Novel Approach to Overcome Osimertinib Resistance Using Bromodomain and Extra-Terminal Domain Inhibitors

Abstract

Osimertinib, a third-generation EGFR-tyrosine kinase inhibitor, is the first-line therapy for lung cancer harboring EGFR mutations. The mechanisms underlying osimertinib resistance are diverse, with approximately half remaining unknown. Epigenetic dysregulation is implicated in drug resistance; however, the mechanisms remain unclear. Therefore, we investigated epigenetic involvement in osimertinib resistance and its therapeutic potential. We established osimertinib-resistant cells and used an assay for transposase-accessible chromatin using sequencing to evaluate chromatin accessibility, finding significant changes post-resistance. Combining the assay for transposase-accessible chromatin and RNA sequencing data, we identified FGF1 as a resistance-related gene regulated by histone modifications. FGF1 induced osimertinib resistance, and its suppression attenuated resistance. Bromodomain and extra-terminal domain inhibitors combined with osimertinib overcame osimertinib resistance by reducing FGF1 expression. Increased FGF1 expression was observed in osimertinib-resistant clinical samples. This combination therapy was effective in cell lines and mouse xenograft models. These results suggest targeting histone modifications using bromodomain and extra-terminal domain inhibitors as a novel approach to overcoming osimertinib resistance.

Keywords: ATAC‐seq; Epigenomics; FGF1; lung neoplasms; osimertinib.

FIGURE 1

FGF1 is selected as a candidate gene for osimertinib resistance. (A) Left: Chemosensitivity assay of PC9 cells and PC9‐OR in the presence or absence of the indicated osimertinib concentrations for 72 h. Right: Bar graph with IC₅₀ values of osimertinib for the cells. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t‐test. (B) Cells were treated with the indicated concentrations of osimertinib for 3 h. Western blots for total and phosphorylated EGFR, Akt, ERK1/2, and β‐actin (loading control) proteins. The results are representative of three independent experiments. (C) Outline for identifying candidate genes by overlapping ATAC‐ and RNA‐seq. Fifty‐nine upregulated genes were identified as candidates via overlapping chromatin accessibility (PC9‐OR/PC9 > log₂FC = 1.0) and transcriptome analysis (PC9‐OR/PC9 > log₂FC = 1.2). Five pathways were extracted from 59 upregulated genes using enrichment analysis. Genes contributing to the enrichment of each pathway and p‐value are shown. The MAPK signaling pathway was selected as a cancer‐related pathway. (D) FGF1 mRNA expression in PC9 cells and PC9‐OR. Data represent the mean ± SD of three independent experiments: *p < 0.05, two‐tailed unpaired Student's t‐test. (E) Protein expression levels of FGF1, total and phosphorylated FGFR1, and β‐actin (loading control) in PC9 cells and PC9‐OR. The results are representative of three independent experiments. (F) Immunocytochemistry for FGF1 in PC9 cells and PC9‐OR. Left: Nuclei were stained with DAPI (blue). Middle panel: FGF1 was stained with Alexa Fluor 488 (green). Right: Merged image. Scale bar = 50 μm. (G) Protein expression levels of FGF1 in PC9 cells and PC9‐OR via immunocytochemistry. Scatter plots show the relative fluorescence intensity of FGF1 protein levels in each cell. Data represent the mean ± SD of 10 HPF: *p < 0.05, two‐tailed unpaired Student's t‐test. The results are representative of three independent experiments.

FIGURE 2

FGF1 induces osimertinib resistance, and treatment with FGFR inhibitors or suppression of FGF1 via siRNA attenuates resistance to osimertinib. (A) Left: Chemosensitivity assay of PC9 cells in the presence or absence of the indicated osimertinib concentrations and in the presence or absence of the recombinant FGF1 for 72 h. Right: Bar graph with IC₅₀ values of osimertinib for the cells. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t‐test. (B) Left: Dot plot of flow cytometry data in PC9 cells treated with osimertinib for 72 h in the presence or absence of the recombinant FGF1 for 72 h and stained with Annexin FITC and PI. Right: Bar graphs show the percentage of apoptotic cells. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t‐test. (C) State of the total and phosphorylated EGFR downstream pathway in PC9 cells treated with the indicated concentrations of osimertinib in the presence or absence of the recombinant FGF1. The results are representative of three independent experiments. (D) Left: Chemosensitivity assay of PC9‐OR treated with or without nintedanib, (E) ponatinib, or (F) pemigatinib in the presence or absence of indicated osimertinib concentrations for 72 h. Right: Bar graph shows the IC₅₀ values. Data represent the mean ± SD of three independent experiments: *p < 0.05, ANOVA and post hoc comparisons. (G) FGF1 mRNA expression in PC9‐OR transfected with two specific siRNAs (#1 and #2) and non‐specific control siRNA (NC). Data represent the mean ± SD of three independent experiments: *p < 0.05, ANOVA and post hoc comparisons. (H) Expression of FGF1 and β‐actin (loading control) proteins transfected with two specific siRNAs (#1 and #2) and non‐specific control siRNA (NC) in PC9‐OR. The results are representative of three independent experiments. (I) Left: Chemosensitivity assay of PC9‐OR transfected with siNC or siFGF1 (#1 and #2) in the presence or absence of the indicated osimertinib concentrations for 72 h. Right: Bar graph shows the IC₅₀ values of osimertinib for the cells. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA and post hoc comparisons. (J) Percentage of apoptotic cells determined using Annexin V staining in PC9‐OR treated with osimertinib following siRNA‐mediated knockdown (NC, #1 and #2) for 72 h. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA and post hoc comparisons.

FIGURE 3

BET inhibitors attenuate resistance to osimertinib. (A) FGF1 mRNA expression in PC9‐OR treated with or without JQ1, I‐BET151, or RVX‐273 for 72 h. Data represent the mean ± SD of three independent experiments: *p < 0.05, ANOVA, and post hoc comparisons. (B) FGF1 and β‐actin (loading control) protein expression. The results are representative of three independent experiments. (C) Left: Chemosensitivity assay of PC9‐OR treated with JQ1, I‐BET151, or RVX273 in the presence or absence of the indicated osimertinib concentrations for 72 h. Right: Bar graph with IC₅₀ values of osimertinib. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA, and post hoc comparisons. (D) Percentage of apoptotic cells, determined using Annexin V staining, in PC9‐OR treated with or without JQ1, I‐BET151, or RVX273 in the presence of osimertinib for 72 h. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA, and post hoc comparisons. (E) State of the total and phosphorylated EGFR downstream pathway in PC9‐OR treated with osimertinib for 3 h in the presence or absence of BET inhibitors (JQ1, I‐BET151, and RVX‐273). The results are representative of three independent experiments. (F) Left: Chemosensitivity assay of PC9‐OR treated with JQ1 ± recombinant FGF1, (G) I‐BET151 ± recombinant FGF1, or (H) RVX273 ± recombinant FGF1 in the presence or absence of the indicated osimertinib concentrations for 72 h. Right: The bar graph shows IC₅₀ values of osimertinib. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t‐test. (I) Percentage of apoptotic cells, determined using Annexin V staining, in PC9‐OR treated with the JQ1 ± recombinant FGF1, (J) I‐BET151 ± recombinant FGF1, or (K) RVX273 ± recombinant FGF1 in the presence of osimertinib for 72 h. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t‐test.

FIGURE 4

Osimertinib resistance with elevated FGF1 expression in human specimens. (A) FGF1 mRNA expression in JFCR338 treated with or without JQ1, I‐BET151, or RVX‐273 for 72 h. Data represent the mean ± SD of three independent experiments: *p < 0.05, ANOVA, and post hoc comparisons. (B) FGF1 and β‐actin (loading control) protein expression. The results are representative of three independent experiments. (C) FGF1 protein expression in JFCR338 via immunocytochemistry. Scatter plots show the relative fluorescence intensity of the protein levels of FGF1 in each cell. Data represent the mean ± SD of 10 HPF: *p < 0.05, ANOVA, and post hoc comparisons. The results are representative of three independent experiments. (D) Left: Chemosensitivity assay of JFCR338 treated with JQ1, I‐BET151, or RVX273 in the presence or absence of the indicated osimertinib concentrations for 72 h. Right: Bar graph shows IC₅₀ values of osimertinib. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA, and post hoc comparisons. (E) Percentage of apoptotic cells determined using Annexin V staining in JFCR338 treated with or without JQ1, I‐BET151, or RVX273 in the presence of osimertinib for 72 h. Data from three independent experiments are presented as the mean ± SD, *p < 0.05, ANOVA, and post hoc comparisons. (F) State of the total and phosphorylated EGFR downstream pathway in JFCR338 treated with osimertinib for 3 h in the presence or absence of BET inhibitors (JQ1, I‐BET151, and RVX‐273). The results are representative of three independent experiments. (G–I) Immunohistochemistry (IHC) staining of FGF1 in cases 1, 2, and 3 (200× magnification). The upper figures show IHC in specimens before osimertinib treatment. The lower figures show IHC in specimens after osimertinib treatment. Scale bar = 50 μm. The results are representative of three independent experiments.

FIGURE 5

Combining the BET inhibitor with osimertinib overcomes osimertinib resistance in EGFR‐positive NSCLC mouse xenograft models. (A) Tumors of the four groups after treatment (1–4: Control, 5–8: JQ1 (50 mg/kg), 9–12: Osimertinib (1 mg/kg), 13–15: Combination of JQ1 and osimertinib). (B, C) The line graph represents the ratio of tumor volume increase and body weight gain from the start of treatment for each group. Data are presented as the mean ± SD, *p < 0.05, two‐tailed unpaired Student's t test.