Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance

Abstract

The role of epithelial-to-mesenchymal transition (EMT) in metastasis is a longstanding source of debate, largely owing to an inability to monitor transient and reversible EMT phenotypes in vivo. Here we establish an EMT lineage-tracing system to monitor this process in mice, using a mesenchymal-specific Cre-mediated fluorescent marker switch system in spontaneous breast-to-lung metastasis models. We show that within a predominantly epithelial primary tumour, a small proportion of tumour cells undergo EMT. Notably, lung metastases mainly consist of non-EMT tumour cells that maintain their epithelial phenotype. Inhibiting EMT by overexpressing the microRNA miR-200 does not affect lung metastasis development. However, EMT cells significantly contribute to recurrent lung metastasis formation after chemotherapy. These cells survived cyclophosphamide treatment owing to reduced proliferation, apoptotic tolerance and increased expression of chemoresistance-related genes. Overexpression of miR-200 abrogated this resistance. This study suggests the potential of an EMT-targeting strategy, in conjunction with conventional chemotherapies, for breast cancer treatment.

Extended Data Figure 1. Characterization of the primary tumor and lung metastasis of Tri-PyMT mice

Sections of primary tumors (a) and lungs (b) from Tri-PyMT mice were immunostained for E-cadherin (E-cad, top), vimentin (Vim, middle) and CD45 (bottom) in white psuedocolor. Representative images are shown (n > 5 mice). Note the co-localization of PyMT with RFP, and CD45 with GFP (as indicated by arrows) in both primary tumors and lung metastases.

Extended Data Figure 2. Characterization of the primary tumor and lung metastasis of Tri-Neu mice

Sections of primary tumors (left panel) and lungs (right panel) from Tri-Neu mice were immunostained for CD45, Neu, E-cadherin, and Vimentin (in white psuedocolor). Representative images are shown (n>5 mice). Note that both primary tumors and lung metastases are largely composed of epithelial RFP+ tumor cells.

Extended Data Figure 3. Characterization of the primary tumor and lung metastasis of Tri-PyMT/Vim mice

Tri-PyMT/Vim mice were obtained by crossing MMTV-PyMT, Vimentin-Cre and Rosa26-GRFP transgenic mice. Sections of primary tumors (a) and lungs (b) were immunostained for PyMT, E-cadherin, and Vimentin (in white psuedocolor). Representative images are shown (n> 5 mice). Note that both primary tumors and lung metastases are largely composed of epithelial RFP+ tumor cells.

Extended Data Figure 4. Characterization of Tri-PyMT cells

a, EMT of Tri-PyMT cells with TGF. RFP+ Tri-PyMT cells were sorted by flow cytometry and cultured in medium containing 2% FBS with or without TGF-β1 (2ng/mL) for 3 days. Plot shows quantification of the percentage of GFP+ cells analyzed by flow cytometry (n=2 biological replicates). b, Cell migration assay of Tri-PyMT cells. The tracing plots show the movement of individual RFP+ and GFP+ cells in 10 hours of live imaging. Quantification plot (right panel) showed the average distance that RFP+ and GFP+ cells have moved during the time frame (n>20, *p<0.01). c, Relative expression of epithelial, mesenchymal, and tumor markers in sorted RFP+ and GFP+ Tri-PyMT cells as determined by Q-RT-PCR with GAPDH as the internal control. n=2 individual experiments. d, EMT of Tri-PyMT cells is reported by fluorescent marker switch. Flow cytometry plot shows E-cadherin- (E-Cad-) and E-cadherin+ (E-Cad+) subpopulations of Tri-PyMT cells (upper panel). Of the E-Cad- and E-Cad+ subsets, the populations were further dissected according to innate fluorescence (lower panel). Numbers indicate the percentage of GFP+, RFP+, or transitioning (Q2) cells in the parental E-Cad- or E-Cad+ subsets, respectively.

Extended Data Figure 5. Establishing an orthotopic model with sorted RFP+ Tri-PyMT cells

a, Flow cytometry plots show Tri-PyMT cells before and after sorting for RFP+ cells. Numbers indicate the percentage and purity of RFP+ cells used for establishing orthotopic breast tumors in mice. b, Schematic of the orthotopic breast tumor model with sorted RFP+ Tri-PyMT cells. Cells are injected into the mammary gland of wild type mice to generate primary breast tumors, resection of primary tumor at 4 weeks and lung metastases evaluation in another 4 weeks. c, Characterization of tumor cells in the primary tumor, disseminated tumor cells (DTCs) and tumor cells in the lung metastasis of the Tri-PyMT orthotopic model. Sections of primary tumors and lungs from Tri-PyMT orthotopic mice were immunostained for E-cadherin and Vimentin (in white pseudocolor). Essentially all RFP+ tumor cells are detected as E-cad+/Vim−, while the scattered GFP+ tumor cells in the primary tumor are E-Cad−/Vim+ (as indicated by arrows in the top panel). Representative images are shown (n=8). d, Plot shows the percentage of GFP+ cells out of total tumor cells (GFP+ plus RFP+, n=6).

Extended Data Figure 6. Characterization of EMT status of orthotopic Tri-Vim-PyMT primary tumor

Sections of Tri-Vim-PyMT orthotopic primary tumors (a) and metastatic lung (b) were immunostained for E-cadherin and vimentin (in white psuedocolor). As expected, RFP+ tumor cells are entirely E-cadherin positive and vimentin negative, GFP+ tumor cells are vimentin positive and E-cadherin negative, and lung metastases are epithelial and RFP+.

Extended Data Figure 7. Dissemination of Tri-PyMT cells in vivo

a, Disseminated tumor cells are RFP+ and epithelial. RFP+ Tri-PyMT cells were injected into the fat pad of mice. The fluorescence of the primary tumor, circulating tumor cells in the blood and disseminated tumor cells in the lung was analyzed by flow cytometry. The flow cytometry plots depicted are the enumeration of RFP+ and GFP+ cells. b, The ratios of detected RFP+ versus GFP+ cells are shown in the chart (n=4 mice). c, Relative expression of miR-200 family microRNAs in Tri-PyMT control and miR-200 expressing cells. n=2 individual experiments. d, Relative expression of EMT markers and tumor markers in Tri-PyMT control and +mir-200 cells as determined by Q-RT-PCR with GAPDH as the internal control, n=2 individual experiments.

Extended Data Figure 8. Effects of CTX therapy on primary tumors

a, Quantification of primary tumor growth after 2 weeks of CTX therapy. For tumor growth data see Source Data File. b, Proliferation status of primary tumor cells as detected by Ki67 staining in control mice and after 2 weeks of CTX therapy. c, Level of apoptosis in primary tumors as detected by active caspase-3 staining in control mice and after 2 weeks of CTX therapy. Representative images of Ki67 (d) and active caspase-3 staining (e) (white pseudocolor) of primary tumors in control mice and CTX-treated mice. f, Proliferation status of RFP+ and GFP+ primary tumor cells as detected by Ki67 staining in control and CTX-treated mice. g, Level of apoptosis in RFP+ and GFP+ primary tumors as detected by active caspase-3 staining in control and CTX-treated mice. h, Percentage of GFP+ tumor cells in control and CTX-treated primary tumors. n=3 mice for all figures described above. Quantification performed utilizing ImageJ software.

Extended Data Figure 9. EMT tumor cells are resistant to CTX treatment both in vitro and in vivo

a, b, Long term CTX treatment in vitro results in GFP+ population. Tri-PyMT cells were subjected to 2 weeks cyclophosphamide (+CTX) treatment (4 µM). Fluorescent imaging (a) and flow cytometry (quantified, b, n=3) exhibit the percentage of GFP+ cells in the CTX-treated culture compared to control untreated cells c, d, EMT status of lung nodules in competitive survival assay. Representative fluorescent images of Tri-PyMT lung metastases in untreated control lungs (c) and CTX-treated lungs (d), depicting RFP+ and GFP+ tumor cells. Immunostaining showing E-cadherin (E-cad), vimentin (Vim) in white pseudocolor. White arrow indicates GFP+ tumor cells with epithelial phenotypes (E-cad+/Vim−), while the yellow arrow indicates GFP+ cells with mesenchymal phenotypes (E-cad−/Vim+). Nuclei were counter stained with DAPI. n =5.

Extended Data Figure 10. Gene expression profile analysis of RFP+ and GFP+ Tri-PyMT cells

RFP+ and GFP+ Tri-PyMT cells were sorted by flow cytometry and subjected to transcriptomic analysis by RNA-sequencing. a, Heatmap of differentially expressed genes (adjusted p<0.05) from RNA-Seq of sorted RFP+ and GFP+ Tri-PyMT cells, biologically duplicated. Genes that are established epithelial markers (Group 1) include E-cad, Dsp, Epcam, Fgfbp1, Krt18, Krt19, Ocln, Tjp3, Krt14, Tjp2; the mesenchymal markers (Group 2) include N-cad, Col23a1, Col3a1, Col5a1, Col6a2, fsp1, Mmp3, Wnt5a, Zeb1. b, Cell cycle (left panel) and chemoresistance-related (right panel) genes alternatively regulated in RFP+ and GFP+ cells. c, GFP+ Tri-PyMT cells were also sorted from 4-OH-cyclophosphamide (CTX, 4 µM) treated samples. Interestingly, a branch of genes related to drug metabolism were significantly elevated in CTX treated GFP+ cells. Group 1 genes are drug transporters including Abcb1a, Abcb1b, Abcc1. Group 2 genes are Phase I drug metabolizing enzymes including Adh7, Aldh1a1, Aldh1a3, Aldh1l1, Aldh1l2, Aldh2, Aldh3a1, Aldh3a2, Aldh3b2, Aldh4a1, Cyp1a1, Cyp2f2, Cyp2j6, Ptgs1, Ptgs2. Group 3 genes are Phase II drug metabolizing enzymes including Aox1, Blvrb, Ces2e, Ces2f, Ces2g, Chst1, Ephx1, Fmo1, Gpx2, Gsta3, Gsta4, Gstm2, Gsto1, Gstp1, Gstt3, Maoa, Mgst1, Mgst2, Nat6, Nat9, Nqo1, Pon3, Ugt1a6a, Ugt1a7c. d, Aldehyde dehydrogenase (ALDH) activity assay. Cell lysates were prepared from flow cytometry sorted RFP+ and GFP+ Tri-PyMT cells. ALDH activity in samples was measured by OD at 450 nm in a kinetic mode (every 3 minute for 60 min). Representative result from two independent experiments depicted. e, EMT tumor cells (GFP+ cells) showed resistance to multiple commonly used chemotherapies. Tri-PyMT cells were subjected to treatment with 4-OH-cyclophosphamide (CTX, 8 µM), doxorubicin (Dox, 2µM), paclitaxel (Taxol, 10µM) and 5FU (1.6µM) for 3 days. Flow cytometry analysis of apoptotic cells was performed after Annexin staining. The percentage of dead cells (Annexin+) in RFP+ and GFP+ cells, respectively, was quantified. n=2 biological replicates.

Figure 1. Establishing an EMT lineage tracing system in triple transgenic mice

a, Schematic of triple transgenic mice carrying Polyoma Middle-T (PyMT) or Neu oncogenes driven by the MMTV promoter, Cre recombinase under the control of the FSP1 promoter, and floxed RFP-STOP followed by GFP under control of the β-actin promoter in the Rosa26 locus. RFP+ epithelial tumor cells undergoing EMT permanently convert into GFP+ cells following activation of Fsp-1 Cre. b and c, Immunofluorescent microscopy images of Tri-PyMT primary tumors (b) and lung metastases (c) (>10 sections from 3 mice), depicting RFP+ and GFP+ cells within the tumor bed, and staining (white, pseudo-colored) for PyMT.

Figure 2. The EMT lineage tracing system reports EMT in tumor cells with high fidelity

a, Scatter plots from flow cytometry analysis of Tri-PyMT primary tumor cells, depicting GFP+ and RFP+ populations in the primary tumor immediately after sorting of RFP+ cells (P1), and after 10 passages in culture with 10% FBS (P10+10%FBS). Numbers indicate the percentage of RFP+ and GFP+ cells in the total population. b, Phase contrast/fluorescent overlay image of Tri-PyMT cells in culture. c, Western blot of sorted RFP+ and GFP+ Tri-PyMT cells for E-cadherin, vimentin and β-actin as a loading control. Representative of 2 individual experiments. For gel source data, see Supplementary Figure 1. d, Representative imaging of GFP+ and RFP+ tumor cells in primary tumors (PT) and lung metastases (LM) in the orthotopic model (n=8 mice). Arrow indicates scattered GFP+EMT tumor cells in the primary tumor. e, Q-RT-PCR analysis of relative expression of EMT markers in RFP+ and GFP+ cells sorted from orthotopic Tri-PyMT primary tumors. GAPDH served as the internal control. Data are reported as mean ± SEM, n=4 primary tumors.

Figure 3. mir-200 inhibition of EMT in Tri-PyMT cells did not impact lung metastasis

a, Flow cytometry analysis of Tri-PyMT control and +mir-200 cells, indicating percentage of RFP+ and GFP+ cells. b, Representative histologic lung images in Tri-PyMT Cont and +mir-200 orthotopic mice (n=5). c, Quantification of lung metastasis formation (number of individual nodules) in Tri-PyMT control and +mir-200 tumor-bearing mice (n=5).

Figure 4. EMT tumor cells are resistant to chemotherapy

a, Schema of cyclophosphamide (CTX) treatment in Tri-PyMT orthotopic model. Mice bearing an RFP+ primary tumor were treated with CTX (100mg/kg, once per week, for 4 weeks, as indicated by blue arrows). After two weeks of treatment, PT was removed (black arrow). Lung metastasis growth was permitted for 4 weeks post CTX treatment. Fluorescent imaging of lungs revealed the contribution of GFP+ tumor cells to lung metastases (n=9 mice). b, Ratio of GFP+ to RFP+ cells in early metastatic lungs (4 weeks post orthotopic injection) of untreated (Cont) and CTX-treated (+CTX) mice as quantified by flow cytometry (n=4, *p<0.05). c, Apoptosis (as measured by Annexin binding) of RFP+ and GFP+ Tri-PyMT cells treated with CTX (n=2 biological replicates). d, Flow cytometry scatter plot showing the proportions of RFP+ and GFP+ Tri-PyMT cells prior to intravenous injection. Mice were treated with CTX (100mg/kg per week for 3 weeks, n=5 mice per group). e, Quantification of flow cytometry data showing the percentage of RFP+ and GFP+ tumor cells (red and green bars, respectively) of total cells in the lung of the Control and CTX treated mice (CTX) (n=5 mice per group, *p<0.05). f, Quantification of flow cytometry data showing the ratio of GFP+ to RFP+ cells in lungs of Control and CTX-treated mice. Black line represents the starting ratio of GFP+ to RFP+ cells prior to injection as derived from the data in Fig. 4d (*p<0.05).

Figure 5. miR-200 overexpression abrogates CTX resistance

a, Sensitivity of Control and +miR-200 Tri-PyMT tumor cells to CTX treatment as measured by CellTiter-Glo®. n=4 biological replicates per condition b, Representative histologic lung images in Tri-PyMT Cont and +mir-200 tumor-bearing mice treated with CTX, (n=5). c, Quantification of lung metastasis formation (number of individual nodules) in Tri-PyMT control and +mir-200 tumor-bearing mice CTX-treated mice (n=5).