Oncogenic cells evade cell competition and evolve into tumors through clone size-dependent, progressive elevation of Yki activity

Oncogenic cells evade cell competition and evolve into tumors through clone size-dependent, progressive elevation of Yki activity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Oncogenic cells evade cell competition and evolve into tumors through clone size-dependent, progressive elevation of Yki activity Sai Katayama, Seiya Nishikawa, Shizue Ohsawa, Atsuko Takamatsu, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9135893/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Normal epithelial tissues often exert tumor-suppressive effects against newly emerged oncogenic cells. In Drosophila epithelia, clones of oncogenic polarity-deficient cells mutant for scribble ( scrib ) are eliminated by cell competition when surrounded by wild-type cells. Here, we show that the win-lose fate of scrib mutant cells depends on their initial population size. Small scrib -knockdown ( scrib KD ) clones are efficiently eliminated from wing imaginal epithelia, whereas larger clones initially behave as losers but later escape elimination and overgrow into tumors. This shift in competitive behavior in large clones is accompanied by a progressive activation of the Hippo pathway effector Yorkie (Yki). To dissect this intriguing growth dynamics, we extend our previously proposed deterministic mathematical model of cell competition to a stochastic framework incorporating experimentally observed sources of variability. The model reproduces the large variability observed in clone growth dynamics. Furthermore, addition of the effect of late-stage, progressive growth acceleration to the mathematical model recapitulated the experimentally observed size-dependent progressive tumorigenesis. Our findings establish the initial size of oncogenic cell population as the key determinant of the win-lose fate and provide a potential mechanism by which oncogenic clones evade cell competition through size-dependent, progressive elevation of Yki activity. Biological sciences/Cancer Biological sciences/Cell biology Biological sciences/Developmental biology cell competition tumorigenesis Yorkie Drosophila epithelium mathematical modeling Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Cells within animal tissues compete with each other for survival, a phenomenon called cell competition. Cell competition is a form of cell-cell interaction whereby cells with higher fitness (“winners”) eliminate neighboring cells with lower fitness (“losers”) by inducing cell death 1 – 3 . Therefore, it is thought that cell competition plays important roles in optimizing tissue and organismal fitness, selecting fitter cells within stem cell population, and eliminating damaged or harmful cells from tissues 4 . Indeed, accumulating evidence in Drosophila and mammalian systems has shown that potentially tumorigenic mutant cells are actively eliminated from epithelial tissues by cell competition when confronted with normal cells 5 . In Drosophila , developing imaginal epithelial tissues mutant for apicobasal polarity genes such as scribble ( scrib ) or discs large ( dlg ) undergo overgrowth and develop into tumors 6 , 7 . Intriguingly, when these oncogenic polarity-deficient cells are induced as mosaics, which are surrounded by wild-type cells, they are eliminated from the tissue 6 , 8 . This phenomenon is considered as tumor-suppressive cell competition, in which oncogenic polarity-deficient cells are eliminated as the “losers”, and surrounding wild-type “winner” cells proliferate and occupy the space vacated by the losers’ death. Thus, small clones of polarity-deficient cells are progressively eliminated from the tissue during development, resulting in the formation of entirely normal adult tissue. Interestingly, it has been reported in Drosophila imaginal epithelia that, while small clones of oncogenic cells mutant for the endocytic protein Rab5 are eliminated from the imaginal epithelium by cell competition 9 , large populations of Rab5-knockdown cells induced by the Gal4/UAS system overproliferate and develop into tumors 10 . This suggests that the initial size of an oncogenic cell population may determine their win-lose fate. However, whether this principle generally applies to tumor-suppressive cell competition and how the initial size of an oncogenic cell population affects their win-lose fate remain unknown. Here, using Drosophila genetics combined with mathematical modeling, we show that the initial size of an oncogenic polarity-deficient cell population in otherwise normal epithelial tissue determines whether these cells are eliminated by cell competition or instead overgrow and develop into tumors. Furthermore, simulation analyses reveal that tumorous overgrowth at later developmental stages occurs only when the initial oncogenic population is sufficiently large, consistent with a late-stage increase in the effective carrying capacity of mutant cells. Genetic analyses further indicate that this late-stage growth advantage is associated with a size-dependent, progressive activation of the Hippo pathway effector Yorkie (Yki) observed at the later developmental stages. Our findings suggest a possible mechanism by which oncogenic mutant cells evade cell competition and evolve into cancer cells through size-dependent, progressive elevation of Yki activity over time. Results The win-lose fate of polarity-deficient clones is determined by their initial size To investigate whether the initial size of oncogenic cell population affects their win-lose fate during cell competition, we induced various sizes of GFP-labeled wild-type or scrib- knockdown ( scrib KD ) clones in Drosophila wing imaginal epithelium using the heat-shock-mediated flip-out system. This experimental system allowed us to control the initial size of GFP-expressing cell clones by modulating the duration of the heat-shock treatment. To monitor the process of cell competition over an extended period, we cultured larvae on food supplemented with erg-2Δ mutant yeast ( erg-2Δ ) 11,12 , which depletes ecdysone from the food and thereby extends the larval period, allowing us to analyze wing discs at 3 days after clone induction (3d ACI) and later time points. When wild-type clones were induced in wing discs by different durations of heat-shock treatment (10 min, 20 min, 30 min, or 60 min), the total size of GFP-expressing cell population (% of the tissue area) increased with the length of the heat-shock period (Fig. 1 ). Importantly, in each heat-shock condition, the ratio of GFP-expressing cell population (% of the tissue area) did not significantly change over time during larval development (Fig. 1 A, E, I, and M, quantified in C, G, K, and O, respectively). Thus, under this experimental condition, the initial size of GFP-expressing cell population can be controled by adjusting the duration of heat-shock treatment. We then analyzed the fate of scrib KD clones induced in wing discs with different initial sizes. When small clones of scrib KD cells were induced by a 10 min heat-shock (estimated to produce an initial clone size of ~ 25% based on the wild-type control), the size of scrib KD clones drastically decreased by 3 days after clone induction (3d ACI) and remained largely unchanged thereafter (Fig. 1 B, quantified in Fig. 1 D). This suggests that scrib KD clones were eliminated as losers of cell competition. When clones were induced by a 20 min heat-shock (estimated initial clone size of ~ 60%), the size of scrib KD cell clones gradually decreased over time until around 6d ACI (Fig. 1 F, quantified in Fig. 1 H), suggesting that these cells also behaved as losers. Intriguingly, when clones were induced by a 30 min heat-shock (estimated initial clone size of ~ 70%), the size of scrib KD clones did not significantly change from 2d ACI to 6d ACI (Fig. 1 J, quantified in Fig. 1 L). In this condition, their size stabilized at ~ 50% throughout this period, which is smaller than the estimated initial size (70%). This suggests that scrib KD clones are initially eliminated as losers but somehow evade loser status over time. Remarkably, when scrib KD clones were induced by a 60 min heat-shock (estimated initial clone size of ~ 85%), the size of scrib KD clones gradually decreased until around 4d ACI and then began to increase, leading to tumorous overgrowth by 6d ACI (Fig. 1 N, quantified in Fig. 1 P). Thus, when induced as large clones, scrib KD clones behave as losers initially but eventually convert to winners. Together, these data indicate that the win-lose fate of scrib KD clones is determined by their initial size. Mathematical modeling recapitulates the initial-size-dependency of the win-lose fate To understand the mechanism underlying the initial size-dependent fate of scrib KD clones, we constructed a mathematical model of cell competition based on our previously published model 13 . The previous model describes deterministic population dynamics of wild-type and mutant cells forming a monolayer tissue (See Supplemental Materials S1). However, the experimental data exhibited substantially larger variability than predicted by the original deterministic model. For example, the relative clone-size measurements showed a spread of approximately 40–50% of the mean value (Fig. 1 ). Therefore, we extended the framework to a stochastic model that incorporates several sources of variability, including stochastic cell division and cell death, variation in heat-shock efficiency, and individual differences in final tissue size (see Supplemental Materials S2.1). Figure 2 shows the results of computer simulations of the stochastic model (lower panels), compared with the corresponding experimental results (upper panels). A total of 100 trials were simulated for each condition. The simulations began at the time of egg laying (time = -2day ACI), when the tissue consisted of 30 wild-type cells. The number of wild-type cells increased to approximately 13×10 2 after 2 days (time = 0), at which point a fraction of wild-type cells converted to mutant cells. The initial clone-size ratio was set to 0.25, 0.60, 0.70, or 0.85, corresponding to the experimentally measured clone-size ratios obtained under the 10-, 20-, 30-, and 60-min heat-shock treatments, respectively (see Supplemental Materials S2.2.1 for details). By incorporating stochastic fluctuations and experimental variability, the simulated data (shown as points) became scattered, closely resembling the experimental results, in contrast to those produced by the deterministic model (shown with thick curves). Under the 10 min heat-shock condition, mutant clones (green triangles) became losers in the deterministic model (thick green and pink lines in Fig. 2 C and 2 D). In contrast, a subset of small clones survived in the stochastic model, reflecting variability introduced by stochastic processes and differences in experimental conditions. Under the 20 min and 30 min heat-shock conditions, the deterministic model predicted an initial decrease in relative clone size until 2 d ACI, followed by an increase thereafter (20 min: Fig. 2 G, 2 H; 30 min: Fig. 2 K, 2 L). In the stochastic model, however, the simulated data were highly scattered: some mutant clones were eliminated, whereas others persisted (Fig. 2 H and 2 L). This variation was most pronounced under the 30-min heat-shock condition (Fig. 2 L). Even under the 60 min heat-shock condition, the relative clone size first decreased until 2 d ACI and then increased thereafter in the model simulations (Fig. 2 O and 2 P). As a consequence of this late-stage increase, most mutant clones ultimately survived. Mathematical simulation predicts a late-stage, progressive increase in growth capacity in mutant cells Intriguingly, although most of the computer simulation results reproduced the corresponding experimental data, we identified several notable discrepancies. In the simulations for the 30 min and 60 min heat-shock conditions, tissue sizes reached saturation by 5–6 d ACI (Fig. 2 K and 2 O), whereas in the experiments some tissues exhibited tumorous overgrowth (Fig. 2 I and 2 M), with sizes deviating up to 250–300×10 3 µm 2 from the majority level of around 150×10 3 µm 2 . This discrepancy was particularly pronounced at 6 d ACI under the 60-min heat-shock condition. These mathematical and experimental observations suggest the existence of an additional mechanism that promotes mutant cell proliferation when a particular condition is fulfilled. A plausible candidate for this condition is a time-dependent process, whereby mutant cells undergo a delayed transition to an overgrowth phenotype. Indeed, incorporating a time-dependent increase in growth capacity at later developmental stages (implemented as an increased carrying capacity, K y ) produced tissue size deviations similar to those observed experimentally (Fig. 2 Q and 2 R). Large scrib KD clones evade cell competition and overgrow via elevated Yki activity We next experimentally investigated the molecular mechanism by which large scrib KD clones acquire tumorigenic potency at later developmental stages. Using the mitotic marker phospho-Histone H3 (PH3), we confirmed that wing disc cells normally ceased proliferation by 5d ACI (Fig. 3 A), as previously reported 14 . In contrast to this developmental arrest, large scrib KD clones exhibited enhanced proliferation at 5d ACI (Fig. 3 B, orange arrow). Small scrib KD clones in the same tissue had already ceased proliferating (Fig. 3 B, orange arrowhead), consistent with our simulation-based prediction that tumorigenic potency depends on clone size. Importantly, at an earlier stage (3d ACI), large scrib KD clones did not exhibit increased proliferation compared with neighboring wild-type tissue; rather, their proliferative activity was comparable to or even lower than that of wild-type cells (Fig. S8A, quantified in S8B). These data suggest that enhanced proliferation is acquired by large scrib KD clones at later developmental stages. We then examined the signaling pathways that might underlie this late-stage proliferative activity. Interestingly, we found that expressions of diap1 and microRNA bantam , which are canonical targets of the Hippo pathway effector Yki, were significantly elevated in scrib KD cells within large clones at 5d ACI (Fig. 3 D, F), while small scrib KD clones, wild-type clones, and large scrib KD clones at earlier developmental stages did not show such inductions (Fig. 3 C, E, Fig. S8C and S8D). To assess the functional contribution of Yki, we knocked down yki simultaneously with scrib knockdown at the time of clone induction and quantified clone size at 3d ACI. Although Yki target gene upregulation was not yet detectable at this stage, yki knockdown significantly reduced the size of scrib KD clones (Fig. 3 H and 3 I, quantified in Fig. 3 K). Importantly, yki knockdown alone also significantly reduced clone size compared with wild-type controls at 3d ACI (Fig. 3 G and 3 J, quantified in Fig. 3 K), confirming its contribution to basal epithelial cell proliferation. Therefore, the reduced growth of scrib KD clones upon yki knockdown likely reflects both general proliferative function and the requirement of Yki in the later acquisition of tumorigenic growth capacity. Together, these findings support a model in which large scrib KD clones progressively engage Yki-dependent growth programs over time, enabling them to escape developmental growth arrest and undergo tumorigenesis. Discussion Our data show that the win-lose fate of scrib KD clones in the growing imaginal disc is determined by their initial clone size. As summarized in Fig. 4 , the outcome of scrib KD clones can be represented as a size-dependent phase transition in clone dynamics. Simulation analysis reveals a size threshold: when the initial mutant clone size remains below approximately 17% of the tissue, scrib KD clones are completely eliminated (Fig. 4 A; the threshold ratio, indicated by the navy dashed line, was estimated using Eq. (S3) in Supplemental Materials S1). In contrast, when the initial clone size exceeds this threshold, a substantial fraction of mutant cells is eliminated during early developmental stages (Fig. 4 B), yet the remaining population subsequently expands and surpasses the normal tissue carrying capacity at later stages (Fig. 4 C). Ultimately, large scrib KD clones persist and progress toward tumorigenic overgrowth (Fig. 4 D). These data are consistent with previous reports showing that small clones of tumorigenic Ras V12 / lgl −/− or Rab5 mutant cells are eliminated by cell competition, whereas larger clones overgrow within the tissue 10 , 15 . In the present study, we quantitatively analyze the win-lose fate of scrib KD clones of different initial sizes and, by integrating experimental data with mathematical modeling, provide a framework explaining how initial clone size determines long-term win-lose fate (Fig. 4 ). Importantly, our experimental data indicate that this late-stage clonal expansion is accompanied by a progressive increase in Yki activity. Elevated expression of Yki target genes is observed specifically in large clones at later developmental stages, suggesting that detectable Yki target induction is not evident at clone initiation but becomes progressively enhanced over time in sufficiently large clone domains. Thus, clone size does not merely influence growth rate but defines whether mutant cells remain susceptible to elimination or transition into a Yki-dependent tumorigenic state. Nevertheless, several limitations should be considered. Because temporally restricted knockdown of yki specifically at later developmental stages was not technically feasible in our system, we cannot formally distinguish between a requirement of basal Yki activity and a stage-specific activation during tumorigenic expansion. Nonetheless, the lack of detectable Yki target induction at early stages argues against strong pathway activation at clone initiation and instead supports a progressive increase in Yki signaling as clone domains expand over time. In this context, the upstream mechanisms that drive the progressive increase in Yki activity within large scrib KD clones remain to be elucidated. One possibility is that growth factors or other signaling molecules produced by scrib KD cells accumulate within sufficiently large clone territories, thereby modifying the local microenvironment and reducing their susceptibility to elimination. Such size-dependent accumulation of signaling inputs could provide a mechanistic link between clone expansion and Yki activation. Notably, scrib KD cells are eliminated by cell competition in mammalian epithelial cell cultures 17 , and the Yki homolog YAP can convert mammalian cells into winners of cell competition 18 . These observations raise the possibility that similar size-dependent win-lose transitions may operate in mammalian tissues and contribute to early tumorigenic progression. Together, our findings suggest that tumorigenic potential may emerge not solely from oncogenic mutation, but from the integration of clone size, competitive context, and progressive signaling activation over time. Methods Fly culture and generation of clones Flies were cultured at 25°C on standard fly food unless otherwise noted. Fly food containing erg2Δ yeast was prepared as discribed previously 14 . Adult flies were allowed to mate for 6 h, and the eggs laid during this period were collected for rearing. The time point 3 h after the onset of mating was defined as 0 h after egg laying (AEL0). Vials containing developing larvae were subjected to a 37°C heat shock at 48h after AEL0. Larvae were dissected at designated time points after the initiation of the heat shock, and wing discs were collected. Fluorescently labeled mitotic clones were produced in larval imaginal discs using the following strains: hsFLP 122 ; Act > y + > Gal4, UAS-GFP (tester), UAS-Dcr2 (BDSC #24646), UAS-scrib-RNAi (HMS01490, BDRC#35748), bantam-LacZ (BDSC #10154), diap1-lacZ/Diap1 j5C8 (BDSC #12093) and UAS-Yki-RNAi (NIG #4005R-2). Histology larvae were dissected in phosphate buffered saline (PBS) and fixed with 4%-Paraformaldehyde (PFA) for 5 min on ice, followed by 20 min at room temperature. The fixed larvae were washed 3 times with PBT (PBS + 0.1% TritonX100) for 20 min each, and were then blocked with PBTn (5% donkey serum + PBT) solution for 30 min. For immunostaining, they were incubated overnight at 4°C with primary antibodies in PBTn. After four 30-min washes with PBT, the samples were re-blocked in PBTn for 30 minutes, then incubated with secondary antibodies for 2 hours at room temperature. Following another set of four 30-minute washes with PBT, the samples were mounted using DAPI-containing SlowFade Gold Antifade Reagent (Invitrogen, #S36939). Images were taken with Leica SP5 confocal microscopy. Primary antibodies used are as follows; Phospho-Histone H3 (Ser10) (1:100; Cell Signaling #9706) and chicken anti-β-gal (1:10; abcam #ab9361). Secondary antibodies used are as follows; Goat anti-rabbit Alexa 546 (1:250; Invitrogen #A11035) and Goat anti-chicken Alexa 647 (1:250; Invitrogen #A21449). Disc size and GFP-labeled clone size were measured using OpenCV (cv2), a Python library for image processing. Statistical analyses were performed using RStudio. Declarations Funding This work was supported in part by the MEXT/JSPS KAKENHI (22H05619 and 24H01394 to S.N., 20H05945, 22H02616 and 21K19257 to S.O., 26114002 and 25127717 to A.T., and 21H05284 and 21H05039 to T.I.), Japan Science and Technology Agency (Moonshot Research & Development: Grant Number JPMJPS2022) to S.O., AMED-CREST, Japan Agency for Medical Research and Development (23gm1710002h0002) to T.I., the Princess Takamatsu Cancer Research Fund to S.O. and the Takeda Science Foundation to S.O. and T.I. Author Contribution S.K., S.N., S.O., A.T. and T.I. designed the study; S.K. and S.O. performed experiments; S.N. and A.T. developed the mathematical model; S.K., S.N., S.O., A.T. and T.I. analyzed the data; S.K., S.N., S.O., A.T. and T.I. wrote the manuscript.All authors reviewed the manuscript. Acknowledgement We thank Igaki and Takamatsu lab members for discussions; K. Baba, M. Tanaka, and M. Koijima for technical support. We also thank the Bloomington Drosophila Stock Center (Indiana), the Vienna Drosophila Resource Center (Vienna), the National Institute of Genetics Stock Center (Mishima), and the Drosophila Genomics and Genetic Resources (Kyoto) for fly stocks. Data Availability The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. References Levayer, R. & Moreno, E. Mechanisms of cell competition: themes and variations. J. Cell. Biol. 200 , 689–698. 10.1083/jcb.201301051 (2013). Vincent, J. P., Fletcher, A. G. & Baena-Lopez, L. A. Mechanisms and mechanics of cell competition in epithelia. Nat. Rev. Mol. Cell. Biol. 14 , 581–591. 10.1038/nrm3639 (2013). Amoyel, M. & Bach, E. A. Cell competition: how to eliminate your neighbours. Development 141 , 988–1000. 10.1242/dev.079129 (2014). 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. 10.1016/j.devcel.2016.08.012 (2016). Baker, N. E. Cell competition. Curr. Biol. 21 , R11–15. 10.1016/j.cub.2010.11.030 (2011). Woods, D. F. & Bryant, P. J. The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions. Cell 66 , 451–464. 10.1016/0092-8674(81)90009-x (1991). Bilder, D. & Perrimon, N. Localization of apical epithelial determinants by the basolateral PDZ protein Scribble. Nature 403 , 676–680. 10.1038/35001108 (2000). Humbert, P., Russell, S., Richardson, H. & Dlg Scribble and Lgl in cell polarity, cell proliferation and cancer. Bioessays 25 , 542–553. 10.1002/bies.10286 (2003). Takino, K., Ohsawa, S. & Igaki, T. Loss of Rab5 drives non-autonomous cell proliferation through TNF and Ras signaling in Drosophila. Dev. Biol. 395 , 19–28. 10.1016/j.ydbio.2014.09.003 (2014). Ballesteros-Arias, L., Saavedra, V. & Morata, G. Cell competition may function either as tumour-suppressing or as tumour-stimulating factor in Drosophila. Oncogene 33 , 4377–4384. 10.1038/onc.2013.407 (2014). Bos, M., Burnet, B., Farrow, R. & Woods, R. A. Development of Drosophila on sterol mutants of the yeast Saccharomyces cerevisiae. Genet. Res. 28 , 163–176. 10.1017/S0016672300016840 (1976). Parkin, C. A. & Burnet, B. Growth arrest of Drosophila melanogaster on erg-2 and erg-6 sterol mutant strains of Saccharomyces cerevisiae. J. Insect. Physiol. 32 , 463–471. 10.1016/0022-1910(86)90007-7 (1986). Nishikawa, S., Takamatsu, A., Ohsawa, S. & Igaki, T. Mathematical model for cell competition: Predator-prey interactions at the interface between two groups of cells in monolayer tissue. J. Theor. Biol. 404 , 40–50. 10.1016/j.jtbi.2016.05.031 (2016). Katsuyama, T. & Paro, R. Innate immune cells are dispensable for regenerative growth of imaginal discs. Mech. Dev. 130 , 112–121. 10.1016/j.mod.2012.11.005 (2013). Menendez, J., Perez-Garijo, A., Calleja, M. & Morata, G. A tumor-suppressing mechanism in Drosophila involving cell competition and the Hippo pathway. Proc. Natl. Acad. Sci. U S A . 107 , 14651–14656. 10.1073/pnas.1009376107 (2010). Chen, C. L., Schroeder, M. C., Kango-Singh, M., Tao, C. & Halder, G. Tumor suppression by cell competition through regulation of the Hippo pathway. Proc. Natl. Acad. Sci. U S A . 109 , 484–489. 10.1073/pnas.1113882109 (2012). Norman, M. et al. Loss of Scribble causes cell competition in mammalian cells. J. Cell Sci. 125 , 59. 10.1242/jcs.085803 (2012). Mamada, H., Sato, T., Ota, M. & Sasaki, H. Cell competition in mouse NIH3T3 embryonic fibroblasts is controlled by the activity of Tead family proteins and Myc. J. Cell. Sci. 128 , 790–803. 10.1242/jcs.163675 (2015). Sidiropoulos, N. et al. SinaPlot: An Enhanced Chart for Simple and Truthful Representation of Single Observations Over Multiple Classes. J. Comput. Graphical Stat. 27 , 673–676. 10.1080/10618600.2017.1366914 (2018). Additional Declarations No competing interests reported. Supplementary Files KatayamaetalSupplementalMaterial.pdf Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 27 Apr, 2026 Reviewers agreed at journal 13 Apr, 2026 Reviewers invited by journal 07 Apr, 2026 Editor assigned by journal 07 Apr, 2026 Editor invited by journal 07 Apr, 2026 Submission checks completed at journal 31 Mar, 2026 First submitted to journal 30 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9135893","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":621937132,"identity":"dbc7c6f5-3b36-4286-99b2-3d67f837114a","order_by":0,"name":"Sai Katayama","email":"","orcid":"","institution":"Kyoto University","correspondingAuthor":false,"prefix":"","firstName":"Sai","middleName":"","lastName":"Katayama","suffix":""},{"id":621937133,"identity":"cd7c9f23-d9db-4bb6-a296-8fb29c46830b","order_by":1,"name":"Seiya Nishikawa","email":"","orcid":"","institution":"Kyoto University","correspondingAuthor":false,"prefix":"","firstName":"Seiya","middleName":"","lastName":"Nishikawa","suffix":""},{"id":621937137,"identity":"5778cb18-e204-4aa2-844b-693ea38b2ffb","order_by":2,"name":"Shizue Ohsawa","email":"","orcid":"","institution":"Nagoya University","correspondingAuthor":false,"prefix":"","firstName":"Shizue","middleName":"","lastName":"Ohsawa","suffix":""},{"id":621937138,"identity":"b10eda71-73d7-40d1-ac55-ec65445fbc24","order_by":3,"name":"Atsuko Takamatsu","email":"","orcid":"","institution":"Waseda University","correspondingAuthor":false,"prefix":"","firstName":"Atsuko","middleName":"","lastName":"Takamatsu","suffix":""},{"id":621937143,"identity":"2fe13e6c-7f1a-4821-b90b-f3e6752102f6","order_by":4,"name":"Tatsushi Igaki","email":"data:image/png;base64,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","orcid":"","institution":"Kyoto University","correspondingAuthor":true,"prefix":"","firstName":"Tatsushi","middleName":"","lastName":"Igaki","suffix":""}],"badges":[],"createdAt":"2026-03-16 09:38:59","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9135893/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9135893/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106934352,"identity":"cb7363bc-ce9c-4ace-87bc-3e0670a977ed","added_by":"auto","created_at":"2026-04-15 02:58:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":316427,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe win-lose fate of polarity-deficient clones is determined by their initial size\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A, B, E, F, I, J, M, N) GFP-labeled wild-type (WT; A, E, I, M) or \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e (B, F, J, N) clones were induced in wing discs by heat-shock for 10 min (A-D), 20 min (E-H), 30 min (I-L), or 60 min(M-P). Wing discs at 2, 3, 4, 5, and 6 days after clone induction (ACI) are shown.\u003c/p\u003e\n\u003cp\u003e(C, D, G, H, K, L, O, P) Quantification of relative clone size for the genotypes shown in (A) (n:day2= 16, day3= 14, day4= 12, day5= 64, day6= 14), (B) (n:day2= 50, day3= 10, day4= 14, day5= 24, day6= 27), (E) (n:day2= 46, day3= 26, day4= 40, day5= 40, day6= 38), (F) (n:day2= 35, day3= 43, day4= 51, day5= 47, day6= 34), (I) (n:day2= 19, day3= 14, day4= 5, day5= 15, day6= 24), (J) (n:day2= 60, day3= 35, day4= 37, day5= 28, day6= 25), (M) (n:day2= 19, day3= 13, day4= 9, day5= 14, day6= 15), and (N) (n:day2= 13, day3= 32, day4= 34, day5= 44, day6= 16). Steel-Dwass test; * p \u0026lt; 0.05, ** p \u0026lt; 0.001. Values are means; ns, not significant. Scale bar, 100µm (A).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/c5230d72fd2d9c7e36d86053.png"},{"id":106961462,"identity":"dde3a425-bb96-4826-ad9a-785195f20334","added_by":"auto","created_at":"2026-04-15 09:25:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":203318,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of computer simulation with experimental results.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePeriods of heat shocks in (A-D), (E-H), (I-L), and (M-R) are 10, 20, 30, and 60 min, respectively. (A, E, I, and M) Time evolution of tissue size and \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clone size, estimated from their area measurements and plotted as SinaPlots\u003csup\u003e19\u003c/sup\u003e with black open circles representing tissue size and green triangles representing clone size. The plot style of SinaPlots is a hybrid of violin plots and jitter plots, in which the jitter points are restricted within the density distribution. Fly eggs were laid at time -2 days, and heat shocks were applied at time 0 days.\u003c/p\u003e\n\u003cp\u003e(B, F, J, and N) Time evolution of relative size of mutant clones, calculated from the data in (A, E, I, and M). These datasets correspond to (D, H, L, P) in Fig. 1.\u003c/p\u003e\n\u003cp\u003e(C, G, K, O, and Q) Time evolution of tissue and mutant clone sizes in computer simulations. Thin and thick curves represent the results of the stochastic Eq. (S2) and the deterministic models Eq. (S1), respectively. Gray, blue, and green curves represent the populations of the whole tissue, wild-type (WT) cell domains, and mutant clone domains, respectively. SinaPlots of tissue size and clone size at each experimental observation time point are superimposed on the curves.\u003c/p\u003e\n\u003cp\u003e(D, H, L, P, and R) Time evolution of the relative mutant clone size in computer simulations. Pink curves are the results of the deterministic model. In Q and R, overgrowth is implemented as a time-dependent increase in the mutant carrying capacity (see Supplemental Materials §S4).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/86ee9a7ecbc24394a1f2e3bb.png"},{"id":106934350,"identity":"1c3aaef8-9618-4329-895e-bc37751d0f37","added_by":"auto","created_at":"2026-04-15 02:58:03","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":221601,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLarge \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003escrib\u003c/strong\u003e\u003c/em\u003e\u003csup\u003e\u003cem\u003e\u003cstrong\u003eKD\u003c/strong\u003e\u003c/em\u003e\u003c/sup\u003e\u003cstrong\u003e clones evade cell competition and overgrow via elevated Yki activity.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A and B) Wing discs bearing GFP-labeled wild-type (A) or \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD \u003c/em\u003e\u003c/sup\u003e(B) clones induced by a 15 min heat-shock were stained with anti-P-histone H3 (PH3) antibody. An orange arrow indicates a tumorigenic \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clone, and an orange arrowhead indicate an eliminating \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones.\u003c/p\u003e\n\u003cp\u003e(C-F) 5d ACI \u003cem\u003eDIAP1-LacZ\u003c/em\u003e/+ (C and D) or \u003cem\u003ebantam-LacZ\u003c/em\u003e/+\u003cem\u003e \u003c/em\u003e(E and F) wing discs bearing GFP-labelled \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD \u003c/em\u003e\u003c/sup\u003e(D and F) or WT (C and E) clones induced by a 30 min (C and D) or 60 min (E and F) heat shock were stained with an anti-b-galactosidase antibody. Orange arrows indicate tumorigenic \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones showing strong expression of \u003cem\u003ediap1\u003c/em\u003e or \u003cem\u003ebantam\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003e(G-K) 3d ACI wing discs bearing GFP-labelled WT (G) (n = 14), \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e (H) (n = 16), \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e + \u003cem\u003eYki\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e (I) (n = 13), or \u003cem\u003eYki\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD \u003c/em\u003e\u003c/sup\u003e(J) (n = 16) clones induced by a 30 min heat shock are shown. The average relative clones size were quantified using custom Python scripts (K). Steel-Dwass test; * p \u0026lt; 0.05, ** p \u0026lt; 0.001. Values represent means. All images are taken at the same magnification. Scale bar, 100µm (A)\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/43eb7458e92e176dd8224452.png"},{"id":106961399,"identity":"88a66e09-5dc7-42dd-b201-9c2370dddc9f","added_by":"auto","created_at":"2026-04-15 09:25:25","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":106086,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic diagrams of initial-state-dependent fate after cell competition\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eEach colored area classifies interaction matter between WT and mutant cells, depending on combinations of WT- and mutant-cell domain sizes (represented as \u003cem\u003ex\u003c/em\u003e and \u003cem\u003ey\u003c/em\u003e, respectively). (A) WT to be a winner after cell competition. The tissue consists of WT cells exclusively, whose size becomes normal. (B) Cell competition effectively works before the tissue size exceeding the normal one. (C) Cell competition still works but proliferation of WT cells stops. (D) Overgrowth of mutant clones. A switch turns on over time and increases the mutant carrying capacity to K\u003csub\u003eover\u003c/sub\u003e . The background graph represents a phase portrait in the x-\u003cem\u003ey\u003c/em\u003e plane. Blue and red lines/curves are nullclines of \u003cem\u003ef(x,y)=0\u003c/em\u003e and \u003cem\u003eg(x,y)=0\u003c/em\u003e, respectively (See Supplemental Materials §S1 for details). The thick black solid curves are solutions of the deterministic equations of Eq. (S1) after heat shocks (clone size ratio controlled by heat shock = 0.25, 0.6, 0.7, and 0.8). The solutions start from the time when heat shocks are applied at day 0. Gray, blue, green, magenta, and red dots on the solution curves denote points of observation time at 2, 3, 4, 5, and 6 days ACI, respectively.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/8eb8777321850e68c37ef7dc.png"},{"id":106963350,"identity":"a36dbc90-03b8-43ea-b6ea-78fb09952321","added_by":"auto","created_at":"2026-04-15 09:43:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1589438,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/eb5067d0-566f-4c7c-9124-b979a95be86d.pdf"},{"id":106934349,"identity":"622b33c3-a901-470e-be2e-cdf20a58cddf","added_by":"auto","created_at":"2026-04-15 02:58:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2697437,"visible":true,"origin":"","legend":"","description":"","filename":"KatayamaetalSupplementalMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9135893/v1/26f03bb25e71aef28da3f24b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Oncogenic cells evade cell competition and evolve into tumors through clone size-dependent, progressive elevation of Yki activity","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCells within animal tissues compete with each other for survival, a phenomenon called cell competition. Cell competition is a form of cell-cell interaction whereby cells with higher fitness (\u0026ldquo;winners\u0026rdquo;) eliminate neighboring cells with lower fitness (\u0026ldquo;losers\u0026rdquo;) by inducing cell death\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Therefore, it is thought that cell competition plays important roles in optimizing tissue and organismal fitness, selecting fitter cells within stem cell population, and eliminating damaged or harmful cells from tissues\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Indeed, accumulating evidence in \u003cem\u003eDrosophila\u003c/em\u003e and mammalian systems has shown that potentially tumorigenic mutant cells are actively eliminated from epithelial tissues by cell competition when confronted with normal cells\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn \u003cem\u003eDrosophila\u003c/em\u003e, developing imaginal epithelial tissues mutant for apicobasal polarity genes such as \u003cem\u003escribble\u003c/em\u003e (\u003cem\u003escrib\u003c/em\u003e) or \u003cem\u003ediscs large\u003c/em\u003e (\u003cem\u003edlg\u003c/em\u003e) undergo overgrowth and develop into tumors\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Intriguingly, when these oncogenic polarity-deficient cells are induced as mosaics, which are surrounded by wild-type cells, they are eliminated from the tissue\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. This phenomenon is considered as tumor-suppressive cell competition, in which oncogenic polarity-deficient cells are eliminated as the \u0026ldquo;losers\u0026rdquo;, and surrounding wild-type \u0026ldquo;winner\u0026rdquo; cells proliferate and occupy the space vacated by the losers\u0026rsquo; death. Thus, small clones of polarity-deficient cells are progressively eliminated from the tissue during development, resulting in the formation of entirely normal adult tissue. Interestingly, it has been reported in \u003cem\u003eDrosophila\u003c/em\u003e imaginal epithelia that, while small clones of oncogenic cells mutant for the endocytic protein Rab5 are eliminated from the imaginal epithelium by cell competition\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, large populations of Rab5-knockdown cells induced by the \u003cem\u003eGal4/UAS\u003c/em\u003e system overproliferate and develop into tumors\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. This suggests that the initial size of an oncogenic cell population may determine their win-lose fate. However, whether this principle generally applies to tumor-suppressive cell competition and how the initial size of an oncogenic cell population affects their win-lose fate remain unknown.\u003c/p\u003e \u003cp\u003eHere, using \u003cem\u003eDrosophila\u003c/em\u003e genetics combined with mathematical modeling, we show that the initial size of an oncogenic polarity-deficient cell population in otherwise normal epithelial tissue determines whether these cells are eliminated by cell competition or instead overgrow and develop into tumors. Furthermore, simulation analyses reveal that tumorous overgrowth at later developmental stages occurs only when the initial oncogenic population is sufficiently large, consistent with a late-stage increase in the effective carrying capacity of mutant cells. Genetic analyses further indicate that this late-stage growth advantage is associated with a size-dependent, progressive activation of the Hippo pathway effector Yorkie (Yki) observed at the later developmental stages. Our findings suggest a possible mechanism by which oncogenic mutant cells evade cell competition and evolve into cancer cells through size-dependent, progressive elevation of Yki activity over time.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eThe win-lose fate of polarity-deficient clones is determined by their initial size\u003c/h2\u003e \u003cp\u003eTo investigate whether the initial size of oncogenic cell population affects their win-lose fate during cell competition, we induced various sizes of GFP-labeled wild-type or \u003cem\u003escrib-\u003c/em\u003eknockdown (\u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cspan type=\"ItalicSmallCaps\" class=\"ItalicSmallCaps\" name=\"Emphasis\"\u003eKD\u003c/span\u003e\u003c/sup\u003e\u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003e)\u003c/span\u003e clones in \u003cem\u003eDrosophila\u003c/em\u003e wing imaginal epithelium using the heat-shock-mediated flip-out system. This experimental system allowed us to control the initial size of GFP-expressing cell clones by modulating the duration of the heat-shock treatment. To monitor the process of cell competition over an extended period, we cultured larvae on food supplemented with \u003cem\u003eerg-2Δ\u003c/em\u003e mutant yeast (\u003cem\u003eerg-2Δ\u003c/em\u003e)\u003csup\u003e11,12\u003c/sup\u003e, which depletes ecdysone from the food and thereby extends the larval period, allowing us to analyze wing discs at 3 days after clone induction (3d ACI) and later time points. When wild-type clones were induced in wing discs by different durations of heat-shock treatment (10 min, 20 min, 30 min, or 60 min), the total size of GFP-expressing cell population (% of the tissue area) increased with the length of the heat-shock period (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Importantly, in each heat-shock condition, the ratio of GFP-expressing cell population (% of the tissue area) did not significantly change over time during larval development (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eA, E, I, and M, quantified in C, G, K, and O, respectively). Thus, under this experimental condition, the initial size of GFP-expressing cell population can be controled by adjusting the duration of heat-shock treatment.\u003c/p\u003e \u003cp\u003eWe then analyzed the fate of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones induced in wing discs with different initial sizes. When small clones of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e cells were induced by a 10 min heat-shock (estimated to produce an initial clone size of ~\u0026thinsp;25% based on the wild-type control), the size of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones drastically decreased by 3 days after clone induction (3d ACI) and remained largely unchanged thereafter (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eB, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). This suggests that \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones were eliminated as losers of cell competition. When clones were induced by a 20 min heat-shock (estimated initial clone size of ~\u0026thinsp;60%), the size of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e cell clones gradually decreased over time until around 6d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eH), suggesting that these cells also behaved as losers. Intriguingly, when clones were induced by a 30 min heat-shock (estimated initial clone size of ~\u0026thinsp;70%), the size of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones did not significantly change from 2d ACI to 6d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eJ, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eL). In this condition, their size stabilized at ~\u0026thinsp;50% throughout this period, which is smaller than the estimated initial size (70%). This suggests that \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones are initially eliminated as losers but somehow evade loser status over time. Remarkably, when \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones were induced by a 60 min heat-shock (estimated initial clone size of ~\u0026thinsp;85%), the size of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones gradually decreased until around 4d ACI and then began to increase, leading to tumorous overgrowth by 6d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eN, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003eP). Thus, when induced as large clones, \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones behave as losers initially but eventually convert to winners. Together, these data indicate that the win-lose fate of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones is determined by their initial size.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMathematical modeling recapitulates the initial-size-dependency of the win-lose fate\u003c/h3\u003e\n\u003cp\u003eTo understand the mechanism underlying the initial size-dependent fate of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones, we constructed a mathematical model of cell competition based on our previously published model\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. The previous model describes deterministic population dynamics of wild-type and mutant cells forming a monolayer tissue (See Supplemental Materials S1). However, the experimental data exhibited substantially larger variability than predicted by the original deterministic model. For example, the relative clone-size measurements showed a spread of approximately 40\u0026ndash;50% of the mean value (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Therefore, we extended the framework to a stochastic model that incorporates several sources of variability, including stochastic cell division and cell death, variation in heat-shock efficiency, and individual differences in final tissue size (see Supplemental Materials S2.1). Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the results of computer simulations of the stochastic model (lower panels), compared with the corresponding experimental results (upper panels). A total of 100 trials were simulated for each condition. The simulations began at the time of egg laying (time = -2day ACI), when the tissue consisted of 30 wild-type cells. The number of wild-type cells increased to approximately 13\u0026times;10\u003csup\u003e2\u003c/sup\u003e after 2 days (time\u0026thinsp;=\u0026thinsp;0), at which point a fraction of wild-type cells converted to mutant cells. The initial clone-size ratio was set to 0.25, 0.60, 0.70, or 0.85, corresponding to the experimentally measured clone-size ratios obtained under the 10-, 20-, 30-, and 60-min heat-shock treatments, respectively (see Supplemental Materials S2.2.1 for details).\u003c/p\u003e \u003cp\u003eBy incorporating stochastic fluctuations and experimental variability, the simulated data (shown as points) became scattered, closely resembling the experimental results, in contrast to those produced by the deterministic model (shown with thick curves). Under the 10 min heat-shock condition, mutant clones (green triangles) became losers in the deterministic model (thick green and pink lines in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eC and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). In contrast, a subset of small clones survived in the stochastic model, reflecting variability introduced by stochastic processes and differences in experimental conditions. Under the 20 min and 30 min heat-shock conditions, the deterministic model predicted an initial decrease in relative clone size until 2 d ACI, followed by an increase thereafter (20 min: Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eG, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eH; 30 min: Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eK, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eL). In the stochastic model, however, the simulated data were highly scattered: some mutant clones were eliminated, whereas others persisted (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eH and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eL). This variation was most pronounced under the 30-min heat-shock condition (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eL). Even under the 60 min heat-shock condition, the relative clone size first decreased until 2 d ACI and then increased thereafter in the model simulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eO and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eP). As a consequence of this late-stage increase, most mutant clones ultimately survived.\u003c/p\u003e\n\u003ch3\u003eMathematical simulation predicts a late-stage, progressive increase in growth capacity in mutant cells\u003c/h3\u003e\n\u003cp\u003eIntriguingly, although most of the computer simulation results reproduced the corresponding experimental data, we identified several notable discrepancies. In the simulations for the 30 min and 60 min heat-shock conditions, tissue sizes reached saturation by 5\u0026ndash;6 d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eK and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eO), whereas in the experiments some tissues exhibited tumorous overgrowth (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eI and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eM), with sizes deviating up to 250\u0026ndash;300\u0026times;10\u003csup\u003e3\u003c/sup\u003e \u0026micro;m\u003csup\u003e2\u003c/sup\u003e from the majority level of around 150\u0026times;10\u003csup\u003e3\u003c/sup\u003e \u0026micro;m\u003csup\u003e2\u003c/sup\u003e. This discrepancy was particularly pronounced at 6 d ACI under the 60-min heat-shock condition. These mathematical and experimental observations suggest the existence of an additional mechanism that promotes mutant cell proliferation when a particular condition is fulfilled. A plausible candidate for this condition is a time-dependent process, whereby mutant cells undergo a delayed transition to an overgrowth phenotype. Indeed, incorporating a time-dependent increase in growth capacity at later developmental stages (implemented as an increased carrying capacity, \u003cem\u003eK\u003c/em\u003e\u003csub\u003e\u003cem\u003ey\u003c/em\u003e\u003c/sub\u003e) produced tissue size deviations similar to those observed experimentally (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eQ and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eR).\u003c/p\u003e \u003cp\u003e \u003cb\u003eLarge\u003c/b\u003e \u003cb\u003escrib\u003c/b\u003e\u003csup\u003e\u003cb\u003eKD\u003c/b\u003e\u003c/sup\u003e \u003cb\u003eclones evade cell competition and overgrow via elevated Yki activity\u003c/b\u003e\u003c/p\u003e \u003cp\u003eWe next experimentally investigated the molecular mechanism by which large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones acquire tumorigenic potency at later developmental stages. Using the mitotic marker phospho-Histone H3 (PH3), we confirmed that wing disc cells normally ceased proliferation by 5d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eA), as previously reported\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. In contrast to this developmental arrest, large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones exhibited enhanced proliferation at 5d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eB, orange arrow). Small \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones in the same tissue had already ceased proliferating (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eB, orange arrowhead), consistent with our simulation-based prediction that tumorigenic potency depends on clone size. Importantly, at an earlier stage (3d ACI), large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones did not exhibit increased proliferation compared with neighboring wild-type tissue; rather, their proliferative activity was comparable to or even lower than that of wild-type cells (Fig. S8A, quantified in S8B). These data suggest that enhanced proliferation is acquired by large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones at later developmental stages.\u003c/p\u003e \u003cp\u003eWe then examined the signaling pathways that might underlie this late-stage proliferative activity. Interestingly, we found that expressions of \u003cem\u003ediap1\u003c/em\u003e and microRNA \u003cem\u003ebantam\u003c/em\u003e, which are canonical targets of the Hippo pathway effector Yki, were significantly elevated in \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e cells within large clones at 5d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eD, F), while small \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones, wild-type clones, and large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones at earlier developmental stages did not show such inductions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eC, E, Fig. S8C and S8D).\u003c/p\u003e \u003cp\u003eTo assess the functional contribution of Yki, we knocked down \u003cem\u003eyki\u003c/em\u003e simultaneously with \u003cem\u003escrib\u003c/em\u003e knockdown at the time of clone induction and quantified clone size at 3d ACI. Although Yki target gene upregulation was not yet detectable at this stage, \u003cem\u003eyki\u003c/em\u003e knockdown significantly reduced the size of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eH and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eI, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eK). Importantly, \u003cem\u003eyki\u003c/em\u003e knockdown alone also significantly reduced clone size compared with wild-type controls at 3d ACI (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eG and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eJ, quantified in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eK), confirming its contribution to basal epithelial cell proliferation. Therefore, the reduced growth of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones upon \u003cem\u003eyki\u003c/em\u003e knockdown likely reflects both general proliferative function and the requirement of Yki in the later acquisition of tumorigenic growth capacity. Together, these findings support a model in which large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones progressively engage Yki-dependent growth programs over time, enabling them to escape developmental growth arrest and undergo tumorigenesis.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur data show that the win-lose fate of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones in the growing imaginal disc is determined by their initial clone size. As summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the outcome of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones can be represented as a size-dependent phase transition in clone dynamics. Simulation analysis reveals a size threshold: when the initial mutant clone size remains below approximately 17% of the tissue, \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones are completely eliminated (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003eA; the threshold ratio, indicated by the navy dashed line, was estimated using Eq. (S3) in Supplemental Materials S1). In contrast, when the initial clone size exceeds this threshold, a substantial fraction of mutant cells is eliminated during early developmental stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003eB), yet the remaining population subsequently expands and surpasses the normal tissue carrying capacity at later stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). Ultimately, large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones persist and progress toward tumorigenic overgrowth (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). These data are consistent with previous reports showing that small clones of tumorigenic Ras\u003csup\u003eV12\u003c/sup\u003e/\u003cem\u003elgl\u003c/em\u003e\u003csup\u003e\u0026minus;/\u0026minus;\u003c/sup\u003e or \u003cem\u003eRab5\u003c/em\u003e mutant cells are eliminated by cell competition, whereas larger clones overgrow within the tissue\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. In the present study, we quantitatively analyze the win-lose fate of \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones of different initial sizes and, by integrating experimental data with mathematical modeling, provide a framework explaining how initial clone size determines long-term win-lose fate (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eImportantly, our experimental data indicate that this late-stage clonal expansion is accompanied by a progressive increase in Yki activity. Elevated expression of Yki target genes is observed specifically in large clones at later developmental stages, suggesting that detectable Yki target induction is not evident at clone initiation but becomes progressively enhanced over time in sufficiently large clone domains. Thus, clone size does not merely influence growth rate but defines whether mutant cells remain susceptible to elimination or transition into a Yki-dependent tumorigenic state.\u003c/p\u003e \u003cp\u003eNevertheless, several limitations should be considered. Because temporally restricted knockdown of \u003cem\u003eyki\u003c/em\u003e specifically at later developmental stages was not technically feasible in our system, we cannot formally distinguish between a requirement of basal Yki activity and a stage-specific activation during tumorigenic expansion. Nonetheless, the lack of detectable Yki target induction at early stages argues against strong pathway activation at clone initiation and instead supports a progressive increase in Yki signaling as clone domains expand over time.\u003c/p\u003e \u003cp\u003eIn this context, the upstream mechanisms that drive the progressive increase in Yki activity within large \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e clones remain to be elucidated. One possibility is that growth factors or other signaling molecules produced by \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e cells accumulate within sufficiently large clone territories, thereby modifying the local microenvironment and reducing their susceptibility to elimination. Such size-dependent accumulation of signaling inputs could provide a mechanistic link between clone expansion and Yki activation. Notably, \u003cem\u003escrib\u003c/em\u003e\u003csup\u003e\u003cem\u003eKD\u003c/em\u003e\u003c/sup\u003e cells are eliminated by cell competition in mammalian epithelial cell cultures\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and the Yki homolog YAP can convert mammalian cells into winners of cell competition\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. These observations raise the possibility that similar size-dependent win-lose transitions may operate in mammalian tissues and contribute to early tumorigenic progression. Together, our findings suggest that tumorigenic potential may emerge not solely from oncogenic mutation, but from the integration of clone size, competitive context, and progressive signaling activation over time.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eFly culture and generation of clones\u003c/h2\u003e \u003cp\u003eFlies were cultured at 25\u0026deg;C on standard fly food unless otherwise noted. Fly food containing \u003cem\u003eerg2Δ\u003c/em\u003eyeast was prepared as discribed previously\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Adult flies were allowed to mate for 6 h, and the eggs laid during this period were collected for rearing. The time point 3 h after the onset of mating was defined as 0 h after egg laying (AEL0). Vials containing developing larvae were subjected to a 37\u0026deg;C heat shock at 48h after AEL0. Larvae were dissected at designated time points after the initiation of the heat shock, and wing discs were collected. Fluorescently labeled mitotic clones were produced in larval imaginal discs using the following strains: \u003cem\u003ehsFLP\u003c/em\u003e\u003csup\u003e\u003cem\u003e122\u003c/em\u003e\u003c/sup\u003e; \u003cem\u003eAct\u0026thinsp;\u0026gt;\u0026thinsp;y\u0026thinsp;+\u0026thinsp;\u0026gt;\u0026thinsp;Gal4, UAS-GFP\u003c/em\u003e (tester), \u003cem\u003eUAS-Dcr2\u003c/em\u003e (BDSC #24646), \u003cem\u003eUAS-scrib-RNAi\u003c/em\u003e (HMS01490, BDRC#35748), \u003cem\u003ebantam-LacZ\u003c/em\u003e (BDSC #10154), \u003cem\u003ediap1-lacZ/Diap1\u003c/em\u003e\u003csup\u003e\u003cem\u003ej5C8\u003c/em\u003e\u003c/sup\u003e (BDSC #12093) and \u003cem\u003eUAS-Yki-RNAi\u003c/em\u003e (NIG #4005R-2).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eHistology\u003c/h3\u003e\n\u003cp\u003elarvae were dissected in phosphate buffered saline (PBS) and fixed with 4%-Paraformaldehyde (PFA) for 5 min on ice, followed by 20 min at room temperature. The fixed larvae were washed 3 times with PBT (PBS\u0026thinsp;+\u0026thinsp;0.1% TritonX100) for 20 min each, and were then blocked with PBTn (5% donkey serum\u0026thinsp;+\u0026thinsp;PBT) solution for 30 min. For immunostaining, they were incubated overnight at 4\u0026deg;C with primary antibodies in PBTn. After four 30-min washes with PBT, the samples were re-blocked in PBTn for 30 minutes, then incubated with secondary antibodies for 2 hours at room temperature. Following another set of four 30-minute washes with PBT, the samples were mounted using DAPI-containing SlowFade Gold Antifade Reagent (Invitrogen, #S36939). Images were taken with Leica SP5 confocal microscopy. Primary antibodies used are as follows; Phospho-Histone H3 (Ser10) (1:100; Cell Signaling #9706) and chicken anti-β-gal (1:10; abcam #ab9361). Secondary antibodies used are as follows; Goat anti-rabbit Alexa 546 (1:250; Invitrogen #A11035) and Goat anti-chicken Alexa 647 (1:250; Invitrogen #A21449). Disc size and GFP-labeled clone size were measured using OpenCV (cv2), a Python library for image processing. Statistical analyses were performed using RStudio.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis work was supported in part by the MEXT/JSPS KAKENHI (22H05619 and 24H01394 to S.N., 20H05945, 22H02616 and 21K19257 to S.O., 26114002 and 25127717 to A.T., and 21H05284 and 21H05039 to T.I.), Japan Science and Technology Agency (Moonshot Research \u0026amp; Development: Grant Number JPMJPS2022) to S.O., AMED-CREST, Japan Agency for Medical Research and Development (23gm1710002h0002) to T.I., the Princess Takamatsu Cancer Research Fund to S.O. and the Takeda Science Foundation to S.O. and T.I.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eS.K., S.N., S.O., A.T. and T.I. designed the study; S.K. and S.O. performed experiments; S.N. and A.T. developed the mathematical model; S.K., S.N., S.O., A.T. and T.I. analyzed the data; S.K., S.N., S.O., A.T. and T.I. wrote the manuscript.All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank Igaki and Takamatsu lab members for discussions; K. Baba, M. Tanaka, and M. Koijima for technical support. We also thank the Bloomington Drosophila Stock Center (Indiana), the Vienna Drosophila Resource Center (Vienna), the National Institute of Genetics Stock Center (Mishima), and the Drosophila Genomics and Genetic Resources (Kyoto) for fly stocks.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLevayer, R. \u0026amp; Moreno, E. Mechanisms of cell competition: themes and variations. \u003cem\u003eJ. Cell. Biol.\u003c/em\u003e \u003cb\u003e200\u003c/b\u003e, 689\u0026ndash;698. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1083/jcb.201301051\u003c/span\u003e\u003cspan address=\"10.1083/jcb.201301051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVincent, J. P., Fletcher, A. G. \u0026amp; Baena-Lopez, L. A. Mechanisms and mechanics of cell competition in epithelia. \u003cem\u003eNat. Rev. Mol. Cell. Biol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 581\u0026ndash;591. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nrm3639\u003c/span\u003e\u003cspan address=\"10.1038/nrm3639\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmoyel, M. \u0026amp; Bach, E. A. Cell competition: how to eliminate your neighbours. \u003cem\u003eDevelopment\u003c/em\u003e \u003cb\u003e141\u003c/b\u003e, 988\u0026ndash;1000. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1242/dev.079129\u003c/span\u003e\u003cspan address=\"10.1242/dev.079129\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Gregorio, A., Bowling, S. \u0026amp; Rodriguez, T. A. Cell Competition and Its Role in the Regulation of Cell Fitness from Development to Cancer. \u003cem\u003eDev. Cell.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 621\u0026ndash;634. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.devcel.2016.08.012\u003c/span\u003e\u003cspan address=\"10.1016/j.devcel.2016.08.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaker, N. E. Cell competition. \u003cem\u003eCurr. Biol.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, R11\u0026ndash;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cub.2010.11.030\u003c/span\u003e\u003cspan address=\"10.1016/j.cub.2010.11.030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoods, D. F. \u0026amp; Bryant, P. J. The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions. \u003cem\u003eCell\u003c/em\u003e \u003cb\u003e66\u003c/b\u003e, 451\u0026ndash;464. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0092-8674(81)90009-x\u003c/span\u003e\u003cspan address=\"10.1016/0092-8674(81)90009-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1991).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBilder, D. \u0026amp; Perrimon, N. Localization of apical epithelial determinants by the basolateral PDZ protein Scribble. \u003cem\u003eNature\u003c/em\u003e \u003cb\u003e403\u003c/b\u003e, 676\u0026ndash;680. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/35001108\u003c/span\u003e\u003cspan address=\"10.1038/35001108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHumbert, P., Russell, S., Richardson, H. \u0026amp; Dlg Scribble and Lgl in cell polarity, cell proliferation and cancer. \u003cem\u003eBioessays\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 542\u0026ndash;553. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/bies.10286\u003c/span\u003e\u003cspan address=\"10.1002/bies.10286\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakino, K., Ohsawa, S. \u0026amp; Igaki, T. Loss of Rab5 drives non-autonomous cell proliferation through TNF and Ras signaling in Drosophila. \u003cem\u003eDev. Biol.\u003c/em\u003e \u003cb\u003e395\u003c/b\u003e, 19\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ydbio.2014.09.003\u003c/span\u003e\u003cspan address=\"10.1016/j.ydbio.2014.09.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBallesteros-Arias, L., Saavedra, V. \u0026amp; Morata, G. Cell competition may function either as tumour-suppressing or as tumour-stimulating factor in Drosophila. \u003cem\u003eOncogene\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, 4377\u0026ndash;4384. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/onc.2013.407\u003c/span\u003e\u003cspan address=\"10.1038/onc.2013.407\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBos, M., Burnet, B., Farrow, R. \u0026amp; Woods, R. A. Development of Drosophila on sterol mutants of the yeast Saccharomyces cerevisiae. \u003cem\u003eGenet. Res.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 163\u0026ndash;176. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/S0016672300016840\u003c/span\u003e\u003cspan address=\"10.1017/S0016672300016840\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1976).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParkin, C. A. \u0026amp; Burnet, B. Growth arrest of Drosophila melanogaster on erg-2 and erg-6 sterol mutant strains of Saccharomyces cerevisiae. \u003cem\u003eJ. Insect. Physiol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 463\u0026ndash;471. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0022-1910(86)90007-7\u003c/span\u003e\u003cspan address=\"10.1016/0022-1910(86)90007-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1986).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNishikawa, S., Takamatsu, A., Ohsawa, S. \u0026amp; Igaki, T. Mathematical model for cell competition: Predator-prey interactions at the interface between two groups of cells in monolayer tissue. \u003cem\u003eJ. Theor. Biol.\u003c/em\u003e \u003cb\u003e404\u003c/b\u003e, 40\u0026ndash;50. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jtbi.2016.05.031\u003c/span\u003e\u003cspan address=\"10.1016/j.jtbi.2016.05.031\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKatsuyama, T. \u0026amp; Paro, R. Innate immune cells are dispensable for regenerative growth of imaginal discs. \u003cem\u003eMech. Dev.\u003c/em\u003e \u003cb\u003e130\u003c/b\u003e, 112\u0026ndash;121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.mod.2012.11.005\u003c/span\u003e\u003cspan address=\"10.1016/j.mod.2012.11.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenendez, J., Perez-Garijo, A., Calleja, M. \u0026amp; Morata, G. A tumor-suppressing mechanism in Drosophila involving cell competition and the Hippo pathway. \u003cem\u003eProc. Natl. Acad. Sci. U S A\u003c/em\u003e. \u003cb\u003e107\u003c/b\u003e, 14651\u0026ndash;14656. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.1009376107\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1009376107\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, C. L., Schroeder, M. C., Kango-Singh, M., Tao, C. \u0026amp; Halder, G. Tumor suppression by cell competition through regulation of the Hippo pathway. \u003cem\u003eProc. Natl. Acad. Sci. U S A\u003c/em\u003e. \u003cb\u003e109\u003c/b\u003e, 484\u0026ndash;489. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.1113882109\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1113882109\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNorman, M. et al. Loss of Scribble causes cell competition in mammalian cells. \u003cem\u003eJ. Cell Sci.\u003c/em\u003e \u003cb\u003e125\u003c/b\u003e, 59. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1242/jcs.085803\u003c/span\u003e\u003cspan address=\"10.1242/jcs.085803\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMamada, H., Sato, T., Ota, M. \u0026amp; Sasaki, H. Cell competition in mouse NIH3T3 embryonic fibroblasts is controlled by the activity of Tead family proteins and Myc. \u003cem\u003eJ. Cell. Sci.\u003c/em\u003e \u003cb\u003e128\u003c/b\u003e, 790\u0026ndash;803. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1242/jcs.163675\u003c/span\u003e\u003cspan address=\"10.1242/jcs.163675\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSidiropoulos, N. et al. SinaPlot: An Enhanced Chart for Simple and Truthful Representation of Single Observations Over Multiple Classes. \u003cem\u003eJ. Comput. Graphical Stat.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e, 673\u0026ndash;676. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/10618600.2017.1366914\u003c/span\u003e\u003cspan address=\"10.1080/10618600.2017.1366914\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"cell competition, tumorigenesis, Yorkie, Drosophila epithelium, mathematical modeling","lastPublishedDoi":"10.21203/rs.3.rs-9135893/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9135893/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNormal epithelial tissues often exert tumor-suppressive effects against newly emerged oncogenic cells. In \u003cem\u003eDrosophila\u003c/em\u003e epithelia, clones of oncogenic polarity-deficient cells mutant for \u003cem\u003escribble\u003c/em\u003e (\u003cem\u003escrib\u003c/em\u003e) are eliminated by cell competition when surrounded by wild-type cells. Here, we show that the win-lose fate of \u003cem\u003escrib\u003c/em\u003e mutant cells depends on their initial population size. Small \u003cem\u003escrib\u003c/em\u003e-knockdown (\u003cem\u003escrib\u003c/em\u003e\u003csup\u003eKD\u003c/sup\u003e) clones are efficiently eliminated from wing imaginal epithelia, whereas larger clones initially behave as losers but later escape elimination and overgrow into tumors. This shift in competitive behavior in large clones is accompanied by a progressive activation of the Hippo pathway effector Yorkie (Yki). To dissect this intriguing growth dynamics, we extend our previously proposed deterministic mathematical model of cell competition to a stochastic framework incorporating experimentally observed sources of variability. The model reproduces the large variability observed in clone growth dynamics. Furthermore, addition of the effect of late-stage, progressive growth acceleration to the mathematical model recapitulated the experimentally observed size-dependent progressive tumorigenesis. Our findings establish the initial size of oncogenic cell population as the key determinant of the win-lose fate and provide a potential mechanism by which oncogenic clones evade cell competition through size-dependent, progressive elevation of Yki activity.\u003c/p\u003e","manuscriptTitle":"Oncogenic cells evade cell competition and evolve into tumors through clone size-dependent, progressive elevation of Yki activity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-15 02:57:59","doi":"10.21203/rs.3.rs-9135893/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-27T09:29:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"69671676767541951123451934687811361668","date":"2026-04-13T04:45:40+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-07T18:11:59+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-07T18:01:37+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-07T14:00:24+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-31T06:44:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-31T02:52:34+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7e73b803-b61f-4ca4-b852-aed5f7c9bff5","owner":[],"postedDate":"April 15th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":66179065,"name":"Biological sciences/Cancer"},{"id":66179066,"name":"Biological sciences/Cell biology"},{"id":66179067,"name":"Biological sciences/Developmental biology"}],"tags":[],"updatedAt":"2026-04-15T02:57:59+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-15 02:57:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9135893","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9135893","identity":"rs-9135893","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}