Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Invading Basement Membrane Matrix Is Sufficient for MDA-MB-231 Breast Cancer Cells to Develop a Stable In Vivo Metastatic Phenotype

  • Mohamed Abdelkarim ,

    Contributed equally to this work with: Mohamed Abdelkarim, Nadejda Vintonenko

    Affiliations Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France, INSERM 553 Endothélium et Angiogénèse Laboratoire d'Hémostase, Paris, France

  • Nadejda Vintonenko ,

    Contributed equally to this work with: Mohamed Abdelkarim, Nadejda Vintonenko

    Affiliations Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France, INSERM 553 Endothélium et Angiogénèse Laboratoire d'Hémostase, Paris, France

  • Anna Starzec,

    Affiliation Université Paris 13, EA4222, Li2P, Bobigny, France

  • Aniela Robles,

    Affiliations Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France, Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • Julie Aubert,

    Affiliation AgroParisTech, UMR 518, Mathématiques et Informatique Appliquées, Paris, France

  • Marie-Laure Martin,

    Affiliations AgroParisTech, UMR 518, Mathématiques et Informatique Appliquées, Paris, France, URGV UMR INRA 1165-CNRS 8114-UEVE, Evry, France

  • Samia Mourah,

    Affiliation Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • Marie-Pierre Podgorniak,

    Affiliation Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • Sylvie Rodrigues-Ferreira,

    Affiliations Inserm, U1016, Institut Cochin, Paris, France, CNRS UMR 8104, Paris, France, Université Paris Descartes, Paris, France

  • Clara Nahmias,

    Affiliations Inserm, U1016, Institut Cochin, Paris, France, CNRS UMR 8104, Paris, France, Université Paris Descartes, Paris, France

  • Pierre-Olivier Couraud,

    Affiliations Inserm, U1016, Institut Cochin, Paris, France, CNRS UMR 8104, Paris, France, Université Paris Descartes, Paris, France

  • Christelle Doliger,

    Affiliation Service Commun d'Imagerie, Institut d'Hématologie, Hôpital Saint Louis, Paris, France

  • Odile Sainte-Catherine,

    Affiliation Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France

  • Nicole Peyri,

    Affiliations Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France, Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • Lei Chen,

    Affiliation INSERM 553 Endothélium et Angiogénèse Laboratoire d'Hémostase, Paris, France

  • Jérémie Mariau,

    Affiliation INSERM 553 Endothélium et Angiogénèse Laboratoire d'Hémostase, Paris, France

  • Monique Etienne,

    Affiliation Université Paris 13, Laboratoire d'Histologie, Bobigny, France

  • Gerard-Yves Perret,

    Affiliation Université Paris 13, EA4222, Li2P, Bobigny, France

  • Michel Crepin,

    Affiliation INSERM 553 Endothélium et Angiogénèse Laboratoire d'Hémostase, Paris, France

  • Jean-Luc Poyet,

    Affiliation Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • Abdel-Majid Khatib,

    Affiliations Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France, INSERM U1029 Université Bordeaux 1, Bordeaux, France

  •  [ ... ],
  • Mélanie Di Benedetto

    dibenedetto22@yahoo.fr

    Affiliations Université Paris 13, CNRS FRE CSPBAT, Laboratoire de Chimie Structurale Biomoléculaire, UFR SMBH, Bobigny, France, Université Paris 7, UMRS 940. Equipe Avenir, IGM, Paris, France

  • [ view all ]
  • [ view less ]

Abstract

Introduction

The poor efficacy of various anti-cancer treatments against metastatic cells has focused attention on the role of tumor microenvironment in cancer progression. To understand the contribution of the extracellular matrix (ECM) environment to this phenomenon, we isolated ECM surrogate invading cell populations from MDA-MB-231 breast cancer cells and studied their genotype and malignant phenotype.

Methods

We isolated invasive subpopulations (INV) from non invasive populations (REF) using a 2D-Matrigel assay, a surrogate of basal membrane passage. INV and REF populations were investigated by microarray assay and for their capacities to adhere, invade and transmigrate in vitro, and to form metastases in nude mice.

Results

REF and INV subpopulations were stable in culture and present different transcriptome profiles. INV cells were characterized by reduced expression of cell adhesion and cell-cell junction genes (44% of down regulated genes) and by a gain in expression of anti-apoptotic and pro-angiogenic gene sets. In line with this observation, in vitro INV cells showed reduced adhesion and increased motility through endothelial monolayers and fibronectin. When injected into the circulation, INV cells induced metastases formation, and reduced injected mice survival by up to 80% as compared to REF cells. In nude mice, INV xenografts grew rapidly inducing vessel formation and displaying resistance to apoptosis.

Conclusion

Our findings reveal that the in vitro ECM microenvironment per se was sufficient to select for tumor cells with a stable metastatic phenotype in vivo characterized by loss of adhesion molecules expression and induction of pro-angiogenic and survival factors.

Introduction

Metastasis in breast cancer patients accounts for over 90% of the deaths. Preclinical studies reveal that many drugs used in the management of primary tumors are not or less effective against metastasis [1]. Although the mechanism by which metastases develop is still not fully understood, it is generally believed that tumor cells acquire features that affect their metastatic potential during the progression of the tumor; these features include increased survival, invasive and migratory abilities. Metastasis is a complex cascade of sequential steps, none of which being fully understood. Many studies implicated the stroma in the development of metastases. Stroma and cancer cell interactions were found to contribute to cell detachment from primary tumors, intravasation into the blood stream, and extravasation at distant sites where tumor cells can seed and form tumor metastases [2]. Previously, fibroblasts, endothelial cells and macrophages and other stroma cells were reported to be implicated in the occurrence of metastases [2][4]. Whereas less is known about the influence of the extracellular matrix (ECM) in the development of metastases [5], [2], ECM appears to be involved, particularly through sequestration of soluble factors secreted by stroma cells and acting on tumor cells in a paracrine manner [6][9]. The ECM can modify the adhesion of cancer cells leading to their invasive behavior. However, it is not clear if the ECM, and particularly the basement membrane, can influence the selection of metastatic cells with a well-defined genotype profile. The basement membrane underlies the endothelium in the vessel wall and is the major barrier to tumor cell extravasation and invasion [10][11]. Matrigel is a basement membrane extract derived from a murine tumor [12]. The components of this tumor basement membrane are similar, both chemically and immunogically, to authentic basement membrane components [13]. Consequently, Matrigel may be used as an experimental model of barrier to identify both inhibitors and activators of invasion [14].

Most of previously reported studies on gene profiling in malignant cells used tumor cells isolated from metastatic foci that were already well established in vivo. Therefore, the contribution of ECM to the selection of metastatic cells has not been rigorously assessed. Here, we report the in vitro selection of two cell populations derived from MDA-MB-231 breast cancer cells [15], on the basis of their high (INV) invasive ability (to migrate through Matrigel). We examined the transcriptomes of these two populations using micro arrays and correlated their gene profiles to their malignant phenotypes.

Materials and Methods

Cell culture

Human breast adenocarcinoma MDA-MB-231 cells were obtained from the American Type Culture Collection (Manassas, VA, USA). The INV and REF cells isolated as described underneath were maintained in 10% FBS-DMEM, 1% sodium pyruvate with 1% penicillin and 1% streptomycin at 37°C in a humidified atmosphere containing 5% carbon dioxide. Human microcapillary brain endothelial hCMEC/D3 cells [16], [17] were grown in collagen type I-coated dishes in EBM-2 media supplemented with 2.5% FBS, hydrocortisone and growth factors (VEGF, IGF, EGF and bFGF EGM-2 Bullet kit, Cambrex-Biowhittaker).

Isolation of MDA-MB-231 derived INV and REF variants

MDA-MB-231 cells (2.5×105) were seeded in 8 µm-pore size Boyden chambers (Becton Dickinson) coated with Matrigel (Falcon, MA, USA) and incubated for 16 h at 37°C in a 5% CO2 atmosphere. Invading cells, referred to as INV cells, were collected from the bottom of the membrane in 0.5 mM EDTA and cultured in a 24-well plate. The selection was repeated eight times sequentially in order to obtain highly invasive cells. To select non-invasive cells (REF), the same procedure was used and the cells that remained on top of the membrane after the first round of selection were collected.

Cell growth, migration and invasion assay

For cell growth assays, INV and REF cells were seeded at 2×105 cells per well in 6-well plates and cultured for 72 h. Cell migration experiments were performed using the Boyden chambers as described above: the cells (2.5×105) were put into the upper chamber and 10% FCS-DMEM was used as a chemoattractant. For invasion assays, chambers were coated with Matrigel as described above. Then, cells on the lower face of the filter were fixed, stained with crystal violet and counted under the optical microscope.

Adhesion and transmigration assays

Tumor cells were harvested in 0.5 mM EDTA and seeded at 5×105 cells per well in 96-well plates pre-coated with 1 µg/ml of fibronectin (Sigma, France) and incubated at 37°C for 1 h. The plates were washed with PBS, and adherent cells were stained with crystal violet/methanol (0.5%/20%) for 1 h at room temperature. The plates were again washed, and were incubated in ethanol: 0.1 M citrate (1∶1) pH 4.2 for 30 min at room temperature and absorbance was measured at 540 nm using an ELISA reader (Labsystem, GMI Inc, Minnesota, USA). To study tumor cell adhesion to endothelial cell monolayers, hCMEC/D3 cells were seeded at 104 cells per well on collagen type I-coated 96-wells plates and left to grow until confluence. The cells were then starved for 48 h and incubated for an additional 24 h with the proinflammatory cytokines IFNγ (100 U/ml-Preprotech Inc, NJ, USA.) and TNFá (200 U/ml-Preprotech Inc.). Next, the INV and REF cells at 80% confluence were harvested using PBS-EDTA and labeled for 30 min at 37°C with CMFDA, a green fluorescent cell tracker (Invitrogen, Cergy Pontoise, France). Then, 105 INV or REV cells were added to the endothelial cell monolayer and incubated for 30 min at 37°C. The unattached cells were removed and adherent cells were lysed with water. Fluorescence was measured at 485/530 nm using a microplate reader (Labsystem, France). For transmigration assays, Boyden chambers were coated with hCMEC/D3 endothelial cells and CMFDA-labeled tumor cells were counted by fluorescent microscopy.

Gelatine zymography

INV and REF cells were cultured in serum-free media for 24 h in 6-well plates. Conditioned media were then lyophilized and normalized to the number of cells. They were mixed with non-reducing Laemmli buffer and run on a 10% SDS-polyacrylamide gel containing 1 mg/ml gelatin. The gel was washed three times at room temperature in a solution containing 2.5% (v/v) Triton X-100 and incubated at 37°C for 24 h in 50 mM Tris/HCL, pH 7.5, 0.2 M NaCl, 5 mM CaCl2 and 0.05% Brij 35. The gel was then stained for 60 min with 0.25% (w/v) R-250 Coomassie blue in 45% (v/v) methanol/10% (v/v) acetic acid and destained in 25% methanol (v/v)/10% acetic acid (v/v).

Cell death detection

Cells (2×105) were cultured for 96 h in a serum-free medium or with 2.5 µM Doxorubicin for 24 h in a serum-containing medium. Then, cells were harvested and apoptotic cells were labeled using the Annexin V-FITC Apoptosis Detection kit (Beckman coulter, Fullerton CA, USA). The samples were then analyzed by flow cytometry (Becton Dickinson, Heidelberg, Germany). For detection of cell death in vivo, tumor sections (5 mm) were deparaffinized and rehydrated, then analyzed for DNA fragmentation using a TumorTACS kit (R&D Systems, Abington, UK). Various fields (10 per section) were selected for analysis and the intratumor apo-necrotic cells were counted as previously described [18]. For AKT phosphorylation assay, cells were treated for 1 h or 72 h with serum free medium and stimulation with IGF-1 at 10 ng/ml for 30 min was used as positive control.

Western blots

INV and REF cell lysates were clarified by centrifugation, and protein concentrations were measured by bicinchoninic acid protein assay (Bio-Rad, Marne-La-Coquette, France). Equal amounts of total proteins were resolved by 6% SDS-PAGE and transferred to nitrocellulose membrane. The membrane was blocked in 5% nonfat dried milk in Tris-buffered saline containing 1% Tween 20 and incubated with primary antibodies against NRP-1 (Santa-Cruz Biotechnologies inc., CA) or ß-actin (Sigma, Saint quentin Fallavier, France), AKT or pAKT (Cell signaling technology, Boston, USA) for 1 h, and then with an appropriate conjugated secondary antibody for 30 min at room temperature. Immunoreactive bands were detected by chemiluminescence according to the manufacturer's instructions.

VEGF quantification

INV or REF cells seeded in 24-well culture plates (2×105/well) in DMEM–10% FCS were allowed to adhere for 24 h. Conditioned media containing growth factors secreted by the cells were collected after 48 h of serum starvation. The VEGF content in these media was measured by quantitative enzyme-linked immunoabsorbent sandwich assay (ELISA) using the Quantikine kit (R&D Systems, Abington, UK) according to the manufacturer's instructions. VEGF contents were estimated from a standard curve established from serial dilutions of VEGF protein standards (R&D Systems). The values reported for VEGF secretion are normalized to cell number.

Real-time RT-PCR analysis

Total RNA (1 µg) was reverse-transcribed using MMLV RT (Invitogen, Carlsbad, CA, USA) and random primers from Roche Applied Science (Roche, Indianapolis, IN, USA). Primers for qRT-PCR were purchased from AnyGenes (Creteil, France). Real-time PCR was carried out with the LightCycler® 2.0 Real-Time PCR System (Roche, Indianapolis, IN, USA) using the LightCycler® FastStart DNA MasterPLUS SYBR Green I Kit. The reactions were cycled 45 times [95°C, 10 min, (94°C, 15 sec; 60°C, 1 min; and 72°C, 1 min)]. Reported results are normalized to the expression of the PPIA, GAPDH, TBP, beta 2 Microglobin, Actin, RPL0 and RPL19 genes that exhibit little variation in this data set. Similar results were obtained for all reference genes, the representative data is normalized to PPIA. Fold induction relative to internal controls was calculated by the delta Ct evaluation method [19].

Gene Expression Profiling

Total RNA was isolated from eight cell cultures (four biological replicates corresponding to four different passages of both INV and REF cells) using the RNeasy mini kit (Qiagen, Valencia, CA, Germany) by direct cell lysis in 10 cm culture dishes containing 600 µl RLT Buffer following the manufacturer's instructions. Purified RNA was quantified with a Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Hudson, NH, USA) and its quality was evaluated by running the Eukaryotic Total RNA Nano Assay in the Bioanalyser 2100 (Agilent Technologies, Chandler AZ, USA) using the Agilent RNA 6000 Nano Kit. Probes (300 ng of the total RNAs) were prepared for hybridization to Affymetrix GeneChip® Human Gene 1.0 ST arrays according to the manufacturer's instructions. After scanning, eight CEL files generated by GCOS (Affymetrix GeneChip® Operating Software) were uploaded into Expression Console Software (Affymetrix, Santa Clara, CA, USA) for normalization and probe set summarization using the RMA (Robust Multichip Analysis) algorithm [20]. The resulting data in terms of probe set intensities were exported into a .txt file for further statistical analysis. The microarray data has been deposited in the Gene Expression Omnibus (GEO) database at http://www.ncbi.nlm.nih.gov/geo/ with the accession number GSE16838.

Microarray data analysis

Analyses were carried out in R, a free software environment available at http://www.r-project.org/. To identify differentially expressed genes, differences between the two conditions were calculated on each day of measurement and then analyzed using a double-sided paired t-test implemented in the anapuce R package (http://www.agroparistech.fr/mia/outil.html). Pairing was carried out to remove any source of variation due to the day of measurement from the comparison of the two conditions. The variance was split between subgroups of genes with homogeneous variance [21]. Statistical raw p-values were adjusted for multiple comparisons using the Benjamini-Hochberg procedure [22] which controls the False Discovery Rate (FDR) after the estimation of the proportion of genes non-differentially expressed using the smoother method [23]. The level of statistical significance was set at 0.05 for all comparisons. Gene annotations and functions were assessed using NetAffx™ Analysis Center (Affymetrix, Santa Clara, CA, USA), Ingenuity pathway Analysis (IPA) software (Ingenuity SystemsInC.) and data mining.

Xenografts in nude mice

All in vivo experiments were carried out with the approval of Département d'Expérimentation Animale, Institut Universitaire d'Hématologie, Hôpital Saint-Louis ethical committee (permit number 75516) and according to the UKCCCR guidelines. The animals were kept in a temperature-controlled room on a 12∶ 12 light-dark schedule with food and water ad libidum. Aliquots of 2×106 of INV (n = 7) or REF cells (n = 7) were inoculated subcutaneously (s.c) near the mammary gland of 4-week-old athymic nude mice (nu/nu, Janvier, France, n = 14). Tumor volume was measured once a week along two major axes using calipers. Tumor volume (mm3) was calculated as follows: V = 4/3πR12R2 where R1<R2

Immunohistochemical analysis

Tumor specimens were fixed with 4% paraformaldehyde and embedded in paraffin using standard procedures. For immunohistochemical studies, 5 µm-thick sections were cut, deparaffinized and rehydrated. Routinely, sections were stained in haematoxylin and eosin. Endogenous peroxidase was inactivated with 3% H2O2 and washed in TBS (0.05 M Tris, 1.5 M NaCl, pH 7.6) followed by preincubation in 10% normal goat serum for 1 h at room temperature. Endothelial cells were specifically labeled with GSL-1 isolectin B4 (Vector Laboratories, Burlingame, CA, U.S.A.) as described previously [24]. For each tumor, 10 randomly selected sections were analyzed.

Intracardiac experimental metastasis assay

We performed mice intracardiac injection of bioluminescent INV and REF cells as also described previously in order to observe large dissemination behavior of each subpopulation [25]. For this study, luciferase-expressing INV and REF cell pools were obtained by stable transfection with a plasmid carrying a G418 resistance gene and the luciferase gene (pEGFPLuc, BD Biosciences, Rungis, France) using Fugen-6 (Roche Diagnostics, Meylan, France). Transfected cells were cultured in 10% FCS-DMEM supplemented with G418 at 1 mg/ml. For experimental metastasis, female nude mice (8–10 weeks) were anesthetized by intraperitoneal injection with 120 mg/kg ketamine and 6 mg/kg xylazine on the day of cell injection or by exposure to 1–3% isoflurane on subsequent imaging days. pEGFPLuc-transfected INV or REF cells (105) in 100 µl of sterile DPBS were injected into the left ventricle of the heart by nonsurgical means. Anesthetized mice were given intraperitoneal injections of D-luciferin (Caliper Life Science) and placed in the IVIS™ Imaging System (Xenogen, Caliper Life Science, MA, USA) and imaged from both dorsal and ventral views. Successful intracardiac injection was revealed on day 0 by systemic bioluminescence distributed throughout the whole animal. Only mice with satisfactory injection were used for subsequent experiments. Metastasis analysis was monitored by in vivo imaging once a week for up to eight weeks.

Bioluminescent Imaging

Bioluminescence images were acquired with the IVIS imaging system (Xenogen) 5 minutes after intraperitoneal injection of D-luciferin into the anesthetized animals. Acquisition time at the beginning of experiment was 5 min and was reduced in keeping with signal strength to avoid saturation. Analysis was performed using LivinImage software (Xenogen) by measurement of photon flux (photon/s/cm2) within a region of interest (ROI) drawn around the bioluminescence signal to detect the metastatic sites.

Statistical analysis

Student's t-test was used for statistical analysis. P<0.05 was considered significant.

Results

Selection of INV subpopulation from MDA-MB-231

Using Matrigel coated Boyden chambers, we established two different sub-populations of MDA-MB-231 cells. These included the invasive MDA-MB-231 subpopulation (INV) that was isolated from a less invasive subpopulation (REF) following eight rounds of invasion assays in the Boyden Chamber. Their invasive phenotype was confirmed by migration and invasion assays (Fig. 1). REF cells were expanded and frozen after the first round. As expected, INV cells were up to 4-fold more invasive than REF cells (Fig. 1A and right panel), and up to 3-fold more motile (Fig. 1B and right panel). These properties were stable and not lost after maintenance in culture or freezing. To assess the invasive capacity of INV and REF cells in more detail, the corresponding conditioned media were collected and analyzed for gelatinase enzymatic activity (Fig. 1C). MMP-9 activity was higher in media conditioned by INV cells as compared to REF cells. In contrast, MMP-2 activity remained unchanged (data not shown). Similarly, the mRNA expression of uPA, uPAR and PAI-1 were found significantly up-regulated in INV cells as evaluated by real time PCR (Fig. 1D).

thumbnail
Figure 1. INV and REF breast cancer cell invasion and migration.

Cells (2.5×105) were added to the upper side of each 8 µm insert of Boyden chambers coated (A) or not (B) with Matrigel for invasion or migration assays respectively; cells were then counted as described in Materials and Methods. MMP-9 activity secreted by REF (lane 1) and INV (lane 2) cells (C) was assessed by subjecting aliquots of lyophilized conditioned media normalized to the number of cells to 10% SDS-polyacrylamide gels containing 1 mg/mL gelatin. For QRT-PCR analysis (D), total RNA (1 µg) was reverse-transcribed using MMLV RT and subjected to qRT-PCR as described in Material and Methods. Representative data was normalised to PPIA is given. Each column represents a mean (± SD) of three independent experiments. *P<0.05.

https://doi.org/10.1371/journal.pone.0023334.g001

Transcriptomic analysis of INV and REF cells

We used the Affymetrix Human Gene 1.0ST arrays that contain 28000 annotated genes (Table 1 and 2), to investigate the expression profile of REF and INV cells. Comparative analysis of REF and INV cells revealed 134 differentially expressed genes (Table S1): 38 were down-regulated and 96 were up-regulated in INV cells with respect to REF cells. Among the known down-regulated genes in INV cells (Table 1), 44% are involved in cell adhesion and cell-cell junctions, and 22% in induction of apoptosis and G-protein coupled receptor signaling (Fig. S1). Of the up-regulated genes in INV (Table 2) with known function, 25% are involved in survival via negative regulation of apoptosis, 17.5% in cytoskeleton reorganization, cell proliferation and chemotaxis, 12.5% in cell adhesion and cell-cell junctions and 10% in G-protein coupled receptor signaling (Fig. S1).

thumbnail
Table 1. List of genes that are down-regulated in INV cells versus REF cells.

https://doi.org/10.1371/journal.pone.0023334.t001

thumbnail
Table 2. List of genes that are up-regulated in INV cells versus REF cells.

https://doi.org/10.1371/journal.pone.0023334.t002

We validated the array results by real-time PCR for selected genes (indicated with asterisks in tables 1 and 2).

Adhesion and transmigration abilities of INV and REF cells

The main differences between the transcriptome profiles of the two cell populations concerned molecules involved in cell-ECM and cell-cell junctions. Therefore, we tested INV and REF cells for adhesion and transmigration through endothelial cells. As observed in Fig. 2A, the morphology of INV cells in monolayer cultures differed from that of REF cells. INV but not REF monolayers cells seemed to present gaps between cells. In addition, REF cells were round whereas INV cells were fusiform.

thumbnail
Figure 2. REF and INV cell adhesion and transmigration.

Morphology of INV and REF cells (A) revealed by optical microscopy at ×400 magnification. Arrows indicate cell junctions and asterisks indicate cell shape. Adhesion to fibronectin (B) and hCMEC/D3 endothelial cell adhesion assay (C). Transmigration assays (D) involved growing hCMEC/D3 endothelial cells to confluence on the upper side of 8 µm insert Boyden chambers. CMFDA-stained cells that transmigrated through the endothelial cells were counted as described in “Materials and Methods”. For adhesion to fibronectin, the data were normalized to BSA adhesion. Each column shows means (± SD) of three independent experiments. *P<0.05.

https://doi.org/10.1371/journal.pone.0023334.g002

We next tested adhesion of INV and REF cells to fibronectin and hCMEC/D3 brain microcapillary endothelial cells (Fig. 2 B and C, respectively). Consistent with findings for the transcriptomes, REF cells adhered to fibronectin whereas INV cells did not (the optical density was the same as for negative controls, equivalent to fold induction of 1) (Fig. 2B). Although both cell populations adhered to the endothelium formed by hCMEC/D3 human brain microcapillary endothelial cells, INV cells again showed a lower adhesion capacity than REF cells both in the presence or absence of cytokines (Fig. 2C) (proinflammatory cytokines have been reported to activate endothelial cells and induce tumor cell attachment to the endothelium [26]).

Next, we compared the ability of INV and REF cells to transmigrate through hCMEC/D3 endothelial cells (Fig. 2D): INV cell transmigration was 1.5 fold higher than that of REF cells (p<0.05).

Increased metastatic sites colonization and reduced survival in INV cell-inoculated mice

In order to study the subclones metastasis behaviour in the whole body, Luc-transfected subclones of INV and REF (1×105) were injected in the circulation (intracardiac route as described previously, [25]) of nude mice (n = 9) and their abilities to colonize secondary sites were evaluated weekly during eight weeks (Fig. 3A). Metastases were detected after two and three weeks in mice injected with INV and REF cells, respectively. Five weeks after injection, multiple metastatic sites were detected in mice inoculated with INV cells (Fig. 3A and C). In general, in each mouse inoculated with REF cells, one metastatic site was detected, mainly in bone rather if only one mouse developed lung metastasis which lead to rapid death at day 37 (data not shown. In addition, ex vivo imaging of various tissues after the final in vivo imaging mainly revealed the presence of lesions in bones for REF cells injections (legs and skull) and brain, lymph nodes and ovary for INV ones (Fig. 3B). Ovarian lesions were only found in INV cell-injected mice (4 of 9 mice) and most smooth tissue lesions were also in mice injected with INV cells. The number of metastatic sites was 2 to 3 fold higher in the INV group than in the REF group of mice (Fig. 3C). INV cell-injected mice died earlier than REF cell-injected mice; only 20% of INV as compared to 80% of the REF group survived to the end of the experiment (Fig. 3D).

thumbnail
Figure 3. REF and INV metastasis formation.

INV and REF cells (105) were injected into the left ventricle of nude mice (n = 7, A). Ex vivo data confirm tissue metastasis from INV (right panel) and REF cells injection (left panel) (B). Quantification of metastatic sites (C). Mouse survival following injection of REF and INV cells (D). Each column shows means (±SD) of representative experiments. *P versus control<0.05.

https://doi.org/10.1371/journal.pone.0023334.g003

Inreased growth of INV xenografts associated with enhanced cell survival

We next explored the tumorigenicity of INV and REF cells in nude mice. We injected REF or INV cells (2×106) subcutaneously near the mammary gland of nude mice (Fig. 4). Both cell populations caused the development of 30 mm3 tumors after one week. After 21 days, the volumes of INV tumors were 2.6 fold higher than those of REF tumors (172 mm3 vs 67 mm3, respectively, Fig. 4A). Surprisingly, in vitro proliferation of INV cells was half that of REF cells after 72 h of cell culture in FCS10%/DMEM (data not shown). Thus, the faster growth of INV tumors was unlikely to have been due to a higher proliferation rate. We therefore evaluated cell survival within REF and INV tumors by in situ DNA fragmentation labeling with biotinylated nucleotides. This technique results in brown staining of apoptotic cell nuclei and diffuse brown staining in the cytoplasm and nuclei of necrotic cells (Fig. 4B). The numbers of dead cells were lower in INV tumors than REF tumors (15% versus 22%, P = 0.046; Fig. 4C). In vitro after 24 h of treatment with 2.5 µM doxorubicin (Fig. 5), 90% of REF and 59% of INV cells had undergone apoptosis as revealed by Annexin-V labeling (Fig. 5A). Similarly, following 96 h of serum deprivation, 40% of REF cells and only 9.5% of INV cells were found to be apoptotic (Fig. 5B). These various in vitro and in vivo analyses demonstrate that INV cells are more apoptosis-resistant than REF cells. This is consistent with the qRT-PCR and microarray evidence of up-regulation of several anti-apoptotic molecules in INV cells including TIMP-1 ([27]), TGF β2 ([28]), and IRAK1 ([29]. Also, some pro-apoptotic molecules and some factors involved in drug-induced apoptosis sensitization, such as cyclin D2 and caspase 4 [30], were found to be down-regulated in INV cells. Analysis of AKT activation showed no activation when INV as well as REF cells where in serum free medium (Fig. 5D, lane 2, 3 and lane 5, 6) as compared to controls stimulated with IGF-1 (lane 1 and 4).

thumbnail
Figure 4. REF and INV tumor growth.

INV and REF cells were injected subcutaneously near mammary gland into nude mice as described in “Materials and Methods”. After 1 week, tumors appeared in each group (A). Each point represents the mean tumor volume (mm3) (± SD, n = 7). Dead cells in REF and INV tumor sections were stained (B). Arrows indicate apoptotic and necrotic cells (magnification ×200). Percentage of cell death was determined as described in “Materials and Methods” (C).

https://doi.org/10.1371/journal.pone.0023334.g004

thumbnail
Figure 5. REF and INV cell death.

Cells (2×105) were cultured for 96 h in serum-free medium or were incubated with 2.5 µM Doxorubicin for 72 h in a serum-containing medium (A) or were cultured for 96 h in serum-free medium (B). For qRT-PCR analysis (C), total RNA (1 µg) was reverse-transcribed and the results were normalized as described above. (D) REF cells (Lanes 1 to 3) and INV cells (lanes 4 to 6) were treated (lane 1 and 4) with Serum free DMEM for 1 h or for 3 days (lanes 2 and 5) or with 10 ng/ml of IGF-1 (lanes 3 and 6), to detected AKT and pAKT by western blotting. Each column shows means (± SD) of three independent experiments. *P<0.05.

https://doi.org/10.1371/journal.pone.0023334.g005

Increased angiogenesis in INV xenografts

We next assessed angiogenesis in both INV and REF tumors using GSL-1 lectin to labelmouse endothelial cells (Fig. 6). Both tumor types exhibited vascular areas (Fig. 6A). The number of endothelial cells per area was significantly higher in INV than REF tumors (132 versus 94, respectively, p<0.05, Fig. 6B). Analysis of genes involved in angiogenesis including those encoding VEGF and their receptors [31][35] by qRT-PCR revealed stronger VEGF, VEGFR2 and NRP-1 expression in INV compared to REF cells (Fig. 6C). The over-production of VEGF and NRP-1 proteins was confirmed by ELISA tests for VEGF (Fig. 6D) and western blot analysis for NRP-1 (Fig. 6E).

thumbnail
Figure 6. REF and INV tumor angiogenesis.

Endothelial cells in REF and INV tumor sections were stained as described in “Materials and Methods”. Microvessels are revealed by brown staining (magnification ×100) and endothelial cells were quantified (B). QRT-PCR was used to quantify VEGF, VEGFR2, VEGFR1 and NRP-1 gene expression in INV and REF cells (C). VEGF (D) and NRP-1 (E) proteins were detected by ELISA tests and by western blotting, respectively. Each column represents a mean (± SD) of three independent experiments. *P<0.05. Gene functions from the GO Description provided by NetAffx™ Analysis Center Affymetrix (https://www.affymetrix.com/analysis/netaffx/index.affx)* Validated by QRT-PCR.

https://doi.org/10.1371/journal.pone.0023334.g006

Discussion

We report that breast cancer cell invasion through Matrigel resulted in the generation of an aggressive subpopulation (INV) with a stable malignant genotypic/phenotypic profile. The main characteristic of these cells was loss of adhesion to ECM, and in particular fibronectin and endothelial cells. Accordingly, in nude mice, INV cells were more competent than the control cells (REF) for extravasation through the endothelium to colonise tissues and form metastases. Metastases formed by INV cells were mainly found in various soft tissue (brain, ovary, lymph node and bone) as compared to REF ones (Exclusively in bones). This result was in accordance with previous work of Jenkins et al [36] which demonstrated that dissemination in other sites revealed a hallmark of increased metastatic potential. This indicates that the ECM is sufficient to influence the enrichment of a stable metastatic subpopulation. We also found that INV cells failed to induce significant platelet adhesion and/or aggregation (data not shown), a process previously reported to be involved in metastasis processes [37]. Previously, cell migration was reported to result from a balance between cell adhesion and de-adhesion, both acquired for motility. Indeed, strong adhesion would tightly attach cancer cells to the substrate and prevent their movement. Reduced cell adhesion to both ECM and endothelial cells, observed for our INV subpopulation, is consistent with their enhanced migration ability, and their capacity to cross the endothelial barrier to invade secondary organs and form metastases. Thus, we found that the cells obtained by transmigration through an extracellular matrix displayed a loss of adhesion behavior. Studies on breast cancer progression indicate that loss of cadherin or cell polarity and the acquisition of a mesenchymal phenotype are associated with an increase in invasiveness [38]. Accordingly, our INV cells, the invasive subpopulation, had a fusiform shape with large spaces between cells and thus resembled mesenchymal cells. Also, we found weak expression of OCLN and strong expression of TGFβ2, an association which was previously demonstrated to be linked to the breast cancer epithelial mesenchymal transition (EMT) [39]. Thus, our results support the notion that the basement membrane may also contribute to the EMT leading to plasticity of metastatic cells.

Using GSEA MsigDB [40] data bases, we found that the functions of 38% of down-regulated and 47% of up-regulated genes in INV are known (see also Fig. S1). Most are involved in the immune/inflammatory response, cell cycle (M phase), adhesion and dorsal root ganglion. Accordingly, 44% of the known down-regulated genes in INV cells are involved in adhesion and cell-cell junction (FERMT3, GABRA3, OCLN, AMIGO2, SCARF1, ASAM, PCDHB2).

Many (25%) of the genes more strongly expressed in INV than REF cells, are linked to cell survival by negative regulation of apoptosis. INV cells also over-expressed the proangiogenic factor (VEGF) and its receptor NRP-1, both involved in the autocrine survival of cells and microvessel permeability, processes required for cell extravasation [41]. Other genes including IRAK1, NKAP, FLNA, XIAP and TGFß2 were also up-regulated: XIAP and TGFβ2 are associated with the NFKB pathways possibly being pivotal potential signaling targets [42], [43]. To note, also the implication of UPA, UPAR and PAR-I in the NFKB pathway and that acted in concert with MMP-9 to enhance INV cell invasion. Thus, further studies on the NFKB signaling in these cell would seem interesting to undergo. Some pro-apoptotic genes were down-regulated in INV cells. However, important pro-apoptotic gene as for example, BCL2 was not found to be upregulated, (Fig. S2). These include rather CASP4 and cyclin D2, previously reported to be involved in genistein and doxorubicin sensitization, respectively [44], [30]. The down-regulation of cyclin D2 in INV cells suggests neosis, an important process that mediates tumor cell resistance to apoptosis, heterogeneity and continuity [45]. These findings are in agreement with reports of the critical role of survival during the metastatic process of tumor cells [46]. Further, AKT analysis showed no differences between the INV and REF cells, suggesting that several survival genes mentioned above acted in concert to increase resistance to death. Interestingly, in vivo, we only observed small differences in cell death between INV cell-derived tumors and REF tumors, suggesting that the observed difference in tumor volumes might result from a dormancy exit. Indeed, several reports reveal the participation of angiogenesis in the exit from dormancy by inhibiting apoptosis [47]. The in vitro identification of anti-apoptotic or survival targets in this work constitutes an interesting and original finding that could lead to better treat aggressive sub populations. In particular, it could be a good strategy to prevent cancer relapse and to better investigate tumor dormancy in term of therapy as well as diagnosis. In basis of these first results, further investigation should address preclinal models to indentify targets in order to treat and also survey long-term cancer survivors minimizing the risk of cancer recurrence.

Conclusions

We demonstrated that using only basement membrane extracellular matrix it is possible to obtain a stable metastatic population without using expensive and longer in vivo procedures. Our results suggest that future strategies for anti-metastatic therapies should target molecules involved in angiogenesis and survival rather than in cell adhesion. We report here the genetic, phenotypic and functional characterization of invasive (INV) cancer cell population. We believe that such cells will provide unique cellular models for in vitro screening of anti-metastatic drugs with clinical potential.

Supporting Information

Figure S1.

Gene expression analysis of REF and INV cells. Down-regulated genes (A) and up-regulated genes (B) were determined using Affymetrix Human Gene 1.0ST arrays as described in “Materials and Methods”.

https://doi.org/10.1371/journal.pone.0023334.s001

(TIF)

Figure S2.

BCL2 mRNA quantification in REF and INV cells. Total RNA (1 µg) was reverse-transcribed using MMLV RT and subjected to qRT-PCR as described in Material and Methods.

https://doi.org/10.1371/journal.pone.0023334.s002

(TIF)

Table S1.

Complete list of genes that are down- and up-regulated in INV cells relative to REF cells. Gene functions are taken from the GO Description provided by NetAffx™ Analysis Center Affymetrix (https://www.affymetrix.com/analysis/netaffx/index.affx). * Validated by qRT-PCR.

https://doi.org/10.1371/journal.pone.0023334.s003

(DOC)

Acknowledgments

We thank N. Setterblat for excellent technical support and Marc Lecouvey for supporting students.

Author Contributions

Conceived and designed the experiments: MDB G-YP MC CN A-MK. Performed the experiments: MA NV AR OS-C SR-F JM LC M-PP CD ME NP AS. Analyzed the data: MDB M-LM JA P-OC CN SM A-MK MC AS. Contributed reagents/materials/analysis tools: MDB SM CN P-OC J-LP. Wrote the paper: MDB CN G-YP AS A-MK.

References

  1. 1. Perret GY, Crepin M (2008) New pharmacological strategies against metastatic spread. Fundamental & Clinical Pharmacology 5: 465–492.
  2. 2. Shekhar MP, Pauley R, Heppner G (2003) Host microenvironment in breast cancer development: extracellular matrix-stromal cell contribution to neoplastic phenotype of epithelial cells in the breast. Breast Cancer Res 5: 130–135.
  3. 3. Haslam SZ, Woodward TL (2001) Reciprocal regulation of extracellular matrix proteins and ovarian steroid activity in the mammary gland. Breast Cancer Res 3: 365–372.
  4. 4. Cunha GR, Alarid ET, Turner T, Donjacour AA, Boutin EL, et al. (1992) Normal and abnormal development of the male urogenital tract. Role of androgens, mesenchymal-epithelial interactions, and growth factors. J Androl 13: 465–475.
  5. 5. Erler JT, Weaver VM (2009) Three-dimensional context regulation of metastasis. Clinical & Experimental Metastasis 26: 35–49.
  6. 6. Shekhar MP, Werdell J, Santner SJ, Pauley RJ, Tait L (2001) Breast stroma plays a dominant regulatory role in breast epithelial growth and differentiation: implications for tumor development and progression. Cancer Res 15: 1320–1326.
  7. 7. Hynes RO (2009) The extracellular matrix: not just pretty fibrils. Science 27: 1216–1219.
  8. 8. Engbring JA, Kleinman HK (2003) The basement membrane matrix in malignancy. J Pathol 200: 465–470.
  9. 9. Yang L (2010) TGF beta and cancer metastasis: an inflammation link. Cancer Metastasis Rev 29: 263–271.
  10. 10. Brekhman V, Neufeld G (2009) A novel asymmetric 3D in-vitro assay for the study of tumor cell invasion. BMC Cancer 30: 415.
  11. 11. Pupa SM, Ménard S, Forti S, Tagliabue E (2002) New insights into the role of extracellular matrix during tumor onset and progression. J Cell Physiol 192: 259–267.
  12. 12. Kleinman HK, Koblinski J, Lee S, Engbring J (2001) Role of basement membrane in tumor growth and metastasis. Surg Oncol Clin N Am 10: 329–38.
  13. 13. Kleinman HK, McGarvey ML, Hassell JR, Star VL, Cannon FB, et al. (1986) Basement membrane complexes with biological activity. Biochemistry 28: 312–318.
  14. 14. Albini A, Iwamoto Y, Kleinman HK, Martin GR, Aaronson SA, et al. (1987) A rapid in vitro assay for quantitating the invasive potential of tumor cells. Cancer Res 15: 3239–3245.
  15. 15. Cailleau R, Young R, Olivé M, Reeves WJ Jr (1974) Breast tumor cell lines from pleural effusions. J Natl Cancer Inst 53: 661–674.
  16. 16. Weksler BB, Subileau EA, Perrière N, Charneau P, Holloway K, et al. (2005) Blood-brain barrier-specific properties of a human adult brain endothelial cell line. FASEB J 13: 1872–1874.
  17. 17. Poller B, Gutmann H, Krähenbühl S, Weksler B, Romero I, et al. (2008) The human brain endothelial cell line hCMEC/D3 as a human blood-brain barrier model for drug transport studies. J Neurochem 107: 1358–1368.
  18. 18. Di Benedetto M, Starzec A, Vassy R, Crépin M, Kraemer M (2003) Inhibition of epidermoid carcinoma A431 cell growth and angiogenesis in nude mice by early and late treatment with a novel dextran derivative. Br J Cancer 88: 1987–1994.
  19. 19. Di Benedetto M, Bièche I, Deshayes F, Vacher S, Nouet S, et al. (2006) Structural organization and expression of human MTUS1, a candidate 8p22 tumor suppressor gene encoding a family of angiotensin II AT2 receptor-interacting proteins, ATIP. Gene 380: 127–136.
  20. 20. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, et al. (2003) Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2: 249–264.
  21. 21. Delmar P, Robin S, VarMixt DaudinJ.-J (2005) Efficient variance modelling for the differential analysis of replicated gene expression data. Bioinformatics 4: 502–508.
  22. 22. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Statist Soc B57:1: 289–300.
  23. 23. Storey JD, Tibshirani R (2003) Statistical significance for genome-wide experiments. Proc Nat Acad Sci USA 100: 9440–9445.
  24. 24. Hamma-Kourbali Y, Di Benedetto M, Ledoux D, Oudar O, Leroux Y, et al. (2003) A novel non-containing-nitrogen bisphosphonate inhibits both in vitro and in vivo angiogenesis. Biochem Biophys Res Comm 310: 816–882.
  25. 25. Abdelkarim M, Guenin E, Sainte-Catherine O, Vintonenko N, Peyri N, et al. (2009) New symmetrically esterified m-bromobenzyl non-aminobisphosphonates inhibited breast cancer growth and metastases. PLoS One 4: e4685.
  26. 26. Orr FW, Wang HH, Lafrenie RM, Scherbarth S, Nance DM (2000) Interactions between cancer cells and the endothelium in metastasis. J Pathol 190: 310–329.
  27. 27. Wang T, Lv JH, Zhang XF, Li CJ, Han X, Sun YJ (2010) Tissue inhibitor of metalloproteinase-1 protects MCF-7 breast cancer cells from paclitaxel-induced apoptosis by decreasing the stability of cyclin B1. Int J Cancer 15: 362–70.
  28. 28. Lei X, Yang J, Nichols RW, Sun LZ (2007) Abrogation of TGFbeta signaling induces apoptosis through the modulation of MAP kinase pathways in breast cancer cells. Exp Cell Res 2007 313: 1687–1695.
  29. 29. De Angelis PM, Fjell B, Kravik KL, Haug T, Tunheim SH, et al. (2004) Molecular characterizations of derivatives of HCT116 colorectal cancer cells that are resistant to the chemotherapeutic agent 5-fluorouracil. Int J Oncol 24: 1279–1288.
  30. 30. Sweeney KJ, Sarcevic B, Sutherland RL, Musgrove EA (1997) Cyclin D2 activates Cdk2 in preference to Cdk4 in human breast epithelial cells. Oncogene 14: 1329–1340.
  31. 31. Kranz , Mattfeldt T, Waltenberger J (1999) Molecular mediators of tumor angiogenesis: enhanced expression and activation of vascular endothelial growth factor receptor KDR in primary breast cancer. Int J Cancer 84: 293–298.
  32. 32. Hicklin DJ, Ellis LM (2005) Role of the vascular endothelial growth factor pathway in tumor growth and angiogenesis. J Clin Oncol 23: 1011–27.
  33. 33. Weigand M, Hantel P, Kreienberg R, Waltenberger J (2005) Autocrine vascular endothelial growth factor signalling in breast cancer. Evidence from cell lines and primary breast cancer cultures in vitro. Angiogenesis 3: 197–204.
  34. 34. Aesoy R, Sanchez BC, Norum JH, Lewensohn R, Viktorsson K, Linderholm B (2008) An autocrine VEGF/VEGFR2 and p38 signaling loop confers resistance to 4-hydroxytamoxifen in MCF-7 breast cancer cells. Mol Cancer Res 10: 1630–1638.
  35. 35. Bachelder RE, Crago A, Chung J, Wendt MA, Shaw LM, et al. (2001) Vascular endothelial growth factor is an autocrine survival factor for neuropilin-expressing breast carcinoma cells. Cancer Res 15: 5736–5740.
  36. 36. Jenkins DE, Hornig YS, Oei Y, Dusich J, Purchio T (2005) Bioluminescent human breast cancer cell lines that permit rapid and sensitive in vivo detection of mammary tumors and multiple metastases in immune deficient mice. Breast Cancer Res 7: 444–454.
  37. 37. Weiss L, Grundmann E, Torhorst J, Moberg I, Eder M, Fenoglio-Preiser CM, et al. (1986) Haematogenous metastatic patterns in colonic carcinoma: an analysis of 1541 necropsies. J Pathol 150: 195–203.
  38. 38. Aokage K, Ishii G, Ohtaki Y, Yamaguchi Y, Hishida T, et al. (2010) Dynamic molecular changes associated with epithelial-mesenchymal transition and subsequent mesenchymal-epithelial transition in the early phase of metastatic tumor formation. Int J Cancer. in press.
  39. 39. Tsuji T, Ibaragi S, Hu GF (2009) Epithelial-mesenchymal transition and cell cooperativity in metastasis. Cancer Res 69: 7135–7139.
  40. 40. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, et al. (2005) Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102: 15545–15550.
  41. 41. Weis S, Cui J, Barnes L, Cheresh D (2004) Endothelial barrier disruption by VEGF-mediated Src activity potentiates tumor cell extravasation and metastasis. J Cell Biol 167: 223–229.
  42. 42. Fan Y, Dutta J, Gupta N, Fan G, Gélinas C (2008) Regulation of programmed cell death by NF-kappaB and its role in tumorigenesis and therapy. Adv Exp Med Biol 615: 223–250.
  43. 43. Berezovskaya O, Schimmer AD, Glinskii AB, Pinilla C, Hoffman RM, et al. (2005) Increased expression of apoptosis inhibitor protein XIAP contributes to anoikis resistance of circulating human prostate cancer metastasis precursor cells. Cancer Res 15: 2378–2386.
  44. 44. Yang S, Zhou Q, Yang X (2007) Caspase-3 status is a determinant of the differential responses to genistein between MDA-MB-231 and MCF-7 breast cancer cells. Biochim Biophys Acta 1773: 903–911.
  45. 45. Rajaraman R, Guernsey DL, Rajaraman MM, Rajaraman SR (2006) Stem cells, senescence, neosis and self-renewal in cancer. Cancer Cell Int 6: 25.
  46. 46. Mehlen P, Puisieux A (2006) Metastasis: a question of life or death. Nat Rev Cancer 6: 449–458.
  47. 47. Rajaraman R, Guernsey DL, Rajaraman MM, Rajaraman SR (2006) Stem cells, senescence, neosis and self-renewal in cancer. Cancer Cell Int 6: 25.