Ligand based design and synthesis of pyrazole based derivatives as selective COX-2 inhibitors

Article abstract of DOI:10.1016/j.bioorg.2019.02.031

The design and synthesis of novel pyrazole based derivatives has been carried out using the ligand based approach like pharmacophore and QSAR modelling of reported pyrazoles from the available literature to investigate the chemical features that are essential for the design of selective and potent COX-2 inhibitors. Both pharmacophore and QSAR models with good statistical parameters were selected for the design of the lead molecule. Also by exploiting the chemical structures of selective and marketed COX-2 inhibitors, celecoxib and SC-558 were used in designing the molecules which are used in the treatment of inflammation and related disorders. The therapeutic action of the Non-Steroidal Anti-inflammatory Agents (NSAIDs) is based primarily on the COX-2 inhibition. With this background we have synthesized some azomethine derivatives of 3-methyl-1-substituted-4-phenyl-6-[{(1E)-phenylmethylene}amino]-1,4-dihydro pyrano[2,3-c]pyrazole-5-carbonitrile 6(a-o) and were characterized by ¹HNMR, ¹³CNMR and Mass spectral techniques. All the synthesized pyrazole derivatives were tested for in vitro membrane stability property in both COX-1 & COX-2 inhibition studies and in vivo anti-inflammatory activity by carrageenan induced rat paw edema model. Among them, compound 6k showed very good activity by in vivo anti-inflammatory activity with 0.8575 mmol/kg as ED₅₀. Similarly compounds 6m, 6o, 6i and 6h exhibited comparable anti-inflammatory activity to standard drugs. Also the active compounds were further screened for ulcerogenic activity and were found be safer with less ulcer index compared to the marketed drugs like aspirin, ibuprofen and celecoxib.

Full text of DOI:10.1016/j.bioorg.2019.02.031

Accepted Manuscript
Ligand based Design and Synthesis of Pyrazole Based Derivatives as Selective
COX-2 Inhibitors
Manikanta Murahari, Vivek Mahajan, Sreenivasulu Neeladri, Maushmi S
Kumar, Y.C. Mayur
PII:
S0045-2068(19)30016-1
DOI:
https://doi.org/10.1016/j.bioorg.2019.02.031
YBIOO 2806
Reference:
Bioorganic Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
4 January 2019
6 February 2019
12 February 2019
Please cite this article as: M. Murahari, V. Mahajan, S. Neeladri, M.S. Kumar, Y.C. Mayur, Ligand based Design
and Synthesis of Pyrazole Based Derivatives as Selective COX-2 Inhibitors, Bioorganic Chemistry (2019), doi:
https://doi.org/10.1016/j.bioorg.2019.02.031
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.

Ligand based Design and Synthesis of Pyrazole Based Derivatives as Selective COX-2
Inhibitors
Manikanta Muraharia, Vivek Mahajanb, Sreenivasulu Neeladrib, Maushmi S Kumarc
and Mayur YCc*
aPharmacological Modelling & Simulation Centre, M.S. Ramaiah University of Applied
Sciences, Bangalore- 560 054, India.
bMedicinal Chemistry Research Division, VL College of Pharmacy, Raichur 584103, India
cDepartment of Pharmaceutical Chemistry, SPP School of Pharmacy & Technology
Management, SVKM’s NMIMS, Mumbai-400 056. India.
*Corresponding Author
Dr. Mayur Yergeri C
Professor
Department of Pharmaceutical Chemistry,
SPPSPTM
SVKM’s NMIMS University
Vile Parle (West), Mumbai- 400 056. India.
Tel: (+91)- 9930713869,
Fax: 0091-22-26132905
E-mail: mayuryc@rediff.com, mayuryc@hotmail.com
1

Abstract
The design and synthesis of novel pyrazole based derivatives has been carried out using the
ligand based approach like pharmacophore and QSAR modelling of reported pyrazoles from the
available literature to investigate the chemical features that are essential for the design of
selective and potent COX-2 inhibitors. Both pharmacophore and QSAR models with good
statistical parameters were selected for the design of the lead molecule. Also by exploiting the
chemical structures of selective and marketed COX-2 inhibitors, celecoxib and SC-558 were
used in designing the molecules which are used in the treatment of inflammation and related
disorders. The therapeutic action of the Non-Steroidal Anti-inflammatory Agents (NSAIDs) is
based primarily on the COX-2 inhibition. With this background we have synthesized some
azomethine derivatives of 3-methyl-1-substituted-4-phenyl-6-[{(1E)-phenylmethylene}amino]-
1,4-dihydro pyrano[2,3-c]pyrazole-5-carbonitrile 6(a-o) and were characterized by
1HNMR,13CNMR and Mass spectral techniques. All the synthesized pyrazole derivatives were
tested for in vitro membrane stability property in both COX-1 & COX-2 inhibition studies and in
vivo anti-inflammatory activity by carrageenan induced rat paw edema model. Among them,
compound 6k showed very good activity by in vivo anti-inflammatory activity with 0.8575
mmol/kg as ED50. Similarly compounds 6m, 6o, 6i and 6h exhibited comparable anti-
inflammatory activity to standard drugs. Also the active compounds were further screened for
ulcerogenic activity and were found be safer with less ulcer index compared to the marketed
drugs like aspirin, ibuprofen and celecoxib.
Key words: Pyrazole, NSAIDs, COX-2, Anti-inflammatory, ligand based design,
Pharmacophore
2

1. Introduction
Most of the currently used non-steroidal anti-inflammatory drugs (NSAIDs) show severe cross
effects and majority of these drugs showing gastrointestinal and renal toxicity [1,2]. There has
been a continuous interest among researchers in discovering newer NSAIDs which are selective
to cyclooxygenase-2 (COX-2) enzyme. Unrelated and non-specific side effects exerted by the
classical NSAIDs is due to inhibition of physiologically important cyclooxygenase-1 (COX-1)
enzyme [3]. Isozyme COX-2 is more responsible for the release of prostaglandins as
inflammatory mediators [4,5]. The heterocyclic class of compounds have been studied widely as
a new class of NSAIDs with lesser gastric side effects, a new series of compounds were
synthesized with a focus to develop safer drugs [6]. Pyrazole derivatives are well established
heterocyclic systems with established biological significance, safe and effective drugs used
therapeutically. Feasible synthetic routes and conjugation with diverse aromatic and heterocyclic
rings are subjected in many research investigations of varied pharmacological profiles such as
anti-inflammatory, antipyretic, antimicrobial, antiviral, antitumor, anticonvulsant, antihistaminic,
and antidepressant activities [6–8]. The pyrazole scaffold is a well-known and widely explored
for anti-inflammatory activity by many research groups [9–14]. Most commonly prescribed
NSAIDs which pyrazole based derivatives like celecoxib, deracoxib, etc., are selective and
potent COX-2 inhibitors [15,16].
In view of above mentioned findings, authors have utilized the strength of pharmacophore
modelling to investigate the structural features of pyrazole derivatives required to selectively
inhibit COX-2 enzyme. In addition to that, we have used pyrazole derivatives celecoxib, SC-558
and antipyrine (2,3-dimethyl-1-pheny-3-pyrazolin-5-one) one of the first identified pyrazolone
derivative prescribed and suggested in the management of pain and inflammation. The drug
Antipyrine was withdrawn because of its toxicity, but due to its potential activity their
derivatives have enticed the attention of researchers across the globe [17–19]. Further, we
employed the most advanced Auto Quantitative Structure Activity Relationship (AutoQSAR)
tool to build a model of experimentally reported pyrazoles which are selective COX-2 inhibitors
to quantify the therapeutic potential of synthesized pyrazoles. In continuation of our previous
study and to explore the potential of new candidates which are selective and less toxic anti-
inflammatory agents, we are reporting the ligand based design and synthesis, evaluation by in
3

vitro and in vivo methods for anti-inflammatory activity of novel pyrazole derivatives [20]. Also
the ulcerogenic activity has been carried out to check the gastric related side effects of these
drugs.
2. Results and Discussion
2.1.Generation of ligand based pharmacophore model
Pharmacophore model of pyrazole derivatives have been generated with 7 active (pIC50 greater
than 6.30) and 11 inactive compounds (pIC50 less than 4.35) from selective COX-2 inhibition
assay results. Other compounds pIC50 which were in the range of 4.35 to 6.30 were considered as
moderately active. All the compounds were screened for selective COX-1 and COX-2 inhibition
assay and identified that substituted pyrazoles were selective to COX-2 enzyme with a pIC50
values in the range of 4.3010 to 6.5376. The trail run could generate two five point and nine four
point hypothesis respectively with all active ligands matching the pharmacophoric features. The
parameters listed were generated by the Phase module and AHHR_1 was nominated as best
hypothesis for screening and identification of selective COX-2 inhibitors (Supplementary Data).
The model AHHR_1 was characterized by survival score of 4.782, inactive score of 2.481 and
selective score of 1.380. The hypothesis represents that one hydrogen bond acceptor, two
hydrophobic groups and one aromatic ring at a specific spatial arrangement of angstroms (Å)
highlighted in Figure 1 are crucial for the selective COX-2 inhibition. Hypothesis results for
pyrazole derivatives are shown (Supplementary Data).
2.2.Validation of pharmacophore model
To investigate the predictability of AHHR_1, enrichment view was reviewed for all the
hypotheses. The selected hypothesis AHHR_1 was found with highest Phase Hypo Score of
1.29, AOC and ROC values as 1 with all seven active compounds matched 4 features of
hypothesis (Figure 2). Interestingly plot generated by Phase for AHHR_1 hypothesis suggests
that model can filter 100% of active molecules from the database screening with true positive
rate (Figure 2). To further validate the efficiency and selectivity of COX-2 inhibitors hypothesis,
a model was screened with two data sets i.e., one external test set of selective COX-2 inhibitors
and standard drugs. One data set included 12 compounds which were evaluated for COX-1 & 2
4

enzyme inhibition assay and in vivo anti-inflammatory activity by the same biological method.
Another data set of standard drugs in market and clinical investigation as non-selective and
selective COX-2 inhibitors. Both the datasets were screened through hypothesis and results are
shown (Supplementary Data). Hypothesis screening of 12 compounds were found in correlation
with experimental results. Experimental results indicated that all 12 compounds were selective to
COX-2 inhibition and pIC50 was observed in the range of 4.337 to 6.347. All the 12 compounds
matched four sites of hypothesis and interestingly most active and inactive compounds have
highest and least Phase score (Supplementary Information). The nominated hypothesis AHHR_1
has good results accomplished by selectively screening the COX-2 inhibitor drugs and leaving
the non-selective COX inhibitor molecules. The model suggests that naproxen and indomethacin
were selective to COX-2 enzyme inhibition. All the selective COX-2 inhibitors have displayed
good scores highlighting Celecoxib and SC-558 with highest fitness score of 2.558 and DuP-697
as least with 1.265 (Table 1). Interestingly both the drugs are of congeneric series sharing
pyrazole as common nucleus (Figure 3). With respect to screening of drugs i.e., 8 of non-
selective COX as inactive and 8 of selective COX-2 as active inhibitors through pharmacophore
model, the sensitivity is 1.25 and specificity is 0.75.
Parameter
Value
4.782
Survival Score
Inactive Score
Site Score
2.481
0.975
0.955
0.627
1.380
Vector Score
Volume Score
Selectivity Score
Figure 1. Pharmacophore model of COX-2 Inhibitors AHHR_1-Acceptor (A), Hydrophobic (H),
Aromatic ring (R) with statistical data of hypothesis
5

6

Figure 2. Percentage screen and ROC plot AHHR_1
7

Figure 3. a) Seven Active pyrazoles with pIC50 greater than 6.30 aligned with pharmacophore
AHHR_1 and b) Standard drugs SC-558 and Celecoxib aligned with pharmacophore AHHR_1
8

Table 1: Validation of Pharmacophore Model with Standard Drugs
S. No.
Drug Name
Matched
ligand sites
Fitness Vector Volume
score score score
1
Aspirin
No sorted hits for match of at least 3 out
of 4 features
2
Ibuprofen
3
Flurbiprofen
Naproxen
4
3
4
1.398
1.831
0.654
0.892
0.397
0.622
5
Indomethacin
Diclofenac
Mefenamic acid
Piroxicam
DuP-697
6
No sorted hits for match of at least 3 out
of 4 features
7
8
9
4
4
4
3
3
3
4
3
1.265
2.558
2.538
2.007
1.960
1.637
2.169
1.684
0.907
0.998
0.998
0.997
0.997
0.983
0.989
0.906
0.325
0.622
0.621
0.524
0.477
0.439
0.442
0.404
10
11
12
13
14
15
16
SC-558
Celecoxib
Valdecoxib
Parecoxib
Rofecoxib
Etoricoxib
Lumiracoxib
2.2. Auto QSAR Statistical results
A congeneric series of 25 pyrazole derivatives were collected from the literature and randomly
categorized as training and test sets in the ratio of 80:20. Trial run of Quantitative Structure
Activity Relationship (QSAR) generated different models and were arranged in ranking order
based on statistical values calculated by Auto QSAR tool of Schrodinger Drug Design Suite. The
best model that identified was based on the values of R2, Q2, Standard deviation and Root Mean
Square Error (RMSE). Study could identify model of KPLS molprint 2D principle with R2 and
Q2 values of 0.7711 and 0.7233 respectively. Also standard deviation and RMSE values were
observed as 0.4578 and 0.4991 respectively. The effectiveness and accuracy of the model can be
explained with small deviation from R2 and Q2 values. Predicted pIC50 values of pyrazoles from
literature with deviation or error from experimental pIC50 is shown in Table 2. Also it has been
9

graphically represented to understand that the deviation between experimental and predicted
pIC50 values in Figure 4. The following kpls_ molprint2D_45 was selected as best model for
prediction of biological activity of unknown compounds.
Table 2: QSAR Results of Pyrazoles Collected from Literature
Best QSAR Model = kpls_molprint2D_45
Ranking score = 0.668606
Standard deviation 0.4578
Training set- 20 compounds (80 %)
Test set- 5 compounds (20 %)
R2
0.7711
0.4991
0.7233
RMSE
Q2
S. No.
Compound Pharma
pIC50
pIC50
Error
code
set
(experimental) (predicted)
1.3a
Training
4.3468
4.3979
5.5229
4.3468
4.3468
4.3468
6.2596
6.3468
4.3468
5.0000
6.3468
4.3468
4.3468
4.3526
6.3188
5.6778
6.5376
6.3665
4.0381
4.2699
4.4973
4.4396
4.1304
4.2699
5.5614
5.9083
4.9543
5.4238
6.0662
4.3604
4.3604
4.8804
6.5532
5.6863
6.7142
5.9012
-0.3086
-0.1281
-1.0256
0.0928
-0.2164
-0.0769
-0.6982
-0.4385
0.6076
0.4238
-0.2806
0.0136
0.0136
0.5277
0.2344
0.0085
0.1766
-0.4653
1
2
1.3b
1.3c
1.3d
1.3e
1.3f
1.5
Test
Training
Training
Training
Test
3
4
5
6
Training
Training
Training
Training
Training
Training
Training
Training
Training
Test
7
1.6a
1.6b
1.6c
1.6d
1.6e
1.6f
2.3a
2.3b
2.3c
2.3d
2.3e
8
9
10
11
12
13
14
15
16
17
18
Training
Test
10

2.3f
2.5
Training
Training
Training
Training
Training
Test
4.3188
4.3010
5.4949
4.3098
6.3872
6.5850
4.3010
4.5009
4.4985
5.8253
4.8262
5.7719
5.5818
4.5677
0.1822
0.1975
0.3304
0.5164
-0.6153
-1.0033
0.2667
19
20
21
22
23
24
25
2.8a
2.8b
2.8c
2.8d
2.9
Training
7.3
6.8
6.3
5.8
5.3
4.8
4.3
3.8
y = 0.7411x + 1.2717
R² = 0.7738
4
4.5
5
5.5
6
6.5
7
Figure 4. Plot of COX-2 Inhibition pIC50 (experimental) vs pIC50 (predicted) for pyrazole
derivatives from literature
2.3. Design of substituted pyrazole derivatives as selective COX-2 Inhibitors
Pharmacophore modelling of substituted pyrazoles as selective COX-2 inhibitors identified from
the literature demonstrated that one hydrogen bond acceptor, two hydrophobic groups and one
aromatic ring are essential. Further validation of model with drugs in market and clinical
investigation as selective COX-1 and COX-2 inhibitors, pharmacophore model could selectively
identify COX-2 inhibitors and concluded SC-558 and celecoxib with highest fitness score.
Considering the structural features of both the drugs and essential pharmacophoric features for
selective COX-2 inhibition, we have made an effort to synthesize a series of fifteen pyrazole
derivatives (Figure 5). The common trifluoromethyl group of both the drugs was replaced with a
nitrile to investigate the significance on anti-inflammatory activity. Many new chemical entities
11

(NCEs) containing nitrile group are currently under clinical investigation emphasizing the safety
and potency of substitution. In addition to structural advantages, literature suggests that nitrile
groups containing compounds will not be metabolized readily, have improved binding affinity
towards the target and also release of cyanogen was not observed [21].
Figure-5. Design approach of pyrazoles as selective COX-2 Inhibitors
2.4. Chemistry
The starting material, unsubstituted\substituted 3-methyl-1-pyrazole-5-ones (2a,b) were prepared
from ethylacetoacetate (1) and hydrazine hydrate\phenyl hydrazine in one step in good yields.
1
The structure of the compounds (2a,b) were confirmed by spectral data (IR, HNMR, MS).
12

1HNMR spectrum of 2a showed δ= 7.1-7.9 ppm (m, 5H, Ar-H), compounds (2a,b) showed δ=
2.0-2.1 ppm (s, 3H, -CH3). The IR spectrum of compound showed characteristic absorption band
at ν= 1790 cm-1, 1807 cm-1 for CO, 3030 cm-1 correspond to aromatic C-H stretching. Mass
spectrum showed the prominent ion peak at m/z 99 and 175 (M+1) respectively. The 3-methyl-1-
unsubstituted\substituted pyrazole-5-ones (2a,b) were converted into 6-amino-3-methyl-1-
substituted-4-phenyl-1,4-dihydropyrano[2,3-c]pyrazol-5-carbonitriles (4a-o) by reacting with
different aromatic aldehyde (3), melononitrile and by using alcohol medium in the presence of
1
piperidine by stirring for 2h at room temperature. The HNMR spectrum of the compounds
showed δ = 5.6-5.7 ppm (s, 1H, H of pyran), 2.3-2.5 ppm (s, 3H, -CH3). In the 3rd step the 6-
amino-3-methyl-1-substituted-4-phenyl-1,4-dihydropyrano[2,3-c]pyrazole-5-carbonitrile (4a-o)
were converted to respective azomethines by refluxing for 3 h with aromatic aldehydes (5) in
alcoholic medium by using glacial acetic acid as catalyst, which yields 3-methyl-1-substituted-4-
phenyl-6-[{(1E)-phenylmethylene}amino]1,4-dihyropyrano[2,3-c] pyrazole-5-carbonitrile (6a-o)
as per Scheme 1. All the intermediates were recrystallized and final pyrazoles derivatives were
purified by column chromatography, purity was checked by HPLC and chemical structures were
1
13
interpreted and confirmed by HNMR, CNMR, Mass and IR spectroscopy techniques (Table
3).
13

Scheme 1: Reagents & Conditions: (i) stirring at room temperature (ii) melononitrile,
piperidine, 2 hr (iii) EtOH, AcOH, 3 h, Reflux
14

Table 3: Physical Characterization data of Pyrazole Derivatives
Compound
Code
R
R1
R2
Molecular
formula
Molecular
weight
Melting Yield %
point (0C)
H
4-OH
OH C21H16N4O3
372.38
175-177
160-162
210-212
180-182
185-187
160-162
280-283
215-218
160-162
225-227
210-212
145-147
120-122
178-180
206-208
30
42
35
30
45
48
42
38
32
40
42
44
40
45
46
6a
H
3,4-diOCH3
2-NO2
2-Cl
H
H
H
C23H20N4O3
C21H15N5O3
C21H15ClN4O
400.43
385.38
374.82
374.37
416.47
432.47
462.50
476.53
522.55
461.57
461.57
450.92
466.92
450.46
6b
6c
6d
6e
6f
H
H
H
4-F
OH C21H15FN4O2
C6H5
H
H
H
C27H20N4O
C27H20N4O2
C6H5 4-OH
6g
6h
6i
C6H5 4-OCH3
C6H5 3,4-diOCH3
OH C28H22N4O3
C29H24N4O3
H
C6H5 3,4,5-triOCH3 OH C30H26N4O5
6j
C6H5 2-NO2
C6H5 3-NO2
C6H5 2-Cl
C6H5 3-Cl
C6H5 4-F
H
H
H
C27H19N5O3
C27H19N5O3
C27H19ClN4O
6k
6l
6m
6n
6o
OH C27H19ClN4O2
OH C27H19FN4O2
15

2.5.Biological activity
The main classical mechanism of anti-inflammatory drugs is mediated by inhibition of
prostaglandins (PG) synthesis by cyclooxygenase enzyme [22]. The action of the enzyme in the
conversion of prostaglandin G2 (PGG2) to PGH2 along with peroxidation is associated with
formation of long channels in membranes. The channel opening occurs due to discharge of
chemical mediators and so arachidonic acid is released form membrane and transformed to
prostaglandin. The extracellular activity of these enzymes is said to be related to acute and
chronic inflammation. The action of NSAIDs is either by hindering these lysosomal enzymes
(cyclooxygenase) or by stabilizing the lysosomal membrane [23,24]. A series of pyrazole
derivatives were prepared and evaluated for in vitro Cyclooxygenase inhibition assay, HRBC
membrane stabilizing method and in vivo anti-inflammatory activity by mercury displacement
method using carrageenan rat paw edema model.
2.5.1. In vitro membrane stabilizing property [25–28]
The in vitro anti-inflammatory activity results (Table 4) indicated that all the compounds (6a-o)
at the concentration of 1 mg/ml exhibited moderate to significant anti-inflammatory activity in
HRBC membrane stabilization method. The standard drug Celecoxib and Ibuprofen has
demonstrated 72 and 75 % of membrane stability. The compounds 6e and 6o with 4-fluoro
substitution exhibited 71.12 % and 6c with 2-nitro substitution showed 70.15 % membrane
stabilization comparable to the standard drug. Similarly, compounds substituted with halogen
(i.e., chloro at 2nd position) 6d and 6m have displayed good stabilization of 69.97 and 67.60 %
respectively. Surprisingly compounds 6l with nitro group at 3rd position has curiously reduced
the activity to 55.15 % and was found to be least among all the compounds. Also compound 6n
with 3-chloro has diminished the activity to 56.12 %. Substitution of compounds with electron
donating groups like hydroxyl and mono, -di, -tri methoxy groups have reduced the membrane
protection. The unsubstituted compound 6f exhibited stabilization of 59.55 % indicating the
significance of substitution of phenyl ring for the activity.
16

2.5.2. In vivo anti-inflammatory study
The in vivo anti-inflammatory activity results indicated that all synthesized pyrazole derivatives
(6a-o) when tested at a dose of 200 mg/kg exhibited considerable in vivo activity in acute
inflammatory models induced by carrageenan in rats. The edema volume of pyrazoles (6a-o) and
standard drug ibuprofen and celecoxib are shown in Table 4. In addition to that, ED50 values of
all the compounds and standard drugs in mg/kg and mmol/kg were determined using Graphpad
Prism from percentage reduction in volume with respect to time (Table 6) [29,30]. Standard
drugs Ibuprofen and Celecoxib have demonstrated ED50 values of 4.606 and 3.060 mg/kg
respectively. Selective COX-2 inhibitor celecoxib has exhibited potential anti-inflammatory
activity compared to non-selective inhibitor ibuprofen. Compounds have exhibited percentage
reduction of edema volume in the range of 21.42 to 33.57% after 4th hr and ED50 in the range of
0.8575 to 1.404 mmol/kg. All the pyrazoles have displayed moderate activity and the compounds
6k and 6m showed the highest in vivo anti-inflammatory activity at concentration of 0.8575 and
0.9254 mmol/kg as ED50. Much differences were not observed in the paw edema volume of
compounds (6a-o). Similarly, 6e and 6j were found to be less active at concentration of 1.2466
and 1.404 mmol/kg among all the compounds.
Observing the Structure Activity Relationship (SAR) of synthesized pyrazoles (6a-o),
compounds substituted with electron with-drawing groups like nitro at 2nd position (6k),
halogens like chlorine at 2nd (6m) and fluoro at 4th position (6o) have potentiated the activity.
Interestingly, five compounds i.e., 6i, 6h 6k, 6m and 6o, have exhibited ED50 less than 1
mmol/kg revealed that unsubstituted phenyl ring is directly linked to nitrogen of pyrazole might
be essential for complementary stacking interactions with amino acids at the active pocket of
COX-2 is benefiting the actual inhibition. In the series of pyrazoles from (6a-e), unsubstituted
pyrazoles especially lacking phenyl ring have greatly reduced the anti-inflammatory activity. For
the compounds substituted with electron-donating groups like methoxy and hydroxyl have
weakened anti-inflammatory activity. The trimethoxy substitution (6j) diminished and mono
substitution (6h) enhanced the in vivo activity. Looking at the effect of dimethoxy substitution,
phenyl substituted pyrazole derivative (6i) has enriched and unsubstituted pyrazole derivative
(6b) has greatly reduced anti-inflammatory activity. Similarly, unsubstituted compound at R1
17

position of phenyl ring for compound 6f displayed least activity stressing the significance of
substitution.
2.5.3. Acute Ulcerogenic Studies
Based on the in vivo anti-inflammatory, compounds 6h, 6i, 6k, 6m and 6o were selected for the
ulcerogenic screening. Aspirin and Ibuprofen were taken as standard drugs for the study. Results
exhibited celecoxib with 1.204 ± 0.06 aspirin with 2.358 ± 0.28 and ibuprofen 2.1 ± 0.16 as
Ulcer Index (Table 5). Comparatively all the five active compounds exhibited lesser ulcerogenic
activity in the range of 0.603 ± 0.15 to 1.991 ± 0.34 and found to be safe. But one of the active
compound 6o demonstrated higher ulcerogenic activity with index of 1.991 ± 0.34. The other
four compounds 6h, 6i, 6m and 6k showed ulecrogenic activity of 0.603, 0.642, 0.732 and 0.794
respectively with good GI safety profile. The most active compound 6k with an ED50 of 0.8575
mmol/kg has shown ulcer index of 0.794 ± 0.09 which is almost 2.96 folds higher than aspirin
and 2.64 folds higher and safer than ibuprofen.
Table 4: In vitro Human Red Blood Cell (HRBC) stability and In vivo anti-inflammatory activity
of synthesized pyrazole derivatives in carrageenan induced paw edema model in rats
Edema volume at (Mean±SEM)
% Membrane
Group Treatment
R1
stabilitya
1 hb
2 hb
3 hb
4 hb
1
---
1.18±0.05
1.22±0.05
1.27±0.05#
1.40±0.02##
Toxicant
controlc
0.96±0.02*
0.90±0.03*
1.00±0.10
0.94±0.12
0.90±0.04** 0.87±0.01**
0.84±0.04** 0.82±0.02**
Ibuprofend
Celecoxibd
2
3
75.03±0.013
72.01±0.010
4
5
4-OH
66.23±0.024
59.92±0.006
1.25±0.02
1.16±0.06
1.16±0.02
1.13±0.05
1.00±0.04*
1.05±0.05
0.96±0.03**
0.99±0.03**
6a
6b
3,4-
diOCH3
2-NO2
6
7
8
9
70.14±0.007
69.94±0.017
71.14±0.008
59.51±0.019
1.06±0.05
1.08±0.08
1.20±0.04
1.10±0.03
1.05±0.05
1.06±0.06
1.20±0.04
1.09±0.08
1.00±0.03*
1.05±0.03*
1.18±0.05
1.00±0.03*
0.95±0.03**
0.94±0.04**
0.98±0.08**
1.00±0.01**
6c
6d
6e
6f
2-Cl
4-F
H
18

10
11
12
4-OH
64.36±0.054
61.82±0.010
63.4±0.007
1.23±0.06
1.07±0.05
1.16±0.06
1.12±0.06
1.13±0.09
1.09±0.08
1.15±0.06
1.05±0.06
1.00±0.03*
0.99±0.03**
0.98±0.04**
0.97±0.01**
6g
6h
6i
4-OCH3
3,4-
diOCH3
3,4,5-
13
64.16±0.013
1.12±0.05
1.30±0.10
1.15±0.05
1.10±0.05**
6j
triOCH3
2-NO2
14
15
16
17
18
65.35±0.010
1.07±0.05
1.15±0.09
1.30±0.10
1.13±0.09
1.11±0.09
1.15±0.03
1.05±0.06
1.05±0.12
1.03±0.05
1.03±0.06*
1.05±0.05
0.93±0.04**
1.00±0.02**
0.94±0.04**
0.99±0.05**
0.95±0.03**
6k
6l
3-NO2
2-Cl
3-Cl
4-F
55.16±0.006455 1.16±0.06
67.63±0.017
56.15±0.017
71.18±0.029
1.12±0.05
1.15±0.09
1.21±0.05
6m
6n
6o
aValues were presented as mean ± SEM (n = 6)
bSynthesized compounds were administered at dose of 200 mg/kg.
cCarrageenan was used at dose 0.1 ml (1% w/v).
dStandard drug Ibuprofen and Celecoxib was administered at dose of 40 mg/kg.
Data are expressed as the mean ± SEM (n = 6/group) by using Student’s t - test (significant at
p<0.05* and 0.01**) for carrageenan induced paw edema results and % membrane stability by
using ordinary one way ANOVA (significant at p<0.0001). The statistical significance of
difference of carrageenan-control compared to vehicle-control group was shown as #p < 0.05,
the statistical significance of difference in synthesized compounds and standard drugs (Ibuprofen
and Celecoxib) treated groups as compared to carrageenan-control was designated as *p < 0.01.
Values of edema volume without symbols like * and # were ns (non-significant).
2.5.3. In vitro Cyclooxygenase (COX) Inhibition Studies
The COX-1 and COX-2 inhibition studies were carried out by Enzyme Immunoassay (EIA)
method with the ovine COX-1 and the human recombinant COX-2. IC50 values i.e. the inhibitory
concentration values were calculated and the experimental results is shown in the Table 5.
Selectivity Index (SI) of compounds was determined by using the formula IC50 of COX-1/COX-
2. The experiment was conducted as per the protocol described in the assay kit. Celecoxib,
19

selective COX-2 inhibitor was screened as standard for both COX-1 and COX-2 inhibition and
the study identified with SI of 147.05. All the tested compounds except 6c and 6n have exhibited
COX-1 inhibition greater than 50 µM signifying that the design approach is selective to COX-2
inhibition. Compounds (6a-o) demonstrated selective COX-2 inhibition in the range of 2.560 to
38.12 µM except 6d. The 2-chloro substituted compound 6d was found least active with COX-2
inhibition of 108.12 µM. Comparison of experimental COX-2 inhibition in pIC50 with
AutoQSAR predicted pIC50 values of compounds (6a-o) have shown satisfactory regression
value of 0.7216 (Figure 6). Further pharmacophore screening of chemical databases and
prediction of COX-2 inhibition by QSAR model will identify potential molecules.
Looking at the SAR of COX-2 inhibition, following compounds (6a-o) can be categorized into
two series. The compounds were categorized as (6a-e) as one series and (6f-o) as another series
with pyrazole ring unsubstituted and substituted with phenyl ring. The second series (6f-o) have
exhibited potent and selective COX-2 inhibition highlighting the significance of phenyl ring
substituted to nitrogen of pyrazole ring. Surprisingly, the least active compound 6d with an IC50
value of 108.12 µM is among the series of (6a-e). Among the compounds, 3,4,5-trimethoxy
derivative 6j and 4-methoxy derivative 6h were found with highest COX-2 inhibition at an IC50
of 2.56 and 4.32 µM. It can be noticed that substitution of methoxy group at 4th position is
essential. Similarly, looking at the other compounds substituted with methoxy group 6b and 6i
(3,4-dimethoxy) presented 6.24 and 11.48 µM. Substitution of chlorine at 3rd position for
compound 6n has demonstrated 7.75 µM and surprisingly substitution at 2nd position for 6m has
diminished the activity with 32.11 µM. Also substitution of fluoro group at 4th position for 6o
has reduced with 21.87 µM. Substitution of nitro group at 2nd and 3rd position for 6k, 6l has
showed 38.12 and 25.87 µM respectively. In the series of (6a-e), compounds 6b, 6e and 6a have
exhibited good COX-2 inhibition with an IC50 values of 6.24, 7.63 and 11.71 µM respectively.
With respect to Selectivity Index (SI), compound 6h has shown higher value of 63.99, among all
the synthesized compounds. The compounds 6i and 6j were observed with 26.19 and 32.11.
Similarly compounds 6b, 6m and 6o were found moderately selective with 16.15, 9.387 and
10.31 respectively. Except 6d, all the compounds exhibited moderate to potent and selective
COX-2 inhibition profile. The overall results suggests that substitution of phenyl group to
nitrogen of pyrazole and methoxy group at 3rd and 4th position, chloro group at 3rd position of R2
are crucial for potent and selective COX-2 inhibition.
20

Table 5: In vitro COX-1 and COX-2 inhibition for pyrazoles (6a-o) and ulcerogenic activity
Compound
Code
COX-2
(IC50 in
µM)
COX-2 Predicted COX-1 COX-1
SIb
Ulcerogenic
activityc
a
pIC50
pIC50
(IC50
in µM)
70.61
pIC50
11.71
4.931
5.204
4.596
3.966
5.117
4.646
5.141
5.364
4.940
5.591
4.418
4.587
4.493
5.110
4.660
ND
5.4088
5.4694
4.7738
4.5940
5.4088
4.7248
5.4405
5.2796
5.5909
6.2462
4.9055
4.9055
4.7248
5.4405
5.4405
ND
4.1511
3.9966
4.4234
3.5485
4.2897
3.9483
4.0569
3.5583
3.5218
4.0850
3.6803
4.1153
3.5208
4.4739
3.6465
ND
6.029
16.15
1.49
NDd
ND
ND
ND
ND
ND
ND
6a
6.240
25.31
108.12
7.630
22.56
7.220
4.320
11.48
2.560
38.12
25.87
32.11
7.750
21.87
ND
100.78
37.72
282.77
51.32
112.63
87.72
276.44
300.72
82.22
208.76
76.67
301.42
33.58
225.68
ND
6b
6c
2.615
6.726
4.992
12.14
6d
6e
6f
6g
63.99 0.603 ± 0.15
26.19 0.642 ± 0.25
6h
6i
32.11
5.476 0.794 ± 0.09
2.963 ND
9.387 0.732 ± 0.18
4.332 ND
10.31 1.991 ± 0.34
ND 2.100 ± 0.16
> 50.00 > 4.301 147.05 1.204 ± 0.06
ND
6j
6k
6l
6m
6n
6o
Ibuprofen
Celecoxib
Aspirin
0.34
6.468
ND
ND
ND
ND
ND
ND
ND
2.358 ± 0.28
aCOX-2 Inhibition values predicted from AutoQSAR model
bSelectivity Index of compounds IC50 of COX-1/IC50 of COX-2
cValues represents the Severity Index ± S.E.M
dND- Not Determined
21

Observed vs Predicted pIC50
6.5
6
5.5
5
y = 0.8942x + 0.8856
R² = 0.7216
4.5
4
3.5
4
4.5
5
5.5
6
Figure 6. Plot of COX-2 Inhibition pIC50 (experimental) vs AutoQSAR pIC50 (predicted) for
synthesized pyrazole derivatives
2.6. Pharmacophore hypothesis screening of synthesized pyrazole derivatives and Correlation
with experimental results
Synthesized pyrazole derivatives were screened through ligand based pharmacophore model to
investigate the specificity towards COX-2 enzyme. Interestingly all the 15 compounds passed
through the nominated and validated hypothesis. The AHHR_1 suggests that compounds
exhibited anti-inflammatory activity by inhibition of COX-2 enzyme. Compounds have
displayed fitness score for the hypothesis in the range of 1.104 to 1.671 except one compound 6n
with a score of 0.551 (Table 6). But the fitness score of standard drugs i.e., selective COX-2
inhibitors was observed in the range of 1.265 to 2.558 indicating that compounds were less or
moderately active inhibitors.
To further investigate, in vitro COX-2 inhibition pIC50 values of synthesized pyrazole derivatives
were correlated with fitness score of hypothesis. The COX-2 inhibition pIC50 values of
compounds 6(a-o) were observed in the range of 3.966 to 5.591. All the values were depicted in
Table 5. The present hypothesis could filter all the compounds as selective COX-2 inhibitors and
experimental COX-2 inhibition values too support the same. But poor correlation was observed
between fitness score and experimental pIC50 values. However, compound 6n was observed with
highest COX-1 inhibition of 33.58 µM and was found with least fitness score of 0.551. The
22

Pharmacophore hypothesis screening, being a qualitative approach helps in identifying the lead
for selective COX-2 inhibition and was validated with marketed anti-inflammatory drugs.
For the comparison of experimental in vitro and in vivo results with in silico data, five
compounds 6k, 6m, 6o, 6i and 6h have exhibited in vivo anti-inflammatory activity below 1
mmol/kg. The human RBC membrane stabilizing property of these compounds was observed in
the range of 61.83 to 71.2. Overall range observed for the compounds (6a-o) was 55.15 to 71.2
and it appears quite impressive in terms of correlation. Similarly, observed COX-2 inhibition
values in pIC50 was in a range of 3.966 to 5.591.The overall results of in vitro and in vivo anti-
inflammatory activities reveals that design approach is selective to COX-2 inhibition and active
molecules are safer than standard NSAIDs in the market. Surprisingly, the compounds like 6j
and 6h have shown potential in vitro COX-2 inhibition but failed to show in vivo anti-
inflammatory activity. But the compounds 6k and 6m exhibited good in vivo anti-inflammatory
activity but poor in vitro COX-2 inhibition. However the in vivo active compounds were found
with lesser ulcerogenic activity related to standard drugs in the market. Extending the in vivo
study for few more hours or conducting the pharmacokinetic study of active compounds might
help in better understanding the SAR and correlation of in vitro vs in vivo activity. Also further
modifications in the structure might produce a potent and safer molecule from this series.
23

Table 6: Correlation of in vivo anti-inflammatory, in vitro COX-2 Inhibition values and
hypothesis screening scores of synthesized pyrazole derivatives
a
Code
ED50
Matched Fitness Vector Volume pIC50
ligand
sites
3
score
score
score
(obs.)
mg/kg mmol/kg
436.5
483.1
422.5
417.3
466.7
499.8
483.1
466.7
451.3
733.7
395.8
499.8
417.3
483.1
422.5
4.606
3.060
1.1721
1.2064
1.0963
1.1133
1.2466
1.2000
1.1170
1.0090
0.9470
1.4040
0.8575
1.0828
0.9254
1.0346
0.9379
1.585
1.155
1.601
1.602
1.671
1.662
1.651
1.106
1.119
1.104
1.63
0.889
0.839
0.889
0.889
0.935
0.889
0.889
0.839
0.839
0.839
0.889
0.889
0.889
0.628
0.889
ND
0.402
0.380
0.419
0.419
0.359
0.480
0.469
0.332
0.344
0.329
0.448
0.427
0.459
0.332
0.459
ND
4.931
5.204
4.596
3.966
5.117
4.646
5.141
5.364
4.940
5.591
4.418
4.587
4.493
5.110
4.660
ND
6a
4
3
6b
6c
3
6d
3
6e
3
6f
3
6g
4
6h
4
6i
4
6j
3
6k
6l
3
1.61
3
1.642
0.551
1.641
ND
6m
4
6n
3
6o
ND
ND
Ibuprofen
Celecoxib
ND
ND
ND
ND
aED50 was calculated for 200mg/kg of compounds after 4 h.
Conclusion
Compounds substituted with electron-withdrawing group nitro and halogens like chloro, fluoro at
ortho position indicated a significant anti-inflammatory activity while the same substitution at
meta position showed a moderate activity. From the design aspects, structural features with
experimental therapeutic and toxicity evaluation, further structural modifications of identified
lead molecule might benefit a potent COX-2 inhibitor from this series. Pharmacophore model
can be exploited for the screening chemical databases of both natural and synthetic compounds
24

for lead identification and repositioning of drugs. With the SAR findings and further structural
modifications of lead molecule with assistance of selected QSAR model can serve in the design
of potent, selective and safe COX-2 inhibitor in future.
3. Experimental Section
3.1.Dataset for Pharmacophore and QSAR Modelling
The chemical structures of 37 substituted pyrazoles as selective COX-2 inhibitors were collected
from the literature as shown in Table 7a-7d [29,30]. From the series, 25 compounds were taken
as one set to build pharmacophore and QSAR modelling and remaining 12 compounds were used
to validate the models. All the compounds were synthesized, characterized, screened for both
COX-1 & COX-2 inhibition assay and in vivo anti-inflammatory study by same method.
Chemical structures of compounds were sketched using Meastro 3D Builder and energy
minimization was performed using Ligprep module[31]. To simplify the input data for in silico
computational studies, IC50 values of all the compounds were converted to pIC50. The dataset of
25 compounds were randomly divided to training and test sets in the ratio of 80:20 for building
QSAR using an advanced tool of Schrodinger Drug Design Suite [32], AutoQSAR.
25

Table 7a: Chemical structures of substituted pyrazoles (Group I)
1.2a-1.2f
1.3a-1.3f
Compound Code
R
Cl
H
H
H
H
H
R1
H
R2
H
R3
H
1.3a
1.3b
1.3c
1.3d
1.3e
1.3f
1.2a
1.2b
1.3c
1.2d
1.2e
1.2f
H
Cl
H
H
Br
H
H
NO2
OCH3
H
H
H
H
CF3
CF3
26

Table 7b: Chemical structures of substituted pyrazoles (Group II)
1.5
1.6a-1.6f
Compound
Compound
Code
R
R
Code
1.6d
1.6a
1.6e
1.6f
1.6b
1.6c
27

Table 7c: Chemical structures of substituted pyrazoles (Group III)
2.2a-2.2f
2.3a-2.3f
Compound Code
Ar
Compound Code
Ar
2.3a
2.3b
2.3c
2.2d
2.2e
2.2f
2.3d
2.3e
2.3f
2.2a
2.2b
2.2c
28

Table 7d: Chemical structures of substituted pyrazoles (Group IV)
2.5
2.8a-2.8d
2.9
Compound
Code
R
Compound
Code
R
4-Br
2-Cl
2.8c
4-NO2
2.8a
2.8d
3,5-diCF3
2.8b
3.2.Building and Validation of Pharmacophore Model
The ligand based pharmacophore models were built for selected 25 pyrazole derivatives from the
published literature and all the synthesized pyrazoles were biologically tested for both COX-1 &
COX-2 inhibition and in vivo anti-inflammatory activity. Chemical structures of the compounds
were sketched with Maestro 3D Builder [33]. IC50 values of all the compounds were added as
separate column and computed to logarithmic values (pIC50) in the project table of Maestro.
Pharmacophore model was built with an option of creating hypothesis with multiple ligands
using Phase module of Schrodinger Drug Design Suite [34]. Study was continued with default
options like generate and minimize 50 conformers for each molecule, generated hypothesis
should match atleast 50 % of all the actives and preferred up to 5 features with a minimum of 4
features in each hypothesis. It was allowed to generate and display top 10 models based on Phase
HypoScore. Hypothesis generated will have features in combination of hydrogen bond acceptor
29