3-Arylidene-2-oxindoles as Potent NRH:Quinone Oxidoreductase 2 Inhibitors

The enzyme NRH:quinone oxidoreductase 2 (NQO2) plays an important role in the pathogenesis of various diseases such as neurodegenerative disorders, malaria, glaucoma, COVID-19 and cancer. NQO2 expression is known to be increased in some cancer cell lines. Since 3-arylidene-2-oxindoles are widely used in the design of new anticancer drugs, such as kinase inhibitors, it was interesting to study whether such structures have additional activity towards NQO2. Herein, we report the synthesis and study of 3-arylidene-2-oxindoles as novel NRH:quinone oxidoreductase inhibitors. It was demonstrated that oxindoles with 6-membered aryls in the arylidene moiety were obtained predominantly as E-isomers while for some 5-membered aryls, the Z-isomers prevailed. The most active compounds inhibited NQO2 with an IC50 of 0.368 µM. The presence of a double bond in the oxindoles was crucial for NQO2 inhibition activity. There was no correlation between NQO2 inhibition activity of the synthesized compounds and their cytotoxic effect on the A549 cell line.


Introduction
Human NRH:quinone oxidoreductase 2 (NQO2) is an enzyme that belongs to the quinone oxidoreductase gene family. The biological activity of NQO2 is quite multifaceted. NQO2 can be considered a detoxifying agent because it catalyzes two-electron reduction of quinones and quinoid compounds into hydroquinones [1]; however, NQO2 activity was also associated with ROS production [2][3][4]. This dual toxifying/detoxifying action of NQO2 is still being debated [5].
NQO2 was linked to the development of various diseases such as neurodegenerative disorders [3,[6][7][8][9], malaria [10], glaucoma [11], COVID-19 [12] and various cancers. It is important to note that the role of NQO2 in cancer pathogenesis is still being discussed. It was reported that NQO2 activity contributed to the proliferation of A549 and H1299 cancer cells [13], promoted the establishment of bone metastases of LNCaP-C4-2B prostate cancer [14] and indirectly influenced the activation of NF-kB which resulted in the suppression of cell apoptosis and protection of cancer cells against chemotherapy [15], so NQO2 is considered to be a potential anticancer drug target [16][17][18]. At the same time due to its high expression in cancerous cells, NQO2 can be used as an antitumor prodrug activator [19] and can act as an endogenous cancer suppressor [20,21]. The development of novel NQO2 inhibitors may elucidate the role of this enzyme in cancer pathogenesis.
Owing to the extensive research of Boutin and his team, it is generally assumed that NQO2 is the third melatonin binding site, MT3 [22]. Melatonin was able to inhibit NQO2 in the 50 µM range and it was speculated that melatonin antioxidant activity is at least partly a result of this interaction [23]. The majority of potent NQO2 inhibitors contain at least two or three fused aromatic rings, since it advantageous to possess a planar structure capable of engaging in π-π stacking interactions with the cofactor in the active site of the protein. Other known NQO2 ligands include flavonoids [5], ammosamide B analogues [24], imidazoacridin-6-ones [25], furan-amidines [26], etc.
Previously, we demonstrated that 2-oxindole derivatives can act as NQO2 inhibitors and bind to NQO2 binding site, analogous to melatonin [11,27]. In this work we report the development of new oxindole-based NQO2 inhibitors containing additional arylidene moiety in order to amplify the π-π stacking interactions with the FAD cofactor in the active center of the enzyme (Figure 1).

Scheme 1. Synthesis of 3-arylidene-2-oxindoles.
The detailed structures of the obtained 3-arylidene-2-oxindoles are presented in Table 4. The simultaneous reduction of the double bond and nitro group in compound 4 was performed using Zn/HCl (Scheme 2). Compound 47 was prepared analogously [29]. Scheme 2. The simultaneous reduction of nitro group and double bond in compound 4.

Determining the Configuration of Obtained Compounds
3-arylidene-2-oxindoles were often obtained as mixtures of isomers in various ratios, so it was important to establish convenient criteria for determining the predominant isomer. For several arylidene derivatives, we have identified the characteristic signals in the NMR spectra.
2.1.1. Correlation of 1H NMR Signals for E/Z Isomers of 4 -Substituted Benzylidene-2-Oxindoles For some 4 -substituted benzylidene-2-oxindoles, it was previously found that the chemical shift for the 2 (6 ) protons of benzylidene moiety in the 1 H NMR spectra are drastically different for Eand Z-isomers, being 7.45-7.84 and 7.85-8.53 ppm, respectively [30]. This significant difference exists due to the fact that 2 and 6 protons are deshielded by the carbonyl group of oxindole in the case of Z-isomers, and at the same time the corresponding protons of the E-isomer are shielded by the benzene ring of the oxindole core structure.
In order to determine the main configuration of our products, we analyzed the NMR spectra of the obtained 4 -substituted benzylidene-2-oxindoles (Table 1). We can confirm that this method for establishing the configuration is very convenient since the characteristic signals in Z-isomers significantly shift downfield and can be easily distinguished from other aromatic signals. We found the similar spectral trend for 3-(pyridin-2-ylmethylidene)-substituted oxindole derivatives. In E-isomers, the 1H NMR signal which belongs to the H 4 proton of the oxindole core has a significant downfield shift because of the deshielding by a nearby pyridine nitrogen. To confirm this hypothesis, we carried out an NOE experiment on compound 1 (Figure 2). This experiment confirmed the assignment of the downfield signal to the H 4 proton via its interaction with the other protons of the oxindole core. The NOE correlation was observed between the downshifted signal of H 4 and H 3 ' of the pyridine ring, but no interaction was detected between the methylene proton H and the H 4 oxindole proton. Thus, we have shown that the structure of compound 1, which was previously incorrectly identified as Z [28], actually has the E configuration. A pair of doublets (H 4 and pyridine H 3 signals) located downfield turned out to be a distinguishing feature of (E)-3-(pyridin-2-ylmethylidene)-2-oxindole derivatives. Thus, we carried out the assignment of signals for all the obtained compounds and identified the predominant isomer in all cases ( Table 2).

Correlation of 1H NMR Signals for E/Z Isomers of Pyrazole Derivatives
We found Z-isomer to be predominant in some pyrazole derivatives. The NOE experiment for 3-(1-methyl-1H-pyrazol-4-ylmethylidene)-2-oxindole 41, obtained as a single isomer, showed the interaction of the double bond proton with the proton H 4 of the oxindole fragment (Figure 3), indicating their cis configuration. Table 2. Preferred configuration of 3-(pyridin-2-ylmethylidene)-2-oxindoles and corresponding characteristic chemical shifts. The configurations of pyrazole derivative 45 was assigned by comparison of its 1H NMR spectrum with one of the reference compounds. The spectral characteristics of the aromatic region of the 1H NMR spectrum of 45 suggested the predominance of the Z-isomer in this compound as well.

The Influence of the Reaction Conditions on Isomer Ratio
As can be seen from the above studies, usually one isomer predominates in the synthesized oxindoles. Since Eand Z-isomers may have different biological activities, we searched for new reaction conditions that could increase the content of the minor isomer. We investigated the solvent effect and the influence of microwave activation (MW) on the yields and isomeric composition of the products. The reactions were carried out using standard thermal activation and a standard protic solvent (ethanol), an aprotic low-polar solvent (dioxane) and an aprotic polar solvent (ethyl acetate). As model experiments, both reactions with donor and acceptor aldehydes were carried out (Table 3). It is interesting that in case of pyrazole derivatives 39 and 44, the ratio of isomers in all conditions indicated in Table 3 was not dependent on solvent or activation method while for both 6-membered aryls, the quantity of the minor Z-isomer increased with changing ethanol to an aprotic solvent (Table 3). When carrying out the reaction with an acceptor aldehyde in an aprotic solvent, intermediate 1a precipitated after 10 min and can be isolated. Interestingly, in the case of dioxane, the reaction went further until complete conversion to compound 1 as a mixture of isomers, while in ethyl acetate the reaction stopped at the stage of formation of product 1a (Scheme 3). Thus, for non-pyrazole aldehydes, changing the solvent and MW can increase the yield of the minor Z-isomer, but does not lead to a complete inversion of the isomeric composition of the products. To obtain the E-isomer, standard reaction conditions remain preferable.

Biological Activity of the Obtained Compounds
The inhibition activity of all synthesized compounds at a concentration of 10 µM was preliminarily tested in vitro using human recombinant NRH:quinone oxidoreductase (NQO2). For active compounds, IC 50 values were also determined. Quercetin [31,32] and melatonin [22] were used as positive controls (Table 4, Figure S21). It was found that the presence of a 3-arylidene moiety was necessary for NQO2 inhibition, while 3-alkyl derivatives were essentially inactive. The most active compounds contain an OH-group in the arylidene moiety (13,(15)(16)(17)(18). Addition of a 5-acylamino or 5-carbamoylamino group may increase the compound's inhibition activity, and an increase in affinity was also caused by the addition of a methyl group to the amide nitrogen of the oxindole ring (as demonstrated by compounds 1 and 2 or 14 and 15).  The mechanism of NQO2 inhibition by one of the lead compounds, 15, was elucidated using a kinetic experiment (Figure 4 and Figure S22). The reaction rate was monitored under a range of inhibitor (0-25 µM) and BNAH substrate concentrations (9.375-150 µM). We found that both maximum rate V max and Michaelis constant K m decreased while V max /K m ratio increased at higher BNAH concentrations. Hence, 15 behaves as a typical mixed-type (noncompetitive) inhibitor.
The cytotoxicity of the synthesized oxindoles was studied using the A549 cell line in which the levels of NQO2 are increased (Table 4). We found that some oxindoles had moderate cytotoxic effects in the micromolar range of concentration, but no correlation between NQO2 inhibition activity and influence on cell viability was observed. The cytotoxicity of the compounds may be due to the inhibition of other enzymes associated with the activation of apoptosis in cancer cells, for example GSK3b or tyrosine kinases, or with the general non-specific toxicity of these compounds. Thus, the influence of NQO2 on A549 cell viability has not been established.

In Silico Studies
To analyze the mode of binding of the obtained oxindole derivatives to the NQO2 active site, we performed ligand-protein (Induced Fit) calculations using the Schrodinger software package. The 3OWH PDB structure [33] was used as a protein model. The calculation results are given for active compound 15 ( Figure 5), and similar patterns can be traced for the other derivatives (see Supplementary Materials, Figure S23 and Table S1). The possibility of π-π stacking of both Eand Z-isomers of 15 with FAD are shown. For the E-isomer, the binding geometry (the first calculated positions according to the docking score) was extremely similar to MCA-NAT ( Figure 5A) (PDB 3OVM [33]) and to our previously reported 2-oxindole analogue [27]. The position of the Z-isomer was similar to the E one and to the classic inhibitors mentioned above ( Figure 5B). This allows us to conclude that there is a possibility of binding both isomers (Z and E) in a manner similar to MCA-NAT.

Chemistry
All solvents were used as received without further purification. The reactions were monitored by thin layer chromatography (TLC) carried out on Merck TLC silica gel plates (60 F254), using UV light for visualization. Flash column chromatography purifications were carried out using silica gel 60 (particle size 0.040-0.060 mm). 1 H and 13 C NMR spectra were recorded at 298 K on a Bruker Avance 300 spectrometer with operating frequencies of 400.13 and 100.6 MHz, respectively, and calibrated using residual CHCl 3 (δH = 7.26 ppm) and CDCl 3 (δC = 77.16 ppm) or DMSO-d 5 (δH = 2.50 ppm) and DMSO-d 6 (δC = 39.52 ppm) as internal references. NMR data were presented as follows: chemical shift (δ ppm), multiplicity (s = singlet, d = doublet, dd = doublet of doublets, t = triplet, td = triplet of doublets, q = quartet, m = multiplet, br. = broad), coupling constant (J) in Hertz (Hz), integration. High-resolution mass spectra (HRMS) were measured on a Thermo Scientific LTQ Orbitrap instrument using nanoelectrospray ionization (nano-ESI). The reactions under microwave irradiation were carried out in a Monowave 300 microwave reactor (Anton Paar, GmbH, Graz, Austria), as well as in a Bosch HMT72M420 domestic microwave oven with a volume of 17 L. The reactions under thermal activation conditions were carried out on laboratory hot plates, as well as in a ChemiStation chemical reactor (EYELA). 1 H and 13 C NMR spectra of all novel compounds (Figures S1-S20) are given in Supplementary Materials.

Influence of solvent on E/Z Isomer Ratio
The general synthesis procedure was used for the synthesis of followed compounds with some variations: 1,4-dioxane or ethyl acetate was used instead of ethanol. organic fraction was dried with Na 2 SO 4 and the solvent was removed under reduced pressure. The compound (0.086 g, 65%) was obtained as brown powder. 1  The activity of recombinant human NQO2 (Sigma #Q0380, St. Louis, MA, USA) was evaluated kinetically using menadione and N-benzyl-dihydronicotinamide (BNAH) as the substrate and co-substrate, respectively. All reagents and test compounds were dissolved in 50 mM Hepes-KOH (pH 7.4) containing 1 mM of β-octyl-D-glucopyranoside, 0.1 mg/mL BSA and 1 µM FAD. In a 96-well black flat-bottom plate, 50 µL of the test compounds were introduced at a final concentration of 10 µM for primary screening, and a range of final concentrations from 10 nM to 100 µM was used to determine IC 50 values. Quercetin was used as a positive control. The solutions of the test compounds were pre-incubated for 5 min with 50 µL of human recombinant NQO2 (final concentration 42 ng/mL). Then, 25 µL of menadione was introduced at a final concentration of 100 µM. The reaction was started with 25 µL of 100 µM BNAH. The fluorescence of the co-substrate BNAH was followed at wavelengths of 370/440 nm at 37 • C using an Infinite M200 Pro microplate reader (Tecan, Grödig, Austria). The fluorescence data was fit according to a one-phase decay nonlinear regression to obtain slope values with Prism 8.0 (GraphPad Inc., San Diego, CA, USA). The activity in sample wells were normalized against negative control and enzyme-blank samples.

Antiproliferative Assay
The effects of the synthesized compounds on cell viability were determined using the MTT colorimetric test. All examined cells were diluted with the growth medium to 3.5 × 10 4 cells per mL and the aliquots (7 × 10 3 cells per 200 µL) were placed in individual wells in 96-multiplates (Eppendorf, Germany) and incubated for 24 h. The next day the cells were treated with the synthesized compounds separately at various concentrations for the determination of CC50 and incubated for 72 h at 37 • C in a 5% CO 2 atmosphere. After incubation, the cells were then treated with 40 µL MTT solution (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, 5 mg mL −1 in PBS) and incubated for 4 h. After an additional 4 h incubation, the medium with MTT was removed and DMSO (150 µL) was added to dissolve the formazan crystals. The plates were shaken for 10 min. The optical density of each well was determined at 560 nm using a microplate reader GloMax Multi+ (Promega, Madison, WI, USA). Each of the tested compounds was evaluated for cytotoxicity in three separate experiments.

In Silico Studies
The in silico studies were carried out according to the standard algorithms of the Schrodinger software package and our previous work [38].

Conclusions
The synthesis and study of 3-arylidene-2-oxindoles as novel NRH:quinone oxidoreductase 2 inhibitors was performed. It was shown that the E-isomer was the predominant product in the synthesis of 3-benzylidene and 3-(pyridylmethylidene) derivatives, while for pyrazole-4-carbaldehyde derivatives, the Z-isomer sometimes predominated. An NMR criterion for determining the (E/Z)-configuration of 3-(pyridin-2-ylmethylidene)-2-oxindoles was proposed. It was shown that for E-isomers, the signals of protons in the 4th position of the oxindole ring are characteristic in the 1H NMR spectra and lie in the region of 8.9-10 ppm. The molecular modeling of binding to the active site of NQO2 demonstrated that both the Eand Z-isomers are capable of π-π stacking with the FAD cofactor and are located in the active site of the enzyme similar to MCA-NAT (a selective inhibitor of NQO2). The presence of hydroxy groups in the arylidene moiety and the introduction of a methyl group to the oxindole nitrogen led to an increase in binding affinity. The most active compounds 15, 17, 18, 24, and 39 inhibited NQO2 with IC 50 values of 0.37-0.62 µM.
No correlation was observed between the enzymatic inhibition activity and the cytotoxicity of 2-oxindole derivatives, so further investigation is required.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/molecules28031174/s1, Figures S1-S20: 1H and 13C NMR spectra of synthesized compounds; Figures S21 and S22: concentration dependence of inhibition activity for the most active compounds and Michaelis-Menten kinetic study for compound 15; Figure S23: docking simulation for compound 24; Table S1: docking score for most active

Data Availability Statement:
The data presented in this study are available on request from the corresponding authors.

Conflicts of Interest:
The authors declare no conflict of interest.