Cell competition—the sensing and elimination of less fit ‘loser’ cells by neighbouring ‘winner’ cells—was first described in Drosophila. Although cell competition has been proposed as a selection mechanism to optimize tissue and organ development, its evolutionary generality remains unclear. Here, by using live imaging, lineage tracing, single-cell transcriptomics and genetics, we identify two cell competition mechanisms that sequentially shape and maintain the architecture of stratified tissue during skin development in mice. In the single-layered epithelium of the early embryonic epidermis, winner progenitors kill and subsequently clear neighbouring loser cells by engulfment. Later, as the tissue begins to stratify, the basal layer instead expels losers through upward flux of differentiating progeny. This cell competition switch is physiologically relevant: when it is perturbed, so too is barrier formation. Our findings show that cell competition is a selective force that optimizes vertebrate tissue function, and illuminate how a tissue dynamically adjusts cell competition strategies to preserve fitness as its architectural complexity increases during morphogenesis.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Single cell RNA sequencing data that support the findings of this study have been deposited in the Gene Expression Omnibus under accession number GSE128241. RNA sequencing data shown in Fig. 4b and Extended Data Fig. 1a were extracted from ref. 19 and were previously deposited in GEO with the accession number GSE75931. All other data are in the manuscript, Supplementary Materials and Source Data or are available from the corresponding author upon reasonable request.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Morata, G. & Ripoll, P. Minutes: mutants of Drosophila autonomously affecting cell division rate. Dev. Biol. 42, 211–221 (1975).
Simpson, P. & Morata, G. Differential mitotic rates and patterns of growth in compartments in the Drosophila wing. Dev. Biol. 85, 299–308 (1981).
Simpson, P. Parameters of cell competition in the compartments of the wing disc of Drosophila. Dev. Biol. 69, 182–193 (1979).
Moreno, E., Basler, K. & Morata, G. Cells compete for decapentaplegic survival factor to prevent apoptosis in Drosophila wing development. Nature 416, 755–759 (2002).
Johnston, L. A., Prober, D. A., Edgar, B. A., Eisenman, R. N. & Gallant, P. Drosophila myc regulates cellular growth during development. Cell 98, 779–790 (1999).
de la Cova, C., Abril, M., Bellosta, P., Gallant, P. & Johnston, L. A. Drosophila myc regulates organ size by inducing cell competition. Cell 117, 107–116 (2004).
Rhiner, C. et al. Persistent competition among stem cells and their daughters in the Drosophila ovary germline niche. Development 136, 995–1006 (2009).
Stine, R. R., Greenspan, L. J., Ramachandran, K. V. & Matunis, E. L. Coordinate regulation of stem cell competition by Slit-Robo and JAK-STAT signaling in the Drosophila testis. PLoS Genet. 10, e1004713 (2014).
Amoyel, M., Simons, B. D. & Bach, E. A. Neutral competition of stem cells is skewed by proliferative changes downstream of Hh and Hpo. EMBO J. 33, 2295–2313 (2014).
Clavería, C. & Torres, M. Cell competition: mechanisms and physiological roles. Annu. Rev. Cell Dev. Biol. 32, 411–439 (2016).
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).
Clavería, C., Giovinazzo, G., Sierra, R. & Torres, M. Myc-driven endogenous cell competition in the early mammalian embryo. Nature 500, 39–44 (2013).
Díaz-Díaz, C. et al. Pluripotency surveillance by myc-driven competitive elimination of differentiating cells. Dev. Cell 42, 585–599.e4 (2017).
Sancho, M. et al. Competitive interactions eliminate unfit embryonic stem cells at the onset of differentiation. Dev. Cell 26, 19–30 (2013).
Kon, S. et al. Cell competition with normal epithelial cells promotes apical extrusion of transformed cells through metabolic changes. Nat. Cell Biol. 19, 530–541 (2017).
Martins, V. C. et al. Cell competition is a tumour suppressor mechanism in the thymus. Nature 509, 465–470 (2014).
Bondar, T. & Medzhitov, R. p53-mediated hematopoietic stem and progenitor cell competition. Cell Stem Cell 6, 309–322 (2010).
Moreno, E. & Basler, K. dMyc transforms cells into super-competitors. Cell 117, 117–129 (2004).
Asare, A., Levorse, J. & Fuchs, E. Coupling organelle inheritance with mitosis to balance growth and differentiation. Science 355, eaah4701 (2017).
Beronja, S., Livshits, G., Williams, S. & Fuchs, E. Rapid functional dissection of genetic networks via tissue-specific transduction and RNAi in mouse embryos. Nat. Med. 16, 821–827 (2010).
Tallquist, M. D. & Soriano, P. Epiblast-restricted Cre expression in MORE mice: a tool to distinguish embryonic vs. extra-embryonic gene function. Genesis 26, 113–115 (2000).
Mesa, K. R. et al. Niche-induced cell death and epithelial phagocytosis regulate hair follicle stem cell pool. Nature 522, 94–97 (2015).
Tirosh, I. et al. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science 352, 189–196 (2016).
Fuchs, E. Scratching the surface of skin development. Nature 445, 834–842 (2007).
Williams, S. E., Beronja, S., Pasolli, H. A. & Fuchs, E. Asymmetric cell divisions promote Notch-dependent epidermal differentiation. Nature 470, 353–358 (2011).
Williams, S. E., Ratliff, L. A., Postiglione, M. P., Knoblich, J. A. & Fuchs, E. Par3-mInsc and Gαi3 cooperate to promote oriented epidermal cell divisions through LGN. Nat. Cell Biol. 16, 758–769 (2014).
Segre, J. A., Bauer, C. & Fuchs, E. Klf4 is a transcription factor required for establishing the barrier function of the skin. Nat. Genet. 22, 356–360 (1999).
Indra, A. K. & Leid, M. Epidermal permeability barrier measurement in mammalian skin. Methods Mol. Biol. 763, 73–81 (2011).
Lee, C. H. et al. A regulatory response to ribosomal protein mutations controls translation, growth, and cell competition. Dev. Cell 46, 456–469.e4 (2018).
Merino, M. M. et al. Elimination of unfit cells maintains tissue health and prolongs lifespan. Cell 160, 461–476 (2015).
Lynch, M. D. et al. Spatial constraints govern competition of mutant clones in human epidermis. Nat. Commun. 8, 1119 (2017).
Martincorena, I. et al. Tumor evolution. High burden and pervasive positive selection of somatic mutations in normal human skin. Science 348, 880–886 (2015).
Brown, S. et al. Correction of aberrant growth preserves tissue homeostasis. Nature 548, 334–337 (2017).
Muzumdar, M. D., Tasic, B., Miyamichi, K., Li, L. & Luo, L. A global double-fluorescent Cre reporter mouse. Genesis 45, 593–605 (2007).
Snippert, H. J. et al. Intestinal crypt homeostasis results from neutral competition between symmetrically dividing Lgr5 stem cells. Cell 143, 134–144 (2010).
Knoepfler, P. S., Cheng, P. F. & Eisenman, R. N. N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev. 16, 2699–2712 (2002).
Srinivas, S. et al. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1, 4 (2001).
Ouspenskaia, T., Matos, I., Mertz, A. F., Fiore, V. F. & Fuchs, E. WNT-SHH antagonism specifies and expands stem cells prior to niche formation. Cell 164, 156–169 (2016).
Picelli, S. et al. Smart-seq2 for sensitive full-length transcriptome profiling in single cells. Nat. Methods 10, 1096–1098 (2013).
Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).
Soneson, C., Love, M. I. & Robinson, M. D. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000 Res. 4, 1521 (2015).
Brennecke, P. et al. Accounting for technical noise in single-cell RNA-seq experiments. Nat. Methods 10, 1093–1095 (2013).
Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).
We thank I. Matos for assistance with live-imaging set-up; F. García-Quiroz for assistance with TEWL experiments; L. Polak, L. Hidalgo, and M. Sribour for animal assistance; R. Eisenman for Mycn floxed mice; A. Asare for validating Gulp1 shRNAs; L. Johnston, E. Bach, M. Simunovic, and all Fuchs laboratory members, but especially M. Laurin, A. Sendoel, S. Baksh, S. Gur-Cohen, S. Liu and K. Stewart for discussion and/or comments on the manuscript; Rockefeller University (RU) Bio-Imaging Resource Center for use of microscopes; RU Comparative Bioscience Center (AAALAC-accredited) for care of mice in accordance with National Institutes of Health (NIH) guidelines; and RU Flow Cytometry Resource Center for assistance with FACS. S.J.E. was supported by fellowships from the Human Frontiers Science Program (LT000907/2015) and RU Women & Science. N.C.G. holds a Burroughs Wellcome Fund Postdoctoral Enrichment Program Award and an NIH Postdoctoral Ruth L. Kirschstein National Research Service Award (F32CA221353). A.F.M. was supported by an Arnold O. Beckman Postdoctoral Fellowship. Y.G. is supported by a NIAMS 1K01AR072132-01A1 Career Development Award and an Irene Diamond Fund/AFAR Postdoctoral Transition Award in Aging. E.F. is an HHMI Investigator. This research was supported by the NIH (R01-AR27883 to E.F.).
Nature thanks Bruce Edgar, Emma Rawlins and the other anonymous reviewer(s) for their contribution to the peer review of this work.
Extended data figures and tables
Extended Data Fig. 1 Mycn is expressed in single-layered embryonic epidermis and its expression is halved in both Mycnfl/+;LV-CreRFP and Mycnfl/+;Meox2Cre mice with no compensatory upregulation of Myc.
a, Mycn and Myc expression in E10.5 epidermis by RNA-seq19. Mycn is expressed at higher levels than Myc. b, Fluorescence activated cell sorting (FACS) strategy to isolate LV-CreRFP-transduced wild-type and Mycn+/– basal keratinocytes from E12.5 embryonic skin. c, d, Quantitative PCR revealed that transcript levels were halved in Mycn+/– cells compared to wild-type at E12.5 using both LV-Cre (n = 2) and Meox2Cre (n = 4, unpaired two-tailed t-test) samples. No significant compensatory change in levels of Myc mRNA was observed. Embryos of each genotype within a single litter were pooled for cDNA collection. This was repeated to generate two biological replicates per genotype. Data shown as mean ± s.e.m. Source data
Extended Data Fig. 2 Live imaging reveals that loser cell death is preceded by increased contact with wild-type cells and suggests that engulfment mediates clearance of loser cell corpses.
a, Schematic showing planes imaged in time-lapse microscopy experiments. b–d, Extracted images showing frames from Supplementary Videos 2–4. Dying GFP+Mycn+/– cells are marked with an arrowhead in the first frame of fluorescence images (top two rows) and by an asterisk in the traced panels (bottom). Wild-type cells are RFP+. e, Images from the suprabasal plane extracted from the first and last time points of Supplementary Video 1, and corresponding to the still images shown in the first and last panels of Fig. 1g. The asterisk marks the position of the dying cell 11 μm below. No new cells enter the suprabasal plane through the time course of the video, supporting the conclusion that the cell dies and is eliminated within the plane of the epidermis and not extruded to the suprabasal plane. Scale bars, 10 μm.
Extended Data Fig. 3 A genetic model to switch the fate of Mycn+/− epidermal cells from losers to winners.
a, R26YFP reporter for Cre activity shows undesired mosaicism in some embryos, which were prescreened and excluded from further analysis. All Meox2Cre embryos in our study were subjected to this analysis. b, TUNEL stain of Mycn+/–;R26YFP embryos revealed neither TUNEL labelling in regions of ubiquitous Cre activity (as indicated by YFP expression), nor any marked differences in proliferation as assayed by EdU incorporation (n denotes number of images analysed from each genotype; control n = 4, Meox2Cre n = 7; two-tailed unpaired t-test). There also was no discernable effect on postnatal viability or growth (P15 males shown in bottom panels, genotypes as indicated). c, d, Extended fields of view (c) from images shown in Fig. 1m of whole-mount E12.5 back skins stained for DAPI (nuclei, blue), RFP (transduced cells, red) and TUNEL (dying cell corpses, white; arrowheads) for each of the indicated genotypes. An additional representative image and quantification (d) is also shown for a second Mycn shRNA (n denotes number of images quantified from two litters per genotype; Mycn+/−/Mycn+/− n = 13, Mycn+/−/Mycn+/−;shMycn 2 n = 12; two-sided Mann–Whitney test). Data show mean ± s.e.m, scale bar, 50 μm. Source data
Extended Data Fig. 4 Elevated BCL-XL expression in early embryonic epidermis is sufficient to block Mycn-dependent apoptotic cell competition and has no long-term consequence for skin development.
a, Doxycycline-inducible activation of K14rtTA by lentivirally delivered TRE-driven transgenes at E12.5 (embryo transduced with TRE-GFP is shown, representative of results obtained in two independent litters). b, Experimental strategy for elevating BCL-XL expression in the early embryonic epidermis. TRE-Bcl2l1 was transduced at a relatively high titre, whereas CreRFP was transduced at a low titre (n = 4 embryos). c, Whole-mount images of E12.5 transduced embryos show that >50% of basal epidermal cells express BCL-XL and surround a small population of cells that express both BCL-XL (green) and CreRFP (red). TUNEL (white, signal inverted in right-most image) was quantified (Fig. 2a) relative to the position of CreRFP-expressing Mycn+/– loser cells that also express BCL-XL. Note that appreciable TUNEL labelling was not observed. Images are representative of data obtained from two litters. d, Proliferation in WT embryos at E12.5 (n = 4 per genotype) was lower in BCL-XL+ cells than in controls. At E17.5 there was no difference in proliferation between controls (n = 11) and BCL-XL+ cells (n = 14; n denotes number of images analysed from embryos from two litters for each genotype; two-sided Mann–Whitney test). e, Induction of BCL-XL expression from E9.5 to E15.5 had no appreciable consequence for epidermal differentiation or thickness at E17.5 (control n = 54; BCL-XL+ n = 68; n denotes number of thickness measurements taken from images of back skin cryosections from two mice per genotype). Representative images of back skin sections are shown immunolabelled for K14 (green) to mark the basal epidermis, and K10, Loricrin, or Filaggrin (red) to mark the differentiating spinous and granular layers. Dashed lines denote epidermal–dermal border; solid line demarcates the skin surface. Scale bars, 50 μm. Data are mean ± s.e.m. Source data
Extended Data Fig. 5 Epidermal cells, and not phagocytic immune cells, mediate clearance of corpses at E12.5.
a, b, CD45+ immune cells do not infiltrate the epidermis at E12.5. Wild-type and Mycnfl/+ embryos were infected with LV-Cre at E9.5 and analysed at E12.5. Whole-mount immunofluorescence and confocal microscopy showed that CD45+ immune cells (mostly macrophages at this time) were confined primarily to the deeper dermis. Immune cells were quantified over image tiles that encompassed a region of 425 μm × 425 μm (n denotes number of images analysed from two embryos per genotype, wild-type n = 5, Mycnfl/+ n = 6; one-way ANOVA with Tukey’s multiple comparisons test). Representative images from quantifications in a are shown in b at subsequent z positions within the tissue as indicated. c, Few uncleared cell corpses in Gulp1 shRNA-treated epidermis are RFP+, suggesting that corpses do not accumulate simply because of the consequences of Gulp1 knockdown on cell viability. Left, quantification of RFP status of TUNEL+ corpses in Gulp1 knockdown epidermis (n = 9 images from 4 embryos, two-tailed students t-test). Scale bar, 50 μm in b; 10 μm in c. Data are mean ± s.e.m. Source data
Extended Data Fig. 6 Generation of a cell competition signature and validation of method to classify wild-type epidermal cells via scRNA-seq.
a, Experimental strategy to identify endogenous winners and losers in the epidermis. b, c, FACS-based sorting strategies to isolate winner and loser populations from the epidermis. d, Differential expression comparing transcriptomes of 39 winner (annexinV–) cells with those of 104 loser (66 Mycn+/– and 38 annexinV+) cells uncovers genes that define the cell competition signature (P values generated from Wilcoxan rank sum test). Top, loser genes; bottom, winner genes. e, Putative winners (annexinV–), putative losers (annexinV+), and Mycn+/– losers fall where expected relative to the wild-type distribution based on expression of cell competition signature genes.
a, Sorting strategy to isolate wild-type epidermal cells. b–f, Principal component analysis (b, via the Jackstraw Method, see Methods) and t-SNE clustering (c). Analysis was performed to confirm that neither cell cycle differences (d) nor batch effects (e) account for the clustering. f, Further subclustering of E17.5 cells reveals distinct lineage populations: interfollicular epidermis (IFE), hair follicle (HF), and differentiating suprabasal cells (SB). Analysis was performed on 227 cells (111 E17.5 cells and 116 E12.5 cells). Heatmap shows the top 30 enriched genes that define each cluster.
Extended Data Fig. 8 Differentiation mediates the context-dependent exit of Mycn+/− loser cells from the basal layer during cell competition in late epidermal development.
a, Fewer Mycn+/– proliferating basal cells than proliferating WT neighbours were observed at E15.5 (n = 15 regions measured, two-sided Mann–Whitney test). b, c, Mycn+/– loser cells were found more often than wild-type cells in the FLG+ layer of the epidermis at E17.5 (n > 10 clones, two-sided Mann–Whitney test). d, Binned distributions extracted from data shown in Fig. 4c of GFP+K14+ clones (left) and GFP+K10+ clones (right) for each of three genotypes at E17.5: wild-type (top, grey), Mycn+/– (middle, pink), and Mycn+/–;shLgn (bottom, teal). e, Tissue-wide loss of one allele of Mycn yields no epidermal differentiation phenotypes or evidence of accelerated differentiation. The thickness of the region labelled with three markers of differentiation in the skin (keratin-10, involucrin, and filaggrin) was measured and no significant difference was found between control and Meox2Cre;Mycnfl/+ embryos (n denotes number of regions measured from cryosections generated from two different animals per genotype; unpaired two-tailed t-test). R26YFP was used as a marker of Cre expression and is shown in the inset. f, Spindle angle data from rose diagrams in Fig. 4d, plotted as a scatter plot for statistical comparison. Wild-type n = 32, Mycn+/– n = 25; two-sided Mann–Whitney test. Scale bars, 50 μm. Data are mean ± s.e.m. Source data
Extended Data Fig. 9 Barrier assays and R26Confetti labelling experiments uncover functional consequences of disrupting cell competition during epidermal development.
a, Strategy to block apoptotic cell competition and measure barrier function. b, Asymmetry score compares TEWL values from embryo’s left and right sides (n denotes number of embryos, F-test to compare variance). c, Experimental strategy for confetti lineage tracing. d, Representative whole-mount images of confetti-marked clones from each genotype at E17.5 (left, examples of clones from three litters of embryos analysed for each genotype). Examples of maximum projections and optical sections of individual epidermal layers are shown for each analysed fluorescent protein. SB, suprabasal. Right, quantification in wild-type embryos shows that approximately equal labelling efficiency is obtained at E17.5 with each of the three confetti fluorescent proteins (RFP n = 9 clones, YFP n = 11 clones, CFP n = 12 clones; Kruskal–Wallis test). e, f, Total clone size (e) and suprabasal clone size (f) dynamics for the genotypes indicated. All data are mean ± s.e.m. except for bar graphs, which show binned distributions. Source data
Less fit loser cells are initially cleared from the developing epidermis by apoptosis and subsequent engulfment by epidermal neighbours. As the tissue begins to stratify and differentiate, at E15.5, losers are instead removed from the basal layer of epidermal progenitors via asymmetric cell division and differentiation into the suprabasal layers of the developing skin. Although mechanistically distinct from one another, both phases of cell competition require cell–cell contact between winners and losers to trigger loser cell elimination.
Video 1 Examples of Mycn+/- cell in contact with WT neighbours undergoing cell death and clearance. Time-lapse imaging of the basal epidermis of an intact E12.5 embryo shows Mycn+/- cells (green, LV-Cre positive) are more likely to die dependent on contact with WT neighbours (red, LV-Cre-negative). Example is representative of death events quantified from 22 videos. Arrowhead in first frame indicates position of cell that will die. Scale bar = 10µm.
Video 2 Example of Mycn+/- cell in contact with WT neighbours undergoing cell death and clearance. Time-lapse imaging of the basal epidermis of an intact E12.5 embryo shows Mycn+/- cells (green, LV-Cre positive) are more likely to die dependent on contact with WT neighbours (red, LV-Cre-negative). Example is representative of death events quantified from 22 videos. Arrowhead in first frame indicates position of cell that will die. Scale bar = 10µm.
Video 3 Example of Mycn+/- cell in contact with WT neighbours undergoing cell death and clearance. Time-lapse imaging of the basal epidermis of an intact E12.5 embryo shows Mycn+/- cells (green, LV-Cre positive) are more likely to die dependent on contact with WT neighbours (red, LV-Cre-negative). Example is representative of death events quantified from 22 videos. Arrowhead in first frame indicates position of cell that will die. Scale bar = 10µm.
Video 4 Example of Mycn+/- cell in contact with WT neighbours undergoing cell death and clearance. Time-lapse imaging of the basal epidermis of an intact E12.5 embryo shows Mycn+/- cells (green, LV-Cre positive) are more likely to die dependent on contact with WT neighbours (red, LV-Cre-negative). Example is representative of death events quantified from 22 videos. Arrowhead in first frame indicates position of cell that will die. Scale bar = 10µm.