Role of Quinone Reductase 2 in the Antimalarial Properties of Indolone-Type Derivatives

Abstract

Indolone-N-oxides have antiplasmodial properties against Plasmodium falciparum at the erythrocytic stage, with IC50 values in the nanomolar range. The mechanism of action of indolone derivatives involves the production of free radicals, which follows their bioreduction by an unknown mechanism. In this study, we hypothesized that human quinone reductase 2 (hQR2), known to act as a flavin redox switch upon binding to the broadly used antimalarial chloroquine, could be involved in the activity of the redox-active indolone derivatives. Therefore, we investigated the role of hQR2 in the reduction of indolone derivatives. We analyzed the interaction between hQR2 and several indolone-type derivatives by examining enzymatic kinetics, the substrate/protein complex structure with X-ray diffraction analysis, and the production of free radicals with electron paramagnetic resonance. The reduction of each compound in cells overexpressing hQR2 was compared to its reduction in naïve cells. This process could be inhibited by the specific hQR2 inhibitor, S29434. These results confirmed that the anti-malarial activity of indolone-type derivatives was linked to their ability to serve as hQR2 substrates and not as hQR2 inhibitors as reported for chloroquine, leading to the possibility that substrate of hQR2 could be considered as a new avenue for the design of new antimalarial compounds.

Keywords: human quinone reductase 2; indolones; inhibitor; malaria; mechanism.

Figure 2

Structural illustrations of human quinone reductase 2 (hQR2) co-crystallized with compounds 8′ and 10. (A,B) Crystal structure of the flavin adenine dinucleotide (FAD) complexed with hQR2, and bound to (A) compound 10 or (B) compound 8′ (cartoon representations were created with the Pymol program); (C,D) Detailed view of the ligands in the binding pocket (FAD, red sticks); (C) compound 10, grey sticks; (D) compound 8′, yellow sticks. Grey mesh represents the electronic density (2fo-fc map) surrounding FAD, compound 8′, and compound 10 (1.5σ).

Figure 3

Free radicals produced by hQR2 metabolization of indolone derivatives. (A) Examples of EPR spectra recorded with the different compounds tested. Actual spectra (dark) were compared with theoretical spectra (red); (B) comparison of the compound potency (IC₅₀ in nM, blue bars P. falciparum, strain FcB1) to the amount of radicals produced in a pure hQR2 system (orange bars).

Figure 4

Free radicals produced by metabolization of indolone derivatives in red blood cells. (A) EPR spectra recorded (left) after adding 7 to red blood cells, compared to the simulated spectrum (right, in red); (B) Comparison of reactive oxygen species (ROS) production for compounds selected from the two sets of active series. Cells (5 × 10⁶) in DPBS buffer were treated with an indolone derivative substrate (100 µM), N-benzyldihydronicotinamide BNAH (100 µM), and DMPO (50 mM) (6% DMSO in the final mix). ROS production was evaluated by double integration of the area under the peaks. In two cases, cells were pre-incubated with S29434 (20 µM) for 30 min before adding compounds 1 and 12; (C) Comparison of the expression of hQR2 expression in healthy versus parasitized red blood cells by western blot.

Figure 5

Free radicals produced by metabolization of indolone derivatives in CHO cells. Radical production was measured after two different CHO cell lines were treated (2 min incubation) with indolone derivatives selected from the two active series. One cell line overexpressed hQR2 (CHO-QR2, orange bars) and the other was not transfected (CHO-NT, blue bars), and therefore, expressed only endogenous QR2 levels. Cells (5 × 10⁶) were treated with the indicated indolone substrate (100 µM), N-benzyldihydronicotinamide (BNAH) (100 µM), and DMPO (50 mM). Radical production was evaluated as the double integration of the area under the EPR peaks. When indicated, cells were pre-incubated for 30 min with S29434 (20 µM).

Figure 6

Rates of ferricyanide reduction reflect the metabolization of indolone derivatives in CHO cells. Ferricyanide reduction rates were measured (mean % reduction ± SD) in two cell lines treated for 60 min with potassium ferricyanide (600 µM), BNAH (100 µM), and the indicated indolone substrates (100 µM). One cell line overexpressed hQR2 (CHO-QR2, orange bars) and the other was not transfected (CHO-NT, blue bars), and therefore, expressed only endogenous QR2 levels. The % reduction was evaluated by comparing absorbance in treated cells to the absorbance measured in the absence of substrates. Cells were pre-incubated for 24 h with the QR2 inhibitor S29434 (20 µM): CHO-QR2 + S29434, yellow bars; CHO-NT + S29434, grey bars.