Discovery of new thymol-3,4-disubstituted thiazole hybrids as dual COX-2/5-LOX inhibitors with in vivo proof

Abstract

New thymol–3,4-disubstitutedthiazole hybrids were synthesised as dual COX-2/5-LOX inhibitors. Compounds 6b, 6d, 6e, and 6f displayed in vitro inhibitory activity against COX-2 (IC₅₀= 0.037, 0.042, 0.046, and 0.039 µM) nearly equal to celecoxib (IC₅₀= 0.045 µM). 6b, 6d, and 6f showed SI (379, 341, and 374, respectively) higher than that of celecoxib (327). 6a–l elicited in vitro 5-LOX inhibitory activity higher than quercetin. 6a–f, 6i–l, 7a, and 7c possessed in vivo inhibition of formalin induced paw edoema higher than celecoxib. 6a, 6b, 6f, 6h–l, and 7b showed gastrointestinal safety profile as celecoxib and diclofenac sodium in the population of fasted rats. Induced fit docking and molecular dynamics simulation predicted good fitting of 6b and 6f without changing the packing and globularity of the apo protein. In conclusion, 6b and 6f achieved the target goal as multitarget inhibitors of inflammation.

Graphical Abstract

Keywords:

• thiazole
• anti-inflammatory
• dual COX-2/5-LOX inhibitors
• induced fit docking
• molecular dynamic simulation

1. Introduction

Inflammation leads to the release of mediators which result in various cardiovascular and renal problems¹. In the inflammatory process, cyclooxygenase (COX) and 5-lipoxygenase (5-LOX) enzymes produce various prostaglandins and leukotrienes (LTs) from arachidonic acid, respectively¹. The COX isoenzymes are endoplasmic reticulum-bound enzymes. The COX isoforms crystal structures are homologous. Each monomer consists of an N-terminal epidermal growth factor-like domain, a membrane-binding domain that anchors to the lipid bilayer, and a C-terminal catalytic domain that includes COX active site^2,3. Moreover, COX-1 is constitutive and is involved in homoeostatic processes in platelets, endothelium, and kidneys. Besides, COX-2 is constitutive in some tissues, like the brain and kidneys while it is inducible in inflammatory processes^4–9. On the other hand, LTs play a major role in inflammatory processes like allergic rhinitis, asthma, osteoarthritis, atopic dermatitis, gastric damage, ulceration ischaemia, atherosclerosis, septic shock, and tumorigenesis^10–23. In addition, LTC4 and LTD4 are potent contractors of blood vessels and bronchi smooth muscles and inducers of mucus secretion. Besides, LTB4 is a potent chemotactic and chemokinetic agent¹⁹. Non-steroidal anti-inflammatory drugs (NSAIDs) have been applied for the treatment of pain, inflammation, arthritis, and fever. NSAIDs inhibit COX enzymes, so they suppress the production of prostaglandins. Dual inhibition of COX-1 and COX-2 by classical NSAIDs leads to various side effects^24–28. Consequently, various selective COX-2 inhibitors like celecoxib, rofecoxib, and valdecoxib showed low gastrointestinal side effects, so they were clinically approved. However, these compounds were withdrawn from the market due to their cardiovascular side effects^29–32. Dual inhibition of COX-1/COX-2 upregulates 5-LOX pathway, so it increases LTs synthesis³³. Consequently, simultaneous inhibition of COX-2 and 5-LOX enzymes could decrease the side effects of anti-inflammatory candidates^34,35. On the other hand, 1,5-diarylpyrazole was a major pharmacophore in various selective COX-2 inhibitors like celecoxib I. Besides, thymol II, a natural phenol derivative, was reported as a COX-2 inhibitor with anti-inflammatory activities^36–38. In addition, Thymus vulgaris essential oil which contains a significant percentage of thymol was proved to suppress 5-LOX activity in human monocytic cell line (THP-1)³⁹. It is worth mentioning that, the hybridisation of 1,5-diarylpyrazole to thymol produced compounds (IIIa–c) possessing in vitro dual inhibitory activities against COX-2 and 5-LOX higher than references celecoxib and quercetin IV (COX-2 IC₅₀ for IIIa–c and celecoxib: 0.043, 0.045, 0.068, and 0.045 µM, respectively) (5-LOX IC₅₀ for IIIa–c and quercetin: 3.05, 1.58, 1.60, and 3.34 µM, respectively)⁴⁰. In addition, compounds IIIa–c possessed in vivo potent anti-inflammatory activity higher than celecoxib and gastrointestinal safety profile as celecoxib⁴⁰. Moreover, hybridisation between thymol and thiazolidinone produced compounds Va,b which elicited in vitro dual COX-2/5-LOX inhibitory activities (COX-2 IC₅₀: 0.058, 0.054 µM; 5-LOX IC₅₀: 3.61, 3.26 µM respectively)⁴¹. In addition, hybridisation between di-tert-butylphenol and thiazolidinone in darbufelone VI resulted in dual COX-2/5-LOX inhibition⁴². Furthermore, thiazole was considered bioisostere to pyrazole moiety, so various substituted thiazole derivatives were reported as COX-2 inhibitors, 5-LOX inhibitors and effective anti-inflammatory agents. Besides, 4-arylthiazole derivative VII showed 5-LOX inhibitory activity (IC₅₀ = 10 μM) with anti-inflammatory activities⁴³. Moreover, combination of 4-arylthiazole with thiazolidinone in compound VIII showed 34% COX-2 inhibition, 5-LOX IC₅₀, 54 µM, and 67.3% protection on in-vivo anti-inflammatory models⁴⁴. On the other hand, 3,4-diarylthiazole derivatives, IX and X revealed dual COX-2/5-LOX inhibition (COX-2 IC₅₀ = 0.08, 0.77 µM; 5-LOX IC₅₀ = 0.36, 0.58 µM, respectively)^45,46 (Figure 1).

Figure 1. Structures of compounds having COX-2, or 5-LOX inhibitory activities.

In the present investigation, 3,4-disubstitutedthiazole pharmacophore, a bioisostere to the COX-2 inhibitor pharmacophore 1,5-diarylpyrazole, was hybridised with thymol, a natural phenolic 5-LOX inhibitor, aiming to act as dual COX-2/5-LOX inhibitors. These compounds could possess in vivo anti-inflammatory activities with better safety profile and fewer side effects than traditional NSAIDs (Figure 2).

Figure 2. Design of thymol—3,4-disubstitutedthiazole hybrids as dual COX-2/5-LOX inhibitors.

2. Experimental

2.1. Chemistry

Griffin melting point apparatus or electrothermal capillary tube melting point apparatus in open glass capillaries were used to measure melting points and were all uncorrected. Perkin-Elmer 1430 Infra-red spectrophotometer and Schimadzu FT-IR Affinity-1 Spectrometer were used to record infra-red spectra (IR) using KBr discs, in the Central Laboratory, Faculty of Pharmacy, Alexandria University, Egypt and in Faculty of Pharmacy, Cairo University, Egypt, respectively. Bruker High Performance Digital FT-NMR Spectrometer Avance III 400 MHz was used to determine nuclear magnetic resonance (¹H-NMR and ¹³C-NMR) in Faculty of Pharmacy, Cairo University, Egypt using deuterated dimethylsulphoxide as a solvent. The data were reported as chemical shifts, or δ values (ppm) relative to tetramethylsilane (TMS) as internal standard. Signals are indicated by the following abbreviations: s = singlet, d = doublet, t = triplet, q = quartette, and m = multiplet. Gas chromatograph/mass spectrophotometer at Al-Azhar University (The Regional Centre of Microbiology and Biotechnology) was used to run electron impact mass spectra (EIMS). The most characteristic fragments were recorded as relative intensity %. Elemental microanalyses were performed at the microanalytical unit, Al-Azhar University (The Regional Centre of Microbiology and Biotechnology), Egypt. Thin-layer chromatography (TLC) on silica gel sheets (60 GF254, Merck) was used to monitor reaction progress. The spots were visualised by exposure to iodine vapour, or UV-lamp at λ 254 nm for few seconds.

Compounds 2^47–50, 3a–d^51,52, 4a–d⁴¹, and 5a–c^53–55 were prepared according to reported procedures.

2.1.1. General procedure for synthesis of 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3,4-disubstituted-2,3-dihydrothiazoles (6a-l)

A mixture of the corresponding hydrazine carbothioamide 4a–d (0.1 mol), appropriate phenacyl bromide 5a–c (0.1 mol) and fused sodium acetate (0.15 mol) in glacial acetic acid (10 ml) was refluxed for 5–8 h. Then the reaction mixture was cooled to RT, poured onto crushed ice, the formed precipitate was filtered, washed with water, air-dried, and then recrystallized from ethanol.

2.1.1.1. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3,4-diphenyl-2,3-dihydrothiazole; 6a

Yellow crystals yield 82%, m.p0.175–177 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1589.34, 1573.91(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.88 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.25-3.28(m, 1H, CH of isopropyl); 6.67(d, 1H, J = 7.68 Hz, C5 of thymol); 6.72(s, 1H, CH of thiazole); 7.10(d, 1H, J = 7.68 Hz, C4 of thymol); 7.18-7.20(m, 3H, C2, C4, C6 of N-phenyl);7.24-7.26(m, 3H, C3, C4, C5 of phenyl); 7.31-7.35 (m, 2H, C3, C5 of N-phenyl); 7.38-7.41(m, 2H, C2, C6 of phenyl); 8.52(s, 1H, CH of CH=N); 12.19(s, 1H, OH, D₂O exchangeable).¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 19.29(CH₃ of 6-methylthymol); 22.90(2CH₃ of isopropyl); 26.32(CH of isopropyl); 101.17(CH of thiazole); 116.15(C1 of thymol); 121.40(C5 of thymol);127.92(C4 of phenyl); 128.72(3 C, C2, C4, C6 of N-phenyl); 128.92(2 C, C2, C6 of phenyl); 129.11 (2 C, C3, C5 of phenyl); 129.32(2 C, C3, C5 of N-phenyl); 129.54(C4 of thymol); 130.88(C1 of phenyl); 133.41(C3 of thymol); 136.02(C1 of N-phenyl); 137.65(C6 of thymol); 140.80(C4 of thiazole); 153.04(CH of CH = N); 156.46(C2 of thymol); 169.07(C = N of thiazole). EIMS, m/z (relative abundance %): 428 (M^+.+1) (29.56); 427 (M^+.) (100.00); 426 (3.5); 413 (3.98); 412 (13.74); 411 (6.69); 410 (12.94); 384 (2.43). Analysis Calcd for C₂₆H₂₅N₃OS (427.57): C, 73.04; H, 5.89; N, 9.83; S, 7.50. Found: C, 72.88; H, 6.11; N, 9.97; S, 7.62.

2.1.1.2. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-chlorophenyl)-4-phenyl -2,3-dihydrothiazole; 6b

Pale yellow crystals yield 87%, m.p0.225–227 °C. IR (KBr, v cm⁻¹): 3488.44(OH); 1593.20, 1573.91(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.88 Hz, 2CH₃ of isopropyl); 2.31(s, 3H, CH₃ of 6-methylthymol); 3.27-3.28(m, 1H, CH of isopropyl); 6.67(d, 1H, J = 7.72 Hz, C5 of thymol); 6.74(s, 1H, CH of thiazole); 7.11(d, 1H, J = 7.72 Hz, C4 of thymol); 7.24-7.29(m, 5H, C3, C4, C5 of phenyl, C5, C6 of chlorophenyl);7.42-7.44(m, 2H, C2, C6 of phenyl); 7.54-7.56 (m, 1H, C4 of chlorophenyl); 7.67-7.69(m, 1H, C3 of chlorophenyl); 8.50(s, 1H, CH of CH=N); 12.18(s, 1H, OH, D₂O exchangeable). ¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 19.29(CH₃ of 6-methylthymol); 22.83 (2CH₃ of isopropyl); 26.34(CH of isopropyl); 100.99(CH of thiazole); 116.06(C1 of thymol); 121.43(C5 of thymol);128.06(C4 of chlorophenyl); 128.75, 128.81(5 C, C2, C3, C5, C6 of phenyl, C6 of chlorophenyl); 129.48 (C4 of thymol); 130.32(C4 of phenyl); 130.68(C2 of chlorophenyl); 131.39(C1 of phenyl); 132.70(2 C, C1, C3 of chlorophenyl); 133.42(C3 of thymol); 135.17(C5 of chlorophenyl); 136.18(C6 of thymol); 140.40(C4 of thiazole); 153.49(CH of CH = N); 156.51(C2 of thymol); 168.36(C = N of thiazole). EIMS, m/z (relative abundance %): 463 (M^+.+2) (39.97); 462 (M^+.+1) (32.90); 461 (M^+.) (100.00); 448 (6.04); 447 (5.51); 446 (16.76); 445 (4.07); 444 (7.79); 428 (2.25); 427 (4.30); 426 (13.11); 409 (4.24). Analysis Calcd for C₂₆H₂₄ClN₃OS (461.13): C, 67.59; H, 5.24; N, 9.10; S, 6.94. Found: C, 67.73; H, 5.47; N, 9.26; S, 6.89.

2.1.1.3. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-pyridyl)-4-phenyl-2,3-dihydrothiazole; 6c

Yellow crystals yield 85%, m.p0.179–181 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1566.20, 1597.06(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.22-3.279(m, 1H, CH of isopropyl); 6.68(d, 1H, J = 7.72 Hz, C5 of thymol); 6.74(s, 1H, CH of thiazole); 7.11(d, 1H, J = 7.72 Hz, C4 of thymol); 7.15-7.17(m, 2H, C2, C6 of phenyl);7.22-7.25(m, 3H, C3, C4, C5 of phenyl); 7.38-7.41 (m, 1H, C4 of 2-pyridyl); 7.74 (d, 1H, J = 7.96 Hz, C3 of 2-pyridyl); 7.97-8.01(m, 1H, C5 of 2-pyridyl); 8.40(d, 1H, J = 3.8 Hz, C6 of 2-pyridyl); 8.54(s, 1H, CH of CH=N); 12.15(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 429 (M^+.+1) (42.82); 428 (M^+.) (100.00); 413 (4.95); 411 (4.09). Analysis Calcd for C₂₅H₂₄N₄OS (428.55): C, 70.07; H, 5.65; N, 13.07; S, 7.48. Found: C, 70.34; H, 5.82; N, 12.89; S, 7.60.

2.1.1.4. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(4-methoxyphenyl)-4-phenyl-2,3-dihydrothiazole; 6d

Khaki crystals yield 85%, m.p0.195–198 °C. IR (KBr, v cm⁻¹): 3425.58(OH); 1589.34, 1573.91(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.18-3.28(m, 1H, CH of isopropyl); 3.75(s, 3H, CH₃ of OCH₃); 6.67(d, 1H, J = 7.68 Hz, C5 of thymol); 6.68(s, 1H, CH of thiazole); 6.92(d, 2H, J = 8.92 Hz, C2, C6 of p-methoxyphenyl); 7.10(d, 1H, J = 7.68 Hz, C4 of thymol); 7.20-7.28(m, 7H, C2, C3, C4, C5, C6 of phenyl, C3, C5 of p-methoxy phenyl); 8.52(s, 1H, CH of CH=N); 12.21(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 458 (M^+.+1) (34.60); 457 (M^+.) (100.00); 443 (3.19); 442 (10.82); 441 (5.30); 440 (9.56); 281 (2.5). Analysis Calcd for C₂₇H₂₇N₃O₂S (457.59): C, 70.87; H, 5.95; N, 9.18; S, 7.01. Found: C, 71.06; H, 6.12; N, 9.41; S, 7.08.

2.1.1.5. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-phenyl-4-(4-bromophenyl) -2,3-dihydrothiazole; 6e

Linen crystals yield 88%, m.p0.211–213 °C. IR (KBr, v cm⁻¹): 3417.86(OH); 1589.34 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.28-3.31(m, 1H, CH of isopropyl); 6.67(d, 1H, J = 7.72 Hz, C5 of thymol); 6.78(s, 1H, CH of thiazole); 7.10(d, 1H, J = 7.72 Hz, C4 of thymol); 7.14(d, 2H, J = 8.52 Hz, C2, C6 of bromophenyl); 7.33-7.37(m, 3H, C3, C4, C5 of phenyl); 7.41(d, 2H, J = 7.64 Hz, C2, C6 of phenyl); 7.46(d, 2H, J = 8.52 Hz, C3, C5 of bromophenyl); 8.52(s, 1H, CH of CH=N); 12.17(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 509 (M^+.+2) (10.04); 508 (M^+.+1) (32.16); 507 (M^+.) (100.00); 506 (37.01); 505 (9.09); 492 (12.74); 491 (9.18); 490 (22.40); 489 (6.10); 488 (11.02); 331 (8.19); 329 (4.19). Analysis Calcd for C₂₆H₂₄BrN₃OS (506.46): C, 61.66; H, 4.78; N, 8.30; S, 6.33. Found: C, 61.52; H, 4.89; N, 8.56; S, 6.49.

2.1.1.6. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-chlorophenyl)-4-(4-bromophenyl)-2,3-dihydrothiazole; 6f

Yellow crystals yield 85%, m.p0.185–188 °C. IR (KBr, v cm⁻¹): 3479.58(OH); 1593.20 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.84 Hz, 2CH₃ of isopropyl); 2.31(s, 3H, CH₃ of 6-methylthymol); 3.27-3.31(m, 1H, CH of isopropyl); 6.67(d, 1H, J = 7.76 Hz, C5 of thymol); 6.81(s, 1H, CH of thiazole); 7.10(d, 1H, J = 7.76 Hz, C4 of thymol); 7.19(d, 2H, J = 8.48 Hz, C2, C6 of bromophenyl); 7.44-7.49(m, 4H, C3, C5 of bromophenyl, C4, C6 of chlorophenyl); 7.56-7.59(m, 1H, C5 of chlorophenyl); 7.70-7.73(m, 1H, C3 of chlorophenyl); 8.50(s, 1H, CH of CH=N); 12.16(s, 1H, OH, D₂O exchangeable). ¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 19.29(CH₃ of 5-methyl thymol); 22.92(2CH₃ of isopropyl); 26.35(CH of isopropyl); 101.72(CH of thiazole); 116.02(C1 of thymol); 121.44(C5 of thymol); 122.98(C4 of bromophenyl); 128.11 (C6 of chlorophenyl); 128.83(C4 of chlorophenyl); 129.51 (C4 of thymol); 130.76(C2 of chlorophenyl); 130.82(2 C, C2, C6 of chlorophenyl); 131.54(C1 of bromophenyl); 131.79(2 C, C3, C5 of bromophenyl); 132.57(C3 of chlorophenyl); 132.71(C5 of chlorophenyl); 133.43(C3 of thymol); 134.92 (C1 of chlorophenyl); 136.22(C6 of thymol); 139.21(C4 of thiazole); 153.69(CH of CH = N); 156.53(C2 of thymol); 168.20(C = N of thiazole). EIMS, m/z (relative abundance %): 545 (M^+.+4)(8.05); 543 (M^+.+2) (22.45); 542(M^+.+1) (34.58); 541 (M^+.) (100.00); 540 (29.41); 539 (76.77); 526 (14.02); 525 (5.50); 524 (14.79); 506 (8.73); 504 (8.18). Analysis Calcd for C₂₆H₂₃BrClN₃OS (540.90): C, 57.73; H, 4.29; N, 7.77; S, 5.93. Found: C, 58.01; H, 4.42; N, 7.92; S, 6.04.

2.1.1.7. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-pyridyl)-4-(4-bromophenyl)-2,3-dihydrothiazole; 6 g

Light yellow crystals yield 79%, m.p0.112–115 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1589.34, 1570.06 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 2.33(s, 3H, CH₃ of 6-methylthymol); 3.28-3.32(m, 1H, CH of isopropyl); 6.68(d, 1H, J = 7.76 Hz, C5 of thymol); 6.81(s, 1H, CH of thiazole); 7.10-7.13(m, 3H, C4 of thymol, C3, C5 of bromophenyl); 7.40-7.47(m, 3H, C2, C6 of bromophenyl, C4 of 2-pyridyl); 7.78(d, 1H, J = 7.96 Hz, C3 of 2-pyridyl); 8.00-8.04(m, 1H, C5 of 2-pyridyl); 8.40 (d, 1H, J = 4.8 Hz, C6 of 2-pyridyl); 8.55(s, 1H, CH of CH=N); 12.12(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 510 (M^+.+2) (9.15); 509 (M^+.+1) (29.52); 508 (M^+.) (100.00); 507 (33.70); 506 (97.12); 493 (11.29); 492 (7.76); 491 (19.78); 490 (5.28); 489 (10.16); 332 (3.11). Analysis Calcd for C₂₅H₂₃BrN₄OS (507.45): C, 59.17; H, 4.57; N, 11.04; S, 6.32. Found: C, 59.39; H, 4.65; N, 11.32; S, 6.49.

2.1.1.8. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(4-methoxyphenyl)-4-(4-bromophenyl)-2,3-dihydrothiazole; 6h

Cream crystals yield 86%, m.p0.255–258 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1589.34 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.88 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.28-3.30(m, 1H, CH of isopropyl); 3.76(s, 3H, CH₃ of OCH₃); 6.67(d, 1H, J = 7.68 Hz, C5 of thymol); 6.74(s, 1H, CH of thiazole); 6.94(d, 2H, J = 8.80 Hz, C2, C6 of p-methoxyphenyl); 7.10(d, 1H, J = 7.68 Hz, C4 of thymol); 7.16(d, 2H, J = 8.40 Hz, C2, C6 of bromophenyl); 7.25(d, 2H, J = 8.80 Hz, C3, C5 of p-methoxy phenyl); 7.48(d, 2H, J = 8.40 Hz, C3, C5 of bromophenyl); 8.52(s, 1H, CH of CH=N); 12.18(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 539 (M^+.+2) (8.58); 538 (M^+.+1) (29.34); 537 (M^+.) (100.00); 536 (32.61); 535 (91.43); 522 (14.57); 521 (11.27); 520 (28.32); 518 (13.43); 363 (25.64); 362 (30.19); 361 (84.06); 360 (29.54); 359 (62.40); 348 (12.34); 347 (9.25); 346 (12.46). Analysis Calcd for C₂₇H₂₆BrN₃O₂S (536.49): C, 60.45; H, 4.89; N, 7.83; S, 5.98. Found: C, 60.71; H, 4.95; N, 7.98; S, 6.12.

2.1.1.9. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-phenyl-4-(4-methoxyphenyl)-2,3-dihydrothiazole; 6i

Light yellow crystals yield 84%, m.p0.208–210 °C. IR (KBr, v cm⁻¹): 3483.44(OH); 1589.34 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 2.31(s, 3H, CH₃ of 6-methylthymol); 3.28-3.30(m, 1H, CH of isopropyl); 3.70(s, 3H, CH₃ of OCH₃); 6.60(s, 1H, CH of thiazole); 6.67(d, 1H, J = 7.72 Hz, C5 of thymol); 6.79(d, 2H, J = 8.80 Hz, C3, C5 of p-methoxyphenyl); 7.09-7.14(m, 3H, C4 of thymol, C2, C6 of p-methoxyphenyl); 7.30-7.43 (m, 5H, C2, C3, C4, C5, C6 of phenyl,); 8.51(s, 1H, CH of CH=N); 12.21(s, 1H, OH, D₂O exchangeable). ¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 19.29(CH₃ of 6-methylthymol); 22.90(2CH₃ of isopropyl); 26.32(CH of isopropyl); 55.59(CH₃ of OCH₃); 99.91(CH of thiazole); 114.13(2 C, C3, C5 of p-methoxyphenyl); 116.17(C1 of thymol); 121.38(C5 of thymol); 123.15(C4 of phenyl); 127.86(C1 of p-methoxyphenyl); 128.71(2 C, C2, C6 of phenyl); 129.39(2 C, C3, C5 of phenyl); 129.54(C4 of thymol); 130.38(2 C, C2, C6 of p-methoxyphenyl); 133.41(C3 of thymol); 135.97(C1 of phenyl); 137.72(C6 of thymol); 140.65(C4 of thiazole); 152.83(CH of CH = N); 156.44(C2 of thymol); 159.79(C4 of p-methoxyphenyl); 169.14(C = N of thiazole). EIMS, m/z (relative abundance %): 458 (M^+.+1) (33.22); 457 (M^+.) (100.00); 442 (10.94); 441 (6.02); 440 (10.72); 382 (2.45); 381 (4.81). Analysis Calcd for C₂₇H₂₇N₃O₂S (457.59): C, 70.87; H, 5.95; N, 9.18; S, 7.01. Found: C, 71.23; H, 6.12; N, 9.40; S, 7.06.

2.1.1.10. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-chlorophenyl)-4-(4-methoxyphenyl)-2,3-dihydrothiazole; 6j

Light yellow crystals yield 85%, m.p0.183–184 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1597.06(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.84 Hz, 2CH₃ of isopropyl); 2.31(s, 3H, CH₃ of 6-methylthymol); 3.28-3.32(m, 1H, CH of isopropyl); 3.70(s, 3H, CH₃ of OCH₃); 6.63(s, 1H, CH of thiazole); 6.67(d, 1H, J = 7.72 Hz, C5 of thymol); 6.81(d, 2H, J = 8.76 Hz, C3, C5 of p-methoxyphenyl); 7.10(d, 1H, J = 7.72 Hz, C4 of thymol); 7.16(d, 2H, J = 8.76 Hz, C2, C6 of p-methoxyphenyl); 7.41-7.46 (m, 2H, C4, C6 of chlorophenyl); 7.55-7.57(m, 1H, C5 of chlorophenyl); 7.65-7.67(m, 1H, C3 of chlorophenyl); 8.48(s, 1H, CH of CH=N); 12.20(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 493 (M^+.+2) (39.29); 492 (M^+.+1) (30.70); 491 (M^+.) (100.00); 478 (6.28); 477 (6.12); 476 (18.69); 475 (4.65); 474 (8.89); 458 (3.55); 457 (6.61); 456 (17.99); 441 (2.94); 440 (3.48); 439 (9.87). Analysis Calcd for C₂₇H₂₆ClN₃O₂S (492.03): C, 65.91; H, 5.33; N, 8.54; S, 6.52. Found: C, 66.12; H, 5.42; N, 8.82; S, 6.63.

2.1.1.11. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3-(2-pyridyl)-4-(4-methoxyphenyl)-2,3-dihydrothiazole; 6k

Lemon chiffon crystals yield 85%, m.p0.150–153 °C. IR (KBr, v cm⁻¹): 3483.44(OH); 1597.06, 1570.08(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.88 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.28-3.32(m, 1H, CH of isopropyl); 3.70(s, 3H, CH₃, OCH₃); 6.62(s, 1H, CH of thiazole); 6.68(d, 1H, J = 7.76 Hz, C5 of thymol); 6.79(d, 2H, J = 8.84 Hz, C3, C5 of p-methoxyphenyl); 7.07-7.12(m, 3H, C4 of thymol, C2, C6 of p-methoxyphenyl); 7.40-7.43(m, 1H, C4 of 2-pyridyl); 7.69(d, 1H, J = 7.92 Hz, C3 of 2-pyridyl); 7.96-8.01(m, 1H, C5 of 2-pyridyl); 8.43-8.45(d, 1H, J = 8 Hz, C6 of 2-pyridyl); 8.52(s, 1H, CH of CH=N); 12.16(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 459 (M^+.+1) (26.96); 458 (M^+.) (100.00); 443 (14.74); 442 (7.95); 441 (18.12); 352 (6.21). Analysis Calcd for C₂₆H₂₆N₄O₂S (458.58): C, 68.10; H, 5.72; N, 12.22; S, 6.99. Found: C, 68.34; H, 5.83; N, 12.47; S, 7.14.

2.1.1.12. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-3,4-di(4-methoxyphenyl)-2,3-dihydrothiazole; 6 l

Ivory crystals yield 80%, m.p0.235–238 °C. IR (KBr, v cm⁻¹): 3479.58(OH); 1589.34, 1573.91 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.19(d, 6H, J = 6.84 Hz, 2CH₃ of isopropyl); 2.32(s, 3H, CH₃ of 6-methylthymol); 3.28-3.30(m, 1H, CH of isopropyl); 3.71, 3.76(2s, 6H, 2OCH₃); 6.57(s, 1H, CH of thiazole); 6.67(d, 1H, J = 7.68 Hz, C5 of thymol); 6.82(d, 2H, J = 8.64 Hz, C2, C6 of N-p-methoxyphenyl); 6.93(d, 2H, J = 8.80 Hz, C3, C5 of p-methoxyphenyl); 7.10(d, 1H, J = 7.68 Hz, C4 of thymol); 7.13(d, 2H, J = 8.64 Hz, C3, C5 of N-p-methoxyphenyl); 7.23(d, 2H, J = 8.80 Hz, C2, C6 of p-methoxyphenyl); 8.50(s, 1H, CH of CH=N); 12.22(s, 1H, OH, D₂O exchangeable). EIMS, m/z (relative abundance %): 488 (M^+.+1) (42.70); 487 (M^+.) (100.00); 472(10.10); 471 (4.84); 470 (10.74). Analysis Calcd for C₂₈H₂₉N₃O₃S (487.62): C, 68.97; H, 5.99; N, 8.62; S, 6.57. Found: C, 69.21; H, 6.15; N, 8.90; S, 6.72.

2.1.2. General procedure for synthesis of 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-4-methyl-3-substituted-2,3-dihyrothiazoles (7a-c)

A mixture of the corresponding hydrazine carbothioamide 4a–c (1 mmol) and chloroacetone (1 mmol) was refluxed with fused sodium acetate (1.5 mmol) in glacial acetic acid for 6–8 h, cooling to RT, pouring on to crushed ice, then filtering the precipitate, air-dried, recrystallization from ethanol.

2.1.2.1. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-4-methyl-3-phenyl-2,3-dihyrothiazole; 7a

Dark goldenrod crystals yield 60%, m.p0.100–101 °C. IR (KBr, v cm⁻¹): 3406.29(OH); 1589.34, 1570.06(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.18(d, 6H, J = 6.92 Hz, 2CH₃ of isopropyl); 1.85(s, 3H, CH₃ of thiazole); 2.28(s, 3H, CH₃ of 6-methylthymol); 3.26-3.29(m, 1H, CH of isopropyl); 6.34(s, 1H, CH of C5 of thiazole); 6.64(d, 1H, J = 7.72 Hz, C5 of thymol); 7.06(d, 1H, J = 7.72 Hz, C4 of thymol); 7.44(d, 2H, J = 7.24 Hz, C2, C6 of phenyl); 7.48-7.52(m, 1H, C4 of phenyl);7.55-7.58(m, 2H, C3, C5 of phenyl); 8.45(s, 1H, CH of CH=N); 12.24(s, 1H, OH, D₂O exchangeable). ¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 14.99(CH₃ of thiazole); 19.26(CH₃ of 6-methylthymol); 22.86(2CH₃ of isopropyl); 26.34(CH of isopropyl); 97.41(CH of C5 of thiazole); 116.22(C1 of thymol); 121.30(C5 of thymol); 127.64(C4 of phenyl); 129.19(2 C, C2, C6 of phenyl); 129.37(C4 of thymol); 130.09(2 C, C3, C5 of phenyl); 133.35(C3 of thymol); 135.78(C1 of phenyl); 136.59(C6 of thymol); 137.04(C4 of thiazole); 152.17(CH of CH = N); 156.39(C2 of thymol); 169.46(C = N of thiazole). Analysis Calcd for C₂₁H₂₃N₃OS (365.50): C, 69.01; H, 6.34; N, 11.50; S, 8.77. Found: C, 69.28; H, 6.43; N, 11.15; S, 8.59.

2.1.2.2. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-4-methyl-3-(2-chlorophenyl)-2,3-dihyrothiazole; 7b

Chocolate crystals yield 58%, m.p0.169–171 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1587.70 (C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.18(d, 6H, J = 6.88 Hz, 2CH₃ of isopropyl); 1.81(s, 3H, CH₃ of thiazole); 2.29(s, 3H, CH₃ of 6-methylthymol); 3.28-3.31(m, 1H, CH of isopropyl); 6.38(s, 1H, CH of C5 of thiazole); 6.64(d, 1H, J = 7.68 Hz, C5 of thymol); 7.07(d, 1H, J = 7.68 Hz, C4 of thymol); 7.56-7.59(m, 2H, C4, C6 of chlorophenyl); 7.64-7.65(m, 1H, C5 of chlorophenyl); 7.73-7.75(m, 1H, C3 of chlorophenyl); 8.45(s, 1H, CH of CH=N); 12.22(s, 1H, OH, D₂O exchangeable). Analysis Calcd for C₂₁H₂₂ClN₃OS (399.94): C, 63.07; H, 5.54; N, 10.51; S, 8.02. Found: C, 62.93; H, 5.36; N, 10.24; S, 7.93.

2.1.2.3. 2-((2-Hydroxy-3-isopropyl-6-methylbenzylidene)hydrazono)-4-methyl-3-(4-methoxyphenyl)-2,3-dihyrothiazole; 7c

Saddle brown crystals yield 52%, m.p0.158–160 °C. IR (KBr, v cm⁻¹): 3421.72(OH); 1589.34, 1573.91(C = N). ¹H-NMR (400 MHz, DMSO-d₆δ ppm): 1.12(d, 6H, J = 6.84 Hz, 2CH₃ of isopropyl); 1.92(s, 3H, CH₃ of thiazole); 2.04(s, 3H, CH₃ of 6-methylthymol); 3.22-3.27(m, 1H, CH of isopropyl); 3.32(s, 3H, CH₃ of OCH₃); 6.34(s, 1H, CH of C5 of thiazole); 6.59(d, 1H, J = 7.64 Hz, C5 of thymol); 6.66(d, 2H, J = 8.76 Hz, C2, C6 of p-methoxyphenyl); 6.88(d, 1H, J = 7.64 Hz, C4 of thymol); 7.09(d, 2H, J = 8.76 Hz, C3, C5 of p-methoxyphenyl); 8.65(s, 1H, CH of CH=N); 12.20(s, 1H, OH, D₂O exchangeable). ¹³C-NMR (100 MHz, DMSO-d₆δ ppm): 14.90(CH₃ of thiazole); 19.27(CH₃ of 6-methylthymol); 22.88(2CH₃ of isopropyl); 26.35(CH of isopropyl); 57.10(CH₃ of OCH₃); 97.91(CH of C5 of thiazole); 115.71(2 C, C3, C5 of p-methoxyphenyl); 118.82(C1 of thymol); 121.35(C5 of thymol); 127.86(C1 of p-methoxyphenyl); 128.58(C4 of thymol); 133.40(2 C, C2, C6 of p-methoxyphenyl); 133.58(C3 of thymol); 136.42(C6 of thymol); 139.89(C4 of thiazole); 150.41(CH of CH = N); 156.22(C2 of thymol); 156.46(C4 of p-methoxyphenyl); 168.83(C = N of thiazole). Analysis Calcd for C₂₂H₂₅N₃O₂S (395.52): C, 66.81; H, 6.37; N, 10.62; S, 8.11. Found: C, 67.12; H, 6.21; N, 10.30; S, 8.37.

2.2. Biological screening

2.2.1. In vitro COX-1 and COX-2 inhibitory assay

Compounds 6a–l and 7a–c were tested for their in vitro COX-1 and COX-2 inhibitory activities using Cayman colorimetric COX (ovine) inhibitor screening assay kit (Catalog No. 560131) supplied by Cayman chemicals, Ann Arbour, MI, USA according to reported method^41,56. (Page: S34, Supp. file)

2.2.2. In vitro 5-LOX inhibitory assay

Compounds 6a–l and 7a–c were tested for their in vitro 5-LOX inhibitory activity using Abnova 5-LOX inhibitor screening assay kit (Catalog No. 760700) according to reported method ^41,57. (Page: S34–35, Supp. file)

2.2.3. In vivo anti-inflammatory activity

2.2.3.1. Formalin-induced paw edoema

Compounds 6a–l and 7a–c were evaluated for their in vivo anti-inflammatory activity using formalin-induced paw edoema screening protocol as an acute inflammation model using celecoxib and diclofenac sodium as reference drugs according to reported procedures^41,58,59 (Approved by AlexU-IACUC)⁶⁰. (Page: S37-39, Supp. file)

2.2.3.2. Gastric ulcerogenic activity

Acute gastric ulcerogenic effect was evaluated in adult female Wistar rats according to reported procedures^41,59,61 (Approved by AlexU-IACUC)⁶⁰. (Page: S40, Supp. file)

2.3. Molecular modelling

2.3.1. Induced fit docking to COX-2 and 5-LOX

Molecular Operating Environment (MOE 2016.08) software (Chemical Computing Group, Montreal, Canada)⁶² was used to perform the molecular modelling studies. The proteins crystal structures were downloaded from the Protein Data Bank (PDB) website. (Page: S41, Supp. file)

2.3.2. Molecular dynamics simulation

Molecular dynamics simulation of protein–ligand complexes was performed by using WebGRO^63,64. (Page: S41-44, Supp. file)

3. Results and discussion

3.1. Chemistry

Scheme 1 outlined the synthesis of the target compounds. It started with the synthesis of 2-formylthymol (2) by reaction of thymol (1) with magnesium methoxide and paraformaldehyde^47–50. The 2-formylthymol (2) was then reacted with various substituted thiosemicarbazides (3a–d)^51,52 in the presence of a few drops of glacial acetic acid in ethanol, to give thiosemicarbazones (4a–d)⁴¹. Compounds 4a-d were further cyclized by refluxing with α-haloketone (namely, substituted phenacyl bromide, or α-chloroacetone) in glacial acetic acid containing fused sodium acetate to afford the corresponding 3,4-disubstitutedthiazole derivatives 6a–l and 7a–c, respectively. Spectroscopic data have been utilised to elucidate the chemical structures of the target thiazole derivatives. The presence of C5-H of thiazole ring in compounds 6a–l was demonstrated by ¹H-NMR and ¹³C-NMR spectroscopy which exhibited singlet signal at (6.60 − 6.74) ppm for one proton, and (99.91–101.72) ppm, respectively. The structures of the prepared compounds 7a–c were confirmed by studying the IR, ¹H-NMR as well as ¹³C-NMR spectra for these compounds. IR spectra were characterised by the absence of absorption bands assigned to (NH) and (C = S) groups, and the presence of absorption bands assigned to (C = N), in addition to (OH) group at expected absorption regions. ¹H-NMR spectra of compounds 7a–c were characterised by the absence of signals assigned to NH protons, and the presence of singlet signals assigned to protons of CH₃ at C4 of thiazole and proton at C5 of thiazole at 1.81–1.85 ppm and 6.30–6.38 ppm, respectively. ¹³C-NMR spectra of compounds 7a and 7c showed the presence of signals of CH₃ at C4 of thiazole, C5 of thiazole and C = N at 137–138.36 ppm, 96.94–97.41 ppm and 169.46–169.73 ppm, respectively.

Scheme 1. Synthesis of the target thymol—3,4-disubstitutedthiazole hybrids.

3.2. Biological screening

3.2.1. In vitro cox-1 and cox-2 inhibitory assay

Compounds 6a–6l and 7a–7c were screened for their ability to inhibit COX-1 and COX-2 enzymes in vitro. The results (Table 1) revealed that all compounds showed lower inhibitory activity against COX-1 with IC₅₀ 8.01 − 14.58 µM compared to diclofenac sodium (IC₅₀ = 3.9 µM). Among the tested compounds, compound 7a was the most active against COX-1. In addition, compounds 6b, 6d, 6e, and 6f showed COX-2 inhibitory activities nearly equal to that of celecoxib. Moreover, compounds 6b, 6d, and 6f showed SI (379, 341, and 374, respectively) higher than that of celecoxib (327) while 6e SI (301) was comparable to celecoxib. Furthermore, compounds 6a, 6c, 6 g, and 6h were slightly less active than celecoxib while compounds 6i–6l showed moderate inhibitory activity against COX-2. On the other hand, compound 7c showed nearly half the activity of celecoxib, while 7a and 7b elicited lower inhibitory activity than celecoxib (Table 1). It is worth mentioning that 4-arylthiazole derivatives (6a–6l) showed higher COX-2 selectivity and inhibitory activity than 4-methylthiazole derivatives (7a–7c). In addition, substitution at the para position of the phenyl moiety at position 4 of thiazole with lipophilic bromo halogen (6e–6h) resulted in increased COX-2 inhibitory activity and selectivity followed by the unsubstituted derivatives (6a–6d). While the substitution with polar methoxy group resulted in decreased activity. Furthermore, substitution at position 3 of thiazole with lipophilic 2-chlorophenyl moiety increased COX-2 selectivity and inhibitory activity (6b and 6f).

Table 1. In vitro COX-1, COX-2, and 5-LOX enzyme inhibitory activities, ^aIC₅₀ values and ^bSI of the tested compounds.

.

Code	R	X	COX-1 IC50 (µM)	COX-2 IC50 (µM)	SI	5-LOX IC50 (µM)
Celecoxib	—	—	14.7 ± 0.07	0.045 ± 0.001	327	—
Diclofenac sodium	—	—	3.9 ± 0.04	0.8 ± 0.009	5	—
Quercetin	—	—	—	—	—	3.34 ± 0.05
6a	Phenyl	H	12.03 ± 0.12	0.077 ± 0.002	156	3.18 ± 0.04
6b	2-Chlorophenyl	H	14.01 ± 0.068	0.037 ± 0.007	379	2.51 ± 0.015
6c	2-Pyridyl	H	13.41 ± 0.068	0.068 ± 0.001	197	2.89 ± 0.026
6d	4-Methoxyphenyl	H	14.32 ± 0.15	0.042 ± 0.001	341	2.47 ± 0.06
6e	Phenyl	Br	13.84 ± 0.088	0.046 ± 0.0016	301	1.75 ± 0.06
6f	2-Chlorophenyl	Br	14.58 ± 0.04	0.039 ± 0.001	374	1.53 ± 0.02
6g	2-Pyridyl	Br	12.87 ± 0.15	0.065 ± 0.002	198	1.98 ± 0.06
6h	4-Methoxyphenyl	Br	11.78 ± 0.13	0.081 ± 0.0016	145	1.91 ± 0.026
6i	Phenyl	OCH3	10.47 ± 0.15	0.10 ± 0.007	105	2.98 ± 0.1
6j	2-Chlorophenyl	OCH3	9.87 ± 0.12	0.13 ± 0.007	76	2.63 ± 0.07
6k	2-Pyridyl	OCH3	8.97 ± 0.12	0.16 ± 0.01	56	2.87 ± 0.07
6l	4-Methoxyphenyl	OCH3	10.18 ± 0.055	0.18 ± 0.007	66	2.69 ± 0.027
7a	Phenyl	—	8.01 ± 0.1	0.22 ± 0.01	36	5.01 ± 0.12
7b	2-Chlorophenyl	—	9.01 ± 0.19	0.18 ± 0.007	50	4.31 ± 0.06
7c	4-Methoxyphenyl	—	10.22 ± 0.15	0.09 ± 0.007	114	3.87 ± 0.12

^aIC₅₀ is the concentration (µM) needed to cause 50% inhibition of COX-1 and COX-2 enzymatic activity. All values are expressed as a mean of three replicates with standard deviation less than 10% of the mean.

^bSI = (COX-1 IC₅₀/COX-2 IC₅₀).

3.2.2. In vitro 5-lox inhibitory assay

Compounds 6a–6l and 7a–7c were screened for their ability to inhibit 5-lipoxygenase enzyme. The results listed in Table 1 revealed that compounds 6a–6l showed 5-LOX inhibitory activities higher than quercetin while compounds 7a–7c elicited slightly less activity than quercetin (Table 1). Moreover, 4-arylthiazole derivatives (6a–6l) showed higher 5-LOX inhibitory activity than 4-methylthiazole derivatives (7a–7c). Furthermore, substitution at the para position of the phenyl moiety at position 4 of thiazole with lipophilic bromo halogen (6e–6h) resulted in increased 5-LOX inhibitory activity over unsubstituted derivatives (6a–6d) and p-methoxy derivatives (6i–6l). In addition, substitution at position 3 of thiazole with lipophilic 2-chlorophenyl moiety (6b, 6f, and 6j) increased 5-LOX inhibitory activity within the same series.

3.2.3. In vivo anti-inflammatory activity

3.2.3.1. Formalin-induced paw edoema test (acute inflammation model)

Compounds 6a–l and 7a–c were subjected to in vivo anti-inflammatory activity (Figure 3 and Table 1 in Supp. File page S38-S39). Compounds (6a–f, 6i–l, 7a, 7c) showed higher % inhibition than celecoxib. Compounds 6a–c, 6e–f, 6i, j showed higher % inhibition than that of diclofenac sodium. Compounds 6b and 6j exhibited double % inhibition exhibited by celecoxib.

Figure 3. % Inhibition of formalin-induced paw edoema.

3.2.3.2. Gastric ulcerogenic activity

Compounds 6a–l and 7a–c were evaluated for their ulcerogenic potential in rats. Gross examination revealed that compounds 6a, 6b, 6f, 6h–l, and 7b showed a superior gastrointestinal safety profile (no ulceration) as the reference celecoxib in the population of fasted rats. On the other hand, compounds 6c–e, 6 g, 7a, and 7c showed variable degrees of hyperaemia (Supp. file page S40). In addition, histopathological examination was performed to confirm gross observations for the most active compounds (6b and 6f) as well as for references and control. Microscopic examination revealed that compounds 6b and 6f showed a superior gastrointestinal safety profile (no ulceration with normal gastric mucosa) as well as the reference celecoxib and DMSO. As regards diclofenac sodium, normal epithelium, and some congested blood vessels were observed (Figure 4).

Figure 4. Histological analysis of rats’ stomach sections stained with H&E (100x magnification).

3.3. Molecular modeling

3.3.1. Molecular docking

The molecular modelling studies were performed using the Molecular Operating Environment (MOE 2016.08) software⁶⁵ for compounds 6b and 6f using the crystal structure of COX-2 enzyme co-crystallized with celecoxib (PDB: 3LN1) and 5-LOX in complex with 4-[(2R,3S)-3-[(3,4-dihydroxyphenyl)methyl]-2-methylbutyl]benzene-1,2-diol (NDGA) (PDB: 6N2W). It is well-known that protein flexibility plays an important role in the protein–ligand recognition process. Induced fit docking considers both the ligand and active site residues flexibility in its calculations which increases the accuracy of predicted binding interactions and binding affinity^66,67. In the present investigation, MOE induced fit docking was used as a docking method. This method continuously sample ligand-receptor conformations as the docking proceeds, allowing for a best fit of the ligand within a potential binding pocket. In addition, five different compound-receptor fit complexes were obtained for each compound, and the best score was chosen.

3.3.1.1 Induced fit docking inside COX-2 active site

Compounds 6b and 6f which showed dual in vitro COX-2/5-LOX inhibiting activity and the reference celecoxib were subjected to induced fit docking into COX-2 active site (PDB: 3LN1)⁶⁸ using Molecular Operating Environment (MOE) version 2016.0802 software⁶². The docking solutions of the compounds 6b, 6f, and celecoxib (Table 2 and Figures 5–8) confirmed the potential activities against COX-2 as the modes of interaction of the best poses were comparable to the reference celecoxib. All the target compounds revealed hydrophobic interaction with the key amino acid Leu338, and polar interaction with the key amino acid Ser339⁶⁹ in the polar side pocket of COX-2 active site. Interaction with this small pocket’s amino acids is essential for the selective inhibition of COX-2 enzyme⁶⁹. In addition, 6b interacted by 2 hydrogen bonds between Cl and Val335, and between imino N and Arg106 (Figure 6). Also, 6f interacted by hydrogen bond between imino N and Arg106 (Figure 7). Moreover, the overlay of compounds 6b, 6f, and celecoxib inside the active site of COX-2 (Figure 8) revealed that the 3,4-diarylthiazole moiety of 6b and 6f was superposed on 1,5-diarylpyrazole moiety of celecoxib with p-substitutedphenyl at position 4 of thiazole interacted at the same position of tolyl moiety of celecoxib. Besides, the substitutedphenyl at position 3 of thiazole interacted at the same position of benzenesulfonamide moiety of celecoxib.

Figure 5. Left: co-crystallized celecoxib, ligand-enzyme interaction (2D). Right: Overlay of co-crystallized celecoxib (yellow) and docked celecoxib (cyan) with RMSD = 0.422 (3D) inside COX-2 active site.

Figure 6. 6b, ligand–enzyme interaction 2D (left) and 3D (right) inside COX-2 active site.

Figure 7. 6f, ligand–enzyme interaction 2D (left) and 3D (right) inside COX-2 active site.

Figure 8. Overlay of 6b (green), 6f (magenta), and celecoxib (cyan) inside COX-2 active site.

Table 2. Induced fit docking results of the active compounds inside COX-2 active site.

Code	Score in Kcal/mol	Hydrogen bonds Interaction amino acids	Hydrophobic Interaction amino acids	Polar Interaction amino acids
Co-crystallized ligand (celecoxib) (Figure 5)	−9.26	Three H-bonds: two bonds between N of SO2NH2 and Ser339, Leu338; O of SO2 with Arg499	Val335, Leu370, Phe367, Leu517, Leu345, Trp373, Met508, Ala502, Phe504, Val509 Ala513, Ile503	His75, Tyr341, Ser516, Tyr371, Gly340, Arg106, Gln178, Gly512
6b (Figure 6)	−7.71	Two H-bonds between CL and Val335*, and between N and Arg106	Val335, Leu517, Leu338, Trp373, Val102, Met508, Ala513, Phe504, Val174, Pro71, Leu78	His75, Tyr341, Tyr371, Arg559, Glu510, Gly512, Ser105, Ser339, Tyr334, Tyr101. Two pi-H bonds between 2-Chlorophenyl and Val509, and between phenyl and Ser516
6f (Figure 7)	−7.91	H-bond between N and Arg106	Val335, Leu338, Trp373, Val102, Met508, Ala513, Phe504, Val74, Pro71, Leu78, Leu370, Val509	His75, Tyr341, Tyr371, Arg559, Glu510, Gly512, Ser339, Ser516, Ser105, Tyr101.

*Amino acids interacted with celecoxib were marketed in bold format.

3.3.1.2. Induced fit docking to 5-LOX active site

Compounds 6b and 6f which showed dual in vitro COX-2/5-LOX inhibiting activity and co-crystallized ligand 4-[(2R,3S)-3-[(3,4-dihydroxyphenyl)methyl]-2-methylbutyl]benzene-1,2-diol (NDGA) were subjected to induced fit docking into 5-LOX active site (PDB: 6N2W)⁷⁰ using Molecular Operating Environment (MOE) version 2016.0802 software⁶². The docking solutions of compounds 6b, 6f, and NDGA (Table 3 and Figure 9–13) supported the potential activities against 5-LOX. It is worth mentioning that, compounds 6b and 6f interacted with all residues interacted with NDGA (Figures 9–12). Furthermore, 5-LOX active site was reported to have several anchors. Polar positively-charged anchor consists of His550, His367, and His372 which interact with Fe²⁺. Polar anchor includes Asp176, Asn180, and Gln363 which interact by both electrostatic and hydrogen-bonding. Besides, Phe177, Leu607, Ile673, Leu414, Phe421, Gly174, Val175, Leu368, and Ile406 were considered hydrophobic anchor ⁷¹. Compounds 6b and 6f interacted with various anchors inside 5-LOX active site. All of them showed polar interaction with the His367 and His372 as positively charged anchors in addition to interaction with the key ferrous ion FE701. Besides, all compounds elicited polar interaction with Gln363 as part of the second polar anchor. Furthermore, all compounds elicited hydrophobic interaction with Leu368, Ile406, Leu414, Leu607, and Ile673 as part of the hydrophobic anchor. Moreover, the overlay of compounds 6b, 6f, and NDGA inside 5-LOX active site (Figure 13) predicted that, thymol moieties of both 6b and 6f were superposed on the first 3,4-dihydroxyphenyl moiety of NDGA. Furthermore, the substituted phenyl moieties at position 4 of thiazole of both 6b and 6f were superposed on the second 3,4-dihydroxyphenyl moiety of NDGA. This result confirmed the importance of the substituted phenyl moiety at position 4 of thiazole in the interaction with 5-LOX active site. In conclusion, the target compounds revealed binding score and fitting to 5-LOX active site better than NDGA. In addition, induced fit docking of 6b and 6f predicted additional polar interaction with the key ferrous ion FE701 which was not found with NDGA. So, the targets 6b and 6f could have a favourable binding affinity towards 5-LOX more than NDGA.

Figure 9. Mode of binding of NDGA inside 5-LOX active site 2D (left) and 3D (right).

Figure 10. Overlay of NDGA (cyan) and docked NDGA (yellow) inside 5-LOX active site with RMSD = 1.2.

Figure 11. Mode of binding of 6b inside 5-LOX active site 2D (left) and 3D (right).

Figure 12. Mode of binding of 6f inside 5-LOX active site 2D (left) and 3D (right).

Table 3. Induced fit docking results of the active compounds in 5-LOX active site.

Code	Score in Kcal/mol	Hydrophobic interaction amino acids	Polar interaction amino acids
NDGA (Figures 9, 10)	−6.27	Phe359, Leu368, Leu373, Ala410, Leu414, Pro569, Trp599, Ala603, Val604, Leu607	Gln363, His367, His372, Asn407, Arg411, Arg596, His600
6b (Figure 11)	−8.06	Phe359*, Leu368, Leu373, Ile406, Ala410, Leu414, Pro569, Trp599, Ala603, Val604, Leu607, Ile673	His360, Gln363, His367, His372, Asn407, Arg411, His432, Arg596, His600, FE701
6f (Figure 12)	−7.86	Phe359, Leu368, Leu373, Ile406, Ala410, Leu414, Pro569, Trp599, Ala603, Val604, Leu607, Ile673	His360, Gln363, His372, Glu376, Asn407, Arg411, His432, Arg596, His600, FE701

*Amino acids interacted with NDGA were marked in bold format.

3.3.2. Molecular dynamics simulation

3.3.2.1. Molecular dynamics simulation of COX-2–ligand complexes

3.3.2.1.1. Root mean square deviation (RMSD)

RMSD plot for the backbone of apo protein and complex structures with the best docked 6b and 6f conformers were calculated (Figure 14). It was observed that, after 10 ns, 6b and 6f complexes together with apo structure attained a stable state with non-significant change in RMSD. This indicated that, the complexes between compounds 6b and 6f with COX-2 retained dynamic conformational stability of the apo protein backbone. Furthermore, RMSD of 6b and 6f ligands were calculated (Supp. file page S41). 6b revealed neglected deviation from the induced docking conformer throughout the time of the test while 6f showed non-significant change in RMSD after 10 ns. These results confirmed the good fitting of the inhibitors 6b and 6f inside COX-2 active site.

Figure 14. RMSD of COX-2 apo (Grey), 6b (Blue), and 6f (red) complex structures.

3.3.2.1.2. Root mean square fluctuation (RMSF)

The interaction residues in COX-2 active site included three parts: part I for residues 335–345, part II residues 367–373, and part III residues 504–516. RMSF predicted that, non-significant variations of apo, 6b and 6f at the interaction residues (part I, II, and III) were noticed (Figure 15, Supp. file page S42). This means that, the interaction between 6b and 6f ligands with the COX-2 active site residues formed stable and low energy complexes due to good fitting inside the active site.

Figure 13. Overlay of docked NDGA (yellow), 6b (green), and 6f (magenta) inside 5-LOX active site.

Figure 15. RMSF of COX-2 -ligand interaction residues, Apo (Grey), 6b (Blue), and 6f (red): (A) Part I and II and (B) Part III.

3.3.2.1.3. Radius of gyration analysis (Rg)

Apo protein, 6b and 6f complexes Rg values predicted non-significant variations, and after 70000 ps all showed nearly the same Rg. This means that the binding of 6b and 6f with COX-2 did not change the packing and globularity of the apo protein. In addition, all complexes were compact and stable along with the apo throughout the simulation (Figure 16).

Figure 16. Rg plot of Apo (Grey), 6b (Blue), and 6f (red).

3.3.2.1.4. Hydrogen bond analysis

There was not much considerable change in the hydrogen bond interaction within the protein of all the complexes. The average values of hydrogen bonds were nearly constant after 15 ns for all complexes (Supp. file page S42). Furthermore, the number of hydrogen bonds was calculated within 0.35 nm between protein and ligand complex structures. 6b complex predicted a maximum number of 7 hydrogen bond pairs after 45 ns, while the 6f complex was 6 pairs after 50 ns (Figure 17). This indicated that, 6b and 6f were able to form hydrogen bonds with a high frequency of occurring a hydrogen bond.

Figure 17. Hydrogen bonds between 6b (Blue) and 6f (red) ligands and protein.

3.3.2.1.5. Solvent-accessible surface area (SASA) calculations

The interaction area between the solvent and protein/protein–ligand complexes was calculated. The mean values for apo, 6b and 6f complexes were 254.23 nm², 253.84 nm², and 259.55 nm². Higher value of SASA implies more hydrophilicity. The SASA value for 6b complex was nearly equal to apo while the 6f complex was more accessible to the solvent than apo and 6b complex (Supp. file page S42). The SASA did not show a significant change between complex structures in comparison with apo protein.

3.3.2.2. Molecular dynamics simulations of 5-LOX–ligand complexes

3.3.2.2.1. Root mean square deviation (RMSD)

RMSD plot for the backbone of apo protein and complex structures with the best docked 6b and 6f conformers were calculated (Figure 18). It was observed that, after 30 ns, 6b and 6f complexes together with apo structure attained a stable state with non-significant change in RMSD. This indicated that the complexes between compounds 6b and 6f with 5-LOX retained dynamic conformational stability of the apo protein backbone. Furthermore, RMSD of 6b and 6f ligands were calculated (Supp. file page S43). 6b revealed stability with neglected deviation from the induced docking conformer after 30 ns of simulation while 6f showed a non-significant change in RMSD after 5 ns till the end of the simulation. These results confirmed the good fitting of the inhibitors 6b and 6f inside 5-LOX active site.

Figure 18. RMSD of 5-LOX apo (Grey), 6b (Blue), and 6f (red) complex structures.

3.3.2.2.2. Root mean square fluctuation (RMSF)

The interaction residues in 5-LOX active site included three parts: part I for residues 359–376, part II residues 406–414, and part III residues 596–607. RMSF predicted that, non-significant variations of apo, 6b and 6f at the interaction residues (part I, II, and III) were noticed (Figure 19, Supp. file page S43). This means that, the interaction between 6b and 6f ligands with the 5-LOX active site residues formed stable and low energy complexes due to good fitting inside the active site.

Figure 19. RMSF of 5-LOX -ligand interaction residues, Apo (Grey), 6b (Blue), and 6f (red): (A) Part I and II and (B) Part III.

3.3.2.2.3. Radius of gyration analysis (Rg)

Apo protein, 6b and 6f complexes Rg values predicted non-significant variations after 20 ns. 6b complex was slightly more compact than apo and 6f complex. This means that the binding of 6b and 6f with 5-LOX did not significantly change the packing and globularity of the apo protein. In addition, all complexes were compact and stable along with the apo throughout the simulation (Figure 20).

Figure 20. Rg plot of Apo (Grey), 6b (Blue), and 6f (red) complexes.

3.3.2.2.4. Hydrogen bond analysis

There was not much considerable change in the hydrogen bond interaction within the protein of all the complexes. The average values of hydrogen bonds were nearly constant after 10 ns for all complexes (Supp. file page S43). Furthermore, the number of hydrogen bonds was calculated within 0.35 nm between protein and ligand complex structures. 6b complex predicted a maximum number of 6 hydrogen bond pairs, while 6f complex was 5 pairs (Figure 21). This indicated that 6b and 6f were able to form hydrogen bonds with 5-LOX active site and a high frequency of occurrence of a hydrogen bond.

Figure 21. Hydrogen bonds between 6b (Blue) and 6f (red) ligands and protein.

3.3.2.2.5. Solvent-accessible surface area (SASA) calculations

The interaction area between the solvent and protein/protein–ligand complexes was calculated. The mean values for apo, 6b, and 6f complexes were 283.31 nm², 286.42 nm², and 290.55 nm². A higher value of SASA implies more hydrophilicity. The SASA value for 6b and 6f complexes were higher than apo which means that 6b and 6f complexes were more accessible to the solvent than apo (Supp. file page S44). The SASA did not show a significant change between complex structures in comparison with apo protein.

3.4. Structure–activity relationship

The in vitro COX-1, COX-2, and 5-LOX inhibitory activities, and in vivo anti-inflammatory activity of the target compounds showed that substitution at position 4 of thiazole with aryl moiety is essential for activity as all 6a–l compounds were more active than the methyl derivatives (7a-c). Furthermore, the type of para substitution on the phenyl moiety at position 4 of thiazole greatly affected COX-2 and 5-LOX inhibitory activities. The para substitution with lipophilic electron withdrawing bromo derivatives (6e–6h) showed the highest dual COX-2/5-LOX inhibitory activities followed by the unsubstituted derivatives (6a–6d) while substitution with polar electron-donating methoxy group (6i–6 l) decreased COX-2 inhibitory activity, and retained 5-LOX inhibitory activity as unsubstituted derivatives. On the other hand, the type of substitution on the aryl moiety at position 3 of thiazole slightly affected the activity. The most active were the derivatives which contained the lipophilic electron-withdrawing 2-chloro substitution (6b and 6f) followed by the unsubstituted phenyl derivatives (6a and 6e) and the 2-pyridyl derivatives (6c and 6 g). In conclusion, the presence of aromatic moiety substituted with electron withdrawing lipophilic halogen at both positions three and four of thiazole had a great effect on COX-2 and 5-LOX inhibitory activities. This could be due to the hydrophobic interaction with the lipophilic anchors in both COX-2 and 5-LOX active sites.

4. Conclusion

There is an urgent need for new safe anti-inflammatory drugs lake the side effects of selective inhibition of COX-2 alone. Consequently, simultaneous inhibition of COX-2 and 5-LOX enzymes could reduce the side effects of anti-inflammatory candidates. In the present investigation, 3,4-disubstitutedthiazole pharmacophore, a bioisostere to the COX-2 inhibitor pharmacophore 1,5-diarylpyrazole, was hybridised with thymol, a natural 5-LOX inhibitor, aiming to act as dual COX-2/5-LOX inhibitors to produce in vivo active anti-inflammatory compounds with better safety profile and fewer side effects. Compounds 6b, 6d, 6e, and 6f displayed in vitro inhibitory activity against COX-2 (IC₅₀= 0.037, 0.042, 0.046, and 0.039 µM) nearly equal to celecoxib (IC₅₀= 0.045 µM). Besides, 6b, 6d, and 6f showed SI (379, 341, and 374, respectively) higher than that of celecoxib (327). Moreover, 6a–l elicited in vitro 5-LOX inhibitory activity higher than reference quercetin. On the other hand, 6a–f, 6i–l, 7a, and 7c possessed in vivo inhibition of formalin induced paw edoema higher than celecoxib. In addition, 6a, 6b, 6f, 6h–l, and 7b showed a superior gastrointestinal safety profile (no ulceration) as celecoxib and diclofenac sodium in the population of fasted rats. Furthermore, induced fit docking together with molecular dynamics simulation of COX-2 and 5-LOX – ligand complexes and apo protein were calculated. RMSD results confirmed the good fitting of the inhibitors 6b and 6f inside both COX-2 and 5-LOX active sites. Moreover, RMSF results predicted that the interaction between 6b and 6f ligands and both COX-2 and 5-LOX active sites’ residues formed stable and low energy complexes due to good fitting. Besides, Rg anticipated that, binding of 6b and 6f with both COX-2 and 5-LOX active sites did not change the packing and globularity of the apo protein, and all complexes were compact and stable. Furthermore, H-bond analysis concluded that 6b and 6f were able to form hydrogen bonds with a high frequency of occurrence. On the other hand, SASA did not show the significant change between complex structures in comparison with apo protein. In conclusion, 6b and 6f elicited multitarget inhibition of inflammation by in vitro simultaneous inhibition of COX-2 and 5-LOX higher than celecoxib and quercetin, in vivo potent anti-inflammatory activity higher than celecoxib, and in vivo gastrointestinal safety profile as celecoxib.

Author contributions

The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript. All authors contributed equally.

Acknowledgement

Ahmed K. Al-Kubeisi would like to thank Alnukhba University College for their free support.