Flower isoforms promote competitive growth in cancer

Abstract

In humans, the adaptive immune system uses the exchange of information between cells to detect and eliminate foreign or damaged cells; however, the removal of unwanted cells does not always require an adaptive immune system1,2. For example, cell selection in Drosophila uses a cell selection mechanism based on ‘fitness fingerprints’, which allow it to delay ageing3, prevent developmental malformations3,4 and replace old tissues during regeneration5. At the molecular level, these fitness fingerprints consist of combinations of Flower membrane proteins3,4,6. Proteins that indicate reduced fitness are called Flower-Lose, because they are expressed in cells marked to be eliminated6. However, the presence of Flower-Lose isoforms at a cell’s membrane does not always lead to elimination, because if neighbouring cells have similar levels of Lose proteins, the cell will not be killed4,6,7. Humans could benefit from the capability to recognize unfit cells, because accumulation of damaged but viable cells during development and ageing causes organ dysfunction and disease8,9,10,11,12,13,14,15,16,17. However, in Drosophila this mechanism is hijacked by premalignant cells to gain a competitive growth advantage18. This would be undesirable for humans because it might make tumours more aggressive19,20,21. It is unknown whether a similar mechanism of cell-fitness comparison is present in humans. Here we show that two human Flower isoforms (hFWE1 and hFWE3) behave as Flower-Lose proteins, whereas the other two isoforms (hFWE2 and hFWE4) behave as Flower-Win proteins. The latter give cells a competitive advantage over cells expressing Lose isoforms, but Lose-expressing cells are not eliminated if their neighbours express similar levels of Lose isoforms; these proteins therefore act as fitness fingerprints. Moreover, human cancer cells show increased Win isoform expression and proliferate in the presence of Lose-expressing stroma, which confers a competitive growth advantage on the cancer cells. Inhibition of the expression of Flower proteins reduces tumour growth and metastasis, and induces sensitivity to chemotherapy. Our results show that ancient mechanisms of cell recognition and selection are active in humans and affect oncogenic growth.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Characterization of human Flower isoforms and competition between Winner and Loser cells.
Fig. 2: Flower isoforms and human cancer.
Fig. 3: Lose isoform in stroma and Win isoform in tumour cooperate to induce aggressive cancer growth.
Fig. 4: hFWE knockdown inhibits tumorigenicity and improves chemotherapy efficiency.

Data availability

All data generated or analysed during this study are included in the paper, its Extended Data and its Supplementary Information. The source data for Figs. 24 and Extended Data Figs. 110 are available online.

References

  1. 1.

    Medzhitov, R. & Janeway, C. A. Jr Decoding the patterns of self and nonself by the innate immune system. Science 296, 298–300 (2002).

  2. 2.

    Vivier, E. et al. Innate or adaptive immunity? The example of natural killer cells. Science 331, 44–49 (2011).

  3. 3.

    Merino, M. M. et al. Elimination of unfit cells maintains tissue health and prolongs lifespan. Cell 160, 461–476 (2015).

  4. 4.

    Merino, M. M., Rhiner, C., Portela, M. & Moreno, E. “Fitness fingerprints” mediate physiological culling of unwanted neurons in Drosophila. Curr. Biol. 23, 1300–1309 (2013).

  5. 5.

    Moreno, E., Fernandez-Marrero, Y., Meyer, P. & Rhiner, C. Brain regeneration in Drosophila involves comparison of neuronal fitness. Curr. Biol. 25, 955–963 (2015).

  6. 6.

    Rhiner, C. et al. Flower forms an extracellular code that reveals the fitness of a cell to its neighbors in Drosophila. Dev. Cell 18, 985–998 (2010).

  7. 7.

    Merino, M. M., Levayer, R. & Moreno, E. Survival of the fittest: essential roles of cell competition in development, aging, and cancer. Trends Cell Biol. 26, 776–788 (2016).

  8. 8.

    Gogna, R., Shee, K. & Moreno, E. Cell competition during growth and regeneration. Annu. Rev. Genet. 49, 697–718 (2015).

  9. 9.

    Di Gregorio, A., Bowling, S. & Rodriguez, T. A. Cell competition and its role in the regulation of cell fitness from development to cancer. Dev. Cell 38, 621–634 (2016).

  10. 10.

    Greaves, M. & Maley, C. C. Clonal evolution in cancer. Nature 481, 306–313 (2012).

  11. 11.

    Jacobs, K. B. et al. Detectable clonal mosaicism and its relationship to aging and cancer. Nat. Genet. 44, 651–658 (2012).

  12. 12.

    Kennedy, S. R., Loeb, L. A. & Herr, A. J. Somatic mutations in aging, cancer and neurodegeneration. Mech. Ageing Dev. 133, 118–126 (2012).

  13. 13.

    Laurie, C. C. et al. Detectable clonal mosaicism from birth to old age and its relationship to cancer. Nat. Genet. 44, 642–650 (2012).

  14. 14.

    López-Otín, C., Blasco, M. A., Partridge, L., Serrano, M. & Kroemer, G. The hallmarks of aging. Cell 153, 1194–1217 (2013).

  15. 15.

    Vanneste, E. et al. Chromosome instability is common in human cleavage-stage embryos. Nat. Med. 15, 577–583 (2009).

  16. 16.

    Vijg, J. Somatic mutations, genome mosaicism, cancer and aging. Curr. Opin. Genet. Dev. 26, 141–149 (2014).

  17. 17.

    Neves, J., Demaria, M., Campisi, J. & Jasper, H. Of flies, mice, and men: evolutionarily conserved tissue damage responses and aging. Dev. Cell 32, 9–18 (2015).

  18. 18.

    Levayer, R., Hauert, B. & Moreno, E. Cell mixing induced by myc is required for competitive tissue invasion and destruction. Nature 524, 476–480 (2015).

  19. 19.

    Klein, C. A. Selection and adaptation during metastatic cancer progression. Nature 501, 365–372 (2013).

  20. 20.

    Maruyama, T. & Fujita, Y. Cell competition in mammals—novel homeostatic machinery for embryonic development and cancer prevention. Curr. Opin. Cell Biol. 48, 106–112 (2017).

  21. 21.

    Moreno, E. Is cell competition relevant to cancer? Nat. Rev. Cancer 8, 141–147 (2008).

  22. 22.

    ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012).

  23. 23.

    Ji, X. et al. Chromatin proteomic profiling reveals novel proteins associated with histone-marked genomic regions. Proc. Natl Acad. Sci. USA 112, 3841–3846 (2015).

  24. 24.

    Moreno, E. & Rhiner, C. Darwin’s multicellularity: from neurotrophic theories and cell competition to fitness fingerprints. Curr. Opin. Cell Biol. 31, 16–22 (2014).

  25. 25.

    Chang, H. F. et al. Cytotoxic granule endocytosis depends on the Flower protein. J. Cell Biol. 217, 667–683 (2018).

  26. 26.

    Petrova, E., López-Gay, J. M., Rhiner, C. & Moreno, E. Flower-deficient mice have reduced susceptibility to skin papilloma formation. Dis. Model. Mech. 5, 553–561 (2012).

  27. 27.

    Xue, L. et al. Voltage-dependent calcium channels at the plasma membrane, but not vesicular channels, couple exocytosis to endocytosis. Cell Rep. 1, 632–638 (2012).

  28. 28.

    Zhang, P. et al. Curcumin synergizes with 5-fluorouracil by impairing AMPK/ULK1-dependent autophagy, AKT activity and enhancing apoptosis in colon cancer cells with tumor growth inhibition in xenograft mice. J. Exp. Clin. Cancer Res. 36, 190 (2017).

  29. 29.

    Davies, A. H., Wang, Y. & Zoubeidi, A. Patient-derived xenografts: A platform for accelerating translational research in prostate cancer. Mol. Cell. Endocrinol. 462, 17–24 (2018).

  30. 30.

    Chang, C. E. et al. Novel application of pluronic lecithin organogels (PLOs) for local delivery of synergistic combination of docetaxel and cisplatin to improve therapeutic efficacy against ovarian cancer. Drug Deliv. 25, 632–643 (2018).

  31. 31.

    Sanjana, N. E., Shalem, O. & Zhang, F. Improved vectors and genome-wide libraries for CRISPR screening. Nat. Methods 11, 783–784 (2014).

  32. 32.

    Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science 343, 84–87 (2014).

  33. 33.

    Prasad, M. et al. Dual TGFβ/BMP pathway inhibition enables expansion and characterization of multiple epithelial cell types of the normal and cancerous breast. Mol. Cancer Res. 17,1556–1570 (2019).

  34. 34.

    Yanagida, J. et al. Accelerated elimination of ultraviolet-induced DNA damage through apoptosis in CDC25A-deficient skin. Carcinogenesis 33, 1754–1761 (2012).

  35. 35.

    Schmittgen, T. D. & Livak, K. J. Analyzing real-time PCR data by the comparative C(T) method. Nat. Protoc. 3, 1101–1108 (2008).

  36. 36.

    Zubeldia-Plazaola, A. et al. Comparison of methods for the isolation of human breast epithelial and myoepithelial cells. Front. Cell Dev. Biol. 3, 32 (2015).

  37. 37.

    Jivrajani, M., Shaikh, M. V., Shrivastava, N. & Nivsarkar, M. An improved and versatile immunosuppression protocol for the development of tumor xenograft in mice. Anticancer Res. 34, 7177–7183 (2014).

  38. 38.

    Gogna, R., Madan, E., Kuppusamy, P. & Pati, U. Chaperoning of mutant p53 protein by wild-type p53 protein causes hypoxic tumor regression. J. Biol. Chem. 287, 2907–2914 (2012).

  39. 39.

    Hadjal, Y., Hadadeh, O., Yazidi, C. E., Barruet, E. & Binétruy, B. A p38MAPK-p53 cascade regulates mesodermal differentiation and neurogenesis of embryonic stem cells. Cell Death Dis. 4, e737 (2013).

  40. 40.

    Madan, E. et al. SCO2 induces p53-mediated apoptosis by Thr845 phosphorylation of ASK-1 and dissociation of the ASK-1-Trx complex. Mol. Cell. Biol. 33, 1285–1302 (2013).

  41. 41.

    Annunziato, S. et al. Modeling invasive lobular breast carcinoma by CRISPR/Cas9-mediated somatic genome editing of the mammary gland. Genes Dev. 30, 1470–1480 (2016).

  42. 42.

    Akhtar, J., Wang, Z., Yu, C. & Zhang, Z. P. Effectiveness of local injection of lentivirus-delivered stathmin1 and stathmin1 shRNA in human gastric cancer xenograft mouse. J. Gastroenterol. Hepatol. 29, 1685–1691 (2014).

  43. 43.

    Zheng, J. Y. et al. Regression of prostate cancer xenografts by a lentiviral vector specifically expressing diphtheria toxin A. Cancer Gene Ther. 10, 764–770 (2003).

  44. 44.

    Madan, E. et al. TIGAR induces p53-mediated cell-cycle arrest by regulation of RB-E2F1 complex. Br. J. Cancer 107, 516–526 (2012).

Download references

Acknowledgements

This study was supported by ERC, SNSF, Josef Steiner Cancer Research Foundation, Swiss Cancer League and Champalimaud Foundation to E.M.; and Swiss Cancer League, LB692, LB506, Seeds of Science, Winthrop P Rockefeller Cancer Institute and Creighton University startup funds to R.G. We thank K. Polyak for experimental suggestions; J. Billheimer and A. Dhiman for help with analysis; A. Gogna for support; and the MTT platform (R. Tomás), the Histopathology Platform (I. Terras Marques, M. I. Romano, S. Casimiro and S. Dias) and the Rodent platform at the Champalimaud Centre for the Unknown.

Author information

R.G. and E. Moreno conceptualized the idea, initiated the project, provided research support, designed, analysed, and interpreted the experiments and wrote the manuscript. E. Madan analysed the experiments, wrote the manuscript, and performed imaging, qPCR, flow-cytometry and mouse studies. R.G. designed and analysed the experiments, wrote the manuscript, created molecular biology design, and performed imaging analysis, clonogenicity assay, FISH-molecular design, IHC, mouse studies, and chemotherapy designs. C.J.P. analysed the experiments, helped in the preparation of figures, formatted and wrote the manuscript and performed statistical analysis, cell culture experiments, calcium imaging, low-density plating and conditioned medium experiments, and knockout and wild-type cell interactions. M.N. analysed the experiments, helped in the preparation of figures, and performed statistical analysis, flow-cytometry experiments and analysis, and FISH. T.M.P. analysed the experiments, helped in preparation of figures, formatted and wrote the manuscript, performed statistical analysis and cell culture experiments, extracted and cultured normal and epithelial cells from breast tissue, and performed laser-capture microdissection, qPCR, flow-cytometry and virus production. R.C.-M., A.G., V.H., Y.Y., T.Y., D.S., S.R.P. and H.Y. assisted in procuring tissue samples and IHC. K.F. performed IHC. K.S., A.M.P. and D.C. assisted with experiments and analysis. C.R. assisted with manuscript writing. M.A.F.P. and A.V. assisted with flow-cytometry experiments. D.A. assisted with confocal microscopy. H.N. provided normal breast cells. L.A.H. provided SCC samples, supervised K.F. for IHC in SCC and edited the manuscript. P.K. provided research support, and helped with data analysis, organization, and editing of the manuscript. C.C. and A.L.B. provided a clinical perspective, and helped with the procurement of tissue samples, design of IHC experiments, and design of clinical experiments such as the inclusion of chemotherapy.

Correspondence to Rajan Gogna or Eduardo Moreno.

Ethics declarations

Competing interests

: The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Knockout of the hFWE gene in MCF-7 cells does not affect their cellular functions.

a, ENCODE data mining for chromatin immunoprecipitation (ChIP) with H3K2Ac antibody shows active euchromatin status of the hFWE promoter. b, ENCODE data mining for DNAase-I hypersensitivity analysis shows footprints of DNA unwinding at the locations of hFWE exons. c, Phylogenetic analysis. Exon sequences are highly conserved amongst vertebrates except exon 5, which is more specific to mammals. d, Pictorial representation of the protocol for the functional analysis of hFWE isoforms in cell culture. MCF-7 hFWEKO cells were infected with lentiviruses encoding each hFWE isoform alongside an independent GFP or RFP tag. Transduced cells were sorted for RFP+ and GFP+ populations, plated together and co-cultured for 24 h. Co-cultured cells were then studied for a further 24 h using live-cell imaging and flow cytometry. e, To obtain human cells expressing single hFWE isoforms, MCF-7 cells (breast cancer origin) were selected and MCF-7 hFWEKO cells were generated. qPCR analysis of relative transcript synthesis of hFWE exon 1 from MCF-7 hFWEWT and MCF-7 hFWEKO cells shows lack of gene product in the knockout cells (n = 3 biologically independent experiments). f, MTT assay shows that knockout of the hFWE gene does not affect cell viability and mitochondrial activity in MCF-7 cells (n = 3 biologically independent experiments). g, BrdU assay shows that knockout of the hFWE gene does not affect cellular proliferation rates of MCF-7 cells (n = 7 biologically independent experiments). h, Cell-cycle distribution was examined by analysing DNA content. Propidium iodide staining and subsequent flow cytometric analysis show that knockout of the hFWE gene does not affect the cell-cycle progression of the MCF-7 cells (n = 3 biologically independent experiments). i, Flow cytometric analysis of ROS via measurement of DCF fluorescence shows that knockout of the hFWE gene does not alter cellular ROS in MCF-7 cells (n = 4 biologically independent experiments). j, Annexin-V staining shows that knockout of the hFWE gene does not affect cellular apoptosis in MCF-7 cells (n = 4 biologically independent experiments). k, The clonogenic assay shows that knockout of the hFWE gene does not affect the ability of MCF-7 cells to form colonies (n = 3). l, Diagrammatic representation of the hFWEX–IRES–GFP/RFP lentiviral constructs. Eight lentiviral constructs were prepared by cloning hFWE1/2/3/4–IRES–GFP/RFP into pD2109-CMV lentiviral vectors. m, Eight additional constructs were prepared by cloning hFWE1/2/3/4 into dual promoter pCDH-CMV-MCS-EF1α-copGFP and pCDH-CMV-MCS-EF1α-copRFP (System Biosciences), respectively. ej, P values shown, two-tailed t-test, mean ± s.d. Source Data

Extended Data Fig. 2 Overexpression of single hFWE isoforms in MCF-7 hFWEKO cells does not alter viability.

a, MCF-7 hFWEKO cells were infected with each hFWE1/2/3/4–GFP/RFP construct independently and the cells were monitored via live-cell imaging. Images at 24 h show that MCF-7 hFWEKO cells homogenously expressing single hFWE isoforms do not undergo cell death (n = 3 biologically independent experiments, P values shown, two-tailed t-test). b, Analysis of live-cell imaging is shown as heat maps to represent the total number of cells, number of cells dying every 1.5 h, average cellular volume and average cellular sphericity; each block represents a gradient scale of low (blue), medium (yellow) and high (red) number or shape or size of the cells (n = 3, analysis performed using manual tool from Fiji and automated tools from Imaris, Genie tool used for representing data as heat maps). c, qPCR analysis of MCF-7 hFWEKO cells expressing each of four hFWE isoforms showed comparable lentiviral-mediated expression of hFWE1, hFWE2, hFWE3 or hFWE4 (n = 3 biologically independent experiments, mean ± s.d., ANOVA showed no significant differences). d, Flow cytometry-based annexin-V staining shows that overexpression of individual hFWE1/2/3/4–GFP/RFP isoforms does not induce apoptosis in MCF-7 hFWEKO cells (n = 3 biologically independent experiments, mean ± s.d., ANOVA showed no significant differences). e, Annexin-V staining and flow cytometry-based cell imaging (image stream) also show that overexpression of individual hFWE isoforms does not induce apoptosis (n = 3). f, Long-term effect of overexpression of single hFWE isoforms with GFP or RFP reporters over 21 days using clonogenic assay. The overexpression of hFWE isoforms does not affect the colony formation ability of MCF-7 hFWEKO cells (n = 3). Source Data

Extended Data Fig. 3 Live-cell imaging of co-culture assay using MCF-7 hFWEKO cells transduced with hFWE isoforms.

a, Results from 24-h live-cell imaging experiments of co-cultures of MCF-7 hFWEKO cells expressing the four hFWE isoforms. In [1], MCF-7 hFWEKO cells expressing hFWE1–IRES–GFP were co-cultured with cells expressing hFWE1–IRES–RFP. GFP+ and RFP+ cells were monitored at 0 and 24 h to follow the effects of hFWE isoforms on cell proliferation. The ratio of GFP+ to RFP+ cells did not vary significantly between 0 and 24 h. In co-culture experiments [2], MCF-7 hFWEKO cells expressing hFWE1–IRES–GFP were co-cultured with cells expressing hFWE2–IRES–RFP. The population of RFP+ cells was significantly higher at 24 h than at 0 h, indicating competition between hFWE1–IRES–GFP and hFWE2–IRES–RFP cells. Each co-culture combination is presented amongst the four hFWE isoforms along with IRES–GFP or IRES–RFP co-expression. Cells expressing hFWE2 or hFWE4 emerged as winners when co-cultured with cells expressing hFWE1 or hFWE3 regardless of GFP or RFP reporter. The ratio of GFP+ to RFP+ cells at 0 and 24 h for each co-culture experiment is presented quantitatively below. The ratios at 0 and 24 h for each combination were compared statistically using a two-tailed t-test assuming unequal variances; n = 4 biologically independent experiments, P values shown, mean ± s.d. b, Analysis of live-cell imaging is shown as heat maps to represent the total number of cells, the number of cells that died every 1.5 h, average cellular volume, and average cellular sphericity. The co-culture combinations are indicated on the left. For example, the co-culture combination of cells expressing hFWE1–IRES–GFP with cells expressing hFWE2–IRES–RFP results in the death of hFWE1–IRES–GFP cells at the expense of an increase in the number of hFWE2–IRES–RFP cells. Analysis of all co-culture combinations is presented. The data support the idea that cells expressing hFWE1 or hFWE3, when co-cultured with cells expressing hFWE2 or hFWE4, undergo cell death accompanied by loss of differentiated cellular architecture, indicated by decreased cellular volume (blue) and increased cellular sphericity (red). Each block represents a gradient scale of low (blue), medium (yellow) and high (red) for the number, size or shape of the cells (n = 4, analysis performed using the manual tool from Fiji and automated tools from Imaris, Genie tool used for representing data as heat maps). Source Data

Extended Data Fig. 4 High-resolution imaging of competition-induced Loser cell death.

a, High-resolution live-cell imaging experiment (28 h) showing cell competition in MCF-7 hFWEKO cells expressing hFWE1–IRES–RFP and hFWE2–IRES–GFP isoforms. The co-culture results show elimination of cells carrying the Lose isoform (hFWE1–IRES–GFP); n = 3 biologically independent experiments with similar results. b, The results of live-cell imaging were confirmed using annexin-V staining and flow-cytometry-based imaging of GFP+ and RFP+ cells for each co-culture combination. Cells were sorted following 24 h of co-culture, and the percentage of apoptotic cells is displayed. The flow-based imaging of these cells is also presented and shows no GFP+ signal in RFP-sorted cells or RFP+ signal in GFP-sorted cells. The annexin-V+ signal is shown in purple. For example, co-culture of cells expressing hFWE1–IRES–GFP with cells expressing hFWE2–IRES–RFP results in apoptosis of hFWE1–IRES–GFP cells, as indicated by the annexin-V+ signal. Together, results from live-cell imaging and flow cytometry demonstrate that cells expressing hFWELose isoforms (hFWE1 or hFWE3) undergo apoptosis when in competition with cells expressing hFWEWin isoforms (hFWE2 or hFWE4) (n = 3 biologically independent experiments, P values shown, two-tailed t-test, was performed for P-value calculations, mean ± s.d.). Source Data

Extended Data Fig. 5 Cells expressing hFWELose isoforms undergo caspase-dependent apoptosis during co-culture with cells expressing hFWEWin isoforms.

a, The effect of incubation with the pan-caspase inhibitor Z-VAD-FMK (20 µM) on co-culture of cells expressing hFWEWin or hFWELose isoforms. Caspase inhibition by Z-VAD-FMK rescued hFWELose cells expressing hFWE1 or hFWE3 from undergoing cell death induced by co-culture with hFWEWin cells expressing hFWE2 or hFWE4. The ratio of GFP+ to RFP+ cells at 0 and 24 h for each co-culture experiment is presented quantitatively below. The ratios at 0 and 24 h for each combination were compared statistically using a two-tailed t-test assuming unequal variances (n = 4 biologically independent experiments, P values shown, mean ± s.d.). b, Annexin-V staining and flow cytometry-based imaging of GFP+ and RFP+ cells for each co-culture combination confirms that incubation with Z-VAD-FMK prevents apoptosis in co-culture of cells expressing hFWEWin or hFWELose isoforms (n = 3 biologically independent experiments, P values shown, two-tailed t-test, mean ± s.d.). Source Data

Extended Data Fig. 6 Characterization of competition in co-culture assay of cells expressing hFWEWin or hFWELose isoforms.

a, Analysis of live-cell imaging is shown as heat maps to represent the total number of cells, number of cells that died every 1.5 h, average cellular volume, and average cellular sphericity. Each block represents a gradient scale of low (blue), medium (yellow) and high (red) number, shape or size of cells. Treatment with Z-VAD-FMK blocked hFWEWin cell-induced death of co-cultured hFWELose cells, resulting in equal ratios of GFP+ and RFP+ populations (n = 4, analysis performed using the manual tool from Fiji and automated tools from Imaris, Genie tool used for representing data as heat maps). b, Imaging results show that MCF-7 hFWEKO cells expressing each hFWE isoform independently do not outcompete co-cultured hFWEKO cells, consistent with a mechanism of hFWE-mediated cell death supported by the presence of both hFWEWin and hFWELose isoforms (n = 3). c, Culling of cells expressing hFWELose isoforms appeared to be contact dependent. Live-cell imaging was performed on low-density-plated MCF-7 hFWEKO cells expressing hFWE1–IRES–GFP and hFWE2–IRES–RFP. Results show cell proliferation and absence of cell death in either cell population under conditions of low-density plating (n = 3 biologically independent experiments). d, The potential effect of secreted factors from cells expressing hFWEWin isoforms was investigated. Conditioned medium from MCF-7 hFWEKO cells expressing hFWE2–IRES–GFP did not affect viability when transferred to a culture of MCF-7 hFWEKO cells expressing hFWE2–IRES–GFP and imaged over 24 h (n = 3). Source Data

Extended Data Fig. 7 Calcium-independent mechanism of hFWE-mediated cell competition.

a, The potential effect of individual hFWE isoforms on intracellular calcium levels was investigated. Live-cell imaging was performed to detect levels of Fluo-4AM in MCF-7 hFWEKO cells and MCF-7 hFWEKO cells overexpressing single hFWE isoforms (n = 3 biologically independent experiments, ANOVA showed no significant differences, mean ± s.d.). b, Live-cell imaging at 5-s intervals over 1 min showed constant levels of Fluo-4AM in MCF-7 hFWEKO cells and MCF-7 hFWEKO cells overexpressing single hFWE isoforms (n = 3 biologically independent experiments, ANOVA showed no significant differences, mean ± s.d.). c, hFWE-mediated cell competition assay carried out in the presence or absence of the calcium chelator BAPTA-AM. Treatment with BAPTA-AM did not affect the ability of MCF-7 hFWEKO cells expressing hFWE2 to kill co-cultured MCF-7 hFWEKO cells expressing hFWE1 (n = 3 biologically independent experiments, P values shown, two-tailed t-test found no significant differences, mean ± s.d.). d, Laser-capture microdissection was performed in increments of 200 µm in distance perpendicular to the defined boundary of breast tissue tumour mass. Representative images are shown for breast cancer tissue section before and after laser capture. hFWELose isoforms were upregulated in tumour-adjacent host tissue nearest to breast cancer (0–400 μm) or colon cancer (0–200 μm). hFWEWin isoforms were highly enriched within breast or colon tumour tissue but not adjacent tissue (n = 3 biologically independent experiments, mean ± s.d., fold change calculated relative to the external reference of the expression of hFWEWin isoforms in normal breast tissue, P values shown, two-tailed t-test). e, qPCR array was used to examine the expression of 354 genes involved in apoptotic pathways in samples of laser-captured tumour-adjacent host tissue. Gene expression heat map shows that host tissue in the immediate vicinity of cancer (0–400 μm) displayed upregulation of genes involved in apoptotic pathways. Host tissue farther away (400–1,000 μm) from the tumour edge displayed basal levels of apoptosis-related genes. The heat map shows apoptotic genes that were induced (red), suppressed (blue) or unchanged (yellow) in expression. Black represents unsuccessful runs (n = 3). f, FISH shows the expression of hFWEWin isoforms in FFPE samples of breast cancer tissue (top) and normal breast tissue (bottom). First column, H&E staining; second column, control DAPI staining; third column, expression of hFWEWin isoforms within tumour tissue; fourth column, overlay of DAPI and hFWEWin isoforms. Magnified images below each panel show expression of hFWEWin isoforms specifically in tumour tissue, as these isoforms are poorly expressed in the stromal tissue surrounding the tumour and normal breast tissue (n = 3 for all staining). g, Schematic depicting the process of epithelial cell isolation from normal breast, colon, and lung tissue. Expression of the four hFWE isoforms was compared in epithelia versus total tissue by qPCR analysis (n = 3 biologically independent experiments, all statistically significant P values shown, two-tailed t-test, mean ± s.d.). h, IHC staining of FFPE samples of breast cancer with newly developed antibodies against hFWE. Breast cancer and stromal regions are shown in the H&E images (left). The anti-hFWEWin antibody is specific to Win isoforms hFWE2 and hFWE4 whereas the anti-hFWE-N-term antibody targets the common N terminus and recognizes all four isoforms. Immunohistochemistry staining shows the abundance of hFWEWin isoforms within the breast cancer samples in human tumours. From this, we conclude that hFWEWin proteins are expressed in the tumour but not in the stroma. To demonstrate the expression of Lose isoforms in the stroma, we compared the staining of all hFWE isoforms using our N-terminal-specific antibody with the poor expression of hFWEWin in the stroma. The anti-hFWE-N-term antibody shows strong immunoreactivity and equal distribution of staining in the tumour and stromal tissue. We find no positive signal for hFWEWin isoforms in the stroma, but we find strong expression of total hFWE (hFWEWin and hFWELose combined) in the stroma near the tumour. From this, we can conclude that Lose isoforms are expressed in the stroma. This experiment was repeated independently three times with similar results. Source Data

Extended Data Fig. 8 Dynamics of hFWE-mediated cell competition between cancer and normal cells.

a, Protocol for observing hFWE-mediated cell competition between hFWEWin-expressing cancer cells and primary cultures of human epithelial cells and fibroblasts. MCF-7 hFWEKO cells overexpressing hFWEWin isoforms outcompeted and induced apoptosis of co-cultured primary breast normal epithelial cells or fibroblasts over 6 days (n = 3 biologically independent experiments, all statistically significant P values are shown, ANOVA, mean ± s.d.). b, hFWEWin isoforms (hFWE2 and hFWE4) and hFWELose isoforms (hFWE1 and hFWE3) are characterized by intra-exonic inclusions. The in silico model shows predicted transmembrane structures of the four hFWE proteins. An antibody was raised against the common N terminus (exon 1), which is included in all four isoforms. Another antibody was raised to specifically recognize the hFWEWin isoforms (which share exon 3). Owing to the similarity in sequences to hFWEWin isoforms and low antigenicity of the peptide sequence at the junction of exons 2 and 4, generation of an antibody specific to hFWELose isoforms was not possible. Source Data

Extended Data Fig. 9 Generation of antibodies against hFWEWin isoforms (hFWE2 and hFWE4) or the common hFWE N terminus and expression in human cancers.

a, Immunocytochemistry of hFWE isoforms in paraffin-embedded and sectioned human MCF-7 cells using novel anti-hFWE antibodies. The specificity was validated by immunocytochemistry of MCF-7 hFWEKO cells with or without overexpression of individual hFWE isoforms. Cells were stained with anti-hFWE-N-term antibody or anti-hFWEWin antibody. Immunocytochemistry results confirm antibody specificity in paraffin sections with MCF-7 hFWEKO cells with or without overexpression of individual hFWE isoforms. The anti-hFWEWin antibody does not stain sections with hFWEKO cells, and recognizes specifically hFWE2 and hFWE4, but not the hFWELose isoforms hFWE1 and hFWE3 (row 1). The anti-hFWE-N-term antibody is also negative for staining in control hFWEKO cells but detected positive membrane staining for each of the four hFWE isoforms (row 2; n = 3). b, Immunohistochemistry of hFWE proteins in sections from clinical breast cancer samples (n = 3). Arrows indicate membrane staining. c, Immunofluorescence staining of hFWE proteins in human SCC tissue shows an abundance of hFWEWin isoforms within defined cancer lesions, whereas the anti-hFWE-N-term antibody stained both stroma and cancer tissue (top; n = 3) when compared with normal skin (bottom; n = 3). d, Tumour volumes were measured and analysed for each of the 16 combinations of hFWEX–IRES–GFP expression in MCF-7 hFWEKO xenografted cells and hFWE isoform expression within FweKO mouse mammary tissue (n = 3). In vivo bioluminescence imaging (using the IVIS system) was used to detect and measure the fluorescence of GFP+ cancer cells in tumours resulting from all genetic combinations at 28 days post-implantation. Substantially reduced tumour growth was observed when mammary tissue expressed hFWEWin isoforms and cancer cells expressed hFWELose isoforms. By contrast, tumour growth is strongly promoted when mammary tissue expresses hFWELose isoforms and cancer cells express hFWEWin isoforms (n = 3). e, Tumorigenic potential of HCT-116 (colon origin) or MCF-7 (breast origin) hFWEKO cells overexpressing hFWE2 in recipient FweWT and FweKO mice. Tumour growth of both FweKO breast and colon cancer cell lines overexpressing the Win isoform hFWE2 was greater in FweWT mice than in FweKO mice. Photos of resected tumours are shown. f, Tumour volumes were measured every week over 42 days for groups shown in a. hFWEKO colon- or breast-derived tumours overexpressing hFWE2 showed significantly higher growth in FweWT mice than in FweKO mice (n = 5, P values shown, one-tailed t-test, mean ± s.d.). g, At the conclusion of tumour growth experiments, mice were examined for the presence or absence of metastases in inguinal lymph nodes (ILN), axillary lymph nodes (ALN), colon, pancreas, prostate, lung, and liver. Heat map scale indicates the probability of metastasis. Results show a marked reduction in the metastatic potential of both breast and colon cancer hFWEKO cells overexpressing hFWE2 when xenografted into FweKO mice as compared to FweWT mice (n = 5 each group). Source Data

Extended Data Fig. 10 Endogenous hFWE in human cancer cell lines and tumorigenic potential.

a, Culture experiments were conducted to examine the dynamics of hFWEWin and hFWELose isoform expression. MCF-7 hFWEKO cells expressing hFWE2–RFP were co-cultured with wild-type MCF-7 cells expressing GFP for 24 h and sorted for analysis of hFWE isoform expression. n = 3 biologically independent experiments with similar results. b, Co-culture of MCF-7 hFWEKO cells expressing hFWE2–RFP with wild-type MCF-7 cells expressing GFP caused upregulation of hFWELose isoforms in wild-type MCF-7 cells. qPCR analysis of the expression of hFWEWin and hFWELose isoforms in GFP+ wild-type MCF-7 cells sorted from co-culture shows a significant increase in hFWELose isoforms (bar 4) when compared with monocultured wild-type MCF-7 cells (bar 2). n = 3 biologically independent experiments with similar results; fold change calculated relative to the expression of hFWEWin isoforms in monocultured GFP+ wild-type MCF-7 cells, all statistically significant P values shown, two-tailed t-test, mean ± s.d. c, MCF-7 cells expressing hFWE2 were xenografted into FweWT mice to assess their tumorigenic potential and the host expression of endogenous hFWE isoforms compared to control. The mouse tissue adjacent to the tumour showed a significant increase in expression of hFWELose isoforms at 21 days post-xenograft. n = 3 biologically independent experiments with similar results; all statistically significant P values shown, one-tailed t-test, mean ± s.d. d, The effect of anti-hFWE shRNA cocktail on the tumorigenic potential of HCT-116 cells. Row 1, growth potential of HCT-116 hFWEWT cells. Row 2, knockout of hFWE in these cells significantly reduced tumour growth. Row 3, treatment of HCT-116 hFWEWT cells with anti-hFWE shRNA reduced tumour volume. Row 4, Rescue experiment in which similar hFWE shRNA-treated tumours to row 3 were infected with lentivirus overexpressing hFWE2 14 days after implantation. These tumours are significantly larger than those in row 3. e, Tumour volumes for experiments shown in Fig. 4a were measured weekly, and growth patterns were analysed over 42 days for groups shown in a. Growth curves show the reduced growth of tumours from HCT-116 hFWEKO cells (blue) and hFWE shRNA-treated HCT-116 hFWEWT cells (red). Green line shows rescue experiment and growth pattern changes in hFWE shRNA-treated HCT-116 hFWEWT tumours expressing with hFWE2 (n = 5, P values shown, one-tailed t-test, mean ± s.d.). f, All mice used in the study were examined for the presence or absence of metastases in ILN, ALN, colon, pancreas, prostate, lung and liver. Heat map scale indicates the probability of metastasis. Metastatic potential was reduced by knockout or knockdown of hFWE in HCT-116 cells (compare column 1 with columns 2 and 3). The rescue of tumour growth by re-introduction of hFWE2 cDNA was accompanied by an increase in metastasis of these cells (compare column 3 with column 4; n = 5 each group). g, A cocktail of shRNAs were designed to knock down all four isoforms of hFWE. All shRNAs were checked for off-target effects. h, Gene expression analysis confirmed deletion of total hFWE in HCT-116 hFWEKO cells. Exogenous expression of hFWE2 cDNA was detectable as the total hFWE expression in wild-type HCT-116 cells co-treated with anti-hFWE shRNA (observed in resected tumours). n = 3 biologically independent experiments, fold change calculated relative to expression of hFWEWin isoforms in MCF-10A cells, P values shown, two-tailed t-test, mean ± s.d. i, Tumorigenic potential of wild-type CCL-218 cells and CCL-218 cells overexpressing hFWE2. Xenografts overexpressing hFWE2 cDNA showed increased tumour volume at 28 days. Photos of resected tumours are shown. n = 3 biologically independent experiments with similar results, all statistically significant P values shown, ANOVA, mean ± s.d. Control qPCR experiment demonstrates the overexpression of hFWE2 in CCL-218 tumours at day 28. n = 3 biologically independent experiments, fold change calculated relative to expression of hFWEWin isoforms in MCF-10A cells, P values shown, two-tailed t-test, mean ± s.d. j, Endogenous expression of the four hFWE isoforms in 19 cancer cell lines of multiple origins (n = 3 biologically independent experiments, fold change is calculated relative to expression of hFWEWin isoforms in MCF-10A cells, P values shown, ANOVA, mean ± s.d.). k, Gene expression analysis shows efficient shRNA-mediated knockdown of total hFWE in HCT-116, DU-145, CCL-218 and OVCAR-8 tumours (observed in resected tumours). n = 3 biologically independent experiments, fold change calculated relative to expression of hFWEWin isoforms in MCF-10A cells, P values shown, ANOVA, mean ± s.d. Source Data

Supplementary information

Reporting Summary

Video 1

Live cell imaging (24 hr) showing the effect of the four hFwe(1/2/3/4)-IRES-GFP and four hFwe(1/2/3/4)-IRES-RFP lentiviral infections during monoculture of MCF-7 hFweKO cells. No significant cell death was observed upon overexpression of hFwe isoforms during monoculture, n=3 biologically independent experiments with similar results.

Video 2

Live cell imaging (24 hr) showing the effect of co-culture of MCF-7 hFweKO cells expressing hFwe1-GFP and hFwe1-RFP. No cell death was observed in either cell population, n=3 biologically independent experiments with similar results.

Video 3

Live cell imaging (24 hr) showing the effect of co-culture of MCF-7 hFweKO cells expressing hFwe1-GFP and hFwe2-RFP. Significant cell death was observed in cells expressing hFwe1 when co-cultured with cells expressing hFwe2, n=3 biologically independent experiments with similar results.

Video 4

High-resolution live cell imaging (24 hr) depicts the elimination of cells carrying hFwe1-IRES-RFP when co-cultured with cells expressing hFwe2-IRES-GFP, n=3 biologically independent experiments with similar results.

Video 5

High-resolution live cell imaging (24 hr) depicts the elimination of cells carrying hFwe1-IRES-GFP when co-cultured with cells expressing hFwe2-IRES-RFP, n=3 biologically independent experiments with similar results.

Video 6

Live cell imaging (24 hr) showing the effect of pan-caspase inhibitor Z-VAD-FMK on co-culture of MCF-7 hFweKO cells expressing hFwe1-GFP or hFwe1-RFP, n=3 biologically independent experiments with similar results.

Video 7

Live cell imaging (24 hr) showing the effect of pan-caspase inhibitor Z-VAD-FMK on co-culture of MCF-7 hFweKO cells expressing hFwe1-GFP or hFwe2-RFP. Caspase inhibition rescued cells expressing hFwe1 from being eliminated when co-cultured with cells expressing hFwe2, n=3 biologically independent experiments with similar results.

Video 8

Live cell imaging (24 hr) showing that hFwe expression is required in both winner and loser cells. MCF-7 hFweKO cells expressing hFwe1/2/3/4-GFP or do not eliminate MCF-7 hFweKO cells, n=3 biologically independent experiments with similar results.

Video 9

Live cell imaging (24 hr) of low-density plated hFwe1-GFP and hFwe2-RFP expressing MCF-7 hFweKO cells showing cell proliferation but not death in either population, n=3 biologically independent experiments with similar results.

Source data

Source Data Fig. 2

Source Data Fig. 3

Source Data Fig. 4

Source Data Extended Data Fig. 1

Source Data Extended Data Fig. 2

Source Data Extended Data Fig. 3

Source Data Extended Data Fig. 4

Source Data Extended Data Fig. 5

Source Data Extended Data Fig. 6

Source Data Extended Data Fig. 7

Source Data Extended Data Fig. 8

Source Data Extended Data Fig. 9

Source Data Extended Data Fig. 10

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.