Synthesis, characterization, molecular docking, ADMET prediction, and anti-inflammatory activity of some Schiff bases derived from salicylaldehyde as a potential cyclooxygenase inhibitor

: A series of Schiff base-bearing salicylaldehyde moiety compounds (1-4) had been designed, synthesized, subjected to insilico ADMET prediction, molecular docking, characterization by FT-IR, and CHNS analysis techniques, and finally to their Anti-inflammatory profile using cyclooxygenase fluorescence inhibitor screening assay methods along with standard drugs, celecoxib, and diclofenac. The ADMET studies were used to predict which compounds would be suitable for oral administration, as well as absorption sites, bioavailability, TPSA, and drug likeness. According to the results of ADME data, all of the produced chemicals can be absorbed through the GIT and have passed Lipinski’s rule of five. Through molecular docking with PyRx 0.8, these synthesized compounds were tested insilico selectivity toward COX-1 and COX-2 and in vitro for their anti-inflammatory efficacy . In vitro testing demonstrated that all of the produced compounds had significantly stronger activity against the COX-2 enzyme than COX-1. Among these, compound 1 displayed the most potent inhibitory activity with an IC 50 value of 0.19 µM compared to standard drug celecoxib (IC 50 = 0.29 µM). The most active derivative compound1 was oriented towards the active site and occupied the target enzyme based on the docking investigation against COX-1 and COX-2. In addition, insilico investigations found that COX-2 has a higher inhibitory activity than COX-1


Introduction:
Inflammation is a common and very important pathological process that can cause tissue or cell damage characterized by redness, edema, fever, heat, pain, and loss of function 1,2 . As a result, inflammatory processes are involved in a variety of diseases such as atherosclerosis, Alzheimer's disease, Parkinson's disease, cancer, asthma, arthritis, etc. 3,4 . Although nonsteroidal anti-inflammatory drugs (NSAIDs) are currently among the most commonly prescribed drugs, they are attributed to their wide range of medical indications as they can be used as analgesic, anti-inflammatory, anti-rheumatic and antipyretic agents 5 . In addition, NSAIDs' pharmacological effect is linked to the inhibition of PG synthesis from arachidonic acid via inhibition of the crucial regulating enzyme cyclooxygenase COX. At least two isoforms of cyclooxygenase have been identified: COX-1, or prostaglandin H1 synthase, and COX-2, or prostaglandin H2 synthase 6 .COX-1 is expressed in most tissues, regulates physiological processes such as gastric cryoprotection, renal function, and platelet aggregation, and is stimulated by growth factors and hormones, while COX-2 mainly supplies prostaglandin E2 (PGE2) and prostacyclin (PGI2). Many NSAIDs interact with both COX-1 and COX-2 isoforms and nonselectively inhibit their enzymatic activity, resulting in a reduction in the production of the prostaglandins PGE2 and PGI2 7,8 . NSAIDs, despite their widespread usage, have been linked to significant adverse effects such as gastrointestinal (GI) problems, hypertension, edema, kidney disease, and heart disease risk 9. Because of the difference in expression profiles between COX-1 and COX-2, a hypothesis was advanced in the 1990s that selective inhibitors(celecoxib)of COX-2 would share the beneficial anti-inflammatory properties of traditional NSAIDs but lack the gastric toxicity 10 . Therefore, it is crucial to synthesize and develop novel and more potent anti-inflammatory drugs with no or fewer side effects. Schiff bases are aldehyde or ketone-like compounds in which the carbonyl group is replaced by an imine or azomethine group(C=N), when reacted with any primary amine under certain conditions 11 . Schiff bases are among the most commonly used organic compounds 12,13 . In recent years, Schiff bases have been widely used to formulate various types of drugs due to their diverse biological activities including antifungal, antibacterial, antimalarial, antiproliferative, antiinflammatory, antiviral, and antipyretic properties 11,[14][15][16][17] . Numerous studies have reported that Schiff's bases of substituted salicylaldehyde (2hydroxybenzaldehyde) are well-known antimicrobial agents, analgesic, and antiinflammatory in free form or as ligands in metal complexes [18][19][20] . In the present study, both experimental and computational approaches are used to discover novel anti-inflammatory drug candidates.

Instrumentation
Synthesized compounds were purified by recrystallization in appropriate solvents and examined through thin layer chromatography (Merck Silica gel 60 F254) and UV light (320 nm). The melting points were determined in open capillary tubes using the Stuart/SMP3 melting point equipment version 5.0. Infrared spectra were recorded without a KBr disk using Thermo Scientific™ Nicolet™ iS™10 FT-IR Spectrophotometer in Pioneer company for pharmaceutical industry-Sulaimani-Iraq. CHNSelemental analysis was used to characterize the produced substances (CHNS-O Elemental Analyzer Vario EL, ELEMENTAR, Hanau-Germany) in the College of Pharmacy-Hamedan University of Medical Sciences-Hamedan-Iran.

Chemical synthesis:
The synthesis of Schiff base derivatives was carried out using the methods illustrated in Scheme 1.

General Procedure for Schiff bases derivatives synthesis (1-4):
Equimolar amounts of salicylaldehyde and the primary amine (2,4-dinitrophenylhydrazine, aniline, semicarbazide, thiosemicarbazone) were dissolved in methanol 10 mL and agitated for 10 minutes at room temperature to yield a transparent solution. After approximately three days of standing. TLC was used to monitor the reaction, then filtering was used to collect the precipitate. The precipitate was recrystallized from methanol, washed three times with methanol, and dried in a vacuum desiccator containing anhydrous CaCl2. Table 1 shows the physicochemical parameters of the compounds (1-4) 21

Computational Studies: In silico ADMET properties
In-silico prediction of ADME properties of the Schiff base derivatives was carried out using the SWISS-ADME online server 22 . while toxicity was predicted using the organ toxicity and endpoint toxicity model of the ProTox-II software 23 . Several critical parameters were predicted, including molecular weight, topological polar surface area (TPSA), number of hydrogen bond donors (HBD), and number of hydrogen bond acceptors (HBA). Organ toxicity is classified in the category: of hepatotoxicity. The endpoints of toxicity are classified: carcinogenicity, immunotoxicity, mutagenicity, and cytotoxicity. The chemical structure of the designed compounds (1-4) was drawn by using theMarvinSketch19.9 application 24 and then converted to the SMILE name. The smile notation for all synthesized compounds was considered as the starting point and input to the SWISS_ADME and ProTox-II webserver and thereby ADME and toxicity predictions were carried out.

Molecular Docking study Ligand Preparation
MarvinSketch19.9 24 was used to draw the compounds (1-4). They were sketched primarily as 2D structures and then converted to 3D format (pdb) using the same program. Ligand energy was minimized by applying the MMFF94 (Merck Molecular Force Field 94) force field algorithm 25 , and the minimized structures were converted into PDBQT format by using PyRx 0.8 26 before performing molecular docking analysis.

Preparation of receptors
The Protein Data Bank 27 was used to get the crystal structures of the cyclooxygenase receptor, the COX-1 enzyme (PDBID:3N8Z), and the COX-2 receptor (PDB ID:1PXX) Fig.1. The Discoverystudio2021 client 28 was used to remove the water molecules, heteroatoms, and co- crystallized ligands. Autodock-Tool-1.5.6 29 was used to add the polar hydrogens and Kollman charges. PyRx was used to convert the PDB files to PDBQT format.

Docking study
The docking tool PyRx (Python Prescription 0.8) was used to dock compounds (1-4) into the previously synthesized proteins COX-1 (ID:3N8Z) and COX-2 (ID:1PXX) 30 . The binding sites were chosen based on the target protein's co-crystallized ligands. PyRx affinity scores (in kcal/mol) for each chemical were obtained and rated using the free energy binding theory (more negative value means greater binding affinity). DS Visualizer and PyMOL Molecular Graphics System (Version 2.3.2 Schrödinger) were used to visualize docked conformations (poses) and receptor-ligand interactions at the molecular level.

Anti-Inflammatory Studies In vitro COX-1 and COX-2 inhibition assay
The assay for COX-1 and COX-2 enzyme inhibitory activity of the synthesized compounds (1-4) was performed based on a slightly modified protocol published by Kezhal M. et al. has been reported 31 . Cayman's COX fluorescent inhibitor screening assay (Catalog number 700100, Cayman Chemical, Ann Arbor, MI, USA), provides a convenient fluorescent-based method for screening both ovine COX-1and human recombinant COX-2 for isozyme-specific inhibitors. All reagents and solutions were prepared according to the protocols established by Cayman Chemical for the COX-1 and COX-2 inhibition assays. The solvent for the stock solutions for the test samples was dimethyl sulfoxide (DMSO). 10 µL of various concentrations of test sample solutions 0.01, 0.1, 1, 10, 50, and 100 µM were added to a series of supplied reaction buffer solutions (960 mL 0.1 M Tris-HCl pH 8.0 with 5 mM EDTA and 2 mM phenol) with either COX-I or COX2 enzyme 10 µL in the presence of heme 10 µL and 10 µL of fluorometric substrate ADHP (10-acetyl-3,7dihydroxyphenoxazine) after a 5 min incubation at 37 °C, Added 10 µL of arachidonic acid 100 µM solution and stopped the COX reaction after 2 minutes with 50 µLof 1 M HCl. When PGG2 and ADHP react, the extremely fluorescent chemical resorufin is formed. With an excitation wavelength of 535 nm and an emission wavelength of 590 nm, the fluorescence of produced resorufin can be detected. The intensity of this fluorescence depends on the amount of resorufin present in the well throughout the incubation period and is proportional to the amount of PGG2 present in the well. From the concentration-inhibition-response curve, the concentration of the test chemical causing 50% inhibition IC50, µM was calculated. The in vitro tests were carried out three times.

Computational Studies:
Analysis of ADMET properties: One of the most important aspects of the drug discovery/development process is predicting absorption, distribution, metabolism, and excretion (ADME) characteristics prior to experimental trials 32 . The bioactivity score of the synthesized compounds was calculated using the SwissADME webserver. SwissADME's freely available web server (http://www.swissadme.ch/) was used to determine ADME, pharmacokinetic parameters, and drug-like features by entering chemical structure followed by SMILES. Table 2, shows the pharmacokinetic and physicochemical data determined, including cLogP (partition coefficient), compound weight, heavy atoms, hydrogen donors, hydrogen acceptors, rotatable bonds, and TPSA values. According to Lipinski's criteria, molecules intended to be taken orally should contain no more than one violation of the following rules: (i) a maximum of 5 hydrogen donors (ii) a maximum of ten hydrogen-bond acceptors. (iii) The molecular weights of the molecules should not exceed 500 m/z. (iv) The lipophilicity (cLogP) should not exceed 5. The compounds do not function as therapeutic candidates if they exhibit more than one harm based on the above criteria 33 . From Table 2, all compounds (1)(2)(3)(4) show no Lipinskis injury and all compounds synthesized have less than ten hydrogen bond acceptors and less than 5 hydrogen bond donors. In addition, the molecule produced has a molecular weight of less than 500 m/z. All of the compounds show a decent drug-like profile based on the parameters given above. The sum of the surface areas of all polar atoms or molecules, usually hydrogen, oxygen, and nitrogen, is known as the topological molecular surface area (TPSA). The TPSA calculation is required to estimate the drug's ability to enter cells. A TPSA value of less than 140 Å 2 indicates that the molecules have good drug transport properties across cell membranes 34 . These values vary from 32.59 Å 2 to 136.26 Å 2 for synthesized compounds. These numbers show that all compounds are able to penetrate cell membranes. According to Lipinski's rule, molecules with a higher number of rotatable bonds have a more flexible structure and are better for interaction. Compound 2 has two rotatable bonds, while compounds 3 and compound 4 have three, and compound 1 has five. This demonstrates that the produced compounds, particularly compound1 with its five rotatable bonds, have a high ability to interact with living cells. The most crucial criterion is bioavailability, and bioactivity score values larger than zero imply that the chemical is very drug-like 35 . Bioactivity levels of 0.55 are found in all synthesized derivatives. The two most significant pharmacokinetic activities to be investigated at various phases of drug development are gastrointestinal (GI) absorption and blood-brain barrier (BBB) characteristics of a molecule. The permeability of the synthesized compounds (1)(2)(3)(4) identified by the BOILED-Egg method may be assessed using gastrointestinal absorption and bloodbrain barrier tests, as illustrated in Fig. 2. The BOILED Egg model (Brain or IntestinaL EstimateD Penetration) is offered as a precise predicting model based on the lipophilicity index (WLOGP) and polarity of the produced chemical. The white area suggests that the compounds are likely to be passively absorbed by the gastrointestinal system, while the yellow area indicates that the compounds can cross through the blood-brain barrier (BBB) and get access to the central nervous system. The gray areas reflect substances that are not expected to be well absorbed or permeate the BBB. The blue dot indicates that the molecule is projected to be a Pglycoprotein substrate (PGP+), whereas the red dot indicates that the molecule is likely to be a Pglycoprotein non-substrate (PGP-) 36 . As shown in Fig. 2, all of the synthesized derivatives (1)(2)(3)(4) are well absorbed in the gastrointestinal system, but they are unable to pass through the BBB, with the exception of compound 2, which shows that the compound has access to the central nervous system (CNS) and can thus be used to treat CNS inflammation.

Toxicity Prediction results:
The endpoints toxicity and organ toxicity were calculated using the online software ProTox-II. All compounds have no carcinogenic profile and no toxicity found in the hepatotoxicity and AMES toxicity assessment, Table 3.

Molecular Docking Study
Recently synthesized compounds (1-4) docked to two receptors including cyclooxygenase-1 (COX-1) enzyme (3N8Z.pdb) and cyclooxygenase-2 protein (COX-2) (1PXX.pdb). All compounds had docking scores between -9.3 and -6.7 kcal/mol Tables 4 and  5. These compounds have a higher potential to form hydrogen bonds and hydrophobic interactions with THR206, ASN382, HIS386, TYR385, HIS388, ALA202, LEU390, and TRP387 were common interacting residues. It is worth mentioning, that compound 1 was the most active compound for each protein (Tables 4 and 5). For this reason, compound 1 was selected for further analysis. The potential energy of all molecules docked with the COX-1 enzyme ranged from -8.5 to -6.7 kcal/mol. Compound1, which had the greatest docking score of -8.5 kcal/mol, has a strong inhibitory effect against the target receptor, as indicated in Table 4. The illustration of the molecular docking for the interactions of compound 1 and receptor was presented in Fig. 3 A, B, and C. The analysis of the best-docked pose of compound1 showed that the amino acid residues including THR206, THR212, and ASN382 had formed hydrogen-bonding interactions.

Anti-Inflammatory Studies In vitro COX inhibition
The in vitro anti-inflammatory activity of the synthesized Schiff base derivatives (1-4) was tested using ovine COX-1 and human recombinant COX-2 (for isozyme-specific inhibitors), Celecoxib, and Diclofenac were also utilized as reference drugs, the results are listed in Table 6. The results showed that the reference drugs; celecoxib and diclofenac exhibited COX-1inhibitory activity with IC50 15.8 and 0.21 μM respectively and COX-2 inhibitory activity with IC50 0.29 and 3.8 μM respectively. Also, the results showed COX-1 inhibitory activities of compounds 1, 2, 3, and 4 with IC50 0.98, 4.58, 11.23, and 17.86μM respectively. Compound 1 was the most effective COX-1 inhibitor in this study, with an IC50 of 0.98 M, which was 15 times greater than celecoxib's IC50 15.8 M. In comparison to the inhibition of COX-1 (IC50 > 0.98M), those compounds showed potent inhibition of COX-2 with 0.19,1.98, 10, and 8 M, respectively. Compound 1 was also found to be a more effective inhibitor of COX-2 than celecoxib IC50 0.29 M, with a greater activity IC50 0.19 M. The effects of compound 1 substituents at the 2 and 4-positions of a dinitro group boosted both COX-1 and COX-2 inhibitory action, with the latter having higher efficacy, resulting in a COX-2-selective inhibitor (SI = 5.157) 38 . In addition, the results showed that the highest COX-2 selectivity index S.I. of all compounds(1-4) is equal to 5.157,2.313,1.123, and 2.23 respectively which is higher than Diclofenac (S.I. = 0.055). i The chemical concentration necessary to produce 50% inhibition of COX-1or COX-2 for means of two measurements and deviation from the mean is 10% of the mean value is known as the IC50 value. Data are presented as mean ± SD (n = 3).

Conclusion:
To summarize, created a series of Schiff bases derived from salicylaldehyde as a cyclooxygenase inhibitor, estimated ADMET parameters in silico, and synthesized them. IR and CHNS elemental analyses were used to describe the title compounds. Compound 1 outperformed the conventional medications celecoxib and diclofenac in vitro in inhibiting COX-1 and COX-2. The findings revealed that the dinitro substitution of compound 1 was critical for cyclooxygenase inhibition. Furthermore, adding NH2 groups to compounds 3 and 4 reduced the inhibitory activity of cyclooxygenase. According to the ADME analysis, all produced compounds satisfied the Lipinski criterion and were absorbed by GIT. Through molecular docking studies, it was found that the most active molecule, compound1, fits into the target enzyme. The results indicated that compound 1 could be a lead molecule because this ligand showed the best computational and experimental results. The cyclooxygenase inhibitory activity of these compounds could be further improved to find a promising therapeutic candidate for the treatment of inflammatory diseases.