Structural and molecular characterization of lopinavir and ivermectin as breast cancer resistance protein (BCRP/ABCG2) inhibitors

Abstract

A current clinical challenge in cancer is multidrug resistance (MDR) mediated by ABC transporters. Breast cancer resistance protein (BCRP) or ABCG2 transporter is one of the most important ABC transporters implicated in MDR and the use of inhibitors is a promising approach to overcome the resistance in cancer. This study aimed to characterize the molecular mechanism of ABCG2 inhibitors identified by a repurposing drug strategy using antiviral, anti-inflammatory and antiparasitic agents. Lopinavir and ivermectin can be considered as pan-inhibitors of ABC transporters, since both compounds inhibited ABCG2, P-glycoprotein and MRP1. They inhibited ABCG2 activity showing IC₅₀ values of 25.5 and 23.4 µM, respectively. These drugs were highly cytotoxic and not transported by ABCG2. Additionally, these drugs increased the 5D3 antibody binding and did not affect the mRNA and protein expression levels. Cell-based analysis of the type of inhibition suggested a non-competitive inhibition, which was further corroborated by in silico approaches of molecular docking and molecular dynamics simulations. These results showed an overlap of the lopinavir and ivermectin binding sites on ABCG2, mainly interacting with E446 residue. However, the substrate mitoxantrone occupies a different site, binding to the F436 region, closer to the L554/L555 plug. In conclusion, these results revealed the mechanistic basis of lopinavir and ivermectin interaction with ABCG2. See also the Graphical abstract(Fig. 1).

Keywords: ABCG2; ABCG2 inhibitors; drug repositioning; ivermectin; lopinavir; molecular modelling.

Table 1. Nucleotide sequences of primers used for RT-qPCR

Figure 1. Graphical abstract

Figure 2. ABCG2 inhibition by flow cytometry. (A) Screening of drugs: hydrocortisone, prednisolone, dexamethasone, ivermectin, lopinavir, hydroxychloroquine, chloroquine and oseltamivir as ABCG2 inhibitors. Drugs were tested at 10 and 100 µM on HEK293-ABCG2 cells by flow cytometry using Hoechst 33342 as substrate at 3 µM. Ko143 at 1 µM was used as reference inhibitor (100 % of inhibition). Data represent the mean ± SD of at least three independent experiments. (B) Representative histograms of different conditions, including control (Ctrl) with Hoechst 33342 (3 µM) alone and together with lopinavir (100 µM), ivermectin (100 µM) and Ko143 (1 µM). (C) Lopinavir and (D) ivermectin IC₅₀ curves of ABCG2 inhibition. Data represent the mean ± SD of at least three independent experiments.

Figure 3. Cytotoxicity and absence of transport mediated by ABCG2 transporter. MTT cell viability assay was performed on HEK293 cells (WT) in red and HEK293-ABCG2 cells (ABCG2) in blue after 72 hours of treatment with (A) hydrocortisone, (B) prednisolone, (C) dexamethasone, (D) lopinavir, (E) vermectin, (F) hydroxychloroquine, (G) chloroquine, (H) oseltamivir, and (I) SN38. Drugs were tested at different concentrations, as indicated in the graphs and the data represent the mean ± SD of at least three independent experiments. Cells treated with the vehicle (DMSO or H₂O) were considered 100 % of viable cells.

Figure 4. ABCG2 inhibition by flow cytometry. (A) Conformational 5D3 antibody binding. The data was normalized by the untreated control. *Significant difference (p > 0.05) according to Kruskal-Wallis test comparing the different groups (Ctrl, Ko143, LPV and IVT). (B) Representative histograms of “5D3 shift” assay: Ko143 (1 µM), lopinavir (100 µM) and ivermectin (100 µM) conditions. (C) mRNA expression levels quantified by qPCR. (D) Protein expression levels quantified by western blot. (E) Representative image of western blot assay.

Figure 5. Kinetic behavior of mitoxantrone ABCG2-mediated efflux by flow cytometry. Intracellular fluorescence was determined using a range of mitoxantrone/inhibitors concentrations. (A) Lopinavir (LPV). (B) Ivermectin (IVT). Comparison of V_max and K_M of (C) lopinavir and (D) ivermectin. Data represent the mean ± SD of at least three independent experiments.

Figure 6. Molecular docking and molecular dynamics simulations. (A) The violin plot shows calculated binding affinity (kcal/mol) of MTX and MTX (+ inhibitor) complexes. (B) Bar chart, with interaction frequency (fraction of simulation time) between important residues and each molecule. (C and D) Sticks representation of three most prevalent populations of IVT (shades of orange), LPV (shades of yellow) and MTX (shades of blue). Residues with frequent interactions are indicated with arrows. Oxygen and nitrogen atoms are colored red and blue, respectively, and non-polar hydrogens are omitted. In the center: Surface representation of 6VXI and ligands in State 1, indicating the DBC, delimited by the leucine 554/555 plug (colored in pale green). MTX is mitoxantrone; LPV is lopinavir and IVT is ivermectin.