Conditioned Medium from Malignant Breast Cancer Cells Induces an EMT-Like Phenotype and an Altered N-Glycan Profile in Normal Epithelial MCF10A Cells

Abstract

Epithelial-mesenchymal transition (EMT) is a key process in cancer development and progression. Communication (crosstalk) between cancer cells and normal (nonmalignant) cells may facilitate cancer progression. Conditioned medium (CM) obtained from cultured cancer cells contains secreted factors capable of affecting phenotypes and the behaviors of normal cells. In this study, a culture of normal breast epithelial MCF10A cells with CM from malignant breast cancer cells (termed 231-CM and 453-CM) resulted in an alteration of morphology. CM-treated MCF10A, in comparison with control cells, showed a reduced expression of the epithelial marker E-cadherin, increased expression of the mesenchymal markers fibronectin, vimentin, N-cadherin, and TWIST1, meanwhile cell proliferation and migration were enhanced while cell apoptosis was decreased. N-glycan profiles of 231-CM-treated and control MCF10A cells were compared by MALDI-TOF/TOF-MS (Matrix-Assisted Laser Desorption/ Ionization Time of Flight Mass Spectrometry) and a lectin microarray analysis. The treated cells showed lower levels of high-mannose-type N-glycan structures, and higher levels of complex-type and hybrid-type structures. Altered N-glycan profiles were also detected in 453-CM-treated and non-treated MCF10A cells by MALDI-TOF/TOF-MS, and we found that the expression of five fucosylated N-glycan structures (m/z 1406.663, 1590.471, 1668.782, 2421.141, and 2988.342) and one high-mannose structure m/z 1743.722 have the same pattern as 231-CM-treated MCF10A cells. Our findings, taken together, show that CM derived from breast cancer cells induced an EMT-like process in normal epithelial cells and altered their N-glycan profile.

Keywords: EMT; N-glycan; breast cancer; cell migration; conditioned medium.

Figure 1

Cell morphology and expression of epithelial-mesenchymal transition (EMT) markers in MCF10A cells following TGF-β and conditioned medium (CM) treatment. (a) The morphological changes. The cells were cultured in 6-well plates and treated with CM for 24 h (Right). The cells cultured in DMEM (dulbecco’s modified eagle medium) with 1% fetal bovine serum (FBS) were used as control. The morphological changes of TGF-β-treated cells (Left) were compared with those of the CM-treated cells. The photos were taken under phase-contrast microscopy (40×); (b) Comparative expression of EMT markers in TGF-β-treated vs. PBS (phosphate buffered saline)-treated cells. Protein (10 μg/well) was subjected to SDS-PAGE, and the expression of EMT markers N-cad, E-cad, FN, Vimentin, and TWIST1 was analyzed by Western blotting (GAPDH (reduced glyceraldehyde-phosphate dehydrogenase) as control). Histograms was used to quantify the Western blot data. *** p < 0.001; (c) Comparative expression of EMT markers in CM-treated vs. DMEM/1% FBS-cultured cells. Western blotting was performed as above. Histograms were used to quantify the Western blot data. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 2

Proliferation and migration of TGF-β-treated and CM-treated MCF10A cells. (a) MTT assay of cell proliferation. Cells were seeded in 96-well plates and treated with 231-CM and 453-CM for 24, 48, or 72 h. DMEM/1% FBS-cultured cells were used as control. Proliferation was compared for TGF-β-treated vs. CM-treated cells. The results shown are mean ± standard error of measurement (SEM) from three independent experiments. * p < 0.05 ,*** p < 0.001; (b) Wound assay of cell migration. The cells were cultured in 24-well plates to high confluence (>80%), scratched with a 200-μL pipette tip at the marked position, washed twice with PBS, cultured in fresh medium with 1% FBS, and treated with TGF-β or 231-CM or 453-CM for 24, 48, or 72 h. Nontreated cells were used as control. Wounds were photographed at the marked position at the above times under phase-contrast microscopy (10×). Histograms was used to quantify the wound assay data. *** p < 0.001; (c) Cell apoptosis analysis. The cells were cultured in 6-well plates and treated with TGF-β or CM for 24 h. A PE (Phycoerythrin) Annexin V Apoptosis Detection Kit I was used to stain the apoptosis cells. Cells stained with PE Annexin V were identified as the early apoptotic cells (7-AAD (7-Aminoactinomycin D) negative, PE Annexin V positive), and cells that were in late apoptosis or were already dead were both PE Annexin V and 7-AAD positive.

Figure 3

The alteration of N-glycan profiles in MCF10A cells was induced by 231-CM. (a) MALDI-TOF/TOF-MS spectra of N-glycans. The cells were cultured in a 10 cm dish, and the N-glycans were separated and desalted as described in Materials and Methods. Lyophilized N-glycans were dissolved in methanol/H₂O (MW), and an aliquot of mixture with DHB (2,5-dihydroxybenzoic acid) solution was spotted on an MTP (maldi target plate) AnchorChip sample target and air-dried. MALDI-TOF/TOF-MS was performed in positive-ion mode. The experiments were performed in triplicate, and representative N-glycan spectra are shown. Peaks (signal-to-noise ratio >5) were selected for a relative proportion analysis. Detailed structures were analyzed using the GlycoWorkbench program. Proposed structures are indicated by m/z value. Top: 231-CM-treated cells. Bottom: DMEM/1% FBS-incubated cells; (b) Relative variation of various types of N-glycans in 231-CM-treated cells. Different colors represent m/z values as indicated.

Figure 4

Lectin microarray analysis of glycan variation in 231-CM-treated MCF10A. (a) Variation of expression of the glycans recognized by 37 lectins as indicated. * p < 0.05; ** p < 0.01; *** p < 0.001; (b) Lectin staining analysis of altered glycan expression. Five lectins (SJA, AAL, LEL, STL, and PTL-II) were applied, and lectin staining was performed as described in Materials and Methods. Signals are shown from a merge image of Cy3-conjugated lectins and DAPI (4′,6-diamidino-2-phenylindole) staining of the nuclei in control (top) and 231-CM-treated (bottom) cells (magnification 60×).