The basic principle of 3D cell cultures is to provide mammary carcinoma cells with properties that recapitulate in-vivo tumor microenvironment. The 3D cellular structures were characterized by establishment of adhesion complexes, tissue polarity, cytoskeletal structures, and cell volume that is significantly different from those found in traditional monolayer culture 2442.

Alteration in the morphology of SUM149 grown on monolayer versus those grown on 3D models were observed after 24 h, 48 h and 72 h. SUM149 cells grown on monolayer adhere to plastic surface after 24 h (Figure 1A, left panel). SUM149 cells form projections towards each other after 48 h (Figure 1A, middle panel) and divide to 70-80% confluence after 72 h (Figure 1A, right panel). On the other hand, SUM149 cells grown on 3D model became connected together after 24 h (Figure 1B, left panel) and formed aggregates after 48 h (Figure 1B, middle panel). Clumps of cells simulate in-vivo emboli like structures were observed after 72 h (Figure 1B, right panel). The morphology acquired by SUM149 cells when seeded in-vitro in 3D culture model resembles in vivo tumor emboli of IBC patients (Figure 1C) where IBC carcinoma cells adhered together forming clusters surrounded by endothelial cells and basement membrane 1011. The adherence of IBC cells within lymphatic vessels forming tumor emboli was assumed to be attributed to over-expression of E-cadherin 8.

Using cytokine antibody array analysis cytokines, chemokines and growth factors secreted by U937 were identified in U937-CM. Cytokine antibody array detected difference in density values of the assessed cytokines, chemokines and growth factors (Figure 2A). Densitometry analysis performed with ImageJ software showed that cytokines, chemokines and growth factors of low density value (0-4000) were: IL-7, IL-3, MCP-2, GCSF, GM-CSF, IL-17, IL-6, IL-12p70, MIG, ICAM-1, IL-2, IL-16, Eotaxin, IL-4, IL-13, IL-15, IL-1β, IL-11, TGF-β, MIP-1δ, IFN-γ, sTNFRI, I-309, TNf-β, IL-12p40, IL-1α, PDGF-BB, Eotaxin-2, MIP-1α, IP-10, and IL-10, respectively. Cytokines, chemokines and growth factors of moderate density values (4001-8000) were: IL-6sR, M-CSF, sTNFRII, MIP-1β, TNF-α, RANTES and TIMP-2, respectively, while cytokines and chemokine of high density value (> 8000) were IL-8 and MCP-1 (CCL2) (Figure 2B).

Results revealed that U937-CM induced the migration and formation of branched like structures by SUM149 cells; which was consistent with previous results of the author and colleagues 41. The level of expression of E-cadherin by SUM149 cells grown in U937-CM, for 48 h at 37°C in humidified CO₂ incubator, in comparison to control ones grown in complete Ham's F-12 media under same conditions was measured by immunoblot. The expression of E-cadherin was not altered by control SUM149 cells grown in complete Ham's F-12 media compared to SUM149 cells grown in U937-CM (Figure 3A). Immunocytochemistry confirmed immunoblot results which showed no difference in the level of expression of E-cadherin by control SUM149 cells grown in complete Ham's F-12 media (Figure 3B) when compared to SUM149 cells grown in U937-CM (Figure 3C).

On the other hand immunoblot analysis showed that the adhesion molecule fibronectin, which is responsible for cell-cell and cell-matrix interaction was weakly expressed by control SUM149 cells grown in complete Ham's F-12 media and over-expressed by SUM149 cells grown in U937-CM for 48 h in 3D culture model (Figure 4A). Similarly, immunocytochemical staining of fibronectin revealed a weak expression of cellular and membranous fibronectin by control SUM149 cells grown in complete Ham's F-12 media (Figure 4B) compared to SUM149 cells grown in U937-CM for 48 h in 3D culture model at 37°C in CO₂ incubator (Figure 4C).

Treatment of SUM149 cells with recombinant MCP-1 and IL-8 induced the expression of fibronectin by SUM149 cells as detected by immunoblot analysis (Figure 5A). Since, PI3K/Akt signaling molecules found to be involved in fibronectin expression 1920 the level of expression of PI3K, Akt and p-Akt in SUM149 cells treated with recombinant IL-8 and MCP-1 were measured. Immunoblot results showed that IL-8 induced the expression of PI3K-p85, total Akt and p-Akt by SUM149 cells. On the contrary, MCP-1 did not alter the level of expression of the PI3K/Akt signaling molecules by SUM149 cells (Figure 5A).

Since the physiological effects of IL-8 are mediated by its specific receptors CXCR1 and CXCR2, the expression of CXCR1 and CXCR2 by SUM149 cells were assessed by immunoblot. Results revealed that SUM149 cells express CXCR1 and CXCR2 regardless the treatment of MCP-1 and IL-8 (Figure 5B). The present results are consistent with other studies which showed that IL-8 specific receptors CXCR1 and CXCR2 may regulate fibronectin expression through PI3k/Akt pathway 4043.

IBC cell line SUM149 was a gift from Dr. Stephen P. Ethier, Barbara Ann Karmanos Cancer Institute, Wayne State University, Detroit, MI, USA. U937 human monocytic cell lines [American Type Culture Collection (ATCC, Manassas, USA], anti-CXCR2 and anti-CXCR2 polyclonal antibodies (From Thermo Scientific Pierce Antibodies, IL. USA) were a gifts from Dr. Bonnie Sloane's lab, Department of Pharmacology, Wayne State University, Detroit, MI, 48201, USA. Ham's F-12 media was purchased from Mediatech (Manassas, VA, USA). Monoclonal mouse E-cadherin antibody and Dispase was purchased from BD Biosciences, Inc. (San Diego, CA, USA). Monoclonal anti-fibronectin antibody was purchased from Sigma-Aldrich (St. Louis, MO, USA) monoclonal anti-human fibronectin Antibody from R&D Systems, Inc. (Minneapolis, MN, USA). PI3 Kinase p85 (PI3K-p85) Antibody, Akt Antibody and phosphorylated Akt (p-Akt) were purchased from cell signaling technology (Danvers, MA, USA). Fluorescein-conjugated affinity-purified donkey anti-rabbit IgG and normal donkey serum were purchased from Jackson ImmunoResearch (West Grove, PA, USA). ChemiArray™ Human Inflammation Antibody Array III kit was from Chemicon (Temecula, CA, USA). Cultrex^® Basement Membrane Extract (BME) was from Trevigen (Gaithersberg, MD, USA). Recombinant Human CCL2/JE/MCP1 and Recombinant Human CXCL8/IL8 form R&D Systems (Minneapolis, USA) were kind gift from Dr. Robert J. Schneider's lab, Department of Microbiology, New York University, New York, NY, 10016, USA. Acrylamide and nitrocellulose membranes were from BioRad (Hercules, CA, USA). Horseradish peroxidase-labeled goat anti-mouse IgG and Membrane peroxidase substrate 3, 3', 5, 5' -Tetramethylbenzidine (TMB) were purchased from Kirkegaard and Perry Laboratories (KPL), Inc. (Gaithersburg, MD, USA). Bradford Reagent for determining protein concentration was from Sigma (Sigma-Aldrich, Germany). ProSieve^® Color Protein Marker was purchased from Lonza (Cologne GmbH, Germany).

For monolayer culture assay, SUM149 cells at density of 250 × 10³/ml were seeded on plastic Petri dishes in Ham's F-12 complete media and examined after 24 h, 48 h and 72 h by Zeiss Axiovert microscope (Carl Zeiss AG, Germany). To examine cell morphological changes on 3D culture, 30 mm Petri dishes were coated with 200 μl basement membrane extract BME (Culterx^®) incubated at 37°C in CO2 incubator for 15 min till solidification. Than SUM149 cells suspension at density of 250 × 10³/ml was mixed with 2% BME overlaid on coated Petri dishes as described before 41. Cells on 3D were examined microscopically after 24 h, 48 h and 72 h and morphological changes were inspected.

Media conditioned by U937 cells secreted cytokines, chemokines and growth factors (designated as U937-CM) used for cytokine profiling and as conditioned media for seeding of SUM149 were prepared as was described before 53. Briefly, U937 cells were seeded at density of 2.5 × 10⁵ cells/ml in Ham's F12 complete media into T-75 tissue culture flasks and grown to approximately 80% confluency. Cells were collected washed twice with phosphate-buffered saline (PBS) and incubated overnight in serum-free medium. Overnight condition media were collected and centrifuged at 700 g at room temperature for 5 min to pellet cells. The supernatant was collected and centrifuged at 2000 g at 4°C for 10 min to remove cell debris. The supernatant representing U937-CM was divided into two aliquots one subjected for cytokine profiling and the second aliquot was 5-fold concentrated using AmiconUltracell10K filters (Millipore, Billerica, MA) for conditioning SUM149 culture media.

One ml of serum free U937-CM was profiled using a ChemiArray™ Human Inflammation Antibody Array III kit that detects 40 different cytokines, chemokines and growth factors. Method was conducted as was described before 53. Washing buffers and antibody dilutions were provided with the kit and prepared as the kit instruction guidelines. Antibody-array membrane was placed in special tray provided with the kit followed by addition of 1 ml blocking buffer for 30 min at room temperature on a horizontal shaker followed by washing. The membrane was incubated with 1 ml U937-CM and placed overnight at 4°C. After washing, the membrane was incubated with primary antibody biotin-conjugated anti-cytokines at room temperature for 2 h while shaking. Washing and incubation with HRP-conjugated streptavidin at room temperature was carried out for 2 h. After a final wash, the membrane signals were detected by adding chemiluminescence detection reagent provided with the kit and the membrane was exposed to X-ray film for different time intervals. Signal intensities representing fold differences between detected cytokines were analyzed by imageJ software (National Institutes of Health, MD, USA) as described elsewhere 54. Relative cytokines, chemokines and growth factors levels were calculated by subtracting the background staining and normalizing density value of each cytokine, chemokine and growth factor spot to the positive control spots on the same membrane. To compare relative U937-CM detected cytokines, chemokines and growth factors the density value detection limit was divided into three ranges as follows: 0-4000 low secreted cytokines; 4001-8000 moderately secreted cytokines and above 8000 were considered as highly secreted cytokines.

Concentrated U937-CM prepared as described above was diluted 1:5 with Ham's F12 complete culture medium containing 5% FBS then used as culture media for SUM149 cells.

To prepare 3D culture model, 30 mm Petri dishes was coated with 200 μl BME. Coated Petri dishes were incubated at 37°C in CO2 incubator for 15 min to solidify. SUM149 cell suspension at density of 250 × 10³/ml was mixed with 2% BME seeded as overlay culture over solidified BME and left at CO₂ incubator for 10-15 min till cells stick as we described before 41. Complete Ham's F-12 media were added and cells were incubated at 37°C in CO₂ incubator. After 72 h, when SUM149 form spheroid like structures, Ham's F-12 complete media containing 5% FBS was changed in the control Petri dishes and replaced by media conditioned by U937-CM in the tested Petri dishes. Culture media were discarded after 48 h and SUM149 cells were washed several times with PBS then serum free media were added and cells were incubated at 37°C in CO₂ incubator overnight.

Serums starved SUM149 cells grown as control and in U937-CM were collected from BME by incubating them in Dispase solution for 2 h at 37°C in humidified CO₂ incubator to dissolve BME. Then the cell suspension was transferred into a tube, mixed up and down several times, followed by addition of EDTA to stop the action of Dispase. Cells were centrifuged at 700 g and the pellet was washed several times and lysed in lysis buffer. Cell lysates were sonicated on ice in a 50 W Ultrasonicator five times for five seconds.

SDS-PAGE was conducted as described elsewhere 53. Briefly, cell lysates were reduced and separated by SDS-PAGE (12% acrylamide). Loading was performed at a concentration of 30 μg protein per well. Gels were then transferred onto nitrocellulose membranes. Immunoblot analysis was performed with primary antibodies against each of E-cadherin (1:2500), fibronectin (1:1000) and β-actin (1:1000). Followed by washing, and the addition of secondary antibody conjugated with horseradish peroxidase (1:10,000) in Tris-buffered saline wash buffer (20 mM Tris, pH 7.5, 0.5 M NaCl) containing 0.5% Tween 20 and 5% (w/v) non-fat dry milk. After washing, bounded antibodies were detected by adding TMB chromagen/substrate solution. Once the color was developed the reaction was stopped by immersing the membrane into water for 20-30 seconds.

To localize and test expression of adhesion molecules E-cadherin and fibronectin by control SUM149 cells cultured in complete culture media in comparison to SUM149 cells cultured in U937-CM, cells were seeded in 24 well plates at density of 30 × 10³ cell/well. Each well was lined with a cover slip coated with 30 μl BME. After 72 h, cells were washed and media conditioned by U937 secretions were added. Ham's F-12 complete media was added to the parallel wells as controls. After 48 h the cells were washed with PBS (37°C), and then incubated overnight in serum free media. For immunocytochemical staining, cells were fixed with -20°C methanol for 5-10 min, at room temperature. Cells were permeabilized with 2% saponin in PBS, blocking was achieved by incubation with 0.5% bovine serum albumin in PBS. E-cadherin and fibronectin staining was carried out as previously described 55. Primary antibodies used were rabbit anti-human E-cadherin (1:50) and fibronectin (1:50). Texas-Red-conjugated affinity-purified donkey anti-mouse IgG (20 mg ml^-1) was used as secondary antibody. Negative control for immunocytochemistry was run similarly but in the absence of primary antibodies. Coverslips were mounted upside-down with antifade reagent on super frost slides. Flouresence labeled cells were detected by Zeiss LSM 510 confocal microscope (Carl Zeiss, Thornwood, NY, USA).

To study the role of individual cytokines MCP-1 and IL-8 on fibronectin production and relevant downstream signaling pathway (PI3k/Akt/p-Akt) SUM149 cells were seeded in monolayer as described above. After 72 h, when SUM149 form 70-80% confluence complete culture media was replaced by serum free media containing different concentrations of each of recombinant MCP-1 and IL-8 (50, 100, 150 and 200 ηg/ml). The results demonstrated that different concentrations of IL-8 and MCP-1 had similar effect on the production of fibronectin by SUM149 cells (data not shown). Thus, a dose of 200 ηg/ml of recombinant IL-8 and MCP-1 was selected as appropriate dose for treatment. SUM149 cells were grown in BME in Ham's F12 complete medium containing 5% FBS as described before. After 72 h, when SUM149 form spheroid like structures, culture media was discarded, cells were washed twice with PBS, cultured for 4 h in serum-free media with each of recombinant IL-8 and MCP-1 (200 ηg/ml). Cells were collected and prepared for SDS-PAGE and immunoblot analysis of fibronectin, PI3k/Akt/p-Akt and specific IL-8 receptors (CXCR1 and CXCR2) as described above.