Synthesis of some novel coumarin isoxazol sulfonamide hybrid compounds, 3D-QSAR studies, and antibacterial evaluation

With the progressive and ever-increasing antibacterial resistance pathway, the need for novel antibiotic design becomes critical. Sulfonamides are one of the more effective antibiotics against bacteria. In this work, several novel sulfonamide hybrids including coumarin and isoxazole group were synthesized in five steps starting from coumarin-3-carboxylic acid and 3-amino-5-methyl isoxazole and assayed for antibacterial activity. The samples were obtained in good to high yield and characterized by FT-IR, 13C-NMR, 1H-NMR, CHN and melting point techniques. 3D-QSAR is a fast, easy, cost-effective, and high throughput screening method to predict the effect of the compound's efficacy, which notably decreases the needed price for experimental drug assay. The 3D-QSAR model displayed acceptable predictive and descriptive capability to find r2 and q2 the pMIC of the designed compound. Key descriptors, which robustly depend on antibacterial activity, perhaps were explained by this method. According to this model, among the synthesized sulfonamide hybrids, 9b and 9f had the highest effect on the gram-negative and gram-positive bacteria based on the pMIC. The 3D-QSAR results were confirmed in the experimental assays, demonstrating that our model is useful for developing new antibacterial agents. The work proposes a computationally-driven strategy for designing and discovering new sulfonamide scaffold for bacterial inhibition.

Due to the biological activity of isoxazoles and sulfonamides, several sulfonamides bearing isoxazole moiety have been designed and synthesized as pharmacological active compounds. Valdecoxib, as an isoxazole sulfonamide hybrid drug, demonstrated as a selective COX-2 inhibitor 31 . Zonisamide was showed anticonvulsant and antiobesity properties 32,33 . Sulfamethoxazole and sulfisoxazole were known as antibacterial isoxazole sulfonamide hybrid drugs 34 (Fig. 2).
Isoxazole coumarin is an important hybrid structure that is an essential and integral part of various therapeutic scaffolds. Recently, Khaled et al., was studied the antioxidant activity of several coumarin isoxazole derivatives using DPPH as radical scavenging 35 . They found that the target compounds (6a-d) have broad range of activities in comparison with silymarin which among them the methoxy derivative (6d) was the most potent one. Also, several antibacterial coumarin-based isoxazoles (7a-d) were synthesized by Suresh et al. 36 . In another work, isoxazoles connected 6-hydroxycoumarin were synthesized and studied their cytotoxic activity 37 . The best and selective activity was observed for compound 8 and 9 against prostate (PC-3) cancer cell line (Fig. 3).
Novel drug introducing process to market entail significant time and money investment. A drug's development and introduction into the market are estimated to spend about 2.6 billion dollars and the price has risen by almost 150 percent in the last ten years. However, the failure rate has risen to almost 90% 38 . There are a divers of computer-aided drug discovery (CADD) methods that can be used to create bioactive molecules. Based on the molecular interaction nature, Three-dimensional quantitative structure-activity relationship (3D-QSAR) analyses. This calculation offers plenty of knowledge about the precise molecular features that prerequisite for biological activity and can be used as a major predictive method for pharmaceutical design. The relationship between the configuration of the ligand and its biological activity was quantified by using QSAR 39 .
In continuance of our works on the synthesis and computational studies of small biologically active compounds [40][41][42][43][44][45] , 3D-QSAR models were performed to predict novel coumarin Isoxazol sulfonamide hybrids possible MIC on the bacteria. Also, the antibacterial activity of the designed and synthesized sulfonamides hybrids was assessed. The predicted MIC and experimental MIC were compared and the accuracy of 3D-QSAR was discussed.  Average of actives analysis results. The identified fields along with shape and hydrophobic interactions depend on the high biological activity, and it stated new molecules which show negative or positive fields in the same or similar area that must be supposed as an active molecule Fig. 6.
Validation of the 3D-QSAR. Predicted versus experimental activity plots of the training and test set's molecules distinctly placed in the ( Figure S43). Both two models of the S. aureus and E. coli displayed that observed and estimated activity were highly correlated in the training and test set's compounds. Also, the compounds' predicted activity was remarkably approach to the experimental activity. The r 2 of training and test sets of all models were beyond 0.6 so all of these data verify that this model is the accurate and applicable to predict MIC of the compounds. All of these observations indicate that the model is capable of accurately predicting the pMIC of compounds.
The compounds predicted activity. All compounds predicted activity had an excellent result, especially about the which was/or were more effective than the other compounds Table 1.  Fig. 7.
To begin, cumarin-3-carboxylic acid chloride was prepared by the reaction of cumarin-3-carboxylic acid with SOCl 2 . In the second step, the obtained cumarin acid chloride was reacted with 3-amino-5-methyl isoxazole to produce the corresponding amide 7 in high yield and purity. In this step, using sodium hydrogen carbonate as base, solvent-free conditions and room temperature are showing that the reaction was carried out under green conditions. Then, chlorosulfonic acid was used for the chlorosulfonating of the obtained amide (7). The corresponding sulfonylchloride (8) was very pure and used directly in next step. Finally, the obtained Isoxazol coumarin sulfonylchloride (8) was combined with different amines to form the corresponding sulfonamides (9a-l). The reaction was carried out in the presence of a sodium hydrogen carbonate as a green base at ambient temperature in the absence of any solvent. All of the coumarin sulfonamide isoxazol hybrid compounds (9a-l) were obtained in high purity after an easy work-up, just by adding water and simple filtration. It means that this step as the same as the second step is completely in agreement with green chemistry. Several aromatic and www.nature.com/scientificreports/ aliphatic amines were tested to extend the synthetic utility of the procedure ( Table 2). As the results is shown, aromatic amines containing electron-donating and electron-withdrawing substituent and also aliphatic amines are reacted with sulfonyl chloride 3 in 10-30 min and the corresponding sulfonamides were obtained in 85-95% yield with great purity ( Figure S1-S41). In comparison, anilines with electron-donating groups such as methoxy and methyl reacted with sulfonyl chloride faster than those with the electron-withdrawing group such as Cl and Br and the products were obtained in higher yields.
Biological evaluation. The antibacterial activity of the synthesized sulfonamide base hybrids (9a-l) was investigated against E. coli and S. aureus by applying the agar diffusion method. As the results was shown, aromatic amines with no substituent (entry 2) and methyl substituent at the 4-position of phenyl ring (entry 6) were the most potent compounds against both E.coli and S. aureus strains (Table 3). Derivatives including halogen groups (Cl, Br) in para position (entry 1, 10), showed higher antibacterial activity against S. aureus than E coli. This indicated that compounds with electron-withdrawing or electron-donating groups at 4-position showed better inhibitory activities than those compounds with no substituent. Compounds with chlorin atom at orto position of phenyl ring (entry 8), could efficiently improve the antibacterial activity against E. coli but this substituent is not suitable against S. aureus. The introduction of methoxy group at the orto or meta positions of phenyl ring (entry 3, 4), led to a reduction in activity. Also, the replacement of aromatic amines with aliphatic amine (entry 12) improved the inhibitory activity against E. coli, while against S. aureus is not tolerable.
MIC. MIC of All the new synthesized compounds was assessed by microdilution broth. MIC results presented that most compounds are moderately active against bacteria except 9b and 9f that have the robust antibacterial effect on the E. coli and S. aureus. 9f had a consequential effect on the E. coli (Table 4).

Conclusion
The researched work focused on the field-based 3D QSAR model expansion on a series of coumarin isoxazol sulfonamide hybrids to investigate the mechanism of sulfonamide inhibition on the bacteria development of the new sulfonamides with enhanced power of the antibacterial effect. The elucidated action mechanism clarified the SAR and, thus, may be accelerate design and improve the structure of novel and potent sulfonamide ligand against bacteria. The structure study and chemical field analyses made it bright to assess which structure of sulfonamides block bacterial growth. The obtained outcomes also suggest potential insight for the area where the active molecules lie and also implicate the average active molecules shape. The positive and negative charges,  www.nature.com/scientificreports/ such as hydrophobic regions, were demonstrated. This utilized method would assist to consider the model predicted compounds and potential modifications assumption. These variables make fitted model to compounds and predict required changes to raise their bioactivity. The good r 2 , q 2 and RMSE values for the training and the CV training set (in both E. coli and S. aureus) gained. Accordingly, some coumarin 3-carbamide sulfonamides derived from 3-amino-5-methyl isoxazole were synthesized and evaluated for the antibacterial activity. The obtained products were characterized using melting point, 13 C NMR, 1 H NMR and FT-IR techniques. All the compounds were synthesized very pure with an easy procedure. Also, the in vitro antibacterial activities of the synthesized sulfonamides have been studied on the S. aureus and E. coli strains. The predicted activity of the compounds was near to the MIC activity and in accordance with the expectation (based on the predicted activity of bioactivity) 9b and 9f compounds had better antibacterial activity than the others, which made them potent inhibitors against bacteria. It seems like methyl substituent at the 4 positions of anilines (9f) and also derivatives without substituent (9b), create the most potency against both gram-negative and positive bacteria. Predicted activity in 3D-QSAR model successfully predicted the arrangement of the efficacy of the synthesized sulfonamide special about the high influence of the 9f and 9b compounds on the gram-negative and gram-positive bacteria. These two compounds had the more predicted activity than the other. On the other hand, our 3D-QSAR capability to predict the bioactivity of sulfonamides not only in the computational model was excellent but also in the in vitro experimental test had acceptable outcomes.

Methods
In silico assay. QSAR, 3D-QSAR data set. Data set was gathered against Dihydropteroate Synthase (CHEMBL4032, CHEMBL1075045) for sulfonamides from ChEMBL 20 database and articles 30,[46][47][48][49][50][51][52] . The database includes compounds that had the antibacterial effect and sulfonamides functional group. The initial dataset derived from MIC was selected for further investigation on the dataset. Compounds which had not the MIC were omitted 53 . Then 65, and 72 Compounds were gained for S. aureus and E. coli, respectively. All collected data set placed in the supplementary 2 file. 3D-QSAR; conformation hunt, pharmacophore generation, alignment, and built model calculations. The accumulated CSV format was transformed to SDF format. All articles extracted compounds were created in the Marvin Sketch program. Discovery Studio program 54 applied to minimize (CHARMm force field) and add hydrogen (in pH = 7) to the Marvin Sketch created compound. All collected compounds (compounds from ChEMBL database and articles) were gathered in the one SDF format dataset. The crystallize targeted protein (PDB = 3TYE) was extracted from the PDB site and separated into protein receptor and ligand references. The bound drug was utilized as a reference ligand to produce field pharmacophores and subsequently as a bioactive reference conformation, while the target receptor was used as a protein receptor excluding volume. The reference conformation was also annotated with its computed field points, which were generated in 3D field point pattern. Filed pharmacophore, bioactive reference conformation, target receptor and references drug were employed to generate GRID. XED (eXtended Electron Distribution) force field was applied to make field points. As a result, three divers molecular fields including 'Shape' (van der Waals), positive and negative electrostatic, and 'hydrophobic' fields (a density function associated with steric bulk and hydrophobicity) were computed. The filed points template provides a standardized description of the electrostatics, compound's shape, and hydrophobicity. Following that, the reference conformer was utilized to align the compounds in the training and test sets by using Maximum Common Substructure (MCS) and customized thresholds. The conformation hunt was carried out by using very accurate and slow calculation procedure with a maximum number of conformations produced for every molecule set to 1000. To create the 3D-QSAR model, the best fitting low energy conformations for the template were used. The compounds data set's experimental behavior was converted to positive logarithmic scale. E. coli and S. aureus pMIC, which is produced via formula of pMIC = −log (MIC), was used as an associate variable. In a ratio of 80 percent and 20%, total molecules were divided into the training and test  Figure 7. Synthetic route of the synthesis of coumarin sulfonamide isoxazol hybrids 9a-l. Validation of the 3D-QSAR. The correlation coefficient (r 2 ), the cross-validation regression coefficient q 2 and the root mean square error (RMSE) were considered to the 3D-QSAR model predictive potential 54 .

Visualization of SAR activity Atlas models.
To qualitatively visualize the training dataset, the Bayesian technique was used. Bayesian metod offered comprehensive interpretation of the electrostatic, hydrophobic and shape characteristics, and constitute selected compounds 'SAR. The three-dimensional analysis of these models will provide valuable detail. The activity atlas study offered a range of interrelated biochemical outcomes which comprising activity regions explore analyses and an average shape of active. The average of actives revealed the Experimental. All commercially chemicals were prepared from the Merck Chemical Company. The structure of synthesized compounds was confirmed using FT-IR spectroscopy (Nicolet 400D, KBr matrix, 4000-400 cm −1 ), melting points obtained with an Amstead Electro Thermal 9200 apparatus, and nuclear magnetic resonance ( 1 H NMR, 13 C NMR) spectra (Bruker DRX-400 Avance spectrometer at 400 and 100 MHz, respectively. The purity of products and also reaction progress evaluation were determined using thin layer chromatography on silica-gel Polygram SILG/UV 254 plates (Merck). (1). A mixture of SOCl 2 (4.8 mmol) and 2-oxo-2H-chromene-3-carboxylic acid (4 mmol) was stirred for 12 h at 75 ºC. After extra SOCl 2 distillation , high purity light yellow solid product was obtained; M.p = 158-160 °C (M.p Lit = 158-161 °C) 45 . (2). A mixture of synthesized chromene-3-carbonyl chloride (3 mmol), sodium hydrogen carbonate (NaHCO 3 , 3 mmol) and 3-amino-5-methyl isoxazole (3 mmol) was triturated at room temperature under solvent-free conditions. After completion of the reaction as it was monitored by TLC, 25 mL distilled water was added, stirred for 15 min, and filtrated. The very pure generated product directly employed for the next step.   www.nature.com/scientificreports/ Synthesis of 3-((5-methylisoxazol-3-yl) carbamoyl)-2-oxo-2H-chromene-6-sulfonyl chloride (3). Chlorosulfonic acid (20 mmol) was slowly added and stirred to compound 2 (2 mmol) at 0 ºC and for 0.5 h. The stirred process was continued for 1 h at room temperature and 8 h at 60 ºC. Then, the mixture was added drop wise to ice bath, stirred for 5 min, filtered and washed with water until neutralized. The product was obtained very pure and characterized by FT-IR, melting point, 13 12.6, 96.7, 116.6, 118.0, 119.5, 127.6, 132.4, 145.0,  148.8, 154.4, 157.6, 160.1, 160.8, 170.9.

Synthesis of N-(5-methylisoxazol-3-yl)-2-oxo-2H-chromene-3-carboxamide
General process for the Isoxazol sulfamoyl chromene carboxamide synthesis (4a-l). A mixture of 3-((5-methylisoxazol-3-yl) carbamoyl)-2-oxo-2H-chromene-6-sulfonyl chloride (3) (1 mmol), NaHCO 3 (1 mmol) and amine (1 mmol) was grinded in a mortar at room temperature. After completion of the reaction (monitored by TLC), the distilled water was added (25 mL), stirred for 5 min, filtered and washed with enough water until neutralized. The final product was dried and characterized by melting point, FT-IR, NMR and CHN analysis methods.    License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.