. J. Pathol., March 1, 2009; 174(3): 746 - 761.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
fj.07-101386v1
22/8/3059    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Becker, R.
Right arrow Articles by Christian, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Becker, R.
Right arrow Articles by Christian, S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS