The interaction of bortezomib with multidrug transporters: implications for therapeutic applications in advanced multiple myeloma and other neoplasias

Abstract

Purpose: Bortezomib is an important agent in multiple myeloma treatment, but resistance in cell lines and patients has been described. The main mechanisms of resistance described in cancer fall into one of two categories, pharmacokinetic resistance (PK), e.g. over expression of drug efflux pumps and pharmacodynamic resistance, e.g. apoptosis resistance or altered survival pathways, where the agent reaches an appropriate concentration, but this fails to propagate an appropriate cell death response. Of the known pump mechanisms, P-glycoprotein (P-gp) is the best studied and considered to be the most important in contributing to general PK drug resistance. Resistance to bortezomib is multifactorial and there are conflicting indications that cellular overexpression of P-gp may contribute to resistance agent. Hence, better characterization of the interactions of this drug with classical resistance mechanisms should identify improved treatment applications.

Methods: Cell lines with different P-gp expression levels were used to determine the relationship between bortezomib and P-gp. Coculture system with stromal cells was used to determine the effect of the local microenvironment on the bortezomib-elacridar combination. To further assess P-gp function, intracellular accumulation of P-gp probe rhodamine-123 was utilised.

Results: In the present study, we show that bortezomib is a substrate for P-gp, but not for the other drug efflux transporters. Bortezomib activity is affected by P-gp expression and conversely, the expression of P-gp affect bortezomib's ability to act as a P-gp substrate. The local microenvironment did not alter the cellular response to bortezomib. We also demonstrate that bortezomib directly affects the expression and function of P-gp.

Conclusions: Our findings strongly support a role for P-gp in bortezomib resistance and, therefore, suggest that combination of a P-gp inhibitor and bortezomib in P-gp positive myeloma would be a reasonable treatment combination to extend efficacy of this important drug.

Fig. 1

Bortezomib acts as a P-glycoprotein substrate and a weak inhibitor. Bortezomib is also not a substrate or inhibitor of other drug efflux pumps. Using isogenic lung cancer cell lines, DLKP which overexpresses MRP-1; DLKP-A which overexpresses P-gp; and DLKP-SQ/Mitox which overexpresses BCRP, we were able to demonstrate that bortezomib was a substrate of P-gp by showing synergistic cytotoxicity in the presence of a P-gp inhibitor, elacridar in DLKP-A cells. a Bortezomib had no synergistic activity when combined with sulindac sulphide, a MRP-1 inhibitor in DLKP cells (b) and elacridar, a BCRP inhibitor in DLKP-SQ/Mitox (c). In DLKP-A cells, when tested in the presence of doxorubicin, a known P-gp substrate, bortezomib was shown to be an ineffective P-gp inhibitor as enhanced cytotoxicity was only demonstrated at the highest dose of bortezomib. All experiments were incubated at the doses indicated for 5 days. Cell survival assessed by acid phosphatase assay was expressed as percentage (mean ± SD) compared to vehicle treated control. ⁺Statistically not significant; *statistically significantly compared to control (P < 0.05)

Fig. 2

The effectiveness of bortezomib to act as a P-glycoprotein substrate is dependent on the expression level of the transporter. Using immunoblot analysis, we were able to determine the baseline P-gp expression level of different cell lines that are known to overexpress P-gp with the parental lines as a negative control (a). The densitometry readings are shown on top of the relevant cell lines. This image was representative of findings from three experiments. The cell lines expression level of P-gp corresponded to the degree of synergy seen when bortezomib was combined with elacridar. The cell lines that exhibit the highest levels of P-gp showed the most synergy when bortezomib is combined with elacridar for 5 days. DLKP-A (b) treated with Bortezomib (filled square without elacridar, open square with elacridar 0.125 μM). A549-taxol (c) treated with bortezomib (filled diamond without elacridar, open diamond with elacridar 0.4 μM). NCI/Adr-res (d) and RPMI-Dox40 (e) treated with bortezomib (filled square without elacridar, open square with elacridar 0.125 μM). All experiments were incubated at the doses indicated for 5 days. Cell survival was expressed as percentage (mean ± SD) compared to vehicle treated control

Fig. 3

Synergy between bortezomib and elacridar persists in the protective effect of the BM microenvironment. Using the compartment- specific bioluminescence imaging (CS-BLI) approach, we evaluated the combination of bortezomib and elacridar on MM RPMI-Dox40-MCherry/luc cells cultured in the presence versus absence of HS-5 stromal cells and treated with bortezomib 4 and 8 nM and elacridar 0.125, 0.25 μM for 5 days. RPMI-Dox40-MCherry/luc viability was expressed as percentage (mean ± SD) compared to vehicle treated control. Experiment was repeated in triplicate. Synergistic cytotoxicity of the combination of bortezomib and elacridar persisted despite coculture with BMSCs (^Δp < 0.05; *p < 0.001)

Fig. 4

Bortezomib is able to reduce the expression and function of P-glycoprotein. When RPMI-Dox40 (a) and DLKP-A (b) cells were treated with 4 and 16 nM bortezomib, respectively, immunoblot analysis demonstrated a reduction in the level of P-gp expression by 24 h. These images were representative of three experiments. GAPDH was employed as a loading control. Bortezomib treatment inhibited rhodamine-123 efflux in RPMI-Dox40 cells with maximum inhibition at 120 min for all doses tested (c). The experiment was performed in triplicate, and results were expressed relative to control at each time point (mean of the median intensity of rhodamine-123 fluorescence ± SEM)

Fig. 5

MM cells co-cultured with stroma shows reduced expression and function of P-glycoprotein. RPMI-Dox40 cells were co-cultured with BMSCs HS-5. Immunoblot analysis demonstrated a reduction in P-gp expression in the MM cells that were co-cultured with BMSCs (a). GAPDH was employed as a loading control. Inhibition of rhodamine-123 efflux was demonstrated in RPMI-Dox40 cells when co-cultured with BMSCs (b). The experiment was performed in triplicate, and results were expressed relative to absence of BMSCs control at each time point (mean of the median intensity of rhodamine-123 fluorescence ± SEM)

Fig. 6

Characterization of stromal cells co-cultured with MM cells. RPMI-Dox40 cells were co-cultured with BMSCs HS-5. Using flow cytometry, we demonstrated that the previously P-gp-negative stroma cells acquired P-gp expression after co-culturing with P-gp-positive RPMI-Dox40 cells; however, this P-gp-positive status was not functional. a RPMI-Dox40 cells prelabelled with CellVue claret were co-cultured with BMSCs HS-5 for 72 h. Analysing only the stroma cell population, there was an increase in the P-gp expression in the previously P-gp-negative stromal cells. Shown in black solid line BMSCs alone with a P-gp–PE antibody. The stroma fraction after co-culture with a IgG-PE (dotted line) and P-gp-PE (grey line) antibody. b RPMI-Dox40 cells prelabelled with CellVue claret were co-cultured with stroma for up to 72 h and Rh-123 assay was performed. The MM cell population was distinguished from the stromal cells, and only the stromal fraction was analysed. The P-gp-positive stromal fraction was not functional as evidenced by the lack of rhodamine 123 extrusion in the absence of verapamil. The experiment was performed in triplicate (average of the mean intensity of rhodamine-123 fluorescence ± SD)