Nanoscale View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
PAPER
Cite this: DOI: 10.1039/c5nr09126h
View Journal
DDB2 (damaged-DNA binding 2) protein: a new modulator of nanomechanical properties and cell adhesion of breast cancer cells Claire Barbieux,a,b Jalal Bacharouche,c,d Charles Soussen,a,b Sébastien Hupont,e Angélina Razafitianamaharavo,f,g Rémi Klotz,a,b Rémi Pannequin,a,b David Brie,a,b Philippe Bécuwe,a,b Grégory Francius*c,d and Stéphanie Grandemange*a,b DDB2, known for its role in DNA repair, was recently shown to reduce mammary tumor invasiveness by inducing the transcription of IκBα, an inhibitor of NF-κB activity. Since cellular adhesion is a key event during the epithelial to mesenchymal transition (EMT) leading to the invasive capacities of breast tumor cells, the aim of this study was to investigate the role of DDB2 in this process. Thus, using low and high DDB2-expressing MDA-MB231 and MCF7 cells, respectively, in which DDB2 expression was modulated experimentally, we showed that DDB2 overexpression was associated with a decrease of adhesion abilities on glass and plastic areas of breast cancer cells. Then, we investigated cell nanomechanical properties by atomic force microscopy (AFM). Our results revealed significant changes in the Young’s Modulus value and the adhesion force in MDA-MB231 and MCF7 cells, whether DDB2 was expressed or not. The cell stiffness decrease observed in MDA-MB231 and MCF7 expressing DDB2 was correlated with a loss of the cortical actin-cytoskeleton staining. To understand how DDB2 regulates these processes, an adhesionrelated gene PCR-Array was performed. Several adhesion-related genes were differentially expressed
Received 22nd December 2015, Accepted 4th February 2016
according to DDB2 expression, indicating that important changes are occurring at the molecular level. Thus, this work demonstrates that AFM technology is an important tool to follow cellular changes during
DOI: 10.1039/c5nr09126h
tumorigenesis. Moreover, our data revealed that DDB2 is involved in early events occurring during metastatic progression of breast cancer cells and will contribute to define this protein as a new marker of
www.rsc.org/nanoscale
metastatic progression in this type of cancer.
1.
Introduction
Due to the high incidence and mortality rate of metastatic breast cancers, it is critical to understand the mechanisms
a Université de Lorraine, Centre de Recherche en Automatique de Nancy, CRAN, UMR 7039, Vandœuvre-lès-Nancy, F-54506, France. E-mail: [email protected], [email protected] b CNRS, Centre de Recherche en Automatique de Nancy, CRAN, UMR 7039, Vandœuvre-lès-Nancy, F-54506, France c Université de Lorraine, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, LCPME, UMR 7564, Villers-lès-Nancy, F-54600, France. E-mail: [email protected] d CNRS, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, LCPME, UMR 7564, Villers-lès-Nancy, F-54600, France e CNRS, FR3209 Biologie Moléculaire Cellulaire et Thérapeutique (BMCT), Plateforme d’Imagerie Cellulaire et Tissulaire PTIBC-IBISA, Biopôle de l’Université de Lorraine, Campus Biologie-Santé, Vandœuvre-lès-Nancy, F-54506, France f Université de Lorraine, Laboratoire Interdisciplinaire des Environnements Continentaux, LIEC, UMR 7360, Vandœuvre-lès-Nancy, F-54500, France g CNRS, Laboratoire Interdisciplinaire des Environnements Continentaux, LIEC, UMR 7360, Vandœuvre-lès-Nancy, F-54500, France
This journal is © The Royal Society of Chemistry 2016
behind metastasis formation. The invasive potential of cancer cells is often associated with different modifications such as morphology, adhesion to extracellular matrix, cytoskeleton organization and cell-to-cell interaction. However, molecular mechanisms involved in the invasive potential of cancer cells are not fully understood. Recently, increasing evidence point out the fact that the protein Damaged DNA Binding 2 (DDB2), originally known for its role in nucleotide excision repair after UV-DNA damages, plays an important role in breast cancer growth by a transcriptional regulation of target genes.1–4 DDB2 is constitutively overexpressed in non invasive breast cancer cells, whose proliferation increases by inhibiting the transcription of sod2 gene encoding the manganese superoxide dismutase and prevents the acquisition of invasive properties by upregulating IκBα gene expression which leads to a lower NFκB activity.2 In parallel, Roy and colleagues have shown that DDB2 overexpression prevented the epithelial to mesenchymal transition (EMT) of aggressive colon cancer cells. Indeed, DDB2 was found to be bound on the promoter of several genes encoding EMT-inducing factors,3 leading to a decrease of their
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
expression. EMT is a key event in invasiveness of cancer cells as well as in tumor progression toward metastasis.5 In accordance with these data, the expression of DDB2 was inversely correlated with the highest colon or breast tumor grade. The Atomic Force Microscopy (AFM) approach has been widely used in biological research, particularly in cancer research area.6–8 This technology allows to get topographic information by scanning cell surfaces on their natural support as well as to measure cell adhesion forces. During tumorigenesis, cells are subject to several modifications and differ from normal cells in terms of growth, morphology, adhesion to extracellular matrix, cytoskeleton organization and cell-to-cell interaction. Most of these changes are associated with particular mechanical properties of cell membranes that can be measured by AFM. In cancer research, AFM has been mainly used for exploratory research on cancer cell morphology,9 cell elasticity10 or to explore the real-time changes in the ultrastructure of the cell membrane following drug interactions.11 In addition, a recent study by Xu et al. demonstrates that cell stiffness, determined by AFM measurement, could be a potential biomarker of ovarian cancer cells.12 Indeed, cell stiffness not only discriminates ovarian cancer cells from non-malignant cells but also distinguishes between high and low tumorigenic invasive cancer cells. In accordance with this work, Plodinec et al. have shown in Nature Nanotechnology that breast cancer biopsies have a particular stiffness profile compared to normal tissues as attested by AFM measurements.13 Thus stiffness measurements by AFM could be an interesting tool in clinical diagnosis as well as in fundamental research. EMT and cell invasiveness are associated with modifications of cancer cell adhesion by loss of epithelial characteristics including cell polarity and cell–cell contact. In this context, we hypothesize that DDB2 could play a role in mechanical and adhesion properties of breast cancer cells. Using AFM technology, we demonstrate that the nuclear DDB2 protein modulates cell adhesion by (i) strongly affecting the nanomechanical properties of breast tumor cells; (ii) promoting a remodeling of the cortical actin-cytoskeleton; (iii) inducing changes in the adhesion-related gene expression.
2. Materials and methods 2.1.
Cell culture, transfection and infection
Two human breast cancer cell lines, MCF7 and MDA-MB231 were purchased from European Collection of Cell Culture. Cells were grown in RPMI 1640 without phenol red (Gibco) supplemented with 10% (v/v) fetal calf serum, 1% penicillin/ streptomycin and 2 mM L-glutamine at 37 °C, 5% CO2. As described previously,2 MCF7 cells, derived from non metastatic breast adenocarcinoma and expressing constitutively high level of DDB2, were infected with lentiviral particles containing either small hairpin RNA targeting DDB2 (MCF7 DDB2−) to inhibit DDB2 expression, or a nontargeting control sequence (MCF7 DDB2+), and puromycin resistance gene for selection (Santa Cruz). Infected cells were selected by puromycin treat-
Nanoscale
ment (1 µg mL−1) for 72 hours. MDA-MB231 cells, derived from metastatic breast adenocarcinoma and expressing low level of DDB2, were stably transfected with pcDNA3(+) expression vector containing the cDNA encoding DDB2 (MDA DDB2+) or not (MDA DDB2−), using Jet PEI reagent (PolyPlus Transfection), as described previously.1 Resistant cells to 400 µg mL−1 G418 were isolated and screened for DDB2 expression by RT-qPCR and Western Blot analysis, as described previously.2 Two days before AFM acquisitions, cells were seeded at 8 × 104 cells per mL in a 10 cm2 dish. At 50% confluency, acquisitions were done on adherent living cells maintained in RPMI 1640 medium. 2.2.
Western blot analysis
Cells were harvested, rinsed with ice-cold PBS and lysed into 10 mM Tris-HCl buffer pH 7.4 containing 5 mM EDTA, 1% Triton X-100 and protease inhibitors. After centrifugation, the total protein concentrations of supernatants were determined according to Bradford protein assay, using bovine serum albumin as a standard (Biorad). A total protein (50 µg) was run on SDS-polyacrylamide gels according to Laemmli and transferred onto a PVDF membrane. After blocking with milk (5%) in 60 mM Tris buffer containing 150 mM Nacl and Tween (0.1%) (TBS-Tween), immunoblot analysis was then carried out by incubating membranes with either rabbit polyclonal antibody anti-DDB2 (Abcam) or rabbit monoclonal antibody anti α-tubulin (Abcam) or mouse polyclonal antibody anti I-CAM1 (Santa Cruz) diluted in TBS-Tween containing 2.5% milk. After washings, bands corresponding to DDB2 or tubulin were detected using an anti-IgG polyclonal antibody conjugated to peroxydase and diluted in TBS-Tween containing 2.5% milk, and were visualized with a ChemiDoc system (Biorad) after exposition to a chemiluminescent substrate. Band intensities were quantified by densitometry with QuantityOne software (Biorad). Relative DDB2 and I-CAM1 expression was the result of the ratio between densitometry corresponding to DDB2 on tubulin. 2.3.
Cell adhesion assay
Cells were seeded at 2.5 × 104 per well in a 96 well-plate directly either on the plastic surface, glass coverslip, or on plate previously coated with fibronectin (10 µg cm−2), collagens I and IV (20 µg cm−2), laminin (4 µg cm−2), vitronectin (0.2 µg cm−2) or matrigel (10.5 µg cm−2). The plate was incubated during 30 min (or 90 min for the adhesion on glass) at 37 °C, and then non-adherent cells were eliminated by washing with PBS containing Ca2+ and Mg2+. Before staining with Crystal Violet (0.1% in ethanol 0.4%), adherent cells were fixed in paraformaldehyde 4% during 30 min at RT. Fixed adherent cells were resuspended in 10% acetic acid during 10 min and absorbance at 595 nm was recorded using a microplate reader (Perkin Elmer). The maximum adhesion (100%) for each experimental condition correspond to the 2.5 × 104 seeded cells, incubated during 30 or 90 min at 37 °C before being fixed and stained. The percentage of adherent cells for each condition was calculated from the maximal adhesion rate.
This journal is © The Royal Society of Chemistry 2016
View Article Online
Paper
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
2.4. Measurements of elastic modulus by AFM nanoindentation Force measurements were carried out using a MFP3D-BIO instrument (Asylum Research Technology, Atomic Force F & E GmbH, Mannheim, Germany). The nanoindentation method provides the Young’s modulus calculated from the force vs. indentation curve. Triangular cantilevers were purchased from Bruker (MLCT, Bruker-nano AXS, Palaiseau, France). The spring constants of the cantilevers measured using the thermal noise method, were calculated in the range of 12–16 pN nm−1. Experiments were performed in RPMI 1640 medium at room temperature and cell plates were changed every 2 hours to avoid morphological and pH changes. Maps of mechanical properties (FVI for Force Volume Image) were obtained by recording a grid map of 50-by-50 force curves at different locations of 80 µm × 80 µm on the Petri dish surface. The maximum loading force was 5 nN. Maps of mechanical and adhesive properties and the corresponding histograms (statistical distribution) were estimated from the analysis of the approach curves according to the Sneddon’s model. Within the framework of Sneddon’s theory, the loading force F depends on the indentation depth δ according to: F¼
2E TanðαÞ 2 δ fBECC πð1 ν2 Þ
where δ is the indentation depth, ν the Poisson coefficient, α is the semi-top angle of the tip and fBECC is the Bottom Effect Cone Correction function that takes into account the stiffness of the film-supporting substrate.14 FVI were performed at 0.5 Hz scan rate and with maximal loading/pulling speed of about 10 µm s−1. Then they were analyzed by means of an automatic Matlab algorithm described elsewhere.6 Briefly, the physicochemical parameters are estimated from the raw data using a nonlinear regression procedure. The signal processing method is based on advanced tools for “segmenting” the force curves i.e., automatically detecting the tip-to-surface contact point with accuracy. 2.5.
Actin cytoskeleton staining
MCF7 and MDA-MB231 cells were seeded on glass coverslips, in 12-well plate at 105 and 1.5 × 105 cells per well, respectively. Twenty four hours after seeding, cells were fixed with paraformaldehyde 4% at room temperature. Then, cells were blocked and permeabilized with PBS containing 3% BSA and 0.2% Triton X-100 during 20 minutes. Actin cytoskeleton was stained with Alexa Fluor 488 phalloidin (Invitrogen) diluted at 1 : 50 in PBS at room temperature in the darkness. After three washes with PBS, nuclei were stained with TO-PRO-3 (Life Technologies) diluted at 1 : 1000 in PBS. Coverslips were mounted in antifading medium (FluorSafe, Merck) and observed with a confocal microscope LEICA TCS SPX AOBS CLSM. Images at a 0.256 micrometer sidelength square pixel size were obtained for each case in 1024 × 1024 matrices at 400× magnification (numerical aperture = 0.8) of the CLSM. Fluorescence emissions were recorded within an Airy disk confocal pinhole
This journal is © The Royal Society of Chemistry 2016
Table 1 List of genes analyzed by Human Extracellular Matrix and Adhesion Molecules PCR-Array. Adhesion-related genes are highlighted in blue and housekeeping genes are highlighted in pink. Wells highlighted in green correspond to internal controls for genomic DNA presence (HGDC), reverse transcription efficiency (RTC) and PCR efficiency (PPC)
Nanoscale
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
setting (Airy 1). Each channel was acquired sequentially. The first detection channel ( phalloidin) was set from 502 to 600 nm with a 492 nm excitation laser line. The second detection channel (TO-PRO-3) was set from 653 to 730 nm with a 643 nm excitation laser line. Confocal 3D Acquisitions (xyz) were performed on 10 micron thick corresponding to eight focal planes. To quantify the cortical actin-cytoskeleton, 10 regions of interest corresponding to 40 pixels, were positioned arbitrarily on peripheral membranes of cells in each picture. The profiles of fluorescence intensity were measured on the 10 regions per picture. The maximum value of fluorescence in these profiles was extracted and the median value was calculated. 2.6.
PCR array
Expression level of 84 adhesion-related genes was evaluated by a quantitative PCR Array (Human Extracellular Matrix and Adhesion Molecules RT2 Profiler PCR Array, Qiagen) and compared to 5 housekeeping genes (the list of these genes is given in Table 1). RNAs were extracted from 70–80% confluent cells with ReliaPrep RNA Miniprep columns (Promega) according to manufacturer’s instructions without modifications. Their quantity and quality were determined spectroscopically using a nanodrop 2000c (Thermo Scientific, France). Reverse transcription was performed using 1 µg of RNA with RT2 First Strand Kit (Qiagen), according to manufacturer’s recommendations. Quantitative PCR was conducted using diluted cDNA in RT2 SYBR Green Mastermix (Qiagen). PCR reactions were run on the iCycler real-time PCR instrument (Opticon 2, BioRad), according to the manufacturer instructions. Specificity of the products was assured by melting curve analysis. The relative transcript level, based on cycle threshold (Ct), was quantified by calculating the 2−ΔΔCt. Briefly, the 2−ΔCt was obtained by calculating the 2−Ct value for each adhesion-related gene and normalized with average of the 2−Ct value of housekeeping genes. The 2−ΔΔCt was finally obtained by dividing the 2−ΔCt in DDB2 overexpressing cells (2−ΔCtDDB2+) by the 2−ΔCtDDB2− in low DDB2 expressing cells, and that for each adhesion-related gene. This ratio was lower than 1 for gene repression and conversely, higher than 1 for gene overexpression in DDB2 overexpressing cells. 2.7.
Statistical analysis
All results are represented as mean value ± SEM. Statistical analyses were performed by using Student’s t test which compares DDB2− versus DDB2+ cells for each cell line (MDA-MB231 and MCF7). Statistically significant results were represented as follows: * p < 0.05, ** p < 0.01, *** p < 0.001.
3. Results and discussion 3.1. DDB2 expression in breast cancer cells correlates with a loss of adhesion properties Since migration and invasiveness are closely linked to cell adhesion modifications, we focused on the ability of DDB2
Nanoscale
protein to modulate adhesion properties in breast cancer cell lines expressing or not DDB2. To do so, two cell lines were used: MDA-MB231 cells expressing an empty vector (MDA DDB2−) versus cells expressing the cDNA encoding DDB2 protein (MDA DDB2+) and MCF7 cells stably transfected with a scramble shRNA (MCF7 DDB2+) versus cells stably transfected with shRNA targeting DDB2 mRNA (MCF7 DDB2−). As shown in Fig. 1, DDB2 expression was increased by 4.8-fold in MDA DDB2+ in contrast to MDA DDB2−, while it was reduced by 2.1-fold in MCF7 DDB2− in contrast to MCF7 DDB2+. As DDB2 protein was shown to negatively control the migratory and invasive abilities of breast cancer cell lines by positively regulating the IκBα gene expression, we hypothesized that DDB2 expression could also be associated with some modifications of cell adhesion. Thus we measured adhesion properties of MDA DDB2+ cells vs. control MDA DDB2− cells as well as MCF7 DDB2− cells vs. control MCF7 DDB2+ cells by counting adherent cells on different substrates including glass, plastic, and several extracellular matrix proteins (fibronectin, collagen I and IV, laminin and vitronectin). The number of adherent cells on neutral support, such as plastic and glass, was significantly decreased in MDA DDB2+ cells as compared to MDA DDB2− cells (Fig. 2A). Conversely, a significant increase in the number of adherent cells was only observed on glass support for MCF7 DDB2− cells compared with MCF7 DDB2+ cells (Fig. 2B). No significant differences in the adhesion of cells on extracellular matrix proteins were observed for both cell lines (data not shown). These differences being observed only on glass and plastic substratum, these data indicate that DDB2 modulate non-specific adhesion properties of cells. As adhesion on this substratum is due to hydrophilic ( polystyrene of tissue culture plate are treated to increase their hydrophilicity) and electrostatic interactions (van der Waals interactions), we can suppose that DDB2 expression could be associated with modifications of the global membrane composition. Indeed, Loomis and colleagues have demonstrated that non-specific adhesion is triggered by membrane glycoprotein and lead to van der Waals interactions.15 Moreover, cholesterol composition can also modify the adhesion properties on neutral substratum such as plastic plates.16,17 Thus, DDB2 expression is associated with changes in adhesion properties of breast cancer cells that could be related to modifications of membrane composition. Therefore, we attempted to link these changes to physicochemical modifications of cellular plasma membrane, as already reported.18 3.2. DDB2 expression is associated with changes in physical and mechanical properties of breast cancer cell plasma membrane Modifications in physical and mechanical properties of plasma membrane are related to different cellular processes such as adhesion to extracellular matrix, cell-to-cell interaction but also in tumorigenesis.19,20 Thus, AFM experiments were performed in order to quantify the mechanical and adhesion properties of the cells overexpressing or not DDB2, as illus-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
Fig. 1 DDB2 expression level in MDA-MB231 and MCF7 breast cancer cell lines. MDA-MB231 cells were stably transfected with an empty vector (MDA DDB2−) or a plasmid containing DDB2 gene (MDA DDB2+). MCF7 cells were infected with lentiviral particles containing a plasmid expressing a shRNA control (MCF7 DDB2+) or a shRNA directed against DDB2 (MCF7 DDB2−). DDB2 encoding gene expression was analyzed by RT-qPCR (A and B) and normalized by 2 housekeeping genes (UBB and RPLP0). DDB2 protein expression was analyzed by Western blot (C) with 50 µg of total protein from each tumor cells. The density of each band was measured and compared with the density of tubulin bands used as loading control. Results are presented as the fold increase or decrease over control cells fixed to 1 and correspond to the mean value of three independent experiments ± SEM.
Fig. 2 Correlation between DDB2 expression and loss of adhesion abilities on culture areas of breast cancer cells. (A) Number of adherent MDA-MB231 and (B) MCF7 cells onto plastic and glass was estimated by cell adhesion assay. For each substrate, DDB2+ cells (grey bars) were compared with DDB2− cells (black bars). Results are expressed as the percentage of adherent cells out of total cells and are presented as a mean value of 3 independent experiments ± SEM. Differences between DDB2+ and DDB2− cells were statistically significant for * p < 0.05, ** p < 0.01, *** p < 0.001.
This journal is © The Royal Society of Chemistry 2016
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
trated in Fig. 3. Mechanical properties were extracted from the approach force curves (red lines in Fig. 3A) using Sneddon’s model whereas the adhesion forces between the cell and AFMtip were measured from the analysis of the retraction force curves (blue lines in Fig. 3A). In addition, the force-volume mode of the AFM was used to spatially perform resolved grid of 50 × 50 force curves on surfaces of 80 × 80 μm2 to entirely scan several cells as illustrated in Fig. 3B. The analysis of spatially resolved force curves allows the construction of several maps of topography (Fig. 4), elasticity, adhesion and work of adhesion of cells. So the cell physical properties can
Nanoscale
be quantified and correlated to their specific location with high resolution, e.g. elasticity or adhesion values can be attributed to both the nucleus and the cytoplasm. This point is crucial since Young’s modulus and adhesion forces calculated on cell surface depend on intracellular structures, such as the cytoskeleton, the presence of underlying organelles,21 as well as cell membrane glycoproteins,22 that contribute to the stiffening/adhesion properties of the cell surface.23,24 The calculation of these parameters was performed on each cell carefully delineated (as indicated in the elasticity map presented Fig. 3B). Furthermore, AFM measurements are the
Fig. 3 Principle of AFM measurements and related data. The upper panel (A) represents applied methods for AFM force-volume acquisitions in the approach (left) and retraction (center) phases. The right subfigure shows representative force curves. The lower panel (B) displays each image reconstructed after some acquisition. (a) An optical view of cell monolayer. (d) An height map representing the cell topography. The other subfigures represent maps of the physico-chemical parameters extracted using the signal processing software. (b) Corresponds to the Young’s modulus of all pixels of delineated cells 1 to 6, obtained in the approach phase. On the other hand the force (e) and work of adhesion (f ) are estimated in the retraction phase. (c) Represents the signal-to-noise ratio (SNR), a factor of confidence of the reconstructed values of Young’s modulus shown in (b). For pixels lying in the cell regions, most of SNR values are greater than 25 dB, which means that the force curve analysis (nonlinear regression algorithm) is reliable, and the related values of the Young modulus are then accurate.
Nanoscale
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
Fig. 4 Height-map cell morphologies are not affected after the modulation of DDB2 expression. Height maps obtained by AFM measurements are shown. Pictures (left column) result from a scan of 80 × 80 µm with a resolution of 50 × 50 pixels. The dotted line identifies the section realized in middle graphs. The right column displays the same picture in three dimensions. Several acquisitions have been performed and most representative pictures were selected.
most suitable technique to investigate and quantify in situ the physico-chemical properties of whole living cells without any treatment, contrary to fluorescence and other microscopies.25,26 The analysis of cell morphology depicted in Fig. 5A (upper panel) shows that MDA-MB231 cells are characterized by an average surface area of about 367 and 319 μm2, respectively for cells expressing DDB2 or not. In the case of MCF7 cells (Fig. 5A, lower panel), an average surface area of 352 and 382 μm2 is evaluated for the spreading cells in which DDB2 is repressed or not, respectively. These results indicate that the modulation of DDB2 expression is neither associated with morphological differences in MDA-MB231 nor in MCF7 cells. The statistical analysis of cell stiffness and adhesion to tip measured by AFM and described in Fig. 5B and C shows important variations when DBB2 is overexpressed. In Fig. 5B, cell stiffness decreased from 5 to 2 kPa and from 3 to 2.5 kPa for MDA-MB231 and MCF7 cells, respectively, when DDB2 is expressed. However, the effect was more intense in MDA-MB231 cells as compared to MCF7. This could be due to the stronger differential expression level of DDB2 in MDA DDB2− versus MDA DDB2+ cells compared to MCF7 DDB2+ versus MCF7 DDB2− as indicated in Fig. 1. Thus these data point out that the expression of DDB2, known to be correlated with an inhibition of the migratory and invasiveness properties of breast cancer cells, is associated with a decrease of the cell
This journal is © The Royal Society of Chemistry 2016
stiffness. Interestingly, some previous works highlighted a negative correlation between invasiveness and stiffness in different types of cancer cells.12,27,28 Nevertheless, experimental conditions need to be considered. For example, cell indentation performed in the perinuclear region of individual cell could lead to extremely different results in contrast to entirely scanned cells. Moreover, culture conditions could also lead to opposite effect as well as if stiffness measurements were recorded on living or fixed cells.29 Indeed, support adhesion processes are associated with multiple modifications of biophysical characteristics of plasma membrane (such as plasma membrane glycoprotein expression/activation, cytoskeleton reorganization and physical interaction with extra-cellular matrix proteins). In this sense, a study from Teng et al. has shown that the adhesion of cells on different extra-cellular matrix compounds is associated with Young’s modulus modifications.28,30 In contrast, the stiffness was increased when the modulation of myoferlin and SV40T expression was associated with an increase of the cell invasiveness.31,32 Interestingly, membrane stiffness analysis was done between the nucleus and the most peripheral border of the normal cells versus SV40T expressing cells to avoid the underlying nucleus rigidity. Furthermore, a recent paper from Wu and colleagues, revealed that a treatment of lung carcinoma cells with TGFβ-1, which is known to induce EMT,33,34 was associated with an increase in the cell stiffness and adhesion force.35
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
Fig. 5 Nanomechanical properties of breast cancer cells according to DDB2 expression level. From the left to the right, the surface (A), the Young’s Modulus value (B) and the adhesion force (C) for MDA-MB231 and MCF7 cells are presented. Low-DDB2 expressing cells (DDB2−) are represented with black bars and DDB2 expressing cells (DDB2+) are represented with grey bars. These physico-chemical parameters were obtained from the analysis of 33 MDA-MB231 DDB2− cells, 38 MDA-MB231 DDB2+, 29 MCF7 DDB2+ cells and 38 MCF7 DDB2−. Data are represented as the average value ± SEM from 3 independent experiments and are indicated statistically significant for * p < 0.05, ** p < 0.01, *** p < 0.001.
To evaluate changes in cell surface adhesion, we measured the cellular adhesive force that depends on the mechanical properties of the existing adhesive molecules into the plasma membrane. In Fig. 5C, a significant decrease in the adhesion to tip was observed in breast tumor cells expressing DDB2: from 0.46 nN in MDA DDB2− to 0.3 nN in MDA DDB2+ cells) and 0.19 nN in MCF7 DDB2+ to 0.14 nN in MCF7 DDB2− cells. These results are in agreement with the cell adhesion measurements to plastic and glass support presented in Fig. 2. Indeed, these two different experiments demonstrate a significant decrease of the adhesion properties of DDB2 expressing cells. The cytoskeleton organization, depending mainly on the actin network, is a common indicator of cell stiffness.36,37 To verify the organization of the actin cytoskeleton, actin network was stained using phalloidin and observed by fluorescence imaging (Fig. 6). A drop in actin staining, particularly in the cortical area under the plasma membrane was observed in DDB2-overexpressing MDA-MB231 cells in contrast to MDA-MB231 DDB2− cells. Conversely and in accordance with these results, we observed a higher staining of cortical actin
Nanoscale
cytoskeleton in DDB2-downregulated MCF7 cells than in control MCF7 that expressed DDB2. Since F-actin is closely linked to the cellular morphology and to the mechanical properties of cells, these data could thus explain that DDB2 expressing cells (MDA-MB231 and MCF7) are less stiff than DDB2-low expressing cells, as indicated above (Fig. 5). 3.3. DDB2 expression is associated with modification of adhesion-related gene expression To investigate if DDB2 could lead to global membrane changes and to better understand the molecular mechanisms that could lead to DDB2-dependent changes in adhesion of breast cancer cells, we performed a PCR array including 84 different adhesion-related genes and 5 housekeeping genes in breast cancer cell lines expressing or not DDB2 (Table 1). Among these 84 genes, 31 encode transmembrane proteins (highlighted in the Table 2). We observed changes in the expression of several genes according to DDB2 expression as reported in Table 2A and B). Among the 84 adhesion-related genes analyzed, only 14 were modulated when DDB2 was overexpressed in MDA-MB231 cells (MDA DDB2+). Most of them were down-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Paper
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Fig. 6 Reorganization of cortical actin-cytoskeleton according to DDB2 expression in breast cancer cells. After the fixation of MDA-MB231 and MCF7 cells expressing or not DDB2, nuclei and F-actin filaments were stained with TO-PRO-3 and Phalloidin, respectively. (A) Cells were observed by confocal microcopy (magnification 400×). The 8 confocal z sections of cell expressing or not DDB2 (left), and the z axis projection (right) are presented. (B) The median value of maximum fluorescence intensity from 10 regions of interest, arbitrarily placed across the cortical actin-cytoskeleton, was calculated for each acquisition. Data are represented as the average value ± SEM from 3 independent experiments and are indicated statistically significant for * p < 0.05.
regulated, except for ICAM1 and SELL which were significantly upregulated in MDA DDB2+ cells. Concerning the analysis of MCF7 cells, 20 out of the 84 genes analyzed were modulated
This journal is © The Royal Society of Chemistry 2016
when DDB2 expression was inhibited, and only 6 were downregulated in MCF7 DDB2+ compared to MCF7 DDB2− cells (Table 2A and B). Probably due to the different phenotype
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper Table 2 Changes in the adhesion-related gene expression according to DDB2 expression. A PCR Array targeting 84 adhesion and extracellular matrix related genes was performed, using total RNA extracted from cells. Quantification of each gene expression level was based on cycle threshold and a fold change between MDA-MB231 (A) and MCF7 (B) DDB2+ and DDB2− cells. Genes are downregulated or upregulated, when the ratio is lower or higher than 1, respectively in DDB2 overexpressing cells, in contrast to the DDB2− cells. Fold changes are presented as a mean value ± SEM of three independent experiments. Differences between DDB2+ and DDB2− cells for differentially expressed genes are statistically significant for p < 0.05, p < 0.01, and p < 0.001
between the epithelial MCF7 cells and mesenchymal MDA cells, only 3 common genes, ICAM1 (intracellular adhesion molecule 1), CTGF (connective tissue growth factor) and TGFβ1 (transforming growth factor beta 1), were modulated similarly in both cell lines, suggesting that DDB2 would modulate adhesion-related gene expression in a cell specific manner. CTGF and TGFβ1 genes are known to be broadly implicated in adhesion as well as invasion and migration of tumor cells.
Nanoscale
Nanoscale
Indeed, CTGF expression seems to be associated with a poor prognosis in breast cancers and its overexpression has been linked to an increase of migratory properties of breast cancer cells.38–40 Furthermore, TGFβ1 is a potent secreted factor that drives cancer progression. While it is known to act as tumor suppressor by inhibiting epithelial cell proliferation, it plays a key role in breast tumor progression toward metastasis by inducing EMT and invasive properties in vitro.41–43 Both TGFβ1 and CTGF have been shown to be associated with an increase of bone and pulmonary metastasis formation in vivo.39,44 CTGF being a TGFβ1 responsive gene, we may speculate that the downregulation of this gene is related to DDB2-dependent TGFβ1 gene downregulation in breast cancer DDB2+ cells.45 However, even if growing evidence show that DDB2 is acting as a transcriptional regulator, its direct involvement in regulation of such adhesion-related gene expression needs to be further investigated. The modulation of cellular adhesive force and adhesion properties in cells expressing or not DDB2 could also be explained by the modification of membrane proteins and other relative amounts of adhesive compounds as shown by Cross et al.46 and Pi et al.47 It is the case of ICAM1, a protein which is also known to be largely involved in tumor cell adhesion, even if the role of this gene in cancer cells is quite controversial. It has been shown that the downregulation of ICAM1 could be associated with a decrease of breast cancer invasion.48 However, other studies have reported an opposite effect of ICAM1 in tumorigenesis, by associating its expression with a relatively good prognosis in breast cancers.49,50 They have shown that a decrease in the ICAM1 expression was associated with breast tumor cell dissemination, favoring development of metastases in order to escape from immune surveillance.51 Consistent with all these data, our results showing the downregulation of CTGF and TGFβ-1 expression and the upregulation of ICAM1 are in accordance with previous knowledge related to higher proliferation rates, an inhibition of the migratory and invasive properties as well as a better outcome of colon and breast cancers overexpressing DDB2.2,3,52 Our data highlighted that both DDB2+ cell lines expressed higher ICAM1 mRNA levels. Among the 31 genes encoded transmembrane proteins, only the gene encoded ICAM1 is modulated similarly in both cell tumor cell lines. To attest that DDB2 dependent adhesion properties, cell stiffness and adhesion force modifications could be due to differences in cellular membrane composition, the level of ICAM1 protein has been evaluated (Fig. 7). According to the transcript level evaluated by PCR array, ICAM1 protein is significantly upregulated in DDB2-expressing cells in contrast to their counterpart. Thus, these results could reflect in part a relation between changes in the cellular membrane composition and the DDB2dependent nanomechanical properties modifications. Moreover, the decrease in the cell stiffness observed in DDB2+ cells could be related to the downregulation of some other adhesion related genes in MDA DDB2+ cells such as those encoding the integrin isoforms ITGA2, ITGA6, ITGB4 and ITGA1 in MCF7 DDB2+ cells (Table 2). Integrins are trans-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
colon and breast cancer via the regulation of EMT inducing factors and IκBα, the cytoplasmic inhibitor of NFκB transcription factor. Taken together, these data indicate a new role of DDB2 in adhesion and mechanical properties of cancer cells that are the early events occurring during tumor progression toward metastasis. Thus our results bring new insight that DDB2 could be a relevant clinical marker of breast cancer progression.
Acknowledgements Fig. 7 Increase of ICAM1 expression level in DDB2 expressing cells. Expression of ICAM1 protein was analyzed by Western blot with 50 µg of total protein from breast tumor cells. The density of each band was measured and compared with the density of tubulin bands used as loading control. Results are presented as the fold increase or decrease over control cells fixed to 1 and correspond to the mean value of three independent experiments ± SEM.
This work was supported by the CNRS and the University of Lorraine (PEPS Mirabelle Bioforce Project) and the “Ligue contre le cancer” (Comité de Meurthe et Moselle). The fellowship of Claire Barbieux was given by the French Research Ministry (contrat doctoral de l’Université de Lorraine). The authors thank Dr S. Kaminski for critical reading of the article.
References membrane proteins well known for their role as adhesion molecules by interacting with extracellular matrix but also by acting on actin cytoskeleton organization and transduction of intracellular signal regulating many cellular functions such as proliferation, migration and invasion.53 Integrins are linked to the actin cytoskeleton via a scaffold protein complex under the plasma membrane.54–56 Thus, the cytoskeletal reorganization observed in DDB2 expressing cells compared to their counterparts could be related to the decrease in the expression of integrins, as referred in Table 2.
4.
Conclusion
In the present work we showed that DDB2, a nuclear protein commonly known for its major role in DNA repair and more recently involved in the control of the invasive abilities of breast cancer cells, is associated with cell adhesion by modulating membrane mechanical properties. We observed upon microscale and nanoscale analysis by AFM, PCR-Array and fluorescence microscopy that expression of DDB2 modulates cancer cell stiffness and strongly impacts the cortical actincytoskeleton organization. Indeed, we demonstrated quantitatively that expression of DDB2 decreases cell adhesion, Young’s modulus of MCF7 and MDA-MB231 cells. These modifications could be related to modulation of specific adhesion related gene expression in DDB2 overexpressing cells as compared to DDB2− cells that could impact the global plasma membrane composition, in accordance with the new role of DDB2 as transcriptional regulator.1–4 Furthermore, these data indicate that nanomechanical properties of cancer cell measured by AFM could be useful to follow the changes occurring during tumorigenesis. It has been already shown that DDB2 plays a role in tumorigenesis by preventing cells to acquire invasive abilities in
This journal is © The Royal Society of Chemistry 2016
1 V. Minig, Z. Kattan, J. van Beeumen, E. Brunner and P. Becuwe, J. Biol. Chem., 2009, 284, 14165–14176. 2 M. Ennen, R. Klotz, N. Touche, S. Pinel, C. Barbieux, V. Besancenot, E. Brunner, D. Thiebaut, A. C. Jung, S. Ledrappier, L. Domenjoud, J. Abecassis, F. Plénat, S. Grandemange and P. Becuwe, Cancer Res., 2013, 73, 5040–5052. 3 N. Roy, P. V. Bommi, U. G. Bhat, S. Bhattacharjee, I. Elangovan, J. Li, K. C. Patra, D. Kopanja, A. Blunier, R. Benya, S. Bagchi and P. Raychaudhuri, Cancer Res., 2013, 73, 3771–3782. 4 R. Zhao, C. Han, E. Eisenhauer, J. Kroger, W. Zhao, J. Yu, K. Selvendiran, X. Liu, A. A. Wani and Q.-E. Wang, Mol. Cancer Res., 2014, 12, 370–380. 5 B. De Craene and G. Berx, Nat. Rev. Cancer, 2013, 13, 97–110. 6 P. Polyakov, C. Soussen, J. Duan, J. F. L. Duval, D. Brie and G. Francius, PLoS One, 2011, 6, e18887. 7 D. C. Lin, E. K. Dimitriadis and F. Horkay, J. Biomech. Eng., 2007, 129, 904–912. 8 L. Sirghi, J. Ponti, F. Broggi and F. Rossi, Eur. Biophys. J., 2008, 37, 935–945. 9 E. Canetta, A. Riches, E. Borger, S. Herrington, K. Dholakia and A. K. Adya, Acta Biomater., 2014, 10, 2043– 2055. 10 M. Lekka, K. Pogoda, J. Gostek, O. Klymenko, S. PrauznerBechcicki, J. Wiltowska-Zuber, J. Jaczewska, J. Lekki and Z. Stachura, Micron Oxf. Engl., 2012, 43, 1259–1266. 11 S. E. Cross, Y.-S. Jin, Q.-Y. Lu, J. Rao and J. K. Gimzewski, Nanotechnology, 2011, 22, 215101. 12 W. Xu, R. Mezencev, B. Kim, L. Wang, J. McDonald and T. Sulchek, PLoS One, 2012, 7, e46609. 13 M. Plodinec, M. Loparic, C. A. Monnier, E. C. Obermann, R. Zanetti-Dallenbach, P. Oertle, J. T. Hyotyla, U. Aebi,
Nanoscale
View Article Online
Paper
14
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
15 16 17
18 19 20 21 22
23
24 25
26
27 28 29 30 31
32
33
M. Bentires-Alj, R. Y. H. Lim and C.-A. Schoenenberger, Nat. Nanotechnol., 2012, 7, 757–765. N. Gavara and R. S. Chadwick, Nat. Nanotechnol., 2012, 7, 733–736. W. F. Loomis, D. Fuller, E. Gutierrez, A. Groisman and W.-J. Rappel, PLoS One, 2012, 7, e42033. M. Sun, N. Northup, F. Marga, T. Huber, F. J. Byfield, I. Levitan and G. Forgacs, J. Cell Sci., 2007, 120, 2223–2231. O. G. Ramprasad, G. Srinivas, K. S. Rao, P. Joshi, J. P. Thiery, S. Dufour and G. Pande, Cell Motil. Cytoskeleton, 2007, 64, 199–216. Y. Shan and H. Wang, Chem. Soc. Rev., 2015, 44, 3617– 3638. D. J. Müller and Y. F. Dufrêne, Trends Cell Biol., 2011, 21, 461–469. S. Suresh, Acta Biomater., 2007, 3, 413–438. C. Roduit, S. Sekatski, G. Dietler, S. Catsicas, F. Lafont and S. Kasas, Biophys. J., 2009, 97, 674–677. C. Roduit, F. G. van der Goot, P. De Los Rios, A. Yersin, P. Steiner, G. Dietler, S. Catsicas, F. Lafont and S. Kasas, Biophys. J., 2008, 94, 1521–1532. D. F. Tschaharganeh, W. Xue, D. F. Calvisi, M. Evert, T. V. Michurina, L. E. Dow, A. Banito, S. F. Katz, E. R. Kastenhuber, S. Weissmueller, C.-H. Huang, A. Lechel, J. B. Andersen, D. Capper, L. Zender, T. Longerich, G. Enikolopov and S. W. Lowe, Cell, 2014, 158, 579–592. F. Guilak, J. R. Tedrow and R. Burgkart, Biochem. Biophys. Res. Commun., 2000, 269, 781–786. T. G. Kuznetsova, M. N. Starodubtseva, N. I. Yegorenkov, S. A. Chizhik and R. I. Zhdanov, Micron Oxf. Engl., 2007, 38, 824–833. M. Lekka, D. Gil, K. Pogoda, J. Dulińska-Litewka, R. Jach, J. Gostek, O. Klymenko, S. Prauzner-Bechcicki, Z. Stachura, J. Wiltowska-Zuber, K. Okoń and P. Laidler, Arch. Biochem. Biophys., 2012, 518, 151–156. S. E. Cross, Y.-S. Jin, J. Rao and J. K. Gimzewski, Nat. Nanotechnol., 2007, 2, 780–783. E. L. Baker, J. Lu, D. Yu, R. T. Bonnecaze and M. H. Zaman, Biophys. J., 2010, 99, 2048–2057. E. A. Corbin, F. Kong, C. T. Lim, W. P. King and R. Bashir, Lab Chip, 2015, 15, 839–847. Y. Teng, J. Qiu, Y. Zheng, X. Luo, L. Zhang, L. Chen and G. Wang, J. Med. Biol. Eng., 2012, 62–68. L.-S. Z. Rathje, N. Nordgren, T. Pettersson, D. Rönnlund, J. Widengren, P. Aspenström and A. K. B. Gad, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 1515–1520. L. I. Volakis, R. Li, W. E. Ackerman, C. Mihai, M. Bechel, T. L. Summerfield, C. S. Ahn, H. M. Powell, R. Zielinski, T. J. Rosol, S. N. Ghadiali and D. A. Kniss, PLoS One, 2014, 9, e86110. J. M. Zarzynska, Mediators Inflammation, 2014, 2014, e141747.
Nanoscale
Nanoscale
34 J. Massagué, Nat. Rev. Mol. Cell Biol., 2012, 13, 616–630. 35 T.-H. Wu, Y.-W. Chou, P.-H. Chiu, M.-J. Tang, C.-W. Hu and M.-L. Yeh, Cancer Cell Int., 2014, 14, 20. 36 A. M. Collinsworth, S. Zhang, W. E. Kraus and G. A. Truskey, Am. J. Physiol.: Cell Physiol., 2002, 283, C1219–C1227. 37 S. T. Buckley, C. Medina, A. M. Davies and C. Ehrhardt, Nanomed. Nanotechnol. Biol. Med., 2012, 8, 355–364. 38 P.-S. Chen, M.-Y. Wang, S.-N. Wu, J.-L. Su, C.-C. Hong, S.-E. Chuang, M.-W. Chen, K.-T. Hua, Y.-L. Wu, S.-T. Cha, M. S. Babu, C.-N. Chen, P.-H. Lee, K.-J. Chang and M.-L. Kuo, J. Cell Sci., 2007, 120, 2053–2065. 39 W. Chien, J. O’Kelly, D. Lu, A. Leiter, J. Sohn, D. Yin, B. Karlan, J. Vadgama, K. M. Lyons and H. P. Koeffler, Int. J. Oncol., 2011, 38, 1741–1747. 40 D. Xie, K. Nakachi, H. Wang, R. Elashoff and H. P. Koeffler, Cancer Res., 2001, 61, 8917–8923. 41 L. Kubiczkova, L. Sedlarikova, R. Hajek and S. Sevcikova, J. Transl. Med., 2012, 10, 183. 42 M. H. Barcellos-Hoff and R. J. Akhurst, Breast Cancer Res., 2009, 11, 202. 43 D. Padua and J. Massagué, Cell Res., 2009, 19, 89–102. 44 A. A. Mourskaia, Z. Dong, S. Ng, M. Banville, J. C. Zwaagstra, M. D. O’Connor-McCourt and P. M. Siegel, Oncogene, 2009, 28, 1005–1015. 45 G. R. Grotendorst, H. Okochi and N. Hayashi, Cell Growth Differ. Mol. Biol. J. Am. Assoc. Cancer Res., 1996, 7, 469–480. 46 S. E. Cross, Y.-S. Jin, J. Tondre, R. Wong, J. Rao and J. K. Gimzewski, Nanotechnology, 2008, 19, 384003. 47 J. Pi, T. Li, J. Liu, X. Su, R. Wang, F. Yang, H. Bai, H. Jin and J. Cai, Micron Oxf. Engl., 2014, 65, 1–9. 48 C. Rosette, R. B. Roth, P. Oeth, A. Braun, S. Kammerer, J. Ekblom and M. F. Denissenko, Carcinogenesis, 2005, 26, 943–950. 49 Y. Ogawa, K. Hirakawa, B. Nakata, T. Fujihara, T. Sawada, Y. Kato, K. Yoshikawa and M. Sowa, Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res., 1998, 4, 31–36. 50 A. C. Donadio, M. M. Remedi, S. Frede, G. R. Bonacci, G. A. Chiabrando and M. C. Pistoresi-Palencia, Clin. Exp. Metastasis, 2002, 19, 437–444. 51 S. Mukai, H. Kagamu, S. Shu and G. E. Plautz, Cell. Immunol., 1999, 192, 122–132. 52 Z. Kattan, S. Marchal, E. Brunner, C. Ramacci, A. Leroux, J. L. Merlin, L. Domenjoud, M. Dauça and P. Becuwe, PLoS One, 2008, 3, e2002. 53 A. Huttenlocher and A. R. Horwitz, Cold Spring Harbor Perspect. Biol., 2011, 3, a005074. 54 K. A. DeMali, K. Wennerberg and K. Burridge, Curr. Opin. Cell Biol., 2003, 15, 572–582. 55 I. Delon and N. H. Brown, Curr. Opin. Cell Biol., 2007, 19, 43–50. 56 C. G. Galbraith, K. M. Yamada and J. A. Galbraith, Science, 2007, 315, 992–995.
This journal is © The Royal Society of Chemistry 2016
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
PAPER
Cite this: DOI: 10.1039/c5nr09126h
View Journal
DDB2 (damaged-DNA binding 2) protein: a new modulator of nanomechanical properties and cell adhesion of breast cancer cells Claire Barbieux,a,b Jalal Bacharouche,c,d Charles Soussen,a,b Sébastien Hupont,e Angélina Razafitianamaharavo,f,g Rémi Klotz,a,b Rémi Pannequin,a,b David Brie,a,b Philippe Bécuwe,a,b Grégory Francius*c,d and Stéphanie Grandemange*a,b DDB2, known for its role in DNA repair, was recently shown to reduce mammary tumor invasiveness by inducing the transcription of IκBα, an inhibitor of NF-κB activity. Since cellular adhesion is a key event during the epithelial to mesenchymal transition (EMT) leading to the invasive capacities of breast tumor cells, the aim of this study was to investigate the role of DDB2 in this process. Thus, using low and high DDB2-expressing MDA-MB231 and MCF7 cells, respectively, in which DDB2 expression was modulated experimentally, we showed that DDB2 overexpression was associated with a decrease of adhesion abilities on glass and plastic areas of breast cancer cells. Then, we investigated cell nanomechanical properties by atomic force microscopy (AFM). Our results revealed significant changes in the Young’s Modulus value and the adhesion force in MDA-MB231 and MCF7 cells, whether DDB2 was expressed or not. The cell stiffness decrease observed in MDA-MB231 and MCF7 expressing DDB2 was correlated with a loss of the cortical actin-cytoskeleton staining. To understand how DDB2 regulates these processes, an adhesionrelated gene PCR-Array was performed. Several adhesion-related genes were differentially expressed
Received 22nd December 2015, Accepted 4th February 2016
according to DDB2 expression, indicating that important changes are occurring at the molecular level. Thus, this work demonstrates that AFM technology is an important tool to follow cellular changes during
DOI: 10.1039/c5nr09126h
tumorigenesis. Moreover, our data revealed that DDB2 is involved in early events occurring during metastatic progression of breast cancer cells and will contribute to define this protein as a new marker of
www.rsc.org/nanoscale
metastatic progression in this type of cancer.
1.
Introduction
Due to the high incidence and mortality rate of metastatic breast cancers, it is critical to understand the mechanisms
a Université de Lorraine, Centre de Recherche en Automatique de Nancy, CRAN, UMR 7039, Vandœuvre-lès-Nancy, F-54506, France. E-mail: [email protected], [email protected] b CNRS, Centre de Recherche en Automatique de Nancy, CRAN, UMR 7039, Vandœuvre-lès-Nancy, F-54506, France c Université de Lorraine, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, LCPME, UMR 7564, Villers-lès-Nancy, F-54600, France. E-mail: [email protected] d CNRS, Laboratoire de Chimie Physique et Microbiologie pour l’Environnement, LCPME, UMR 7564, Villers-lès-Nancy, F-54600, France e CNRS, FR3209 Biologie Moléculaire Cellulaire et Thérapeutique (BMCT), Plateforme d’Imagerie Cellulaire et Tissulaire PTIBC-IBISA, Biopôle de l’Université de Lorraine, Campus Biologie-Santé, Vandœuvre-lès-Nancy, F-54506, France f Université de Lorraine, Laboratoire Interdisciplinaire des Environnements Continentaux, LIEC, UMR 7360, Vandœuvre-lès-Nancy, F-54500, France g CNRS, Laboratoire Interdisciplinaire des Environnements Continentaux, LIEC, UMR 7360, Vandœuvre-lès-Nancy, F-54500, France
This journal is © The Royal Society of Chemistry 2016
behind metastasis formation. The invasive potential of cancer cells is often associated with different modifications such as morphology, adhesion to extracellular matrix, cytoskeleton organization and cell-to-cell interaction. However, molecular mechanisms involved in the invasive potential of cancer cells are not fully understood. Recently, increasing evidence point out the fact that the protein Damaged DNA Binding 2 (DDB2), originally known for its role in nucleotide excision repair after UV-DNA damages, plays an important role in breast cancer growth by a transcriptional regulation of target genes.1–4 DDB2 is constitutively overexpressed in non invasive breast cancer cells, whose proliferation increases by inhibiting the transcription of sod2 gene encoding the manganese superoxide dismutase and prevents the acquisition of invasive properties by upregulating IκBα gene expression which leads to a lower NFκB activity.2 In parallel, Roy and colleagues have shown that DDB2 overexpression prevented the epithelial to mesenchymal transition (EMT) of aggressive colon cancer cells. Indeed, DDB2 was found to be bound on the promoter of several genes encoding EMT-inducing factors,3 leading to a decrease of their
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
expression. EMT is a key event in invasiveness of cancer cells as well as in tumor progression toward metastasis.5 In accordance with these data, the expression of DDB2 was inversely correlated with the highest colon or breast tumor grade. The Atomic Force Microscopy (AFM) approach has been widely used in biological research, particularly in cancer research area.6–8 This technology allows to get topographic information by scanning cell surfaces on their natural support as well as to measure cell adhesion forces. During tumorigenesis, cells are subject to several modifications and differ from normal cells in terms of growth, morphology, adhesion to extracellular matrix, cytoskeleton organization and cell-to-cell interaction. Most of these changes are associated with particular mechanical properties of cell membranes that can be measured by AFM. In cancer research, AFM has been mainly used for exploratory research on cancer cell morphology,9 cell elasticity10 or to explore the real-time changes in the ultrastructure of the cell membrane following drug interactions.11 In addition, a recent study by Xu et al. demonstrates that cell stiffness, determined by AFM measurement, could be a potential biomarker of ovarian cancer cells.12 Indeed, cell stiffness not only discriminates ovarian cancer cells from non-malignant cells but also distinguishes between high and low tumorigenic invasive cancer cells. In accordance with this work, Plodinec et al. have shown in Nature Nanotechnology that breast cancer biopsies have a particular stiffness profile compared to normal tissues as attested by AFM measurements.13 Thus stiffness measurements by AFM could be an interesting tool in clinical diagnosis as well as in fundamental research. EMT and cell invasiveness are associated with modifications of cancer cell adhesion by loss of epithelial characteristics including cell polarity and cell–cell contact. In this context, we hypothesize that DDB2 could play a role in mechanical and adhesion properties of breast cancer cells. Using AFM technology, we demonstrate that the nuclear DDB2 protein modulates cell adhesion by (i) strongly affecting the nanomechanical properties of breast tumor cells; (ii) promoting a remodeling of the cortical actin-cytoskeleton; (iii) inducing changes in the adhesion-related gene expression.
2. Materials and methods 2.1.
Cell culture, transfection and infection
Two human breast cancer cell lines, MCF7 and MDA-MB231 were purchased from European Collection of Cell Culture. Cells were grown in RPMI 1640 without phenol red (Gibco) supplemented with 10% (v/v) fetal calf serum, 1% penicillin/ streptomycin and 2 mM L-glutamine at 37 °C, 5% CO2. As described previously,2 MCF7 cells, derived from non metastatic breast adenocarcinoma and expressing constitutively high level of DDB2, were infected with lentiviral particles containing either small hairpin RNA targeting DDB2 (MCF7 DDB2−) to inhibit DDB2 expression, or a nontargeting control sequence (MCF7 DDB2+), and puromycin resistance gene for selection (Santa Cruz). Infected cells were selected by puromycin treat-
Nanoscale
ment (1 µg mL−1) for 72 hours. MDA-MB231 cells, derived from metastatic breast adenocarcinoma and expressing low level of DDB2, were stably transfected with pcDNA3(+) expression vector containing the cDNA encoding DDB2 (MDA DDB2+) or not (MDA DDB2−), using Jet PEI reagent (PolyPlus Transfection), as described previously.1 Resistant cells to 400 µg mL−1 G418 were isolated and screened for DDB2 expression by RT-qPCR and Western Blot analysis, as described previously.2 Two days before AFM acquisitions, cells were seeded at 8 × 104 cells per mL in a 10 cm2 dish. At 50% confluency, acquisitions were done on adherent living cells maintained in RPMI 1640 medium. 2.2.
Western blot analysis
Cells were harvested, rinsed with ice-cold PBS and lysed into 10 mM Tris-HCl buffer pH 7.4 containing 5 mM EDTA, 1% Triton X-100 and protease inhibitors. After centrifugation, the total protein concentrations of supernatants were determined according to Bradford protein assay, using bovine serum albumin as a standard (Biorad). A total protein (50 µg) was run on SDS-polyacrylamide gels according to Laemmli and transferred onto a PVDF membrane. After blocking with milk (5%) in 60 mM Tris buffer containing 150 mM Nacl and Tween (0.1%) (TBS-Tween), immunoblot analysis was then carried out by incubating membranes with either rabbit polyclonal antibody anti-DDB2 (Abcam) or rabbit monoclonal antibody anti α-tubulin (Abcam) or mouse polyclonal antibody anti I-CAM1 (Santa Cruz) diluted in TBS-Tween containing 2.5% milk. After washings, bands corresponding to DDB2 or tubulin were detected using an anti-IgG polyclonal antibody conjugated to peroxydase and diluted in TBS-Tween containing 2.5% milk, and were visualized with a ChemiDoc system (Biorad) after exposition to a chemiluminescent substrate. Band intensities were quantified by densitometry with QuantityOne software (Biorad). Relative DDB2 and I-CAM1 expression was the result of the ratio between densitometry corresponding to DDB2 on tubulin. 2.3.
Cell adhesion assay
Cells were seeded at 2.5 × 104 per well in a 96 well-plate directly either on the plastic surface, glass coverslip, or on plate previously coated with fibronectin (10 µg cm−2), collagens I and IV (20 µg cm−2), laminin (4 µg cm−2), vitronectin (0.2 µg cm−2) or matrigel (10.5 µg cm−2). The plate was incubated during 30 min (or 90 min for the adhesion on glass) at 37 °C, and then non-adherent cells were eliminated by washing with PBS containing Ca2+ and Mg2+. Before staining with Crystal Violet (0.1% in ethanol 0.4%), adherent cells were fixed in paraformaldehyde 4% during 30 min at RT. Fixed adherent cells were resuspended in 10% acetic acid during 10 min and absorbance at 595 nm was recorded using a microplate reader (Perkin Elmer). The maximum adhesion (100%) for each experimental condition correspond to the 2.5 × 104 seeded cells, incubated during 30 or 90 min at 37 °C before being fixed and stained. The percentage of adherent cells for each condition was calculated from the maximal adhesion rate.
This journal is © The Royal Society of Chemistry 2016
View Article Online
Paper
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
2.4. Measurements of elastic modulus by AFM nanoindentation Force measurements were carried out using a MFP3D-BIO instrument (Asylum Research Technology, Atomic Force F & E GmbH, Mannheim, Germany). The nanoindentation method provides the Young’s modulus calculated from the force vs. indentation curve. Triangular cantilevers were purchased from Bruker (MLCT, Bruker-nano AXS, Palaiseau, France). The spring constants of the cantilevers measured using the thermal noise method, were calculated in the range of 12–16 pN nm−1. Experiments were performed in RPMI 1640 medium at room temperature and cell plates were changed every 2 hours to avoid morphological and pH changes. Maps of mechanical properties (FVI for Force Volume Image) were obtained by recording a grid map of 50-by-50 force curves at different locations of 80 µm × 80 µm on the Petri dish surface. The maximum loading force was 5 nN. Maps of mechanical and adhesive properties and the corresponding histograms (statistical distribution) were estimated from the analysis of the approach curves according to the Sneddon’s model. Within the framework of Sneddon’s theory, the loading force F depends on the indentation depth δ according to: F¼
2E TanðαÞ 2 δ fBECC πð1 ν2 Þ
where δ is the indentation depth, ν the Poisson coefficient, α is the semi-top angle of the tip and fBECC is the Bottom Effect Cone Correction function that takes into account the stiffness of the film-supporting substrate.14 FVI were performed at 0.5 Hz scan rate and with maximal loading/pulling speed of about 10 µm s−1. Then they were analyzed by means of an automatic Matlab algorithm described elsewhere.6 Briefly, the physicochemical parameters are estimated from the raw data using a nonlinear regression procedure. The signal processing method is based on advanced tools for “segmenting” the force curves i.e., automatically detecting the tip-to-surface contact point with accuracy. 2.5.
Actin cytoskeleton staining
MCF7 and MDA-MB231 cells were seeded on glass coverslips, in 12-well plate at 105 and 1.5 × 105 cells per well, respectively. Twenty four hours after seeding, cells were fixed with paraformaldehyde 4% at room temperature. Then, cells were blocked and permeabilized with PBS containing 3% BSA and 0.2% Triton X-100 during 20 minutes. Actin cytoskeleton was stained with Alexa Fluor 488 phalloidin (Invitrogen) diluted at 1 : 50 in PBS at room temperature in the darkness. After three washes with PBS, nuclei were stained with TO-PRO-3 (Life Technologies) diluted at 1 : 1000 in PBS. Coverslips were mounted in antifading medium (FluorSafe, Merck) and observed with a confocal microscope LEICA TCS SPX AOBS CLSM. Images at a 0.256 micrometer sidelength square pixel size were obtained for each case in 1024 × 1024 matrices at 400× magnification (numerical aperture = 0.8) of the CLSM. Fluorescence emissions were recorded within an Airy disk confocal pinhole
This journal is © The Royal Society of Chemistry 2016
Table 1 List of genes analyzed by Human Extracellular Matrix and Adhesion Molecules PCR-Array. Adhesion-related genes are highlighted in blue and housekeeping genes are highlighted in pink. Wells highlighted in green correspond to internal controls for genomic DNA presence (HGDC), reverse transcription efficiency (RTC) and PCR efficiency (PPC)
Nanoscale
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
setting (Airy 1). Each channel was acquired sequentially. The first detection channel ( phalloidin) was set from 502 to 600 nm with a 492 nm excitation laser line. The second detection channel (TO-PRO-3) was set from 653 to 730 nm with a 643 nm excitation laser line. Confocal 3D Acquisitions (xyz) were performed on 10 micron thick corresponding to eight focal planes. To quantify the cortical actin-cytoskeleton, 10 regions of interest corresponding to 40 pixels, were positioned arbitrarily on peripheral membranes of cells in each picture. The profiles of fluorescence intensity were measured on the 10 regions per picture. The maximum value of fluorescence in these profiles was extracted and the median value was calculated. 2.6.
PCR array
Expression level of 84 adhesion-related genes was evaluated by a quantitative PCR Array (Human Extracellular Matrix and Adhesion Molecules RT2 Profiler PCR Array, Qiagen) and compared to 5 housekeeping genes (the list of these genes is given in Table 1). RNAs were extracted from 70–80% confluent cells with ReliaPrep RNA Miniprep columns (Promega) according to manufacturer’s instructions without modifications. Their quantity and quality were determined spectroscopically using a nanodrop 2000c (Thermo Scientific, France). Reverse transcription was performed using 1 µg of RNA with RT2 First Strand Kit (Qiagen), according to manufacturer’s recommendations. Quantitative PCR was conducted using diluted cDNA in RT2 SYBR Green Mastermix (Qiagen). PCR reactions were run on the iCycler real-time PCR instrument (Opticon 2, BioRad), according to the manufacturer instructions. Specificity of the products was assured by melting curve analysis. The relative transcript level, based on cycle threshold (Ct), was quantified by calculating the 2−ΔΔCt. Briefly, the 2−ΔCt was obtained by calculating the 2−Ct value for each adhesion-related gene and normalized with average of the 2−Ct value of housekeeping genes. The 2−ΔΔCt was finally obtained by dividing the 2−ΔCt in DDB2 overexpressing cells (2−ΔCtDDB2+) by the 2−ΔCtDDB2− in low DDB2 expressing cells, and that for each adhesion-related gene. This ratio was lower than 1 for gene repression and conversely, higher than 1 for gene overexpression in DDB2 overexpressing cells. 2.7.
Statistical analysis
All results are represented as mean value ± SEM. Statistical analyses were performed by using Student’s t test which compares DDB2− versus DDB2+ cells for each cell line (MDA-MB231 and MCF7). Statistically significant results were represented as follows: * p < 0.05, ** p < 0.01, *** p < 0.001.
3. Results and discussion 3.1. DDB2 expression in breast cancer cells correlates with a loss of adhesion properties Since migration and invasiveness are closely linked to cell adhesion modifications, we focused on the ability of DDB2
Nanoscale
protein to modulate adhesion properties in breast cancer cell lines expressing or not DDB2. To do so, two cell lines were used: MDA-MB231 cells expressing an empty vector (MDA DDB2−) versus cells expressing the cDNA encoding DDB2 protein (MDA DDB2+) and MCF7 cells stably transfected with a scramble shRNA (MCF7 DDB2+) versus cells stably transfected with shRNA targeting DDB2 mRNA (MCF7 DDB2−). As shown in Fig. 1, DDB2 expression was increased by 4.8-fold in MDA DDB2+ in contrast to MDA DDB2−, while it was reduced by 2.1-fold in MCF7 DDB2− in contrast to MCF7 DDB2+. As DDB2 protein was shown to negatively control the migratory and invasive abilities of breast cancer cell lines by positively regulating the IκBα gene expression, we hypothesized that DDB2 expression could also be associated with some modifications of cell adhesion. Thus we measured adhesion properties of MDA DDB2+ cells vs. control MDA DDB2− cells as well as MCF7 DDB2− cells vs. control MCF7 DDB2+ cells by counting adherent cells on different substrates including glass, plastic, and several extracellular matrix proteins (fibronectin, collagen I and IV, laminin and vitronectin). The number of adherent cells on neutral support, such as plastic and glass, was significantly decreased in MDA DDB2+ cells as compared to MDA DDB2− cells (Fig. 2A). Conversely, a significant increase in the number of adherent cells was only observed on glass support for MCF7 DDB2− cells compared with MCF7 DDB2+ cells (Fig. 2B). No significant differences in the adhesion of cells on extracellular matrix proteins were observed for both cell lines (data not shown). These differences being observed only on glass and plastic substratum, these data indicate that DDB2 modulate non-specific adhesion properties of cells. As adhesion on this substratum is due to hydrophilic ( polystyrene of tissue culture plate are treated to increase their hydrophilicity) and electrostatic interactions (van der Waals interactions), we can suppose that DDB2 expression could be associated with modifications of the global membrane composition. Indeed, Loomis and colleagues have demonstrated that non-specific adhesion is triggered by membrane glycoprotein and lead to van der Waals interactions.15 Moreover, cholesterol composition can also modify the adhesion properties on neutral substratum such as plastic plates.16,17 Thus, DDB2 expression is associated with changes in adhesion properties of breast cancer cells that could be related to modifications of membrane composition. Therefore, we attempted to link these changes to physicochemical modifications of cellular plasma membrane, as already reported.18 3.2. DDB2 expression is associated with changes in physical and mechanical properties of breast cancer cell plasma membrane Modifications in physical and mechanical properties of plasma membrane are related to different cellular processes such as adhesion to extracellular matrix, cell-to-cell interaction but also in tumorigenesis.19,20 Thus, AFM experiments were performed in order to quantify the mechanical and adhesion properties of the cells overexpressing or not DDB2, as illus-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
Fig. 1 DDB2 expression level in MDA-MB231 and MCF7 breast cancer cell lines. MDA-MB231 cells were stably transfected with an empty vector (MDA DDB2−) or a plasmid containing DDB2 gene (MDA DDB2+). MCF7 cells were infected with lentiviral particles containing a plasmid expressing a shRNA control (MCF7 DDB2+) or a shRNA directed against DDB2 (MCF7 DDB2−). DDB2 encoding gene expression was analyzed by RT-qPCR (A and B) and normalized by 2 housekeeping genes (UBB and RPLP0). DDB2 protein expression was analyzed by Western blot (C) with 50 µg of total protein from each tumor cells. The density of each band was measured and compared with the density of tubulin bands used as loading control. Results are presented as the fold increase or decrease over control cells fixed to 1 and correspond to the mean value of three independent experiments ± SEM.
Fig. 2 Correlation between DDB2 expression and loss of adhesion abilities on culture areas of breast cancer cells. (A) Number of adherent MDA-MB231 and (B) MCF7 cells onto plastic and glass was estimated by cell adhesion assay. For each substrate, DDB2+ cells (grey bars) were compared with DDB2− cells (black bars). Results are expressed as the percentage of adherent cells out of total cells and are presented as a mean value of 3 independent experiments ± SEM. Differences between DDB2+ and DDB2− cells were statistically significant for * p < 0.05, ** p < 0.01, *** p < 0.001.
This journal is © The Royal Society of Chemistry 2016
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
trated in Fig. 3. Mechanical properties were extracted from the approach force curves (red lines in Fig. 3A) using Sneddon’s model whereas the adhesion forces between the cell and AFMtip were measured from the analysis of the retraction force curves (blue lines in Fig. 3A). In addition, the force-volume mode of the AFM was used to spatially perform resolved grid of 50 × 50 force curves on surfaces of 80 × 80 μm2 to entirely scan several cells as illustrated in Fig. 3B. The analysis of spatially resolved force curves allows the construction of several maps of topography (Fig. 4), elasticity, adhesion and work of adhesion of cells. So the cell physical properties can
Nanoscale
be quantified and correlated to their specific location with high resolution, e.g. elasticity or adhesion values can be attributed to both the nucleus and the cytoplasm. This point is crucial since Young’s modulus and adhesion forces calculated on cell surface depend on intracellular structures, such as the cytoskeleton, the presence of underlying organelles,21 as well as cell membrane glycoproteins,22 that contribute to the stiffening/adhesion properties of the cell surface.23,24 The calculation of these parameters was performed on each cell carefully delineated (as indicated in the elasticity map presented Fig. 3B). Furthermore, AFM measurements are the
Fig. 3 Principle of AFM measurements and related data. The upper panel (A) represents applied methods for AFM force-volume acquisitions in the approach (left) and retraction (center) phases. The right subfigure shows representative force curves. The lower panel (B) displays each image reconstructed after some acquisition. (a) An optical view of cell monolayer. (d) An height map representing the cell topography. The other subfigures represent maps of the physico-chemical parameters extracted using the signal processing software. (b) Corresponds to the Young’s modulus of all pixels of delineated cells 1 to 6, obtained in the approach phase. On the other hand the force (e) and work of adhesion (f ) are estimated in the retraction phase. (c) Represents the signal-to-noise ratio (SNR), a factor of confidence of the reconstructed values of Young’s modulus shown in (b). For pixels lying in the cell regions, most of SNR values are greater than 25 dB, which means that the force curve analysis (nonlinear regression algorithm) is reliable, and the related values of the Young modulus are then accurate.
Nanoscale
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
Fig. 4 Height-map cell morphologies are not affected after the modulation of DDB2 expression. Height maps obtained by AFM measurements are shown. Pictures (left column) result from a scan of 80 × 80 µm with a resolution of 50 × 50 pixels. The dotted line identifies the section realized in middle graphs. The right column displays the same picture in three dimensions. Several acquisitions have been performed and most representative pictures were selected.
most suitable technique to investigate and quantify in situ the physico-chemical properties of whole living cells without any treatment, contrary to fluorescence and other microscopies.25,26 The analysis of cell morphology depicted in Fig. 5A (upper panel) shows that MDA-MB231 cells are characterized by an average surface area of about 367 and 319 μm2, respectively for cells expressing DDB2 or not. In the case of MCF7 cells (Fig. 5A, lower panel), an average surface area of 352 and 382 μm2 is evaluated for the spreading cells in which DDB2 is repressed or not, respectively. These results indicate that the modulation of DDB2 expression is neither associated with morphological differences in MDA-MB231 nor in MCF7 cells. The statistical analysis of cell stiffness and adhesion to tip measured by AFM and described in Fig. 5B and C shows important variations when DBB2 is overexpressed. In Fig. 5B, cell stiffness decreased from 5 to 2 kPa and from 3 to 2.5 kPa for MDA-MB231 and MCF7 cells, respectively, when DDB2 is expressed. However, the effect was more intense in MDA-MB231 cells as compared to MCF7. This could be due to the stronger differential expression level of DDB2 in MDA DDB2− versus MDA DDB2+ cells compared to MCF7 DDB2+ versus MCF7 DDB2− as indicated in Fig. 1. Thus these data point out that the expression of DDB2, known to be correlated with an inhibition of the migratory and invasiveness properties of breast cancer cells, is associated with a decrease of the cell
This journal is © The Royal Society of Chemistry 2016
stiffness. Interestingly, some previous works highlighted a negative correlation between invasiveness and stiffness in different types of cancer cells.12,27,28 Nevertheless, experimental conditions need to be considered. For example, cell indentation performed in the perinuclear region of individual cell could lead to extremely different results in contrast to entirely scanned cells. Moreover, culture conditions could also lead to opposite effect as well as if stiffness measurements were recorded on living or fixed cells.29 Indeed, support adhesion processes are associated with multiple modifications of biophysical characteristics of plasma membrane (such as plasma membrane glycoprotein expression/activation, cytoskeleton reorganization and physical interaction with extra-cellular matrix proteins). In this sense, a study from Teng et al. has shown that the adhesion of cells on different extra-cellular matrix compounds is associated with Young’s modulus modifications.28,30 In contrast, the stiffness was increased when the modulation of myoferlin and SV40T expression was associated with an increase of the cell invasiveness.31,32 Interestingly, membrane stiffness analysis was done between the nucleus and the most peripheral border of the normal cells versus SV40T expressing cells to avoid the underlying nucleus rigidity. Furthermore, a recent paper from Wu and colleagues, revealed that a treatment of lung carcinoma cells with TGFβ-1, which is known to induce EMT,33,34 was associated with an increase in the cell stiffness and adhesion force.35
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper
Nanoscale
Fig. 5 Nanomechanical properties of breast cancer cells according to DDB2 expression level. From the left to the right, the surface (A), the Young’s Modulus value (B) and the adhesion force (C) for MDA-MB231 and MCF7 cells are presented. Low-DDB2 expressing cells (DDB2−) are represented with black bars and DDB2 expressing cells (DDB2+) are represented with grey bars. These physico-chemical parameters were obtained from the analysis of 33 MDA-MB231 DDB2− cells, 38 MDA-MB231 DDB2+, 29 MCF7 DDB2+ cells and 38 MCF7 DDB2−. Data are represented as the average value ± SEM from 3 independent experiments and are indicated statistically significant for * p < 0.05, ** p < 0.01, *** p < 0.001.
To evaluate changes in cell surface adhesion, we measured the cellular adhesive force that depends on the mechanical properties of the existing adhesive molecules into the plasma membrane. In Fig. 5C, a significant decrease in the adhesion to tip was observed in breast tumor cells expressing DDB2: from 0.46 nN in MDA DDB2− to 0.3 nN in MDA DDB2+ cells) and 0.19 nN in MCF7 DDB2+ to 0.14 nN in MCF7 DDB2− cells. These results are in agreement with the cell adhesion measurements to plastic and glass support presented in Fig. 2. Indeed, these two different experiments demonstrate a significant decrease of the adhesion properties of DDB2 expressing cells. The cytoskeleton organization, depending mainly on the actin network, is a common indicator of cell stiffness.36,37 To verify the organization of the actin cytoskeleton, actin network was stained using phalloidin and observed by fluorescence imaging (Fig. 6). A drop in actin staining, particularly in the cortical area under the plasma membrane was observed in DDB2-overexpressing MDA-MB231 cells in contrast to MDA-MB231 DDB2− cells. Conversely and in accordance with these results, we observed a higher staining of cortical actin
Nanoscale
cytoskeleton in DDB2-downregulated MCF7 cells than in control MCF7 that expressed DDB2. Since F-actin is closely linked to the cellular morphology and to the mechanical properties of cells, these data could thus explain that DDB2 expressing cells (MDA-MB231 and MCF7) are less stiff than DDB2-low expressing cells, as indicated above (Fig. 5). 3.3. DDB2 expression is associated with modification of adhesion-related gene expression To investigate if DDB2 could lead to global membrane changes and to better understand the molecular mechanisms that could lead to DDB2-dependent changes in adhesion of breast cancer cells, we performed a PCR array including 84 different adhesion-related genes and 5 housekeeping genes in breast cancer cell lines expressing or not DDB2 (Table 1). Among these 84 genes, 31 encode transmembrane proteins (highlighted in the Table 2). We observed changes in the expression of several genes according to DDB2 expression as reported in Table 2A and B). Among the 84 adhesion-related genes analyzed, only 14 were modulated when DDB2 was overexpressed in MDA-MB231 cells (MDA DDB2+). Most of them were down-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Paper
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Fig. 6 Reorganization of cortical actin-cytoskeleton according to DDB2 expression in breast cancer cells. After the fixation of MDA-MB231 and MCF7 cells expressing or not DDB2, nuclei and F-actin filaments were stained with TO-PRO-3 and Phalloidin, respectively. (A) Cells were observed by confocal microcopy (magnification 400×). The 8 confocal z sections of cell expressing or not DDB2 (left), and the z axis projection (right) are presented. (B) The median value of maximum fluorescence intensity from 10 regions of interest, arbitrarily placed across the cortical actin-cytoskeleton, was calculated for each acquisition. Data are represented as the average value ± SEM from 3 independent experiments and are indicated statistically significant for * p < 0.05.
regulated, except for ICAM1 and SELL which were significantly upregulated in MDA DDB2+ cells. Concerning the analysis of MCF7 cells, 20 out of the 84 genes analyzed were modulated
This journal is © The Royal Society of Chemistry 2016
when DDB2 expression was inhibited, and only 6 were downregulated in MCF7 DDB2+ compared to MCF7 DDB2− cells (Table 2A and B). Probably due to the different phenotype
Nanoscale
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Paper Table 2 Changes in the adhesion-related gene expression according to DDB2 expression. A PCR Array targeting 84 adhesion and extracellular matrix related genes was performed, using total RNA extracted from cells. Quantification of each gene expression level was based on cycle threshold and a fold change between MDA-MB231 (A) and MCF7 (B) DDB2+ and DDB2− cells. Genes are downregulated or upregulated, when the ratio is lower or higher than 1, respectively in DDB2 overexpressing cells, in contrast to the DDB2− cells. Fold changes are presented as a mean value ± SEM of three independent experiments. Differences between DDB2+ and DDB2− cells for differentially expressed genes are statistically significant for p < 0.05, p < 0.01, and p < 0.001
between the epithelial MCF7 cells and mesenchymal MDA cells, only 3 common genes, ICAM1 (intracellular adhesion molecule 1), CTGF (connective tissue growth factor) and TGFβ1 (transforming growth factor beta 1), were modulated similarly in both cell lines, suggesting that DDB2 would modulate adhesion-related gene expression in a cell specific manner. CTGF and TGFβ1 genes are known to be broadly implicated in adhesion as well as invasion and migration of tumor cells.
Nanoscale
Nanoscale
Indeed, CTGF expression seems to be associated with a poor prognosis in breast cancers and its overexpression has been linked to an increase of migratory properties of breast cancer cells.38–40 Furthermore, TGFβ1 is a potent secreted factor that drives cancer progression. While it is known to act as tumor suppressor by inhibiting epithelial cell proliferation, it plays a key role in breast tumor progression toward metastasis by inducing EMT and invasive properties in vitro.41–43 Both TGFβ1 and CTGF have been shown to be associated with an increase of bone and pulmonary metastasis formation in vivo.39,44 CTGF being a TGFβ1 responsive gene, we may speculate that the downregulation of this gene is related to DDB2-dependent TGFβ1 gene downregulation in breast cancer DDB2+ cells.45 However, even if growing evidence show that DDB2 is acting as a transcriptional regulator, its direct involvement in regulation of such adhesion-related gene expression needs to be further investigated. The modulation of cellular adhesive force and adhesion properties in cells expressing or not DDB2 could also be explained by the modification of membrane proteins and other relative amounts of adhesive compounds as shown by Cross et al.46 and Pi et al.47 It is the case of ICAM1, a protein which is also known to be largely involved in tumor cell adhesion, even if the role of this gene in cancer cells is quite controversial. It has been shown that the downregulation of ICAM1 could be associated with a decrease of breast cancer invasion.48 However, other studies have reported an opposite effect of ICAM1 in tumorigenesis, by associating its expression with a relatively good prognosis in breast cancers.49,50 They have shown that a decrease in the ICAM1 expression was associated with breast tumor cell dissemination, favoring development of metastases in order to escape from immune surveillance.51 Consistent with all these data, our results showing the downregulation of CTGF and TGFβ-1 expression and the upregulation of ICAM1 are in accordance with previous knowledge related to higher proliferation rates, an inhibition of the migratory and invasive properties as well as a better outcome of colon and breast cancers overexpressing DDB2.2,3,52 Our data highlighted that both DDB2+ cell lines expressed higher ICAM1 mRNA levels. Among the 31 genes encoded transmembrane proteins, only the gene encoded ICAM1 is modulated similarly in both cell tumor cell lines. To attest that DDB2 dependent adhesion properties, cell stiffness and adhesion force modifications could be due to differences in cellular membrane composition, the level of ICAM1 protein has been evaluated (Fig. 7). According to the transcript level evaluated by PCR array, ICAM1 protein is significantly upregulated in DDB2-expressing cells in contrast to their counterpart. Thus, these results could reflect in part a relation between changes in the cellular membrane composition and the DDB2dependent nanomechanical properties modifications. Moreover, the decrease in the cell stiffness observed in DDB2+ cells could be related to the downregulation of some other adhesion related genes in MDA DDB2+ cells such as those encoding the integrin isoforms ITGA2, ITGA6, ITGB4 and ITGA1 in MCF7 DDB2+ cells (Table 2). Integrins are trans-
This journal is © The Royal Society of Chemistry 2016
View Article Online
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
Nanoscale
Paper
colon and breast cancer via the regulation of EMT inducing factors and IκBα, the cytoplasmic inhibitor of NFκB transcription factor. Taken together, these data indicate a new role of DDB2 in adhesion and mechanical properties of cancer cells that are the early events occurring during tumor progression toward metastasis. Thus our results bring new insight that DDB2 could be a relevant clinical marker of breast cancer progression.
Acknowledgements Fig. 7 Increase of ICAM1 expression level in DDB2 expressing cells. Expression of ICAM1 protein was analyzed by Western blot with 50 µg of total protein from breast tumor cells. The density of each band was measured and compared with the density of tubulin bands used as loading control. Results are presented as the fold increase or decrease over control cells fixed to 1 and correspond to the mean value of three independent experiments ± SEM.
This work was supported by the CNRS and the University of Lorraine (PEPS Mirabelle Bioforce Project) and the “Ligue contre le cancer” (Comité de Meurthe et Moselle). The fellowship of Claire Barbieux was given by the French Research Ministry (contrat doctoral de l’Université de Lorraine). The authors thank Dr S. Kaminski for critical reading of the article.
References membrane proteins well known for their role as adhesion molecules by interacting with extracellular matrix but also by acting on actin cytoskeleton organization and transduction of intracellular signal regulating many cellular functions such as proliferation, migration and invasion.53 Integrins are linked to the actin cytoskeleton via a scaffold protein complex under the plasma membrane.54–56 Thus, the cytoskeletal reorganization observed in DDB2 expressing cells compared to their counterparts could be related to the decrease in the expression of integrins, as referred in Table 2.
4.
Conclusion
In the present work we showed that DDB2, a nuclear protein commonly known for its major role in DNA repair and more recently involved in the control of the invasive abilities of breast cancer cells, is associated with cell adhesion by modulating membrane mechanical properties. We observed upon microscale and nanoscale analysis by AFM, PCR-Array and fluorescence microscopy that expression of DDB2 modulates cancer cell stiffness and strongly impacts the cortical actincytoskeleton organization. Indeed, we demonstrated quantitatively that expression of DDB2 decreases cell adhesion, Young’s modulus of MCF7 and MDA-MB231 cells. These modifications could be related to modulation of specific adhesion related gene expression in DDB2 overexpressing cells as compared to DDB2− cells that could impact the global plasma membrane composition, in accordance with the new role of DDB2 as transcriptional regulator.1–4 Furthermore, these data indicate that nanomechanical properties of cancer cell measured by AFM could be useful to follow the changes occurring during tumorigenesis. It has been already shown that DDB2 plays a role in tumorigenesis by preventing cells to acquire invasive abilities in
This journal is © The Royal Society of Chemistry 2016
1 V. Minig, Z. Kattan, J. van Beeumen, E. Brunner and P. Becuwe, J. Biol. Chem., 2009, 284, 14165–14176. 2 M. Ennen, R. Klotz, N. Touche, S. Pinel, C. Barbieux, V. Besancenot, E. Brunner, D. Thiebaut, A. C. Jung, S. Ledrappier, L. Domenjoud, J. Abecassis, F. Plénat, S. Grandemange and P. Becuwe, Cancer Res., 2013, 73, 5040–5052. 3 N. Roy, P. V. Bommi, U. G. Bhat, S. Bhattacharjee, I. Elangovan, J. Li, K. C. Patra, D. Kopanja, A. Blunier, R. Benya, S. Bagchi and P. Raychaudhuri, Cancer Res., 2013, 73, 3771–3782. 4 R. Zhao, C. Han, E. Eisenhauer, J. Kroger, W. Zhao, J. Yu, K. Selvendiran, X. Liu, A. A. Wani and Q.-E. Wang, Mol. Cancer Res., 2014, 12, 370–380. 5 B. De Craene and G. Berx, Nat. Rev. Cancer, 2013, 13, 97–110. 6 P. Polyakov, C. Soussen, J. Duan, J. F. L. Duval, D. Brie and G. Francius, PLoS One, 2011, 6, e18887. 7 D. C. Lin, E. K. Dimitriadis and F. Horkay, J. Biomech. Eng., 2007, 129, 904–912. 8 L. Sirghi, J. Ponti, F. Broggi and F. Rossi, Eur. Biophys. J., 2008, 37, 935–945. 9 E. Canetta, A. Riches, E. Borger, S. Herrington, K. Dholakia and A. K. Adya, Acta Biomater., 2014, 10, 2043– 2055. 10 M. Lekka, K. Pogoda, J. Gostek, O. Klymenko, S. PrauznerBechcicki, J. Wiltowska-Zuber, J. Jaczewska, J. Lekki and Z. Stachura, Micron Oxf. Engl., 2012, 43, 1259–1266. 11 S. E. Cross, Y.-S. Jin, Q.-Y. Lu, J. Rao and J. K. Gimzewski, Nanotechnology, 2011, 22, 215101. 12 W. Xu, R. Mezencev, B. Kim, L. Wang, J. McDonald and T. Sulchek, PLoS One, 2012, 7, e46609. 13 M. Plodinec, M. Loparic, C. A. Monnier, E. C. Obermann, R. Zanetti-Dallenbach, P. Oertle, J. T. Hyotyla, U. Aebi,
Nanoscale
View Article Online
Paper
14
Published on 05 February 2016. Downloaded by University of California - San Diego on 16/02/2016 12:17:32.
15 16 17
18 19 20 21 22
23
24 25
26
27 28 29 30 31
32
33
M. Bentires-Alj, R. Y. H. Lim and C.-A. Schoenenberger, Nat. Nanotechnol., 2012, 7, 757–765. N. Gavara and R. S. Chadwick, Nat. Nanotechnol., 2012, 7, 733–736. W. F. Loomis, D. Fuller, E. Gutierrez, A. Groisman and W.-J. Rappel, PLoS One, 2012, 7, e42033. M. Sun, N. Northup, F. Marga, T. Huber, F. J. Byfield, I. Levitan and G. Forgacs, J. Cell Sci., 2007, 120, 2223–2231. O. G. Ramprasad, G. Srinivas, K. S. Rao, P. Joshi, J. P. Thiery, S. Dufour and G. Pande, Cell Motil. Cytoskeleton, 2007, 64, 199–216. Y. Shan and H. Wang, Chem. Soc. Rev., 2015, 44, 3617– 3638. D. J. Müller and Y. F. Dufrêne, Trends Cell Biol., 2011, 21, 461–469. S. Suresh, Acta Biomater., 2007, 3, 413–438. C. Roduit, S. Sekatski, G. Dietler, S. Catsicas, F. Lafont and S. Kasas, Biophys. J., 2009, 97, 674–677. C. Roduit, F. G. van der Goot, P. De Los Rios, A. Yersin, P. Steiner, G. Dietler, S. Catsicas, F. Lafont and S. Kasas, Biophys. J., 2008, 94, 1521–1532. D. F. Tschaharganeh, W. Xue, D. F. Calvisi, M. Evert, T. V. Michurina, L. E. Dow, A. Banito, S. F. Katz, E. R. Kastenhuber, S. Weissmueller, C.-H. Huang, A. Lechel, J. B. Andersen, D. Capper, L. Zender, T. Longerich, G. Enikolopov and S. W. Lowe, Cell, 2014, 158, 579–592. F. Guilak, J. R. Tedrow and R. Burgkart, Biochem. Biophys. Res. Commun., 2000, 269, 781–786. T. G. Kuznetsova, M. N. Starodubtseva, N. I. Yegorenkov, S. A. Chizhik and R. I. Zhdanov, Micron Oxf. Engl., 2007, 38, 824–833. M. Lekka, D. Gil, K. Pogoda, J. Dulińska-Litewka, R. Jach, J. Gostek, O. Klymenko, S. Prauzner-Bechcicki, Z. Stachura, J. Wiltowska-Zuber, K. Okoń and P. Laidler, Arch. Biochem. Biophys., 2012, 518, 151–156. S. E. Cross, Y.-S. Jin, J. Rao and J. K. Gimzewski, Nat. Nanotechnol., 2007, 2, 780–783. E. L. Baker, J. Lu, D. Yu, R. T. Bonnecaze and M. H. Zaman, Biophys. J., 2010, 99, 2048–2057. E. A. Corbin, F. Kong, C. T. Lim, W. P. King and R. Bashir, Lab Chip, 2015, 15, 839–847. Y. Teng, J. Qiu, Y. Zheng, X. Luo, L. Zhang, L. Chen and G. Wang, J. Med. Biol. Eng., 2012, 62–68. L.-S. Z. Rathje, N. Nordgren, T. Pettersson, D. Rönnlund, J. Widengren, P. Aspenström and A. K. B. Gad, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 1515–1520. L. I. Volakis, R. Li, W. E. Ackerman, C. Mihai, M. Bechel, T. L. Summerfield, C. S. Ahn, H. M. Powell, R. Zielinski, T. J. Rosol, S. N. Ghadiali and D. A. Kniss, PLoS One, 2014, 9, e86110. J. M. Zarzynska, Mediators Inflammation, 2014, 2014, e141747.
Nanoscale
Nanoscale
34 J. Massagué, Nat. Rev. Mol. Cell Biol., 2012, 13, 616–630. 35 T.-H. Wu, Y.-W. Chou, P.-H. Chiu, M.-J. Tang, C.-W. Hu and M.-L. Yeh, Cancer Cell Int., 2014, 14, 20. 36 A. M. Collinsworth, S. Zhang, W. E. Kraus and G. A. Truskey, Am. J. Physiol.: Cell Physiol., 2002, 283, C1219–C1227. 37 S. T. Buckley, C. Medina, A. M. Davies and C. Ehrhardt, Nanomed. Nanotechnol. Biol. Med., 2012, 8, 355–364. 38 P.-S. Chen, M.-Y. Wang, S.-N. Wu, J.-L. Su, C.-C. Hong, S.-E. Chuang, M.-W. Chen, K.-T. Hua, Y.-L. Wu, S.-T. Cha, M. S. Babu, C.-N. Chen, P.-H. Lee, K.-J. Chang and M.-L. Kuo, J. Cell Sci., 2007, 120, 2053–2065. 39 W. Chien, J. O’Kelly, D. Lu, A. Leiter, J. Sohn, D. Yin, B. Karlan, J. Vadgama, K. M. Lyons and H. P. Koeffler, Int. J. Oncol., 2011, 38, 1741–1747. 40 D. Xie, K. Nakachi, H. Wang, R. Elashoff and H. P. Koeffler, Cancer Res., 2001, 61, 8917–8923. 41 L. Kubiczkova, L. Sedlarikova, R. Hajek and S. Sevcikova, J. Transl. Med., 2012, 10, 183. 42 M. H. Barcellos-Hoff and R. J. Akhurst, Breast Cancer Res., 2009, 11, 202. 43 D. Padua and J. Massagué, Cell Res., 2009, 19, 89–102. 44 A. A. Mourskaia, Z. Dong, S. Ng, M. Banville, J. C. Zwaagstra, M. D. O’Connor-McCourt and P. M. Siegel, Oncogene, 2009, 28, 1005–1015. 45 G. R. Grotendorst, H. Okochi and N. Hayashi, Cell Growth Differ. Mol. Biol. J. Am. Assoc. Cancer Res., 1996, 7, 469–480. 46 S. E. Cross, Y.-S. Jin, J. Tondre, R. Wong, J. Rao and J. K. Gimzewski, Nanotechnology, 2008, 19, 384003. 47 J. Pi, T. Li, J. Liu, X. Su, R. Wang, F. Yang, H. Bai, H. Jin and J. Cai, Micron Oxf. Engl., 2014, 65, 1–9. 48 C. Rosette, R. B. Roth, P. Oeth, A. Braun, S. Kammerer, J. Ekblom and M. F. Denissenko, Carcinogenesis, 2005, 26, 943–950. 49 Y. Ogawa, K. Hirakawa, B. Nakata, T. Fujihara, T. Sawada, Y. Kato, K. Yoshikawa and M. Sowa, Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res., 1998, 4, 31–36. 50 A. C. Donadio, M. M. Remedi, S. Frede, G. R. Bonacci, G. A. Chiabrando and M. C. Pistoresi-Palencia, Clin. Exp. Metastasis, 2002, 19, 437–444. 51 S. Mukai, H. Kagamu, S. Shu and G. E. Plautz, Cell. Immunol., 1999, 192, 122–132. 52 Z. Kattan, S. Marchal, E. Brunner, C. Ramacci, A. Leroux, J. L. Merlin, L. Domenjoud, M. Dauça and P. Becuwe, PLoS One, 2008, 3, e2002. 53 A. Huttenlocher and A. R. Horwitz, Cold Spring Harbor Perspect. Biol., 2011, 3, a005074. 54 K. A. DeMali, K. Wennerberg and K. Burridge, Curr. Opin. Cell Biol., 2003, 15, 572–582. 55 I. Delon and N. H. Brown, Curr. Opin. Cell Biol., 2007, 19, 43–50. 56 C. G. Galbraith, K. M. Yamada and J. A. Galbraith, Science, 2007, 315, 992–995.
This journal is © The Royal Society of Chemistry 2016
