Comprehensive Evaluation of Multiple Approaches Targeting ABCB1 to Resensitize Docetaxel-Resistant Prostate Cancer Cell Lines

Abstract

Docetaxel (DTX) is a mainstay in the treatment of metastatic prostate cancer. Failure of DTX therapy is often associated with multidrug resistance caused by overexpression of efflux membrane transporters of the ABC family such as the glycoprotein ABCB1. This study investigated multiple approaches targeting ABCB1 to resensitize DTX-resistant (DTXR) prostate cancer cell lines. In DU145 DTXR and PC-3 DTXR cells as well as age-matched parental controls, the expression of selected ABC transporters was analyzed by quantitative PCR, Western blot, flow cytometry and immunofluorescence. ABCB1 effluxing activity was studied using the fluorescent ABCB1 substrate rhodamine 123. The influence of ABCB1 inhibitors (elacridar, tariquidar), ABCB1-specific siRNA and inhibition of post-translational glycosylation on DTX tolerance was assessed by cell viability and colony formation assays. In DTXR cells, only ABCB1 was highly upregulated, which was accompanied by a strong effluxing activity and additional post-translational glycosylation of ABCB1. Pharmacological inhibition and siRNA-mediated knockdown of ABCB1 completely resensitized DTXR cells to DTX. Inhibition of glycosylation with tunicamycin affected DTX resistance partially in DU145 DTXR cells, which was accompanied by a slight intracellular accumulation and decreased effluxing activity of ABCB1. In conclusion, DTX resistance can be reversed by various strategies with small molecule inhibitors representing the most promising and feasible approach.

Keywords: ABCB1; P-glycoprotein; chemoresistance; docetaxel; elacridar; glycosylation; prostate cancer; siRNA; tariquidar; tunicamycin.

Figure 1

RNA and protein expression of selected ABC transporters, which are known to be involved in taxane resistance and/or MDR, in DU145 and PC-3 DTXR cells compared to CTRL cells. (a) The relative transcript levels of eight selected ABC transporters were plotted on a heat map. Cells with an expression of the respective ABC transporter beyond the detection threshold (CP ≥ 35) are marked with an asterisk (*). (b) Exemplary Western blots for detecting ABCB1 and ABCG2 proteins in CTRL and DTXR cells. GAPDH served as a reference protein. (c,d) Exemplary flow cytometry analysis for (c) ABCB1 and (d) ABCG2 proteins via specific antibodies conjugated with the fluorescent phycoerythrin (PE) in CTRL and DTXR cells. The vertical black lines indicate the threshold of unspecificity according to the pooled isotype controls. The adjacent graphs to the right depict the median fluorescence intensities (MFI) for ABCB1 and ABCG2 protein, respectively. Data are depicted as mean ± standard deviation (SD) of three to four independent biological replicates. Non-paired t-test with Welch’s correction was performed for comparison of CTRL vs. DTXR cells: *** p < 0.001, ns: not significant.

Figure 2

Influence of ABCB1 inhibitors elacridar and tariquidar on ABCB1 efflux activity of the fluorescent ABCB1 substrate Rh123 in DU145 and PC-3 DTXR cells compared to CTRL cells. (a) DU145 and (b) PC-3 cells were treated with 0.5 µg/mL Rh123 in the presence or absence of an ABCB1 inhibitor (50 nM elacridar or tariquidar) for 15 min. After several washing steps, RPMI 1640 medium with or without one of the inhibitors was added and subsequently the cells were imaged in 10 min intervals. As examples, the first image (t₀) and an image taken after 60 min are depicted for CTRL and DTXR cells (scale bar 50 µm). Longitudinal data were normalized to the corresponding treatment of CTRL cells at t₀ (indicated by ^a, ^b or ^c in the table; black line in the graph). Data are depicted as the mean of four to five independent biological replicates + SD in the graphs for better readability or ± SD in the tables. Non-paired t-test with Welch’s correction was performed for comparison of ^d CTRL + elacridar/tariquidar vs. CTRL cells and ^e DTXR + elacridar/tariquidar vs. DTXR cells at 60 min. Ela.: elacridar, Tari.: tariquidar.

Figure 3

Influence of the ABCB1 inhibitors elacridar and tariquidar in combination with DTX on cellular viability and clonogenic survival of DU145 and PC-3 DTXR cells compared to CTRL cells. (a) DU145 and PC-3 CTRL and DTXR cells were treated with a serial dilution of DTX in the presence or absence of an ABCB1 inhibitor (50 nM elacridar or tariquidar). After treatment, the metabolic activity was measured by the WST-1 assay and IC₅₀ values for DTX were calculated. The adjacent graphs to the right depict IC₅₀ values calculated from individual experiments. Ela.: elacridar, Tari.: tariquidar. (b) CTRL and DTXR cells were treated with either 50 nM elacridar or 10 nM DTX alone or with a combination of both. Thereafter, cell colony formation assays were seeded and cell colonies were counted after 10–12 d. Exemplary colonies are shown in the right panels. All data were normalized to the indicated control (black line). Data are depicted as mean ± SD of four to six independent biological replicates. (a) Non-paired t-test with Welch’s correction for comparison of two treatment groups and (b) one-sample t-test for comparison to the untreated control: ** p < 0.01, *** p < 0.001, ns: not significant.

Figure 4

Influence of siRNA-mediated knockdown of ABCB1 in combination with DTX on cellular viability and clonogenic survival of DU145 and PC-3 DTXR cells. (a) The relative transcript and protein levels of ABCB1 in DTXR cells 96 h after siRNA transfection start are depicted. An exemplary Western blot for the detection of ABCB1 protein with GAPDH as reference protein is shown. (b) DU145 and PC-3 DTXR cells were treated with a serial dilution of DTX after transfection with either siR-CON or siR-ABCB1 (20 nM). After treatment, the metabolic activity was measured by the WST-1 assay and IC₅₀ values for DTX were calculated. The adjacent graphs to the right depict IC₅₀ values calculated from individual experiments. (c) DTXR cells were transfected with either siR-CON or siR-ABCB1 (20 nM) and 96 h after transfection start the cells were treated with 50 nM DTX. Thereafter, cell colony formation assays were seeded and cell colonies were counted after 10–12 d. Exemplary colonies are shown in the right panels. All data were normalized to the indicated control (black line). Data are depicted as mean ± SD of three to five independent biological replicates. (a,c) One-sample t-test for comparison to the siR-CON-treated controls and (b,c) non-paired t-test with Welch’s correction for comparison of two treatment groups: * p < 0.05, *** p < 0.001, ns: not significant.

Figure 5

Influence of PNGase F digestion and tunicamycin treatment on the glycosylation status and localization of ABCB1 protein in DU145 and PC-3 DTXR cells. (a) Protein lysates of DU145 and PC-3 DTXR cells were treated with PNGase F for 2 h and compared to untreated protein lysates (UT). An exemplary Western blot for the detection of ABCB1 protein with GAPDH as reference protein is shown. (b) DU145 and PC-3 DTXR cells were pre-treated with 100 ng/mL tunicamycin (TUN) for 96 h and compared to untreated cells (UT). An exemplary Western blot for detecting ABCB1 protein with α-tubulin as reference protein is shown. (a,b) The protein bands were analyzed densitometrically and then normalized to the respective reference protein. The different glycosylation statuses are depicted as part of the total ABCB1 amount in %. (c) Immunofluorescence staining of ABCB1 in DU145 and PC-3 CTRL cells as well as DTXR cells with or without pre-treatment with 100 ng/mL tunicamycin (TUN) for 96 h. Cell nuclei were visualized with DAPI (blue), ABCB1 was detected using an anti-ABCB1 antibody and an Alexa Fluor 488-conjugated secondary antibody (green) and then the images were merged (scale bar 40 µm). Exemplary images and data, which are depicted as mean ± SD, are taken from four to five independent biological replicates.

Figure 6

Influence of the glycosylation status of ABCB1 protein on ABCB1 efflux activity of the fluorescent ABCB1 substrate Rh123 in DU145 and PC-3 DTXR cells. (a) DU145 and (b) PC-3 DTXR cells with or without pre-treatment with 100 ng/mL tunicamycin (TUN) for 96 h were incubated with 0.5 µg/mL Rh123 for 15 min. After several washing steps, RPMI 1640 medium was added and subsequently the cells were imaged in 10 min intervals. As examples, the first image (t₀) and images taken after 10 and 20 min are depicted for DTXR cells with or without tunicamycin pre-treatment (scale bar 50 µm). Longitudinal data were normalized to DTXR cells without tunicamycin pre-treatment at t₀ (black line in the graph). Data are depicted as mean ± SD of four to five independent biological replicates. Non-paired t-test with Welch’s correction was performed for comparison of DTXR cells with tunicamycin pre-treatment vs. DTXR cells without tunicamycin pre-treatment at each timepoint: * p < 0.05, ** p < 0.01, ns: not significant.

Figure 7

Influence of tunicamycin-mediated deglycosylation of ABCB1 protein in combination with DTX on cellular viability of DU145 and PC-3 DTXR cells. DU145 and PC-3 DTXR cells with or without pre-treatment with 100 ng/mL tunicamycin (TUN) for 96 h were treated with a serial dilution of DTX. After treatment, the metabolic activity and the total adherent cell mass were determined by (a) the WST-1 assay and (b) the crystal violet assay, respectively, and IC₅₀ values for DTX were calculated. The adjacent graphs to the right depict IC₅₀ values calculated from individual experiments. All data were normalized to the indicated control (black line). Data are depicted as mean ± SD of four to six independent biological replicates. Non-paired t-test with Welch’s correction for comparison of two treatment groups: * p < 0.05, ns: not significant.