Fig. 2: Lapatinib-induced ChIPseq and RNAseq changes are associated with FOX family transcription factors.
From: FOXA1 and adaptive response determinants to HER2 targeted therapy in TBCRC 036

a Lapatinib induces 30 SEs, as determined by BRD4 ChIPseq density. Lapatinib in combination with JQ1 suppresses BRD4 binding in all regions, MED1 in all but one region, but does not affect H3K27Ac. Box plots: median, upper/lower quartile, and 1.5 IQR. b, c Lapatinib-mediated changes are strongly correlated with changes in H3K27Ac and MED1. d The log2 fold change binding in response to lapatinib for regions as in (a) and the log2 fold change in mRNA expression for the proximal genes as determined by RNAseq was used for unsupervised hierarchical clustering. Genes included based on ChIPseq density displayed increased chromatin association within 200 kb of transcription start site. Genes with shared induction in all instances (log2 fold change ≥1) were used as input for Dreme-TOMTOM analysis to identify enriched binding motifs. A single motif identified, GT(A/C)AACA, recognized by FOXA1 and FOXO3. Normalized RNAseq reads are shown for the FOX family of transcription factors in SKBR-3 cells after 48 h treatment with DMSO or 300 nM lapatinib. e Genes from (d) were used to query Enrichr. ENCODE TF ChIPseq data identified FOXA1 and p300 as potential factors. f ChIPseq analysis of FOXA1 binding in SKBR-3 cells with the regions of highest density annotated by proximal gene. g Unsupervised hierarchical clustering of FOXA1 ChIPseq density in SKBR-3 cells treated with DMSO, 300 nM lapatinib, and 300 nM JQ1, alone or in combination, at all genomic loci. Lapatinib and JQ1 each result in regions of increased FOXA1 chromatin binding. The combination of lapatinib and JQ1 results in broad disruption of FOXA1. h SKBR-3 cells were treated with siRNA pools, nontargeting (NT) control or FOXA1, prior to DMSO or 300 nM lapatinib treatment for 24 h and subsequent BRD4 ChIPseq analysis. The top 500 regions in lapatinib are plotted. Log2 fold change of lapatinib vs. DMSO with a control siRNA pool are plotted on the x-axis. Log2 fold changes of lapatinib vs. DMSO with FOXA1 knockdown are plotted on the y-axis. Dots are scaled relative to their lapatinib-induced ChIPseq density.