The EMT-activator Zeb1 is a key factor for cell plasticity and promotes metastasis in pancreatic cancer

Journal name:
Nature Cell Biology
Volume:
19,
Pages:
518–529
Year published:
DOI:
doi:10.1038/ncb3513
Received
Accepted
Published online

Abstract

Metastasis is the major cause of cancer-associated death. Partial activation of the epithelial-to-mesenchymal transition program (partial EMT) was considered a major driver of tumour progression from initiation to metastasis. However, the role of EMT in promoting metastasis has recently been challenged, in particular concerning effects of the Snail and Twist EMT transcription factors (EMT-TFs) in pancreatic cancer. In contrast, we show here that in the same pancreatic cancer model, driven by Pdx1-cre-mediated activation of mutant Kras and p53 (KPC model), the EMT-TF Zeb1 is a key factor for the formation of precursor lesions, invasion and notably metastasis. Depletion of Zeb1 suppresses stemness, colonization capacity and in particular phenotypic/metabolic plasticity of tumour cells, probably causing the observed in vivo effects. Accordingly, we conclude that different EMT-TFs have complementary subfunctions in driving pancreatic tumour metastasis. Therapeutic strategies should consider these potential specificities of EMT-TFs to target these factors simultaneously.

At a glance

Figures

  1. Zeb1 depletion reduces invasion and metastasis in pancreatic cancer.
    Figure 1: Zeb1 depletion reduces invasion and metastasis in pancreatic cancer.

    (a) Scheme of the genetic mouse models for pancreatic cancer. The colour code (blue, KPC; red, KPCZ) is used for all results. (b) Tumour-free survival (n = 28 KPC, 18 KPCZ; log-rank (Mantel–Cox) test) and tumour volume (0, start of magnetic resonance imaging (MRI) measurements; n = 23 KPC, 27 KPCZ; error bars show mean ± s.e.m.; multiple t-tests with correction for multiple comparison using the Holm–Sidak method). NS, not significant. (c) Representative haematoxylin and eosin (HE)-stained sections for the grading of the respective tumours (n = 48 KPC, 29 KPCZ independent tumours). Scale bar, 250μm and 125μm for higher magnifications. (d) Grading and local invasion of the respective tumours (n = 48 KPC, 29 KPCZ independent tumours; error bars show mean  ±  s.d.; Mann–Whitney test (two tailed), chi-square test (two tailed) for grade 3/4 tumours), P < 0.0001. (e) Representative immunohistochemical stainings of consecutive sections showing nuclear Zeb1 in tumour cells (arrows) of invasive tumour regions in KPC, but not in KPCZ mice (n = 15 KPC, 13 KPCZ independent tumours). Asterisks mark Zeb1 expression in stromal cells, cen (central) and inv (invasive tumour regions). Scale bar, 75μm. (f) Numbers and grading of metastasized tumours (n = 52 KPC, 29 KPCZ independent tumours; error bars, mean ± s.d.; chi-square test (two tailed) for metastasis, Mann–Whitney test (two tailed) for grading). (g) Representative images of differentiated (KPC and KPCZ) and undifferentiated (KPC) primary tumours (PT) and corresponding metastases (Met) with the same phenotype (L, liver) (n = 19 KPC, 4 KPCZ independent tumours and corresponding metastases). Scale bar, 150μm.

  2. Depletion of Zeb1 affects phenotypic variability of tumour cells.
    Figure 2: Depletion of Zeb1 affects phenotypic variability of tumour cells.

    (a) Anti-E-cadherin and anti-vimentin immunofluorescence stainings showing variable expression in KPC cell lines, and homogeneous E-cadherin and lack of vimentin expression in all KPCZ cell lines. DAPI, 4,6-diamidino-2-phenylindole. Scale bar, 100μm. (b) Relative messenger RNA expression levels of indicated marker genes in the isolated tumour cells. (c) Relative mRNA expression levels for EMT transcription factors and epithelial microRNAs. The mRNA level of cell line 661 was set to 1. n = 3 biologically independent experiments, error bars, mean ± s.e.m. P < 0.05,P < 0.01, NS, not significant, Mann–Whitney test (one tailed) (b,c). (d) Immunoblots of indicated marker genes (unprocessed scans of immunoblots are shown in Supplementary Fig. 8). (e) 5-bromodeoxyuridine (BrdU) proliferation assay for the isolated tumour cell lines (n = 3 biologically independent experiments, error bars, mean ± s.e.m.). The colour code for the isolated cell lines as depicted in b is valid for all corresponding results.

  3. Depletion of Zeb1 affects stemness, tumorigenic and colonization capacities.
    Figure 3: Depletion of Zeb1 affects stemness, tumorigenic and colonization capacities.

    (a) Representative images of macroscopic and HE-stained lungs, 18 days after intravenous (i.v.) injection of tumour cells in syngeneic mice. Quantification of lung colonies (left, cell lines grouped by genotype; right, individual cell lines (for ac,e), normalized to 20mm2 lung area). n = 3 mice per cell line, n = 4 mice for line 524, error bars, mean  ±  s.d.; P < 0.0001, Mann–Whitney test (two tailed). Scale bar, 200μm. (b) Number of green fluorescent protein (GFP)+ cells per visual field 2h after i.v. injection (n = 3 mice per cell line, error bars, mean ± s.d. Mann–Whitney test (two tailed)). (c) Quantification after i.v. injection of KPC, KPCS and KPCZ tumour cells in nude mice (n = 13 mice for KPC, n = 8 for KPCS, n = 6 for KPCZ—four mice per cell line, Mann–Whitney test (two tailed), P < 0.01, NS, not significant). Relative mRNA expression levels in KPCS cell lines; mRNA levels of KPC661 (expressing low levels of Snail) set to 1 (average of n = 2 biologically independent experiments, error bars, mean ± s.d.). Immunoblot for the indicated proteins with KPC701 as control expressing high Snail levels. (d) Number of lung colonies after i.v. injection of KPC shcontrol (ctr) and KPC shZeb1 tumour cells in nude mice (normalized to 20mm2 lung area) (n = 3 mice per cell line, error bars mean ± s.d.; Mann–Whitney test (two tailed), NS, not significant). Immunoblots and corresponding quantifications, showing short hairpin RNA (shRNA)-mediated partial reduction of Zeb1 (n = 3 biologically independent experiments, error bars, mean ± s.e.m.; unpaired Students t-test (two tailed), P < 0.01). (e) Quantification of sphere-forming capacity (n = 3 biologically independent experiments, error bars, mean ± s.d.; P < 0.05, Mann–Whitney test (two tailed)). (f) Relative mRNA expression levels and immunoblots of stem-cell genes (n = 3 biologically independent experiments, error bars, mean ± s.e.m.; P < 0.05, Mann–Whitney test (one-tailed)). mRNA levels of line 661 set to 1. (g) HE and immunohistochemical staining for Sox2 in tumours grown subcutaneously (n = 51) or in the lung (n = 36) after i.v. injection (l.c.) of indicated cell lines. Scale bar, 100μm. (h) Immunoblot for indicated proteins on overexpression of Mir200c. For source data for c,d,f see Supplementary Table 5; unprocessed scans of immunoblots are shown in Supplementary Fig. 8.

  4. Depletion of Zeb1 reduces ADM- and PanIN-precursor lesions.
    Figure 4: Depletion of Zeb1 reduces ADM- and PanIN-precursor lesions.

    (a,b) Consecutive sections of representative HE- and periodic acid/Schiffs (PAS)-stained sections showing precancerous PanIN (a) and ADM lesions (b) in the pancreas of 6-month-old KC and KCZ mice. Specific dark blue/purple PAS staining indicates the mucin-rich PanIN lesions; arrows indicate ADMs. Squares mark the magnified regions; scale bars, 1mm and 150μm for higher magnifications in a and 75μm in b. Quantification of the ADM and PanIN areas and PanIN grading is given. n = 12 KC and 7 KCZ independent mice, error bars, mean ± s.d.; P < 0.01,P < 0.0001; unpaired Students t-test (two tailed) with Welchs correction for ADM and PanIN areas and Mann–Whitney test (two tailed) for grading.

  5. Depletion of Zeb1 reduces phenotypic variability.
    Figure 5: Depletion of Zeb1 reduces phenotypic variability.

    (a) PCA of the KPC and KPCZ cell-line transcriptomes. The plot depicts the first two principal components using all samples, accounting for about 44% and about ~17% of the variance, respectively. (b) GSEAs of transcriptome data from KPCZ versus KPC cells reveal enrichment of gene signatures associated with Kras dependency and the classical type of pancreatic cancer, as well as a reduction of genes associated with metastasis in KPCZ cell lines. NES, normalized enrichment score; FDR, false-discovery rate. (c) Relative mRNA expression levels (quantitative PCR with reverse transcription, qRT–PCR) and immunoblots of indicated genes associated with metastasis in the isolated tumour cells (n = 3 biologically independent experiments, error bars, mean ± s.e.m.; P < 0.05, P < 0.01, Mann–Whitney test (one-tailed)). The mRNA level of cell line 661 was set to 1. Unprocessed scans of immunoblots are shown in Supplementary Fig. 8.

  6. Depletion of Zeb1 reduces TGF[beta]-induced cellular plasticity.
    Figure 6: Depletion of Zeb1 reduces TGFβ-induced cellular plasticity.

    (a) Anti-E-cadherin and anti-vimentin immunofluorescence staining of two epithelial KPC and two KPCZ cancer cell lines treated with TGFβ1 for 3 and 21 days. Scale bar, 100μm. (b) Immunoblots for indicated marker genes of the same lines as in a. Unprocessed scans of immunoblots are shown in Supplementary Fig. 8. (c) PCA of transcriptome signatures of the KPC and KPCZ cell lines on TGFβ treatment. TGFβ-induced shifts in expression of the cell lines shown in a are marked with coloured boxes (microarrays carried out in duplicate, referred to as TGFβ_1 and TGFβ_2). Note a great shift towards a mesenchymal pattern for KPC cell lines but not for KPCZ lines (upper panel). Venn diagram showing a number of genes significantly up- or downregulated (cutoff: adjusted P < 0.05 and log2FC > 0.5) by 14 days of TGFβ treatment of cell lines shown in a. Moderated t-test (lower panel). (d) Relative mRNA expression levels (qRT–PCR) of indicated genes (including the metastasis set in Fig. 5c) in KPC and KPCZ cell lines treated for different times with TGFβ (times: 0, 6 h, 1, 3, 7, 14, 21 days) (n = 3 biologically independent experiments, error bars, mean ± s.e.m.). mRNA levels of cell line 661 at day 0 were set to 1. Statistical analysis is shown for the comparison of TGFβ treated and untreated samples (grey bars) of each individual cell line. P < 0.05, P < 0.01, unpaired Students t-test (one-tailed). (e) Anti-E-cadherin and anti-vimentin immunofluorescence staining of two epithelial KPC and two KPCZ cancer cell lines treated with TGFβ for more than 21 days (>21 d) followed by 14 days TGFβ withdrawal (−14 d). Scale bar, 100μm. (f,g) Immunoblots (f) and relative mRNA expression levels (qRT–PCR) (g) of indicated marker genes of the same cell lines as in e (n = 3 biologically independent experiments, error bars, mean ± s.e.m.; P < 0.05, P < 0.01, P < 0.001, unpaired Students t-test (one-tailed)). mRNA levels of cell line 661 at day 0 were set to 1. For source data for Fig. 5d, f see Supplementary Table 5; unprocessed scans of immunoblots are shown in Supplementary Fig. 8.

  7. Depletion of Zeb1 reduces metabolic and phenotypic plasticity.
    Figure 7: Depletion of Zeb1 reduces metabolic and phenotypic plasticity.

    (a) Mitochondrial stress test (MST) showing the oxygen consumption rate (OCR) as an indicator for OXPHOS and deduced levels for basal respiration and ATP production. (b) Glycolysis stress test (GST) showing the extracellular acidification rate (ECAR) as an indicator for glycolysis and the OCR after glucose stimulation, blocking of oxidative phosphorylation with oligomycin and blocking of glycolysis with 2-deoxy-glucose (2DG), as well as deduced glycolytic capacity and glycolytic reserve (n = 7 biologically independent experiments; error bars, ± s.e.m. for MST and GST and ± s.d. for other parameters; for MST and GST a multiple t-test with correction for multiple comparison using the Holm–Sidak method was used; for other parameters an unpaired Students t-test (two tailed) was used; P < 0.05, P < 0.01, P < 0.001,P < 0.0001). Note a complete lack of a glycolytic reserve (upper arrow) after blocking oxidative phosphorylation (lower arrow) in KPCZ cells. KPC661 and 792 as well as all KPCZ cell lines were used. (c) Representative images of consecutive sections of immunohistochemical stainings for Ck19 and Zeb1, comparing the plasticity of Zeb1 expression in central and invasive tumour regions. Shown are tumours derived from one KPC and one KPCZ cell line. Asterisks label Zeb1 expression in stromal cells; arrows indicate Zeb1-positive tumour cells at the invasive front. Ck19 expression is shown to identify cancer cells. n = 15 KPC, 13 KPCZ independent tumours, Scale bars, 50μm and 150μm for higher magnifications.

  8. Characterisation of KPC, heterozygously and homozygously Zeb1 depleted KPC tumours.
    Supplementary Fig. 1: Characterisation of KPC, heterozygously and homozygously Zeb1 depleted KPC tumours.

    (a) Representative Zeb1-immunolabeling of a GFP lineage-traced primary tumour showing Zeb1/GFP double-positive tumour cells (arrows). n = 5 independent tumors. Scale bar, 50μm. (b) Representative consecutive sections of HE and indicated immunohistochemical stainings of four Zeb1 expressing KPC tumours demonstrating the heterogeneity in phenotype, grading and marker expression. A representative differentiated Zeb1-negative KPCZ tumour is shown for comparison. Arrows indicate Zeb1 positive tumour cells in the differentiated KPC tumour. n = 15 KPC, 13 KPCZ independent tumours. Scale bar, 100μm. (c) Tumour-free survival of KPC mice vs. KPC mice with a heterozygous deletion of Zeb1 (KPCz) (n = 15 KPC, 16 KPCz independent tumours); log-rank (Mantel-Cox) test); tumour volume (0 = start of MRI measurements, n = 12 KPC, 14 KPCz independent tumours); error bars show mean ± S.E.M.; multiple t-tests with correction for multiple comparison using the Holm-Sidak method; grading, local invasion and relative ECM deposition of the respective tumours (n = 31 KPC, 17 KPCz; Mann-Whitney test (two-tailed); percentage of metastasized tumours (n = 35 KPC, 17 KPCz independent tumours; Chi-square test (two-tailed); n.s. = not significant.

  9. Characterisation of KPC vs. KPCZ tumours.
    Supplementary Fig. 2: Characterisation of KPC vs. KPCZ tumours.

    Representative images of immunohistochemical and histological stainings of KPC and KPCZ tumours and quantifications of the indicated markers are given. Asterisks label Zeb1-expressing stroma cells in KPCZ tumours. Specific blue MTS staining labels collagen fibres. Scale bars, 100μm, for lower left image 50μm. n = 48 KPC, 29 KPCZ independent tumours for Zeb1 and MTS; n = 15 independent tumours for KPC, 13 independent tumours for KPCZ for all other markers, error bars show mean ± S.D.; p < 0.0001, n.s. = not significant, Chi-square test (two-tailed) for Zeb1, E-cadherin and Sox2, unpaired Students t-test (two-tailed) for Ki67 and Casp3 (with Welchs correction), Mann-Whitney test (two-tailed) for ECM and CD31.

  10. Characterisation of differentiaton markers in KPC vs. KPCZ tumours.
    Supplementary Fig. 3: Characterisation of differentiaton markers in KPC vs. KPCZ tumours.

    (a) Representative images of positive and negative immunohistochemical stainings and statistical analysis for the indicated EMT-TFs. Scale bar, 150μm. n = 14 independent tumours for KPC, 13 independent tumours for KPCZ, Chi-square test (two-tailed); n.s. = not significant. (b) Representative images of immunohistochemical stainings and statistical analysis for expression of Gata6. Scale bar, 150μm. n = 14 independent tumours for KPC, 13 independent tumours for KPCZ; error bars show mean ± S.D.; Mann-Whitney test (two-tailed), p < 0.001. (c) Representative images of differentiated KPCZ and undifferentiated KPC primary tumours (PT) and corresponding metastases (Met) with the same phenotype. Immunohistochemical labelling of Zeb1 expressing tumour cells in the KPC PT and Met (arrows). L = liver or lung tissue. n = 19 KPC, 4 KPCZ independent tumours and corresponding metastases. Scale bar, 100μm.

  11. Characterisation of KPC vs. KPCZ tumour derived cell lines.
    Supplementary Fig. 4: Characterisation of KPC vs. KPCZ tumour derived cell lines.

    (a) Bright field image of primary cell lines from KPC and KPCZ tumours as well as HE stainings of the respective tumours after grafting in syngeneic mice and of the respective primary tumours are shown. Scale bars, 100μm for bright field, 75μm for HE stainings. (b) MTT viability assay for the isolated tumour cell lines after treatment with the indicated doses of gemcitabine and erlotinib. The calculated IC50 values for gemcitabine are shown. n = 3 biologically independent experiments, error bars show mean ± S.E.M. (c) Tumour onset after subcutaneous injection of 1 × 105 KPC and KPCZ cells into syngeneic mice. n = 4 mice per cell line, error bars show mean ± S.E.M. (d) Tumour grading, grading at invasive regions and relative ECM deposition of one representative tumour/cell line analysed in c) (n = 6 tumours for KPC, n = 5 tumours for KPCZ); error bars show mean ± S.D.;p < 0.05, p < 0.01, Mann-Whitney test (two-tailed).

  12. Depletion of Zeb1 affects tumour promoting capacities.
    Supplementary Fig. 5: Depletion of Zeb1 affects tumour promoting capacities.

    (a) Representative images of one visual field (n = 6 fields/cell line) showing GFP + cells/cell clusters in the lungs (green dots) 2 h after i.v. injection of KPC and KPCZ tumour cells and control lungs. Scale bar, 500μm. (b) No. of tumours after subcutaneous injection of the indicated cell numbers for the KPC and KPCZ tumour cell lines and calculated fraction of tumourigenic cells. inf = infinite, Chi-square test. (c) Representative images showing spheres of KPC and KPCZ tumour cells. Scale bar, 500μm and 50μm for higher magnifications. (d) Percentage of cells in KPC and KPCZ lines positive for the indicated markers or marker combinations; n = 2 biologically independent experiments, error bars show ± S.D. Source data see Supplementary Table 5, Statistics Source Data. Relative mRNA expression levels (qRT-PCR) of indicated genes, mRNA levels of KPC661 was set to 1; n = 3 biologically independent experiments, Mann-Whitney test (two-tailed), p < 0.05, p < 0.01, error bars show mean ± S.E.M.

  13. Depletion of Zeb1 reduces early PanIN lesions.
    Supplementary Fig. 6: Depletion of Zeb1 reduces early PanIN lesions.

    (a) Consecutive sections showing representative HE and PAS stainings of precancerous PanIN lesions in the pancreas of two different 6 month old KC and of one KCZ mice. Specific dark blue PAS staining indicates the mucin-rich PanIN lesions. Scale bars, 2.5 mm and 150μm for higher magnifications. Quantification of the PanIN area (% of pancreas area). n = 12 KC and 7 KCZ independent mice, error bars show mean ± S.D.; p < 0.01, unpaired Students t-test (two tailed) with Welchs correction. (b) Gene set enrichment analyses (GSEA) of transcriptome data from KPCZ vs. KPC cells reveals reduction of gene signatures associated with cancer mesenchymal transition and Zeb1 targets in KPCZ vs. KPC cell lines. NES = normalized enrichment score; FDR = false discovery rate.

  14. Depletion of Zeb1 reduces tumour cell plasticity.
    Supplementary Fig. 7: Depletion of Zeb1 reduces tumour cell plasticity.

    (a) Relative mRNA expression levels (qRT-PCR) of indicated genes in KPC and KPCZ cell lines treated for different times with TGFβ (time points: 0, 6 h, 1, 3, 7, 14, 21 days). mRNA levels of cell line 661 at day 0 were set to 1. n = 3 biologically independent experiments, error bars show mean ± S.E.M. Statistical analysis is shown for the comparison of TGFβ treated to untreated samples (grey bars) of each individual cell line p < 0.05, p < 0.01, p < 0.001, p < 0.0001, unpaired Students t-test (one-tailed) Source data see Supplementary Table 5, Statistics Source Data. (b) Table showing log2FC in mRNA expression levels (microarray) of genes previously determined as common ZEB1/YAP targets in KPC and KPCZ cell lines upon TGFβ treatment for 14 days. (cut-off: adj. p-value < 0.05 and log2FC > 0.5). (c) Representative images of consecutive sections of immunohistochemistry for Ck19 and Zeb1 comparing the plasticity of Zeb1 expression in central and invasive tumour regions. Tumours derived from one KPC and one KPCZ cell line are shown. Asterisks label Zeb1 expression in stroma cells, arrows indicate Zeb1 expression in tumour cells at the invasive front. Ck19 expression is shown to identify cancer cells. n = 15 KPC, 13 KPCZ independent tumours, Scale bars, 50μm and 150μm for higher magnifications.

Accession codes

Primary accessions

Gene Expression Omnibus

References

  1. Brabletz, T. To differentiate or not—routes towards metastasis. Nat. Rev. Cancer 12, 425436 (2012).
  2. Kalluri, R. & Weinberg, R. A. The basics of epithelial-mesenchymal transition. J. Clin. Invest. 119, 14201428 (2009).
  3. Nieto, M. A., Huang, R. Y.-J., Jackson, R. A. & Thiery, J. P. EMT: 2016. Cell 166, 2145 (2016).
  4. Brabletz, S. & Brabletz, T. The ZEB/miR-200 feedback loop–a motor of cellular plasticity in development and cancer? EMBO Rep. 11, 670677 (2010).
  5. Vandewalle, C., Van Roy, F. & Berx, G. The role of the ZEB family of transcription factors in development and disease. Cell. Mol. Life Sci. 66, 773787 (2009).
  6. Zhang, P., Sun, Y. & Ma, L. ZEB1: at the crossroads of epithelial-mesenchymal transition, metastasis and therapy resistance. Cell Cycle 14, 481487 (2015).
  7. Fischer, K. R. et al. Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance. Nature 527, 472476 (2015).
  8. Zheng, X. et al. Epithelial-to-mesenchymal transition is dispensable for metastasis but induces chemoresistance in pancreatic cancer. Nature 527, 525530 (2015).
  9. Hingorani, S. R. et al. Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell 7, 469483 (2005).
  10. Rhim, A. D. et al. EMT and dissemination precede pancreatic tumor formation. Cell 148, 349361 (2012).
  11. Higashi, Y. et al. Impairment of T Cell Development in deltaEF1 Mutant Mice. J. Exp. Med. 185, 14671480 (1997).
  12. Brabletz, S. et al. Generation and characterization of mice for conditional inactivation of Zeb1. Genesis http://dx.doi.org/10.1002/dvg.23024 (2017).
  13. Martinelli, P. et al. GATA6 regulates EMT and tumour dissemination, and is a marker of response to adjuvant chemotherapy in pancreatic cancer. Gut http://dx.doi.org/10.1136/gutjnl-2015-311256 (2016).
  14. Laklai, H. et al. Genotype tunes pancreatic ductal adenocarcinoma tissue tension to induce matricellular fibrosis and tumor progression. Nat. Med. 22, 497505 (2016).
  15. Erkan, M. et al. The activated stroma index is a novel and independent prognostic marker in pancreatic ductal adenocarcinoma. Clin. Gastroenterol. Hepatol. 6, 11551161 (2008).
  16. Özdemir, B. C. et al. Depletion of carcinoma-associated fibroblasts and fibrosis induces immunosuppression and accelerates pancreas cancer with reduced survival. Cancer Cell 25, 719734 (2014).
  17. Rhim, A. D. et al. Stromal elements act to restrain, rather than support, pancreatic ductal adenocarcinoma. Cancer Cell 25, 735747 (2014).
  18. Tsai, J. H., Donaher, J. L., Murphy, D. A., Chau, S. & Yang, J. Spatiotemporal regulation of epithelial-mesenchymal transition is essential for squamous cell carcinoma metastasis. Cancer Cell 22, 725736 (2012).
  19. Ocaña, O. H. et al. Metastatic colonization requires the repression of the epithelial-mesenchymal transition inducer Prrx1. Cancer Cell 22, 709724 (2012).
  20. Korpal, M. et al. Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization. Nat. Med. 17, 11011108 (2011).
  21. Raj, D., Aicher, A. & Heeschen, C. Concise review: stem cells in pancreatic cancer: from concept to translation. Stem Cells 33, 28932902 (2015).
  22. Vannier, C., Mock, K., Brabletz, T. & Driever, W. Zeb1 regulates E-cadherin and Epcam (epithelial cell adhesion molecule) expression to control cell behavior in early zebrafish development. J. Biol. Chem. 288, 1864318659 (2013).
  23. Dosch, J. S., Ziemke, E. K., Shettigar, A., Rehemtulla, A. & Sebolt-Leopold, J. S. Cancer stem cell marker phenotypes are reversible and functionally homogeneous in a preclinical model of pancreatic cancer. Cancer Res. 75, 45824592 (2015).
  24. Wellner, U. et al. The EMT-activator ZEB1 promotes tumorigenicity by repressing stemness-inhibiting microRNAs. Nat. Cell Biol. 11, 14871495 (2009).
  25. Herreros-Villanueva, M. et al. SOX2 promotes dedifferentiation and imparts stem cell-like features to pancreatic cancer cells. Oncogenesis 2, e61 (2013).
  26. Sanada, Y. et al. Histopathologic evaluation of stepwise progression of pancreatic carcinoma with immunohistochemical analysis of gastric epithelial transcription factor SOX2: comparison of expression patterns between invasive components and cancerous or nonneoplastic intraductal components. Pancreas 32, 164170 (2006).
  27. Singh, S. K. et al. Antithetical NFATc1–Sox2 and p53–miR200 signaling networks govern pancreatic cancer cell plasticity. EMBO J. 34, 517530 (2015).
  28. Muller, P. A. J. & Vousden, K. H. Mutant p53 in cancer: new functions and therapeutic opportunities. Cancer Cell 25, 304317 (2014).
  29. Rivlin, N., Koifman, G. & Rotter, V. p53 orchestrates between normal differentiation and cancer. Semin. Cancer Biol. 32, 1017 (2015).
  30. Morton, J. P. et al. Mutant p53 drives metastasis and overcomes growth arrest/senescence in pancreatic cancer. Proc. Natl Acad. Sci. USA 107, 246251 (2010).
  31. Lehmann, W. et al. ZEB1 turns into a transcriptional activator by interacting with YAP1 in aggressive cancer types. Nat. Commun. 7, 10498 (2016).
  32. Mock, K. et al. The EMT-activator ZEB1 induces bone metastasis associated genes including BMP-inhibitors. Oncotarget 6, 1439914412 (2015).
  33. Brabletz, S. et al. The ZEB1/miR-200 feedback loop controls Notch signalling in cancer cells. EMBO J. 30, 770782 (2011).
  34. Singh, A. et al. A gene expression signature associated with “K-Ras addiction reveals regulators of EMT and tumor cell survival. Cancer Cell 15, 489500 (2009).
  35. Nakamura, T., Fidler, I. J. & Coombes, K. R. Gene expression profile of metastatic human pancreatic cancer cells depends on the organ microenvironment. Cancer Res. 67, 139148 (2007).
  36. Collisson, E. A. et al. Subtypes of pancreatic ductal adenocarcinoma and their differing responses to therapy. Nat. Med. 17, 500503 (2011).
  37. Weissmueller, S. et al. Mutant p53 drives pancreatic cancer metastasis through cell-autonomous PDGF receptor β signaling. Cell 157, 382394 (2014).
  38. Diepenbruck, M. & Christofori, G. Epithelial–mesenchymal transition (EMT) and metastasis: yes, no, maybe? Curr. Opin. Cell Biol. 43, 713 (2016).
  39. Ye, X. & Weinberg, R. A. Epithelial–mesenchymal plasticity: a central regulator of cancer progression. Trends Cell Biol. 25, 675686 (2015).
  40. Bellomo, C., Caja, L. & Moustakas, A. Transforming growth factor [beta] as regulator of cancer stemness and metastasis. Br. J. Cancer 115, 761769 (2016).
  41. Korpal, M. & Kang, Y. Targeting the transforming growth factor-β signalling pathway in metastatic cancer. Eur. J. Cancer 46, 12321240 (2010).
  42. Lehuédé, C., Dupuy, F., Rabinovitch, R., Jones, R. G. & Siegel, P. M. Metabolic plasticity as a determinant of tumor growth and metastasis. Cancer Res. 76, 52015208 (2016).
  43. Nieto, M. A. Epithelial plasticity: a common theme in embryonic and cancer cells. Science 342, 1234850 (2013).
  44. Brabletz, T. et al. Variable beta-catenin expression in colorectal cancer indicates a tumor progression driven by the tumor environment. Proc. Natl Acad. Sci. USA 98, 1035610361 (2001).
  45. Chaffer, C. L. et al. Normal and neoplastic nonstem cells can spontaneously convert to a stem-like state. Proc. Natl Acad. Sci. USA 108, 79507955 (2011).
  46. Puisieux, A., Brabletz, T. & Caramel, J. Oncogenic roles of EMT-inducing transcription factors. Nat. Cell Biol. 16, 488494 (2014).
  47. Chaffer, C. L. et al. Poised chromatin at the ZEB1 promoter enables breast cancer cell plasticity and enhances tumorigenicity. Cell 154, 6174 (2013).
  48. Gruber, R. et al. YAP1 and TAZ control pancreatic cancer initiation in mice by direct up-regulation of JAK–STAT3 signaling. Gastroenterology 151, 526539 (2016).
  49. Zhang, W. et al. Downstream of mutant KRAS, the transcription regulator YAP Is essential for neoplastic progression to pancreatic ductal adenocarcinoma. Sci. Signal. 7, ra42-ra42 (2014).
  50. Moffitt, R. A. et al. Virtual microdissection identifies distinct tumor- and stroma-specific subtypes of pancreatic ductal adenocarcinoma. Nat. Genet. 47, 11681178 (2015).
  51. Galvan, J. A. et al. Expression of E-cadherin repressors SNAIL, ZEB1 and ZEB2 by tumour and stromal cells influences tumour-budding phenotype and suggests heterogeneity of stromal cells in pancreatic cancer. Br. J. Cancer 112, 19441950 (2015).
  52. Bronsert, P. et al. Prognostic significance of Zinc finger E-box binding homeobox 1 (ZEB1) expression in cancer cells and cancer-associated fibroblasts in pancreatic head cancer. Surgery 156, 97108 (2014).
  53. Singh, A. & Settleman, J. EMT, cancer stem cells and drug resistance: an emerging axis of evil in the war on cancer. Oncogene 29, 47414751 (2010).
  54. Eser, S., Schnieke, A., Schneider, G. & Saur, D. Oncogenic KRAS signalling in pancreatic cancer. Br. J. Cancer 111, 817822 (2014).
  55. Ni, T. et al. Snail1-dependent p53 repression regulates expansion and activity of tumour-initiating cells in breast cancer. Nat. Cell Biol. 18, 12211232 (2016).
  56. Turajlic, S. & Swanton, C. Metastasis as an evolutionary process. Science 352, 169175 (2016).
  57. Caramel, J. et al. A switch in the expression of embryonic EMT-inducers drives the development of malignant melanoma. Cancer Cell 24, 466480 (2013).
  58. Denecker, G. et al. Identification of a ZEB2-MITF-ZEB1 transcriptional network that controls melanogenesis and melanoma progression. Cell Death Differ. 21, 12501261 (2014).
  59. Ye, X. et al. Distinct EMT programs control normal mammary stem cells and tumour-initiating cells. Nature 525, 256260 (2015).
  60. Tiwari, N. et al. Sox4 is a master regulator of epithelial-mesenchymal transition by controlling Ezh2 expression and epigenetic reprogramming. Cancer Cell 23, 768783 (2013).
  61. Tran, H. D. et al. Transient SNAIL1 expression is necessary for metastatic competence in breast cancer. Cancer Res. 74, 63306340 (2014).
  62. Olive, K. P. et al. Mutant p53 gain of function in two mouse models of Li-Fraumeni syndrome. Cell 119, 847860 (2004).
  63. Jackson, E. L. et al. Analysis of lung tumor initiation and progression using conditional expression of oncogenic K-ras. Genes Dev. 15, 32433248 (2001).
  64. Hingorani, S. R. et al. Preinvasive and invasive ductal pancreatic cancer and its early detection in the mouse. Cancer Cell 4, 437450 (2003).
  65. Chan, I. T. et al. Conditional expression of oncogenic K-ras from its endogenous promoter induces a myeloproliferative disease. J. Clin. Invest. 113, 528538 (2004).
  66. Novak, A., Guo, C., Yang, W., Nagy, A. & Lobe, C. G. Z/EG, a double reporter mouse line that expresses enhanced green fluorescent protein upon Cre-mediated excision. Genesis 28, 147155 (2000).
  67. Murray, S. A., Carver, E. A. & Gridley, T. Generation of a Snail1 (Snai1) conditional null allele. Genesis 44, 711 (2006).
  68. Meidhof, S. et al. ZEB1-associated drug resistance in cancer cells is reversed by the class I HDAC inhibitor mocetinostat. EMBO Mol. Med. 7, 831847 (2015).
  69. Dunning, M. J., Smith, M. L., Ritchie, M. E. & Tavare, S. beadarray: R classes and methods for Illumina bead-based data. Bioinformatics 23, 21832184 (2007).
  70. Johnson, W. E., Li, C. & Rabinovic, A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 8, 118127 (2007).

Download references

Author information

  1. These authors jointly supervised this work.

    • Marc P. Stemmler &
    • Thomas Brabletz

Affiliations

  1. Department of Experimental Medicine 1, Nikolaus-Fiebiger-Center for Molecular Medicine, FAU University Erlangen-Nürnberg, Glückstrasse 6, 91054 Erlangen, Germany

    • Angela M. Krebs,
    • María Lasierra Losada,
    • Simone Brabletz,
    • Marc P. Stemmler &
    • Thomas Brabletz
  2. University of Freiburg, Faculty of Biology, Schaenzlestrasse 1, 79104 Freiburg, Germany

    • Angela M. Krebs
  3. German Cancer Consortium (DKTK), 79106 Freiburg, Germany

    • Angela M. Krebs,
    • Melanie Boerries,
    • Hauke Busch,
    • Wilfried Reichardt &
    • Peter Bronsert
  4. German Cancer Research Center (DKFZ), 69121 Heidelberg, Germany

    • Angela M. Krebs,
    • Melanie Boerries,
    • Hauke Busch,
    • Wilfried Reichardt &
    • Peter Bronsert
  5. Department of General and Visceral Surgery, University of Freiburg Medical Center, Hugstetter Strasse 55, 79106 Freiburg, Germany

    • Julia Mitschke &
    • Otto Schmalhofer
  6. Systems Biology of the Cellular Microenvironment Group, Institute of Molecular Medicine and Cell Research (IMMZ), Albert-Ludwigs-University Freiburg, Stefan-Meier-Strasse 17, 79106 Freiburg, Germany

    • Melanie Boerries &
    • Hauke Busch
  7. Department of Internal Medicine 5, Hematology and Oncology, Universitätsklinikum, FAU University Erlangen-Nürnberg, Ulmenweg 18, 91054 Erlangen, Germany

    • Martin Boettcher &
    • Dimitrios Mougiakakos
  8. Department of Radiology Medical Physics, University Medical Center, Freiburg, Breisacher Strasse 60a, 79106 Freiburg, Germany

    • Wilfried Reichardt
  9. Institute of Pathology and Comprehensive Cancer Center, University Medical Center, Freiburg, Breisacher Strasse 115a, 79106 Freiburg, Germany

    • Peter Bronsert
  10. Edinburgh Cancer Research Centre, Institute of Genetics & Molecular Medicine, University of Edinburgh, Crewe Road South, Edinburgh EH4 2XR, UK

    • Valerie G. Brunton
  11. Department of Surgery, Universitätsklinikum, FAU University Erlangen-Nürnberg, Ulmenweg 18, 91054 Erlangen, Germany

    • Christian Pilarsky
  12. Chair of Genetics, Nikolaus-Fiebiger-Center for Molecular Medicine, Department Biology, FAU University Erlangen-Nürnberg, Glückstrasse 6, 91054 Erlangen, Germany

    • Thomas H. Winkler

Contributions

A.M.K. planned and carried out experiments and wrote the manuscript. J.M. carried out mouse experiments. M.L.L. carried out drug studies. O.S. generated the floxed Zeb1 allele. M.B. and H.B. carried out bioinformatics analyses. M.B. and D.M. carried out metabolic tests. W.R. carried out MRI analyses. P.B. carried out histological analyses. V.G.B. established mouse models. C.P. generated cell lines. T.H.W. carried out mouse experiments. S.B. generated the floxed Zeb1 allele, and planned and carried out experiments. M.P.S. generated the floxed Zeb1 allele, planned and carried out mouse experiments, was involved in coordination and wrote the manuscript. T.B. planned and coordinated the project, analysed data and wrote the manuscript. M.P.S. and T.B. contributed equally and share senior authorship.

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

Author details

Supplementary information

Supplementary Figures

  1. Supplementary Figure 1: Characterisation of KPC, heterozygously and homozygously Zeb1 depleted KPC tumours. (351 KB)

    (a) Representative Zeb1-immunolabeling of a GFP lineage-traced primary tumour showing Zeb1/GFP double-positive tumour cells (arrows). n = 5 independent tumors. Scale bar, 50μm. (b) Representative consecutive sections of HE and indicated immunohistochemical stainings of four Zeb1 expressing KPC tumours demonstrating the heterogeneity in phenotype, grading and marker expression. A representative differentiated Zeb1-negative KPCZ tumour is shown for comparison. Arrows indicate Zeb1 positive tumour cells in the differentiated KPC tumour. n = 15 KPC, 13 KPCZ independent tumours. Scale bar, 100μm. (c) Tumour-free survival of KPC mice vs. KPC mice with a heterozygous deletion of Zeb1 (KPCz) (n = 15 KPC, 16 KPCz independent tumours); log-rank (Mantel-Cox) test); tumour volume (0 = start of MRI measurements, n = 12 KPC, 14 KPCz independent tumours); error bars show mean ± S.E.M.; multiple t-tests with correction for multiple comparison using the Holm-Sidak method; grading, local invasion and relative ECM deposition of the respective tumours (n = 31 KPC, 17 KPCz; Mann-Whitney test (two-tailed); percentage of metastasized tumours (n = 35 KPC, 17 KPCz independent tumours; Chi-square test (two-tailed); n.s. = not significant.

  2. Supplementary Figure 2: Characterisation of KPC vs. KPCZ tumours. (1,366 KB)

    Representative images of immunohistochemical and histological stainings of KPC and KPCZ tumours and quantifications of the indicated markers are given. Asterisks label Zeb1-expressing stroma cells in KPCZ tumours. Specific blue MTS staining labels collagen fibres. Scale bars, 100μm, for lower left image 50μm. n = 48 KPC, 29 KPCZ independent tumours for Zeb1 and MTS; n = 15 independent tumours for KPC, 13 independent tumours for KPCZ for all other markers, error bars show mean ± S.D.; p < 0.0001, n.s. = not significant, Chi-square test (two-tailed) for Zeb1, E-cadherin and Sox2, unpaired Students t-test (two-tailed) for Ki67 and Casp3 (with Welchs correction), Mann-Whitney test (two-tailed) for ECM and CD31.

  3. Supplementary Figure 3: Characterisation of differentiaton markers in KPC vs. KPCZ tumours. (724 KB)

    (a) Representative images of positive and negative immunohistochemical stainings and statistical analysis for the indicated EMT-TFs. Scale bar, 150μm. n = 14 independent tumours for KPC, 13 independent tumours for KPCZ, Chi-square test (two-tailed); n.s. = not significant. (b) Representative images of immunohistochemical stainings and statistical analysis for expression of Gata6. Scale bar, 150μm. n = 14 independent tumours for KPC, 13 independent tumours for KPCZ; error bars show mean ± S.D.; Mann-Whitney test (two-tailed), p < 0.001. (c) Representative images of differentiated KPCZ and undifferentiated KPC primary tumours (PT) and corresponding metastases (Met) with the same phenotype. Immunohistochemical labelling of Zeb1 expressing tumour cells in the KPC PT and Met (arrows). L = liver or lung tissue. n = 19 KPC, 4 KPCZ independent tumours and corresponding metastases. Scale bar, 100μm.

  4. Supplementary Figure 4: Characterisation of KPC vs. KPCZ tumour derived cell lines. (891 KB)

    (a) Bright field image of primary cell lines from KPC and KPCZ tumours as well as HE stainings of the respective tumours after grafting in syngeneic mice and of the respective primary tumours are shown. Scale bars, 100μm for bright field, 75μm for HE stainings. (b) MTT viability assay for the isolated tumour cell lines after treatment with the indicated doses of gemcitabine and erlotinib. The calculated IC50 values for gemcitabine are shown. n = 3 biologically independent experiments, error bars show mean ± S.E.M. (c) Tumour onset after subcutaneous injection of 1 × 105 KPC and KPCZ cells into syngeneic mice. n = 4 mice per cell line, error bars show mean ± S.E.M. (d) Tumour grading, grading at invasive regions and relative ECM deposition of one representative tumour/cell line analysed in c) (n = 6 tumours for KPC, n = 5 tumours for KPCZ); error bars show mean ± S.D.;p < 0.05, p < 0.01, Mann-Whitney test (two-tailed).

  5. Supplementary Figure 5: Depletion of Zeb1 affects tumour promoting capacities. (396 KB)

    (a) Representative images of one visual field (n = 6 fields/cell line) showing GFP + cells/cell clusters in the lungs (green dots) 2 h after i.v. injection of KPC and KPCZ tumour cells and control lungs. Scale bar, 500μm. (b) No. of tumours after subcutaneous injection of the indicated cell numbers for the KPC and KPCZ tumour cell lines and calculated fraction of tumourigenic cells. inf = infinite, Chi-square test. (c) Representative images showing spheres of KPC and KPCZ tumour cells. Scale bar, 500μm and 50μm for higher magnifications. (d) Percentage of cells in KPC and KPCZ lines positive for the indicated markers or marker combinations; n = 2 biologically independent experiments, error bars show ± S.D. Source data see Supplementary Table 5, Statistics Source Data. Relative mRNA expression levels (qRT-PCR) of indicated genes, mRNA levels of KPC661 was set to 1; n = 3 biologically independent experiments, Mann-Whitney test (two-tailed), p < 0.05, p < 0.01, error bars show mean ± S.E.M.

  6. Supplementary Figure 6: Depletion of Zeb1 reduces early PanIN lesions. (817 KB)

    (a) Consecutive sections showing representative HE and PAS stainings of precancerous PanIN lesions in the pancreas of two different 6 month old KC and of one KCZ mice. Specific dark blue PAS staining indicates the mucin-rich PanIN lesions. Scale bars, 2.5 mm and 150μm for higher magnifications. Quantification of the PanIN area (% of pancreas area). n = 12 KC and 7 KCZ independent mice, error bars show mean ± S.D.; p < 0.01, unpaired Students t-test (two tailed) with Welchs correction. (b) Gene set enrichment analyses (GSEA) of transcriptome data from KPCZ vs. KPC cells reveals reduction of gene signatures associated with cancer mesenchymal transition and Zeb1 targets in KPCZ vs. KPC cell lines. NES = normalized enrichment score; FDR = false discovery rate.

  7. Supplementary Figure 7: Depletion of Zeb1 reduces tumour cell plasticity. (647 KB)

    (a) Relative mRNA expression levels (qRT-PCR) of indicated genes in KPC and KPCZ cell lines treated for different times with TGFβ (time points: 0, 6 h, 1, 3, 7, 14, 21 days). mRNA levels of cell line 661 at day 0 were set to 1. n = 3 biologically independent experiments, error bars show mean ± S.E.M. Statistical analysis is shown for the comparison of TGFβ treated to untreated samples (grey bars) of each individual cell line p < 0.05, p < 0.01, p < 0.001, p < 0.0001, unpaired Students t-test (one-tailed) Source data see Supplementary Table 5, Statistics Source Data. (b) Table showing log2FC in mRNA expression levels (microarray) of genes previously determined as common ZEB1/YAP targets in KPC and KPCZ cell lines upon TGFβ treatment for 14 days. (cut-off: adj. p-value < 0.05 and log2FC > 0.5). (c) Representative images of consecutive sections of immunohistochemistry for Ck19 and Zeb1 comparing the plasticity of Zeb1 expression in central and invasive tumour regions. Tumours derived from one KPC and one KPCZ cell line are shown. Asterisks label Zeb1 expression in stroma cells, arrows indicate Zeb1 expression in tumour cells at the invasive front. Ck19 expression is shown to identify cancer cells. n = 15 KPC, 13 KPCZ independent tumours, Scale bars, 50μm and 150μm for higher magnifications.

PDF files

  1. Supplementary Information (9.5 MB)

    Supplementary Information

Excel files

  1. Supplementary Table 1 (15 KB)

    Supplementary Information

  2. Supplementary Table 2 (385 KB)

    Supplementary Information

  3. Supplementary Table 3 (11 KB)

    Supplementary Information

  4. Supplementary Table 4 (12 KB)

    Supplementary Information

  5. Supplementary Table 5 (30 KB)

    Supplementary Information

Additional data