Fig 2: Single-cell RNA sequencing reveals intratumor TEff/EM heterogeneity(A) Independent tSNE plots of sorted live TCR-ß+CD44+CD62L-CD69-CD103- tumor TEff/EM, TCR-ß+CD44+CD62L+CD69-CD103- tumor TCM, TCR-ß+CD44+CD62L-CD69+CD103+ tumor TRM, and TCR-ß+CD44+CD69+CD103+ distant mucosa TRM single-cell transcriptomes obtained from 10 tumors and matched distant mucosa. Each dot represents a cell; each color indicates a distinct CD8+ T cell cluster.(B–D) TEff/EM CD8+ clusters are labeled p1–p4, and the expression of (B) effector molecules and activation markers, (C) transcription factors, and (D) proliferation markers in the tumor TEFF/EM clusters are shown.

Fig 3: RICH1 inhibits stem cell-like properties in breast cancer and maintains the normal epithelial architecture of MCF10A cells.A RICH1 knockdown in MCF10A and overexpression in SUM159 and BT549 cells. B–D Flow cytometry analysis and representative images of ALDHhi and CD44+CD24- subpopulations in MCF10A, SUM159, and BT549 cells. Numbers in the flow cytometry charts indicate the CSC percentages. E Quantitation and representative images of primary and secondary tumorsphere formation in RICH1-knockdown MCF10A and RICH-overexpressing breast cancer cells. F, G. Relative mRNA levels of CSC-TFs (OCT4, SOX2, and NANOG) after RICH1 knockdown (F) and overexpression (G). H Morphogenesis and aberrant acini formation rate of MCF10A cells plated on matrigel 3D culture. I NC- or shRICH1-transfected MCF10A cells in matrigel 3D culture stained with E-cadherin on day 15. All the experiments were repeated three times independently with similar results, and the data of one representative experiment are shown. Data represent mean ± SD. Statistical significance was determined by two-tailed unpaired t-test (B, C, E–H). Abbreviation: CSC-TFs CSC properties-associated transcription factors; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig 4: The lipase catalytic activity of LIPG is relevant to cell growth, but not to the motility, stemness and EMT of breast cancer cells.(A) Western blot analysis of LIPG and tubulin protein expression in vector-control (CTRL), wild-type LIPG-overexpressing (LIPG) and LIPG mutant-overexpressing (LIPGS149A) MCF7 lines. (B) Loss of the lipase catalytic activity of LIPG attenuates cell growth. 2 × 105 cells of three MCF7 lines as described in (A) were seeded for the four-day cell growth study (n = 3 for each time point). (C) LIPGS149A is potent to promote migration of MCF7 cells. Triplicate transwell-based migration experiments were performed on three MCF7 lines as described in (A). The scale bar indicates 100 µm. (D) LIPGS149A is competent to enhance the primary and secondary CSC sphere formation of MCF7 but fails to promote sphere cell proliferation. Triplicate primary and secondary sphere formation experiments were performed on three MCF7 lines as described in (A). The scale bar indicates 50 µm. (E) LIPGS149A overexpression increases the CSC-enriched CD24-/CD44+ cell subset in MCF7 cells. Triplicate FACS analyses of the CD24/CD44 profile were performed on three MCF7 lines as described in (A). The representative FACS profiles of CD24/CD44 are shown on the left panel and the bar graph is shown on the right panel. (F) LIPGS149A overexpression decreases the EpCAMhigh cell population in MCF7 cells. Triplicate FACS analyses of EpCAM vs. FSC were performed on three MCF7 lines as described in (A). The representative FACS profiles of EpCAM/FSC are shown on the left panel and the bar graph is shown on the right panel. (G) LIPGS149A overexpression in MCF7 cells leads to the increased expression of basal/EMT and stem cell genes and downregulation of E-cadherin. qRT-PCR analysis of basal/EMT and stem cell genes was performed on three MCF7 lines as described in (A). ¥ p<0.01 versus the control dataset (MCF7-CTRL); # p<0.01 versus the control dataset (MCF7-CTRL); n = 3. (H) LIPGS149A overexpression downregulates E-cadherin expression. Immunofluorescent analysis of E-cadherin protein expression was performed on three MCF7 lines as described in (A). Cellular DNA was stained with DAPI. The scale bar indicates 25 µm. Error, standard deviation (SD); *p<0.05; **p<0.01; ***p<0.001.10.7554/eLife.31334.015Figure 3—source data 1.Source data for Figure 3. The source data file contains numerical data that were used to generate data graphs presented in sub-figures (3B, C, D, E, F, G) of Figure 3.

Fig 5: ETV4 is required for breast cancer cell stem-like characteristics.A GSEA plot displaying that the ES1 target genes were strongly enriched in TCGA breast cancer tissues with high ETV4 expression. B Heatmap of the expression signature of PEA3 subfamily members between stem cell and non-stem cell subpopulations of MCF10A and SUM149 cells using gene expression data from GSE15192 and GSE132083. C Quantitative RT-PCR measurement of cancer stem cell markers (OCT4 and NANOG) and PEA3 subfamily members (ETV1, ETV4 and ETV5) in monolayer MDA-MB-231 cells and their spheres, monolayer MCF7 cells and their spheres. D Immunoblots of ETV4 and cancer stem cell markers (OCT4, NANOG and LIN28) in monolayer MCF7 cells and MCF7-derived spheres. E Quantitative RT-PCR measurement of ETV4, OCT4 and NANOG in MCF7 spheres and re-adherent MCF7 cells. F, G Representative images of spheres formed by the indicated cells with or without ETV4 depletion (F) and histograms showing the diameters and numbers of spheres (G). Scale bar =100 µm. H Representative images of spheres of MCF7 cells with or without ETV4 overexpression (left panel) and boxplots or histograms showing the diameters and numbers of spheres (right panel). A box represents the distribution of sphere diameter and a upper (lower) whisker represents the max (min) diameter. Scale bar =100 µm. I The subpopulation of CD44+/CD24- cells in control and ETV4-silenced MDA-MB-231 cells as assessed by flow cytometry. J Immunoblots of C-MYC, OCT4, NANOG and LIN28 in MDA-MB-231 and MCF7 cells with or without ETV4 knockdown. Data are presented as mean ± SD, n = 3. *P < 0.05, **P < 0.01, ***P < 0.001.