Synthesis, 3D-QSAR and Molecular Docking Study of Nopol-Based 1,2,4-Triazole-Thioether Compounds as Potential Antifungal Agents

Cytochrome bc 1 complex is an important component of cellular respiratory chain, and it is also an important target enzyme to inhibit the growth of plant pathogens. Using cytochrome bc 1 complex as the target enzyme, twenty-three novel nopol-based 1,2,4-triazole-thioether compounds were designed and synthesized from natural preponderant resource β-pinene, and their structures were confirmed by FT-IR, NMR, ESI-MS and elemental analysis. The in vitro antifungal activity of the target compounds 5a-5w was preliminarily evaluated against eight plant pathogens at the concentration of 50 µg/ml. The bioassay results showed that the target compounds exhibited the best antifungal activity against Physalospora piricola, in which compounds 5b (R= o-CH3 Ph), 5e (R= o-OCH3 Ph), 5h (R= o-F Ph), 5m (R= o-Br Ph), 5o (R= m,m-OCH3 Ph), and 5r (R= p-OH Ph) had inhibition rates of 91.4, 83.3, 86.7, 83.8, 91.4 and 87.3%, respectively, much better than that of the positive control chlorothalonil. Also, compound 5a (R= Ph) had inhibition rate of 87.9% against Rhizoeotnia solani, and compound 5b (R= o-CH3 Ph) had inhibition rates of 87.6 and 89% against Bipolaris maydis and Colleterichum orbicala, respectively. In order to develop novel and promising antifungal compounds against P. piricola, the analysis of three-dimensional quantitative structure-activity relationship (3D-QSAR) was carried out using the CoMFA method on the basis of their antifungal activity data, and a reasonable and effective 3D-QSAR model (r 2 = 0.944, q 2 = 0.685) has been established. In addition, the theoretical study of molecular docking revealed that the target compounds could bind to and interact with the site of cytochrome bc 1 complex.


INTRODUCTION
Plant diseases caused by phytopathogens have always been one of the main reasons for crop yield reduction, and the use of fungicides is the most critical method to effectively control crop diseases in agriculture . However, the frequent use and misuse of many traditional pesticides have caused environmental pollution, pesticide residual toxicity, and the emergence of resistant phytopathogenic fungi isolates (Jiao et al., 2021;Qi et al., 2021). Therefore, the development of novel effective antifungal agents is of great significance.
Cytochrome bc 1 complex (EC 1. 10.2.2), also known as complex III, is an important component of cellular respiratory chain, as well as a target enzyme used in the development of fungicides owing to its ability to inhibit the growth of plant pathogens (Chen et al., 2016;Cheng et al., 2018). At present, about twenty cytochrome bc 1 complex fungicides have been successfully employed in the world market and more are still being developed according to the statistics from the Fungicide Resistance Action Committee (FRAC, 2021) including flufenoxystrobin, fluoxastrobin, famoxadone, pyrametostrobin and so on (http://www.frac.info/ (accessed Aug 12, 2021). These compounds inhibit mitochondrial respiration of plant pathogens by binding at the Qo site of a membrane-bound homodimeric cytochrome bc 1 complex and blocking the generation of adenosine triphosphate (ATP), leading to the inhibition of the energy production which is essential for survival (Wang et al., 2018;Zhu et al., 2019). Therefore, cytochrome bc 1 complex inhibitors have been widely investigated by a large number of chemists to discover novel fungicides. In the past decade, some kinds of potential cytochrome bc 1 complex inhibitors were designed and synthesized by Yang research team and some target compounds exhibited good in vitro inhibitory activities against downy mildew and powdery mildew (Wang et al., 2011;Hao et al., 2012;Hao et al., 2015).
Turpentine oil, with a productivity of around 330,000 tons for the most recent decades, is a natural renewable biomass resource which can be obtained by the steam distillation of the oleoresin exudate from living pine trees (García et al., 2020). It is one of the most widely produced secondary metabolites of plants, and its two major components are α-pinene and β-pinene, which have received much attention in recent years due to their various biological activities (Nie et al., 2014). The content of β-pinene in turpentine oil of Pinus elliottii can be as high as about 30% (Xu et al., 1992). Nopol, in which molecular scaffold of the natural α-pinene is maintained, can be prepared by the Prins reaction of β-pinene with paraformaldehyde under catalysis of Lewis acid (Yadav and Jasra, 2006;Jadhav et al., 2010;Vrbková et al., 2020). Nopol and its derivatives exhibited a broad spectrum of biological activities, such as antifungal Jin et al., 2017;Feng et al., 2019;Chen et al., 2021), repellent (Han et al., 2008), antifeedant (Han et al., 2007), and treatment of diabetes and gastrointestinal irritable syndrome activities (He et al., 2008;Majouli et al., 2017). Herein, nopol deserves further study for agrochemical or pharmaceutical uses based on its bioactive property and chemical reactivity.
In continuation of our interest in developing natural productbased bioactive compounds Wang X. et al., 2018;Lin et al., 2018;Kang et al., 2019;Liu et al., 2020;Zhu et al., 2020;Chen et al., 2021;He et al., 2021), a series of novel nopol derivatives containing 1,2,4-triazole and thioether moieties were designed and synthesized by the strategy of molecular docking-based virtual screening based on the crystal structure of cytochrome bc 1 complex. In addition, a three-dimensional structure-activity quantitative relationship (3D-QSAR) model was established by the comparative molecular field analysis (CoMFA) method.

General
All other materials and reagents were purchased from commercial suppliers and used as received. β-Pinene (GC purity 98%) was provided by Jiangxi Xuesong Natural Medicinal Oil Co., Ltd. 5-Substituted-1,2,4-triazole-3-thiones were prepared by our laboratory according to the literature (Liu et al., 2012). The GC analysis was performed on an Agilent 6890 GC equipped with an HP-1 (30 m, 0.530 mm, 0.88 µm) column. NMR spectra were recorded using tetramethylsilane (TMS) as the internal standard and deuterated chloroform (CDCl 3 ) as a solvent on a Bruker Avance III HD 600 MHz spectrometer. Mass spectra were obtained by means of the electrospray ionization (ESI) method on TSQ Quantum Access MAX HPLC-MS instrument. The IR spectra were recorded by employing a Nicolet iS50 FT-IR spectrometer using the KBr pellet method. Elemental analyses were measured using a PE 2400 II elemental analyzer.

Antifungal Assay
All the target compounds were evaluated for their in vitro antifungal activities against eight plant fungi by the agar dilution method according to the literature (Su et al., 2013). Each tested compound was dissolved in acetone and diluted with Sorporl-144 (200 mg/L) as an emulsifier to prepare the solution with a concentration of 500 mg/L. The stock solution (1 mL) was added to the Potato-Sugar-Agar (PSA, 9 mL) culture medium. Then, mycelion dishes of 5 mm diameter were cut along the external edge of the mycelium was transferred to the center of flat containing the tested compound and put in equilateral triangular style in triplicate. After culturing 48 h at (24 ± 1)°C in the incubator, the expanded colony diameters of strains were measured and compared with that treated with aseptic distilled water, and then calculated the relative inhibition percentage. The corresponding aseptic distilled water without the test samples was served as a blank control and the commercial protective fungicide chlorothalonil was employed as the positive control. The test for each target compound was repeated three times. The inhibitory rate of the tested compound was calculated by the formula: inhibitory rate [(the average extended mycelium diameter of the blank assay−the average extended mycelium diameter after treatment with emulsion)/the average extended mycelium diameter of the blank assay] × 100%.

3D-QSAR Study
The 3D-QSAR model was built using the CoMFA method of Sybyl-X 2.1.1 software to investigate the relationship between the antifungal activity and the compound structure according to our previous report (Zhao et al., 2021). The 3D structures of compounds 5a-5r were built on the sketch module and minimized by the conjugate gradient method in the program based on the Tripos force field using termination convergence energy of 0.005 kcal/ (mol*Å), a maximum of 1,000 iterations, and Gasteiger-Hückel charge as the parameters. The compound 5b with the best activity against P. piricola was selected as the template compound, of which the atoms marked with an asterisk (Figure 1) were used as the common superimposed skeleton. 16  compounds using as the training set were aligned to form a superimposed model (Figure 2) and the remaining three compounds bearing aromatic R groups constitute the test set. Their inhibitor rates against P. piricola were converted to the active factor (AF) values by the formula: where I was the inhibition rate at 50 µg/mL and MW was the molecular weight. The built 3D-QSAR model was checked by the partial least squares method. The modeling predictive capability was indicated by a correlation coefficient squared r 2 , a crossvalidated squared q 2 , a standard deviation S, and a Fisher validation value F.

Molecular Docking Study
To understand the binding interactions of the target compounds with the active site of cytochrome bc 1 complex, the molecular docking procedures were carried out using AutoDock 4.2.6 software according to the reported paper Du et al., 2020 The grid box was generated by applying default parameters. The docking calculations were performed using the Lamarckian genetic algorithm (GA) and the number of GA runs was set to 20 conformations. The lowest binding energy in the maximum cluster for the docked conformations was chosen as the representative binding energy. The binding energy between the docked compound and the enzyme was calculated using the AutoGrid program with a grid spacing of 0.375 Å and the Lamarckian genetic algorithm as a searching method. When the docking results were generated, the binding energies of all docked compounds were also automatically obtained in this program.

Chemistry
The synthetic route of nopol-based 1,2,4-triazole-thioether compounds 5a-5w was illustrated in Scheme 1. At first, nopol 2 was prepared by Prins reaction of β-pinene 1 with paraformaldehyde under catalysis of Lewis acid ZnCl 2 (Yi et al., 2000). Then, nopyl chloroacetate 3 was prepared by alcoholysis reaction of chloroacetyl chloride with nopol 2 in a good yield , followed by nucleophilic substitution reaction of nopyl chloroacetate 3 with selfprepared 5-substituted-1,2,4-triazole-3-thiones 4 to afford a series of novel nopol-based 1,2,4-triazole-thioether compounds 5a-5w (Liu et al., 2007). The structures of all the synthesized compounds and the key intermediates 2 and 3 were characterized by FT-IR, 1 H NMR, 13 C NMR, ESI-MS, and elemental analysis, and the related spectra can be found in Supplementary Material. In the IR spectra of the target compounds, the weak absorption bands at about 3,025 cm −1 and about 1,610 cm −1 were attributed to the stretching vibrations of C-H and C C in the nopol moiety, respectively. The strong absorption bands at about 1,730 cm −1 and about 1,470 cm −1 were assigned to the stretching vibrations of C O in the ester group and C N in the 1,2,4-triazole moiety, respectively. Also, the absorption bands at 605-712 cm −1 revealed the presence of C-S-C in thioether moiety. In the 1 H NMR spectra, the olefinic protons of nopol scaffold showed signals at about 5.33 ppm, and the other protons

In Vitro Antifungal Activity of Target Compounds
The antifungal activity of the target compounds 5a-5w was evaluated using the agar dilution method against Fusarium   Frontiers in Chemistry | www.frontiersin.org October 2021 | Volume 9 | Article 757584 9 oxysporum f. sp. cucumerinum, Cercospora arachidicola, Physalospora piricola, Alternaria solani, Gibberella zeae, Rhizoeotnia solani, Bipolaris maydis, and Colleterichum orbicalare at a concentration of 50 µg/mL, using the commercial antifungal drug chlorothalonil as positive control. The results were listed in Table 1.
It was found that, at the concentration of 50 µg/mL, the target compounds exhibited the best antifungal activity against P. piricola, in which compounds 5b (R o-CH 3 Ph), 5e (R o-OCH 3 Ph), 5h (R o-F Ph), 5m (R o-Br Ph), 5o (R m, m-OCH 3 Ph), and 5r (R p-OH Ph) had inhibition rates of 91.4, 83.3, 86.7, 83.8, 91.4 and 87.3%, respectively, much better than that of the positive control chlorothalonil. Moreover, compound 5a (R Ph) showed inhibition rate of 87.9% against R. solani, and compound 5b (R o-CH 3 Ph) showed inhibition rate of 87.6% against B. maydis. It was also found that compound 5a (R Ph), 5b (R o-CH 3 Ph), 5e (R o-OCH 3 Ph), and 5m (R o-Br Ph) displayed inhibition rates of 81.9, 89.0, 81.9, and 84.3%, respectively, against C. orbicalare. On the whole, compound 5b (R o-CH 3 Ph) is worthy of further study due to its better broad-spectrum antifungal activity against the eight tested plant pathogens. It was also found that the R groups displayed a noticeable influence on antifungal activity, a 3D-QSAR study was then carried out.

3D-QSAR Study
The 3D-QSAR study was performed by CoMFA mothed to investigate the influence on the activity against P. piricola for the R groups. According to the report , the inhibition rates of the compounds 5a-5r were converted to the available active factor (AF) values, which were listed in Table 3. The validity for the 3D-QSAR model was checked by the partial least squares (PLS) analysis including a correlation coefficient squared r 2 (close to 1), a cross-validated squared q 2 (>0.5), a standard deviation S (close to 0), a Fisher validation value F (>100), and an optimal number of component (ONC). Their values for the built 3D-QSAR model were shown in Table 2. The r 2 was 0.944, q 2 was 0.685, S was 0.099, F was 133.642, and ONC was 5, suggesting that the built 3D-QSAR model was valid. The experimental AF values, the predicted AF′ values, and their residue values were presented in Table 3, and the scatter diagram of AF vs. AF′ was shown in Figure 3. All data were concentrated near the X Y line, also indicating that the 3D-QSAR model was reliable and had a good predictive ability.
The relative contributions of the steric and electrostatic fields of the 3D-QSAR model were 55.9 and 44.1%, respectively. In Figure 4A, the steric field contours were represented with two different colors: the green contour indicated that the R group embedding at 2-position of benzene ring was beneficial to increase the antifungal activity but yellow was the opposite. For instance, compounds 5b (R o-CH 3 Ph) and 5e (R o-OCH 3 Ph) possessed better antifungal activity against P. piricola than compounds 5d (R p-CH 3 Ph) and 5g (R p-OCH 3 Ph). In Figure 4B, the electrostatic contours were represented in two distinguishable colors: the blue enclosed volume represented that the R group with an electrondonating surface embedding in this area will favor the increase   in activity, while red defines the opposite. The presence of electron-withdrawing groups such as fluorine or chlorine atom as well as bromine atom at 2-position of benzene ring was favorable for higher antifungal activity. For example, compounds 5h (R o-F Ph), 5k (R o-Cl Ph), and 5m (R o-Br Ph) showed a higher inhibitor rate against P. piricola than the unsubstituted compound 5a (R Ph). In contrast, the introduction of the electron-donating group at the 3-position of benzene ring played a crucial role in the antifungal activity. For instance, 5c (R m-CH 3 Ph) and 5f (R m-OCH 3 Ph) displayed better antifungal activity against P. piricola. The 3D-QSAR study of these title compounds can provide useful information for the further rational design of novel nopol derivatives. Herein, based on the results of 3D-QSAR analysis above, two novel unsynthesized compounds ( Figure 5) were designed and the predicted ED values were calculated by the established CoMFA model. As a result, the designed compounds A (AF′ −1.445) and B (AF′ −1.589) showed potential excellent antifungal activities with inhibition rates of 94.68 and 92.93%, respectively, indicating that the antifungal activities of the proposed molecules were better than that of the compounds containing aromatic rings.

Molecular Docking Analysis
Using the cytochrome bc 1 complex (PDB ID 1SQB) as the target protease, the molecular docking study for all the target compounds was performed to investigate the correlation between the binding energy and antifungal activity. There were 20 conformations docked for each target compound (Figure 6), and the lowest binding energy in the maximum cluster for the docked conformations was chosen as the representative binding energy for the corresponding compound. The scatter plot of AF values versus binding energies for the title compounds was shown in Figure 7. All of the data were concentrated near the line Y 0.1564X-1.395, illustrating that there was a clear positive monotonic association between AF values and binding energies. In addition, Spearman's rank correlation coefficient analytical approach was carried out using IBM SPSS STATITICS 22 software to investigate the correlation between AF values and binding energies. The result were listed in Table 4 and Table 5. It was found that the correlation was significant at 0.001 (at 0.01 level). The Spearman correlation coefficient was 0.626, indicating that there was a significant positive correlation, namely, the activity gradually increased with the increase in binding energies. It suggested that the cytochrome bc 1 complex was a potential target protease.
In order to preliminarily explore the possible binding mode of compound 5b possessing the best antifungal activity with the typical cytochrome bc 1 complex inhibitor Azoxystrobin, molecular docking analysis was carried out on Sybyl X 2.1.1 program and AutoDock 4.2.6 software. The result was showed in Figures 8, 9. In Figures 8A1, 9B1, it can be observed that the compound 5b was well embedded into the active domain where the ligand azoxystrobin was found. The binding conformation of compound azoxystrobin and 5b in the active site of cytochrome bc 1 complex, as well as the 2D interactions map of compound azoxystrobin and 5b with residues were presented in Figures  8A2,B2,A3,B3. The 3D interactions map of compound azoxystrobin with residues were showed in Figure 8A4, and compound 5b with residues in Figure 9B4. It was found that compound 5b was buried into the binding pocket consisting of the residues LEU121, MET124, PHE274, PRO270, PHE128, TYR131, LYS269, ILE146, VAL145, etc. Meanwhile, two π-π stacking interactions (distance 2.12 and 2.90 Å) were observed between compound 5b and the residues PHE128 and PHE274, respectively. In addition, two π-π stacking interactions were observed between azoxystrobin and the residues PHE274 and TYR278, respectively. A H-bond was formed by the oxygen atom of methoxyacrylate moiety and the residue GLU271 in cytochrome bc 1 complex binding site (distance 2.55 Å), which was essential to stabilize the binding between cytochrome bc 1 complex inhibitors and cytochrome bc 1 complex. Based on these results, the binding domain and the binding mode of compound 5b were both similar to that of azoxystrobin, suggesting that they shared a similar action mode.

CONCLUSION
Using molecular docking-based virtual screening in the assumption of cytochrome bc 1 complex as the target enzyme, and natural preponderant resource β-pinene as starting material, twenty-three novel nopol-based 1,2,4-triazole-thioether compounds were designed and synthesized. Their structures were confirmed by FT-IR, NMR, ESI-MS, and elemental analysis. The in vitro antifungal activity of the target compounds 5a-5w was preliminarily evaluated against eight plant pathogens at the concentration of 50 µg/mL. The bioassay results indicated that some of the target compounds showed excellent inhibitory activity Frontiers in Chemistry | www.frontiersin.org October 2021 | Volume 9 | Article 757584 against the tested fungi, especially against P. piricola. Compound 5b (R o-CH 3 Ph) exhibited better broad-spectrum antifungal activity against the tested fungi. There was a significant positive Spearman's rank correlation between the antifungal activity and the docking-based binding free energy. Molecular docking study revealed that there was hydrophobic interactions between the target compounds and the key favorable residues of cytochrome bc 1 complex. In addition, a reasonable and effective 3D-QSAR model (r 2 0.944, q 2 0.685) has been established for the further development of novel and promising antifungal compounds.

DATA AVAILABILITY STATEMENT
The crystallographic data of the target protein cytochrome bc 1 complex (PDB ID 1SQB, 2.69 Å in resolution) used in the molecular docking study of our article was downloaded from the RCSB PDB web (https://www.rcsb.org/).

AUTHOR CONTRIBUTIONS
XW carried out the experimental work, participated in the discussion of biological activities, and wrote the paper; WD and GL constructed the target compound structure, designed the experimental scheme, directed and supervised the whole experimentation, 3D-QSAR study and theoretical calculation, discussed the biological activities, and revised the paper; BL, MC, and FL participated in the synthesis of target compounds and the discussion of biological activities.