Design, Synthesis and Fungicidal Activity of Ester Derivatives of 4-(3,4-Dichloroisothiazole) 7-Hydroxy Coumarin

The development of new fungicides is vital for safeguarding crops and ensuring sustainable agriculture. Building on our previous finding that 4-(3,4-dichloroisothiazole)-7-hydroxy coumarins can be used as fungicidal leads, 44 novel coumarin ester derivatives were designed and synthesized to evaluate whether esterification could enhance their fungicidal activity. In vitro fungicidal bioassays indicated that compound 2ai displayed good activity against Alternaria solani, Botrytis cinereal, Cercospora arachidicola, Physalospora piricola and Sclerotinia sclerotiorum, with an EC50 value ranging from 2.90 to 5.56 μg/mL, comparable to the lead compound 1a, with its EC50 value ranging from 1.92 to 9.37 μg/mL. In vivo bioassays demonstrated that compounds 1a, 2ar and 2bg showed comparable, excellent efficacy against Pseudoperonospora cubensis at a dose of 25 µg/mL. Our research shows that the esterification of 4-(3,4-dichloroisothiazole) 7-hydroxycoumarins results in a fungicidal activity equivalent to that of its lead compounds. Furthermore, our density functional theory (DFT) calculations and 3D-QSAR modeling provide a rational explanation of the structure–activity relationship and offer valuable insights to guide further molecular design.


Introduction
Agrochemicals are important strategical materials for crop protection and food safety production [1,2]. However, the unreasonable and unscientific use of some pesticides has also resulted in a series of risks [3,4], including increasing pest resistance [5], toxicity against non-target organisms and potential environmental pollution [6]. Therefore, the development of green and high-activity agricultural chemicals with good environmental behavior is always crucial for plant protection.
Natural products are usually metabolites or secondary metabolites of plants, microorganisms and animals, which have the advantages of structural diversity, biocompatibility, low resistance risk, a unique mode of action and a specific target [7][8][9][10][11]. They are widely used in the development of novel medicines and agrochemicals [12,13]. Coumarins are a highly significant class of natural products with bactericidal [14], anticoagulant [15], anti-inflammatory [16], antiviral [17], anti-HIV [18] and antitumor activity [19,20]; they are considered as phytoalexins for plant defense against pathogen attacks [1,21]. Several coumarin derivatives, such as osthole and coumoxystrobin, have been reported as fungicides [22,23] (Figure 1). Lead optimization is an effective approach for the development of novel pesticides [24,25]. Among the different methods for novel pesticide development, intermediate derivatization is effective and widely adopted [26,27]. Previously, our group discovered 4-(3,4dichloroisothiazole) 7-hydroxy coumarins as fungicidal leads by combining the naturally occurring coumarin and a substructure (i.e., 3,4-dichloroisothiazole) of the compounds with systemic acquired resistance. Although compounds 1a and 1b displayed favorable in vitro activity, their in vivo performance was suboptimal [1]. Subsequently, a series of 3,4-dichloroisothiazolocoumarin-containing strobilurins were developed and found to have poor solubility [2]. To continue the exploration of fungicidal leads and improve the physicochemical properties of a weak acidic moiety of 7-hydroxy group on the coumarins, the esterification of 7-hydroxy coumarins was carried out (Figure 2), the in vitro and in vivo fungicidal activity of the target compounds were simultaneously evaluated, and notably, compounds 2ai and 2ar were found to show good activity.

Chemical Synthesis
The synthetic route for the target compounds 2aa-2bv is outlined in Scheme 1. The target compounds 2aa-2bv were prepared via esterification using compound 1 with the corresponding acyl chloride. The specific structures of the target compounds are shown in Table 1.
The crystal structure cell stacking diagram of 2be, shown in Figure 3, was identified via X-ray diffraction analysis for structure validation. The crystal belongs to the monoclinic system with the space group P2 1/c , and each unit cell contains four molecules. The unit cell parameters are a = 16.9633 (6)  In addition, the torsional angles (all near 180 • or 0 • ) and the measured dihedral angles between the benzene and pyranone rings indicate that the entire coumarin ring was coplanar. Each molecule has a coumarin ring and an isothiazole moiety with a measured dihedral angle of 55.1 • between these two moieties. There are no hydrogen bonds to be found in the crystal structure. The additional data for 2be were alternatively obtained from the Cambridge Crystallographic Data Center (www.ccdc.cam.ac.uk, accessed on 1 June 2023), denoted as CCDC 2256560.

In Vivo Fungicidal Activity
The in vivo fungicidal activity of the target compounds was tested against Pseudoperonospora cubensis. The results are shown in Table 4. Compounds 2ai, 2aj, 2ar and 2bg showed excellent activity against P. cubensis at 200 and 100 µg/mL, results which were consistent with compounds 1a and 1b. When the concentration was reduced to 50 µg/mL and 25 µg/mL, the compounds 2ar maintained 80% and 75% inhibition rates, and compound 2bg maintained 88% and 65% inhibition rates, being equivalent to the positive control compound 1a. In short, 2ar and 2bg showed excellent in vivo activity against P. cubensis, and they could be potential candidates for further research.

3D-QSAR Analysis
The analysis of the statistical parameters of the established 3D-QSAR model was carried out on the basis of the optimal molecular conformation of fungicidal activity, and the statistical results are shown in Table 5. The CoMFA model includes the effects of the steric field and electrostatic field. When the optimum component was five, the parameters (Q 2 = 0.797, R 2 = 0.979, SEE = 0.096, F = 166.514) were obtained. For CoMFA models, when Q 2 is greater than 0.5 and R 2 is greater than 0.8, this suggests that the models have high reliability and a good prediction ability. The predicted versus experimental values ( Figure 4) obtained using this 3D-QSAR model indicate that the predictive capability of this model is robust.  In the steric field equipotential diagram of the CoMFA model, the green area indicates that the large-volume group is beneficial for improving the molecular activity, and the yellow area indicates that the large-volume group is not beneficial for improving the molecular activity. Figure 5A indicates that increasing the group volume of the 3-position and 7-position on the coumarin ring is conducive to improving the inhibitory activity of the target compounds against P. piricola. In the electrostatic field equipotential diagram of the CoMFA model, the blue area indicates that the positively charged group is beneficial for improving the molecular activity, and the red area indicates that the negatively charged group is beneficial for improving the molecular activity. Figure 5B indicates that increasing the positively charged group in the 7-position on the coumarin ring is conducive to improving the inhibitory activity of the target compounds against P. piricola.
As mentioned above, a larger group and positively charged group at the 7-position of the coumarin ring and a larger group at the 3-position of the coumarin ring are beneficial for improving fungicidal activity against P. piricola.

DFT Calculations of 2ai, 2aq and 1a
The molecular total energy (MTE), frontier orbital energy (FMO), energy gaps between the HOMO and LUMO of compounds with high (2ai) or low (2aq) fungicidal potency and the positive control 1a are listed in Table 6. The energy gaps between the HOMO and LUMO were 4.17, 3.83 and 4.00 eV, respectively (Table 6). LUMO and HOMO affect electron transition by accepting and providing electrons, respectively, according to the FMO theory. In Figure 6, the HOMO of 2ai and 1a is mainly located on the benzopyranone ring, while the LUMO is mainly located on the benzopyranone ring and the isothiazole ring, suggesting that the electrons of both compounds were transferred from the benzopyranone ring to the isothiazole ring. However, the HOMO of 2aq with low fungicidal potency is primarily distributed in the naphthalene ring, and the LUMO is distributed in the ester bond and benzopyranone and isothiazole rings, indicating that the direction of electron transition is different from that of 2ai with high fungicidal potency and the positive control 1a.
The aforementioned results regarding the energy and electron transition of 2ai are in approximate agreement with those of 1a, whereas that for 2aq was different, explaining their difference in fungicidal potency on the basis of their chemical structures.

Instruments and Materials
All solvents and reagents used were obtained from commercial sources and were not further purified. The melting points were measured using an X-4 digital display micromelting point apparatus (Gongyi, China) and were not corrected. 1 H NMR (400 MHz), 13 C NMR (101 MHz) or 19 F NMR (376 MHz) spectra were measured using TMS as the internal standard with a Bruker AV 400 spectrometer in CDCl 3 . High-resolution mass spectrometry (HRMS) data were measured using a Varian QFT-ESI instrument. Crystal structures were determined using a Rigaku 007 Saturn 70 diffractometer.

General Procedures for Preparation of the Target Compounds 2aa-2bv
According to the reported method, intermediates 1a and 1b were synthesized [1]. The compounds 2aa-2bv were synthesized according to the reported method with minor modifications [28].
To a mixture of compound 1 (0.25 mmol), 4-DMAP (0.50 mmol) in dichloromethane (15 mL) and triethylamine (0.50 mmol), acyl chloride (0.75 mmol) was slowly added under stirring, and the reaction mixture was then stirred overnight at room temperature. After the reaction was complete, the mixture was extracted with dichloromethane (3 × 15 mL). The organic phase was sequentially washed with saturated NaHCO 3 (20 mL) and brine (20 mL) and dried over anhydrous Na 2 SO 4 . The solution was then filtered and concentrated under reduced pressure. The crude product was purified using column chromatography with a mobile phase of dichloromethane/methanol (500:1-200:1, v/v) to yield compounds 2aa-2bv.

X-ray Diffraction
The crystal structure of compound 2be was cultured using the fusion method. Using a hot stage to heat the compound until it melted, the temperature was then meticulously controlled just below the melting point during the slow crystal growth process. The crystal structure was determined using a Rigaku Saturn 724 CCD diffractometer equipped with graphite-monochromatic Mo Kα radiation (λ = 0.71073 Å) at 113 K. The atomic coordinates for 2be were deposited in the Cambridge Crystallographic Data Centre (CCDC).

In Vitro Fungicidal Activity
The fungicidal activity of compounds 2 against A. solani (A. s), B. cinerea (B. c), C. arachidicola (C. a), Fusarium graminearum (F. g), P. piricola (P. p), R. solani (R. s) and S. sclerotiorum (S. s) were evaluated in vitro at a concentration of 50 µg/mL according to the mycelium growth inhibition method. Pure drugs of osthole, isotianil and compounds 1a and 1b were chosen as positive controls. The median effective concentration (EC 50 ) of the active target compounds was also assessed according to the mycelium growth inhibition method. To prepare the test solution, a highly concentrated mother solution of the test substance was carefully created at a concentration of 25,000 µg/mL using N, N-dimethylformamide as the solvent. A precise amount of the mother solution was added to sterile water containing 0.1% Tween 80 to create a test solution with a concentration of 500 µg/mL. This solution was then mixed with potato dextrose agar (PDA) to create a culture with a controlled concentration of 50 µg/mL of the test compound. A blank control was also included using pure DMF without the target compound. To ensure accuracy, three replicates were included for each treatment. The EC 50 was determined using seven concentration gradients ranging from 1.56 to 100 µg/mL, and advanced DPS software was used to process the data.

In Vivo Fungicidal Activity
The in vivo activity of compounds 2 against Pseudoperonospora cubensis on cucumber was evaluated at Shenyang Sinochem Agrochemicals R&D Co., Ltd., Shenyang, China. Pure osthole and cyzofamid were used as a positive control. The compounds with an inhibition rate of 100% were chosen for further tests with a concentration gradient. To prepare the test solution, 10 mg of the compound to be tested was accurately weighed and dissolved in 0.5 mL of N, N-dimethylformamide. The resulting solution was then diluted with distilled water containing 1% Tween 80 to create a stock concentration of 500 µg/mL. Potted plants with uniform growth were carefully selected as test plants, and the stock solution was further diluted with distilled water to the desired test concentration. The test solution was then sprayed onto the surface of the plant and allowed to dry for 2 h to allow for solvent evaporation. After 24 h, the treated test colonies were properly inoculated on the leaves of the plant for three replicates. A blank control was also included, which involved spraying the plants with plain distilled water containing no test solution. The treated plants were then cultured normally in the greenhouse, and the antifungal efficacy was assessed after 7 d.

3D-QSAR Molecular Modelling
Next, 3D-QSAR studies were carried out with 29 target compounds (listed in Table 7) with fungicidal activity against P. piricola. The construction of the 3D-QSAR model was completed in the CoMFA modules in SYBYL X 2.1.1. All the structures were constructed using the molecular fragment library of the SYBYL software. The Powell method was used in the optimization and conformation search process. The gradients were limited to 0.005 kJ·mol −1 , and the maximum number of iterations was 1000. The selected smallmolecular force field was the Tripos force field, and the charge carried by the molecule was the Gasteiger-Hückel charge. The potent compound 2aj was chosen as a template molecule. The biological activity was expressed in terms of D (Table 7) using the following formula [29]: D = log 10 a 1 − a + log 10 M w where a is the percentage inhibition against P. piricola at 50 µg/mL in vitro, M w is the molecular weight of the target compounds, and D is the biological activity index.

Density Functional Theory (DFT) Calculation
Based on their fungicidal activity, the highly active compound 2ai, the slightly active compound 2aq and the lead compound 1a were selected for molecular orbital calculations using Gaussian 09W and drawn in Gaussview 6.0. The structures were preprocessed and geometrically optimized, and frequency calculations were carried out to generate geometrically stable structures using the DFT-B3LYP/6-31G (d) method. Single-point energy calculations were then performed using the DFT-B3LYP/6-31G (d, p) method.

Conclusions
A series of new coumarin derivatives were designed and synthesized. The bioassay results indicated that compound 2ai exhibited broad-spectrum fungicidal in vitro activity, and compounds 2ar and 2bg showed good fungicidal activity against P. cubensis in vivo. Overall, we successfully adjusted the solubility of this series of compounds via esterification, and the activity was equivalent to that of the lead compound. The 3D-QSAR analysis showed that a larger group and positively charged group at the 7-position of the coumarin ring and a larger group at the 3-position of the coumarin ring were beneficial for improving fungicidal activity against P. piricola, the results of this study provide guidance for further coumarin lead based fungicide development.