Synthesis, Antiviral Bioactivity of Novel 4-Thioquinazoline Derivatives Containing Chalcone Moiety

A series of novel 4-thioquinazoline derivatives containing chalcone moiety were designed, synthesized and systematically evaluated for their antiviral activity against TMV. The bioassay results showed that most of these compounds exhibited moderate to good anti-TMV activity. In particular, compounds M2 and M6 possessed appreciable protection activities against TMV in vivo, with 50% effective concentration (EC50) values of 138.1 and 154.8 μg/mL, respectively, which were superior to that of Ribavirin (436.0 μg/mL). The results indicated that chalcone derivatives containing 4-thioquinazoline moiety could effectively control TMV. Meanwhile, the structure-activity relationship (SAR) of the target compounds, studied using the three-dimensional quantitative structure-activity relationship (3D-QSAR) method of comparative molecular field analysis (CoMFA) based on the protection activities against TMV, demonstrated that the CoMFA model exhibited good predictive ability with the cross-validated q2 and non-cross-validated r2 values of 0.674 and 0.993, respectively. Meanwhile, the microscale thermophoresis (MST) experimental showed that the compound M6 may interaction with the tobacco mosaic virus coat protein (TMV CP).


Introduction
Tobacco mosaic virus (TMV) is a positive-sense single stranded RNA virus that infects members of nine plant families and at least 125 species, including tobacco, tomato, pepper, cucumbers, and a number of ornamental flowers. Once infected with TMV, leaves tend to become mosaic, yellow, necrosis, and ruguse, which would affect crop production and quality, and caused $100 million in economic losses in worldwide [1]. However, there are no effective antiviral agents for controlling TMV. Therefore, it is a challenge to the development of novel, potent, and structurally concise antiviral agents.
Tobacco mosaic virus coat protein (TMV CP) assembly systems include helix and four-layer aggregate disk systems. The helix forms of TMV CP mainly exist in the presence of TMV RNA. In these helix forms, TMV CP has an important role in the self-assembly of TMV through an initial RNA recognition reaction that triggers the assembly, which is believed to be necessary for virus assembly initiation and elongation. The four-layer aggregate disk forms consisting of 34 subunit aggregates of TMV CP are crystallized as a dimer of bilayer disks with 17 subunits per layer in the absence of TMV RNA [24].
In this work, a series of novel 4-thioquinazoline derivatives containing chalcone moiety were designed, synthesized and systematically evaluated their antiviral activities against TMV. Biological results showed that some of the title compounds displayed moderate to good antiviral activity. Among the title compounds, compounds M 2 and M 6 showed appreciable protection activities against TMV in vivo, which were better than that of the commercial agricultural antiviral agent Ribavirin. In addition, the structure-activity relationship (SAR) of the compounds was also discussed using comparative molecular field analysis (CoMFA) of the three-dimensional quantitative structure-activity relationship (3D-QSAR) method based on the protection activities against TMV. The microscale thermophoresis (MST) experimental showed that the compound M 6 may interactions with the TMV CP. To the best of our knowledge, this is the first report on the synthesis, antiviral activity, 3D-QSAR, and interaction study of 4-thioquinazoline derivatives containing chalcone moiety.

Chemistry
The summary of the synthetic route designed for the 4-thioquinazoline derivatives containing chalcone moiety were shown in Scheme 1. Substituted 2-aminobenzoicacid, using as the starting materials, was reacted with formamide for 4.5 h at 140 to 145 °C to obtain the intermediate 4(3H)-quinazolinone (1). Substituted 4-chloroquinazoline (2) was prepared by chlorination reaction with SOCl2 [8]. Then, using ethanol as the solvent, nucleophilic substitution reaction of 2 and carbamimidothioic acid, transformed from thiourea via keto-enol tautomerism equilibrium, was reacted 8 h at 87 °C to obtain the intermediate quinazolin-4-yl carbamimidothioate, then quinazolin-4-yl carbamimidothioate was reacted with potassium hydrate, and adjusted pH value to 7 by acetic acid to get the key intermediate quinazoline  The structures of all the title compounds confirmed through IR, 1 H-NMR, 13 C-NMR spectral and elemental analyses. The IR spectra of the title compounds M 1 -M 21 exhibited characteristic absorption bands at 1690-1640 cm −1 , which indicated the presence of C=O. The stretching frequency at 1620-1600 cm −1 was assigned to C=N vibrations. 1 H-NMR indicated that all of the phenyl protons showed multiplets at 8.27-6.93 ppm. The main characteristic of the 1 H-NMR spectra for the target compounds was the presence of a high-frequency downfield singlet δH 9.00-9.02 for Qu-2-H. And -O-CH2-and -S-CH2-showed a triplet at 4.40-4.45 ppm and 3.80-3.84 ppm, respectively. Moreover, the chemical shifts shown at nearly 188.40-189.40, 66.50-66.60, and 28.22-28.33 ppm in 13 C-NMR were also confirmed the presence of C=O, -OCH2-, and -SCH2-, respectively.

Antiviral Activity and Structure Activity Relationship against TMV
As an extension of this approach, the structure-activity relationships were deduced on the basis of activity values in Tables 1 and 2. Some of the compounds showed potency against TMV. When the R were 2-CH3-Ph (M 1 ), 4-CF3-Ph (M 2 ), 3,4-di-OCH3-Ph (M 6 ), and 2-CF3-Ph (M 10 ) groups, the corresponding target compounds exhibited good curative activity. When R group was substituted with fused ring, it was disfavored for the anti-TMV following the order of M 5 (Naphthalene-2-yl) < M 7 (Thiophene-2-yl). And, when the R were 4-CF3-Ph (M 2 ) and 3,4-di-OCH3-Ph (M 6 ) groups, the corresponding compounds exhibited good protection activities against TMV which was surpassed that of Ribavirin. However,

3D-QSAR Study
In order to analyze the SAR base on the protection activity against TMV, the CoMFA of 3D-QSAR model [28] with a total of twenty-one target compounds were developed by using Sybyl 7.3 software [29] from Tripos Inc. (St. Louis, MO, USA). Predicted pEC50 [30] values of compounds in both the training and testing sets were presented together with their actual pEC50 values in Table 2, and correlations between predicted and actual pEC50 in CoMFA model were presented in Figure 1. Overall, predicted EC50 values were very close to the corresponding actual values for compounds in both the training and testing sets. The mostly linear correlations in Figure 1 demonstrated high predictive power of the CoMFA model. Meanwhile, as shown in Table 3, the non-cross-validated PLS analysis was repeated with the optimum number of components, as determined by the cross-validated analysis. To obtain statistical confidence limits, the non-cross-validated analysis was repeated, which yielded an r 2 value of 0.993, and the q 2 value of highly predictive CoMFA was 0.674 with 8 ONC, which suggested that the model has good predictive ability (r 2 > 0.9, q 2 > 0.5). Meanwhile, as shown in Table 3, the SEE was 0.020 and the F value was 161.503, respectively. The relative contributions to bioactivity from steric and electrostatic fields in the CoMFA model were 0.478 and 0.522, respectively, suggesting that bioactivity was mainly determined by electrostatic interactions.  The CoMFA contour map of the steric was shown in Figure 2A. Green contours in the CoMFA steric field indicated regions where bulky groups would increase activity, whereas yellow contours indicated regions where bulky groups would decrease activity. As shown in Figure 2A  3D-QASR results indicated that introduction of small and electron withdrawing groups at 2-position of aromatic ring could largely improve the activity. Bulky groups at 3-and 4-position of the aromatic ring played a favorable role in improving the activity.

Binding Sites of M6, M15 and Ribavirin to TMV CP
In order to study the interactions between target compounds and TMV CP, the MST analysis method was used. The MST results indicated that the compounds M 6 , M 15 , and Ribavirin binding to TMV CP protein yielded Kd values of 31.1 ± 1.83, 346 ± 23.9, and 511 ± 33.1 μM, respectively ( Figure 3). As predicted in MST, M 6 shares, indeed, moderate affinity, in the contrary to M 15 and Ribavirin, which share weak affinity. The results showed that the combining capacity in the order of M 6 > M 15 > Ribavirin is consistent with the trend of antiviral activity screening. The experimental results showed that the compound M 6 may interactions with the TMV CP. As shown in the Figure 3 and Table 4, bulky groups at 3-and 4-position of aromatic ring played a favorable role in improving the activity. The bioactivity was mainly determined by electrostatic interactions. As shown in Figure 3 and Table 4

Instruments
1 H-NMR and 13 C-NMR spectra were obtained at 500 MHz using a JEOL-ECX500 NMR spectrometer at room temperature using tetramethylsilane as an internal standard (solvent CDCl3). Elemental analysis was performed on an Elementar Vario-III CHN analyzer. The melting points of the products were determined under an XT-4 binocular microscope (Beijing Tech. Instrument Co., Beijing, China) and left untouched. Analytical thin-layer chromatography (TLC) was conducted on a silica gel GF254 (400 mesh). Column chromatographic operations were performed on silica gel (200-300 mesh). Tobacco seeds were provided by the Guizhou Institute of Tobacco.

General Procedure for Preparation of Intermediates Z 1-21
A mixture of 4 (4.95 mmol) and K2CO3 (9.99 mmol) in DMF (10 mL) was stirred at 80 °C for 1 h. Then, 1,2-dibromoethane (14.84 mmol) was added dropwise. The mixture was stirred until the TLC showed the reaction finished. The reaction mixture was poured into water (20 mL). The mixture was extracted with EtOAc (3 × 10 mL) and the combined organic layers were dried over anhydrous Na2SO4 and concentrated under reduced pressure. The crude product was purified by column chromatography over silicagel by using petroleum ether and ethyl acetate (v/v = 2:1) as eluent to give Z 1-21 as a solid.

General Procedure for Preparation of Title Compounds (M 1 -M 21 )
The target compounds M 1 -M 21 were synthesized as schematized in Scheme 1. A 25 mL round-bottomed flask equipped with a magnetic stirrer was charged with intermediates 3 (0.74 mmol), KOH (0.89 mmol), DMF (5 mL). The flask was stirred at 40 °C for 1 h, and then Z (0.59 mmol) in DMF (2 mL) was dropwise into the flask. The resulting mixture was stirred at 40 °C for 8 to 10 h. TLC monitored the progress of the reaction. Upon completion of the reaction (as indicated by TLC), the reaction mixture was poured into saturated brine, the solid was filtered off and then dissolved with dichloromethane and washed by 10% KOH. The organic fraction was evaporated under reduced pressure. The solid was recrystallized from ethyl acetate/petroleum ether (3:1, v/v) to obtain the title compounds M 1 -M 21 with the yields from 56% to 78%. The physical characteristics, IR, 1 H-NMR, 13 C-NMR, and elemental analysis data, for all the synthesized compounds were reported in Supplementary data and the representative data of M 6 were shown below.  13

Purification of TMV
Using Gooding's method [21], the upper leaves of N. tabacum cv. K326 inoculated with TMV were selected, ground in phosphate buffer, and then filtered through a double-layer pledget. The filtrate was centrifuged at 10,000 g, treated twice with PEG, and then centrifuged again. The entire experiment was conducted at 4 °C. Absorbance values were estimated at 260 nm using an ultraviolet spectrophotometer.

Curative Effects of the Target Compounds against TMV in Vivo
Growing 5-6-leaf stage Nicotiana tabacum L. tobaccos were selected. TMV (concentration of 6 × 10 −3 mg/mL) was dipped and inoculated using a brush on the whole leaves, which were previously then dried. The compound solution was smeared on the left side of the leaves, scattered with silicon carbide. The leaves were then washed with water after inoculation for 0.5 h and the solvent was smeared on the right side for control. All plants were cultivated in an incubator at a temperature of 23 ± 1 °C and an illumination of 10,000 Lux. The number of local lesions was counted and recorded 3 to 4 days after inoculation. Three repetitions were conducted for each compound [27].

Protection Effects of the Target Compounds against TMV in Vivo
The compound solution was smeared on the left side, whereas the solvent was smeared on the right side of Nicotiana tabacum L. leaves of the same age to serve as the control. The leaves were inoculated with the virus after 12 h. A brush was dipped in 6 × 10 −3 mg/mL TMV to inoculate the leaves which were previously scattered with silicon carbide. Subsequently, the leaves were washed with water and rubbed softly along the nervature once or twice. All plants were cultivated in an incubator at a temperature of 23 ± 1 °C and an illumination of 10,000 Lux. The number of local lesions was counted and recorded 3 to 4 days after inoculation. Three repetitions were conducted for each compound [27].

Inactivation Effects of the Target Compounds against TMV in Vivo
The virus was inhibited by mixing with the compound solution at the same volume for 30 min. The mixture was then inoculated on Nicotiana tabacum L. leaves, and the right side of the leaves was inoculated with solvent and virus mixture for control. All of the leaves were previously scattered with silicon carbide. All plants were cultivated in an incubator at a temperature of 23 ± 1 °C and an illumination of 10,000 Lux. The number of local lesions was counted and recorded 3 to 4 days after inoculation. Three repetitions were conducted for each compound [27].
The inhibitory rate of the compounds was calculated according to the following formula ("av" denotes average):

3D-QSAR Study
The protection activity used in study was expressed as pEC50 listed in Table 2, 18 molecules of total compounds were randomly chosen as the training set for CoMFA and the other three compounds (asterisk labeled) were used as the testing set.

Molecular Modeling and Alignment
Molecular modeling, CoMFA analysis was performed using Sybyl 7.3 (Tripos Inc., St. Louis, MO, USA) software. The 3D structures of all molecules were built using the "Sketch Molecule" function in Sybyl. Initial optimization of the structures were carried out using the Gasteiger-Hückel charge, Tripos force field, and Powell conjugate gradient algorithm with a convergence criterion of 0.005 kcal/mol·Å [28]. The 3D structures of the 21 molecules were aligned on a common template molecule with M 6 for the CoMFA modeling study.

Partial Least-Squares Analysis
The partial least squares (PLS) analysis was used to derive the 3D-QSAR models. In which, molecules were placed in a rectangular grid, the steric and electrostatic fields were calculated using a volume-dependent lattice with a 2.0 Å grid spacing [29], and the CoMFA descriptors was used as the independent variables, and the experimental pEC50 values were presented as the dependent variables. Then, 3D-QSAR analysis was carried out using the PLS technique. The cross-validation and the ONC were used to evaluate the performance of the models, ONC was determined with the highest cross-validated q 2 [30,31]. Then, the non-cross-validated correlation coefficient r 2 value, standard error of estimate (SEE), and F value and standard error were calculated according to the definitions in Sybyl 7.3 package, and as factors for estimating. The contour maps and standard deviations values of CoMFA was generated by the PLS coefficients.

MST Studies
TMV CP was purified according to a previously described method [24]. A range of concentrations of the required compounds (range from 0.1 to 2 mM) were incubated with 0.1 mM of purified recombinant TMV CP for 5 min with the Monolith NT Protein Labeling Kit Red (Nano Temper Technologies, München, Germany) in assay buffer (10 mM Tris/HCl and 100 mM sodium chloride, pH 7.4). The sample was loaded into the NanoTemper glass capillaries and microthermophoresis carried out using 50% LED power and 40% MST. The Kd values were calculated from the duplicate reads of three separate experiments using the mass action equation in the NanoTemper software [32].

Conclusions
In summary, a series of 4-thioquinazoline derivatives containing chalcone moiety were prepared and evaluated for their antiviral activities against TMV using half-leaf method in vivo. Bioassay results indicated that compounds M 2 and M 6 possessed appreciable protection activities against TMV in vivo, with the EC50 values of 156.4 and 138.1 μg/mL, respectively, which were superior to that of Ribavirin (436.0 μg/mL). Meanwhile, the CoMFA model was generated base on the protection activities against TMV and exhibited good predictive abilities with the cross-validated q 2 and non-cross-validated r 2 values of 0.674 and 0.993, respectively. The MST experimental showed that the compound M 6 may interaction with the TMV CP. The model provided a practical tool for the modification and optimization of 4-thioquinazoline derivatives containing chalcone moiety to further improve the antiviral activity.