Identification of Novel Anthracycline Resistance Genes and Their Inhibitors

Abstract

Differentially expressed genes have been previously identified by us in multidrug-resistant tumor cells mainly resistant to doxorubicin. In the present study, we exemplarily focused on some of these genes to investigate their causative relationship with drug resistance. HMOX1, NEIL2, and PRKCA were overexpressed by lentiviral-plasmid-based transfection of HEK293 cells. An in silico drug repurposing approach was applied using virtual screening and molecular docking of FDA-approved drugs to identify inhibitors of these new drug-resistant genes. Overexpression of the selected genes conferred resistance to doxorubicin and daunorubicin but not to vincristine, docetaxel, and cisplatin, indicating the involvement of these genes in resistance to anthracyclines but not to a broader MDR phenotype. Using virtual drug screening and molecular docking analyses, we identified FDA-approved compounds (conivaptan, bexarotene, and desloratadine) that were interacting with HMOX1 and PRKCA at even stronger binding affinities than 1-(adamantan-1-yl)-2-(1H-imidazol-1-yl)ethenone and ellagic acid as known inhibitors of HMOX1 and PRKCA, respectively. Conivaptan treatment increased doxorubicin sensitivity of both HMOX1- and PRKCA-transfected cell lines. Bexarotene treatment had a comparable doxorubicin-sensitizing effect in HMOX1-transfected cells and desloratadine in PRKCA-transfected cells. Novel drug resistance mechanisms independent of ABC transporters have been identified that contribute to anthracycline resistance in MDR cells.

Keywords: RNA sequencing; cancer; chemotherapy; drug resistance; transfection.

Figure 1

Effect of MDR-related mechanisms on cancer progression.

Figure 2

Transfection of HEK293 cells with HMOX1, NEIL2, and PRKCA plasmid constructs. (A) Verification of the clones after EcoRI digestion. (B) Verification of transfection with GFP signal observation under a fluorescent microscope. (C) Verification of protein overexpression for HMOX1, NEIL2, and PRKCA. β-actin was used as a loading control. NC, non-transfected control.

Figure 3

Effect of HMOX1, NEIL2, and PRKCA overexpression on doxorubicin and daunorubicin resistance. Dose response curves for (A) doxorubicin and (B) daunorubicinIC₅₀ values for (C) doxorubicin and (D) daunorubicin treatment.

Figure 4

Similarity-based analysis of (A) top 10 FDA-approved drugs after HMOX1 screening and (B) top 10 FDA-approved drugs after PRKCA screening.

Figure 5

Molecular docking poses of conivaptan (blue) and bexarotene (green) on (A) HMOX1 and that of conivaptan (blue) and desloratadine (green) on (B) PRKCA. The known HMOX1 inhibitor 1-(adamantan-1-yl)-2-(1H-imidazol-1-yl)ethenone and the known PRKCA inhibitor ellagic acid are displayed in red. Amino acid residues forming hydrogen bonds are displayed in bold.

Figure 5

Molecular docking poses of conivaptan (blue) and bexarotene (green) on (A) HMOX1 and that of conivaptan (blue) and desloratadine (green) on (B) PRKCA. The known HMOX1 inhibitor 1-(adamantan-1-yl)-2-(1H-imidazol-1-yl)ethenone and the known PRKCA inhibitor ellagic acid are displayed in red. Amino acid residues forming hydrogen bonds are displayed in bold.

Figure 6

Effect of candidate inhibitors of HMOX1 or PRKCA on doxorubicin cytotoxicity in HMOX1- or PRKCA-transfected cell lines (cell viability, % of control: y-axis; concentration in µM: x-axis). (A) Conivaptan alone, (B) conivaptan with or without doxorubicin in HMOX1-transfected cells, (C) conivaptan with or without doxorubicin in PRKCA-transfected cells, (D) bexarotene alone, (E) bexarotene with or without doxorubicin in HMOX1-transfected cells, (F) desloratadine alone, and (G) desloratadine with or without doxorubicin in PRKCA-transfected cells.