The strategy of rational approaches to the search for selective COX-2 inhibitors as potential antiinflammatory agents has been proposed and elaborated. It is based on the use of PASS-prediction and molecular docking. The choice of the basic structure of 4-amino-3-thio-1,2,4-triazole as a promising object of chemical modification has been substantiated. Using a modification of the primary molecule, a virtual library of S-derivatives of 5-substituted 4-amino(pyrrol)3-thio-4H-1,2,4-triazoles in the amount of 100 compounds (ten groups) has been obtained by introducing various pharmacophore fragments. Based on the analysis of the results of the PASS-prediction and molecular docking, six of the ten planned groups of compounds have been selected for the synthesis as promising selective COX-2 inhibitors. The reliability of the prediction results has already been confirmed for one of the promising group 4-amino-5-(pyridine-4-yl)-1,2,4-triazole (4Н)-3-yl-thioacetamides.

The synthesis of new structureswith the predicted activity is expedient to carry out in that class of chemicalcompounds where substances with a certain directed action have been alreadyfound (Oliveira et al., 2019).Scientists refer the heterocyclic system of 1,2,4-triazole to the privilegedstructure (“privileged scaffold”), since most of the derivatives of thisheterocyclic synthesized exhibit some pharmacological activity, including theantiinflammatory (Zhuang et al., 2017). The combination of factorsaffecting biochemical processes, as well as the practical absence of toxiceffects on the body, indicates the feasibility of further targeted search fornew biologically active substances among derivatives of 1,2,4-triazole (Mioc etal., 2017). Acareful study of the literature data on the spectrum of pharmacologicalproperties of the heterocyclic system of 1,2,4-triazole allows us toconfidently assert that the presence of this cycle in the structure ofsubstances determines the manifestation of the antiinflammatory activity (Moise etal., 2009). Previousresearch by Sondhi et al. (2007) obtained a large group of derivatives of1,2,4-triazole, pyridine, pyrrol, and other related heterocyclic compounds,among them selective cyclooxygenase-2 (COX-2), 5-lipoxygenase (5-LOX), andCOX-2/5-LOX double inhibitors were identified. Moreover, some studies suggestthat the presence of the 1,2,4-triazole cycle causes selective inhibition ofCOX-2 (Cai et al., 2015; Jiang etal., 2010).

In addition, 1,2,4-triazolederivatives are low-toxic, rather simple in synthesis and highly reactivesubstances. It allows to introduce various pharmacophore fragments into theirstructure (Khanage et al., 2012).Our analysis of the scientific literature has shown that despite a large numberof publications devoted to functional derivatives of 1,2,4-triazole the pharmacologicalpotential of this class of compounds at the present stage is not exhausted.

The compound of 4-amino-3-thio-1,2,4-triazolehas a high synthetic potential, significant opportunities in terms of theintroduction of pharmacophore fragments and, correspondingly, the expansion ofthe spectrum for searching the biological activity. The presence of a sulfuratom in the basic structure increases lipophilicity and, therefore, can improveabsorption and bioavailability of the compounds synthesized on its basis (Jin et al.,2007). Inaddition, thio and amino groups are reaction centers for the introduction ofadditional pharmacophores into the molecule.

The introduction of different pharmacophorefragments into the molecule by modifying the thio group and obtaining S-derivatives of 5-substituted4-amino(pyrrol)3-thio-4H-1,2,4-triazoles (groups of compounds I-X).This allows increasing the chances of finding new effective compounds in thisseries, as presented in Figure 1. The following modification have been planned as follows:

1.Alkylation of5-pyridine-4-amino(pyrrol)3-thio-1,2,4- triazole derivatives by haloalkanes(1-bromopropane, 1-bromohexane, 1-bromoheptan, 1-bromononane, 1-bromodecane) (I) in order to increase lipophilicityof the initial 5-pyridine-4-yl-4-amino-3-thio-4Н-1,2,4-triazole.

2.Introduction of the arylideneaniline fragment (II) to alkylderivatives of group (I), it willincrease the number of unsaturated bonds in the molecule and may increase theactivity.

3.Oxidation of 5-(pyridine-4)-4-amino-3-thio-1,2,4-triazolederivatives to alkyl sulfonyl derivatives (III)as in the structure of modern oxicams there is a sulfonyl group.

4.Introduction of the5-(pyridine-4)-4-amino-3-thio-1,2,4-triazole fragment of alkyl urea (IV) into the basic structure sincethere are literature data concerning the effect of the acetamide residue theincrease of the antiinflammatory activity.

5.Alkylation of5-(pyridine-4(2,3))-4-amino-3-thio-1,2,4-triazoles (V-VI) byα-chloroacetamide’s taking into account the literature data concerning theeffect of the acetamide residue on the increase of the antiinflammatoryactivity.

6.Alkylation of5-(furyl-2)-amino-3-thio-1,2,4-triazoles (VII)by α-chloroacetamide’s.

7.Replacement of the amino group in S-alkyl derivatives of5-(pyridine-2,4)-4-amino-3-thio-4Н-1,2,4-triazoles (VIII-IX) and5-(furyl-2)-4-amino-3-thio-4Н-1,2,4-triazoles (X) on the pyrrol residue.

X=N, Y=C X=C, Y=N

Figure 1. General formulas for the planned groups of new structures of5-substituted 4-amino(pyrrol)3-thio-4H-1,2,4-triazoles from compounds I to X

A more real possibility of a comprehensive study ofthe biological activity of substances is the use of new computer predictiontechnologies in silico and theirapplication to the assessment of the spectrum of activity of chemical compoundswith subsequent testing of the substances studied according to the results ofthe prediction (Parasuraman, 2011). We have proposed amethodology for assessing the activity using the existing calculation programs.

The logical and structural assessment of thepossible biological effect was performed using the Prediction of ActivitySpectra for Substances (PASS) computer system online (http://www.pharmaexpert.ru/PassOnline). This system predicts 565 types of the biological activity by thestructural formula of the chemical compound. These activities include the mainand side pharmacological effects, mechanisms of their implementation and toxicmanifestations, such as mutagenicity, carcinogenicity and teratogenicity (Lagunin et al., 2000).

The prediction result is presented in the form ofconfidence “to be active” – Pa and “to be inactive” – Pi by the types ofactivity. It should be noted that the data obtained using the PASS program canserve only as an indicative characteristic when selecting promising moleculesand help to conduct the primary sample of probable promising groups. It isknown that the Pa value reflects, first of all, the similarity of the moleculewith the most typical known drugs of the training sample (Assyl et al., 2014).

In order tooptimize the targeted search for COX-2 inhibitors as potential antiinflammatoryagents and substantiate the feasibility of the experimental screening for theantiinflammatory activity the docking studies were also conducted (Laube et al., 2016). Conducting the docking studieshas allowed us to investigate the affinity of a definite group of compounds tothis biological target, predict the ability of substances to inhibit thecatalytic activity of 1СХ2 and 6СОХ, i.e. to identify their inhibitor, which isa key link in the pathogenesis of the disease. The use of in silico methods also makes it possible to save laboratory animalsin the case of complete absence of affinity to potential biological targets (Leelananda& Lindert, 2016).

Generation of the3D-structure of the planned substances and their optimization using the methodof molecular mechanics MM+ and semi-empirical quantum mechanical method PM3

Objects of thisstudy were derivatives of4-amino-5-(pyridine-2(3)-yl)-1,2,4-triazole(4Н)-3-yl-thioacetamides;4-amino-5-(pyridine-4-yl)-1,2,4-triazole(4H)-3-yl-thioacetamides; and theirpyrrol derivatives of2-((4-amino-5-(furan-2-yl)-1,2,4-triazole(4h)-3-yl)-sulfonyl)-N-acetamides and their pyrrolderivatives, as well as the known selective COX-2 inhibitors (celecoxib, natriumdiclofenac) and non-selective inhibitor aspirin.

The choice ofantiinflammatory biological targets and determination of their active sites

Crystallographicmodels of COX-2 was obtained from Protein Data Bank (www.rcsb.org), with PDB ID 6COX and PDB ID 1СХ2(Kurumbail et al., 1997). Their difference lies in the different spatial groups of ligandbinding. In both crystallographic models of COX-2 PDB ID 1СХ2 (chains A, B, C,D) and PDB ID 6COX (chains A, B) the following ligands were isolated: selectiveCOX-2 inhibitor SC-558, protoporphyrin IX (contains Fe), N-acetyl-D-glucosamine in accordance with the literature data that1СХ2 and 6COX crystallographic models represent such spatial groups as P 2121 2 and I 2 2 2. (Kurumbail et al., 1997). These data indicate thedifference in protein conformations, and, therefore, the structural differenceof the binding regions; it affects the protein-ligand interactions. Thus, weconsidered it expedient to take into account the data obtained as a result ofthe use of both crystallographic forms of COX-2 in the docking studies. Afteraddition hydrogen atoms to 1СХ2 and 6СОХ molecules, removal of water molecules,and isolation of chains B, C, D and B, respectively, the active sites ofproteins 1CX2 were isolated as presented in Figure 2.

Figure 2. The active site of 1СХ2 protein

Conducting theactual molecular docking

The docking studies of hypothetical compounds wereconducted using a SCIGRESS software package (Fujitsu, Fukuoka, Japan with license742F6852C191). The parameter used is binding free energy (kcal/mol), where thevalue of binding free energy of all test compounds is compared with comparativecompounds such as celecoxib, aspirin, and natrium diclofenac. The geneticalgorithm, an evolutionary search algorithm applied to solve optimization andmodeling problems by sequential selection, combination and variation of theparameters studied was used for automatic docking. The use of the geneticalgorithm makes it possible to effectively study the entire available space forthe ligand (Torres et al., 2019).

According to the prediction results the probabilityof detection of the antiinflammatory activity in the group of S-alkyl derivatives of5-(pyridine-4-yl)-3-thio-4-amino-4Н-1,2,4-triazole (I) is insignificant, the average value of Ра does not exceed 0.39 forthe antiinflammatory activity and 0.32 for the analgesic activity. Afterreplacing the amino group in position 4 of the 1,2,4-triazole cycle on thesubstituted benzylidene fragment an insignificant increase in theantiinflammatory and analgesic activity of compounds in group II is predicted. After thismodification the highest probability of detection of the antiinflammatoryactivity is predicted for 4-methoxy, 4-nitro, and 4-butylphenyl derivatives;however, for these compounds, the Ра value does not exceed 0.4 for theantiinflammatory activity and 0.49 for the analgesic activity. The averagevalue of Ра ≤ 0.500, hence, the group is not promising for the synthesis.

Despite the fact that the presence of a sulfonylgroup can increase the antiinflammatory activity, which has been proven inmodern COX inhibitors including piroxicam, tenoxicam, celecoxib, etc. (Cordero et al., 2001), the probability of its manifestation is insignificant according tothe results of the computer prediction for sulfonyl derivatives of5-(pyridine-4-yl)-3-thio-4-amino-4Н-1,2,4-triazole (III). The Ра value does not exceed 0.35, in addition, there is aclear regularity: with an increase in the number of carbon atoms in alkylradicals, the Ра value decreases and, therefore, there is the probability ofdetecting the antiinflammatory activity. Therefore, compounds I to III were not tested by molecular docking.

With a low degree of probability (Ра ≤ 0.3), themanifestation of the antiinflammatory activity for group IV of thiourea derivatives of4-amino-3-thio-5-(pyridine-4-yl)-1,2,4-triazole(4Н) is predicted. According tothe data of the PASS computer program online, the antiinflammatory andanalgesic activities are predicted for those groups of compounds that contain thepyridine cycle in position 5 (the activity indices are in the range from 0.51to 0.66).

By the computer prediction data,4-amino-5-(pyridine-2(3)-yl)-1,2,4-triazole(4Н)-3-yl-thioacetamides (group V) are promising compounds. Furtherchemical modification by moving the nitrogen atom to the para-position of thepyridine cycle in4-amino-5-(pyridine-4-yl)-1,2,4-triazole(4H)-3-yl-thioacetamides – thestructure of general formula from group VI,and the introduction of the furyl fragment (group VII) were highly appropriate. For compounds in group VII, the indices of theantiinflammatory activity are the highest among all planned groups (0.63-0.65).But for compounds of these groups the indices of the analgesic activity do notsignificantly decrease and are in the range from 0.58 to 0.59.

Another variant of the expedient optimization wasthe chemical modification of the amino group in position 4. The analysis of thecomputer prediction data obtained for S-alkylderivatives, in which the amino group in 5-(pyridine-2,4)-4-amino3-thio-4Н-1,2,4-triazoles (VIII-IX) and 5-(furyl-2)-4-amino3-thio-4Н-1,2,4-triazoles (X) wasreplaced by a pyrrol residue showed that the greatest probability of activitywas also at the level of approximately 60%.

Thus, according to the results of the in siliсo studies among ten plannedgroups of S-derivatives of5-(R)-4-amino-3-thio-4H-1,2,4-triazole (groups of compounds I-X),only six groups that are likely to be antagonists of prostaglandins, and canexhibit antiexudative and analgesic effects have been selected as promisingcompounds (Chalenko & Syrovaya, 2017).

The dockingstudies of the sample for derivatives of 4-amino-5-(pyridine-2(3)-yl)-1,2,4-triazole(4Н)-3-yl-thioacetamides,4-amino-5-(pyridine-4-yl)-1,2,4-triazole (4Н)-3-yl-thioacetamides and theirpyrrol derivatives,2-((4-amino-5-(furan-2-yl)-1,2,4-triazole(4Н)-3-yl)-sulfonyl)-N-acetamides and their pyrrolderivatives were conducted by the method of a flexible molecular docking intothe active zone of the COX-2 enzyme (Mozziconacci et al., 2005). As a result of the docking studies the value of Consensus scoringfunctions, which assess certain characteristics of the ligand-protein complexand indicate the possibility of their comparison, has been obtained. Consensusfunctions allow forming a rating of compounds according to the values of allscoring functions and analyzing data on the choice of potentialagonists/antagonists of the biological target selected (Sliwoski et al., 2014). The average values of the calculated scoring functions of moleculardocking for promising groups of compounds are given in Table I.

Table I. The values of scoringfunctions for promising groups of compounds and reference drugs obtained whenconducting molecular docking

The values ofscoring functions for almost all compounds in complexes with the COX-2 enzymeexceed the values of these functions for aspirin and natrium diclofenac. Forgroup VI the average values ofscoring functions for binding free energy exceed the values for celecoxib.Visualization of docking results for the most promising compound of group VI is presented in Figure 3.

Figure 3. The promising compound in the protein active site (1СХ2crystallographic model)

The data obtained are an argument for studying theantiinflammatory activity for all six groups of compounds selected. Based onthese data it has been determined that 4-amino-5-(pyridine-4-yl)-1,2,4-triazole(4Н)-3-yl-thioacetamidesand their pyrrol derivatives are characterized by the highest level of affinitycalculated to all targets studied compared with other groups of compounds.Their affinity value is comparable to classical agonists, and it opens theopportunity of identifying new receptor agonists. The reliability of theprediction results has already been confirmed for one of the promising groups4-amino-5-(pyridine-4-yl)-1,2,4-triazole(4Н)-3-yl-thioacetamides. The plannedsubstances have been synthesized and tested for antiinflammatory activity,besides it has also been published scientifically (Chalenko et al., 2019).

In order to search for potential antiinflammatoryagents 1,2,4-triazol-3-thiones have been selected as promising objects ofchemical modification. The strategy of rational approaches to the search forselective COX-2 inhibitors as potential antiinflammatory agents has beenproposed and elaborated. It is based on the use of the PASS-prediction andmolecular docking. Based on the results of the PASS-prediction and moleculardocking, six of the ten planned groups of compounds have been selected for thesynthesis as promising selective COX-2 inhibitors.

The authors gratefully acknowledgethe financial support from the National University of Pharmacy. The authorshave declared no conflict of interest.

Assyl, B.,Georgiyants, V.A., & Perekhoda, L.O. (2014). Study of acyl substitution ondocking properties of substituted dihydrogercetines as anti-inflammatoryagents. Journal of Chemical andPharmaceutical Research, 6(4), 749-753.

Cai, H.,Huang, X., Xu, S., Shen, H., Zhang, P., Huang, Y., Jiang, J., Sun, Y., Jiang,B., Wu, X., Yao, H., & Xu, J. (2016). Discovery of novel hybrids ofdiaryl-1,2,4-triazoles and caffeic acid as dual inhibitors of cyclooxygenase-2and 5-lipoxygenase for cancer therapy. EuropeanJournal of Medicinal Chemistry, 108, 89-103. doi: 10.1016/j.ejmech.2015.11.013

Chalenko,N.M., Demchenko, A.M., & Syrovaya, G.O. (2019). Synthesis and biologicalactivity of 4-amino-5-(pyridin-4-yl)-1,2,4-triazole(4H)-3-il-thioacetamides. Actual problems of experimental and clinicalbiochemistry: materials of the international scientific-practical conference(p. 79). Kharkiv, Ukraine: Kharkiv National Medical University.

Chalenko,N.M. & Syrovaya, G.O. (2017). The feasibility of using the prediction ofactivity spectra for substances program for targeted synthesis of potentialnon-steroidal class 3-thio-1,2,4-triazoles. Bioethicsand Biosafety: Multidisciplinary Aspects: Materials of a Scientific andPractical Conference with International Participation dedicated to the 105thAnniversary of V.K. Viskovich (pp. 160-161). Kharkiv, Ukraine: KharkivNational Medical University.

Cordero, J.A., Camacho, M., Obach, R., Domenech,J., & Vila, L. (2001). In vitro based index of topical anti-inflammatoryactivity to compare a series of NSAIDs. EuropeanJournal of Pharmaceutics and Biopharmaceutics, 51(2), 135-142. doi: 10.1016/s0939-6411(00)00149-1

Jiang, B.,Zeng, Y., Li, M.J., Xu, J.Y., Zhang, Y.N., Wang, Q.J., Sun, N.Y., Lu, T., &Wu, X.M. (2010). Design, synthesis, and biological evaluation of1,5-diaryl-1,2,4-triazole derivatives as selective cyclooxygenase-2 inhibitors.Archiv der Pharmazie, 343(9),500-508. doi: 10.1002/ardp.200900227

Jin, J.Y., Zhang, L.X., Zhang, A.J., Lei, X.X., & Zhu, J.H. (2007).Synthesis and Biological Activity of Some Novel Derivatives of4-Amino-3-(D-galactopentitol-1-yl)-5-mercapto-1,2,4-triazole. Molecules, 12(8), 1596-1605. doi: 10.3390/12081596

Khanage,S.G., Raju, S.A., Mohite, P.B., & Pandhare, R.B. (2012). Synthesis andPharmacological Evaluation of Some New Pyrimidine Derivatives Containing1,2,4-Triazole. Advanced PharmaceuticalBulletin, 2(2), 213-222. doi: 10.5681/apb.2012.033

Kurumbail, R.G., Stevens, A.M., Gierse, J.K., McDonald, J.J., Stegeman,R.A., Pak, J.Y., Gildehaus, D., Miyashiro, J.M., Penning, T.D., Seibert, K.,Isakson, P.C., & Stallings, W.C. (1997). Correction: Structural basis forselective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature, 385, 555. doi: 10.1038/385555b0

Lagunin, A.,Stepanchikova, A., Filimonov, D., & Poroikov, V. (2000). PASS: predictionof activity spectra for biologically active substances. Bioinformatics, 16(8), 747-748. doi: 10.1093/bioinformatics/16.8.747

Laube, M., Kniess, T., & Pietzsch, J.(2016). Development of Antioxidant COX-2 Inhibitors as Radioprotective Agentsfor Radiation Therapy—A Hypothesis-Driven Review. Antioxidants, 5(2), 14. doi: 10.3390/antiox5020014

Leelananda,S.P. & Lindert, S. (2016). Computational methods in drug discovery. Beilstein Journal of Organic Chemistry, 12,2694-2718. doi: 10.3762/bjoc.12.267

Mioc, M., Soica, C., Bercean, V., Avram, S.,Balan-Porcarasu, M., Coricovac, D., Ghiulai, R., Muntean, D., Andrica, F.,Dehelean, C., Spandidos, D.A., Tsatsakis, A.M., & Kurunczi, L. (2017).Design, synthesis and pharmaco-toxicological assessment of5-mercapto-1,2,4-triazole derivatives with antibacterial and antiproliferativeactivity. International Journal ofOncology, 50(4), 1175-1183. doi: 10.3892/ijo.2017.3912

Moise, M., Sunel, V., Profire, L., Popa, M., Desbrieres, J., &Peptu, C. (2009). Synthesis and Biological Activity of Some New1,3,4-Thiadiazole and 1,2,4-Triazole Compounds Containing a PhenylalanineMoiety. Molecules, 14(7), 2621-2631. doi: 10.3390/molecules14072621

Mozziconacci, J.C., Arnoult, E., Bernard, P., Do, Q.T., Marot, C., &Morin-Allory, L. (2005). Optimization and validation of a docking-scoringprotocol; application to virtual screening for COX-2 inhibitors. Journal of Medicinal Chemistry, 48(4),1055-1068. doi: 10.1021/jm049332v

Oliveira, J.W.F., Rocha, H.A.O., de Medeiros, W.M.T.Q., & Silva,M.S. (2019). Application of Dithiocarbamates as Potential NewAntitrypanosomatids-Drugs: Approach Chemistry, Functional and Biological. Molecules, 24(15), 2806. doi: 10.3390/molecules24152806

Parasuraman, S. (2011). Prediction of activity spectra for substances. Journal of Pharmacology andPharmacotherapeutics, 2(1), 52-53. doi: 10.4103/0976-500X.77119

Sliwoski, G., Kothiwale, S., Meiler, J., &Lowe Jr., E.W. (2014). Computational Methods in Drug Discovery. Pharmacological Reviews, 66(1), 334-395. doi: 10.1124/pr.112.007336

Sondhi, S.M., Dinodia, M., Singh, J., &Rani, R. (2007). Heterocyclic Compounds as Anti-Inflammatory Agents. Current Bioactive Compounds, 3(2),91-108. doi: 10.2174/157340707780809554

Torres, P.H.M., Sodero, A.C.R., Jofily, P., & Silva Jr., F.P.(2019). Key Topics in Molecular Docking for Drug Design. International Journal of Molecular Sciences, 20(18), 4574. doi: 10.3390/ijms20184574

Zhuang, C.,Zhang, W., Sheng, C., Zhang, W., Xing, C., & Miao, Z. (2017). Chalcone: APrivileged Structure in Medicinal Chemistry. Chemical Reviews, 117(12), 7762-7810. doi: 10.1021/acs.chemrev.7b00020