Discovery of new indomethacin-based analogs with potentially selective cyclooxygenase-2 inhibition and observed diminishing to PGE2 activities

Article abstract of DOI:10.1016/j.ejmech.2017.09.056

New ring-extended analogs of indomethacin were designed based on the structure of active binding site of both COX-1 and COX-2 isoenzymes and the interaction pattern required for selective inhibition of COX-2 to improve its selectivity against COX-2. The strategy adopted for designing the new inhibitors involved i) ring extension of indomethacin to reduce the possibility of analogs to be accommodated into the narrow hydrophobic tunnel of COX-1, ii) deletion of carboxylic acid to reduce the possibility of inhibitor to form salt bridge with Arg120 and eventually prevent COX-1 inhibition, and iii) introduction of methylsulfonyl group to increase the opportunity of the analogs to interact with the polar side pocket that's is crucial for inhibition process of COX-2. The three series of tetrahydrocarbazoles involving 4, 5, 9, 10 and 12 were synthesized in quantitative yields adopting limited number of reaction steps, and applying laboratory friendly reaction conditions. In vitro and in vivo assays for data profiling the new candidates revealed the significant improvement in the potency and selectivity against COX-2 of 6-methoxytetrahydrocarbazole 4 (IC₅₀ = 0.97 μmol) to verify the effect of ring extension in comparison to indomethacin (IC₅₀ = 2.63 μmol), and 6-methylsulfonyltetrahydrocarbazole 10a (IC₅₀ = 0.28 μmol) to verify the effect of ring extension and introduction of methylsulfonyl group. 9-(4-chlorobenzoyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-one 12a showed the most potential and selective activity against COX-2 (IC₅₀ = 0.23 μmol) to be with superior potency to Celecoxib (IC₅₀ = 0.30 μmol). Consistently, 12a was the most active with all the other anti-inflammatory test descriptors and its activity in diminishing the PGE2 with the other analogs confirmed the elaboration of new class of selective COX-2 inhibitors beyond the diarylsulfonamides as a previously common class of selective COX-2 inhibitors. Molecular docking study revealed the high binding score of compound 12a (?30.78 kcal/mol), with less clash contribution (7.2) that is close to indomethacin. Also, 12a showed low conformation entropy score (1.40). Molecular dynamic (MD) simulation identified the equilibrium of both potential and kinetic energies.

Full text of DOI:10.1016/j.ejmech.2017.09.056

Accepted Manuscript
Discovery of new indomethacin-based analogs with potentially selective
cyclooxygenase-2 inhibition and observed diminishing to PGE2 activities
Shaymaa E. Kassab, Mohammed A. Khedr, Hamed I. Ali, Mohamed M. Abdalla
PII:
S0223-5234(17)30771-7
10.1016/j.ejmech.2017.09.056
EJMECH 9771
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 25 June 2017
Revised Date: 21 September 2017
Accepted Date: 26 September 2017
Please cite this article as: S.E. Kassab, M.A. Khedr, H.I. Ali, M.M. Abdalla, Discovery of new
indomethacin-based analogs with potentially selective cyclooxygenase-2 inhibition and observed
diminishing to PGE2 activities, European Journal of Medicinal Chemistry (2017), doi: 10.1016/
j.ejmech.2017.09.056.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Discovery of New Indomethacin-based analogs with Potentially Selective Cyclooxygenase-2 Inhibition and
Observed Diminishing to PGE2 activities.
Shaymaa E. Kassaba , Mohammed A. Khedrb, Hamed I. Alib,c, Mohamed M. Abdallad
aPharmaceutical Chemistry Department, Faulty of Pharmacy, Damanhour University, Damanhour, El-Buhaira
22516, Egypt.
bPharmaceutical Chemistry Department, Faculty of Pharmacy, Helwan University, Ein Helwan, Cairo 11795,
Egypt.
cDepartment of Pharmaceutical Sciences, Texas A&M University Irma Lerma Rangel College of Pharmacy
Kingsville78363, Texas, USA.
dResearch Unit, Saco Pharm. Co., 6th of October City, Giza 68330, Egypt.
Abstract
New ring-extended analogs of indomethacin were designed based on the structure of active binding site of both
COX-1 and COX-2 isoenzymes and the interaction pattern required for selective inhibition of COX-2 to
improve its selectivity against COX-2. The strategy adopted for designing the new inhibitors involved i) ring
extension of indomethacin to reduce the possibility of analogs to be accommodated into the narrow hydrophobic
tunnel of COX-1, ii) deletion of carboxylic acid to reduce the possibility of inhibitor to form salt bridge with
Arg120 and eventually prevent COX-1 inhibition, and iii) introduction of methylsulfonyl group to increase the
opportunity of the analogs to interact with the polar side pocket that’s is crucial for inhibition process of COX-2.
The three series of tetrahydrocarbazoles involving 4, 5, 9, 10 and 12 were synthesized in quantitative yields
adopting limited number of reaction steps, and applying laboratory friendly reaction conditions. In vitro and in
vivo assays for data profiling the new candidates revealed the significant improvement in the potency and
selectivity against COX-2 of 6-methoxytetrahydrocarbazole 4 (IC50 = 0.97 µmol) to verify the effect of ring
extension in comparison to indomethacin (IC50 = 2.63 µmol), and 6-methylsulfonyltetrahydrocarbazole 10a
(IC50 = 0.28 µmol) to verify the effect of ring extension and introduction of methylsulfonyl group. 9-(4-
chlorobenzoyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-one 12a showed the most potential and
selective activity against COX-2 (IC50 = 0.23 µmol) to be with superior potency to Celecoxib (IC50 = 0.30
µmol). Consistently, 12a was the most active with all the other anti-inflammatory test descriptors and its activity
in diminishing the PGE2 with the other analogs confirmed the elaboration of new class of selective COX-2
inhibitors beyond the diarylsulfonamides as a previously common class of selective COX-2 inhibitors.
Molecular docking study revealed the high binding score of compound 12a (-30.78 kcal/mol), with less clash
contribution (7.2) that is close to indomethacin. Also, 12a showed low conformation entropy score (1.40).
Molecular dynamic (MD) simulation identified the equilibrium of both potential and kinetic energies.
Keywords: Indomethacin, Tetrahydrocarbazoles, COX-2 inhibitors, Molecular docking, PGE2
1

ACCEPTED MANUSCRIPT
1. Introduction
Prostaglandins (PGs) and glucocorticoids are mediators that potentially implicated to the inflammation process.
Non-steroidal anti-inflammatory drugs (NSAIDs) are potent inhibitors of prostaglandins production. NSAID is
pharmacologically targeting cyclooxygenase (COX), or PGH synthase, which catalyzes the first step of
arachidonic-acid metabolism[1, 2]. Two isoforms of the membrane COX proteins are identified[3, 4]: COX-1,
which is constitutively expressed in most tissues, to which the production of prostaglandins is attributed to; and
COX-2, which is induced by cytokines, mitogens and endotoxins in inflammatory cells[5], is implicated to the
elevated levels of prostaglandins during the inflammation. The enzymatic activity of COX involves bis-
oxygenation of arachidonic acid to PGG2, which then reduced to PGH2 in a peroxidase reaction by the same
protein[6]. NSAIDs act at the cyclooxygenase active site, and most inhibit both COX-1 and COX-2 with
minimal specificity[7], leading to serious complications such as gastric ulcers and renal toxicity[8]. It is worthy
to mention that both COX-1 and COX-2 isoenzymes are of ~60% sequence identity[7, 9]. Consistent with the
high sequence identity, overall COX-1 and COX-2 are structurally conserved (RMSD < 1.0 Å for all Cα
atoms)[10] with high significance. The structure consists of three distinct domains: N-terminal epidermal growth
factor (EGF) domain, a membrane-binding motif, and the C-terminal catalytic domain which involves the COX
and peroxidase active sites[11]. The COX active site, termed as the lobby of hydrophobic pocket, is found at C-
terminal. The active site structures between human COX-1 (hCOX-1) and hCOX-2 are quite different; in COX-
1, it is long tunnel-like space, and in COX-2, it has an accessibly extra space due to presence of what is called
side pocket[12]. This is attributed to the amino acid residue Ile 522 (Ile523 in Murine COX-1) in hCOX-1 which
is analogous to Val 509 (Val523 in Murine COX-2) in hCOX-2[13, 14] but bulkier (Figure 1)[15] in the
hydrophobic space of COX-1[16]. In COX-2, the less hindered Valine side chain and the conformational
changes at Tyr355 leaves the hydrophobic segment of the polar side pocket widely open (Figure 1).
2

ACCEPTED MANUSCRIPT
Fig. 1. Schematic comparison between COX-1 and COX-2 active binding sites.
The attempts for elaboration of selective COX-2 inhibitors are continued hoping to discover new inhibitors with
less cardiovascular (CVS) side effects as a common problem of the whole class[12, 17, 18] that made Rofecoxib
be withdrawn from the market. Thus, we selected indomethacin for the present study to be the lead compound
upon which the design of the new analogs with improved selectivity against COX-2 is based.
Indomethacin, a classic non-selective COX inhibitor, binds deeply into the COX active site. Moreover,
indomethacin penetrates furthest into the hydrophobic tunnel (Figure 2a)[19]. A 4-bromobenzyl analog, which
lacks the benzoyl oxygen, shows high selectivity to COX-2 compared to indomethacin[20], suggesting that the
benzoyl oxygen is important in enhancing the affinity to COX-1.
On the other hand, SC-558 (Figure 2b)[19], a diaryl heterocyclic inhibitor with a selectivity ratio > 375 for
COX-1 over COX-2, has central pyrazole ring substituted by a sulfonamide substituent and attached to one of
the aryl rings. In COX-2, there is a channel branches off at the Celecoxib binding site which leads from
membrane to the COX active site[10]. One branch forms a cavity that contains the bromophenyl (tolyl in case of
celecoxib) ring of SC-558, whereas the other represents a cavity not observed in COX-1 structure, and
accommodates the entire phenylsulfonamide moiety (Figure 2b). Beyond the hydrophobic pocket, the
sulfonamide group extends into the relatively polar side (selective) pocket near the surface of COX-2 (Figure
2b). The sulfonamide interacts with Gln178 (Gln192 in Murine COX-2), Leu338 (Leu352 in Murine COX-2)
and Ser339 (Ser353 in Murine COX-2) amino acids of the polar side pocket via hydrogen bonding[16].
Reports recently recorded that the selectivity of COX-2 inhibitors is attributed to the phenyl sulfonamide
moiety, which binds to the polar pocket that is not accessible in COX-1 but more accessible in COX-2, and
remains vacant in complexes of COX-2 with non-selective inhibitors[15, 21-23]. In contrary, the carboxylate
group that is a common group of various non-selective COX-inhibitors, is responsible for the conformational
change of COX-1/2 via formation of salt bridge with the basic nitrogen of Arg120 (Figure 2a) and consequently
non-selective inhibition of the enzyme is resulted[24].
3

ACCEPTED MANUSCRIPT
Fig.2. Schematic representation of SC-558 (COX-2 selective inhibitor) binding to COX-2 (a) and indomethacin (non-selective
inhibitor) (b).
Interestingly, many potent diaryl-heterocyclic selective COX-2 inhibitors carried methylsulfonyl group[25,
26]instead of sulfonamide to confirm that the sulfone radical is the important fragment for interaction with the
hydrophilic side pocket[23] and the oxidation state of the sulfur atom in sulfone form not sulfoxide as well.
Consistently, it was reported that replacement of methoxy group with sulfamate group to one of indomethacin
analogs at position 5 led to significant improvement in the selective COX-2 inhibition[27].
In view of the above findings, we have chosen the design of analogs to indomethacin prototype in which the
acetic acid moiety is cyclized into cyclohexen-one/cyclohexene and in return the indole ring is extended into
tetrahydrocarbazoles as a strategy and consequently deletion of carboxylic acid moiety, for improving the
indomethacin selective COX-2 inhibition activity. The ring extension of indole ring of indomethacin attracted
our attention aiming at proper binding of the new analogs into the wider hydrophobic[3][3][2a][28][3] lobby of
COX-2 active binding site and reducing the opportunity of these analogs to bind into the narrower binding site
counterpart in COX-1. Furthermore, salt bridge formation with Arg120 will never be accessible into COX-1
after the deletion of carboxylic acid. Thus, it has been suggested to generate 9-(4-
chlorobenzoyl/benzyl/phenylsulfonyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-ones [Figure
3 (Series I)]. Other derivatives of the analogs in which 6-methoxy group is replaced by 6-methylsulfonyl group
affording
9-(4-chlorobenzoyl/benzyl/phenylsulfonyl)-6-methylsulfonyl-2,3,4,9-tetrahydro-1H-carbazoles
[Figure 3 (Series II)] and 9-(4-chlorobenzoyl/benzyl/phenylsulfonyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-
4H-carbazol-4-ones [Figure 3 (Series III)]. This is to investigate the detrimental role of methylsulfonyl group
in the conformational change of COX-2 in the inhibition process coupled with the ring extension in improving
the selectivity of inhibition.
4

ACCEPTED MANUSCRIPT
2. Results and discussion
2.1. Chemistry
The target COX-2 inhibitors 9-(4-chlorobenzoyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-
one 4, 9-(4-chlorobenzyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one 5a and 9-(4-
chlorophenyl)sulfonyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one 5b were synthesized
according to scheme 1. The starting 6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one 3 was
previously synthesized by Weng et al adopting a multi-step method to afford the tetrahydrocarbazole in a
quantitative yield after long reaction time (12 hours) under reflux[28]. For us, we adopted another simple and
one-step method starting with commercially available starting materials in which the 4-
Methoxyphenylhydrazine hydrochloride 1 reacted with dimedone 2 in glacial acetic acid at room temperature
for 2 hours and the reaction mixture continued under reflux for further 2.5 hours. The resulting
tetrahydrocarbazole 3 was in a convenient yield 22.5% that was enough to obtain the starting material in a short
time and submit it to the next step. The melting point of 3 was measured to record 2350C that was too close to
the reported melting point (2400C).
N-benzoyl derivative 4 of the prepared tetrahydrocarbazole 3 was prepared by reacting 3 with 4-chlorobenzoyl
chloride using 4-N,N-dimethylaminopyridine (DMAP) and trimethylamine (TEA) as mild base catalysts in
dichloromethane (DCM) to afford the target product 4 in a quantitative yield after stirring at room temperature
for 2 hours.
5

ACCEPTED MANUSCRIPT
Tetrahydrocarbazole derivatives 5a,b were prepared by reacting the starting tetrahydrocarbazole 3 with 4-
chlorobenzyl bromide and 4-chlorophenylsulfonyl chloride in anhydrous dimethylformamide (DMF) using
sodium hydride as base catalyst to give the corresponding N-benzyltetrahydrocarbazole 5a and N-
phenylsulfonyltetrahydrocarbazole 5b in a quantitative yield after stirring for 1 hour at room temperature.
The starting 6-(Methylsulfonyl)-2,3,4,9-tetrahydro-1H-carbazole 8 was reported previously by Mittapalli et
al[29] in 2012 without any published spectral data about the compound but it has been mentioned that it was
prepared by one-step reaction of cyclohexanone with 4-Methylsulfonylphenyl hydrazine hydrochloride in
refluxing acetic acid overnight. It is worthy to mention that we adopted a different laboratory friendly method
and kind of green chemistry as well; in which the ketone 7 and phenyl hydrazine HCl 6 were dissolved in hot
10% aqueous sulfuric acid and left under reflux for only 4 hours as shown in scheme 2 to afford the target
tetrahydrocarbazole 8 as a heavy microcrystalline precipitate separated from the reaction mixture in 76% yield.
The product was confirmed by IR, 1H-NMR, 13C-NMR, Mass spectrometry and microanalysis.
Tetrahydrocarbazole 8 reacted with 4-chlorobenzoyl chloride in DCM in the presence of TEA and DMAP at
room temperature for 4 hours to afford the product 9 in 53% yield. The starting material was not completely
converted into the corresponding product but it was easy to purify the product by fractional crystallization in
boiling methanol.
Reaction of tetrahydrocarbazole 8 with both 4-chlorobenzyl chloride and 4-chlorophenylsulfonyl chloride went
efficiently and completely in anhydrous DMF in the presence of NaH at room temperature and stirring overnight
to afford the corresponding N-benzyl derivative 10a and N-phenylsulfonyl derivative 10b in 68% and 81%
yields respectively.
6

ACCEPTED MANUSCRIPT
The target compounds 12 were prepared as shown in scheme 3 in which 6-(Methylsulfonyl)-2,3,4,9-tetrahydro-
1H-carbazole 8 was exposed to 2 equivalents of DDQ in aqueous tetrahydrofuran (THF) that is an oxidizing
agent very specific for the type of oxidation we aimed to achieve[30]. The reaction gave immediately after few
seconds of stirring at room temperature a white fibrous crystalline precipitate of 6-(Methylsulfonyl)-1,2,3,9-
tetrahydro-4H-carbazol-4-one 11 and the reaction was continued stirring at room temperature for further 2 hours
to get it complete. The product was used for the next step without further crystallization.
Using NaH as a base catalyst in anhydrous DMF was chosen to prepare the N-benzoyl, N-benzyl and N-
phenylsulfonyl 12a, 12b and 12c derivatives respectively. Interestingly, N-benzoyl-tetrahydrocarbazole-4-one
11 couldn’t be obtained efficiently applying the previous reaction conditions that were used to prepare the N-
benzoyl derivatives 4 and 9. This is attributed to the reaction that was not complete the same as 9 but couldn’t
be purified by fractional crystallization and the TLC check showed a lower conversion of the starting
tetrahydrocarbazole 11 into the target product 12a.
1
All the final products 4, 9 and 12 were confirmed by IR, H-NMR, 13C-NMR, Mass spectrometry and
microanalysis.
7

ACCEPTED MANUSCRIPT
2.2. Biological activity:
2.2.1. Human COX-1 and COX-2 enzymatic assay
The tetrahydrocarbazoles 4, 5, 9, 10 and 12 generated for exploring it as selective COX-2 inhibitors, were tested
against both hCOX-1 and hCOX-2 isoenzymes as an in vitro preliminary study to investigate the potential of the
compounds to inhibit selectively COX-2 using Indomethacin and Celecoxib as standard non-selective and
selective inhibitors respectively.
The results shown in (Table 1) revealed the discovery of new analogs of indomethacin 10a (IC50 = 0.28), 10b
(IC50 = 0.34) and 12a (IC50 = 0.23) with potential COX-2 inhibition and comparable activities to Celecoxib (0.3
µmol). It is worthy to emphasize that 9-(4-chlorobenzoyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-
one 12a performed 1.3 times more potency than Celecoxib against COX-2 and 400 times less potency (IC50
=
104 µmol) than Indomethacin (IC50 = 0.26 µmol) against COX-1 to infer a future candidate with impressively
selective COX-2 inhibition activity. Moreover, the superior selectivity ratio of 12a (452.1739) reflected the
outstanding selective COX-2 inhibition activity and made the minimal gastric irritation side effect be greatly
expected.
Interestingly, the 6-methoxy derivatives of the generated tetrahydrocarbazoles 4 (IC50 = 0.97 µmol), 5a (IC50
=
0.91 µmol), and 5b (IC50 = 0.88 µmol) showed more potent inhibition activity against COX-2 than
indomethacin (IC50 = 2.63 µmol) by 3.23, 3.03, and 2.93 times respectively but less potent than Celecoxib
against the same enzyme. Thus, there might be another mechanism by which the new tetrahydrocarbazoles 4,5
inhibited COX-2 other than the interaction with the hydrophilic side pocket like what happens in case of
Celecoxib and the formation of salt bridge with Arg120 like what happens in case of indomethacin. Moreover,
the impressive decrease in the inhibition activity of the corresponding analog 4 to indomethacin against COX-1
(IC50 = 201 µmol) made us conclude that our hypothesis regarding the design of ring-extended analogs of
indomethacin was successfully verified. Thus, 9-(4-Chlorobenzoyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-
8

ACCEPTED MANUSCRIPT
4H-carbazol-4-one 4 is expected to perform more anti-inflammatory activity with much less gastric irritation
side effect than the original indomethacin prototype.
Table 1. In Vitro hCOX-2 and hCOX-1 enzymes inhibitory activities of compounds 4, 5, 9, 10, and 12.
Approximate selectivity ratiob COX-1/
Test compound
COX-2 IC50a (µmol)
COX-1 IC50a (µmol)
COX-2
2.63±0.0021
0.3±0.0015
0.26±0.0043
100±3.45
0.098859
Indomethacin
Celecoxib
333.3333
207.2165
195.6044
189.7727
187.6543
0.97±0.0093
0.91±0.0082
0.88±0.0074
0.81±0.0063
0.28±0.0043
201±5.44
178±4.55
167±3.76
152±4.67
118±3.93
4
5a
5b
9
10a
10b
12a
12b
421.4286
352.9412
452.1739
236.2069
0.34±0.0024
0.23±0.0034
0.58±0.0051
0.79±0.0072
120±1.06
104±4.54
137±2.97
145±3.88
12c
183.5443
aThe in vitro test compound concentration required to produce 50% inhibition of COX-1 or COX-2.
bThe in vitro COX-2 selectivity ratio (COX-1 IC50/COX-2 IC50).
2.2.2. Carrageenan-induced rat paw edema
The synthesized analogs, 6-methoxytetrahydrocarbazoles 4, 5 and 6-methylsulfonyltetrahydrocarbazoles 9,10,
and 12 were subjected to in vivo testing implementing Carrageenan-induced rat paw edema bioassay.
According to the data results expressed in %protection; listed in (Table 2), 6-methylsulfonyltetrahydrocarbazole
12a (88.05668%) showed the best %protection against Carrageenan-induced paw edema in rats to be more than
that was shown by indomethacin (74.49%) and comparable to that of Celecoxib (96.73%). Interestingly, the
results went aligned with what we got from the in vitro testing against hCOX-1 and hCOX-2 enzymatic assay
(Table 1) except for the Celecoxib that gave %protection more than 12a. It is important to emphasize that the
extremely high %protection from the inflammation, sometimes, is not in the favor of the tested compound to be
a future anti-inflammatory agent because it might magnify the cardiovascular (CVS) side effects of selective
COX-2 inhibitors as a common class effect and vice versa. Thus, it is expected for compound 12a to be
potentially selective COX-2 inhibitor with less CVS problems than that reported for Celecoxib. Other
indomethacin analogs exhibited higher %protection (Table 2) than the indomethacin prototype, to prove the
significant improvement in the anti-inflammatory potential of the corresponding ring-extended new candidates.
Table 2. Effects of the tested compounds 4,5,9,10, and 12 on Carrageenan-induced rat paw edema
(mL), percentage protection and activity relative to indomethacin.
Test compounds
Increase in paw edema (mL) ± SEMa,b
0.988±0.0006
% Protection
0
Control
0.25±0.0005
74.49±4.44
Indomethacin
9

ACCEPTED MANUSCRIPT
0.291±0.00053
0.201±0.00024
0.194±0.00015
0.191±0.00026
0.188±0.00037
0.123±0.00022
0.145±0.00034
0.118±0.00043
0.167±0.00015
0.179±0.00026
96.73±3.44
Celecoxib
79.65587±2.43
80.36437±3.54
80.66802±4.45
80.97166±3.66
87.55061±2.35
85.32389±2.63
88.05668±4.34
83.09717±3.74
81.88259±4.55
4
5a
5b
9
10a
10b
12a
12b
12c
a SEM denoted the standard error of the mean.
b All data are significantly different from control (P < 0.001).
2.2.3. Estimation of plasma prostaglandin E2 (PGE2)
Measuring the percentage plasma levels of prostaglandin E2 (PGE2) after treating the animals with the COX-2
inhibitors using Indomethacin and Celecoxib as standard inhibitors is one of the important parameters to assess
the anti-inflammatory potencies of the COX-2 inhibitors in vivo. The results are shown in (Table 3), revealed
the highest anti-inflammatory potency with compound 12a to record %inhibition of plasma PGE2 = 91.29 to be
of superior PGE2-diminishing activity to Celecoxib ((%inhibition = 77.25). All other inhibitors performed
higher potential as PGE2 lowering agents than Celecoxib but quite lower than indomethacin (%inhibition =
98.29) especially for 6-methoxy congeners 4, 5a and 5b (80.24%, 81.16% and 82.16% respectively).
Table 3. Anti-inflammatory potencies of indomethacin analogs 4,5,9, and
10, and 12 (%inhibition of plasma PGE2).
Test compound
Inhibition of plasma PGE2 [%]±SEMa,b
98.29±7.5
77.25±6.6
Indomethacin
Celecoxib
80.24±3.5
81.16±4.4
82.16±3.6
83.19±2.5
89.43±3.8
88.20±4.9
91.29±4.7
86.39±5.8
84.11±4.7
4
5a
5b
9
10a
10b
12a
12b
12c
a SEM denoted the standard error of the mean.
b All data are significantly different from control (P < 0.001).
2.2.4. Human, Rat, and Dog Microsomal COX Assays
10

ACCEPTED MANUSCRIPT
The resulted indomethacin analogs 4, 5, 9, 10 and 12 in this study were tested against human, dog and rat
microsomal COXs to evaluate their nanomolar inhibition activities using Celecoxib as standard inhibitor. As
shown in (Table 4), compound 12a (IC50 = 23 nM) was the most potential human COX inhibitor and it
performed lower potential activity (IC50= 34 nM, and 48 nM) against dog and rat microsomal COX respectively.
Additionally, Compound 12a and all the other indomethacin analogs 4, 5, 9, and 10 were significantly and in a
selective fashion in the favor of human microsomal COX to be more potent against human, dog and rat
microsomal COX than the standard inhibitor (IC50 = 89 nM, 112 nM and 132 nM) respectively.
Table 4. Effect of indomethacin analogs 4,5,9,10, and 12 on Human, Dog, and Rat Microsomal COX
Activities
IC50 nM
Test compound
Human
89±5.6
50±3.6
44±3.8
41±2.7
38±1.9
27±2.7
28±1.6
23±1.6
33±2.7
34±1.8
Dog
Rat
112±9.7
80±5.5
77±4.7
61±8.6
56±4.4
39±1.6
40±2.5
34±1.7
45±3.4
53±3.3
132±8.5
141±9.6
132±8.8
128±9.7
116±7.8
56±4.7
79±3.6
48±2.6
93±5.7
104±6.9
Celecoxib
4
5a
5b
9
10a
10b
12a
12b
12c
Values were calculated from the mean values of data from three separate experiments and presented as mean
value ± SEM.
All results are significantly different from control values at p ≤ 0.005.
2.2.5. Cotton pellet-induced granuloma bioassay
New indomethacin analogs 4, 5, 9, 10, and 12 were subjected to Cotton pellet-induced rat granuloma bioassay
and the inhibition activities are listed in (Table 5) expressed in ED50. Consistently, 6-
methylsulfonyltetrahydrcarbazole-4-one 12a (ED50 = 12.38 µmol) performed the highest anti-inflammatory
activity as it inhibited the Cotton pellet-induced rat granuloma 0.75 times of indomethacin (ED50 = 9.568 µmol)
and 6.95 times of Celecoxib (ED50 = 86.11 µmol) though the comparable activity of 12a to Celecoxib in the
enzymatic assay (Table 1). This might be attributed to the metabolic enzymes effect on compound 12a that led
to more active metabolite. Or, the bioavailability of 12a might have played a significant role in such in vivo
bioassay results.
The other tetrahydrocarbazoles 4, 5, 9, 10, and 12b,c gave significantly higher inhibition activities than
Celecoxib (Table 5). Based on the ED50 recorded for the new analogs, we could conclude that we have
interesting compounds, combine between the potential anti-inflammatory (Table 5) and selective activities
(Table 1), excel the anti-inflammatory profile of both indomethacin and Celecoxib.
11

ACCEPTED MANUSCRIPT
Table 5. Anti-inflammatory potencies of indomethacin
analogs 4,5,9,10, and 12 (inhibition of Cotton pellet-induced
granuloma in ED50, µmol).
Test compounds
ED50 (µmol)
9.568±0.87
86.11±1.23
28.27±1.90
17.27±1.56
26.88±1.36
25.67±1.67
14,56±0.99
18.67±1.34
12.38±0.89
19.20±1.22
22.33±1.34
Indomethacin
Celecoxib
4
5a
5b
9
10a
10b
12a
12b
12c
2.2.6. Ulcerogenic effects
The percentage of ulcerogenic activity of indomethacin analogs 4, 5, 9, 10 and 12 were calculated after exposure
of the experimental animals to the tested compounds using indomethacin and Celecoxib as standard drugs.
According to the data recorded for each compound in (Table 6), compound 12a exhibited the lowest
ulcerogenic activity (3.28%) to be 30.48 times safer than indomethacin (100%) and 0.64 times as safe as
Celecoxib (2.11%). All the other analogs showed significant reduction in the ulcerogenic activity when
compared to indomethacin and the maximum percentage of ulceration was for 6-methoxytetrahydrocarbazole 5b
(20.12%). This indicates that the safety margin of the new analogs became impressively wider than the original
prototype to reflect the real existence of compounds with potential anti-inflammatory activity (Tables 2, 3, 5)
and greatly lower ulcerogenic activity, specifically those are 6-methylsulfonylphenyltetrahydrocarbazole
derivatives 10a,b and 12a,b though it didn’t excel the Celecoxib’s safety against ulcerogenic activity.
Table 6. The percentage ulcerogenic activitya of
Indomethacin analogs 4, 5, 9 ,10, and 12
Test compounds
% Ulceration
100±
Indomethacin
2.11±0.23
13.18±0.23
11.20±0.12
20.12±0.33
12.45±0.24
4.48±0.54
2.90±0.63
Celecoxib
4
5a
5b
9
10a
10b
12

ACCEPTED MANUSCRIPT
12a
12b
12c
3.28±0.12
5.40±0.52
11.29±0.45
a All data (tested compounds, indomethacin and
Celecoxib were significantly different from control (P < 0.001)
2.3. Molecular docking studies:
Molecular docking was done in order to predict the molecular orientation of the synthesized compounds into the
COX-2 binding site and to interpret the biological activities results. In addition to the docking score, some other
parameters such as; the lipophilic contribution score, clash score and conformation entropy score were
computed to find out a suitable correlation to the biological results (Table 7).
Table 7. Molecular docking results of the synthesized compounds using Leadit 2.1.2
Docking score
Test compound
Lipo score
Clash score
Rot Score
Kcal/mol
-18.98
-15.60
-21.85
-23.92
-23.36
-23.40
-30.78
-23.05
-23.95
-31.23
-35.41
-16.85
-17.31
-16.21
-16.65
-17.14
-15.21
-15.41
-15.85
-16.75
-14.66
-15.59
10.34
12.00
10.78
9.34
8.90
8.36
7.21
8.65
9.28
7.03
9.67
4.20
4.20
1.40
1.40
4.20
4.20
1.40
4.20
4.20
4.20
4.20
4
5a
5b
9
10a
10b
12a
12b
12c
Indomethacin
Celecoxib
Docking score (Kcal/mol): free binding energy value. Lipo score: Lipophilic contribution score.
Clash score: Contribution of the clash penalty. Rot score: Ligand conformational entropy score
Computing of the docking score was able to compare the binding free energies of the synthesized compounds
and comparing them to both Indomethacin and Celecoxib. Compound 12a showed the best docking score (-
30.78 kcal/mol) which is very close to that of both Indomethacin and Celecoxib (-31.23 and -35.41 kcal/mol
respectively). The lipophilic contribution score values (Lipo score) listed in (Table 7) inferred its insignificant
effect. The clash penalty score should be a small value to avoid the steric repulsion of the generated
conformations. It was observed that compound 12a and Indomethacin shared the common feature of low clash
score 7.21 and 7.03 respectively. The Ligand conformational entropy score (Rot score) is implemented in Leadit
software to compute the effect of the degree of freedom of the flexible compounds that can affect their rotations
and orientations. It is much better to have a neglected or a small value for this score. From all the docked
compounds, derivatives 5b, 9, and 12a only showed the lowest value (1.40) for the Ligand conformational
entropy score which confirmed their stability in the active site. All compounds featured a common orientation
13

ACCEPTED MANUSCRIPT
mode in which the p-chlorobenzyl or p-chlorobenzoyl moieties were superimposed to that of Indomethacin. The
docking scores illustrated in (Table 7) were proportional to the inhibitory IC50 values of COX-2 (Figure 4).
Fig. 4. Correlation between the docking scores and the IC50 of new COX-2 inhibitors.
Celecoxib is a selective COX-2 inhibitor which has a unique binding mode in a polar side pocket found in COX-
2 isozyme. The reason is correlated to its SO2-NH2 group that has the ability to form many interactions with
residues found in this side pocket such as Arg499, Ser339, Leu338, and Gln178 (Figure 5A). The crystal
structure of the most active compound 12a in complex with COX-2 that was subjected to MD simulations was
used to inspect the interactions of this compound with these specified residues. Upon visualization of the formed
interactions we found that the methyl sulfonyl group CH3-SO2 of this compound has a hydrogen bond with
Arg499 and Ser339 (Figure 5B). It was clear that the impact of the –NH2 in SO2-NH2 side chain of Celecoxib is
higher than the CH3- group in SO2-CH3 in our compound and that may be a start point for an optimization
process in the future for our lead compound.
14

ACCEPTED MANUSCRIPT
Fig. 5. A) Best binding mode of Celecoxib to COX-2. B) Best binding pose of (12a) into COX-2 binding site.
The binding site is represented as cartoon in grey color. Celecoxib was built as stick form (Yellow color). Compound (12a)
was built as ball and stick (Green color). Hydrogen bonds of ligand atoms with the amino acid residues of binding site are
in light blue dotted lines.
The docking results of compounds 4, 5a and 5b with the lowest IC50 values in the in vitro COX-2 inhibitory
assay revealed that these compounds showed also low predicted docking score; -18.98, -15.60, and -21.85
kcal/mol respectively. In addition, the clash contribution score of these compounds (10.34, 12.00, and 10.78
respectively) was higher than the most active compound 12a (7.21).
15

ACCEPTED MANUSCRIPT
Fig. 6. The best pose resulted from the docking of A) compound (4), B) compound (5a) and C) compound (5b) into COX-2 binding
site. The binding site is represented in cartoon (grey color), Indomethacin is in stick form (green color). The synthesized
derivatives are in ball and stick form (golden color). Hydrogen bonds of ligand atoms with the amino acid residues of
binding site are in light blue dotted lines.
Compound 4 showed the same orientation as indomethacin (Fig. 6A), its CO of the cyclohexyl ring interacts via
hydrogen bond formation with Arg120 while its benzoyl-CO showed hydrogen bond with Ala527. The same
interaction of the cyclohexane-CO was also observed in both 5a (Fig. 6B) and 5b (Fig. 6C).
2.4. Generalized Born Poisson Boltzmann (MM/GBVI)
Molecular mechanics can be used to compute the non-bonded interactions. Generalized Born Poisson
Boltzmann (MM/GBVI) was used to calculate the binding strengths of the non-bonded interactions in kcal/mol
(Table 8). It is used to estimate the relative binding strengths via estimation of the change of enthalpy takes
place upon binding. It was clear that compounds 10a, 12b, 10b, 5b, showed -21.13 kcal/mol, -18.44 kcal/mol, -
15.78 kcal/mol, and -13.78 kcal/mol respectively. When these values were compared to that of Celecoxib (-
16.37 kcal/mol), it seemed that they were close to each other in this kind of interactions. On the other hand,
compound 5b showed -26.90 kcal/mol and was the only compound to be close to Indomethacin’s value (-25.61
kcal/mol).
16

ACCEPTED MANUSCRIPT
Table 8. Some calculated parameters for the synthesized compounds
Affinity
MM/GBVI
Kcal/mol
Test compound
TPSA
ClogP
pki
48.30
31.23
48.30
56.14
39.07
73.21
73.21
56.14
90.28
71.36
77.98
5.14
5.77
5.14
4.26
4.89
3.81
3.90
4.53
3.45
2.59
3.82
-11.94
-26.70
-13.78
-11.11
-21.13
-15.78
-12.36
-18.44
-15.75
-25.61
-16.37
10.68
9.88
4
5a
9.26
5b
9.46
9
10a
9.00
9.56
10b
10.67
10.16
10.46
10.21
11.47
12a
12b
12c
Indomethacin
Celecoxib
TPSA: Topological Polar Surface Area. ClogP: calculated logarithm of
compound partition coefficient. MM/GBVI: Generalized Born Poisson
Boltzmann
All compounds had a ClogP value within the accepted range of the drug-like properties. The affinity pki value
of Celecoxib was the highest (11.47). Compounds 4 and 12a showed the top two values of pki; 10.68 and 10.67
respectively.
2.5. Topological Polar Surface Area (TPSA)
The limited number of the synthesized compounds (only 9 compounds) made it difficult to perform 3D QSAR
study in order to find out the descriptors to which the biological data may be correlated. Instead, another 2D
descriptor can be used to measure the polar surface area. TPSA, was calculated (Table 8). It measures the polar
surface area that can be useful when predicting an agent’s bioavailability. Compounds with low values (i.e.,
<75) were identified to be associated with various side effects. According to the results, Celecoxib had a value
of 77.98 which confirmed its high polar surface area and COX-2 selectivity toward binding to the polar cavity
found in COX-2. It was clear that Compound 12c was the highest value (90.28). Compounds 10b, and 12a
showed the same TPSA (73.21). Compounds 9 and 12b had the same as well (56.14) and finally, both
compounds 4 and 5b showed similar TPSA (48.30). The inhibitors’ IC50 values for COX-2 were proportional to
TPSA values (Figure 7).
17

ACCEPTED MANUSCRIPT
Fig. 7. Correlation between IC50 of new COX-2 inhibitors and TPSA
2.6. Molecular Dynamic simulation.
Molecular dynamic relaxation was done to validate the stability of the most active compound 12a within the
active site of COX-2. Time of the dynamics process was computed over 10000 picoseconds (ps) period of time,
the potential energy of the atomic system in kcal/mol, and the kinetic energy of atoms in kcal/mol. The
simulations were done in two steps. First, NAMD was obtained from university of Illinois
(www.ks.uiuc.edu/researc/namd.) The simulations’ running started and after 10000 ps, the trajectory file was
imported and analyzed. Here, the potential energy (Kcal/mol) showed a sharp increase with time then a decline
in which the steady state was not observed (Figure 8A). Regarding the kinetic energy (Kcal/mol) of the atomic
system, it did not give an ideal curve where the start was low kinetic value and a steady state was absent as well
(Figure 8B). The ligand-COX-2 complex resulted from NAMD MD was used in the second step that was
conducted by MOE molecular dynamic simulations. Here, the potential energy showed almost a typical mode of
minimization where the energy increased with time until a steady state was reached (Figure 9A). The kinetic
energy of the 12a-COX-2 complex started by a scattered mode of atoms followed by a steady state as well
(Figure 9B). The ligand-COX 2 complex now can be considered to be in a stable form where we can study the
effects of MD on the interactions of the most active compound. The main difference between the potential
energy profile and that of kinetic energy profile of 12a-COX-2 complex can be observed in the figures where
the equilibrium was reached in case of potential energy at 15 ps (Figure 9A) while the oscillations in case of the
kinetic energy reached the equilibrium at 20 ps (Figure 9B) which is a very small difference.
18

ACCEPTED MANUSCRIPT
Fig. 8. NAMD showing the relation between both the A) Potential energy, B) Kinetic
energy and time
Fig. 9. MD simulations represents the points of oscillations of the complex (12a-COX-2) atoms showing: A)
Potential energy (Kcal/mol) and B) Kinetic energy (Kcal/mol) profile resulted
from MD simulation.
19

ACCEPTED MANUSCRIPT
3. Conclusion:
The present study introduced a new perspective and novel strategy that hasn’t been discussed or even suggested
before for the improvement of the selectivity of indomethacin against COX-2 isoenzyme. The new strategy
involved i) ring extension, ii) Deletion of carboxylic acid and iii) introduction of methylsulfonyl group to reduce
the possibility of COX-1 inhibition by the new candidates and increase the opportunity of COX-2 inhibition as
well. The biological activity data profile of the generated ring-extended analogs verified the success of the
strategy provided, that confirmed the efficacy of the analogs as anti-inflammatory agents against COX-2 in a
potentially
selective
fashion
excelling
the
selectivity
ratio
of
Celecoxib
with
6-
methylsulfonypheyltetrahydrocarbazoles 10a and 12a. In addition, all the new analogs showed high safety
margin against gastric ulceration in comparison to indomethacin and the safety was so close to Celecoxib with
6-methylsulfonypheyltetrahydrocarbazoles 10a,b and 12a,b to confirm the necessity of methylsulfonyl group in
improving the selectivity of new inhibitors against COX-2. The COX-1 and COX-2 inhibition potencies of 6-
Methoxytetrahydrocarbazoles 4 and 5 proved the significant effect of ring extension and deletion of carboxylic
acid on increasing the selectivity ratio and COX-2 inhibition potency as well. The high PGE2 lowering potential
of 12a with the other anti-inflammatory descriptors data profile highlights this analog as a very potential and
selective COX-2 inhibitor. The docking scores obtained for the new candidates confirmed the potential
selectivity of 12a by having the highest binding affinity. The molecular dynamic study indicated the stability of
12a into the catalytic binding site. Conclusively, the study provided the field with new class of potentially
selective COX-2 inhibitors is superior to Celecoxib based on the in vitro, in vivo biological data profiling
recorded. It is worthy to mention that the current investigation paved the road for the researchers including us to
elaborate in the near future other new classes of COX-2 inhibitors beyond the diarylsulfonamides class applying
the new strategy that is introduced by us.
4. Experimental:
3.1. Chemistry
Melting points were determined on digital Gallen-Kamp MFB- 595 instrument using open capillary tubes and
are uncorrected. IR spectra were recorded as potassium bromide discs on Schimadzu FT-IR 440 spectrometer.
1H NMR spectra were recorded on Bruker spectrophotometer at 300 MHz in DMSO-d6; values (δ) are given in
parts per million (ppm) downfield from tetramethylsilane (TMS) as internal reference. 13C NMR spectra were
1
recorded using the same spectrophotometer that used for recording H NMR. Mass spectra were recorded on
Thermo Triple stage quadropole mass spectrometer. The elemental analyses were performed at the
Microanalytical Center, Cairo University, Cairo, Egypt. Reactions were followed up by thin layer
chromatography (TLC) using Merck Silica gel/TLC cards with fluorescent indicator UV254 using Hexane:Ethyl
acetate (EtOAc) 1:1 as the eluting system and the spots were visualized using Spectroline E series dual
wavelength UV lamp at λ=254 nm.
Synthesis of 6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one (3)
20

ACCEPTED MANUSCRIPT
A mixture of 4-methoxypenylhydrazine HCl (3 mmol, 0.524 g) and dimedone (3 mmol, 0.42 g) were suspended
in acetic acid glacial (15 mL) and left stirring at room temperature for 2 hours. The temperature of the reaction
mixture was raised to reflux for 2.5 hours. The solvent was evaporated under reduced pressure to give dark
brown sticky residue. The brown residue was subjected to column chromatography and eluted with Hexane:
EtOAc 1:1 to give beige powder of the title compound in a pure form.
Yield: 22.5%, m.p.: 235 0C. IR (KBr) ʋmax/cm-1: 3172 (NH), 1616 (CO). 1H NMR (DMSO-d6,300 MHz): δ 1.08
(s, 6H, 2,2-di-CH3), 2.31 (s, 2H, 3-H), 2.82 (s, 2H, 1-H), 3.76 (s, 3H, CH3O) 6.80 (d, 1H, 7-H), 7.29 (d, 1H, 8-
H), 7.45 (s, 1H, 5-H), 11.65 (s, 1H, NH).
Synthesis of 9-(4-Chlorobenzoyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one (4)
4-Chlorobenzoyl chloride (43 mg, 0.247 mmol) was added to a stirred suspension of tetrahydrocarbazole (50
mg, 0.205 mmol) in DCM (5 mL), DMAP (14.9 mg, 0.151 mmol) and TEA (0.15 mL) were then added. The
reaction mixture got dissolved into solution after addition of the base catalysts, was left to stir at room
temperature until the reaction got complete after 2 hours according to TLC test. The solvent was removed under
reduced pressure and the resulting residue was washed several times with diethyl ether to give white needle
crystals of the target compound with no need to further purification.
0
1
Yield: 40%, m.p.: 190-192 C; IR (KBr) ʋmax/cm-1: 1696 (CO ketone), 1658 (CO amide). H NMR (DMSO-
d6,300 MHz): δ 7.78 (d, J=8.1 Hz, 2H, 2,6-H), 7.67 (d, J=8.2 Hz, 2H, 3,5-H,), 7.59 (s, 1-H, 5-H), 7.06 (d, J=9.0
Hz, 1H, 8-H) 6.83 (d, J=9.0 Hz, 1H, 7-H), 3.79 (s, 3H, CH3O), 2.76 (s, 2H, 1-H), 2.42 (s, 2H, 3-H), 1.02 (s, 6H,
2,2-di-CH3). 13C NMR (DMSO-d6,75 MHz): δ 27.86 (CH3), 35.18 (C-2), 38.31 (C-1), 51.30 (C-3), 55.37
(CH3O), 103.12, 112.75, 114.97, 115.05, 125.99, 129.15, 130.73, 131.71, 132.55, 138.64, 151.09, and 156.58
(Ar-C), 167.78 (CO-N), 194.31 (CO).MS LRMS (ESI): (calc) 381.11 (found) 382.15 (MH)+. Anal. Calcd for
C22H20ClNO3: C,69.20; H, 5.28; N, 3.67. Found: C, 69.15; H, 5.44; N, 3.46.
Synthesis of 9-(4-Chlorobenzyl)-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one (5a)
Tetrahydrocarbazole (50 mg, 0.205 mmol) was dissolved in dry DMF (1 mL) and the stirred solution was
treated with NaH in oil dispersion 60% (17.8 mg, 0.266 mmol). The reaction mixture was stirred at room
temperature for 10 min until the effervescence was ceased and then treated with 4-chlorobenzyl chloride (40 mg,
0.246 mmol), the stirring was continued at room temperature for 2 hours and poured onto ice-cold water. The
resulting precipitate was filtered off and recrystallized from ethanol affording beige crystals of the title
compound.
Yield: 54%; m.p.: 163-165 0C. IR (KBr) ʋmax/cm-1 1640 (CO). 1H NMR (DMSO-d6,300 MHz): δ 7.52 (s, 1H, 5-
H), 7.38 (d, J=6.7 Hz, 3H, 8-H, 2,6-H), 7.08 (d, J=8.1 Hz, 2H, 3,5-H), 6.80 (d, J=8.8 Hz, 1H, 7-H) 5.46 (s, 2H,
4-Cl-C6H4-CH2), 3.77 (s, 3H, CH3O), 2.84 (s, 2H), 2.35 (s, 2H), 1.07 (s, 6H, 2,2-di-CH3). 13C NMR (DMSO-
d6,75 MHz): δ 28.14 (CH3), 34.93 (C-2), 35.23 (C-1), 45.57 (4-Cl-C6H4-CH2), 51.53 (C-3), 55.31 (CH3O),
102.62, 110.53, 111.40, 111.74, 124.94, 128.25, 128.69, 131.71, 131.99, 136.09, 151.40, and 155.71 (Ar-C),
192.14 (CO). MS LRMS (ESI): (calc) 367.13 (found) 368.10 (MH)+. Anal. Calcd for C22H22ClNO2: C,71.83; H,
6.03; N, 3.81. Found: C, 71.91; H, 6.20; N, 3.62
21

ACCEPTED MANUSCRIPT
Synthesis of 9-[(4-Chlorophenyl(sulfonyl)]-6-methoxy-2,2-dimethyl-1,2,3,9-tetrahydro-4H-carbazol-4-one
(5b)
Tetrahydrocarbazole (50 mg, 0.205 mmol) was dissolved in dry DMF (1 mL) and the stirred solution was
treated with NaH in oil dispersion 60% (17.8 mg, 0.266 mmol). The reaction mixture was stirred at room
temperature for 10 min until the effervescence was ceased and then treated with the appropriate 4-
chlorophenylsulfonyl chloride (52 mg, 0.246 mmol), the stirring was continued at room temperature for 2 hours
and poured onto ice-cold water. The resulting precipitate was filtered off and recrystallized from ethanol
affording beige crystals of the title compound.
0
1
Yield: 62%; m.p.: 175 C. IR (KBr) ʋmax/cm-1 1656 (CO). H NMR (DMSO-d6,300 MHz) δ 8.01 (t, J=8.4 Hz,
3H), 7.70 (d,J=8.3 Hz, 2H), 7.54 (s,1H), 7.00 (d, J=9.3 Hz, 1H), 3.78 (s, 3H, CH3O), 3.24 (s, 2H), 2.43 (s, 2H),
1.07 (s, 6H, 2,2-di-CH3). 13C NMR (DMSO-d6,75 MHz): δ 27.75 (CH3), 34.81 (C-2), 37.35 (C-1), 51.02 (C-3),
55.37 (CH3O), 103.80, 113.80, 114.74, 126.31, 128.52, 129.49, 130.29, 135.82, 140.27, 150.55, and 157.19 (Ar-
C), 194.21 (CO). MS LRMS (ESI): (calc) 417.08 (found) 418.12 (MH)+. Anal. Calcd for C22H22ClNO2: C,
60.36; H, 4.82; N, 3.35. Found: C, 60.24; H, 4.67, N, 3.52
Synthesis of 6-(Methylsulfonyl)-2,3,4,9-tetrahydro-1H-carbazole (8)
A mixture of 4-methylsulfonylphenylhydrazine HCl (1.11 g, 5 mmol) and cyclohexanone (5 mmol, 0.50 mL)
0
were heated at 80 C for 5 min, 10% aq.H2SO4 (20 mL) was added to the solid mixture followed by vigorous
stirring until complete dissolution. The temperature was raised to reflux; greenish-light brown solid crystalline
product was separated out from the solution after 1 h of the reaction time, and the reflux continued to 4 hours.
The reaction mixture was left to cool, diluted with 20 mL H2O for more precipitation of the solid product, the
solid product was filtered off, washed with H2O (10 mL) and left to dry under vacuum. No further
crystallization for purification was required and the product submitted directly to the next step.
0
Yield: 76%, m.p.: 169 C, IR (KBr) ʋmax/cm-1 3321 (NH), 1287 (CH3SO2). 1H NMR (DMSO-d6,300 MHz) δ
11.28 (s, 1H, NH), 7.91 (s, 1H, 5-H), 7.51 (d, J=8.60 Hz, 1H, 7-H), 7.43 (d, J=8.4 Hz, 1H, 8-H), 3.12 (s, 1H,
CH3SO2), 2.70 (m, 4H, 1,4-H), 1.83 (m, 4H, 2,3-H). 13C NMR (DMSO-d6,75 MHz): 20.30 (C-4), 22.57 (C-2,3),
22.68 (C-1), 44.68 (CH3SO2), 109.79, 110.90, 117.15, 118.45, 130,34 and 137.61, 137.78 (Ar-C). MS LRMS
(ESI): (calc) 249.08 (found) 250.03 (MH)+. Anal. Calcd for C12H14N2O2S: C,62.63; H, 6.06; N, 5.62. Found: C,
62.47; H, 6.23; N, 5.79.
Synthesis of 9-(4-Chlorobenzoyl)-6-methylsulfonyl-2,3,4,9-tetrahydro-1H-carbazole (9)
4-Chlorobenzoyl chloride (0.26 mL, 2.0 mmol) was added to a stirred suspension of tetrahydrocarbazole (249
mg, 1.0 mmol) in DCM (10 mL), DMAP (80 mg, 0.67 mmol) and TEA (0.75 mL) were then added. The
reaction mixture got dissolved into solution after addition of the base catalysts, was left stirring at room
temperature overnight. The solvent was removed under reduced pressure and the resulting residue was washed
several times with diethyl ether. The solid residue recrystallized from ethanol to give off white microcrystalline
product of the target compound.
22

ACCEPTED MANUSCRIPT
0
Yield: 53%, m.p.: 207 C, IR (KBr) ʋmax/cm-1 1685 (CO), 1314 (CH3SO2). 1H NMR (DMSO-d6,300 MHz) δ
8.04 (s, 1H, 5-H), 7.70 (dd, J=23.3, 8.5 Hz, 5H, 2,6-H, 3,5-H, 7-H), 7.49 (d, J=8.8 Hz, 1H, 8-H), 3.21 (s, 3H,
CH3SO2), 2.72 (t, 2H, 4-H), 2.47 (t, 2H, 1-H), 1.76 (m, 4H, 2,3-H). 13C NMR (DMSO-d6,75 MHz): 20.31 (C-4),
21.48 (C-3), 22.71 (C-2), 25.09 (C-1), 44.03 (CH3SO2), 114.69, 117.15, 117.43, 121.71, 129.05, 129.29, 131.26,
133.45, 135.06, 137.90, 138.17 (Ar-C), and 167.66 (CO-N). MS LRMS (ESI): (calc) 381.11 (found) 382.15
(MH)+. Anal. Calcd for C12H14N2O2S: C,61.93; H, 4.68; N, 3.61. Found: C, 62.11; H, 4.81; N, 3.45.
Synthesis of 9-(4-Chlorobenzyl)-6-methylsulfonyl-2,3,4,9-tetrahydro-1H-carbazole (10a)
Tetrahydrocarbazole (249 mg, 1.0 mmol) was dissolved in dry DMF (3 mL) and the stirred solution was treated
with NaH in oil dispersion 60% (52 mg, 1.3 mmol). The reaction mixture was stirred at room temperature for 1
hour and then treated with 4-chlorobenzyl chloride (193 mg, 1.2 mmol), the stirring was continued at room
temperature overnight and then poured onto ice-cold water to give solid product. The solid product filtered off,
dried under vacuum and suspended in ether with vigorous shaking then filtered off to give white powder of the
corresponding N-benzyltetrahydrocarbazole in a pure form.
Yield: 68%, m.p.: 153 0C, IR (KBr) ʋmax/cm-1 1302 (CH3SO2). 1H NMR (DMSO-d6,300 MHz) δ 8.00 (s, 1H, 5-
H), 7.62 (d, J=8.60 Hz, 1H, 7-H), 7.59 (d, J=8.60 Hz, 1H, 8-H), 7.36 (dd, J=8.1 Hz, 2H, 2,6-H), 7.03 (dd, J=8.1
Hz, 2H, 3,5-H), 5.43 (s, 2H, 4-Cl-C6H4-CH2), 3.15 (s, 3H, CH3SO2), 2.72 (m, 4H, 1,4-H), 1.81 (m, 4H, 2,3-H).
13C NMR (DMSO-d6,75 MHz): 20.35 (C-4), 21.57 (C-2,3), 22.40 (C-1), 44.49 (CH3SO2), 45.11 (4-Cl-C6H4-
CH2), 109.85, 110.76,117.52, 118.88, 126.30, 128.21, 128.65, 131.04, 131.82, 136.91, 138.09, and 138.59 (Ar-
C). MS LRMS (ESI): (calc) 387.07 (found) 388.12 (MH)+. Anal. Calcd for C12H14N2O2S: C,64.25; H, 5.39; N,
3.75. Found: C, 64.42; H, 5.26; N, 3.57
Synthesis of 9-[(4-Clorophenyl)sulfonyl)]-6-methylsulfonyl-2,3,4,9-tetrahydro-1H-carbazole (10b)
Tetrahydrocarbazole (249 mg, 1 mmol) was dissolved in dry DMF (3 mL) and the stirred solution was treated
with NaH in oil dispersion 60% (21 mg, 0.521 mmol). The reaction mixture was stirred at room temperature for
1 hour and then treated with 4-chlorophenylsulfonyl chloride (274 mg, 0.522 mmol), the stirring was continued
at room temperature overnight and then poured onto ice-cold water to give solid product. The solid product
filtered off, dried under vacuum and suspended in ether with vigorous shaking then filtered off to give white
powder of the corresponding N-phenylsulfonyltetrahydrocarbazole.
Yield: 81%, m.p.: 221 0C, IR (KBr) ʋmax/cm-1 1302 (CH3-SO2). 1H NMR (DMSO-d6,300 MHz) δ 8.25 (d, J=8.8
Hz, 1H), 8.01 (s, 1H), 7.93 (d, J=8.4 Hz, 2H), 7.84 (d, J=8.8 Hz, 1H), 7.66 (d, J=8.5 Hz, 2H), 3.21 (s, 3H,
CH3SO2), 2.99 (t, J=6.3 Hz 2H, 1-H), 2.63 (t, J=5.5 Hz, 2H, 4-H), 1.85 (m, 2H, 3-H), 1.75 (m, 2H, 2-H). 13C
NMR (DMSO-d6,75 MHz): 20.32 (C-4), 21.21 (C-3), 22.51 (C-2), 24.05 (C-1), 43.90 (CH3SO2), 114.25,
117.99, 118.75, 122.57, 128.35, 129.74, 130.24, 136.10, 136.13, 137.35, 137.63 and 138.13, 139.86 (Ar-C). MS
LRMS (ESI): (calc) 423.04 (found) 424.03 (MH)+. Anal. Calcd for C12H14N2O2S: C, 53.83; H, 4.28; N, 3.30.
Found: C, 53.66; H, 4.47; N, 3.49
6-(Methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-one (11)
23

ACCEPTED MANUSCRIPT
Tetrahydrocarbazole (249 mg, 1.0 mmol) was dissolved in 5 mL THF:H2O (9:1). While the solution of
tetrahydrocarbazole was cooled in ice bath, DDQ (454 mg, 2.0 mmol) dissolved in THF (3 mL) was added
dropwise with stirring. The temperature of the reaction mixture raised to room temperature, white solid product
precipitated out from the solution after 10 min of the stirring and the reaction continued stirring for 2 hours until
completion. The precipitate was filtered off and washed with methanol (1 mL) to give fibrous white crystals of
the title compound in a pure form with no need for further crystallization.
Yield: 68%, m.p.: 327 0C, IR (KBr) ʋmax/cm-1 3139 (NH), 1634 (CO), 1291 (CH3-SO2). 1H NMR (DMSO-d6,300
MHz) δ 12.37 (s, 1H, NH), 8.47 (s, 1H, 5-H), 7.71 (d, J=8.6 Hz, 1H, 7-H), 7.63 (d, J=8.5 Hz, 1H, 8-H), 3.17 (s,
3H, CH3SO2), 3.02 (t, J=5.9 Hz, 2H, 1-H), 2.48 (m, 2H, 3-H), 2.15 (m, 2H, 2-H). 13C NMR (DMSO-d6,75
MHz): 22.70 (C-2), 23.19 (C-1) (, 37.62 (C-3), 44.47 (CH3SO2), 112.36, 119.67, 121.13, 123.90, 133.91,
138.25, and 154.94 (Ar-C), 193.13 (CO). MS LRMS (ESI): (calc) 263.06 (found) 264.02 (MH)+.Anal. Calcd
for C12H14N2O2S: C, 59.30; H, 4.98; N, 5.32. Found: C, 59.44; H, 5.10; N, 5.20
Synthesis of 9-(4-chlorobenzoyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-one (12a)
Tetrahydrocarbazole (100 mg, 0.38 mmol) was dissolved in dry DMF (1 mL) and the stirred solution was
treated with NaH in oil dispersion 60% (19.7 mg, 0.494 mmol). The reaction mixture was stirred at room
temperature for 1 hour and then treated with 4-chlorobenzoyl chloride (0.6 mL, 0.456 mmol), the stirring was
continued at room temperature overnight and then poured onto ice-cold water to give solid product. The solid
product filtered off, dried under vacuum and suspended in ether with vigorous shaking then filtered off to give
white powder of the target N-benzoyltetrahydrocarbazole.
0
Yield: 41%, m.p.: 230 C, IR (KBr) ʋmax/cm-1 1707 (CO ketone), 1661 (CO amide), 1303 (CH3-SO2). 1H NMR
(DMSO-d6,300 MHz) δ 8.62 (s, 1H, 5-H), 7.83 (t, J=8.9 Hz, 3H, 7-H, 2,6-H), 7.70 (d, J=8.1 Hz, 2H, 3,5-H),
7.56 (d, J=8.7 Hz, 1H, 8-H), 3.23 (s, 3H, CH3SO2), 2.80 (t, J=5.8 Hz, 2H, 1-H), 2.55 (t, J=6.0 Hz, 2H, 3-H),
2.10 (quintet, J=5.8 Hz, 2H, 2-H). 13C NMR (DMSO-d6,75 MHz): 23.27 (C-2), 25.17 (C-1), 37.33 (C-3), 44.08
(CH3SO2), 115.18, 115.65, 119.81, 123.02, 125.04, 129.25, 132.03, 136.59, 138.31, 139.13 and 154.76 (Ar-C),
167.71 (CO-N), 194.69 (CO).MS LRMS (ESI): (calc) 437.02 (found) 438.07 (MH)+. Anal. Calcd for
C12H14N2O2S: C, 59.78; H, 4.01; N, 3.49. Found: C, 59.61; H, 3.89; N, 3.68.
Synthesis of 9-(4-chlorobenzyl)-6-(methylsulfonyl)-1,2,3,9-tetrahydro-4H-carbazol-4-one (12b)
Tetrahydrocarbazole (249 mg, 1 mmol) was dissolved in dry DMF (4 mL) and the stirred solution was treated
with NaH in oil dispersion 60% (52 g, 0.61 mmol). The reaction mixture was stirred at room temperature for 1
hour and then treated with 4-chlorobenzyl chloride (115 mg, 1.4 mmol), the stirring was continued at room
temperature overnight and then poured onto ice-cold water to give oily product. Methanol was added to the
resulting oily product dropwise until complete dissolution of the oil in MeOH/H2O and left standing overnight at
room temperature to give off light brown crystals, filtered off and washed with ether (5 mL) affording the
corresponding N-benzyltetrahydrocarbazole.
Yield: 69%, m.p.: 205-207 0C, IR (KBr) ʋmax/cm-1 1647 (CO), 1300 (CH3-SO2). 1H NMR (DMSO-d6,300 MHz)
δ 8.55 (s, 1H, 5-H), 7.81 (d, J=8.6 Hz, 1H, 7-H), 7.75 (d, J=8.6 Hz, 1H, 8-H) 7.40 (dd, J=8.0 Hz, 2H, 2,6-H),
24

ACCEPTED MANUSCRIPT
7.16 (dd, J=8.0 Hz, 2H, 3,5-H), 5.60 (s, 2H, 4-Cl-C6H4-CH2), 3.19 (s, 3H, CH3SO2), 3.02 (t, J=5.5 Hz, 2H, 3-H),
2.16 (m, 2H, 2-H). 13C NMR (DMSO-d6,75 MHz): 21.75 (C-2), 22.77 (C-1), 37.33 (C-3), 44.32 (CH3SO2),
45.99 (Ar-CH2), 111.52, 112.48, 119.85, 121.41, 123.75, 128.55, 128.83, 132.28, 134.64, 135.46, 138.87 and
155.34 (Ar-C), 193.18 (CO). MS LRMS (ESI): (calc) 387.07 (found) 387.99 (MH)+. Anal. Calcd for
C12H14N2O2S: C, 61.93; H, 4.68; N, 3.61. Found: C, 61.77; H, 4.87; N, 3.79.
Synthesis of 9-[(4-chlorophenyl)sulfonyl)-6-(methylsulfonyl)]-1,2,3,9-tetrahydro-4H-carbazol-4-one (12c)
Tetrahydrocarbazole (249 mg, 1 mmol) was dissolved in dry DMF (4 mL) and the stirred solution was treated
with NaH in oil dispersion 60% (52 mg, 1.3 mmol). The reaction mixture was stirred at room temperature for 1
hour and then treated with 4-chlorophenylsulfonyl chloride (274 mg, 1.3 mmol), the stirring was continued at
room temperature overnight and then poured onto ice-cold water to give oily product. Methanol was added to
the resulting oily product dropwise until complete dissolution of the oil in MeOH/H2O and left standing
overnight at room temperature to give off white fibrous crystals, filtered off and washed with ether (5 mL)
affording the corresponding N-phenylsulfonyltetrahydrocarbazole.
0
Yield: 58%, m.p.: 253 C, IR (KBr) ʋmax/cm-1 1659 (CO), 1301 (CH3-SO2). 1H NMR (DMSO-d6,300 MHz) δ
8.58 (s, 1H, 5-H), 8.32 (d, J=8.9 Hz, 1H, 7-H), 8.14 (dd, J=8.3 Hz, 2H, 2,6-H), 7.94 (d, J=8.8 Hz, 1H, 8-H),
7.74 (dd, J=8.3 Hz, 2H, 3,5-H), 3.41 (t, J=6.0 Hz, 2H, 3-H), 3.23 (s, 3H, CH3SO2), 2.56 (t, J=6.0 Hz, 2H, 1-H),
2.22 (quintet, J=6.0 Hz, 2H, 2-H). 13C NMR (DMSO-d6,75 MHz): 22.48 (C-2), 23.89(C-1), 37.06 (C-3), 43.96
(CH3SO2), 114.59, 116.67, 120.18, 124.02, 125.28, 129.03, 130.52, 135.37, 137.08, 137.51, 140.78 and 153.96
(Ar-C), 194.50 (CO). MS LRMS (ESI): (calc) 437.02 (found) 438.05 (MH)+. Anal. Calcd for C12H14N2O2S: C,
52.11; H, 3.68; N, 3.20. Found: C, 52.29; H, 3.76; N, 3.18.
3.2. Biological activity:
Experimental animals were obtained from Theodor Bilharz Research Institute (TBRI), Egypt and approval of
the institutional animal ethical committee for the animals’ studies was obtained from the Office of
Environmental Health and Radiation Safety, ACUC Protocol 1096-5. The animals were maintained according to
accepted standards of animal care.
3.2.1. Human COX-1 and COX-2 enzymatic assay
Human COX-1 and COX-2 activities were determined as described by Wakitani et al[31]. Human COX-1 (0.3
mg protein/assay) or COX-2 (1 mg protein/assay) was suspended in 0.2 ml of 100 mmol trise HCl buffer (pH 8)
containing the cofactors, 2 mmol of hematin and 5 mmol of tryptophan. With each compound under
investigation individually, the reaction mixture was pre-incubated for 5 min at 24oC. 30 mmol of [14C]-
arachidonic acid (100.00 dpm) was added to the mixture and then incubated for 2 min with COX-1 and 45 min
with COX-2 at 24oC. 400 µl solution of Et2O/MeOH/1 M citric acid (30:4:1, v/v/v) was added to stop the
reaction. The mixture was centrifuged at 1700X/g for 5 min at 4oC, then 50 µl of the upper phase was applied to
a plate of thin layer chromatography (TLC). TLC was exposed at 4oC to a solvent system of Et2O/MeOH/AcOH
(90:2:0.1, v/v/v). Radiometric photographic system was used to determine the percent conversion of arachidonic
25

ACCEPTED MANUSCRIPT
acid to PGH2 and its decomposition products, and eventually from which the enzyme activity was calculated.
The concentration of the compound causing 50% inhibition (IC50) was calculated.
3.2.2. Carrageenan-induced rat paw edema
Male albino rats weighing 120-150 g were kept in an animal house under standard conditions regarding light,
temperature and free access to food and water. Groups of six rats each were subjected to induction of paw
edema by subplantar injection of 50 ml of 2% carrageenan solution in saline (0.9%). Indomethacin and the
compounds under investigation were dissolved in DMSO then injected subcutaneously in a dose of 10 µmol/kg
body weight, 1 h earlier than the carrageenan injection. Control group was injected with DMSO only.
Plethysmometer were used immediately after carrageenan injection and 4h later to measure the volume of paw
edema. The increase in paw volume from 0 to 4 h was measured[32]. The %protection against inflammation was
calculated as follows:
Vc –VdX100/Vc where Vc is the increase in paw volume in the absence of test compound (control) and Vd is
the increase in paw volume after injection of the test compound. Data were expressed as the mean ± SEM.
Student’s t-test and P values was used to determine the significance in difference between the control and the
groups injected with the test compound. The difference in results was considered significant when P < 0.001.
Taking indomethacin as reference standard compound, the relative anti-inflammatory activity of the test
compounds was also calculated.
3.2.3. Estimation of plasma prostaglandin E2 (PGE2)
8 Heparinized blood samples were collected from rats and centrifuged to separate the plasma at 12,000 g for 2
min at 40oC, then immediately frozen, and stored at 20ºC until use. The estimation procedure was designed to be
competitive immune assay to determine PGE2 quantitatively in the biological fluids using EIA PGE2 kit
(Aldrich, Steinheim, Germany). A monoclonal antibody to PGE2 is used by kit to bind competitively to the
PGE2 in the sample after a simultaneous incubation at room temperature. The substrate was added after washing
the excess reagents away. After a short incubation time, a yellow color generated and got read on a microplate
reader DYNATech, MR 5000 at 405 nm (Dynatech Industries Inc., McLean, VA, USA) after stopping the
enzyme reaction. The intensity of the bound yellow color is inversely proportional to the concentration of PGE2
in either standard or samples.
3.2.4. Human, Rat, and Dog Microsomal COX Assays
In 50 mM potassium phosphate buffer, pH 7.1, containing 0.1 M NaCl, 2 mM EDTA, and 1 mM
phenylmethylsulfonyl fluoride (homogenization buffer), 1–10 g of tissue obtained from the whole kidney were
suspended. A hand-held tissue homogenizer (Biospec Products, Inc., Bartlesville, OK) at maximum setting was
used to homogenize the samples for 2 min on ice, then the homogenized products were sonicated for 10 s using
a micro-ultrasonic cell disrupter (Kontes, Vineland, NJ). Tissue homogenates were then exposed to
centrifugation at 100,000g for 1h at 4°C. The 100,000g microsomal pellet was re-suspended in homogenization
buffer and was sonicated (2 3 10 s) on ice. Microsomal suspensions of human, rat, and dog kidney got diluted to
26

ACCEPTED MANUSCRIPT
protein concentrations of approximately 6, 10, and 12 mg/ml, respectively. Aliquots of microsomal preparations
were stored at 280°C and thawed on ice immediately before assays.
The microsomal preparation from rat, dog, or human kidneys was pre-incubated with the reference standard
drug, Celecoxib at room temperature for 5 or 15 min. The pre-incubation buffer was composed of protein
concentration of 0.12 mg/ml, 0.1 M Trise HCl, pH 7.4, 10 mM EDTA, 0.5mM phenol, 1 mM reduced
glutathione, and 1 mM hematin. The final concentration adjusted at 2 mM, then arachidonic acid and the
samples were incubated at room temperature for 40 min. The reaction was stopped by the addition of 25 ml of 1
N HCl with mixing after the incubation period. Radioimmunoassay was used to analyze the amount of PGE2
after neutralization of the samples with 25 ml of 1 N NaOH.
Assays were repeated two or three times. Ethanol was used as a vehicle for the control reaction mixtures instead
of arachidonic acid. In the control reaction mixture, the levels of PGE2 in samples from human, dog, and rat
kidney microsomes were approximately 1.5 ng/mg protein, 0.1 ng/mg protein, and 6.7 ng/mg protein,
respectively. In the presence of arachidonic acid, levels of PGE2 in these preparations increased to
approximately 4.2 ng/mg protein, 1.2 ng/mg protein, and 22 ng/mg protein, respectively. COX activity in the
absence of test compounds is defined as the difference between PGE2 levels in samples incubated in the
presence of arachidonic acid or ethanol vehicle.
3.2.5. Cotton pellet-induced granuloma bioassay
120-140 g of adult male Spraguee Dawley rats were acclimated one week earlier before use and allowed to
unlimited access to standard rat chow and water. Before starting the experiment, the animals were randomly
grouped into six animals for each. Cotton pellet (35 ± 1 mg) cut from dental rolls were wet with 0.2 ml
(containing 0.01 mmol) of a solution of the compound under the test in chloroform and the solvent was allowed
to evaporate. Each cotton pellet was subsequently injected with 0.2 ml of an aqueous solution of antibiotics (1
mg penicillin G and 1.3 mg dihydrostreptomycin/ml). Two pellets, one in each axilla of the rat were implanted
under mild general anesthesia, subcutaneously. One group of animals received the standard reference
indomethacin and the antibiotics at the same concentration. Control rats were similarly implanted with the
pellets containing only the antibiotics. The animals were sacrificed after 7 days and the two cotton pellets, with
adhering granulomas, were removed, dried for 48 h at 60 ◦C and weighed. The difference between the initial and
final weight was taken as a measure of granuloma ± SEM for each group. The percentage reduction in dry
weight of granuloma from control value was also calculated. The dose-response curves were set using the doses
4, 7, 10 and 15 µmol of each compound. Dose-response curve was used to determine the ED50 values.
3.2.6. Ulcerogenic effects
Indomethacin was used as reference standard to test the ulcerogenic potential[33] of all target compounds. 100-
120 g of Male albino rats were fasted for 12 h before administration of the compounds. Water was given ad
libitum. The animals were randomly grouped into six rats for each. 1% gum acacia were given orally to the
control group. Indomethacin or test compounds were given orally to the test groups in two equal doses at 0 and
12 h for three successive days at a dose of 30 µmol/kg body weight per day. After six hours from the
administration of the last dose, animals were sacrificed by diethyl ether and the stomach was removed. An
27

ACCEPTED MANUSCRIPT
opening was made at the greater curvature, and the stomach was washed with cold saline to be inspected by a
3_magnifying lens for any signs of hyperemia, hemorrhage, definite hemorrhagic erosion or ulcer.
The ulcer index was calculated using an arbitrary scale to measure the severity of stomach lesions[33]. The
%ulceration for each group was calculated as follows: % Ulceration =Number of animals bearing ulcer in a
groupX100/Total number of animals in the same group.
3.3. Molecular docking studies:
All compounds were built and saved as Mol2. The crystal structure of COX-2 enzyme in complex with
indomethacin was downloaded from Protein Data Bank (PDB code entry; 4COX). The protein was loaded into
Leadit 2.1.2[34] and the receptor components were selected. The binding site was defined by selecting
indomethacin as a reference ligand to which all coordinates were computed. Amino acids within radius 6.5 Ao
were selected at the binding site. All chemical ambiguities associated with the residues were left as default.
Ligand binding was driven by enthalpy (classic triangle matching). For scoring, all default settings were
restored. Intra-ligand clashes were computed by using clash factor of 0.6. The maximum number of solutions
per iteration was 200. The maximum value of solution per fragmentation was 200. The base placement method
was used as a docking strategy.
3.4. Molecular Dynamics
The best conformation from each docking process of each compound was kept inside the active site. The quality
of the temperature-related factors, protein geometries, and electron density was tested. All hydrogens were
added and energy minimization was computed. The solvent molecules that were in the system were deleted
before solvation; salt atoms were added to ensure complete neutralization of the biomolecular system. The
solvent atoms were added to surround the biomolecular system (protein-ligand complex) in a spherical shape.
Amber 10:EHT was selected as a force field in the potential setup step. All Van der Waals forces, electrostatics,
and restraints were enabled. The heat was adjusted in order to increase the temperature of the system from 0-300
K which was followed by equilibration and production for 300 ps; cooling was then initiated until to 0 K was
reached.
3.5. MM/GBVI and TPSA calculation
These were calculated after molecular docking with MOE[35]. Placement method was used as Alpha Triangle,
timeout (seconds) = 300, minimum iterations = 8000 and max iterations = 500000. Rescoring was done by
affinity dG, hydrogen bonds = -0.65, hydrophobic contacts = -0.01235, ionic contact =1, Metal ligation =1, and
hydrophobic –polar = 0.02497. The refinement was selected as Grid Mn, in which elec. Cutoff =5.5, VDW
cutoff = 4, and VDW potential was enabled. After that the best pose was kept inside the active site. No ligands
other that the best conformation were kept. TPSA, Affinity pki, were then computed.
[1] P.J. Loll, Structure of prostaglandin H2 synthase-1 (COX-1) and its NSAID binding sites, in: N.
Bazan, J. Botting, J. Vane (Eds.) New Targets in Inflammation: Inhibitors of COX-2 or Adhesion
Molecules Proceedings of a conference held on April 15–16, 1996, in New Orleans, USA, supported
by an educational grant from Boehringer Ingelheim, Springer Netherlands, Dordrecht, 1996, pp. 13-
21.
28

ACCEPTED MANUSCRIPT
[2] G. Rimon, R.S. Sidhu, D.A. Lauver, J.Y. Lee, N.P. Sharma, C. Yuan, R.A. Frieler, R.C. Trievel,
B.R. Lucchesi, W.L. Smith, Coxibs interfere with the action of aspirin by binding tightly to one
monomer of cyclooxygenase-1, Proceedings of the National Academy of Sciences of the United
States of America, 107 (2010) 28-33.
[3] J.R. Vane, J.A. Mitchell, I. Appleton, A. Tomlinson, D. Bishop-Bailey, J. Croxtall, D.A.
Willoughby, Inducible isoforms of cyclooxygenase and nitric-oxide synthase in inflammation,
Proceedings of the National Academy of Sciences of the United States of America, 91 (1994) 2046-
2050.
[4] N. Zidar, K. Odar, D. Glavac, M. Jerse, T. Zupanc, D. Stajer, Cyclooxygenase in normal human
tissues is COX-1 really a constitutive isoform, and COX-2 an inducible isoform?, Journal of Cellular
and Molecular Medicine, 13 (2009) 3753-3763.
[5] N.S. Kirkby, A.K. Zaiss, W.R. Wright, J. Jiao, M.V. Chan, T.D. Warner, H. R. Herschman, J.A.
Mitchell, Differential COX-2 induction by viral and bacterial PAMPs: Consequences for cytokine and
interferon responses and implications for anti-viral COX-2 directed therapies, Biochemical and
Biophysical Research Communications, 438 (2013) 249-256.
[6] W.L. Smith, Y. Urade, P.-J. Jakobsson, Enzymes of the cyclooxygenase pathways of prostanoid
biosynthesis, Chemical Reviews, 111 (2011) 5821-5865.
[7] A. Zarghi, S. Arfaei, Selective COX-2 inhibitors: A review of their structure-activity relationships,
Iranian Journal of Pharmaceutical Research, 10 (2011) 655-683.
[8] J.P. Peesa, P.R. Yalavarthi, A. Rasheed, V.B.R. Mandava, A perspective review on role of novel
NSAID prodrugs in the management of acute inflammation, Journal of Acute Disease, 5 (2016) 364-
381.
[9] C.A. Rouzer, L.J. Marnett, Cyclooxygenases: structural and functional insights, Journal of Lipid
Research, 50 (2009) S29-S34.
[10] R.G. Kurumbail, A.M. Stevens, J.K. Gierse, J.J. McDonald, R.A. Stegeman, J.Y. Pak, D.
Gildehaus, J.M. Miyashiro, T.D. Penning, K. Seibert, P.C. Isakson, W.C. Stallings, Structural basis
for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents, Nature, 384 (1996) 644-
648.
[11] D. Picot, P.J. Loll, R.M. Garavito, The X-ray crystal structure of the membrane protein
prostaglandin H2 synthase-1, Nature, 367 (1994) 243-249.
[12] A.L. Blobaum, L.J. Marnett, Structural and functional basis of cyclooxygenase inhibition,
Journal of Medicinal Chemistry, 50 (2007) 1425-1441.
[13] R. Anana, P.N.P. Rao, Q.-H. Chen, E.E. Knaus, Synthesis and biological evaluation of linear
phenylethynylbenzenesulfonamide regioisomers as cyclooxygenase-1/-2 (COX-1/-2) inhibitors,
Bioorganic & Medicinal Chemistry, 14 (2006) 5259-5265.
[14] J.K. Gierse, J.J. McDonald, S.D. Hauser, S.H. Rangwala, C.M. Koboldt, K. Seibert, A single
amino acid difference between cyclooxygenase-1 (COX-1) and -2 (COX-2) reverses the selectivity of
COX-2 specific inhibitors, The Journal of Biological Chemistry, 271 (1996) 15810-15814.
[15] K.M. Knights, A.A. Mangoni, J.O. Miners, Defining the COX inhibitor selectivity of NSAIDs:
Implications for understanding toxicity, Expert Review of Clinical Pharmacology, 3 (2010) 769-776.
[16] A.K. Dwivedi, V. Gurjar, S. Kumar, N. Singh, Molecular basis for nonspecificity of nonsteroidal
anti-inflammatory drugs (NSAIDs), Drug Discovery Today, 20 (2015) 863-873.
[17] S. Moncada, E.A. Higgs, J.R. Vane, Human arterial and venous tissues generate prostacyclin
(prostaglandin X), a protein inhibitor of platlet aggregation, The Lancet, 309 (1977) 18-21.
[18] J.N. Topper, J. Cai, D. Falb, M.A. Gimbrone, Jr., Identification of vascular endothelial genes
differentially responsive to fluid mechanical stimuli: Cyclooxygenase-2, manganese superoxide
dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar
shear stress, Proceedings of the National Academy of Sciences of the United States of America, 93
(1996) 10417-10422.
[19] M.S. Estevão, L.C.R. Carvalho, M. Freitas, A. Gomes, A. Viegas, J. Manso, S. Erhardt, E.
Fernandes, E.J. Cabrita, M.M.B. Marques, Indole based cyclooxygenase inhibitors: Synthesis,
biological evaluation, docking and NMR screening, European Journal of Medicinal Chemistry, 54
(2012) 823-833.
[20] W.C. Black, C. Bayly, M. Belley, C.C. Chan, S. Charleson, D. Denis, J.Y. Gauthier, R. Gordon,
D. Guay, S. Kargman, C.K. Lau, Y. Leblanc, J. Mancini, M. Ouellet, D. Percival, P. Roy, K. Skorey,
29