Fig. 6: ZEB1 induces different EMT states in breast cancer cells with CD151 as a potential marker of early EMT.

a–h show different aspects of scRNA-seq analysis of MCF7-V cells expressing ZEB1 after induction with DOX for 5 weeks to obtain EpCAM^high and EpCAM^low cell populations. a Visualization of the distribution of the merged data sets of both EpCAM^high (red rectangle) and EpCAM^low (blue rectangle) cell populations, based on the comparison of the transcriptomes of individual cells; the graph was generated by a Uniform Manifold Approximation and Projection (UMAP); the clusters are color-coded and numbered, and each dot represents a single cell. b Single cells expressing EPCAM in different clusters. c Violin plot of the EPCAM transcript levels across different clusters; note that the steps at low levels of expression are due to rounding off after normalization. d Single cells expressing ESR1 in different clusters. e Violin plot for ESR1 transcripts with the same color code and numbering as in a. f–h Expression of top markers of epithelial (f), hybrid EMT (g), and mesenchymal-like (h) states in breast cancer cells upon induction of an EMT by ZEB1 expression. The legend shows a color gradient of normalized expression. The accompanying scheme of the EMT on the right was created with BioRender.com. i Expression of the mRNA for the CD151 cell surface marker across various subpopulations. j Relative proliferation of −DOX and +DOX cells with or without CD151 knockdown exposed to veh or E2 for 72 h (means ± SEM, n = 4 biologically independent experiments averaged from a total of n = 16 technical replicates). Statistical significance was determined with two-way ANOVA. p values are indicated above the bars. k Quantification of the effect of CD151 knockdown on the migration of MCF7-V-ZEB1 cells treated with veh or E2 in a wound-healing assay; the Y axis shows the remaining wound area into which cells have not yet migrated (means ± SEM and n = 3 biological replicates each including n = 2 technical replicates). Statistical significance was determined with a two-way ANOVA. p values are indicated above the bars. Source data are provided as a Source Data file.