Fig 2: Id2 promotes anoikis resistance and metastatic colonization in E-cadherin deficient breast cancer cells.Id2 controls anchorage-independent survival. Shown are the levels of anoikis resistance upon knockdown of Id2 in mouse mILC (A) or E-cadherin mutant human SKBR3 (B) after 24 or 96 h in suspension respectively. Scrambled shRNA (shScr) and Doxycycline (DOX) controls are depicted in white bars. The Student’s t test was used to determine statistical significance. ***p < 0.001, **p < 0.01, *p < 0.05. Results are expressed as mean and error bars represent standard deviation (SD). All experiments were repeated at least 3 times. C, D Id2 expression is necessary for lung colonization. Shown are bioluminescence signals (relative counts) from luciferase-tagged mILC cells injected intra-venously (iv) in the tail vain of recipient Nude mice (D). The effect of Id2 knockdown (shId2) was assessed and compared to controls (shScr). The number of mice showing lung colonization 4 weeks post injection are depicted below, together with the total number of mice injected in parenthesis. A two-tailed Fisher’s exact test was used to determine statistical significance. ***p < 0.001. Panel D shows a representative H&E stained lung sections (magnification 100X) of the mice depicted in C. Arrows point to metastatic outgrowth. For all mouse experiments, animals were randomized, and observers blinded for the experimental conditions.

Fig 3: Increased cytosolic localization of Id2 characterizes ILC.A–C Cytosolic Id2 expression is prevalent in ILC. Shown are representative tissue micro array (TMA) cores (top panels) from IDC-NST and ILC stained for Id2 expression (A) (brown). A total of 148 invasive breast cancer samples were scored as binary output and quantified in B. Note the overall increase in Id2 expression when comparing ILC versus IDC-NST in B, and the increase in the number of cases that express cytosolic Id2 when comparing ILC versus IDC-NST (C). Pearson Chi-square tests were used to determine statistical significance. P-values are stated in the table. D Shown are Id2 mRNA expression levels in the TCGA dataset. Wilcoxon rank tests were used to assess statistical significance. In total 533 breast cancer samples were analyzed (388 IDC-NST versus 145 ILC). E Id2 controls sustained indolent proliferation in ILC. Cartoon depicting a proposed model for E-cadherin dependent regulation of cell cycle progression. Show are E-cadherin expressing (left panel) and E-cadherin negative breast cancer (right panel). Note the consequences of E-cadherin loss (right panel), inducing cytosolic and nuclear translocation of p120 and a subsequent de-repression of Id2 expression by Kaiso. Id2 will accumulate in the cytosol, where it binds hypo-phosphorylated Rb and inhibits phosphorylation of Rb by CDK4/6. Cell cycle progression will be dampened, underpinning the indolent and low-grade features of E-cadherin mutant ILC. Under anchorage independent conditions (dissemination), Id2 expression is further increased due to a nuclear influx of p120 [26], leading to a full prevention of Rb phosphorylation, clustering E2F, and inhibition of cell cycle progression in G0/G1.

Fig 4: Id2 controls inhibition of cell cycle progression.A–C Id2 controls viability through inhibition of cell cycle progression under anchorage-independent culture conditions. FUCCI-2-tagged mILC-1 cells were cultured in suspension for 24 h and stained for apoptotic cells using cleaved Caspase-3 (blue; A). Cell cycle stages of mILC-1 cells in which either Id2 or Rb expression were inhibited, were quantified in B and viability was assessed in parallel using FACS analysis for Annexin-V and propidium iodine incorporation C. C The Student’s t test was used to determine statistical significance. **p < 0.01, *p < 0.05. Error bars depict the standard deviation. Size bar = 5 µm. D, E Id2 depletion causes an increase in cycling cells in a Rb and CDK4/6-dependent manner. Phosphorylated Rb (pRb Ser807/811) IF staining of Adh and Sus (24 h) mILC-1 cells (D). Note the absence of p-Rb signals (blue signals) and the reduction of G2/M cells in the Id2-depleted condition in the presence of the CDK4/6 inhibitor ribociclib (CDK4/6i; 3 µM). The number of cells expressing p-Rb from D are quantified in E. Outlines indicate cells with high pRb staining in G2/M phase. E HPF High Power Fields, showing one microscope image in high resolution. Five HPFs containing at least 120 cells per cell cycle state were quantified. E The Student’s t test with Welch’s correction was used to determine statistical significance. **p < 0.01, *p < 0.05. Error bars depict the standard deviation. (All) Representative experiments are shown from at least three biological independent replicates.

Fig 5: Id2 is upregulated in PDO-PDX human ILC models.A Western blot analysis showing E-cadherin levels in the classical ILC KCL320 and in the E-cadherin expressing non-classical ILC PDO model P008 (top panel). Akt was used as loading control (bottom panel). B Schematic representation of the CDH1 mutation S180Y and the consequences for cell-cell contacts. Note that although E-cadherin S180Y is unable to establish homotypic in trans interactions with adjacent cells, it retains p120-catenin at the cell membrane. C Immunofluorescence showing expression and localization of E-cadherin (red; top panels), p120-catenin (p120; green, middle panels). Merged images are shown in the bottom panels. Note the cytosolic p120 expression in E-cadherin deficient cells versus the co-localization of E-cadherin and p120 in E-cadherin-expressing P008 ILC cells. Isogenic MCF7 and MCF::?CDH1 cell lines were used as a controls. Size bar = 5 µm. D RT-qPCR analysis of Id2 mRNA upregulation in anchorage-independent cells (Sus/168 h). Note the 4-fold upregulation of Id2 in KCL320 suspended cells compared to adherent conditions (Ad). Relative Id2 mRNA expression levels were normalized to either GAPDH or ACTB. The Student’s t test was used to determine statistical significance. ****p < 0.0001, ***p < 0.001. Error bars show the standard deviation. E Western blot analysis showing E-cadherin levels in the isogenic P008 and P008::?CDH1 PDO models (top panel). Akt was used as a loading control (bottom panel). F Representative differential interference contrast microscopy images of P008 and P008::?CDH1 cultured in 3D basement membrane extract-derived matrix. Note the noncoherent “grape-like” phenotypic appearance of both PDO models. Size bar = 50 µm. G RT-qPCR analysis of Id2 mRNA upregulation in anchorage-independent cells. Note the approximate 6-fold upregulation in P008::?CDH1 cells compared to P008 suspended cells. Both organotypic cell lines were kept in suspension for 7 days before harvesting and analysis. Relative Id2 mRNA expression levels were normalized to either GAPDH or ACTB. The Student’s t test was used to determine statistical significance. ****p < 0.0001, ***p < 0.001. Error bars show the standard deviation. (All) Representative results are shown from at least three independent experiments.