Novel FOXM1 inhibitor identified via gene network analysis induces autophagic FOXM1 degradation to overcome chemoresistance of human cancer cells

Abstract

FOXM1 transcription factor is an oncogene and a master regulator of chemoresistance in multiple cancers. Pharmacological inhibition of FOXM1 is a promising approach but has proven to be challenging. We performed a network-centric transcriptomic analysis to identify a novel compound STL427944 that selectively suppresses FOXM1 by inducing the relocalization of nuclear FOXM1 protein to the cytoplasm and promoting its subsequent degradation by autophagosomes. Human cancer cells treated with STL427944 exhibit increased sensitivity to cytotoxic effects of conventional chemotherapeutic treatments (platinum-based agents, 5-fluorouracil, and taxanes). RNA-seq analysis of STL427944-induced gene expression changes revealed prominent suppression of gene signatures characteristic for FOXM1 and its downstream targets but no significant changes in other important regulatory pathways, thereby suggesting high selectivity of STL427944 toward the FOXM1 pathway. Collectively, the novel autophagy-dependent mode of FOXM1 suppression by STL427944 validates a unique pathway to overcome tumor chemoresistance and improve the efficacy of treatment with conventional cancer drugs.

Fig. 1. STL treatment causes dose-dependent suppression of FOXM1 protein levels in cancer cell lines of different etiology.

a Structural formula of STL. b-e Various cell lines representing human prostate (b), HGSOC (c), colorectal (d), or NSCLC (e) cancer were treated with increasing concentrations of STL for 24 h. Total protein samples obtained from treated cells were analyzed for FOXM1 protein levels via immunoblotting, β-actin was used as internal loading control.

Fig. 2. STL inhibits FOXM1 expression on protein level via autophagy-dependent mechanism.

a C3-luc cells stimulated with doxycycline to induce expression of EGFP-FOXM1 fusion protein were treated with increasing concentrations of STL for 24 h. Total protein samples were analyzed via immunoblotting for FOXM1 and GFP expression, β-actin was used as an internal loading control. b Doxycycline-stimulated C3-luc cells were treated with 50 μM STL for 6 or 24 h and 10 μg/mL ActD for 6 h. Total RNA samples were analyzed for FOXM1, GFP, and MCL1 transcript levels via RT-qPCR, 18 S rRNA was used as a reference transcript. Data are presented as means ± S.D. and individual datapoints, N = 4, * – exact p = 0.02857 (Mann–Whitney U test, two-tailed). c C3-luc cells were treated with indicated concentrations of doxycycline, STL, or CHX for 24 h. Total protein samples were analyzed via immunoblotting for FOXM1, NPM, and MCL1 expression, β-actin was used as an internal loading control. d Doxycycline-stimulated C3-luc cells were treated with STL for 24 h in the presence of bortezomib or MG132. Total protein samples were analyzed via immunoblotting for FOXM1 expression, β-actin was used as an internal loading control. e Doxycycline-stimulated C3-luc cells were treated with STL for 24 h in the presence of bafilomycin A1. Total protein samples were analyzed via immunoblotting for FOXM1 expression, β-actin was used as an internal loading control.

Fig. 3. STL-dependent FOXM1 degradation is autophagy-mediated and is prevented by nuclear-export arrest.

a C3-luc cells were treated with indicated concentrations of doxycycline and STL for 24 h. Total protein samples were analyzed via immunoblotting for FOXM1, LAMP1, and LC3 expression, β-actin was used as an internal loading control. b C3-luc cells were treated with indicated concentrations of doxycycline, STL, bafilomycin A1, CQ, and LMB for 24 h. Total protein samples were analyzed via immunoblotting for FOXM1 and LC3 expression, β-actin was used as an internal loading control.

Fig. 4. STL promotes active lysosome formation and FOXM1 translocation from the nucleus to the cytoplasmic autophagosomes.

a Doxycycline-stimulated C3-luc cells expressing EGFP-FOXM1 fusion protein were treated with vehicle (“Control”, panels 1–3), 50 μM STL alone (“STL”, panels 4–6), or 50 μM STL in combinations with 25 nM BafA1 (“STL + BafA1”, panels 7–9), 40 μM CQ (“STL + CQ”, panels 10–12), and 25 nM LMB (“STL + LMB”, panels 13–15) for 12 h. Lysosomes were stained with vital LysoView 540 dye (red), cell morphology was analyzed using differential interference contrast (DIC) microscopy. b Doxycycline-stimulated C3-luc cells were treated with vehicle (“Control”, panels 1–4), 50 μM STL alone (“STL”, panels 5–8), or 50 μM STL in combinations with 40 μM CQ (“STL + CQ”, panels 9–12) and 25 nM LMB (“STL + LMB”, panels 13–16) for 24 h. Cells were stained for LC3 protein, nuclei were counterstained with DAPI. EGFP-FOXM1 (green), LC3 (red) and DAPI (blue) fluorescence was analyzed using confocal laser-scanning microscopy.

Fig. 5. STL enhances the cytotoxic effect of conventional chemotherapeutic drugs through suppression of FOXM1.

a NSCLC and HGSOC cell lines were treated with indicated concentrations of carboplatin and STL alone or in combination for 48 h. b PEO1 cells with stable FOXM1 knockdown were treated with indicated concentrations of carboplatin and STL alone or in combination for 48 h. c Colorectal cancer cell lines were treated with indicated concentrations of 5-FU and STL alone or in combination for 24 h. d Prostate cancer and NSCLC cell lines were treated with indicated concentrations of docetaxel, paclitaxel, and STL alone or in combination for 24 h. In all cases, total protein samples were obtained from cells immediately after treatment and analyzed for FOXM1, cleaved caspase-3 and LC3 levels via immunoblotting, β-actin was used as an internal loading control.

Fig. 6. STL-induced transcriptome changes suggest strong antitumor effect mediated through FOXM1-p21–p53 regulatory networks.

a Heatmap representation of the STL effect on global gene expression changes. HCT116 and doxycycline-stimulated C3-luc cells were treated with 50 μM STL for 24 h. RNA samples obtained from treated cells were subjected to RNA-seq, non-protein-coding genes were excluded from analysis. Data represent the average of two biological replicates for each condition. b Venn diagrams representing the numbers of DE genes (all, up- or downregulated) in HCT116 and C3-luc cells. Only genes with significant expression changes (2-fold or higher change, FDR < 0.1) were analyzed. c Ingenuity-predicted regulatory network facilitating STL treatment effects in C3-luc cells based on “STL signature” changes. d Results of GSEA performed for “STL signature” genes in C3-luc cells.