Fig. 4: HER2 is required for endocrine-therapy-resistant growth of ER⁺ MLH1⁻ breast cancer cells.

Knockdown of endogenous HER2 using siRNA against HER2 or a scrambled control in MCF7 shLuc and shMLH1 cells validated by Western blotting (B) and followed by 2D growth assays for dose response to fulvestrant treatment (A). For shLuc vs. shMLH1 with siScr, p = 0.0001. Supporting data demonstrating similar response to tamoxifen and estrogen deprivation in Fig. S5A, B. C Growth of MCF7 shLuc and shMLH1 cells in response to specified therapeutic combinations represented as a bar graph. For shLuc vs. shMLH1 with fulvestrant treatment, p = 0.0009, and for shMLH1 fulvestrant vs. lapatinib + fulvestrant treatment, p = 3.15e − 05. Supporting data from T47D in Fig. S5C. D Dose curve demonstrating response to neratinib and fulvestrant in MCF7 shLuc and shMLH1 cells. Supporting data demonstrating similar results in T47D cells and in response to tamoxifen in Fig. S5D, E. E 3D growth in Matrigel of MCF7 shLuc and shMLH1 cells in response to specified treatments. Representative images for each treatment group (except lapatinib, which showed no visible difference from vehicle-treated) shown alongside quantification. For both shLuc vs. shMLH1 fulvestrant-treated, and shMLH1 fulvestrant vs. fulvestrant + lapatinib-treated, p = 0.0002. Supporting data in T47D cells in Fig. S6A. For dose curve experiments A, C, D three independent biological replicates were quantified for each group and each dose point. IC50 values were determined over three independent experiments and compared for statistical differences. Circles (A, D) represent mean growth relative to vehicle-treated cells over 7 days of treatment and error bars the standard deviation. Box plots show median, quartiles, minima and maxima, and outliers at 1.5× IQR. All statistical comparisons used the two-sided Student’s t test. All experiments were conducted >2 times. Source data for this figure available with paper.