Regulation of epithelial-mesenchymal transition in breast cancer cells by cell contact and adhesion

Abstract

Epithelial-mesenchymal transition (EMT) is a physiological program that is activated during cancer cell invasion and metastasis. We show here that EMT-related processes are linked to a broad and conserved program of transcriptional alterations that are influenced by cell contact and adhesion. Using cultured human breast cancer and mouse mammary epithelial cells, we find that reduced cell density, conditions under which cell contact is reduced, leads to reduced expression of genes associated with mammary epithelial cell differentiation and increased expression of genes associated with breast cancer. We further find that treatment of cells with matrix metalloproteinase-3 (MMP-3), an inducer of EMT, interrupts a defined subset of cell contact-regulated genes, including genes encoding a variety of RNA splicing proteins known to regulate the expression of Rac1b, an activated splice isoform of Rac1 known to be a key mediator of MMP-3-induced EMT in breast, lung, and pancreas. These results provide new insights into how MMPs act in cancer progression and how loss of cell-cell interactions is a key step in the earliest stages of cancer development.

Keywords: breast cancer; cell contact; epithelial-mesenchymal transition; extracellular matrix; mammary epithelial cells; matrix metalloproteinases.

Figure 1

Regulation of EMT characteristics by cell density in MCF10A cells. (A) Phase contrast micrographs of cells plated at indicated densities in 35-mm plates and imaged after 24 hours. Scale bar 200 μm. (B) Area of cells at indicated densities (n > 20 for each condition; values displayed as means ± SEM; ANOVA P < 0.001 for trend). (C–E) Quantitative PCR assessment of E-cadherin (C; ANOVA P = 0.042 for trend), N-cadherin (D; ANOVA P = 0.12 for trend), and vimentin (E; ANOVA P < 0.001 for trend) expression in the cell cultures.

Figure 2

Analysis of genes differentially expressed by density in MCF10A cells. (A–D) Genes upregulated more than two-fold in cells cultured at 800K density vs 50K density (A and B; n = 1444 features mapped to 1131 genes) or downregulated more than two-fold in cells cultured at 800K density vs 50K density (C and D; n = 1658 features mapped to 1303 genes); all genes are normalized to 50K expression and displayed as line graphs (A and C; colored by expression at 800K) or box-and-whisker plots (B and D). (E and F) Overlap of dataset of genes differentially regulated two-fold in MCF10A cells cultured at 800K density vs 50K density with datasets of genes differentially regulated between MDA-MB-231 cells and MCF10A cells (showing negative correlation; E) and of genes differentially regulated between MCF10A cells cultured on differentiating conditions vs 2D monolayers (showing positive correlation; F).

Figure 3

Effects of MMP-3 treatment on SCp2 mouse mammary epithelial cells cultured at different cell densities. Either 50K (top row), 100K (middle row), or 250K (bottom row) SCp2 cells were plated in 35-mm plates, and then treated as controls (left column) or with MMP-3 (middle column). Cell area measurements indicate significantly increased cell spreading with MMP-3 treatment at all three densities (right column). Scale bars, 250 μm in large views and 50 μm in insets.

Figure 4

Clustering of genes differentially regulated by density and by MMP-3 treatment in SCp2 cells. (A–H) Differentially expressed genes (n = 7056) in the SCp2 dataset were identified as FC > 2 in any of 50K control vs 50K MMP-3, 250K control vs 250K MMP-3, or 50K control vs 250K control. K-means clustering was performed on the SCp2 differentially regulated gene set using eight groups, Pearson-centered similarity measure, and 1000 iterations. The eight-derived clusters could be generally identified as genes upregulated by density and downregulated by MMP-3 (n = 626; A), genes downregulated by density and upregulated by MMP-3 (n = 806; B), genes upregulated by density and little affected by MMP-3 (n = 517; C), genes downregulated by density and little affected by MMP-3 (n = 801; D), genes upregulated by MMP-3 and little affected by density (n = 923; E), genes downregulated by MMP-3 and little affected by density (n = 572; F), genes upregulated by density in the absence of MMP-3 and unregulated by density in the presence of MMP-3 (n = 1528; G), and genes downregulated by density in the absence of MMP-3 and unregulated by density in the presence of MMP-3 (n = 1283; H). For each cluster, the associated genes are shown as a line graph (left panel) and as a box-and-whisker plot (center panel), and expression data for a representative gene from each cluster are shown in the right panel (boxes indicate variation in replicated experiments for 50K, 100K, and 250K).

Figure 5

Enrichment of mRNA processing genes in clusters of genes regulated by only density in the absence of MMP-3. (A–C) Genes upregulated by density in the absence of MMP-3 and unregulated by density in the presence of MMP-3 annotated in Gene Ontology as associated with mRNA processing (n = 56 features mapped to 37 genes). (D–F) Genes downregulated by density in the absence of MMP-3 and unregulated by density in the presence of MMP-3 annotated in Gene Ontology as associated with mRNA processing (n = 80 features mapped to 56 genes). Gene sets are displayed as line graphs (left) and box-and-whisker plots (right; A and D). Expression data for a representative gene from each cluster are shown (B and E; boxes indicate variation in replicated experiments for 50K, 100K, and 250K). The list of each gene set is displayed (C and F).

Figure 6

Density-dependent differences in expression of gene splicing factors and Rac1b in SCp2 and MCF10A cells. (A) Expression in SCp2 cells of the gene encoding hnRNPA1, known to inhibit expression of Rac1b, according to cell density and MMP-3 treatment (gene expression from normalized microarray data, scaled to 50K control, and displayed as means ± SEM; ANOVA P = 0.0575 for trend in control; P = ns for trend in MMP-3 treated). (B) Expression in SCp2 cells of Rac1b according to cell density and MMP-3 treatment (gene expression from QPCR, normalized to GAPDH and scaled to 50K control, and displayed as means ± SEM; ANOVA P = 0.0475 for trend in control; P = ns for trend in MMP-3 treated). (C) Expression in MCF10A cells of genes encoding ESRP1, known to inhibit expression of Rac1b and ESRP2, according to cell density (gene expression from normalized microarray data, scaled to 50K control). (D) Expression in MCF10A cells of Rac1b according to cell density treatment (gene expression from QPCR, normalized to GAPDH and scaled to 50K control, and displayed as means ± SEM; ANOVA P < 0.001 for trend).