Fig 2: KLF4 modulates gene expression through changes in the enhancer landscape.a Scatterplot showing the relation between differential KLF4 binding at enhancer anchors identified by HiChIP, with target gene expression, in PAEC exposed to 15 dyn/cm2 of LSS vs ST conditions for 24 h. b Scatterplot showing the loss of enhancer anchors identified by HiChIP upon KLF siRNA (marked in red), and their impact on target gene expression. HiChIP enhancer anchors were derived from PAEC treated with KLF RNAi for 48 h prior to exposure to 15 dyn/cm2 of LSS vs ST conditions for an additional 24 h. c Scatterplot showing the relation between KLF4 binding at enhancer anchors identified by ABC, with target gene expression. n = 2 experimental replicates. d ABC, HiChIP and ChIP-Seq tracks illustrating a KLF4 bound enhancer loop targeting DUSP5, which is lost with KLF siRNA (left panel). DUSP5 mRNA was determined by RT-qPCR in PAEC exposed to 15 dyn/cm2 of LSS vs ST conditions for 24 h (top right panel). Silencing the enhancer by CRISPRi decreased DUSP5 mRNA levels (bottom right panel). e ABC, HiChIP, and ChIP-Seq tracks illustrating an enhancer loop targeting SMAD5, which was lost with KLF RNAi (left panel). SMAD5 mRNA was determined by RT-qPCR in PAEC exposed to 15 dyn/cm2 of LSS vs ST conditions for 24 h, and shown normalized to ST expression levels (top right panel). Silencing the enhancer by CRISPRi, decreased SMAD5 mRNA levels (bottom right panel). f ABC, HiChIP and ChIP-Seq tracks illustrating enhancer loops that target EDN1 under ST conditions, which are lost with LSS but are present with KLF siRNA (left panel). EDN1 mRNA was determined by RT-qPCR in PAEC exposed to 15 dyn/cm2 of LSS vs ST conditions for 24 h (top right panel). Silencing the enhancer increased EDN1 mRNA levels (bottom right panel). For a and b, N = 2 experimental replicates for KLF4 ChIP-Seq; n = 3 experimental replicates for HiChIP. For d–f, data are shown as the mean +/− s.e.m. For mRNA expression n = 5 experimental replicates; for CRISPRi experiments n = 3 replicates. P values were determined by Student’s two-tailed t-test. Source data are provided as a Source Data file.

Fig 3: KLF4 interacts with the SWI-SNF nucleosome remodeling complex to increase chromatin accessibility.a Heatmap showing the spectral counts obtained by Affinity Purification followed by Mass Spectrometry (AP-MS) of PAEC transduced with adenoviral vectors encoding a Flag-tagged mutant of KLF4 or GFP control. AP was performed using anti-Flag antibodies (FLAG) and anti-KLF antibodies (KLF). Proteins were inferred from the peptides against the human UniProt using an FDR of 1%. n = 2 experimental replicates. b Representative images of Proximity Ligation Assays (PLA) (left panel) of PAEC exposed to 15 dyn/cm2 of LSS or ST conditions for 24 h, show an interaction of KLF4 with BRG1 and SMARCC2 under LSS (magenta). Nuclei were stained with DAPI (blue). Number of foci per nucleus were quantified in 10 non-overlapping random fields of view per replicate (right panel). n = 4 experimental replicates. Data shown as the mean +/− s.e.m. P values were determined by Student’s two-tailed t-test. Scale bar, 20 μm. c Representative image of KLF4-BRG1 PLA in healthy rat lung tissue. The original PLA protocol was modified with longer incubation times to allow the reagents to fully penetrate the 350 μM thick sections. Note the interaction of KLF4 with BRG1 at straight sites of the vasculature that are exposed to LSS (upper left insert), while sites at bifurcations which are exposed to disturbed shear stress (DSS) do not show the KLF4-BRG1 interaction (bottom right insert). Bar graphs show the average number of foci per nucleus quantified in regions of DSS or LSS at 6 random bifurcations. KLF4-BRG1 interaction (magenta); vWF (gray, pseudo-color); αSMA (green). Nuclei were stained with DAPI, (blue). Scale bar, 20 and 10 μm in the higher magnifications. Data shown as the mean +/− s.e.m. P value was determined by Student’s two-tailed t-test. Source data are provided as a Source Data file.

Fig 4: KLF4 represses migration, invasion, and long-term cell growth of breast epithelial cells. a Western blot analysis and quantitation of KLF4 expression across eight different breast cell lines: MCF10A (10A), MCF-7, T47D, SKBR3, SUM159, HCC70, MDA-MB-468 (468), and MDA-MB-231 (231). b RT-qPCR analysis of KLF4 expression levels in MDA-MB-231 and MDA-MB-468 cell lines 3 days after infection with AdGFP (AdG) or AdKLF4 (AdK). c Western blot analysis and quantitation showing KLF4 expression in MDA-MB-231 and MDA-MB-468 following infection with AdGFP (G) or AdKLF4 (K) and compared to endogenous levels in MCF-10A cells. Two biological replicates were performed. Cells were infected with AdG or AdK for 2 days and then allowed to d migrate for 6 h or e invade into Matrigel for 24 h before they were stained and counted. f MDA-MB-231 and MDA-468 cells were infected with AdG or AdK and plated. After 24 h, they were counted using trypan blue exclusion. g RT-qPCR and h western blot analysis of the KLF4 expression in MCF10A cells 3 days post-infection with a non-targeting siRNA (siNS) or siRNA targeting KLF4 (siKLF4). Cells were infected with siNS or siKLF4 and allowed to i migrate for 6 h or j invade for 24 h before they were stained and counted. Cells were k, l infected with AdG or AdK or m transfected with siNS or siKLF4, and the cell number was counted at days 3–5 using trypan blue exclusion assay. Relative values were normalized versus the control, and error bars represent the standard deviation. Experiments were performed three independent times and in triplicate with *p < 0.05

Fig 5: KLF4 represses the EGFR signaling pathway. a Western blot analysis of KLF4, tEGFR, pEGFR (Y1068), tAKT, pAKT (S473), tERK1/2, and pERK1/2 (Y202/Y204) protein levels 3 days after AdGFP or AdKLF4 infection of MDA-MB-231 cells. b Graph depicting western blot quantification. All protein levels were normalized to total protein in the lane to control for loading, and phosphorylated proteins were normalized to their respective unphosphorylated proteins. AdKLF4 lanes were then expressed relative to AdGFP and are graphed with the horizontal bars indicating the median of the values of three replicate blots. Each dot represents the mean value from a different experiment, *p < 0.05. c, d Similar to a and b, but with lysates from infected MDA-MB-468 cells. e Western blot analysis of the same proteins in a, collected 3 days after transfection of MCF10A cells with siNS or siKLF4. f Quantitation and analysis of protein levels were completed as described in b. All western blots were completed in three independent experiments, each in triplicate, *p < 0.05