Antimicrobial, antioxidant and cytotoxic evaluation of diazenyl chalcones along with insights to mechanism of interaction by molecular docking studies

Background In continuation of our work, new diazenyl chalcones scaffolds (C-18 to C-27) were efficiently synthesized from substituted acetophenone azo dyes (A–E) by base catalyzed Claisen–Schmidt condensation with different substituted aromatic/heteroaromatic aldehydes. Methodology The synthesized chalcones were assessed for their in vitro antimicrobial potential towards several pathogenic microbial strains by tube dilution method and further evaluated for antioxidant potential by DPPH assay. These derivatives were also assessed for the cytotoxicity towards the human lung cancer cell line (A549) and normal cell line (HEK) by MTT assay. The most active antimicrobial compounds were docked using Schrodinger v18.1 software with the various potential bacterial receptors to explore the mechanism of interaction. Results The derivative C-22 exhibited high antibacterial activity with very low MIC (1.95–3.90 µg ml−1) and MBC (3.90–7.81 µg ml−1) values. The derivatives C-23, C-24 and C-27 have demonstrated good antioxidant potential (IC50 = 7–18 µg ml−1) correlated to the ascorbic acid (IC50 = 4.45 µg ml−1). The derivative C-25 had shown comparable cytotoxicity to camptothecin against A549 cell line. The docking studies predicted the bacterial dihydrofolate reductase (PDB ID: 3SRW) and bacterial DNA gyrase (PDB ID: 4ZVI) as the possible targets for most of the active antimicrobial compounds. These derivatives affirmed their safety by presenting less cytotoxicity towards HEK cells. Further the ADME prediction by qikprop module of the Schrodinger proved that these compounds exhibited drug-like attributes. Conclusion Hence, these compounds have shown their potential as lead for future expansion of novel antimicrobial and cytotoxic drugs. Electronic supplementary material The online version of this article (10.1186/s13065-019-0596-5) contains supplementary material, which is available to authorized users.


Introduction
The use of antimicrobial agents in clinical practice is one of the remarkable achievements of modern medicine. Several new antimicrobials became available to treat many infectious diseases in the last half of the 20th century. Also, due to the wide range of these antimicrobials, the clinicians are provided with a wide scope of choices to treat different infectious diseases [1,2]. It is significant from the drop in death rate, i.e. from 797 per lakh to 36 per lakh humans, a decrease by a factor of greater than 20 during the period spanning from 1900 to 1980. It is a clear cut proof of the efficacy and testimony of these antibiotics. However, the microorganisms have developed resistance due to the irrational use of antimicrobials over a period of time. This is evident from the rise of drugresistant microbial pathogens like methicillin-resistant Staphylococcus aureus (MRSA), vancomycin resistant enterococci, and multi-drug resistant tuberculosis [3][4][5]. Microbial resistance has been emerged as the most pressing health issue and has become a significant challenge as far as the delivery of healthcare is concerned. The emergence of new strains of microorganisms is also contributing to the observed resistance to the antimicrobial compounds [6,7]. This spread of resistance has thus restricted the treatment alternatives for some severe and life-threatening infectious disorders [8,9]. Therefore, in parallel to the discovery and expansion of new antimicrobial drugs, there is also a requirement for the proper use of current antimicrobials in consideration to decrease the spread of antimicrobial resistance. As there is an urgency for new antimicrobial agents, the current scenario is that many pharmaceutical companies are opting out of the antimicrobial section, especially antibacterial drug development [10]. The present situation has been further worsened as other malignant diseases such as cancer in which the host has immune-compromised or concomitant illness frequently accompanied by the microbial infections [11,12]. The chances of infections in cancer patients are higher in comparison to others due to interrupted epithelial barriers, decrease in neutrophils, compromised host defence, and changes in the microbial flora, etc. [13,14]. Among cancers, lung cancer is the utmost prevalent form and the most common cause of death in men and women worldwide. Non-small cell lung cancer (NSCLC) cells are inherently resistant to some commonly used cytotoxic drugs; however, small-cell lung cancer (SCLC) cells can develop resistance on continuous use of the drugs. Most of the patients with lung cancer have already developed metastatic disease, at the time of diagnosis, hence limiting the use of another therapeutic option, such as radiation and surgery. The current drugs used for the treatment of lung cancers have also developed resistance, hence conferred the limited treatment scope [15,16]. All the above facts necessitate the need to understand the basic mechanism of drug interaction with the possible targets on the molecular level for the development of new agents having antimicrobial and cytotoxic potential.
Chalcone derivatives, the main class of compounds, have gained immense interest from bioorganic and medicinal chemistry research. The chalcones are characterized by three carbon α, β-unsaturated carbonyl system joined by two aromatic rings. Chalcones also constitute an important class of natural products having considerable pharmacological potential [17,18]. The various biological activities like antimicrobial, antibacterial, antifungal, anti-inflammatory, antimalarial, antileishmanial, antioxidant and antitubercular etc. have been reported for the compounds comprising of chalcone backbone. The antimicrobial property of the chalcones is generally correlated with the reactive α, β-unsaturated keto function in the molecule [19,20].
Azo dyes are the most widely used class of colouring materials owing to various applications in various fields such as dying textile fibres, the colouring of different materials, biomedical studies and in advanced organic synthesis [21,22]. Azo compounds and their derivatives are also known for their use as antibacterial, antifungal, antidiabetics, antineoplastics, anti-inflammatory, antiseptic and other useful chemotherapeutic agents [23,24]. Several azo compounds particularly synthesized from β-naphthol, m-resorcinol, tyrosine, aspirin, paracetamol etc. have been frequently reported and exhibited impressive biocidal effects [25,26].
In the light of above facts, herein, we further extend our synthesis to some novel diazenyl chalcones and evaluation of their antimicrobial, antioxidant and cytotoxic potential against lung cancer cell line (A549) and safety study against normal cell line (HEK 293). To further explore their antimicrobial mechanism, we dock the active antimicrobial compounds from this series and the already synthesized compounds by Kaur and Narasimhan [27] against several potential receptors in bacteria.

Chemistry
The synthesis of diazenyl chalcones (C18-C27) has been presented in Scheme 1. The anilines with mono or di-substitution (generally with chloro and nitro groups) were diazotized in the presence of HCl and NaNO 2 , then coupled with o, p and m hydroxy substituted acetophenone derivative in the presence of ethanolic alkaline solution to give azo dyes (A-D). The azo dye E was synthesized by diazotization of p-aminoacetophenone followed by coupling with resorcinol in ethanolic solution. The diazenyl chalcones (C18-C27 were synthesized by the Claisen-Schmidt condensation of the α-methyl group (adjacent to keto group) present in azo dye with aldehyde group present in various reactants in the presence of alkali solution (Scheme 1). The target derivatives structures were confirmed by UV-Vis, FTIR, mass spectroscopy, NMR spectroscopy, and elemental analysis.

Characterization by spectroscopy
The FTIR spectra of dyes (A-E) and diazenyl chalcones (C18-C27) were recorded by the KBr pellet method. The aliphatic C-H stretch in dyes and diazenyl chalcones were observed at 3100-3035 cm −1 . The -C=C-stretching vibration was assigned to bands at 1615-1565 cm −1 . A broad peak was noticed for the phenolic group at 3450-3150 cm −1 . The carbonyl stretching vibrations for the enones was found between 1640 and 1685 cm −1 . The C-O stretching vibration was assigned in the range of 1300-1340 cm −1 . The -N=N-linkage was observed at 1400-1471 cm −1 . Another peaks noticed were Ar-O stretching Substituted aniline derivatives at 1120-1290 cm −1 , -C=C-bending vibration at 670-780 cm −1 , the C-N stretching at 1050-1360 cm −1 . Two strong bands at 1520-1530 cm −1 and 1335-1350 cm −1 attested the existence of NO 2 stretch. The band observed at 1050-550 cm −1 affirmed the absorption of C-X (halogen). The NMR spectra ( 1 H NMR and 13 C NMR) of the derivatives (C18-C27) were recorded in deuterated DMSO. The protons of the double bond of the enone moiety in most of the derivatives appeared as doublets or multiplet in the range of 5.0-6.21 ppm and 5.67-6.72 ppm. A broad peak in the range of 10-14 Hz was observed for the proton of the phenolic group. The protons of the -OCH 3 group were found as a singlet in the range of 3.6 ppm. The other aromatic ring protons appeared in the range of 6.50-8.52 ppm. The carbon signals of carbonyl (C=O) carbon in different diazenyl chalcones came in the range of 202-190 ppm. The signal of the hydroxy substituted carbon atom was noticed at 156-165 ppm. The carbon signals of other aromatic carbons were observed in the range of 135-110 ppm depending on the nature of substituents. The methoxy and methyl carbons were detected at 56 ppm and 23-27 ppm, respectively. The proton and carbon spectra of selected compounds have been provided in Additional file 1. The mass spectroscopy was done to confirm the molecular mass of the synthesized compounds. The % of carbon, oxygen nitrogen and hydrogen in the diazenyl chalcones was detected within the marked limits.

Antioxidant evaluation
DPPH scavenging activity results displayed that title compounds showed very low to high antioxidant activity and presented in Table 3. The % inhibition of DPPH was plotted against the logarithmic values of the concentration of test samples and the standard, to find the IC 50 (µg ml −1 ), which is the quantity of compound required to inhibit the absorbance of free radical DPPH by 50%. These graphs are presented in Fig. 1. Among all, the derivative C-27 had shown the highest antioxidant activity with IC 50 of 7.12 µg ml −1 comparable to the ascorbic acid (IC 50 = 4.45 µg ml −1 ). The other derivatives C-23 and C-24 had also shown the good antioxidant activity with the IC 50 of 15-18 µg ml −1 as compared to the controls (dyes and chalcone derivative). The most of other tested derivatives exhibited less antioxidant activity by presenting IC 50 > 100 µg ml −1 .
The DPPH assay results indicated that the compounds with the substituted resorcinol ring exhibited the highest antioxidant activity.

Cytotoxicity results
The cytotoxicity of the synthesized diazenyl chalcone derivatives (C-19, C-22, C-23, C-24, C-25 and C-27) was evaluated against A-549 cell line by MTT assay using reference drug camptothecin (CPT). The IC 50 values were calculated from the cell viability graphs and presented in Table 4. The cytotoxicity results revealed that the derivative C-25 exhibited good cytotoxic potential having IC 50 of 18.01 µg ml −1 towards A549 cell line in comparison to the standard drug (IC 50 = 8.7 µg ml −1 ). Figure 2 indicated that the diazenyl chalcone C-25 had reduced the number and clumping of A549 cells to a significant extent. The other tested derivatives were found to be less active against tested lung cancer cell line with IC 50 > 100 µg ml −1 . The compounds were also evaluated for the possible cytotoxicity against normal cell line (HEK) by MTT assay. The tested compounds revealed their safety by exhibiting low cytotoxicity towards HEK cell line with IC 50 > 100 µg ml −1 . The test derivative C-25 presented the higher selectivity index by exhibiting IC 50 ratio of normal cell line (HEK293) to carcinoma cell line 15 times as compared to the camptothecin.

Molecular docking results
With the view to elucidate the mode of interaction of active compounds with the bacterial targets, the active compounds from this series and from already synthesized series (structures presented in Fig. 3) [27] were Table 2 [32]. The GLIDE module was used to carry out the molecular docking study, and the findings were examined based on glide energy, and docking score and presented in Tables 5 and 6 respectively. The obtained docking poses were examined visually, and the interactions of the molecules with the residues of the binding pocket were studied with the help of ligand-interactions diagrams. The docking scores have been presented in terms of negative values, lower the docking score and glide energy, best would be the binding affinity. The docking scores of the test derivatives and standard drugs were analyzed for the various bacterial protein targets, and the majority of the active compounds showed high docking scores against dihydrofolate reductase enzyme in comparison to trimethoprim. The highest docking score and binding energy was shown by the deriva- The docking poses were visualized for the active derivatives, and it was noticed that the H-bonding was the most predominant interactions in the active derivatives. The derivative C-8 exhibited three hydrogen bonds through -OH groups with the essential residues: Ala8 (2.24 Å), Phe93 (1.65 Å), Leu29 (2.16 Å) of the binding pocket of dihydrofolate reductase (PDB ID: 3SRW). The π-π stacking was also observed for the aromatic ring with the Phe93 (4.33 Å) residue. Similarly, C-6 derivative exhibited four hydrogen bonds through the -OH groups with the essential residues Leu21 (2.28 Å), Glu20 (1.59 Å), Ala8 (2.29 Å), Phe93 (2.06 Å) while the standard drug trimethoprim exhibited three H-bonds, two donor bonds through amino groups with the Asp28 (2.48 Å) and Ala8 (1.90 Å) residues and one acceptor bond through the pyridyl nitrogen with the Ala8 (2.09 Å) residue of the binding pocket. The ligand interaction diagram and interacting residues of the binding pocket of dihydrofolate reductase with the active derivatives and standard drug trimethoprim had been shown in Fig. 4 and Table 7, respectively. The derivative C-22 had shown two H-bonds, one through -OH groups with the essential residues Glu50 (2.18 Å) and another H-bond through the > C=O of the enone moiety with the Asn (2.00 Å) of the binding cavity of the DNA gyrase (PDB ID: 4ZVI). The derivative C-22 also exhibited one salt bridge (3.27 Å) through the NO 2 group with the Asp49 residue. The standard drug ciprofloxacin displayed two H-bonds one through -OH group with the Asp73 (1.82 Å) and one through the piperazine nitrogen with the Glu50 (2.19 Å) and exhibited one salt bridge (4.48 Å) through the piperazine nitrogen with the Glu50 presented in Fig. 5. The interaction of the active derivatives with the vital residues of the catalytic pocket of the DNA gyrase has been presented in Table 8. UDP-N-acetylmuramoyl-l-alanine:dglutamate ligase (MurD) is one of the crucial enzymes which participates in the peptidoglycan biosynthesis of the cell wall and hence presents a possible target to combat bacterial drug-resistance in search of new antibacterial agents [33]. The C-21 had shown H-bonding interaction with the essential residues Asn421 (2.23 Å), Lys420   Fig. 6 and Table 9. Hence, the synthesized chalcones have shown their potential as dihydrofolate reductase inhibitors, DNA gyrase inhibitor and the cell wall protein synthesis inhibitors.

ADME properties
Most of the newly designed molecules having selectivity and optimized binding affinity for their receptors generally failed during the clinical trials due to deficient pharmacokinetic parameters. Therefore, the screening of promising molecules should not only be restricted to their improved selectivity and increased binding affinity with their targets; but also require significant consideration to the pharmacokinetic parameters such as lipophilicity, blood-brain barrier coefficient, oral bioavailability etc. The QikProp, version 3.5, Schrödinger was used for the prediction of the ADME parameters in comparison to those of 95% known drugs. Total fortyfour significant descriptors were predicted for the substituted diazenyl chalcones. The selected descriptors critical for defining the drug-like properties of the synthesized compounds have been reported in Table 10. The Lipinski rule is an important filter which predicts the druggability of the compounds. This rule has been followed by most of the compounds with MW < 500 Da, H-bond acceptor < 10 and donor < 5 and octanol/water partition coefficient (QPlogPo/w) < 5, The synthesized derivatives exhibited oral absorption in the range of 80-100%.

Conclusion
In this investigation, novel diazenyl chalcones (C18-C27) with various substituted aromatic or heteroaromatic rings were effectively synthesized. These chalcone analogues were evaluated for their in vitro antimicrobial, antioxidant and cytotoxic potential. The results revealed that the chalcone derivative C-22 was the most active antibacterial agent among all the derivatives by exhibiting very low MIC and MBC. The derivatives with substituted resorcinol ring had shown good antioxidant potential. The derivative C-25 had shown

Materials and methods
The procurement of all the chemicals/reagents required for the study was done from Merck Chemicals (India) and Hi-Media Laboratories (India). The required strains of microorganisms were procured from IMTECH, Chandigarh. The progress of the chemical reaction was monitored using TLC performed with pre-coated plates of silica gel (60 F 254 ). The purification of the diazenyl chalcones was done by recrystallization techniques and chromatography by column composed of silica gel of mesh size 100-200, using solvents ethyl acetate and hexane. The infrared spectroscopy (Bruker 12060280, the spectrophotometer) was performed to confirm the presence of functional groups in diazenyl chalcones. The NMR spectroscopy (Bruker Avance II 400 NMR spectrometer), namely 1 H NMR and 13 C NMR was carried out to estimate the carbon/proton signals, in DMSO (deuterated) at a frequency of 400 MHz and 100 MHz, respectively. The % composition of hydrogen, carbon, nitrogen, oxygen and sulphur of the synthesized chalcones was estimated on the elemental analyzer (Flash EA1112N series, Thermo Finnigan, Italy). The mass spectroscopy was carried out on the Advion expression CMS, USA mass spectrometer to confirm the molecular mass of the synthesized compounds.

Synthesis of diazenyl chalcones (C18-C27)
A solution (10 ml) of substituted anilines (0.01 M) was prepared in 0.2 N HCl. Subsequently, a cold solution of NaNO 2 (0.01 M) in H 2 O over 10-15 min at 0 °C to form the diazonium salt solution. This solution was stirred at 4-5 °C for half an hour to complete the diazotization and then added dropwise to the ice-cold solution of hydroxy substituted acetophenone derivative in ethanol at 0-5 °C in 30 min. The solution was additionally stirred for 1 h at 4-5 °C. Afterwards, NaOH (10%) solution was added dropwise to neutralize the acid for the precipitation of the azo dyes (A-D). The azo dye E was synthesized in the same manner by diazotization of p-aminoacetophenone followed by coupling with resorcinol in an alkaline solution. The resultant precipitates of azo dyes were filtered, air-dried and recrystallized from ethanol. These azo dyes (A-E) were used for the synthesis of diazenyl chalcones (C18-C27) by reaction with various mono or di-substituted aromatic/heteroaromatic aldehydes in the presence of the catalytic amount of alkali (Scheme 1). In a 250 ml conical flask, 0.001 M of aldehyde and 0.001 M of azo dye were dissolved in the presence of ethanol and stirred vigorously on a magnetic stirrer. After 30 min, 2 ml of 10% alcoholic KOH was added dropwise to the reaction mixture with rapid stirring keeping the reaction temperature maintained at 25-30 °C with continuous stirring for a period of 18-24 h. The progress of the reaction was monitored by TLC. Afterwards, the contents  of the flask were transferred into the ice-chilled water and neutralized by 0.1-0.2 N HCl, resulting in the precipitation of diazenyl chalcone derivatives. The precipitates obtained were air dried after filtration, and subject to purification by recrystallization from ethanol and by column chromatography (solvent system of 5-25% ethyl acetate and hexane) [27].  Table 7 The interaction of highly active compounds with binding pocket of bacterial dihydrofolate reductase (PDB ID 3SRW)

Antioxidant evaluation
The derivatives (C18-C27) were assessed for their antioxidant potential in the presence of the stable free radical DPPH using ascorbic acid as the standard compound [36,37]. The assay was performed in 96 well plates using ELISA reader with samples aliquots at a series of concen

Cytotoxicity evaluation
The cell lines employed in the investigation were purchased from the National Centre for Cell Sciences (NCCS), Pune, India. The cell lines were cultured in DMEM fortified with 10% FBS, l-glutamine, sodium bicarbonate, and solution of antibiotics (streptomycin 100 μg ml −1 + penicillin-100 U ml −1 ). The culture of cell lines was stored under 5% CO 2 incubation at 37 °C.

Cell proliferation study by MTT assay
The compounds showing maximum activity (antimicrobial and antioxidant potential) were screened for cytotoxicity against A549 and HEK cell lines using MTT assay [38,39]. Initially, 96-well microculture plate was seeded with 1 × 10 4 cells in 100 µl/well DMEM/MEM, supplemented with 10% FBS followed by incubation at a temperature of 37 °C for a period of 24 h under 5% CO 2 atmosphere. After that, the test compounds and camptothecin were added to the cells to achieve a concentration of 10, 25, 50 and 100 µg. MTT reagent (10 µl) of concentration of 5 mg ml −1 added to all the wells after the duration of 48 h. The plate was incubated further for 4 h followed by careful removal of supernatant from each well. To dissolve the formazan crystals, 100 µl of DMSO was added, and the ELISA reader was used to check the % decolouration of DPPH = Control Abs − Sample Abs × 100/Control Abs absorbance at 570 nm wavelength [40]. In order to calculate the IC 50 concentration, the following equation, i.e. A = Bx + C, was used. In this case, A = 50, while the values B and C values were obtained from the survival curve plot. This study was carried out in duplicate.

Molecular docking
The active antimicrobial compounds found from the previous series synthesized by Kaur and Narasimhan (C-6, C-7, C-8, C-15, C-16, C-17) [27] and this series (C-20, C-21, C-22, C-23, C-24, C-25 and C-27) were docked for the various potential bacterial targets such as dihydrofolate reductase, dihydropteroate synthase, cell wall synthesis proteins, DNA gyrase and topoisomerase using Schrodinger Glide software [41]. The 3-dimensional crystal structures of all the proteins such as S. aureus dihydrofolate reductase (PDB ID: 3SRW, resolution 1.7 Å), E. coli gyrase B (PDB ID: 4ZVI, resolution 2.2 Å), E. coli UDP-N-acetylmuramoyl-l-alanine:d-glutamate ligase (PDB ID: 1UAG, resolution 1.95 Å), E. coli topoisomerase IV (PDB ID: 3FV5, resolution 1.8 Å), and dihydropteroate synthase from Streptococcus pneumonia ((PDB ID: 3FV5, resolution 2.4 Å) were used for the molecular modelling and were accessed from the website of Protein Data Bank (http://www.rcsb.org/pdb/home/home) [42]. The derivatives under investigation were studied for the mode of theoretical binding to comprehend the possible intermolecular interactions between the ligand and the receptor. The required protein structures were prepared (preprocessed, optimized and minimized) by the Protein Preparation Wizard available in the Schrodinger software graphical user interface Maestro v11. 5. Crystallographic H 2 O molecules with one or two H-bonds were omitted. To set the pH of the protein to a value of 7.0, hydrogen atoms were added to the structure. To attain the RMSD cut off 0.30 Å, the restrained minimization of the heavy atoms was performed. The ligands (data set) were prepared using the LigPrep module of Schrodinger v11.5. An active site encompassing a radius of 20 Å was defined around the ligand in the crystal structure of the proteins. Also, a grid box was created around the centroid of the defined active site. All the ligands with their low-energy conformations were docked into different proteins in their respective catalytic pockets using extra precision mode (Glide, Schrödinger 2018-1) in the absence of constraints. The structures showing the best results were selected based on docking scores and binding energies [43,44].

ADME prediction
The most prevalent reason for the failure of drugs in clinical trial phase is the lack of knowledge regarding its