(A) Workflow diagram in which 50 μg of protein was extracted from control ESCs and ESC-CMs, as well as ESCs and ESC-CMs treated with 0.01 μmol/L of doxorubicin for 48 hours and analyzed using isobaric labeling mass spectrometry. (B) PCA plot of protein abundance in ESCs (circles) and ESC-CMs (triangles) that were untreated (green) or treated with 0.01 μmol/L of doxorubicin (blue). (C) Volcano plot showing that low-dose doxorubicin caused differential expression of protein coding genes only in ESCs, and not in ESC-CMs, at Benjamini-Hochberg–adjusted P (FDR) = 0.01. Dashed line: nominal differential expression at P 0.05 and absolute logFC 0.5; green data points: significant differential expression at 1% FDR. (D) UMAP projection of a reprocessed public scRNA-Seq data set (29) at multiple time points of cardiac differentiation corresponding to the gene expression profiles of different cell populations at different stages of development (e.g., cluster 18 shows undifferentiated cells and clusters 5 and 6 correspond to cardiomyocyte-like cells), onto which we mapped the bulk proteomics data to deconvolve the cell type composition of each sample in the mass spectrometry experiment. (E) Proportion of each cell type in the bulk proteomics data as estimated with the aid of scRNA-Seq data. The proteomics profile of both the treated and untreated ESC-CM samples in the mass spectrometry experiments corresponded closely with the single-cell gene expression profiles of the cardiomyocyte clusters (5 and 6) with no significant differences in cell composition profiles after doxorubicin treatment. However, doxorubicin treatment in the ESC samples caused a shift in protein abundance that corresponded to a more mesoderm-like gene expression profile, consistent with preferential removal by doxorubicin of cells with high expression of pluripotency markers.