Characterization of Potent ABCG2 Inhibitor Derived from Chromone: From the Mechanism of Inhibition to Human Extracellular Vesicles for Drug Delivery

Abstract

Inhibition of ABC transporters is a promising approach to overcome multidrug resistance in cancer. Herein, we report the characterization of a potent ABCG2 inhibitor, namely, chromone 4a (C4a). Molecular docking and in vitro assays using ABCG2 and P-glycoprotein (P-gp) expressing membrane vesicles of insect cells revealed that C4a interacts with both transporters, while showing selectivity toward ABCG2 using cell-based transport assays. C4a inhibited the ABCG2-mediated efflux of different substrates and molecular dynamic simulations demonstrated that C4a binds in the Ko143-binding pocket. Liposomes and extracellular vesicles (EVs) of Giardia intestinalis and human blood were used to successfully bypass the poor water solubility and delivery of C4a as assessed by inhibition of the ABCG2 function. Human blood EVs also promoted delivery of the well-known P-gp inhibitor, elacridar. Here, for the first time, we demonstrated the potential use of plasma circulating EVs for drug delivery of hydrophobic drugs targeting membrane proteins.

Keywords: ABCG2 transporter; chromones; drug delivery; extracellular vesicles; inhibitors.

Figure 1

Chemical structure of acridone MBLI-87, chromones MBL-II-141 and chromone 4a (C4a).

Figure 2

In silico interaction of C4a with ABCG2 and P-gp. Molecular docking analysis in human ABCG2 (PDB: 6HCO) (A–C) and P-gp (PDB: 6QEX) (D–F). Amino acid residues are colored according to the atom involved in the interaction (carbon, light grey (chain A of ABCG2 and P-gp) and pale blue (chain B of ABCG2); oxygen, red; nitrogen, blue). (A,D) display an overview of the transporter structure, highlighting the position of the docking pose box as an orange square. The original conformation of substrates estrone-3-sulfate (E3S) for ABCG2 display small conformational changes when compared to the highest scored docking pose from molecular docking (B), while redocking of taxol within P-gp is much more expressive (E). Best scored pose of the interaction of C4a with ABCG2 (C) and P-gp (F).

Figure 3

Membrane-based studies of the interaction of C4a with ABCG2 and P-gp. (A) Effect of C4a on ATPase activity using High-Five ABCG2 and P-gp total membrane vesicles. EC₅₀ value is the concentration giving a half-maximal effect. (B) Thermostabilization assay with C4a at saturating concentration of 10 μM on the ATPase activity of ABCG2 and P-gp. The data are the mean ± SD of three independent experiments performed in duplicate.

Figure 4

Cell-based studies of the interaction of C4a with ABCG2. Transport assay using stable transfected cell lines overexpressing ABCG2, P-gp and MRP1. For ABCG2 transporter, mitoxantrone (10 µM) and Ko143 (1 µM) were used as substrate and reference inhibitor, respectively. For P-gp transporter, rhodamine 123 (10 µM) and elacridar (1 µM) were used as substrate and reference inhibitor, respectively. For MRP1 transporter, calcein-AM (0.2 µM) and verapamil (35 µM) were used as substrate and reference inhibitor, respectively. C4a was used at 10 µM. (A) Representative histograms obtained by flow cytometry. (B) Data of mean ± SD of three independent experiments performed in duplicate. (C) Cell-based ABCG2 inhibition transport assay by flow cytometry using different substrates: bodipy-prazosin (0.2 µM), mitoxantrone (10 µM) and pheophorbide a (10 µM). (D) Cell-based ABCG2 inhibition transport assay by confocal microscopy using hoechst 33342 (2 µM) as substrate. (E) 5D3 shift assay by flow cytometry using C4a (10 µM) and Ko143 (2 µM).

Figure 5

Molecular dynamic simulations: (A) Representative snapshots of the last frames for molecular dynamics simulations highlight the superposition of all poses. (B) SN38 substrate—red, (C) Ko143 inhibitor—orange and (D) C4a-blue. ABCG2 amino acid residues are colored according to the atom involved in the interaction (carbon, light grey (chain A) and pale blue (chain B); oxygen, red; nitrogen, blue). Hydrogen bond interactions and water bridges are represented by dashed yellow lines. (E,F) Two-dimensional representation of the interaction frequency for SN38 (E) and C4a (F) observed along with the molecular dynamics simulations. Each number represents the mean of at least five independent simulations (5 × 200 ns).

Figure 6

In vitro studies of ABCG2 inhibition using Lip-C4a (C4a loaded into liposomes composed of DSPC/DOPE). Cell-based transport assay using stably transfected cell lines overexpressing ABCG2. Hoechst 33342 (3 µM) and Ko143 (1 µM) were used as substrate and reference inhibitor, respectively. C4a or Lip-C4a were used at 10 µM. (A) Representative histograms were obtained by flow cytometry (30 min of incubation with inhibitor-upper part; 24 h of incubation with inhibitor-bottom part). (B) Data of mean ± SD of three independent experiments.

Figure 7

In vitro studies of ABCG2 inhibition using EVs G. intestinalis-C4a (C4a loaded into EVs from G. intestinalis). Cell-based transport assay using stably transfected cell lines overexpressing ABCG2. Hoechst 33342 (3 µM) and Ko143 (1 µM) were used as substrate and reference inhibitor, respectively. (A) C4a free in the supernatant or EVs G. intestinalis-C4a were tested at 0.1, 1 and 10 µM. Data of mean ± SD of three independent experiments. (B) Cell viability assay on wild-type cells (HEK293-WT) and HEK293-ABCG2 cells. SN38 (10 nM) was used as chemotherapeutic drug transported by ABCG2. Ko143 (1 µM) was used as a reference inhibitor. Co-treatment of SN38 with either C4a free in the supernatant or EVs G. intestinalis-C4a were tested at 0.1, 1 and 10 µM. Data of mean ± SD of three independent experiments.

Figure 8

In vitro studies of ABCG2 and P-gp inhibition using EVs human-C4a (C4a loaded into EVs from blood human cells and plasma). Cell-based transport assay using stably transfected cell lines overexpressing ABCG2 and P-gp. Data of mean ± SD of three independent experiments. For ABCG2, hoechst 33342 (5 µM) and C4a (1 µM) were used as substrate and reference inhibitor, respectively. (A) C4a free in the supernatant or EVs human blood cells-C4a were tested at 0.1, 1 and 10 µM. (B) C4a free in the supernatant or EVs human plasma-C4a were tested at 0.1, 1 and 10 µM. For P-gp, rhodamine 123 (10 µM) and elacridar (1 µM) were used as substrate and reference inhibitor, respectively. (C) Elacridar free in the supernatant or EVs human blood cells–elacridar were tested at 0.1, 1 and 10 µM. (D) Elacridar free in the supernatant or EVs human plasma–elacridar were tested at 0.1, 1 and 10 µM.