Novel derivatives of eugenol as potent anti-inflammatory agents via PPARγ agonism: rational design, synthesis, analysis, PPARγ protein binding assay and computational studies

Eugenol is a natural product abundantly found in clove buds known for its pharmacological activities such as anti-inflammatory, antidiabetic, antioxidant, and anticancer activities. It is well known from the literature that peroxisome proliferator-activated receptors (PPARγ) have been reported to regulate inflammatory responses. In this backdrop, we rationally designed semi-synthetic derivatives of eugenol with the aid of computational studies, and synthesized, purified, and analyzed four eugenol derivatives as PPARγ agonists. Compounds were screened for PPARγ protein binding by time-resolved fluorescence (TR-FRET) assay. The biochemical assay results were favorable for 1C which exhibited significant binding affinity with an IC50 value of 10.65 μM as compared to the standard pioglitazone with an IC50 value of 1.052 μM. In addition to the protein binding studies, as a functional assay, the synthesized eugenol derivatives were screened for in vitro anti-inflammatory activity at concentrations ranging from 6.25 μM to 400 μM. Among the four compounds tested 1C shows reasonably good anti-inflammatory activity with an IC50 value of 133.8 μM compared to a standard diclofenac sodium IC50 value of 54.32 μM. Structure–activity relationships are derived based on computational studies. Additionally, molecular dynamics simulations were performed to examine the stability of the protein–ligand complex, the dynamic behavior, and the binding affinity of newly synthesized molecules. Altogether, we identified novel eugenol derivatives as potential anti-inflammatory agents via PPARγ agonism.


Introduction
Peroxisome proliferator-activated receptors (PPARs) are wellcharacterized type II nuclear receptors and transcription factors. PPARs represent a group of three receptors PPARa, PPARb/d, and PPARg which causes inhibition of NF-kB activation as depicted in (Fig. 1). 1 Nuclear factor kappa-B (NF-kB) signaling is an important part of the immune response as it plays an important role in inammatory processes by inhibiting factor-a (TNF-a), interleukin 1b (IL-1b), interleukin 6 (IL-6) and nitric oxide. The target protein, PPARg, is reported to regulate inammatory responses in physiological systems. 2,3 Fig. 1 Activated PPARg inhibits the activity of NF-kB which is responsible for inflammatory reactions.

Design of eugenol derivatives as PPARg agonists
The PPARg agonists possess structural features such as a heterocyclic head usually thiazolidinedione followed by a benzyloxy trunk, two carbon linkers, and a lipophilic tail. Considering these structural features, we designed novel eugenol derivatives that possess a benzyloxy trunk, two carbon linkers, and a lipophilic tail (Fig. 2). However, the heterocyclic head is replaced with an aliphatic chain with a terminal double bond. This was the only difference between existing PPARg agonists and the present eugenol derivatives.

Synthesis
The acylated amines 2 were synthesized by treating substituted aromatic amines 1 with chloroacetyl chloride using dichloromethane as a solvent. The acylated amines were then linked to eugenol in presence of anhydrous potassium carbonate and dry acetone to obtain novel eugenol derivatives 20 (1A-D) (Scheme 1). The characterization of the synthesized compounds was done by IR, 1 H-NMR, 13 C-NMR, and mass spectrometry. The melting points of the synthesized derivatives were determined in open capillaries using melting point apparatus and are uncorrected. TLC was performed using n-hexane and ethyl acetate in the ratio of (8 : 2) as mobile phase on aluminium plates which are precoated with silica gel G. Compounds were named following the IUPAC rules. The IR spectra were obtained using a KBr pellet technique on a Shimadzu infrared FTIR spectrophotometer and are given in cm 1 . The IR spectrum shows the disappearance of the band attributed to the -OH group at around 3250 cm À1 . The band at 3460 cm À1 indicates the presence of the (NH) group and the presence of (C]O) at 1690 cm À1 . The solvent used for 1 H-NMR and 13 C-NMR was CDCl 3 . The chemical shis are quanti-ed in parts per million (d ppm), with the notation s ¼ singlet, d ¼ doublet, t ¼ triplet, and m ¼ multiplet. The 1 H-NMR spectra of all the newly synthesized compounds exhibit a singlet at around d 8.90-9.00 ppm which is attributed to the characteristic of NH in addition to aromatic protons. One proton of CH] gives multiplet at around d 5.9-6.0 ppm and two protons of ]CH 2 give doublet at around d 5.1 ppm. Similarly, all the 13 C NMR spectra show distinguishing signals d 166.7-172.1 ppm due to the ketonic carbon, in addition to other normal signals of carbon. The molecular mass of all the nal compounds was determined by LC-MSMS. All these facts conrm the formation of nal compounds (1A-D) Thus, the spectral data conrm the presented structures of all newly synthesized eugenol derivatives.

ADMET and TOPKAT prole
ADMET and toxic prediction of the compounds is essential parameters that need to be assessed the efficacies or risks of small compounds. 21 The pharmacokinetics and dynamic proles of synthesized compounds are tabulated ( Table 1). The potential drug must cross certain barriers in the biological system to show ideal ADME properties. Hence, it was qualitatively assessed for the eugenol derivatives. Furthermore, the analysis depicted all the compounds obey the oral drug-likeness Pzer rule and are free from mutagenicity and carcinogenicity 22 as tabulated in (Tables 1 and 2).

Structure-based drug designing
The perception of structure-based drug designing was implemented for known compounds and drug target proteins. For PPAR-gamma receptor-ligand interaction in the ligand-binding domain region as Phe282, Cys285, Gln286, Arg288, His323, Tyr327, Leu330, Gly338, His449, and Leu469 is favorable for eugenol derivatives. Further, the binding energies calculated using the implicit solvent model PBSA are tabulated (Table 3). During interaction probing it was observed that eugenol derivatives interact with the one or more amino acid residues of the ligandbinding domain of PPAR-gamma. The amino acids Arg288, Ser289, and His323 interact to form hydrogen bond interactions (Fig. 3) and other amino acids are favorable for hydrophobic interactions. Besides, binding energies are signicantly higher with 1C and standard as compared with other complexes.

Structure-activity relationship (SAR)
A structure-activity relationship (SAR) is a key tool and information pool in organic chemistry en routes to synthesize many compounds for the better and more potent acquisition of compounds for various drug targets. 23 It is a guide used to predict the compound's biological activities from chemical structure, the functional moieties attached to the compounds play a signicant role. In this study receptor-ligands complex of 1A-1D was used for interaction pharmacophore analysis as shown in (Fig. 5). Interestingly, the 4-allyl-2-methoxyphenoxy group is common in all the compounds sharing two hydrophobic and one aromatic hydrophobic. The potential high active compound 1C (IC 50 ¼ 10.65 mM) possesses halogen  group (cl) in the para position of benzene enhances biological activity. But, OCH 3 substitution in the para position reduces the biological activity in compound 1B. Hence the position of the parafunctional group is considered a crucial pharmacophore for compounds to inuence the PPAR gamma protein binding and is directly proportional to the biological activity ( Fig. 6).
In both drug targets, many van der Walls interactions offered by residues within 5Å and molecular docking of lysozyme egg protein with eugenol derivatives have a tendency to form hydrogen and hydrophobic interactions (Fig. 4).
Similarly, the lysozyme enzyme interaction complex was probed for SAR analysis depicted (Fig. 6) the halogen, chlorine attached at the para position of the benzene ring shows good biological activity than other eugenol derivatives. Substitution of OCH 3 (1B) or CH 3 (1A) reduced the biological activity. Whereas, CH 3 at the meta position of the benzene ring in compound 1D shows better activity than 1B and 1A. Thus, SAR analysis shows the importance of the functional moiety pharmacophore of the compounds.

Time-dependent parameter conformational analysis of the complex
The binding modes of best-docked molecules of receptorligand were further studied using a molecular dynamics simulation study for a simulation time of 1000 ps. 24 The geometric features of the protein-ligand complexes, such as root mean square deviation (RMSD) and radius of gyration (R g ), were determined to assess the system's stability. The RMSD is used to measure the root mean square deviation of atomic positions of each conformation. The average distance between the atoms of various structural conformations of protein and ligand over a period of time. The average RMSD of the Ca atoms of peroxisome proliferator-activated receptor (PPAR) gamma and heavy atoms of 1C was determined to be 2.490 AE 0.1105Å. In contrast,  the standard drug pioglitazone with peroxisome proliferatoractivated receptor (PPAR) gamma complex had an average RMSD of 2.440 AE 0.07039Å (Fig. 7A). Similar analysis of lysozyme bound complex average root mean square deviations are 2.054 AE 0.06697Å and 1.994 AE 0.04284 of 1C and standard drug (diclofenac) respectively (Fig. 7B). Hence, a conformational change of the compounds has a direct inuence on biological activity. R g is the root mean square distance between each atom  in a system and its center of mass. The R g values for proteinligand complexes: peroxisome proliferator-activated receptor (PPAR) gamma 1C and pioglitazone show uctuations between 18.6Å to 18.9Å while, lysozyme bound complexes R g values are between 13.65Å to 13.85Å. The energy parameter analysis of all the complexes shows fewer variations (Table 4) which indicates energetically all three complexes are good. Overall, peroxisome proliferator-activated receptor (PPAR) gamma 1D complex shows less deviations with time-dependent parameter analysis and 1C lysozyme complex shows best in inammatory activity.

TR-FRET assay
Based on time-resolved uorescence resonance energy transfer (TR-FRET), the nuclear receptor competitive binding assay was utilized to identify PPARg ligands. When a uorescent ligand (tracer) is bound to the receptor, energy transfer from the antibody to the tracer occurs, and a high 520/495 ratio is detected. 25 The tracer is displaced from PPARg-LBD by a chemical under test, resulting in a decrease in the FRET signal and a low TR-FRET ratio (Fig. 8). The assay results indicate that 1C shows an IC 50 value of 10.65 mM whereas standard pioglitazone shows an IC 50 value of 1.052 mM. This suggests that 1C has a binding affinity for the target protein that is similar to that of typical pioglitazone. The IC 50 values of standard drug pioglitazone and synthesized compounds are tabulated in (Table 5).

Statistical analysis
All the results were expressed as means AE standard deviation (SD) and the data were analyzed using one-way ANOVA using Graph pad prism 8.0 soware. P values < 0.05 were considered signicant.

2.9
In vitro anti-inammatory activity 2.9.1. Albumin denaturation assay. The protein denaturation has long been acknowledged as a factor of inammation. Inammatory diseases and disorders such as diabetes, rheumatoid arthritis, and cancer are associated with the denaturation of protein. The ability of a substance to prevent protein denaturation helps to prevent inammatory disorders. 26 This assay nds its importance as part of preclinical studies to establish the potency and efficacy of the new molecules in the process of drug discovery. [27][28][29] In this assay egg albumin is used as protein. Protein denaturation is achieved by keeping the reaction mixture at 70 C in a water bath for 10 minutes. As a part of the investigation of the mechanism of the anti-inammatory activity, the ability of the synthesized compounds and standard diclofenac to inhibit protein denaturation was studied. [30][31][32][33] At various concentrations, it proved efficient in preventing heat-induced albumin denaturation, as shown in (Table 6). The novel synthesized compounds signicantly (p < 0.05) inhibited the albumin denaturation was shown in (Fig. 9) and maximum inhibition of 48.08 AE 1.143 at 200 mM was observed for 1C when compared with standard diclofenac 48.28 AE 3.139 at 150 mM. IC 50 values of standard diclofenac and synthesized compounds on inhibition of albumin denaturation were statistically signicant (Table 7 and Fig. 9).
Albumin denaturation is a process leading to the loss of secondary and tertiary structure of proteins due to external stress such as strong acid or base, organic solvent, or heat. When biological proteins are denatured, they lose their biological function. Hence, a ligand's ability to inhibit the denaturation of protein signies the potential for anti-inammatory activity. The anti-inammatory potential was depicted clearly in the albumin denaturation assay. The present research shows that newly synthesized compounds can limit the formation of autoantigens caused by protein denaturation and stabilize lysosomal membranes in vivo. This research provides the scientic groundwork for various inammatory diseases. In vivo studies can also be applied to understand the mechanism of the anti-inammatory activity of newly synthesized eugenol derivatives.

Summary and conclusions
The titled compounds were designed based on pharmacodynamics and pharmacokinetics requirements. The pharmacokinetics disclosed that newly synthesized compounds 1A-D obey Lipinski's rule and show promising drug scores. The docked structures at the binding sites were found to be stable using molecular dynamics simulations. The average RMSD of Ca atoms of PPAR-gamma and heavy atoms of 1C was found to be 2.490 AE 0.1105Å compared to the standard drug pioglitazone with PPARg complex had an average RMSD of 2.440 AE 0.07039 A. The R g values for protein-ligand complexes: PPAR gamma 1C and pioglitazone show uctuations between 18.6Å to 18.9Å while, lysozyme bound complexes R g values are between 13.65Å to 13.85Å. PPAR gamma 1D complex shows fewer deviations with time-dependent parameter analysis and compound 1C agrees with biological activity. We reported a simple, yet efficient method to synthesize some eugenol derivatives using substituted aromatic amines and characterized using spectroscopic techniques (FTIR, 1 H NMR, 13 C NMR, and mass spectrometry). TR-FRET assays validated our in silico prediction   Step 1: Synthesis of substituted acylated amine. Substituted aromatic amine (1 equivalent) and triethylamine (1.05 equivalent) along with dichloromethane (80-100 ml) were transferred into a ask tted with a guard tube. While the above mixture was stirred under iced conditions (0-5 C), 1.05 equivalents of chloroacetyl chloride were added drop by drop for 30 minutes. The reaction mixture was again stirred for about 10-12 h at room temperature. The reaction was monitored with   thin-layer chromatography (TLC) using n-hexane and ethyl acetate as the mobile phase. Aer completion of the reaction, the mixture was treated with water and dilute HCl and transferred to a separating funnel, and allowed to separate. The water layer was removed and the DCM layer was passed through anhydrous sodium sulfate. The solvent was evaporated to obtain acylated amines. 4.1.2.
Step 2: Coupling eugenol with substituted acylated amines. A mixture of acylated amines (1 eq.), nely powdered anhydrous potassium carbonate (K 2 CO 3 3 eq), eugenol (1.2 eq.) along with 80 ml of dry acetone was stirred at 45 C for about 26 h. The reaction progress was monitored by checking the spots, from time to time during the reaction using TLC with nhexane and ethyl acetate as a mobile phase. Acetone was evaporated and the reaction mixture was extracted with ethyl acetate. The ethyl acetate layer was washed three times with a 10% NaOH solution, once with water, and then with brine solution before being dried on anhydrous sodium sulfate as described in Scheme 1. The ethyl acetate layer was evaporated to obtain the nal compound as illustrated in Table 8.

ADMET, TOPKAT and drug likeness
The pharmacokinetics and dynamics properties of compounds were analyzed through a small molecular protocol (BIOVIA, Discovery Studio 2019) to understand the molecular behavior. Further, mutagenicity and carcinogenicity, and the set dosage range of the compounds were analyzed using the Bayesian and regression model. In addition, Lipinski's rule of 5 was carried out to nd out the oral bioavailability of the compounds.

Molecular docking
The structure of drug target protein and compounds employed for docking was prepared using the macromolecule tool and small molecule protocol in Discovery studio 2019. The bound ligand coordinates 49.720X À36.98Y 19.294Z of radius 8.4Å for PPAR-Gamma (2PRG) and 26.63X 5.50Y 14.22Z of radius 12Å for egg lysozyme (3WXU) with an equal grid spacing of 0.5Å with 90-degree grid angles is dened as binding sites for docking. CDOCKER algorithm was employed to study receptor-ligand interaction, as it is a powerful CHARMm-based docking approach that has been demonstrated to produce extremely accurate docked poses. All the parameters were set as default while executing the docking process. The best poses are further probed for binding energy calculation with implicit solvent model PBSA. The energetically stable complex was taken for molecular dynamics simulation and nonbonded interaction analysis. [34][35][36]

Molecular dynamics simulation
The standard drugs and top two compounds of interaction with drug targets PPAR-gamma and egg lysozyme from the docking study were subjected to 1000 picosecond of molecular dynamics simulations using Dassault Systems BIOVIA, Discovery Studio 2019 Modeling Environment. It was carried out in ve cascade steps beginning with two steps of 500 cycles of energy minimization of the complex with the steepest descent and conjugate gradient. Following heating, the system gradually forms 50 K to 300 K with 100 ps of simulation time and equilibration for 500 ps to attain degrees of freedom. Finally, the production with the canonical ensemble was subjected to equal T mass and P mass at 300 K. Additionally, the SHAKE constraint xes all bonds involving hydrogen in the simulation, within this cutoff distance of 12-10Å. All the atom velocities and positions are calculated at time points using the leap-frog verlet algorithm. At last, the deviations in the conformation of the complexes were analyzed by RMSD and R g .

TR-FRET competitive ligand displacement assay
We performed a time-resolved uorescence resonance energy transfer (TR-FRET) assay using a black at bottom 384-well plate and a buffer containing 20 mM potassium phosphate (pH 7.4), 0.5 mM EDTA, 50 mM KCl, and 0.01% Tween-20, 5 mM TCEP. For the TR-FRET coregulator interaction assay, each well contained 400 nM FITC-labeled TRAP220 or NCoR1 peptides, 1 nM LanthaScreen Elite Tb-anti-His Antibody, 4 nM 6xHis-PPARg LBD protein and 400 nM peptide in TR-FRET buffer in 22.5 mL total volume per well. Each well (22.5 L per well) included 1 nM 6xHis-PPAR LBD protein, 1 nM LanthaScreen Elite Tb-anti-His Antibody, and 5 nM Fluormone Pan-PPAR Green tracer ligand in TR-FRET buffer for the ligand displacement experiment. Ligand stocks were made by serial dilution in DMSO, then added to wells in triplicate to a nal DMSO concentration of 1%, then incubated at room temperature for 1 hour, and read using a BioTek Synergy Neo multimode plate reader. The Tb donor was stimulated at 340 nm, its uorescence emission was measured at 490 nm, and the FITC emission of the acceptor was detected at 520 nm. The signal at 520 nm/ 490 nm was used to compute the TR-FRET ratio. 37,38 4.6 Anti-inammatory activity 4.6.1. Albumin denaturation assay. The reaction mixture containing 0.2 ml of egg albumin (from fresh hen's eggs), 2.8 ml of phosphate-buffered saline (pH 6.4), and were mixed with 2 ml of the synthesized compounds in varying concentrations. Then the samples were incubated at 37 C for 20 min and then heated to 70 C for 20 min. Aer cooling the samples, the turbidity was measured spectrophotometrically at 660 nm. Diclofenac sodium was used as a reference standard. The following formula was used to calculate the percentage inhibition of protein denaturation.