Spheroid Model of Mammary Tumor Cells: Epithelial-Mesenchymal Transition and Doxorubicin Response

Abstract

Breast cancer is the most prevalent cancer among women worldwide. Therapeutic strategies to control tumors and metastasis are still challenging. Three-dimensional (3D) spheroid-type systems more accurately replicate the features of tumors in vivo, working as a better platform for performing therapeutic response analysis. This work aimed to characterize the epithelial-mesenchymal transition and doxorubicin (dox) response in a mammary tumor spheroid (MTS) model. We evaluated the doxorubicin treatment effect on MCF-7 spheroid diameter, cell viability, death, migration and proteins involved in the epithelial-mesenchymal transition (EMT) process. Spheroids were also produced from tumors formed from 4T1 and 67NR cell lines. MTSs mimicked avascular tumor characteristics, exhibited adherens junction proteins and independently produced their own extracellular matrix. Our spheroid model supports the 3D culturing of cells isolated from mice mammary tumors. Through the migration assay, we verified a reduction in E-cadherin expression and an increase in vimentin expression as the cells became more distant from spheroids. Dox promoted cytotoxicity in MTSs and inhibited cell migration and the EMT process. These results suggest, for the first time, that this model reproduces aspects of the EMT process and describes the potential of dox in inhibiting the metastatic process, which can be further explored.

Keywords: breast cancer; cell migration; doxorubicin; epithelial–mesenchymal transition; spheroids; three-dimensional cell culture.

Figure 1

Production of spheroids using MCF-7 cell line. Representative images showing growth and morphology of spheroids with different densities for 10 days by phase-contrast microscopy. Differences in spheroid size were observed depending on the initial number of cells introduced into the wells (A). Quantitative data are shown for area and diameter, respectively; all spheroids, except the one with an initial inoculum of 25,000 cells, grew over time, with prominent growth between 5 and 7 days of culture (B,C). Data are expressed as mean ± SD of one experiment (n = 10). ** p < 0.01; *** p < 0.001. One-way ANOVA test and Bonferroni post hoc test were used.

Figure 2

Mammary tumor spheroid mimics characteristics of avascular tumors. PI staining showing a necrotic core in spheroids with distinct densities: 3125 (A), 6250 (B), 12,500 (C) or 25,000 initial cells (D). Giemsa staining showing a necrotic center and a large spheroid (6250) (E). Immunofluorescence showing Ki-67 (F), E-cadherin (E-cad) (G), and laminin (LN) (H) expression, in green. DNA staining with DAPI can be observed in blue. Ki-67 expression was most prominent at the periphery of the spheroid (F). Spheroids presented immunoreactivity for E-cadherin, a well-known epithelial marker, and were able to produce their own matrix, revealed by laminin staining, respectively (G,H).

Figure 3

Production of spheroids from mice tumors. Phase-contrast microscopy showing monolayer (2D) culture at 24 h and 72 h (A–D) and three-dimensional (3D) culture (E–H) of 4T1 (A,C,E,G) and 67NR (B,D,F,H) tumor cells. Tumors from both cell lines were dissociated and were capable of adhering in the plastic dishes and growing up as monolayers over time (A–D), forming spheroids from cells directly isolated from tumors (E,F) and from cell monolayers (G,H) after one day of culture.

Figure 4

Spheroid migration assay reproduces in vitro features of metastasis. Phase-contrast representative images showing “migration clusters” (asterisk), suggesting collective migration of cells (A,B). Cells at the periphery of the migration exhibit typical features of motility, such as filopodia (arrow) (B). Immunofluorescence staining for fibronectin (FN) (C), E-cadherin (E-cad) (D,F) and vimentin (VIM) (E), in green. A “migration cluster” (asterisk) with decreased fibronectin expression, with arrows depicting migrating cells at the periphery with spindle-like morphology (arrow) (C). EMT representative process (D,E); as cells distance from the spheroid, there is a loss of the E-cadherin marker (D) and a concurrent acquisition of vimentin (VIM) expression (E). Zoom of a cell with flatted-like morphology at the spheroid periphery (inset), a typical feature of mesenchymal cells (F). Nuclei were stained with DAPI.

Figure 4

Spheroid migration assay reproduces in vitro features of metastasis. Phase-contrast representative images showing “migration clusters” (asterisk), suggesting collective migration of cells (A,B). Cells at the periphery of the migration exhibit typical features of motility, such as filopodia (arrow) (B). Immunofluorescence staining for fibronectin (FN) (C), E-cadherin (E-cad) (D,F) and vimentin (VIM) (E), in green. A “migration cluster” (asterisk) with decreased fibronectin expression, with arrows depicting migrating cells at the periphery with spindle-like morphology (arrow) (C). EMT representative process (D,E); as cells distance from the spheroid, there is a loss of the E-cadherin marker (D) and a concurrent acquisition of vimentin (VIM) expression (E). Zoom of a cell with flatted-like morphology at the spheroid periphery (inset), a typical feature of mesenchymal cells (F). Nuclei were stained with DAPI.

Figure 5

Doxorubicin decreases the spheroid size. Representative images of phase-contrast microscopy of untreated spheroids (ctr) (A) and dox treated with 1 μM (B), 2 μM (C) and 4 μM (D) after 48 h of treatment. In addition to a reduction in spheroid size, altered morphology was observed. Graphic representation of spheroid reduction diameter over time along with dox treatment (compared with the untreated control) (E). Quantitative data are expressed as the mean ± SD of three experiments, analyzing five spheroids per experimental condition. * p < 0.05; *** p < 0.001. Two-way ANOVA test and Bonferroni post hoc test were used.

Figure 6

Doxorubicin induces cytotoxicity in spheroids. Representative dot plot of 7-AAD labeling by flow cytometry (A). Percentage of 7-AAD-positive cells showed an increase in cell death during dox treatment (B). Alamarblue^® assay data revealed reduced cell viability during dox treatment at all analyzed time points (C). Quantitative data are expressed as the mean ± SD of three experiments, analyzing five spheroids per experimental condition. * p < 0.05; ** p < 0.01; *** p < 0.001. Two-way ANOVA test and Bonferroni post hoc test were used.

Figure 7

Doxorubicin inhibits cell migration. Representative images of phase-contrast microscopy of untreated spheroids (ctr) (A) and treated with 1 μM (B), 2 μM (C) and 4 μM (D) of dox at 48 h of treatment. Graphic showing a decrease in cell spreading from spheroids over time, along with dox treatment, compared with the untreated control (E). DNA staining with DAPI in blue, showing a reduction in “migration clusters” (asterisk) (inset) and the number of cells going out of spheroids in the treated condition (G) compared with untreated (F). Quantitative data are expressed as the mean ± SD of three experiments, analyzing five spheroids per experimental condition. ** p < 0.01; *** p < 0.001. Two-way ANOVA test and Bonferroni post hoc test were used.

Figure 8

Doxorubicin inhibits aspects of the EMT process. Immunoblotting for evaluation of E-cadherin (E-cad) (A) and vimentin (VIM) (B) expression during cell migration after 24 h of treatment. Densitometric analyses revealed an increase in E-cad content and a decrease in VIM during dox treatment (A,B). Immunofluorescence assay showing E-cad staining (D) also at the peripheral zone of migration and a decrease in VIM staining (F) when cells were treated with dox (D,F) compared with the untreated condition (C,E). Nuclei were labeled with DAPI. Quantitative data are expressed as mean ± SD of three experiments. ** p < 0.01; Mann–Whitney test.

Figure 8

Doxorubicin inhibits aspects of the EMT process. Immunoblotting for evaluation of E-cadherin (E-cad) (A) and vimentin (VIM) (B) expression during cell migration after 24 h of treatment. Densitometric analyses revealed an increase in E-cad content and a decrease in VIM during dox treatment (A,B). Immunofluorescence assay showing E-cad staining (D) also at the peripheral zone of migration and a decrease in VIM staining (F) when cells were treated with dox (D,F) compared with the untreated condition (C,E). Nuclei were labeled with DAPI. Quantitative data are expressed as mean ± SD of three experiments. ** p < 0.01; Mann–Whitney test.