Chem Biol Drug Des 2016; 87: 784–793 Research Article
Synthesis of Novel Substituted Thiourea and Benzimidazole Derivatives Containing a Pyrazolone Ring as Anti-Inflammatory Agents
Ashraf A. Moneer, Khaled O. Mohammed and inhibited by paracetamol (1,9,10). Most NSAIDs like Hala B. El-Nassan* diclofenac I and indomethacin II (Figure 1) are non-selec- tive and inhibit both COX-1 and COX-2 with variable Pharmaceutical Organic Chemistry Department, Faculty of degrees of selectivity. Selective COX-2 inhibitors like cele- Pharmacy, Cairo University, Cairo 11562, Egypt coxib III and meloxicam may reduce the GIT side-effects *Corresponding author: Hala B. El-Nassan, associated with inhibition of COX-1 (11–16). However, they [email protected] exhibit cardiovascular side-effects (14,17–19). Selective COX-2 inhibitors decrease the level of the atheroprotective { Two series of new 1-(alkyl/aryl)-3- 2-[(5-oxo-4,5-dihy- prostacyclin (PGI2), but not the COX-1-derived thrombox- H } – dro-1 -pyrazol-3-yl)amino]phenyl thioureas 2a h and ane A2 which is a proaggregatory and vasoconstrictor 5-[2-(substituted amino)-1H-benzimidazol-1-yl]-4H-pyr- mediator (19,20). This increases the risk of heart attack azol-3-ols 3a–i were designed and synthesized as anti- and stroke. Indeed, Rofecoxib (Vioxx©) was withdrawn inflammatory agents. The cyclooxygenase inhibitory activity of the newly synthesized compounds was from the market in 2004 by Merck & Co., Inc. (Kenilworth, investigated. All the compounds showed non-selective NJ, USA) due to its adverse cardiovascular side-effects inhibition of COX-1 and COX-2 enzymes which was (21). Thus, the development of safe and efficient NSAID consistent with their docking results. Compounds 2c, still represents a challenge to the medicinal chemists 2f, 2g, 3b, and 3g that showed the highest COX-2 inhi- especially after the recently reported use of selective COX- bitory activity were selected for further in vivo testing 2 inhibitors in cancer chemotherapy (1,22–25) and neuro- as anti-inflammatory agents using diclofenac as a ref- logical diseases such as Parkinson (1,26) and Alzheimer’s erence drug. Two of the test compounds (2c and 3b) diseases (1,27,28). showed potent anti-inflammatory activity comparable to that of diclofenac with lower ulcerogenic effect rela- Selective COX-2 inhibitors can be classified into tricyclics tive to indomethacin. SAR study of the two series as like Coxibs such as Celecoxib and Rofecoxib (also known cyclooxygenase inhibitors and anti-inflammatory as ortho-diaryl heterocycles or carbocycles) and non-tri- agents was also provided. cyclics. The tricyclic compounds contain 1,2-diaryl substi- tution on a central hetero or carbocyclic ring system, and Key words: anti-inflammatory activity, benzimidazole, COX one of the aryl rings carries methanesulfonyl, methanesul- inhibitors, pyrazolone, thiourea fonamide, sulfonamido, azido, or tetrazole pharmacophore which is responsible for COX-2 selectivity. The non-tricyc- Received 8 June 2015, revised 3 November 2015 and accepted for publication 26 November 2015 lics lack the cyclic central core; however, they contain acyclic central core such as olefinic, iminic, azo, acetyle- nic, and a,b-unsaturated ketone structures (1). Pain, fever, and inflammation are usually treated with non- steroidal anti-inflammatory drugs (NSAIDs). These drugs Recently, a series of 3-{2-[6-(4-substituted amino)pyridazin- exert their action by competitive inhibition of cyclooxyge- 3-ylamino]phenylamino}-1H-pyrazol-5(4H)-ones were syn- nase enzymes (COX-1, COX-2, and COX-3) (1–3). COX thesized by our laboratory as anti-inflammatory agents (29). enzymes mediate the biosynthesis of prostaglandins (PGs) Among the prepared compounds, compound IV (Figure 1) from arachidonic acid (AA) (3,4). COX-1 is a constitutive showed promising anti-inflammatory activity comparable to enzyme responsible for the production of the cytoprotec- diclofenac (edema inhibition = 62%) together with good tive PG in gastrointestinal tract and kidney, and therefore, analgesic and low ulcerogenic activity in vivo. It also exhib- its inhibition resulted in peptic ulcer and renal impairment. ited good inhibition effect on the serum production of TNF-a COX-2 is an inducible enzyme that is expressed at the site and IL-6. This compound exerted its actions via inhibition of = l = l of injury in response to the release of proinflammatory both COX-1 (IC50 2.86 M) and COX-2 (IC50 0.39 M). mediators (2,5–8). While, COX-3 was recently discovered Molecular modeling of this series indicated that the pyra- in 2002, and it is present mainly in the brain and was zolone NH and the pyrazolone carbonyl made H-bonding to
784 ª 2015 John Wiley & Sons A/S. doi: 10.1111/cbdd.12712 Synthesis of Thioureas and Benzimidazoles
O
Cl Cl OH F O H C H C 3 F NH 3 CH3 N F N N O
OO O Cl S O Celecoxib Diclofenac Indomethacin H2N III I II
O SNHR N N N NH NH N NHR NH N NH N N N NH N N H O O O H 2a-h 3a-i Figure 1: Design of the target compounds. IV the COX key amino acids Tyr 385 and Ser530, respectively. (methyl and methoxy) and electron withdrawing group This resembles the H-bonding made by diclofenac carboxy- (chloro) were prepared. late group which reflected the importance of the pyrazolone ring for the COX inhibitory activity. All the synthesized compounds were tested for their COX-1 and COX-2 inhibitory activities, and the most active inhibi- From these results, it can be concluded that such deriva- tors within the two series were further tested in vivo for their tives belonged to the tricyclic series where the phenyl ring anti-inflammatory and ulcerogenic activities using diclofe- acts as a central carbocyclic ring and the 1H-pyrazol-5 nac as a reference drug. Docking simulations study was (4H)-one ring mimicked the carboxylate group in diclofe- carried out to determine the binding mode of the designed nac (Figure 1). compounds into the active site of COX-2 enzyme.
Encouraged by these findings, further work was carried Experimental Part out on 3-{(2-(substituted)-phenylamino}-1H-pyrazol-5(4H)- ones 2a–h through introducing different alkyl and aryl General thiourea derivatives instead of the pyridazine ring, while the Melting points were determined using a Griffin apparatus pyrazolone ring was kept constant to ensure the enzyme (Fisher Scientific, Leicestershire, UK) and were uncorrected. inhibitory activity. IR spectra were recorded on Shimadzu (Kyoto, Japan) IR 435 spectrophotometer, and values were represented in Besides, 2-(alkyl/arylamino)benzimidazoles 3a–i were pre- cm 1. H NMR and 13C NMR spectra were carried out on pared to examine the effect of structure rigidification of Bruker 400 MHz (Bruker BioSpin AG, Fa¨ llanden, Switzer- 1,2-diaminophenyl into benzimidazole ring on COX inhibi- land) and 100 MHz spectrophotometer, respectively, Fac- tion and hence anti-inflammatory activity. Searching the lit- ulty of Pharmacy, Cairo University, Cairo, Egypt, using TMS erature indicated that analogous benzimidazole derivatives as an internal standard and chemical shifts were recorded might exhibit anti-inflammatory activity (30–33). in p.p.m. on d scale. Elemental analyses were carried out at the regional center for mycology and biotechnology, Al- In both series, the alkyl groups selected were ethyl and Azhar University, Cairo, Egypt. Analytical thin layer chro- butyl groups to study the effect of increasing the alkyl matography on silica gel plates containing UV indicator was chain on COX inhibitory activity. Allyl group was prepared employed routinely to follow the course of the reactions to examine the effect of unsaturation, while cyclohexyl and and to check the purity of the products. All reagents and benzyl groups were designed to study the effect of bulki- solvents were purified and dried by standard techniques. ness and hydrophobicity on enzyme inhibition. Regarding the aryl groups, both unsubstituted phenyl moiety and 5-[(2-Aminophenyl)amino]-2,4-dihydro-3H-pyrazol-3-one phenyl ring substituted with electron-donating groups (1) (34) was prepared according to the published method.
Chem Biol Drug Des 2016; 87: 784–793 785 Moneer et al.
General procedure for the synthesis of 1-alkyl-3- 7.55 (m, 4H, Ar–H), 12.49 (s, 1H, NH, D2O exchangeable), 13 {2-[(5-oxo-4,5-dihydro-1H-pyrazol-3-yl)amino] 13.69 (s, 1H, OH, D2O exchangeable); C NMR (DMSO- phenyl}thioureas 2a–e and 1-(substituted phenyl)- d6) d p.p.m. 13.4, 39.1 (CH3CH2), 26.4 (pyrazolyl CH2), 3-{2-[(5-oxo-4,5-dihydro-1H-pyrazol-3-yl)amino] 111.6, 118.9, 121.7, 122.6, 134.8, 143.4, 148.8, 149.0, phenyl}thioureas 2f–h 167.0 (aromatic carbons); Anal. Calcd for C12H13N5O: C, A mixture of 5-[(2-aminophenyl)amino]-2,4-dihydro-3H-pyr- 59.25; H, 5.39; N, 28.79. Found: C, 59.51; H, 5.38; N, azol-3-one (1) (1 mmol) and the appropriate alkyl or aryl 29.04. isothiocyanate (1 mmol) in absolute ethanol (25 mL) was stirred at room temperature for 3 h. The solid formed was filtered, dried, and crystallized from ethanol. 5-{2-[(4-Chlorophenyl)amino]-1H-benzimidazol-1- yl}-4H-pyrazol-3-ol (3f) Yield: 67%; mp: 254–256 °C; IR (cm 1): 3421 (OH), 3232 1 1-Ethyl-3-{2-[(5-oxo-4,5-dihydro-1H-pyrazol-3-yl) (NH), 2918, 2852 (CH-aliphatic); H NMR (DMSO-d6) d amino]phenyl}thiourea (2a) p.p.m. 4.19 (s, 2H, pyrazole CH2), 7.03–7.69 (m, 8H, Ar– 1 Yield: 53%; mp: 209–211 °C; IR (cm ): 3261, 3194, H), 12.22 (s, 1H, NH, D2O exchangeable), 13.95 (s, 1H, 1 13 3151 (NH), 2968, 2856 (CH-aliphatic), 1676 (C=O); H OH, D2O exchangeable); C NMR (DMSO-d6) d p.p.m. NMR (DMSO-d6) d p.p.m. 1.10 (t, 3H, CH3CH2, 26.7 (pyrazolyl CH2), 118.8, 120.2, 122.0, 129.0, 129.1, J = 7.04 Hz), 3.91 (q, 2H, CH3CH2, J = 7.04 Hz), 4.00 (s, 129.2, 129.6, 130.5, 132.7, 134.5, 148.7, 148.9, 168.5 2H, pyrazolyl CH2), 7.20–7.63 (m, 4H, Ar–H), 9.36 (s, 1H, (aromatic carbons); Anal. Calcd for C16H12ClN5O: C, 58.99; NH, D2O exchangeable), 10.21 (s, 1H, NH, D2O H, 3.71; N, 21.50. Found: C, 59.14; H, 3.65; N, 21.68. exchangeable), 10.25 (s, 1H, NH, D2O exchangeable), 13 10.53 (s, 1H, NH, D2O exchangeable); C NMR (DMSO- The data of compounds 2b–e, 2g–h, 3b–e, and 3g–i were d6) d p.p.m. 15.1, 38.8 (CH3CH2), 34.3 (pyrazolyl CH2), represented in the supporting information. 73.8, 118.7, 122.7, 137.2, 149.5, 154.6, 157.0 (aromatic carbons), 166.8 (C=O), 172.1 (C=S); Anal. Calcd for The experimental procedures of molecular modeling and
C12H15N5OS: C, 51.97; H, 5.45; N, 25.25. Found: C, biological testing were represented in the supporting infor- 52.16; H, 5.49; N, 25.43. mation.
1-{2-[(5-Oxo-4,5-dihydro-1H-pyrazol-3-yl)amino] Results and Discussion phenyl}-3-phenylthiourea (2f) Yield: 58%; mp: 185–187 °C; IR (cm 1): 3375, 3250, Docking study 3209 (NH), 2922, 2852 (CH-aliphatic), 1695 (C=O); 1H Examining the structures of COX-1 and COX-2 enzymes
NMR (DMSO-d6) d p.p.m. 3.91 (s, 2H, pyrazolyl CH2), indicated that both isoforms are structurally homologous 7.10–7.48 (m, 9H, Ar–H), 9.83 (s, 1H, NH, D2O exchange- with about 60% sequence identity (1). Thus, both enzymes able), 10.59 (br s, 2H, two NH, D2O exchangeable), 12.49 consist of a membrane-binding domain (MBD), N-terminal (br s, 1H, NH, D2O exchangeable); Anal. Calcd for epidermal growth factor (EGF)-like domain, and large C- C16H15N5OS: C, 59.06; H, 4.65; N, 21.52. Found: C, terminal globular catalytic domain which contains the COX 59.34; H, 4.67; N, 21.79. active site (35). The latter is a long hydrophobic narrow channel that extends from the MBD to the core of the cat- alytic domain (36,37). AA-binding site is present in the General procedure for the synthesis of 5-[2- upper half of the channel, from Arg-120 to Tyr-385. (substituted amino)-1H-benzimidazol-1-yl]-4H- pyrazol-3-ols 3a–i The selectivity of COX-2 over COX-1 is due to the differ- A mixture of 5-[(2-aminophenyl)amino]-2,4-dihydro-3H-pyr- ence in only three amino acids that results in an additional azol-3-one (1) (1 mmol) and the appropriate alkyl or aryl side pocket that can accommodate larger rigid groups. isothiocyanate (1 mmol) in absolute ethanol (25 mL) was These amino acids are the smaller Val-523 and Val-524 (in heated under reflux for 3 h. The solvent was concentrated COX-1, they are the more bulky Ile-523 and Ile-524) and under reduced pressure, and the solid formed was filtered, the more polar Arg-513 (in COX-1, it is the less polar dried, and crystallized from ethanol. His-513) (13,18,38–40). Selective COX-2 inhibitors usually contain bulky rigid groups (such as sulfonamide and alkylsulfonanilides) that prevent the compound fitting inside 5-[2-(Ethyl amino)-1H-benzimidazol-1-yl]-4H- the narrower COX-1 channel while fit nicely inside the pyrazol-3-ol (3a) COX-2 side pocket (13,18,38–41). Yield: 40%; mp: 233–235 °C; IR (cm 1): 3414 (OH), 3334 1 (NH), 2976, 2856 (CH-aliphatic); H NMR (DMSO-d6) d COX-2 active site can be divided into three main regions: p.p.m. 1.05 (t, 3H, CH3CH2, J = 7.08 Hz), 3.96 (q, 2H, the mouth of the active site formed by three hydrophilic CH3CH2, J = 7.08 Hz), 4.43 (s, 2H, pyrazolyl CH2), 7.15– residues (Arg120, Glu524, and Tyr355) that are arranged
786 Chem Biol Drug Des 2016; 87: 784–793 Synthesis of Thioureas and Benzimidazoles to form a hydrogen bond network. A hydrophobic pocket energy score to diclofenac (data not shown). For the below the heme group formed by the residues Ala201, thiourea series 2a–h, two binding modes were observed: Tyr248, Leu352, Phe381, Tyr385, Trp387, and Phe518. Ser530 is present below Tyr385. Finally, a larger side pocket defined by several residues such as His90, The central phenyl ring bound to Arg120 by arene–ca- Arg513, and Met522 (42). tion interaction, while Tyr355 made one or two H-bonds to the thiourea NH (Figures 2A and S2a). Non-selective NSAIDs such as flurbiprofen, indomethacin, and diclofenac are oriented toward the mouth of the active The pyrazolone C=O and NH groups formed one or two site and form H-bonds with Arg120 and Tyr355 (in case of H-bonds to Ser530 and Tyr385 as shown in Figures 2B flurbiprofen and indomethacin) or Tyr 385 and Ser530 (in and S2b. This resembled the interactions made by the case of diclofenac). However, selective COX-2 NSAIDs like carboxylate group of diclofenac. celecoxib form extra H-bonds with His90 and Arg513 as well as hydrophobic interactions with Phe518, Leu352, Two binding modes were also observed for the benzimida- Leu359, Trp387, and Met522 located in the side pocket zole series 3a–i: (39,41,43,44). Arg120 made arene–cation interaction to the phenyl ring of In this work, all the designed compounds were subjected to benzimidazole, while N-3 of benzimidazole made H-bond to docking study to explore their possible mode of binding to Tyr385 or Ser530 as exhibited by compound 3f (Figure 3A) or COX-2 enzyme using MOLECULAR OPERATING ENVIRONMENT to Tyr355 as exhibited by compound 3e (Figure S3a). (MOE, 10.2008) software (Chemical Computing Group, Montreal, Canada). The crystal structure of COX-2 enzyme The pyrazolone C=O group formed one or two H-bonds complexed with flurbiprofen (a non-selective COX inhibitor) to Ser530 and Tyr385 as shown in Figures 3B and S3b. was obtained from Protein Data Bank (PDB: 3PGH). Diclofe- nac was chosen as a lead compound and its docking into COX-2 active site indicated that the carboxylate group Thus, it seemed that the two series bound similarly to the formed three H-bonds to Tyr385 and Ser530 (Figure S1). active site and they bound mainly to the mouth of the This was consistent with its reported mode of binding (44). channel with the pyrazolone ring extended inside the Then, the target compounds were docked in COX-2 active hydrophobic pocket. However, there are no signs of bind- site, and the results indicated that all the designed com- ing to the side pocket of COX-2. Therefore, both series pounds bind to the COX-2 active site with comparable might act as non-selective COX inhibitors and this was
Figure 2: 2D representation of the expected binding modes of compounds (A) 2a, (B) 2f docked into the active site of COX-2 enzyme.
Chem Biol Drug Des 2016; 87: 784–793 787 Moneer et al.
Figure 3: 2D representation of the expected binding modes of compounds (A) 3f, (B) 3b docked into the active site of COX-2 enzyme. consistent with the results obtained from the enzyme inhi- azol-3-one (1) (34) and alkyl or aryl isothiocyanates at bition study and the ulcerogenesis assay. room temperature. IR spectra of compounds 2a–h indicated the presence of NH bands at 3498–3151 cm 1 and C=O band at 1699–1672 cm 1. 1H NMR spectra of Chemistry compounds 2a–h showed three or four exchangeable sin- Scheme 1 outlines the synthesis of the target compounds glet signals at d 7.92–12.52 p.p.m. corresponding to four 2a–h and 3a–i. NH protons. Moreover, 13C NMR spectra of compounds 2a, b, d, e revealed the presence of C=S carbon at d Compounds 2a–h were prepared via stirring equimolar 172.1–181.4 p.p.m. as well as signals corresponding to amounts of 5-[(2-aminophenyl)amino]-2,4-dihydro-3H-pyr- the introduced aliphatic or aromatic carbons.
N N NHR NHR N - H S N 2 SNHR NH2 N N N NH RNCS NH RNCS NH O HO stirr 3h. reflux 3h. 3a-i NH N N N O H
N H H - RNH O 2 N S 1 X 2a-h N
N NH R= C H , C H , C H , -CH CH=CH , -CH C H 3* 2 5 4 9 6 11 2 2 2 6 5 O C H , 4-ClC H , 3-CH C H , 4-CH C H , 4-CH OC H 6 5 6 4 3 6 4 3 6 4 3 6 4
Scheme 1: Preparation of the target compounds.
788 Chem Biol Drug Des 2016; 87: 784–793 Synthesis of Thioureas and Benzimidazoles
On the other hand, refluxing a mixture of compound 1 and Table 1: Results of in vitro cyclooxygenase (COX) inhibition of the appropriate alkyl or aryl isothiocyanate in absolute compounds 2a–h and 3a–i ethanol afforded 5-[2-(substituted amino)]-1H-benzimida- SNHR zole derivatives 3a–i. NH N Literature review of the reaction of 1,2-phenylenediamine NHR N and isothiocyanates indicated two possible cyclization NH pathways to give benzimidazole derivative as shown in N Scheme 1 (45–47). In this work, the probability of forma- N N tion of compound via elimination of amines was ruled 3* N out based on NMR spectral data. Thus, 1H NMR spectra O H HO of compounds 3a–i showed an exchangeable singlet sig- nal at d 12.18–12.49 p.p.m. corresponding to RNH proton 2a-h 3a-i together with the introduced aliphatic or aromatic protons. 13 Besides, C NMR spectra of compounds 3a–i revealed IC50 in nM the disappearance of C=S signals and the appearance of Selectivity the introduced aliphatic or aromatic carbons. Compound no. R COX-1 COX-2 index (SI) 2a C H 0.3083 0.0715 4.25 – 2 5 Revising the IR spectra of compounds 3a i indicated the 2b n-C4H9 0.4340 0.1790 2.43 absence of C=O band and the presence of a broad band 2c C6H11 0.1340 0.0283 4.73 1 = at 3421–3232 cm corresponding to OH bond. This 2d -CH2CH CH2 0.2346 0.0533 4.36 might be attributed to the keto–enol tautomerism of the 2e -CH2C6H5 0.3369 0.0947 3.60 3H-pyrazol-3-one ring into 4H-pyrazol-3-ol (Scheme 1). 2f C6H5 0.1597 0.0300 5.32 2g 4-ClC H 0.1734 0.0462 3.75 Searching the literature pointed out that such tautomerism 6 4 2h 3-CH C H 0.3505 0.1054 3.32 was common in pyrazolone ring especially when position 4 3 6 4 3a C2H5 0.4615 0.1522 3.03 was unsubstituted (48–52). Examining the 1H NMR spectra 3b n-C4H9 0.2272 0.0469 4.84 – of compounds 3a i indicated the presence of an 3c -CH2CH=CH2 0.3531 0.0903 3.91 exchangeable singlet signal at d 13.67–13.95 p.p.m. 3d -CH2C6H5 0.4112 0.1290 3.18 attributed to OH protons which supported this postulation. 3e C6H5 0.3091 0.1133 2.73 3f 4-ClC6H4 0.3204 0.1324 2.42 3g 3-CH3C6H4 0.1664 0.0370 4.50 3h 4-CH C H 0.2297 0.0680 3.44 Measurement of in vitro COX enzymes inhibition 3 6 4 3i 4-CH OC H 0.2603 0.0587 4.44 The synthesized compounds 2a–h and 3a–i were tested 3 6 4 for their ability to inhibit COX-1 and/or COX-2 enzymes using Cayman’s colorimetric COX (ovine) assay. The assay measures the peroxidase activity colorimetrically by moni- Careful examination of the influence of the substituents (R toring the appearance of oxidized N,N,N0,N0-tetramethyl-p- group) along both series on the enzyme inhibitory activity phenylenediamine at 590 nm as reported by Kulmacz and showed the following: Lands (53).