Figures and data
Competition induced compaction of HRasV12-transformed mutants in vivo and in vitro.
A. Schematic representation of the genetic makeup and experimental workflow pertaining to the cell competition mice model. B. Heterogenous competing population constituted by wildtype and HRasV12-expressing transformed MDCK cells (40:1, number ratio) cultured on a 4 kPa polyacrylamide gel. Induction with doxycycline (dox) shows progressive GFP-HRasV12 expression and concomitantly, the transformed cells round up to finally extrude out of the monolayer (transformed colony is marked by blue dotted circle, also shown in the insets), hpi-hours post induction. Scale bar, 50µm. C. Representative immunofluorescence images of intestinal villi epithelium from villin-Cre-ERT2; LSL-hrasV12-IRES-egfp (villin-HRasV12-GFP) mice after tamoxifen injection, as detailed in A, depicting selective competitive extrusion of HRasV12-expressing transformed (villin-HRasV12-GFP) but not wildtype cells. Scale bar, 50 µm. D. 3-dimensional rendered view of a heterogenous competing population of MDCK cells cultured over 4 kPa polyacrylamide gel. Orange arrow heads point towards the extruding transformed cell groups. Scale bar, 50 µm. E. cell spreading area preferred by HRasV12 and simultaneously EGFP expressing transformed cells in villin-HRasV12-GFP genetic background under homeostatic conditions (referred to as 𝛥ℎ) or under wildtype epithelium and transformed competition condition (referred to as 𝛥𝑐). Bottom panels depict the corresponding segmented cell masks. Scale bar, 10 µm. F. Cell spreading area preferred by HRasV12-transformed MDCK cells under homeostatic conditions (referred to as 𝛥ℎ) or under competition condition (referred to as 𝛥𝑐). Scale bar, 20 µm.
Competition dependent compressive stress at the HRasV12-transformed mutant locus.
A. Traction force microscopy generated map of the forces exerted by MDCK-wildtype (-WT) cells during competition between HRasV12-transformed mutant MDCK and wildtype cells (transformed colony is marked by red dotted circle), analysis done post 2 hours of doxycycline (dox) induction. Scale bar, 50 µm. B. Infographic illustration describing the principle underlying monolayer stress microscopy. C. Stress microscopy generated mean normal stress map depicting compressive stress at the transformed loci (marked by red dotted circle) during competition between WT and MT cells, analysis done post 2 hours of induction with dox. Scale same as in A. D. Infographic illustration describing Bayesian force inference. Bayesian force inference generated E. interfacial tension map and F. relative intracellular pressure map during competition between WT and MT cells (transformed colony marked by black dotted circle), inference performed post 2 hours induction with dox. Scale same as in A. G. normalized relative intracellular pressure characterized with Bayesian force inference, plotted from a pool of 3 independent experiments, analysis performed post 2 hours induction with dox. Data are mean±sem. Statistical significance was calculated using Unpaired t-test with Welch’s correction. H. Relative compaction, 𝛥, defined as 𝛥 = (𝛥𝑐 – 𝛥ℎ)/ 𝛥ℎ is determined for competing HRasV12-transformed cells under different times of induction (and thus experiencing different levels of compressive stress (plotted in Δσ). Data is represented as the relative compaction with the maximum compressive stress obtained for transformed colonies revealing a nearly linear relationship between compressive stress and relative compaction. Data are mean±sem from three independent experiments. Statistical significance was calculated using Unpaired t-test with Welch’s correction.
Perturbed adherens-based cell-cell junctions as distinct mechano-phenotype of HRasV12 transformed cells in vivo and in vitro.
A,B. Representative immunofluorescence images of mice intestinal epithelium depicting endogenous E-cadherin in wildtype cells under villin-HRasV12-GFP background (A) or HRasV12 and simultaneously EGFP (HRasV12-IRES-EGFP) expressing cells in villin-HRasV12-GFP background (B). Insets are shown corresponding to each image from the two scenarios, Scale bar, 10 µm for all image panels and 2 µm for all insets. C,D. Super-resolution spinning-disk confocal microscopy using optical photon reassignment (SoRa) image panels to show endogenous E-cadherin in the intestinal villi epithelium in wildtype cells (C) or in HrasV12 and simultaneously EGFP expressing cells (D) under villin-HRasV12-GFP background. Insets are shown corresponding to each image from the two scenarios. Scale bar, 10 µm for all image panels and 2 µm for all insets. Representative panels to show immunostained endogenous E-cadherin in (E) MDCK-wildtype (-WT) and or (F) MDCK-HRasV12 transformed cells cultured in homogenous conditions. Insets are shown (marked by red boxes) to represent one cell in each condition. Scale bar, 10 µm for all image panels and 2 µm for all insets. G,H. Respective intensity line-plot (Magenta line refers to the intensity plot line) for WT (G) or transformed (H) cells. DAPI intensity (marked by cyan line) is plotted for reference to the nucleus. Pink boxes refer to the cytosol in each cell. Intensity values are normalized to the highest value within each dataset. E-cadherin shows distinct junctional accumulation in WT cells but diffused cytoplasmic signal in transformed cells. I,J. Super-resolution spinning-disk confocal microscopy using optical photon reassignment (SoRa) image panels to show endogenous E-cadherin in WT (I) or transformed (J) cells. Junctional E-cadherin is shown in insets from the two cell-types to depict E-cadherin loss in the transformed cells. Scale bar, 10 µm for all image panels and 2 µm for all insets.
In silico self-propelled Voronoi model recapitulates competition-dependent mechanical features to discover differential collective compressibility, empirically tested by Gel Compression Microscopy.
A. Junctional tension at the edges shared by cells in a circular island of transformed mutant (M) cells (which constitute 12% of the total number of cells in a randomly generated ensemble, with the island’s diameter being approximately 40% of the length of the box) is significantly lower than that of their surrounding wild-type (WT) counterparts. This is due to the single-cell parameters of the M cells, which includeKMA = 0.7, AM0 = 0.8, PM0 = 3.8, which corresponds to a shape index of 
Theoretical modeling reveals different regimes of mechanical phenotypes in cell competition.
In the computational simulations, we assessed both the bulk and shear moduli of the cell colonies. The bulk modulus provided insights into a tissue’s mechanical response to isotropic compression, while the shear modulus was essential for evaluating the tissue’s reaction to shear forces. These mechanical properties were analyzed for a wide range of parameters 𝑘, AM0 and 𝛿𝑞0. Here, 𝑘 encoded competition in cell area elasticities between the cell types. When 𝑘 < 1, the transformed mutant (M) colony was more compressible than the wildtype (WT) cells surrounding them. On the other hand, 𝛿𝑞0encoded the competition in cellular junctional properties; when 𝛿𝑞0 > 0, the transformed cells had weakened cellular junctions compared to the WT cells. A. Behavior of the differential bulk modulus 𝛥 between WT and M cells in the 𝑘 − 𝛿𝑞0 plane for AM0 = AWT0 = 1. The black dotted line corresponds to 𝛥 = 0. B. 𝛥 0 0 plotted as a function of 𝑘 AM0 for a range of 𝛿𝑞0 values. The inset shows 𝛥 − 𝑏 𝛿𝑞0 as a function of 𝑘 AM0 for the same data. Here we used 𝑏 = 1.7 ± 0.2. C. The behavior of the differential colony tension 𝛥𝜏 in the 𝑘 − 𝛿𝑞0 plane for AM0= 1. The white dotted line corresponds to 𝛥𝜏 = 0. D. 𝛥𝜏 plotted as a function of 𝛿𝑞0 for a range of 𝑘 values (and AM0 = 1). The inset shows 𝛥𝜏/𝑘𝑑𝑑 as a function of 𝛿𝑞0 exhibiting a scaling collapse. Here we find 𝑑 = 0.53 ± 0.12. Both the scaling parameters in B and D are obtained by minimizing error functions which we defined to measure the closeness of the respective data sets. E. based on the analysis of Fig. 5A-D, we predict a phase diagram which qualitatively distinguishes between mechanical phenotypes that are favored in non-proliferative mechanical cell competition. Specifically, a colony of cells with lower 𝛿𝑞0and higher 𝑘 AM0 will have a competitive advantage, as they exhibit lower compressibility and higher rigidity. The predicted phase diagram serves as a useful tool for identifying such mechanical phenotypes.
Compromised E-cadherin based cell-cell junctions in the HRasV12-transformed mutants underlie competition induced compressive stress as well as rescued collective compressibility.
A,B. Representative image panels to show immunostained MDCK-wildtype (-WT) or MDCK-HRasV12-transformed mutant cells for E-cadherin extracellular domain marked by DECMA-1. The immunostaining is performed A, WITH detergent (TritonX-100) or B, WITHOUT detergent to demonstrate the inherent instability in MT-MT junctions but not WT-WT junctions whereas the extracellular domain being intact in either case. Scale bar, 10 µm. C. E-cadherin live dynamics are studied in WT or MT cells mildly overexpressing E-cadherin-GFP. Representative differential interference contrast (DIC) and E-cadherin-GFP image panels are shown. Temporal dynamics for each cell-type is shown as a color projection map where each time stamp is differentially colored in the ‘yellow hot’ color panel and superimposed. Each time frame refers to 5 second imaging window. Scale bar, 5 µm. D. Super-resolution spinning-disk confocal microscopy using optical photon reassignment (SoRa) image panels to show endogenous E-cadherin in WT, MT or MT overexpressing E-cadherin-GFP. Junctional E-cadherin is shown in insets from the three scenarios to depict E-cadherin loss in transformed cells and rescue in transformed cells overexpressing exogenous E-cadherin-GFP. Scale bar, 10 µm for all image panels and 2 µm for all insets. E. Traction force microscopy and stress microscopy panels for cell competition between WT cells and MT cells overexpressing E-cadherin-GFP. Competition induced compressive stress as seen in Fig. 2C, is evidently rescued when E-cadherin is supplied exogenously in the transformed cells. Scale bar, 10 µm. F. Relative compressibility measurement based on Fig. 4I, shows a distinct and lower value for WT population whereas MT population shows no distinction compared to the polyacrylamide gel control, as also seen in Fig. 4J. Importantly, MT cells overexpressing E-cadherin-GFP show a distinctly similar value to the WT case, showcasing genetic rescue to the mechanical phenotype of compressibility differential. Data are mean±sem from three-four independent experiments. Statistical significance was calculated using Unpaired t-test with Welch’s correction.
