Design, synthesis, DFT, docking studies and ADME prediction of some new coumarinyl linked pyrazolylthiazoles: Potential standalone or adjuvant antimicrobial agents

The control of antimicrobial resistance (AMR) seems to have come to a dead end. The major consequences of the use and abuse of antibacterial drugs are the development of resistant strains due to genetic mutability of both pathogenic and nonpathogenic microorganisms. We, herein, report the synthesis, characterization and biological activities of coumarin-thiazole-pyrazole (CTP) molecular hybrids with an effort to explore and overcome the increasing antimicrobial resistance. The compounds were characterized by analyzing their IR, Mass, 1H and13C NMR spectral data and elemental analysis. The in vitro antimicrobial activity of the synthesized compounds was investigated against various pathogenic strains; the results obtained were further explained with the help of DFT and molecular orbital calculations. Compound 1b and 1f displayed good antimicrobial activity and synergistic effects when used with kanamycin and amphotericin B. Furthermore, in vitro cytotoxicity of compounds 1b and 1f were studied against HeLa cells (cervical cancer cell) and Hek-293 cells. The results of molecular docking study were used to better rationalize the action and prediction of the binding modes of these compounds.


Introduction
Over the past few decades, globally antibiotic repository has become less effective due to both antibiotic overuse and an increase in antimicrobial drug resistance. Even common bacterial (e.g. Staphylococcus aureus, Staphylococcus enteric, Escherichia coli) and fungal (e.g. Candida albicans, Cryptococcus neoformans) pathogens are developing resistance against widely prescribed antibacterial and antifungal drugs like amphotericin B (Amp B), fluconazole (FLC), PLOS  chloroamphenicol and penicillin [1][2][3].This is particularly dangerous considering that the major bacterial as well as fungal pathogens such as Candida sp. account for~75% of all infections and represent the 4 th leading cause of nosocomial infections [4]. Thus microbial drug resistance has emerged as a major public health challenge across the world. Consequently, strategies are needed to develop new therapeutic compounds that could either act as new drugs or could complement the existing drug therapy more effective by acting as an adjuvant [5][6][7]. In recent past, azoles have garnered considerable attention especially in the design and synthesis of compounds with significant biological activities [8][9][10][11]. In particular, there are various drugs incorporating thiazole nucleus (Fig 1) which are being used as the antimicrobial (sulfathiazole), antiinflammatory (fanetizole) and antifungal (abafungin) agents [12][13][14][15][16][17]. Similarly, compounds containing pyrazole ring have been found to exhibit a variety of biological properties and many of them such as celecoxib, pyrazofurin etc. (Fig 1) are used clinically besides using for agricultural pest management [18][19][20][21][22]. Especially, 3,5-dimethyl-4-aryldiazenylpyrazole derivatives have been reported to exhibit cytotoxic [23], antioxidant [24], analgesic [25] and significant antimicrobial potential [26]. On the other side, compounds containing coumarin moiety (Fig 1) have also been proved as a potent antimicrobial, antitumor, anticoagulant, antiviral, and anti-inflammatory agents [23,24,[27][28][29][30][31]. It has been widely accepted that structural properties of thiazole, pyrazole and coumarin derivatives and their stability under in vivo conditions are the primary factors for their superior pharmacological activities [32][33][34].
A drug discovery program involves the planning of novel chemical entities using molecular hybridization approach where two active pharmacophoric units of known drugs were connected through a covalent bond in a single matrix [35][36][37]. The selection of the active units in the dual drug is usually established on their observed (or anticipated) synergistic or additive pharmacological activities. In prospect of the above discussed facts and in continuation of our research towards the development of novel molecular hybrids as potential biologically active agents [38][39][40][41][42], we have decided to build new molecular hybrids based on coumarin-thiazolepyrazole (CTP).

Chemistry
General. Melting points were determined by open capillary method and are uncorrected. Infrared spectrum was obtained neat on a Thermo Scientific Fischer spectrometer. Multinuclear NMR ( 1 H, 13 C) spectra were recorded on a Bruker avance at 300 MHz. Chemical shifts are reported in parts per million (ppm). Tetramethylsilane served as an internal standard in 13 C and 1 H NMR (δ 0.00 ppm). CHNS analyses were obtained on a Perkin Elmer Model 2400 CHNS elemental analyzer and Mass spectroscopy data of synthesized compound was recorded on waters QQ-TOF micro Mass Spectrometer. All the chemicals used for the synthesis of target compounds have been purchased from Sigma Aldrich and were used as received. All the compounds gave C, H and N analysis within ± 0.5 of the theoretical values. Scanned Spectra for all compounds are shown in S3-S19 Figs.

Antimicrobial assay
The microbial strains (bacterial and fungal) used in the study were obtained from the Institute of Microbial Technology ( were cultured in yeast extract-peptone-dextrose (YEPD broth, HiMedia, India) and RPMI 1640 media (HiMedia, India). For agar plates, 2.5% (w/v) bacteriological agar (HiMedia, India) was added to the medium. The strains were stored with 15% glycerol at -80˚C as frozen stocks. The cells were freshly revived on respective agar plates from the stock before each experiment [43,44].

Antibiotic susceptibility testing
Antibacterial activity. All the bacterial strains (E. coli, S. aureus, E. faecalis, V. cholerae, S. pyogenes) were grown overnight and were diluted in Mueller-Hinton broth (MHB) to a cell density of 10 5 CFU/mL. 100 μL of this culture and compounds 1a-1g (250-0.112 μM) dissolved in DMSO, were added into the 96-well flat bottomed microtiter plate (Genexy-HiMedia, India). The plate was incubated at 37˚C without shaking for 24 h. The visual and optical density was measured using microplate reader at 600 nm (Thermo, Model 680). The minimum inhibitory concentration (MIC) was defined as the concentration of the antimicrobial agent that inhibit >99% growth. kanamycin (KAN), a known antibacterial drug is used as a positive control [45].
Antifungal activity. The antifungal activities of compounds 1a-1g against fungal species (C. albicans, C. glabrata, C. krusei, C. parapsilosis, C. keyfer, C. tropicalis) were performed according to the Clinical and Laboratory Standards Institute (CLSI, formerly NCCL) in RPMI-1640 medium by broth micro-dilution methods [46]. The concentrations of compounds 1a-1g ranged between 250 μM and 0.097 μM. The 96 well flat bottom microtiter plates were incubated without shaking at 30˚C for 48 h. The visual and optical density was measured using microplate reader at 492 nm (Thermo, Model 680) to study the inhibition in fungal growth. The MIC was defined as the lowest concentration of antifungal drug which resulted in >99% inhibition of growth compared to that of the untreated control. amphotericin B, a known antifungal drug is used as a positive control.
Checkerboard assay of compound 1b and 1f. The interaction of compound 1b and 1f with well-known antimicrobials kanamycin, and amphotericin B and fluconazole were evaluated by the checkerboard method as described previously and represented as the fractional inhibitory concentration (FIC) index (FICI) i.e. sum of the FIC for each compounds. MICs of drug X and Y when acting alone and C X comb and C Y comb are concentrations of drugs X and Y at the isoeffective combinations, respectively. The interaction was described as synergistic when the FICI was 0.5, and when > 4.0, it was interpreted as antagonist and any value in between as in different [45].

Scanning electron microscopy (SEM) of 1f treated S. aureus and C. keyfer cells.
The cell suspensions of S. aureus and C. keyfer in exponential phase were prepared in MHB and RPMI 1640 medium (pH 7), respectively. Compound 1f (15.67μM) was added to the S. aureus cells (~1 x 10 5 CFU/mL), whereas 1f (62.25 μM) was added to the C. keyfer (~1 x 10 4 CFU/mL) and incubated at 37˚C (S. aureus for 12 h) and 30˚C (C. keyfer for 12 h). After treatment the samples for SEM study were prepared as described previously and observed under the SEM (JEOL-6100F, JEOL, Japan) [47].

Computational methods
DFT studies. Khon-Sham's DFT method that undergo the gradient-corrected hybrid density functional B3LYP were used for all theoretical calculations [49]. The Becke's three parameters functional of Lee et al [50], is a combination of non-local exchange potential with the non-local correlation function. A full geometry optimization was performed for all structures, using this function [50] and the 6-31+g(d,p) bases set as implemented by Gaussian 03 package [51]. All geometries were visualized using Avogadro 1.2 software package.
Molecular docking. The protein-ligand crystal structure has been optimized to its lower energy conformation using protein preparation wizard where all the water molecules were removed during preprocess and missing side chains of residues have been added using prime. Keeping in mind the appropriate ionization states for the acidic as well as basic amino acids, hydrogen atoms were added to the protein structure corresponding to the physiological pH 7.0. Finally energy minimization with root-mean-square deviation (RMSD) value of 0.30 Å was carried out using optimized potentials for liquid simulations (OPLS-2005) force field after assigning charge and protonation state. Ligand preparation was done by using the Schrodinger LigPrep utility (Schrodinger, LLC, USA) which generates a low energy 3d structures.
The active site of protein was defined by a bounding box (grid) that was centered on the native ligand in the crystal complex. Extra-precision glide docking (Glide XP) which docks ligands flexibly was used to rank the docking poses and to gauze the binding affinity of these ligands toward the protein.
The 1 H NMR spectra of compounds (1a-1g) revealed two singlets in the range δ 8.39-8.29 and 8.00-8.10 ppm corresponds to coumarin 4-H and thiazole 5-H, respectively, thereby confirming the formation of 1. Two Singlets in the range δ 2.19-2.32 and 2.69-2.86 ppm appeared due to 3-CH 3 and 5-CH 3 group attached at pyrazole ring. It has been well established that 5-CH 3 is more deshielded compared to 3-CH 3 due to a lone pair of electrons present on the nitrogen atom of the thiazole nucleus and therefore appeared at about δ 2.80ppm [41,55]. Moreover, the appearance of other signals in the NMR spectra corresponding to aromatic protons and carbons furhter establish the structure of the synthesized compounds and complete assignment of 1 H NMR along with other spectral data has been given in experimental section.

Biological evaluation
Antimicrobial activity. The antimicrobial activity of compounds (1a-1g) was evaluated in vitro according to Clinical and Laboratory Standard Institute (CLSI) guidelines using five clinical bacterial strains (Escherichia coli, Staphylococcus aureus, Enterococcus faecalis, Vibrio cholera and Streptococcus pyogens) and six fungal strains (Candida albicans, Candida krusei, Candida glabrata, Candida keyfer, Candida tropicalis, Candida parapsilosis). The antimicrobial activity of synthesized compounds (1a-1g) ranged from 250 μM to 15.67 μM and results have been summarized in Tables 1 and 2. The activities and efficacies of the compounds were compared with kanamycin, a routinely used standard antibiotic, and amphotericin B, a well-established antifungal drug [56][57][58]. Interestingly, the compounds 1b and 1f have been found to be the most potent antimicrobial agents among all tested compounds. When used as a standalone antibacterial agent, the MIC of 1b was either comparable (S. aureus) or at least 50% reduced (S. pyogens) compared to KAN. In case of 1f, the MIC is 50% less compared to that of KAN when tested against S. aureus, E. faecalis and S. pyogens.
Though MIC was much higher in case of antifungal screening, the compounds 1b and 1f as standalone found to be more active in comparison to other candidates when Amp B was used as reference drug. The results of antimicrobial activity study revealed that incorporation of halogen atoms (Cl and Br) on coumarinyl and phenyl moieties linked to pyrazolylthiazoles may be responsible for higher potential of 1b and 1f. Generally, substitution of hydrogen by halogen in coumarin derivatives compounds increases the lipophilicity and thus increases the rate of cell penetration, which is an important feature of antimicrobial drug efficiency [59]. Substitution of chlorine at position-6 of coumarin ring in 1b and 1f significantly increased activity against the bacterial and fungal strains. The results demonstrated that the antimicrobial activity of the investigated compounds was influenced by the physicochemical properties of the type and position of substituent on the aromatic ring.
Furthermore, it is noteworthy here that S. aureus and candida sp. are responsible for major opportunistic infections in clinical settings. There are growing concerns about the ability of S. aureus to acquire resistance, even towards the newest antibacterial available in the market [60]. Similarly, C. keyfer infections mostly occur in patients with hematologic malignancies or carcinomas who are already undergoing highly cytotoxic chemotherapy [61]. KAN has been used as a standard drug in infections with S. aureus [62]. Likewise, Amp B is a one of the primary standard drugs used in case of candida infections [61,63]. Emergence of resistance against these front line drugs has been a major public health problem. An important strategy to reduce the emergence of drug resistant strains is by reducing the effective amount of drug required to kill the infectious bacterial or fungal agent [64,65]. Therefore, combinatorial therapies or use of adjuvant is encouraged or used to modulate the antimicrobial activity and mitigate the possibility of development of drug resistance [46]. CTP hybrids: Potential standalone or adjuvant antimicrobial agents Synergistic effect. As discussed above, we further evaluated whether the compounds 1b or 1f can act as an adjunct to potentiate a reduction in MIC of KAN and Amp B against clinical strains S. aureus and C. Keyfer, respectively. A checkerboard dilution assay has been used to determine the fraction inhibitory concentration index (FICI) for the compounds 1b and 1f, when used in combination with KAN and Amp B against S. aureus and C. Keyfer, respectively. As evident from Table 3, we observed an FICI of 0.5 in all drug combinations reported here indicating a synergistic interaction between the compounds and known drugs. There was a substantial 8-fold reduction in the MIC of KAN against S. aureus with compounds 1b, and also with 1f. Likewise, in case of Amp B, there is a 5-fold reduction in MIC with compound 1b and~8 fold reduction with compound 1f against fungal pathogen C. keyfer.

Cytotoxic evaluation. Evaluation of cytotoxicity using HeLa cells (cervical cancer cell) and Hek-293 cells (normal embryonic kidney cells) indicated that the compounds 1b and 1f
have comparable cytotoxicity (> 80% survival) at MIC concentrations (Fig 3) and negligible cytotoxicity (> 90% survival) at the concentrations used in combination. Their cytotoxicity was comparable with Novobiocin, a coumarin derivative used as a positive control [48]. Most potent compounds emerging out of our screen 1b and 1f have comparable cytotoxicity at their respective MIC concentrations.  Since compounds 1b and 1f are active as an adjuvant at much lower concentrations (5-8 time less than their MIC values), we believe that these compounds can act as an ideal adjuvant to increase the efficacy of KAN, Amp B or possibly another drugs too. Our experiments with another antifungal agent fluconazole showed that usage of 1b and 1f as adjunct reduced the MIC~5-fold with 1b and 8-fold with 1f against C. keyfer substantiating the broad spectrum effects of these derivatives as adjunct. Compound 1f consistently demonstrated the most potent activity as a standalone as well as in combination to other drugs as an adjunct. So we evaluated this compound for further analysis including a time kill kinetic assay.

Antibacterial/ Antifungal drug (μM) Compd. (μM) Conc. (μM) of the drug in combination with 1b or 1f Fraction inhibitory conc. index (FICI) a Effect
Time kill kinetic. Time kill kinetic is an important assay to find out the efficacy and killing rate of antimicrobial compounds [66]. In the time kill assay, the Colony Forming Units  5). The untreated S. aureus and C. keyfer cells serve as control [67].
As a result, compounds 1b and 1f in our screen displayed significant broad spectrum antimicrobial activity against clinical strains of bacteria and fungi without any significant mammalian cell cytotoxicity. Comparing the antibacterial activity of coumarin derivatives, 1b showed the similar level of activity with MIC 31.25 μM and 1f showed an excellent level of activity with MIC 15.67 μM when compared with standard drug, Kanamycin. Among the synthesized compounds, the 1f showed the best activity against S. aureus (MIC 15.67 μM). The MIC value of Amp B against C. keyfer is 12.50 μM. More promisingly, when used as an adjunct to standard

DFT studies: Electrostatic results
The electron donating and receiving ability of a molecule can be determined by considering HOMO and LUMO energy of that molecule. Higher the energy of HOMO higher will be its electron donating ability. A full geometry optimization was performed for all compounds (1a-1g), using Khon-Sham's DFT method that undergo the gradient-corrected hybrid density functional B3LYP method and the 6-31+g(d,p) bases set [49]. It has been found that the energy of HOMO was comparable for all compounds. However, the LUMO orbital energies were affected by the presence of electron donating or withdrawing or both groups on the structure. According to Mabkhot et al. [68], HOMO-LUMO energy gap (E gap) is an established parameter to measure the extent of the intramolecular charge transfer and was used in pharmaceutical studies. In this study, the compounds exhibited different antimicrobial activity, this may be due to the difference in LUMO energy levels. Table 4 summarized the theoretical electronic parameters for investigating the coumarin-thiazole-pyrazole (CTP) based molecular hybrids.
Correlation of biological assay with electrostatic results. The electrostatic results presented in Table 4 were plotted against the experimental biological activity of synthesized compounds (Fig 6). The biological activity for excellent to low biological active compounds has been correlated with energy of their LUMO. The compound 1b and 1f showed illustrious LUMO energy of -2.647 eV and -2.648 eV respectively and also exhibited higher antibacterial & antifungal activity than the other compounds. It was observed that compounds with more stabilized LUMO orbitals have magnificent biological activity (Fig 6).  The effect of LUMO energy on biological activity may be due to the presence of electron withdrawing bromo and chloro groups in compound 1b and 1f. Presence of chloro group stabilized LUMO to a higher extent than bromo which was evidenced by the comparison of MIC of compounds 1b and 1f. Halogen substitution on coumarin ring stabilized LUMO energy more than on phenyl ring. Relating LUMO energy with biological activity was not the only factor responsible for biological activity. There was a strong demand to look into other aspects like the placement of these molecular orbitals on the molecule (Fig 6). Such investigations revealed that location of LUMO along with their energy together influenced the biological activity. The location of LUMO was particularly affected by the substitution of halogen on coumarin ring which was clearly observed in the correlation of LUMO of compounds 1a, 1b and 1f (Fig 6). It has been revealed from electrostatic results that biological activity of coumarinthiazole-pyrazole (CTP) hybrids was due to substitution of coumarinyl ring with electron withdrawing groups which consequently lowered its LUMO energy. Therefore, a theoretically suggestion about different biological activity of coumarin-thiazole-pyrazole (CTP) hybrids was made based on the localization and energy of LUMO orbital.
Global reactivity descriptors. The chemical reactivity indices like chemical hardness (η), electronegativity (χ), electronic chemical potential (μ), and electrophilicity Index (ω), were also calculated for all the synthesized molecules. The stability and reactivity of a chemical system is represented by chemical hardness which is given by η = (E LUMO -E HOMO )/2. This descriptor is used as a measures of resistance to change in the electron distribution or charge in a molecule [69]. Electronegativity is given by expression χ = -(E HOMO + E LUMO ) /2, which is defined as the power of an atom in a molecule to attract electrons towards it [70]. The negative term of electronegativity of a molecule is defined as Chemical potential [71] which is determined using equation μ = (E HOMO +E LUMO )/2. Electrophilicity index (ω), is a measures of the propensity or capacity of a species to accept electrons was introduced by Parr and is calculated using the electronic chemical potential and chemical hardness and given as ω = μ 2 /2η [72]. The  Table 5. The low chemical hardness (1.76 eV) with high negative value of chemical potential (-4.41 eV) indicates that compound 1a and 1f are comparatively soft molecule with high polarizability than other compounds of this series. The high electronegativity (4.41 eV) and electrophilicity (5.51 eV) of 1a and 1f indicates its high power to withdrawn electrons and hence to act as an electrophile.

Docking studies
Molecular docking is the most useful technique to explore the possible binding mode between ligand and protein complex. The re-docking of native ligand into the crystal complex is a measure of accuracy of a molecular docking. The comparison between lowest energy pose and the experimentally determined binding mode can be used to as a tool for determining the precision of binding mode. First of all the validation of the docking protocol was done by re-docking the native ligand into the active site of DNA gyrase B (PDB ID: 5L3J) with a root-meansquare deviation (RMSD) of <1Å. This lower value of RMSD represents the accuracy and reliability of the docking procedure in reproducing the experimentally observed results (Fig 7).
In the present manuscript, an attempt has been made to study the interactions of novel coumarin-thiazole-pyrazole (CTP) hybrids with DNA gyrase B. For this purpose molecular docking studies of all the synthesized compounds 1a-1g were performed in the binding pocket of DNA gyrase B (PDB ID:5L3J). Four basic parameters-Glide score, Glide energy, H-bonds and non-bonded interactions (van der Waals and Coulombic) were used to define the binding affinity of ligands 1a-1g for DNA gyrase B. Glide score with more negative value represents minimum energy for the formation of complex between ligand and receptor (Glide energy). The docked poses for compounds 1b, 1g and KAN in the active site of DNA gyrase B are shown in Figs 8-10 ( Table 6). The most active compounds showed substantial binding affinities towards the DNA gyrase B structures (Glide energy range -69.60 kcalmol -1 to -84.78 kcalmol -1 ) and the energy ranges are more than that of reference compounds KAN and CPF (Table 6). It has been revealed from binding studies that all the synthesized compounds 1a-1g fit snugly making various close contacts with the residues lining the active site of DNA gyrase. In the binding of molecules 1a-1g to DNA gyrase B, the contribution made by van der waals and columbic interactions are more prevalent over the electrostatic interactions.  (CTP) based hybrids toward active site of DNA gyrase B opens up a new platform structurebased design efforts. Pharmacokinetic effect of synthesized compounds 1a-1g. Pharmacodynamics results indicated that compounds (1a-1g) may prove good lead as antibacterial drugs by targeting DNA gyrase B. For drug formulation pharmacokinetic properties are likewise a significant prospect and should be brought into thoughtfulness. The pharmacokinetic parameters of all synthesised hybrids have been studied along with standard drug KAN and CPF using QikProp [73]. The major parameters studied for different synthesised compounds (1a-1g) were seems to be acceptable and were incorporated in Table 7. The % oral drug absorption predicted for all the test compounds 1a-1g was highly satisfactory with a high percentage (>80%) of Human Oral Absorption as compared to the reference, indicating their possibilities in oral drug formulation for the treatment of S. aureus infection.
The bowel-blood barrier was predicted using Caco-2 cell permeability (QPPCaco) as model along with Lipinski's rule of 5 [74] all the test compounds and showed very good values compared to the standard Kanamycin and ciprofloxacin. Further, the prediction for human serum albumin binding using QPlogKhsa, shows that the values for all the inhibitors lie within the expected range (-1.5 to 1.5). Likewise, the brain/blood partition coefficient (QPlogBB) and IC50 value of HERG K+ channel blockage (QPlogHERG) for all compounds shows satisfactory values as compared with reference compounds (Table 7).

Conclusion
In conclusion, some new coumarinyl linked pyrazolylthiazoles have been synthesized and explored for their potential as standalone or adjuvant antimicrobial agents against pathogenic strains. Compounds 1b and 1f displayed significant antimicrobial activity without expressing any significant mammalian cell cytotoxic effect. Compound 1b showed the similar level of antibacterial activity with MIC 31.25 μM and 1f found best agent with MIC 15.67 μM particularly against S. aureus when compared with standard drug, kanamycin. Interestingly, as an adjuvant to standard antimicrobial drugs, compounds 1b and 1f significantly enhanced the efficacy of the standard drugs KAN, Amp B and FLC by significantly reducing the MIC values of these drugs.
The results demonstrated that the antimicrobial activity of the investigated compounds was influenced by the physicochemical properties of the type and position of substituent on the aromatic ring. The results based on electron microscopic investigations of cell morphology and cellular structural integrity suggest that the antimicrobial activity of these compounds may in part be attributed to initial cell wall damage, leading to action on the intracellular components of cells, ultimately leading to cell death. The result of DFT studiy of compounds were correlated with the observed antimicrobial activities which indicate that coumarin moiety plays an important role in antimicrobial activities. Moreover, the molecular docking studies further helped in supporting the experimental results and the ADME properties predicted further strengthen the path towards oral drug formulation. Therefore, it is concluded that the compounds 1b and 1f   may act as lead compounds for the development of potent antibacterial and antifungal agents and as adjuvant they have potential to unlock the microbial resistance to antibiotics.