Fig 2: BMP signaling is required for RA sensitivity in a hyper-sensitive NB cell line.A ATRA IC50 values for all non-hematological cancer cell lines screened by GDSC1. Each dot represents a cell line (n = 759 in total). Box bounds show 25th and 75th percentiles, horizontal lines within the interquartile range (IQR) boxes indicate the median values for each group, and whiskers represent 1.5*IQR from the IQR values. Cancer types with ≤ 3 cell lines are not included. B Quantified flow cytometry results showing apoptosis in CHP-134, NB13, and D283 Med cell lines treated with ATRA (2 µM). Apoptotic cells were stained positive with annexin V and dead cells were stained positive with PI. Samples at day 9 for CHP-134 P = 1.67 × 10−7, NB13 P = 3.32 × 10−6; samples at day 6 for D283 Med P = 1.35 × 10−6. C Western blots for cleaved PARP (c-PARP) expression in CHP-134, NB13, and D283 Med cell lines treated with ATRA (2 µM). D Enrichr clustergram of CRISPR screen results (for BioPlanet 2019 pathways and GO Molecular Function 2021). Input: top 1% of genes with lowest negative and positive RRA score (191 + 191 genes). The genes and their directionality are listed in Supplementary Data 2. Red indicates screen hits. E Scatter plot highlighting top CRISPR screen hits involved in BMP signaling, ranked by positive RRA score (left, top knockouts cause resistance to ATRA) or negative RRA score (right, top knockouts sensitize to ATRA). RARA is also highlighted. F Viability of CHP-134 cells under 3-day (left) or 6-day (right) treatment with ATRA in the presence of DMSO or 0.5 µM K02288. Cell viability was measured with MTS. Data represent the mean ± SD of 4 independent replicates. For arrow indicated points, P = 3.89 × 10−6 for 3 days, P = 1.91 × 10−8 for 6 days. G. Western blots for c-PARP expression in CHP-134 treated with indicated compounds for 3 days (2 µM ATRA and 0.5 µM K02288). H Quantified flow cytometry results showing apoptosis in CHP-134 treated as in (G) for 3 days (left) and 6 days (right). For panel B and H, data represent the mean ± SD of 3 independent replicates. For panels C and G, GAPDH and β-Actin were used as loading control, the representative results of three independent experiments are shown. Two-sided t-test was used for panels B, F, and H. Source data are provided as a Source Data file.

Fig 3: Downstream BMP and RARA transcription factors cooperate to determine cell fate, following exposure to RA.A Heatmap showing ChIP-seq peaks for RARA, SMAD4, and SMAD9 genome-wide in CHP-134 cells following treatment with DMSO control or 2 µM ATRA for 1 day. The color scale represents binding peak intensity, scaled between 0 and 5 (see Methods). Each genomic region (y-axis) was centered on the binding peak, displaying 1 kb upstream and downstream, sorted by average peak intensity across all transcription factors. Two biologically independent replicates are shown. B Venn diagram showing the number and proportion of overlapped binding regions between RARA, SMAD4, and SMAD9 in DMSO-treated CHP-134 cells. C GSEA plot showing the top 4 enriched pathways (defined in BioPlanet 2019) of SMAD9 bound genes that gain RARA peaks following 2 µM ATRA treatment for 1 day in CHP-134 cells (see Methods). Running Enrichment Scores (colored lines). The vertical-colored lines indicate the members of each gene set, ranked by baseline SMAD9 binding intensity (y-axis, bottom panel). P-values were calculated by a one-sided empirical permutation test. D Volcano plot showing the differentially expressed genes (red or blue points) after 1, 3, or 6 days of 2 µM ATRA exposure in CHP-134 cells. Differential expression cutoffs: log2 fold change > 1 or < -1 and FDR < 0.05. E Bar plot showing the percentage of up or down-regulated genes that are bound by SMAD4 and RARA at 1, 3, and 6 days (x-axis) following 2 µM ATRA exposure in CHP-134 cells. P-values indicate the enrichment of RARA and SMAD4 bound genes among the differentially expressed genes, calculated using a two-sided Fisher’s exact test. P-values from left to right are 3.78 × 10−8, 5.98 × 10−8, 7.53 × 10−16, and 1.60 × 10−8. F Bar plot showing the number of up or down (blue) regulated genes (y-axis), bound by both RARA and SMAD4, following at 1-, 3-, and 6-days exposure 2 µM ATRA. Genes are grouped by their functional annotation (see Methods) in processes related to apoptosis (apop), cell cycle (cycle), and differentiation (diff). G Bar blot showing the expression (y-axis) of selected BMP signaling pathway genes (x-axis) in CHP-134 cells treated with ATRA for 1, 3, or 6 days (colors). Data represents the mean ± SD of two independent replicates. H Top-ranked consensus motif enrichments in the RARA binding regions of 1-day ATRA-treated CHP-134 cells, which are all basic helix-loop-helix (bHLH) factors. I Western blots (upper panel) of c-PARP expression in wild-type (−) and ID2-overexpressed (+) CHP-134 cells following 3 days of ATRA treatment at concentrations indicated. A representative result of three independent experiments is shown. GAPDH was used as a loading control. J Viability of wild-type (WT) and ID2-overexpressing CHP-134 cells after ATRA treatment for 6 days. Data represents the mean ± SD of three independent replicates. P = 1.43 × 10−5 at the arrow indicated point (two-sided t-test). K A conceptual model for how cells with high or low BMP signaling are directed into different fates upon RA exposure. *P < 0.05, **P < 0.01, ***P < 0.001. Source data are provided as a Source Data file.