Integrin-β1 polyclonal antibody was obtained from Cell Signaling Technology, Inc. (Danvers, MA, USA); antibody for hyaluronan receptor (clone G44–26) was from BD Biosciences (San Jose, CA, USA); antibody for neuronal cell adhesion molecule (Nr-CAM) (clone A27), marimastat and matrix metalloproteinase (MMP) inhibitor II were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA); non-specific antibodies were obtained from Rockland Immunochemicals, Inc. (phycoerythrin-conjugated) (Gilbertsville, PA, USA), and Santa Cruz Biotechnology, Inc.

The cell cultures were maintained at 37°C in a humidified atmosphere containing 5% CO₂. U87MG cells (American Type Culture Collection, Manassas, VA, USA) were maintained in minimum essential medium (Mediatech, Herndon, VA, USA) supplemented with 10% fetal bovine serum (Atlanta Biologicals, Inc., Flowery Branch, GA, USA), 2 mM L-glutamine (HyClone, Logan, UT, USA), 50 U/ml penicillin and 50 μg/ml streptomycin (both from Lonza, Rockland, ME, USA). Neural progenitor cells from newborn C57BL/6 mice (Harlan Laboratories, Indianapolis, IN, USA) were obtained, after euthanasia following a protocol approved by the Animal Care and Use Committee of our Institution, through incubation of sections of the forebrain in neurobasal medium supplemented with B27 (both from Invitrogen Life Technologies, Carlsbad, CA, USA), 20 ng/ml epidermal growth factor, 10 ng/ml basic fibroblast growth factor (both from R&D Systems, Minneapolis, MN, USA), 2 mM L-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin. Fluorescent marker-expressing cells, obtained by clonal expansion (U87MG cells) or fluorescence-activated cell sorting (neural progenitor cells) after transduction with retroviral vectors as previously described (10), were co-cultured in neurobasal medium. Tissue culture trypsinization was carried out with Hank’s balanced salt solution containing 0.05% trypsin (Mediatech), and terminated by the addition of serum-supplemented culture medium (U87MG cells) or by dilution (neural progenitor cells).

Stably transduced reporter cells were obtained by cell culture incubation in medium containing lentiviral particles with luciferase expression vector (Cignal Finder Cancer Reporter Array; SABiosciences, Frederick, MD, USA), and the selection of the population of transduced cells through the addition of puromycin (0.5 μg/ml; Sigma-Aldrich, St. Louis, MO, USA).

Luciferase expression by cells transduced with reporter vector, containing either non-inducible TATA-box promoter or TATA-box promoter joined to one of 10 different transcriptional-response elements, was quantified in 96-well plate adherent cultures. For this purpose, a 1:1 mixture of ONE-Glo™ Luciferase reagent (Promega Corp., Madison, WI, USA) and cell culture medium supplemented with 2% fetal bovine serum was added to the cells (50 μl/cell culture) and, 10 min later, the luminescence of the resulting cell lysate was measured in a Turner Biosystems 20/20 luminometer (Turner Biosystems, Inc., Sunnyvale, CA, USA).

In a series consisting of 5 individual cultures of lentiviral vector-transduced cells set up with cell densities decreasing, from 59,200 cells/cm², by a 2-fold serial dilution factor, luciferase reporter gene expression was determined after 48 h from cell seeding, at which time cell confluency was expected to be 64, 32, 16, 8 and 4% (control) (based on 34 h doubling time and 250,000 cells/cm² in a confluent culture). The fold change in cell expression of reporter was determined by dividing the value of luminescence by the average value of the corresponding control cultures (4% cell confluency) and by the corresponding dilution. A statistical threshold for transcriptional activation was established from all the average ratios obtained as the median ± SD. Average ratios above the threshold were considered indicative of promoter activation.

Total protein amount per cell culture was estimated using the BCA protein assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA), after rinsing each culture with phosphate-buffered saline (PBS) and lysing the cells with cold lysis buffer (pH 7.4) containing 20 mM Tris, 0.5% Triton X-100, 150 mM NaCl and protease inhibitor cocktail (1:100; EMD Millipore, Billerica, MA, USA).

Untreated polystyrene plates were coated with the adhesion protein, retronectin (Takara, Otsu, Japan). Retronectin was added at different dilutions in PBS to non-tissue culture-treated plates (BD Biosciences), removing the solutions after 2 h of coating at room temperature. The coated surfaces were rinsed with PBS before seeding the cells. The retronectin coating concentrations were determined as indicated by the manufacturer.

Cell DNA content was quantified by measuring the emission intensity of fluorescent dye-stained DNA in a flow cytometer (Guava EasyCyte™ Plus System; EMD Millipore). For this purpose, the cell culture was trypsinized and the cell suspension passed through a cell strainer (70 μm) (BD Biosciences). After blocking trypsinization, 2×10⁴ cells were centrifuged and resuspended in 0.5 ml of PBS. Using a fine tip pipette, the suspension was drawn up and expelled repeatedly. Subsequently, upon the addition of a 10-fold larger volume of buffer, the cells were centrifuged and resuspended in 200 μl of PBS. The suspension was added drop by drop to 10 ml of ice-cold 70% ethanol for cell fixation. The fixed cells were centrifuged at high speed (450 × g), resuspended in PBS containing 1% bovine serum albumin, re-centrifuged and resuspended in 200 μl of Guava cell cycle reagent (EMD Millipore). The distribution of propidium iodide-stained cells was determined using Guava CytoSoft 5.3 software (Guava Technologies, Inc., Hayward, CA, USA).

The fraction of binucleated cells over the total cells was quantified in 3-day cultures on a retronectin-coated surface (seeding density, 15,000 cells/cm²). Each culture was trypsinized and re-seeded in a regular culture well and, soon after attachment, the cells were fixed with 4% p-formaldehyde and stained with 4′,6-diamidino-2-phenylindole (1:1,000; Sigma-Aldrich). Binucleated cells were counted in randomly selected areas using an Olympus CKX41 fluorescence microscope (Olympus, Center Valley, PA, USA).

The cell surface expression of hyaluronan receptor was quantified by flow cytometric analysis. To this aim, 2×10⁴ cells/sample were resuspended in PBS containing 0.1% bovine serum albumin and, then, half of the cells were incubated for 30 min at 4°C with phycoerythrin-conjugated antibody for hyaluronan receptor and the other half with phycoerythrin-conjugated non-specific antibody (antibody concentration, 0.5 μg/ml). Following incubation with antibody, the cells were washed twice and resuspended in 200 μl of PBS for the analysis of fluorescence intensity in a Guava flow cytometer.

The Student’s t-test was used to determine p-values. p<0.05 was considered to indicate a statistically significant result.

The assessment of transcription of vectors of a pathway reporter array in 2-fold serial dilution cultures revealed both the activation of Elk-1/SRF- and RBPJ-driven transcription in the cell cultures at 32% confluency and further enhanced transcription in the cell cultures at 64% confluency (Fig. 1), indicating that intercellular adhesion elicited signal transduction through either extracellular signal-regulated kinase (ERK), Notch or both pathways. Given that cell adhesion-induced ERK activation is determined by integrin binding to protein ligands (11), we subsequently assessed cell density-dependent transcriptional activation in cultures containing integrin-β1 antibody. Antibody targeting of integrin-β1 hindered the induction of both ERK and Notch target transcription (Table I), suggesting that integrin-mediated signal transduction activated ERK and that, concomitantly, there was either a crosstalk between the ERK and Notch signaling pathways or an interaction between the integrin and Notch receptors (12). Since Nr-CAM is a binding partner of integrins in cell-cell adhesion (13), we also determined cell density-dependent transcriptional activation in cultures containing a specific antibody for that adhesion receptor. This corroborated that signaling decreased (Table I) upon the antibody targeting of a ligand of integrins at the cell surface. Conversely, there was signaling activation in the cultures containing non-specific antibody or blocking antibody for hyaluronan receptor (Table I), a mediator of cell-cell adhesion that has been related to cell fusion (7,14), which indicated that cell confluence and signal transduction were not impaired by the side-effects of the antibody. In summary, these results revealed that intercellular integrin binding elicited adhesion signaling in U87MG cells that activated both ERK and Notch target transcription.

We assessed the level of both ERK and Notch target transcription following integrin-based cell adhesion to surfaces coated with retronectin, a recombinant adhesion protein with integrin binding sites of fibronectin, which is a pericellular matrix component robustly expressed by U87MG cells (15). Cell attachment to coating at a concentration of 0.5 μg/cm² involved transcriptional activation (Table II), providing the induction of both ERK and Notch target transcription. We then determined the rate of proliferation and the percentage of cells with a 4-fold increase in the amount of cellular DNA, corresponding to the G2/M peak of the cells with duplicated tumor genome (9), of cultures on coating that promoted integrin adhesion signaling. The fold increase in the number of cells in the 24–48 h interval from cell seeding, 2.8±0.9, was almost equal to that of the control cultures (15 μg retronectin/cm²), 2.9±0.7 (means ± SD of 3 different experiments). The relative number of cells with duplicated tumor genome, by contrast, increased significantly (Fig. 2). In agreement with this result, the cultures contained a markedly enhanced proportion of binucleated cells (Table III). This linked the transcriptional activation induced by integrin adhesion to an increased rate of cell polyploidization. Hence, we quantified the cells with a duplicated tumor genome in regular cultures containing antibody that prevented signaling activation. The cultures containing antibody for hyaluronan receptor, which prevented the polyploidization of breast carcinoma and melanoma cells, showed an unaltered frequency of polyploidization (Table IV). By contrast, the cultures containing integrin-β1 antibody showed a reduced frequency (Table V). These findings indicated that the induction of both ERK and Notch target transcription resulted in tumor-genome duplication.

Of note, the U87MG cells on coating that promoted integrin adhesion signaling extended thin protruding adhesive structures resembling matrix-degrading invadopodia (Fig. 3), which suggested that the induction of genome duplication was mediated by intracellular invasion. Hence, we quantified the cells with a duplicated tumor genome in the regular cell cultures containing specific inhibitors of cancer cell invasion-associated MMPs. Cultures containing MMP inhibitors had markedly low percentages of polyploid cells (Table VI). Since these findings indicated that adhesion-stimulated genome duplication was mediated by protease-driven tumor invasion, we wished to investigate whether intracellular invasion by U87MG cells resulted in the merging of cell contents. In co-cultures in non-coated dishes of U87MG cells expressing red fluorescent protein and non-tumor neural cells expressing green fluorescent protein, intracellular fluorescence was observed that revealed that the cytoplasmic content of tumor cells was released through fusion pores. Representative images are presented in Fig. 4. These observations revealed that intracellular invasion by U87MG cells resulted in the merging of cell contents, further suggesting that tumor invasion mediated the increases in DNA content.