Dofequidar fumarate sensitizes cancer stem-like side population cells to chemotherapeutic drugs by inhibiting ABCG2/BCRP-mediated drug export

Abstract

The ATP-binding cassette (ABC) transporters (ABC-T) actively efflux structurally and mechanistically unrelated anticancer drugs from cells. As a consequence, they can confer multidrug resistance (MDR) to cancer cells. ABC-T are also reported to be phenotypic markers and functional regulators of cancer stem/initiating cells (CSC) and believed to be associated with tumor initiation, progression, and relapse. Dofequidar fumarate, an orally active quinoline compound, has been reported to overcome MDR by inhibiting ABCB1/P-gp, ABCC1/MDR-associated protein 1, or both. Phase III clinical trials suggested that dofequidar had efficacy in patients who had not received prior therapy. Here we show that dofequidar inhibits the efflux of chemotherapeutic drugs and increases the sensitivity to anticancer drugs in CSC-like side population (SP) cells isolated from various cancer cell lines. Dofequidar treatment greatly reduced the cell number in the SP fraction. Estimation of ABC-T expression revealed that ABCG2/breast cancer resistance protein (BCRP) mRNA level, but not the ABCB1/P-gp or ABCC1/MDR-associated protein 1 mRNA level, in all the tested SP cells was higher than that in non-SP cells. The in vitro vesicle transporter assay clarified that dofequidar had the ability to suppress ABCG2/BCRP function. Dofequidar treatment sensitized SP cells to anticancer agents in vitro. We compared the antitumor efficacy of irinotecan (CPT-11) alone with that of CPT-11 plus dofequidar in xenografted SP cells. Although xenografted SP tumors showed resistance to CPT-11, treatment with CPT-11 plus dofequidar greatly reduced the SP-derived tumor growth in vivo. Our results suggest the possibility of selective eradication of CSC by inhibiting ABCG2/BCRP.

Figure 1

The presence of side population (SP) cells in various cancer cell lines. (A) Cells were stained with Hoechst33342 in the presence (+reserpine) or absence (control) of reserpine, and were analyzed using FACS Vantage. The trapezia in each panel indicate the SP area. (B) Summary of the presence of SP cells in various human cancer cell lines. Cells were stained and the presence of SP cells examined, as described in (A). +, Presence of SP cells; −, absence of SP cells.

Figure 2

Characteristics of side population (SP) cells. (A) SP and non‐SP (NSP) fractions were sorted from HeLa cells. The sorting purities were confirmed by immediate reanalysis. After 30‐day in vitro culture of the separated cells, cells were restained and analyzed (re‐stained). (B) Two thousand sorted HeLa SP and NSP cells were incubated for 72 h, and the viable cell number was assessed by 3‐(4,5‐dimethylthiazol‐2‐yl)‐5‐(3‐carboxymethoxyphenyl)‐2‐(4‐sulfophenyl)‐2H‐tetrazolium (MTS) assay. No significant difference was observed between SP and NSP (NS, P = 0.21). (C) The indicated numbers of sorted HeLa SP and NSP cells were s.c. injected into BALB/c nude mice (n = 6 or 5). Tumor formation was assessed after 60 days (left). The growth of xenografts derived from one hundred sorted HeLa SP and NSP cells (n = 6, *P < 0.01) (right). (D) The xenografted HeLa‐derived SP and NSP tumors were cut into 2‐mm cubes and transplanted into other mice. When secondary tumors reached approximately 100 mm³ in volume, CPT‐11 was administrated intravenously at 67 mg/kg on days 0, 4, and 8 (arrows). The graphs show the relative tumor volume (n = 6). (^† P < 0.05, *P < 0.001).

Figure 3

Dofequidar reduced the cell number in the side population (SP) fraction. (A) Chemical structure of dofequidar. (B) HeLa cells were stained with 5 μg/mL Hoechst33342 in the presence of the indicated concentrations of dofequidar, and then analyzed. (C) Cells were stained and analyzed as in (B). The relative ratios of SP cell numbers in dofequidar‐treated samples is shown by comparing with them with SP cell numbers in dofequidar‐untreated samples.

Figure 4

Efflux of Hoechst33342 by ATP‐binding cassette (ABC) G2/ABCG2/breast cancer resistance protein (BCRP), ABCB1/P‐gp/multidrug resistance (MDR) 1, and ABCC1/MDR‐associated protein (MRP) 1. (A) KB‐3‐1 and the stable transfectants were stained with Hoechst33342 and analyzed using FACS Vantage (left). The cell lysates from KB‐3‐1 and the stable transfectants were immunoblotted with the indicated antibodies (right). (B) mRNA was extracted from the sorted HeLa SP and NSP cells (solid and hatched columns, respectively). The graphs indicate relative mRNA expression levels of each ABC‐transporter normalized with GAPDH. The expression level of ABCG2 mRNA in side population (SP) cells versus that in non‐SP (NSP) cells was significantly different (P < 0.001). The expression levels of ABCB1 and ABCC1 mRNA were not significantly different (NS). (C) HeLa cells were stained with 5 μg/mL Hoechst33342 in the presence (+FTC) or absence (control) of 3 μM fumitremorgin C (FTC) and were analyzed. (D) HeLa cells were transfected with control siRNA or ABCG2 siRNA (si‐ABCG2‐1 and si‐ABCG2‐2). After transfection for 48 h, cells were stained with Hoechst33342 and analyzed (upper panels). The expression level of ABCG2/BCRP protein in the same samples were analyzed (lower panel).

Figure 5

Dofequidar inhibits ATP‐binding cassette (ABC) G2/ABCG2/breast cancer resistantance protein (BCRP)in vitro and in cells. (A) Membrane vesicles from ABCG2/BCRP‐overexpressing insect cells were incubated with ³H‐labeled methotrexate ([³H]MTX) and ATP together with vehicle (ATP only) or the indicated concentrations of dofequidar, fumitremorgin C (FTC), or verapamil at 37°C. In some experiments, membrane vesicles were incubated with [³H]MTX and AMP (AMP only). Membrane vesicles from control insect cells were also incubated with [³H]MTX and ATP or AMP (control vesicles). After incubation for 5 min, the incorporated [³H]MTX was assayed by liquid scintillation. (B) K562/BCRP cells were incubated with 3 μM mitoxantrone (MXR) together with vehicle or the indicated concentrations of dofequidar (MXR+Dofequidar, left panel) or FTC (MXR+FTC, right panel). In some experiments, K562/BCRP cells were incubated without drugs (control). After incubation for 30 min, the fluorescence of MXR at 670 nm was analyzed. (C) KB‐3‐1 and KB/BCRP cells were cultured in medium containing the indicated concentration of MXR with or without dofequidar or FTC for 3 days. Cell viability was evaluated using the 3‐(4,5‐dimethylthiazol‐2‐yl)‐5‐(3‐carboxymethoxyphenyl)‐2‐(4‐sulfophenyl)‐2H‐tetrazolium (MTS) method.

Figure 6

Sensitization by dofequidar of cancer stem‐like SP cells to chemotherapeutic drugs in cells and in vivo. (A) SP and NSP cells from HeLa, HBC‐5 or BSY‐1 cells were cultured in medium containing the indicated concentration of mitoxantrone (upper left panel) or topotecan (upper right panel). In some experiments, side population (SP) cells were cultured in the presence of 3 μM dofequidar (SP+Dofequidar). The lower panel gives a summary of 50% growth inhibition (GI₅₀) values in each experiment. (B) The xenografted HeLa‐derived SP and non‐SP (NSP) secondary tumors were treated with 200 mg/kg dofequidar, 67 mg/kg CPT‐11, or both on days 0, 4, and 8 (arrows) (n = 6). Graphs show relative tumor volume. The data of control and CPT‐11‐treated groups were the same as in Figure 2(D). The xenografted SP tumor size treated with CPT‐11 versus control (^†) or that treated with CPT‐11 plus dofequidar versus control (*) was significantly different (P < 0.05 and P < 0.01, respectively). (C) The mice bearing HeLa SP (left) or NSP (right) tumors were photographed at 14 days after first treatment. a, Dofequidar‐treated; b, CPT‐11‐treated; c, CPT‐11 plus dofequidar‐treated; d, control.