Vicinal diaryl azole-based urea derivatives as potential cholesterol lowering agents acting through inhibition of SOAT enzymes

Article abstract of DOI:10.1016/j.ejmech.2017.02.038

A novel series of vicinal diaryl azole-urea derivatives were synthesized and evaluated for their potential to inhibit SOAT enzyme. Among the reported compounds, compound (12d) emerged as the most potent compound with an IC₅₀value of 2.43?μM. In polaxamer-407 induced lipoprotein lipase inhibition model, compound (12d) reduced triglyceride turnover in?vivo. Compound (12d) also showed dose-dependent prevention of serum total cholesterol and prevention of LDL-C elevation at a dose of 30?mg/kg. Furthermore, compound (12d) showed potential to stop falling levels of serum HDL-C dose-dependently and improved the atherogenic index. Effect of 12d on body weight, plaque formation and development of atherogenic lesions were studied. Toxicological study of compound (12d) indicated that at a dose of 2000?mg/kg, 12d was devoid of any signs of toxicity or mortality.

Full text of DOI:10.1016/j.ejmech.2017.02.038

European Journal of Medicinal Chemistry 130 (2017) 107e123
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Vicinal diaryl azole-based urea derivatives as potential cholesterol
lowering agents acting through inhibition of SOAT enzymes
Palash Pal, Hardik P. Gandhi, Ashish M. Kanhed, Nirali R. Patel, Niraj N. Mankadia,
Satish N. Baldha, Mahesh A. Barmade, Prashant R. Murumkar, Mange Ram Yadav*
Faculty of Pharmacy, Kalabhavan Campus, The Maharaja Sayajirao University of Baroda, Vadodara, 390001, India
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of vicinal diaryl azole-urea derivatives were synthesized and evaluated for their potential
Received 23 September 2016
Received in revised form
13 February 2017
Accepted 14 February 2017
Available online 20 February 2017
to inhibit SOAT enzyme. Among the reported compounds, compound (12d) emerged as the most potent
compound with an IC50 value of 2.43 mM. In polaxamer-407 induced lipoprotein lipase inhibition model,
compound (12d) reduced triglyceride turnover in vivo. Compound (12d) also showed dose-dependent
prevention of serum total cholesterol and prevention of LDL-C elevation at a dose of 30 mg/kg.
Furthermore, compound (12d) showed potential to stop falling levels of serum HDL-C dose-dependently
and improved the atherogenic index. Effect of 12d on body weight, plaque formation and development of
atherogenic lesions were studied. Toxicological study of compound (12d) indicated that at a dose of
2000 mg/kg, 12d was devoid of any signs of toxicity or mortality.
Keywords:
Atherosclerosis
Cholesterol lowering agents
SOAT inhibitors
© 2017 Elsevier Masson SAS. All rights reserved.
Vicinal diaryl compounds
1. Introduction
and storage of cholesterol in plasma and steroidogenic tissues like
the adrenals [8]. Cholesterol levels in the body are governed by
Cardiovascular diseases (CVD), a leading cause of deaths among
the human population arises due to high serum cholesterol levels
[1,2]. Other important risk factors include obesity, diabetes, hy-
pertension, cigarette smoking (traditional and electronic) and lack
of physical activity. Atherosclerosis is a CVD characterized by for-
mation of lipid-rich plaques in the inner walls of arteries. Chronic
hypercholesterolemia predisposes an individual towards the
development of atherosclerosis [2]. Technically, the entire process
is a remodeling event in the inner walls of medium and large ar-
teries that leads to subendothelial deposition of fatty substances
[3]. The end result of such a process is formation of thick, mature
and obstructive plaque. Rupturing of the plaque leads to hemor-
rhage and formation of emboli or thrombosis at the site of the
lesion [4]. Occlusion of the vessels may lead to myocardial infarc-
tion [5], the main cause of morbidity and mortality in the affected
persons.
various factors such as absorption of dietary cholesterol, in situ
cholesterol synthesis and biliary cholesterol excretion [5]. Several
therapeutic agents have been developed for lowering of lipid levels
[9,10a] which are in current use. These include two novel PCSK9
(proprotein convertase subtilisin/kexin type 9) inhibitors approved
in 2015 [10b]. Given the limitations of various lipid-lowering agents
such as statins, fibrates and bile acid sequestrants, the search for
novel lipid lowering agents still continues. There is a need for the
development of non-statin based cholesterol lowering agents due
to occurrence of/or concerns regarding adverse effects of statins.
The adverse effects associated with statins include, but are not
limited to, induction of diabetes mellitus, interstitial lung disease,
hemorrhagic stroke, decreased cognition, tendon rupture and
muscle-related symptoms. Muscle symptoms and tendinopathy are
the most common adverse effects of statins. Muscle symptoms
leading to elevation of creatine kinase (CK) levels upto 10 times the
upper limit of normal levels are serious concerns that frequently
lead to discontinuation of medication mandating the need for
alternative therapies. Another problem is that all patients do not
respond similarly to statins [10c]. Patients intolerant to statins or
those not responding adequately may need to switch to alternative
therapies to manage dyslipidemia and prevent atherosclerosis
[10d].
Cholesterol is a vital constituent required by the human body for
the viability of all types of cells [6]. However, excess of the
cholesterol is toxic to the cells causing cell death [7]. Formation of
cholesterol esters as neutral lipids is imperative to the transport
* Corresponding author.
E-mail address: mryadav11@yahoo.co.in (M.R. Yadav).
http://dx.doi.org/10.1016/j.ejmech.2017.02.038
0223-5234/© 2017 Elsevier Masson SAS. All rights reserved.

108
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
Cholesterol lowering therapies are vital to several car-
diometabolic conditions. Dietary reduction of cholesterol is
seldom complied with by the patients, and monotherapy with
statins alone does not achieve the ultimate treatment goals
[9,11]. Hence, an alternative strategy is to target the absorption
of intestinal cholesterol. During the past thirty years or so,
various therapeutic agents have been reportedly used to limit
cholesterol absorption from the intestine. Chemically diverse
entities such as guar gum, surfactants, sucrose polyesters,
neomycin and plant sterols have been reported to inhibit ab-
sorption of cholesterol through unknown or not so well estab-
lished mechanisms [12,13].
Esterification of cholesterol in the intestine is the rate limiting
step for its absorption. Sterol O-acyltransferase (SOAT), also
known as Acyl CoA: Cholesterol O-acyltransferase (ACAT), is
responsible for this process in the intestinal mucosal cells [14e18].
SOAT is a small family of enzymes comprising of homologous
members namely SOAT-1 and SOAT-2. Presence of SOAT-1 and
SOAT-2 is critical in the body for in vivo cholesterol homeostasis.
Although cholesterol is absorbed exclusively in the non-esterified
form, it appears in the lymph in the esterified form [12,19]. SOAT
is involved in the conversion of cholesterol to cholesteryl esters.
Thus, cholesterol is esterified by various fatty acids during its
absorption process [20,21]. Prevention of absorption of dietary
cholesterol by inhibiting SOAT has been reported to be beneficial
to the affected subjects [12,13]. Studies also imply that inhibition
of SOAT in the liver may result in lowering of plasma lipids
[12,22]. Foam cell formation and accumulation of cholesterol es-
ters within the artery walls get decreased reportedly with the
administration of systemically bioavailable SOAT inhibitors [23].
Because reducing of the elevated cholesterol levels is vital to
several cardiometabolic conditions, SOAT could prove to be a
promising target.
Although the three dimensional structures of SOAT family of
membrane-bound enzymes still remain unknown, a large number
of structurally diverse SOAT inhibitors were discovered in the past
using traditional techniques. In the last few years alone a number of
SOAT inhibitors have been reported [23,24]. A majority of them can
be classified into two major groups, the fatty acid anilides which
mimic the structure of Acyl CoA, and the phenylurea derivatives
and related substances. Certain compounds derived on the basis of
SAR of phenylureas, like avasimibe (I) and related moieties were
found to be very potent SOAT inhibitors but unfortunately they
failed miserably in clinical trials [25e28]. The DuPont Merck
Research Laboratories reported a series of diphenyl substituted
heterocycle-based ureas as SOAT inhibitors. DU-128 (IIa) having
diphenylimidazole moiety was found to be 50 times more potent
than the monoaryl derivative.
N
F
N
O
N
H
N
N
3
H
C7H15
F
(III)
Certain interesting common structural features are present in
the compounds of series II and III. All of them are urea derivatives.
Vicinal diphenyl pyridazine scaffold of series III is replaced by
vicinal diphenyl imidazole moiety in the most active compounds of
series II. In both of the series, di-/tri-substituted urea moiety is
directly attached to the phenyl ring while the heterocyclic scaffold
is separated by an alky linker chain having amino/thioether
grouping. Keeping the above structural features in mind, it was
planned to synthesize di-substituted urea derivatives in which the
urea moiety was directly linked to the heterocyclic moiety instead
of the linker chain, and the pyridazino/imidazolo moieties were
replaced by thiazolo/oxadiazolo heterocyclic rings. It was also
envisaged in some of the compounds to move the urea moiety from
the five membered heterocyclic rings to one of the phenyl rings of
the vicinal diphenyl ring system. Such systems would offer more
diverse structures and the impact of these variations on SOAT
inhibitory activity would be worth watching. Apart from the syn-
thesis of potential SOAT inhibitors and evaluating them biologically,
it was also planned to explore the three-dimensional structures of
SOAT sub-types, SOAT-1 and SOAT-2 using molecular modeling
techniques. The enzyme models were thought to be developed by
using the threading method as no homologous structures were
available to be used as template. These models were planned to be
validated by statistical methods available in I-TASSER and also by
developing ligand based pharmacophore models for the inhibitors
of SOAT enzymes using the data of the synthesized compounds and
comparing the pharmacophoric features with the docking results
obtained by docking of the most potent compounds in the pro-
posed active site of the developed SOAT enzymes' models. Though
these models may have scope for further improvement, this is the
first report of its kind in the direction of exploring the 3D structures
of the sub-types of SOAT enzymes for getting an in depth structural
insight for the development of more potent and safe SOAT in-
hibitors in a more rationale fashion.
2. Results and discussion
2.1. Chemistry
Syntheses of the targeted compounds were carried out as per
Schemes 1e5. For the synthesis of the final vicinal diaryl thaizoles
Pr.i
O
i.Pr
i.Pr
F
R1
R2
O
O
N
O
S
S
O
N
N
N
N
H
5
O
H
H
H
O
O
Br
O
i.Pr
i.Pr
F
a
b
(I)
(II) a: R1 = R2 = Ph
b: R1 = Ph; R2 = H
c: R1 = R2 = H
X
Y
O2N
Y
X
Y
(2a-2h)
(1a-1h)
(3a-3c)
Comp.
X
Y
Comp.
X
Y
(1a, 2a, 3a) NO2 Me
(1b, 2b, 3b) NO2
(1c, 2c, 3c) NO2 Cl
(1d, 2d) NO2 OMe (1h, 2h)
(1e, 2e) Cl Me
(1f, 2f) Cl
(1g, 2g) Cl OMe
(IIb) and approximately 300 times more potent than the cor-
responding 4,5-unsubstituted imidazole (IIc) as SOAT inhibitors in
in vitro assay. Diphenylpyridazineureido derivative (III) structur-
ally related to DuP128 (IIa) showed good inhibitory activity
F
F
H
H
against hSOAT-1 (IC50 0.94
forms [29].
mM) and hSOAT-2 (IC50 1.74 mM) iso-
Scheme 1. Synthesis of the starting derivatives (2a-2h, 3a-3c).
Reagents and conditions: (a) Br2, HOAc, rt; (b) SeO2, DMSO, mw.

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
109
Y
X
Y
Y
O
N
S
N
S
a
b
H
H
NH2
N
N
R
Br
O
X
X
(2a-2h)
(4a-4h)
(5a-5z')
Comp.
X
Y
Comp.
X
Y
(2a, 4a) NO2 Me (2e, 4e) Cl Me
(2b, 4b) NO2 (2f, 4f) Cl
(2c, 4c) NO2 Cl (2g, 4g) Cl OMe
(2d, 4d) NO2 OMe (2h, 4h) H
F
F
H
Scheme 2. Synthesis of 2-thiazolylurea derivatives (5a-5z’).
Reagents and conditions: (a) Thiourea, EtOH, reflux; (b) R-NCO, toluene, reflux.
Y
Y
Y
N
S
N
S
N
H
H
H
H
a
H
H
b
N
N
N
N
N
N
R
R
R
S
O
O
O
R'
O2N
H2N
N
H
(5a-5k)
(6a-6k)
(7a-7t)
Comp.
(5a, 6a) Me 2,4-F2Ph
(5b, 6b) Me 2,6-(Et)2Ph (5h, 6h) Cl
Y
R
Comp.
(5g, 6g) Cl
Y
R
n.Butyl
n.Dodecyl
(5c, 6c)
(5d, 6d)
(5e, 6e)
(5f, 6f)
Me n.Butyl
(5i, 6i) OMe 2,4-F2Ph
(5j, 6j) OMe n.Butyl
(5k, 6k) OMe n.Heptyl
F
F
2,4-F2Ph
n.Butyl
Cl 2,4-F2Ph
Scheme 3. Synthesis of derivatives (7a-7t).
Reagents and conditions: (a) Fe powder, NaCl solution, MeOH, reflux; (b) (i) Acetic anhydride or mesyl chloride, THF or (ii) Isopropyl bromide/n.dodecyl bromide, K2CO3, DMF.
and oxadiazoles, the required starting derivatives were prepared
using Scheme 1. The 1,2-diarylethanones (1a-1h) were prepared by
Friedal-Craft acylation of the suitably substituted benzenes with
the required phenylacetyl chloride derivatives. The ethanones (1a-
1h) were either brominated to the desired bromo derivatives (2a-
2h) or oxidized to the benzil derivatives (3a-3c) using selenium
dioxide in DMSO [30a,b].
To prepare the 2-ureathiazoles (5a-5z’), Scheme 2 was followed.
To construct the thiazole ring, the 2-bromoethanones (2a-2h) were
reacted with thiourea to obtain the 2-aminothiazole derivatives
(4a-4h) which when reacted with the desired type of isocyanates,
the required urea derivatives (5a-5z’) were obtained. 1H NMR
To Shift the urea functionality from the second position of the
thaizole to the para position of the 5-phenyl ring, two different
approaches were utilized (Scheme 4). The bromoethanones (2a-2d)
were reacted with thioactamide to yield the 2-methylthiazoles (8a-
8d). Reduction of the 4-nitro group of these derivatives offered the
amines (9a-9d). In one approach, the amines (9a-9d) on direct
reaction with the desired isocyanates offered the targeted urea
derivatives (12a-12x). In the second approach, the amines (9a-9d)
were first reacted with phenyl chloroformate to obtain phenyl
carbamates (10a-10d), which were further reacted with the
required amines to obtain the desired 1,2-disubstituted urea de-
rivatives (11a-11l). 1H NMR spectrum of a representative compound
spectra of a representative compound (5t) offered signals at
d
10.59
(11d) showed signals at d 6.94e7.50 (m, 11H, Ar-H), 6.06 (s, 1H, NH),
(bs, 1H, NH), 8.85 (bs, 1H, NH), 7.49e7.47 (d, 2H, Ar-H), 7.43e7.24
(m, 8H, Ar-H), 7.04e7.00 (t, 1H, Ar-H), 6.85e6.83 (d, 2H, Ar-H) and
3.77 (s, 3H, OCH3).
5.97 (s, 1H, NH), 3.31e3.34 (m, 2H, diisopropyl-CH), 2.82 (s, 3H,
CH3), 1.21e1.35 (m, 12H, diisopropyl-CH3). Among the directly
synthesized urea derivatives from the desired isocyanates the most
The nitro group in compounds (5a-5k) was reduced to the basic
amino group by reduction with iron power. The resulting amino
derivatives (6a-6k) were reacted with acetyl chloride/meth-
anesulfonyl chloride or alkyl halides to obtain acidic or basic lipo-
philic functionalities in compounds (7a-7t) (Scheme 3). 1H NMR
potent compound (12d) offered proton signals at d 7.12e7.40 (m,
8H, Ar-H), 6.59 (bs,1H, NH), 4.82 (bs, 1H, NH), 3.21e3.3(q, 2H,
NHCH2), 2.72 (s, 3H, thiazole-CH3), 2.32 (s, 3H, Ar-CH3), 1.45e1.52
(m, 2H, CH2), 1.32e1.41(m, 2H, CH2) and 0.91e0.95 (t, 3H, CH3) in its
1H NMR spectrum.
spectra of compound (7o) offered signals at
6.66e8.22 (m, 11H, Ar-H), 3.79 (s, 3H, OCH3), 3.57e3.79 (m, 1H,
NHCH) and 1.22e1.24 (d, 6H, CH3).
d
10.63 (bs, 1H, NH),
For the synthesis of the oxadiazole derivatives, benzils (3a-3c)
as the starting materials were oximated using hydroxylamine hy-
drochloride in a basic solvent like pyridine. The dioximes (13a-13c)

110
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
Y
Y
Y
O
N
S
N
S
a
b
Me
Me
Br
O2N
O2N
H2N
(2a-2d)
(8a-8d)
(9a-9d)
d
c
Y
Y
Y
N
S
N
S
e
N
S
Me
Me
Me
O
O
O
R
R'
N
N
PhO
N
N
N
H
H
H
H
H
(11a-11l)
(10a-10d)
(12a-12x)
Comp.
Y
(2a, 8a, 9a, 10a) Me
(2b, 8b, 9b, 10b) F
(2c, 8c, 9c, 10c) Cl
(2d, 8d, 9d, 10d) OMe
Scheme 4. Synthesis of compounds (11a-11l and 12a-12x).
Reagents and conditions: (a) Thioacetamide, MeOH, reflux; (b) Fe powder, NaCl solution, MeOH, reflux; (c) R-NCO, toluene, rt. (d) Phenyl chloroformate, THF, 0 ꢁC; (e) R-NH2, DMAP,
DMF.
oxadiazole series (16a-16o), offered signals at
d 8.41 (bs, 1H, NH),
Y
Y
Y
7.21e7.61 (m, 8H, Ar-H), 5.91 (bs, 1H, NH), 3.62e3.81 (m, 2H,
NHCH2), 2.61e2.67 (m, 4H, CH2), 1.2e1.4 (t, 3H, CH3) in its proton
NMR spectrum.
NOH
NOH
O
O
N
N
a
b
O
O2N
O2N
O2N
2.2. Biological
(3a-3c)
(13a-13c)
(14a-14c)
2.2.1. In vitro SOAT inhibition assay of the test compounds
The synthesized compounds were evaluated for their potential
to inhibit SOAT activity in vitro using rat microsomal SOAT. The test
c
Y
Y
compounds were evaluated at a fixed concentration (10 mM) for
N
N
N
N
pan-SOAT activity as per the reported method [31] and were not
evaluated for activity against individual isoforms. The results of this
preliminary screening are depicted in Tables 1e4. The preliminary
screening data showed that several of these compounds exhibited
d
O
O
O
R'
H2N
N
N
H
H
good inhibition (some of them ꢀ 60% at 10
mM concentration)
(15a-15c)
(16a-16o)
which was comparable to that of the standard pan-SOAT inhibitor
avasimibe (1). This data was used further to select potent com-
pounds for determination of their IC50 values. A threshold value of
35% inhibition was chosen arbitrarily and all those compounds
showing a minimum of 35% inhibition of microsomal SOAT activity
were selected for IC50 determination at five different concentra-
tions. Since it was also aimed to develop a credible 3D-QSAR model
to validate the pharmacophore models for the compounds reported
in this paper, some compounds with lower inhibition values were
also chosen for their IC50 determinations, as inclusion of full
spectrum of low to high potency compounds was a basic require-
ment for developing a good 3D-QSAR. The IC50 values are depicted
in Table S3, Supplementary material.
Comp.
Y
(3a, 13a, 14a, 15a) Me
(3b, 13b, 14b, 15b)
(3c, 13c, 14c, 15c) Cl
F
Scheme 5. Synthesis of oxadiazole derivatives (16a-16o).
Reagents and conditions: (a) Hydroxylamine HCl, pyridine, reflux; (b) Succinic anhy-
dride, 180 ꢁC; (c) Fe powder, NaCl solution, MeOH, reflux; (d) R-NCO, toluene, rt.
so obtained were dehydrated using succinic anhydride to obtain
the oxadiazoles (14a-14c). The nitro group in these derivatives
(14a-14c) was reduced and the resulting amino derivatives (15a-
15c) were reacted with the required isocyanates to obtain the tar-
geted 1,2-disubstituted urea derivatives (16a-16o) having oxadia-
zole ring in place of the thiazole ring system (Scheme 5).
Looking at the contents of these tables (Tables 1e4), it could be
inferred that placing the urea moiety at the 4th position of the 5-
phenyl ring (Table- 3) was best suited for good SOAT inhibitory
activity and it offered more number of potent compounds as
compared to the 2-ureidothiazole (Tables 1 and 2) analogs.
Compound (16h),
a representative compound among the

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
111
Table 1
The synthesized urea derivatives (5a-5z’) and their in vitro SOAT enzyme inhibitory activity.
Comp.
X
Y
R
% Inhibition
Comp.
X
Y
R
%
(at 10
m
M conc)
Inhibition
(at 10 mM conc)
(5a)
(5b)
(5c)
(5d)
(5e)
(5f)
(5g)
(5h)
(5i)
(5j)
(5k)
(5l)
(5m)
(5n)
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
Cl
Me
Me
Me
F
F
Cl
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
2,4-F2Ph
n.Butyl
2,4-F2Ph
n.Butyl
n.Dodecyl
2,4-F2Ph
n.Butyl
3.86
1.03
3.86
3.59
2.98
3.28
3.02
1.77
3.40
3.12
1.52
9.58
3.93
1.52
(5o)
(5p)
(5q)
(5r)
(5s)
(5t)
(5u)
(5v)
(5w)
(5x)
(5y)
(5z)
(5z’)
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
H
Me
F
F
F
F
OMe
OMe
OMe
OMe
H
n.Butyl
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Dodecyl
Ph
2,4-F2Ph
n.Butyl
n.Heptyl
Ph
36.37
42.34
1.58
3.30
1.91
56.86
2.98
27.58
1.73
10.56
3.03
2.56
Cl
Cl
OMe
OMe
OMe
Me
Me
Me
n.Heptyl
Ph
2,4-F2Ph
2,6-(Et)2Ph
H
H
H
H
H
H
2,4-F2Ph
n.Butyl
n.Dodecyl
Cl
Cl
1.20
Table 2
The prepared urea derivatives (7a-7t) and their in vitro SOAT enzyme inhibitory activity.
Comp.
Y
R0
R
% Inhibition
Comp.
Y
R0
R
% Inhibition
(at 10
m
M conc)
(at 10 mM conc)
(7a)
(7b)
(7c)
(7d)
(7e)
(7f)
(7g)
(7h)
(7i)
Me
Me
Me
Me
F
F
F
F
Cl
Cl
COMe
SO2Me
COMe
SO2Me
COMe
SO2Me
COMe
SO2Me
COMe
SO2Me
2,4-F2Ph
2,4-F2Ph
n.Butyl
3.06
2.17
5.23
32.67
23.32
18.9
15.4
0.43
2.03
6.39
(7k)
(7l)
Cl
Cl
COMe
SO2Me
COMe
SO2Me
i.Pr
n.Dodecyl
COMe
SO2Me
i.Pr
n.Butyl
n.Butyl
2,4-F2Ph
2,4-F2Ph
2,4-F2Ph
2,4-F2Ph
n.Butyl
n.Butyl
n.Butyl
n.Butyl
3.75
5.63
(7m)
(7n)
(7o)
(7p)
(7q)
(7r)
(7s)
(7t)
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
38.98
19.63
51.73
1.77
39.45
5.45
n.Butyl
2,4-F2Ph
2,4-F2Ph
n.Butyl
n.Butyl
2,4-F2Ph
2,4-F2Ph
4.67
3.36
(7j)
n.Dodecyl
Table 3
The phenylurea derivatives (11a-11l and 12a-12x) and their in vitro SOAT enzyme inhibitory activity.
Comp.
Y
R
% Inhibition
Comp.
Y
R
%
(at 10
m
M conc)
Inhibition (at 10 mM conc)
(11a)
(11b)
(11c)
(11d)
(11e)
(11f)
Me
Me
Me
F
F
F
2,6-(i.Pr)2Ph
11.19
21.88
18.96
42.12
41.98
29.32
(11g)
(11h)
(11i)
(11j)
(11k)
(11l)
Cl
Cl
Cl
OMe
OMe
OMe
2,6-(i.Pr)2Ph
12.64
24.44
12.31
30.13
25.44
21.46
3,4,5-(OMe)3Ph
Morpholinoethyl
2,6-(i.Pr)2Ph
3,4,5-(OMe)3Ph
Morpholinoethyl
3,4,5-(OMe)3Ph
Morpholinoethyl
2,6-(i.Pr)2Ph
3,4,5-(OMe)3Ph
Morpholinoethyl
Comp.
Y
R′
% Inhibition
Comp.
Y
R′
%
(at 10
m
M conc)
Inhibition
(at 10 mM conc)
(12a)
(12b)
(12c)
(12d)
(12e)
(12f)
(12g)
(12h)
(12i)
(12j)
(12k)
(12l)
Me
Me
Me
Me
Me
Me
F
F
F
F
F
Ph
13.36
26.63
30.94
64.91
33.41
46.19
16.81
4.97
46.68
59.91
62.96
34.47
(12m)
(12n)
(12o)
(12p)
(12q)
(12r)
(12s)
(12t)
(12u)
(12v)
(12w)
(12x)
Cl
Cl
Cl
Cl
Cl
Cl
OMe
OMe
OMe
OMe
OMe
OMe
Ph
6.88
1.3
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Heptyl
n.Dodecyl
Ph
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Heptyl
n.Dodecyl
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Heptyl
n.Dodecyl
Ph
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Heptyl
n.Dodecyl
63.32
60.71
52.18
5.75
11.37
22.27
31.62
40.02
22.21
27.52
F
Comparing the activities of the compounds obtained from the
thiazole and the oxadiazole ring systems, it became clear that
thiazole ring was better suited to SOAT inhibition than the oxa-
diazole ring (Table 4) because the thiazole ring yielded more
number of compounds with higher potency.
moiety and methoxy group (Y) on the phenyl ring offered more
potent compounds (5t, 7m, 7o, 7q). Polar groups like amino group
on the other phenyl ring (6b) offered low potency compound.
Substituting an acetyl/methanesulfonyl or an alkyl on the amino
group did not improve the activity profile of the resulting com-
pounds beyond a certain point.
Apart from the fact that 2-ureidothiazole scaffold offered low to
medium active compounds, aryl substituents as R on the urea
In the 2-methylthiazole series having the substituted urea

112
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
Table 4
In vitro SOAT enzyme inhibitory activity of compounds (16a-16o).
Comp.
Y
R0
% Inhibition
Comp.
Y
R0
% Inhibition
(at 10
m
M conc)
(at 10 mM conc)
(16a)
(16b)
(16c)
(16d)
(16e)
(16f)
(16g)
(16h)
Me
Me
Me
Me
Me
F
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
n.Heptyl
n.Dodecyl
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
13.52
0.34
2.83
18.65
9.15
47.06
44.85
54.5
(16i)
(16j)
F
F
n.Heptyl
n.Dodecyl
2,4-F2Ph
2,6-(Et)2Ph
n.Butyl
30.93
50.07
21.23
8.25
23.03
4.3
(16k)
(16l)
(16m)
(16n)
(16o)
Cl
Cl
Cl
Cl
Cl
n.Heptyl
n.Dodecyl
F
F
18.51
moiety attached to the para position of the 5-phenyl ring (com-
pounds 11 and 12), non-polar and bulky groups are preferred at
both the positions (Y and R’; 11d, 12d, 12f, 12i-12k, 12o-12q). Polar
groups like morpholinoethyl (at R0 position) or methoxy (at Y po-
sition) are not preferred for higher potency. Methyl or chloro
groups at position Y, and n.butyl, n.heptyl or 2,6-diethylphenyl
groups at R0 position yielded the most potent compounds. Com-
pound (12d) with a combination of methyl at Y position and n.butyl
at R0 position was found out to be the most potent compound (IC50
seemed promising and comparable to the standard inhibitors of
SOAT like avasimibe and pactimibe but majority of these com-
pounds were retracted from further perusal due to prevalence of
different types of problems with them [32,34,37,38].
The compounds reported in this paper are all di-substituted
urea derivatives. Altering the basic scaffold by exchanging hetero-
aryl moieties i.e., thiazole/oxadiazole in place of imidazole, and
altering the substituent pattern have been tried in this study to
generate a range of novel compounds which offered some potent
SOAT inhibitors. As is evident from the results (Table S3, Supple-
mentary material), compounds (12d, 12i, 12k, 12o, 12p and 12q)
were the most potent inhibitors, comparable in activity to the
standard avasimibe (I) [39]. Considering their in vitro potency it was
planned to evaluate these compounds further in some in vivo ani-
mal model(s).
2.43
mM) with SOAT inhibition activity which was better than the
standard drug avasimibe (I; IC50 4.01
mM).
In the oxadiazole series (compounds 16), the most potent
compounds were observed to have fluoro group at position Y. Bulky
lipophilic groups like 2,6-diethylphenyl, n.butyl, n.dodecyl or 2,4-
difluorophenyl at R0 position offered more potent compounds
than the rest in the series.
It can be summarized that thiazole ring is better suited to SOAT
inhibitory activity over the oxadiazole ring and the urea moiety
should not be directly linked to the thaizole ring system in the
vicinal diarylthiazoles. On the contrary, the urea moiety should be
attached to one of the phenyl rings and the substituents on the urea
moiety and the other phenyl ring should be bulky and nonpolar in
nature for obtaining potent SOAT inhibitors.
Several urea-based compounds reported in the literature have
been reported to show good microsomal SOAT inhibitory activity.
These reported compounds are usually di- or tri-substituted urea
derivatives with an array of substituents on either sides of the urea
moiety. Several researchers have reported SAR studies of their own
di-substituted urea derivatives showing promising in vitro SOAT
inhibition and good anti-hypercholesterolemic activity. These
compounds were found to exhibit [32e36] microsomal SOAT in-
2.3. Poloxamer-407 induced lipoprotein lipase inhibition model for
evaluating potent compounds
Increase in triglyceride production rates has been observed after
inhibition of plasma lipoprotein lipase causing temporal changes in
plasma triglyceride (TG) levels [40e42]. Poloxamer-407 or P-407
(Pluronic® F-125), a non-ionic detergent was found to be an in-
hibitor of lipoprotein lipase. It has been reported [43,44] that
known inhibitors of SOAT prevent the cholesteryl ester input in
plasma and reduce hepatic TG secretion rate. Fig. 1 shows the ef-
fects of the selected compounds (12d,12i,12k, 12o,12p and 12q) on
TG turnover following P-407 administration. In commensurate
with the in vitro data, compound (12d) was found to be the most
active compound in this in vivo model also, amongst the selected
lot. Administration of compound (12d) caused only moderate
changes in TG turnover (~20 fold) over a period of 24 h. Similar
hibition (IC50) in the range of 0.007e5.3
mM. These activities
Fig. 1. Effects of test compounds and standard drug on 24 h triglyceride turnover measured as fold-change in TG at 0 h and 24 h after administration of P-407.

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
113
effect was also observed in the avasimibe-treated group. While the
control group animals showed a mean 80-fold change in base-line
TG levels, the other derivatives showed 50e90 fold changes in TG
turnover. This experiment revealed that there could be a correla-
tion between SOAT inhibition and TG turnover as avasimibe was
also found to be active against TG turnover mediated through LPL
inhibition. It is known [45] that chylomicrons, which form a sig-
nificant chunk of TG, after losing triacylglycerols acquire cholesteryl
esters from the lipoproteins. Since administration of SOAT in-
hibitors can prevent de Novo formation of cholesteryl esters, the
pool of cholesteryl esters that remains available for acquisition to
chylomicrons is depleted. This biochemical inhibition thus corre-
lates well with the observations of the present study where
administration of SOAT inhibitors like avasimibe and the test
compound (12d) resulted in reduced TG turnover in vivo.
and kept on rising towards the terminal stage of the study in the
progression control group. A dose-dependent prevention of serum
total cholesterol elevation was observed in the groups (Fig. 2) that
were co-administered compound (12d). A dose of 10 mg/kg was not
found to be significant in preventing cholesterol elevation but the
higher dose (30 mg/kg) group showed a dramatic prevention
(P < 0.01) of cholesterol elevation with serum total cholesterol
levels remaining constant at 138 9.50 mg/dl at the end of the
study. On the contrary as expected, HDL-C values in the control
group showed a downward trend as the study progressed and a 2-
fold reduction in serum HDL-C levels was observed at the end of the
study. Treatment with compound (12d) was able to stop the falling
levels of HDL-C dose-dependently in the treated animals (Fig. 2).
SOAT inhibition causes significant changes in availability of
cholesterol pools. This leads to mobilization of intracellular
cholesterol from the SOAT-available compartment to that of
apolipoprotein-mediated HDL assembly. This could be the possible
mechanism by which 12d prevented a decline in HDL-C levels
[46b]. Minor elevations in triglyceride levels between the initial
and final stages of the study were observed but the changes were
not statistically significant (P value 0.251).
2.4. Evaluation of compound (12d) in animal models of diet-
induced atherogenesis
2.4.1. Effects on serum lipids and lipoproteins
An animal model of diet-induced atherogenesis in rats has been
reported [46] to be a valid model for the study of compounds
effective against atherosclerosis. It was decided to observe the ef-
fect of compound (12d) on serum levels of lipids and lipoproteins in
diet-induced atherogenesis in rats.
Before the initiation of the study, the mean cholesterol values
for all the animals were found to average around 91.16 4.54 mg/dl.
It was observed that cholesterol levels started rising dramatically
after about 2-weeks of the initiation of high cholesterol diet (HCD)
Analysis of lipoprotein cholesterol levels (LDL-C and VLDL-C)
were based on the Friedewald's formula [47a] which showed that
VLDL-C elevations were not significant among the treated groups
(Fig. 3a). However, a sharp rise in LDL-C (more than 12-fold) was
observed in the progression control group (P < 0.001). The test
compound-treated lower dose (10 mg/kg) group showed only
modest prevention in LDL-C elevation but the higher dose (30 mg/
kg) treated group indicated a significant prevention (P < 0.001;
Fig. 2. Effect of compound (12d) on (a) Total cholesterol; (b) HDL levels; (c) Triglycerides.
Fig. 3. Effects of compound (12d) on (a) VLDL-C, (b) LDL levels, (c) Atherogenic index.

114
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
Fig. 3b) of LDL-C elevation as compared to the progression control
group.
Atherogenic index, a surrogate marker of coronary artery dis-
ease, is determined by ratio of triglycerides to HDL [47b]. The
atherogenic index indicated the prevalence of non-HDL cholesterol
over HDL-C This data indicated that after 8 weeks of administration
of HCD, the atherogenic index peaked to 19.59 5.8 in the control
group animals. Such a high value of the atherogenic index indicated
that the animals were prone to the development of atherosclerotic
lesions. Co-administration of compound (12d) in the diet reduced
the atherogenic index to 7.55 0.9 and 1.82 0.34 respectively in
the 10 mg/kg and 30 mg/kg treated groups (Fig. 3c).
Fig. 5. Lipid accumulation lesion area in the aortae.
2.5. Effect of compound (12d) on body weight
The curve of body weight remained relatively unaffected over
the whole treatment period for all the groups. No significant
changes in body weights of the animals were observed between the
final and initial stages of the study. At the end of the study there
was no significant difference between the progression control
(290.3 8.1 g) and the test compound (12d) treated groups at doses
intermittent fatty streaks are evident in aorta of the lower dose
(10 mg/kg) treated (with compound 12) animals. Absence of such
fatty streaks could be seen in the aorta of animals treated with the
higher dose (30 mg/kg) of the test compound (12d). These obser-
vations are indicative of the atheroprotective role of compound
(12d).
Based on the results of in vivo evaluation studies of compound
(12d), it was concluded that the test compound (12d) possessed
significantly high potential in modulating the pathological devel-
opment of atherosclerotic lesions positively.
of 10 mg/kg (294
14.3g) and 30 mg/kg (299
6.6 g). We hy-
pothesize that SOAT inhibition-mediated cholesterol reduction
might lead to hyperphagic response owing to modulation of leptin
levels on chronic therapy. This mechanism might be responsible for
maintaining the body weight in spite of significant cholesterol
reduction. Fig. 4 shows the difference in body weights among the
groups.
2.7. Effect of compound (12d) on lumenal lipid plaque formation
Based on these results it was decided to evaluate compound
(12d) in an atherosclerosis model because the test compound (12d)
also showed favorable effect on serum triglycerides apart from
microsomal SOAT inhibition.
Lumenal plaque formation in the control and test compound
treated groups was studied by taking multiple sections of the
aortae and then staining them with Sudan Red IV to stain the lipid
core of the plaque. Representative images from different groups are
shown in Fig. 7. The images indicate that long-term consumption of
atherogenic diet has led to the development of a plaque core which
obscured the lumen of the aortic root (also known as aortic sinus).
Aortic root is the most putative site for the formation of lipid-rich
plaques, although a plaque may be formed in any anatomical re-
gion of the artery. Such observations of thickened intima in pro-
gression control groups have been reported [48]. No evidence for
the development of lipid-rich cores was found in the test com-
pound (12d)-treated animals.
2.6. Effect of compound (12d) on lipid accumulation lesion area in
the aortae
Lesion area in the aortic strips were characterized by en face
lipid staining with Sudan red IV. Photographs of pinned aortae
upon area analysis by ImageJ revealed that 22.35 7.9% area from
the progression control group was stained as compared to the
16.21 1.07% (P > 0.05) in the lower dose (10 mg/kg) treated-group
and 8.42 2.7% (P < 0.05) in the higher dose (30 mg/kg) treated
group (Fig. 5). The results indicate that treatment with 30 mg/kg of
compound (12d) shows nearly 3-fold reduction in the lesion area.
Representative images of the aortae are shown in Fig. 6.
It is worthy to mention here that the protective effects of
compound (12d) reported in this study were obtained without any
evidence of cutaneous or adrenal toxicities, which are the
commonly reported serious side effects of SOAT inhibitors [48e50].
Accumulation of lipids in the intimal was indicated by the for-
mation of fatty streaks in the aortic strips which got stained in
maroon color with Sudan red IV. These types of streaks are evident
along the length of the aorta of the control group animals. Some
2.8. Toxicological evaluation of compound (12d)
2.8.1. Single dose acute oral toxicity
Earlier studies on different urea-based derivatives have shown
fatal toxicities of the test compounds upto the doses of 2000 mg/kg
[49,50]. Some compounds have been reported to exhibit adrenal
toxicity at high doses which could be an indication of non-specific
cytotoxicity [37,51]. To ensure the safety of the test compound
(12d), it was administered at 2000 mg/kg to a group of animals. The
post-treatment examination period was 14 days from the date of
dosing the animals. Body weights of the animals were recorded on
days 0, 7 and 14. Slight fluctuations were observed in the body
weight of the animals during the observation period. The animals
were closely watched during the first 6 h of dosing. The animals
were starved during this period with free access to water. No sig-
nificant changes were recorded during this period. This part coin-
cided with the light cycle and most of the animals were asleep.
When awake, the animals showed normal grooming behavior and
Fig. 4. Effect on body weights of groups.

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
115
Fig. 6. Accumulation of intimal lipids in control and compound (12d) treated groups. Formation of fatty streaks in progression control group (A); in test compound treated group at
a dose of 10 mg/kg (B); at dose of 30 mg/kg (C).
Fig. 7. Cross-sectional areas from aortic roots after sacrificing animals of different groups. Streaks in the intimal area were not observed in any group but development of plaque was
evident in the HCD progression control (A). Minor damage to the arterial wall was evident in the region where plaque development was initiated. Intimal linings appear normal and
there was no evidence of diameter obscuration in the treated group (B and C). The scale bar in figure C represents 500 mm.
the water intake was also normal. During the entire post-treatment
observation period, special attention was paid to observe alteration
of skin color or fur, abnormal locomotion or breathing and changes
in the eyes. No untoward observations were made in this regard till
the terminal day of the study. No mortality was observed till day 14.
At the end of the study period, the animals were euthanized and
major organs like brain, heart, lung, liver, kidney, spleen and ad-
renals were harvested. Gross necropsy was performed by an indi-
vidual blinded to the groups. No macroscopic lesions were
observed. Viscera, gastrointestinal tract and mucus of the lungs
appeared normal. Adrenal glands exhibited normal shape and size
and any signs of shrinkage were absent.
not suggest any major digression from normal values. Urinary
output and hematological data appeared normal.
It was concluded from this study that repeat dose oral admin-
istration of the test compound (12d) at a dose of 60 mg/kg was safe.
2.10. Computational studies
All members of MBOAT family have several transmembrane
domains and majority of them possess two conserved residues
which may be the active-site residues, a conserved histidine (His)
embedded in a hydrophobic stretch of residues and an asparagine
(Asn) or histidine within a more hydrophilic region, some 30e50
residues upstream [24,54]. Till date the 3D structures of SOAT-1 and
SOAT-2 isoforms have not been reported. As there is no homologous
protein present for the conserved part of SOAT, it is very difficult to
generate the 3D structure of SOAT. So, it was planned to generate a
3D structure of the conserved part of SOAT using de novo protein
modeling. These generated protein models were compared with
the results obtained from pharmacophore and docking studies.
With the help of NCBI Blast, the MBOAT conserved region of the
isoform SOAT-1 (amino acids 286 to 520) has been identified which
represents the functional unit of the target protein (Fig. S1, sup-
plementary material).
Administration of 2000 mg/kg of the test compound (12d)
showed no signs of toxicity or mortality during the test period. The
LD50 of compound (12d) in rats was thus found to be > 2000 mg/kg.
2.9. Repeat dose oral toxicity
Repeat dose toxicity study for 28 days at 60 mg/kg oral dose of
compound (12d) was performed on rats. No untoward observations
were made regarding body weight, food intake or normal behavior
at the end of the study. The most common side effect related to the
mechanism of action of SOAT inhibitors upon repeated dose
administration is cutaneous xanthomatosis [48,52,53]. This effect
was not evident in any of the subjects under study even after an
extended period of 14-days of the study. Biochemical estimates did
The obtained sequence was submitted to I-TASSER (Iterative
Threading ASSEmbly Refinement) [55e57], a hierarchical method
for prediction of the protein structure and function. In this method

116
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
the structural templates are initially identified from the PDB by
multiple threading approach LOMETS (Local Meta-Threading-
Server) which is a locally installed PDB library for meta-threading
approach having multiple threading programs in I-TASSER server.
Templates of the highest significance in the threading alignments
were identified using LOMETS by measuring the Z-score (Table S1,
supplementary material).
After the threading alignment, thousands of conformations
(called decoys) were generated and to finalize the model, SPICKER
program was used in I-TASSER to cluster all the decoys based on the
pair-wise structure similarity, and report up to five models, which
corresponded to the five largest structure clusters. From these
models the confidence of each model was quantitatively measured
by C-score. With the best C-Score model, TM-align program was
used to match it with all the structures in PDB. From the TM-align,
top ten proteins with the maximum structural similarities to the
predicted I-TASSER model were obtained (Table S2, supplementary
material).
receptor interactions. The occlusion maps are represented by blue
and red colors indicating favorable and unfavorable interactions
respectively. The occlusion maps (donor, hydrophobic and elec-
tronegative) with the most active compound (12d) for the devel-
oped 3D-QSAR model are shown in Fig. S5A (supplementary
material). As shown in Fig. S5A for the most active compound (12d),
the blue regions are observed near the NH group of urea linker,
explaining its high activity because of its hydrogen bond donating
tendency. Lower potency of compounds (11i and 12h) could also be
explained very well because the urea NH moiety in these com-
pounds was overlapping the non-favorable i.e. red region. In
docking study of the most active compound (12d) the said NH
moiety was observed to be oriented towards the eCONH2 of
Asn421 of SOAT1, while in SOAT2 the same group appeared towards
C]O group of Leu275 and Val398; whereas in the less potent
compound like (11i), this was not the case and the change in
orientation was mainly observed because of the morpholine group
which was separated from the urea moiety by an ethylene linker.
Fig. S5B shows the hydrophobic volume occlusion map. Here, the
high activity of 12d could be explained pretty well, as the alkyl
(methyl) and the aromatic groups were in blue regions causing
possible hydrophobic interactions. The binding affinity of 12d to the
As the highest matching model 4cadC represents the integral
membrane protease Rce1 and our target SOAT-1 is also a membrane
bound protein, Model1 was selected for determining the 3D
structure of MBOAT conserved domain of the SOAT-1.
As the isoforms 1 and 2 are having almost 55% overall sequence
identity and more than 60% MBOAT sequence identity, the 3D-
structure of SOAT-2 was developed using MBOAT sequence (Fig. S2,
supplementary material) and the developed SOAT-1 model as the
template structure in MODELLER, with the help of homology
modeling. As the probable active site residues His-460 for SOAT-1
and His-434 for SOAT-2 had been reported, these were used to
identify the active sites in both of the developed receptor models.
The pharmacophore and 3D-QSAR models for SOAT inhibitors
were developed using our own 32 test compounds reported in this
paper as SOAT inhibitors by using PHASE module of Schrodinger.
The SOAT inhibitory activity in IC50 (M) was converted into 1/
logIC50 values (pIC50) which was used as the biological activity
parameter for the development of the pharmacophore model
(Table S3, supplementary material).
PHASE [58] employs tree-based partitioning algorithm and de-
tects thoroughly the spatial arrangements of functional groups that
are common and necessary for the biological activity for a set of
high-affinity ligands. The data set having 32 molecules was divided
into a training set containing 26 molecules and test set having 6
compounds for the development of 3D-QSAR models. For the dis-
tribution of dataset into training and test set, structural diversity
over the full range of activity was considered and the models were
developed.
With a prerequisite that all active compounds should match all
the features, five featured common pharmacophore hypotheses
were developed from different variants by means of a tree-based
partition algorithm. These generated hypotheses were assessed
on the basis of ‘Survival’ and ‘Survival-inactive’ scores. The best
model was found to be linked with the five-point hypothesis having
a hydrogen bond acceptor group, a hydrogen bond donor group, a
hydrophobic group and two aromatic rings (ADHRR). The inter-site
distances and angles in the pharmacophoric features were aligned
on the best fit compound which also happened to be the most
active compound (12d, Fig. S3, supplementary material). The graph
for actual verses predicted activity of training set and test set are
shown in Fig. S4 (supplementary material). To evaluate the pre-
dictive ability of the model, the test set compounds were used.
Based on the R2, Q2, SD, RMSE and Pearson-R values, the developed
3D-QSAR model was found to be predictive and accurate (results
are shown in Table S4, supplementary material).
enzyme is further strengthened by p-p stacking of the thiazole ring
with His460 and by hydrophobic interactions of the aromatic rings
with Phe453, Phe486 and Trp499 of SOAT1. In SOAT2 model the
hydrophobic interactions of the said groups are observed with
Val406, Phe427, His434, Met463 and Trp473 residues of the
enzyme. The urea group was seen in the red region exhibiting hy-
drophilic interactions. This observation was supported by the
docking study as the said urea group was observed to be placed in
an area covered by polar regions of Val424 and Asn421 of SOAT1,
and in case of SOAT2 the polar groups of Val398 supported similar
observations. Deviations from favorable interactions were observed
in case of a majority of the less active compounds. Low activity of
compound (12s) could be explained as most of the hydrophobic
groups in the ligand were oriented into the red regions. High ac-
tivity of 12d was further proved by the presence of carbonyl and
thiazole ring-N in the blue regions, as seen in the electronegative
maps (Fig. S5C, supplementary material).
Apart from the 3D-QSAR validation parameters, other widely
reported statistical validation techniques were also applied [59,60].
To derive these statistical results, data set was divided into two sets.
Compounds with pIC50 4.9 and above were considered as ‘active’
and those with activity below this value were considered as ‘inac-
tive’, and modified R2 i.e. (R2m), sensitivity, specificity, accuracy,
precision and Matthew's correlation coefficient (MCC) were
calculated (results are shown in Table S5, supplementary material).
Further, docking of the most active ligand (12d) was done on the
predicted active sites of SOAT1 and SOAT2. Results of the docking
studies further validated existence of pharmacophoric features
projected in the generated pharmacophore models. The binding
patterns of the most active compound (12d) with SOAT-1 and
SOAT-2 are shown in Fig. 8 and Fig. 9 respectively. The active
compound (12d) shows almost similar affinity to the proposed
active site residues of both the isoforms. The docking score for
compound (12d) with SOAT-1 was ꢂ6.15, whereas it was observed
to be ꢂ6.67 with SOAT-2. In SOAT-1 the test compound (12d) in-
teracts mostly by van der Waals or hydrophobic interactions. The
thiazole ring forms
p-p stacking with His460, whereas the two
aromatic rings show hydrophobic interactions with Asn421,
Phe453 and Trp499. In addition to these interactions the butyl
chain fits into the hydrophobic pocket of Leu304, Ile377, Thr380,
Phe381 and Tyr417. While in case of SOAT-2 the proposed active
The volume occlusion maps obtained from 3D-QSAR model
recognized the important features necessary for the ligand-
site residue His343 shows strong
phenyl ring attached to the thiazole ring. The thiazole ring and the
p-p interaction with the p-methyl

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
117
Fig. 8. Docking of the most active compound (12d) in the active site of the proposed SOAT-1 model.
Fig. 9. Docking of the most active compound (12d) in the active site of the of proposed SOAT-2 model.
two aromatic rings attached to the thiazole ring are observed to be
stabilized by hydrophobic interactions with Val406, Asp409,
Phe427, Ile456, Leu460, Met463, Gln467 and Trp473. The butyl
chain is strongly stabilized by hydrophobic part of Leu275, Ile354,
Tyr391, Trp394 and Val398. In 11i, a less potent compound, the
morpholine ring of the compound possessing higher topological
polar surface area and poor lipophilicy appears in the polar region
of the target protein as a complimentary region for interaction
hence, it should be better stabilized. But, in docking studies with
SOAT1, this group was observed to be surrounded by more
nonpolar amino acid residues like Leu275, Ile301, Ile354 and
Trp394, supporting the observed poor activity of compound (11i).
In SOAT2, the same group was observed to be surrounded by amino
acids Leu275 and Trp394.
and SOAT-2 enzymes could prove to be of immense help for the
development of novel potent SOAT inhibitors.
3. Conclusion
Novel series of vicinal diaryl thiazole- and oxadiazole-based
urea derivatives have been synthesized and evaluated for their
SOAT inhibitory activity. To get structural insights for SOAT inhib-
itory activity, the urea group was attached to different positions of
the scaffold in the synthesized molecules i.e. directly to the azole
ring and then to one of the phenyl rings. All of the synthesized
compounds were evaluated for their in vitro SOAT inhibitory ac-
tivity. Based on the data obtained from the preliminary biological
evaluation it was concluded that the thiazole containing com-
pounds were more potent than the oxadiazoles as exemplified by
compound (12d). As far as the safety of compound (12d) was
This pharmacophore based 3D-QSAR study and the accompa-
nying docking study based on the proposed 3D models of SOAT-1

118
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
concerned, it was found to be devoid of any adrenal or cutaneous
toxicity in an animal model of chronic toxicity. Three dimensional
structures of SOAT-1 and SOAT-2 enzymes were also developed by
using homology modelling technique. Further, pharmacophore and
3D-QSAR models for SOAT inhibitors were also developed by using
the compounds reported in the paper. The developed 3D-QSAR
models were validated using different statistical methods. Docking
study of the most active compound (12d) was performed in the
predicted active sites of both the enzymes. Compound (12d)
showed promising activity worthy of consideration for further
studies for its development as a cholesterol lowering agent.
acetate, 7:3); IR (KBr, cmꢂ1): 3416, 3118, 1720, 1610, 1564, 1505,
1432, 1297, 1198, 829; 1H NMR:
10.71 (bs, 1H, NH), 8.83 (bs, 1H,
NH), 8.07 (s, 1H, Ar-H), 6.82e7.40 (m, 10H, Ar-H); MS: m/z 459.9
[Mþ1]þ; Anal. Calcd for C22H13ClF3N3OS: C, 57.46; H, 2.85; N, 9.14.
Found: C, 57.51; H, 2.76; N, 9.21.
d
4.2.3. 1-Butyl-3-[5-(4-chlorophenyl)-4-(4-methoxyphenyl)thiazol-
2-yl]urea (5v)
Prepared as per the general procedure using compound (4g) and
n-butyl isocyanate to obtain the desired compound (5v).Yield 55%;
mp ¼ 203e205 ꢁC; TLC Rf ¼ 0.58 (n-Hexane: Ethyl acetate, 7:3);
UVmax (MeOH): 250.2 nm; IR (KBr, cmꢂ1): 3403, 2931, 1692, 1609,
4. Experimental
1566,1532,1446,1293,1249,1093, 831; 1H NMR:
d 10.6 (bs,1H, NH),
6.8e7.4 (m, 8H, Ar-H), 6.6 (bs, 1H, NH), 3.75 (s, 3H, OCH3), 3.1 (m,
2H, NHCH2), 1.39 (m, 2H, CH2), 1.25 (m, 2H, CH2), 0.92 (t, 3H, CH3,
4.1. Chemistry
J ¼ 7.2 Hz); MS: m/z 414.97 [Mþ]; 13C NMR:
d 158.67, 158.01, 153.90,
All solvents and chemicals were purified before use. Melting
points were determined on a Veego melting point apparatus and
are uncorrected. The IR spectra were obtained (wave numbers in
cmꢂ1) on Bruker ALPHA-T (Germany) FT-IR spectrophotometer
using KBr discs. 1H NMR and 13C NMR spectra were recorded on a
Bruker Advance-II 400 MHz spectrometer (some samples were run
on 500 MHz spectrometer also) in CDCl3 or DMSO-d6 solvents.
143.87, 131.98, 131.53, 130.57, 129.58, 128.71, 127.15, 120.93, 113.54,
54.93, 31.59, 19.44, 13.64, 13.56; Anal. Calcd for C21H22ClN3O2S: C,
60.64; H, 5.33; N, 10.10. Found: C, 60.76; H, 5.28; N, 10.23.
4.3. Synthesis of compounds (7a-7t)
Chemical shifts are given in
d
values (ppm) using tetramethylsilane
4.3.1. 1-[5-(4-Acetamidophenyl)-4-(4-methylphenyl)thiazol-2-yl]-
3-(2,4-difluorophenyl)urea (7a)
(TMS) as the internal standard; coupling constants (J) are given in
Hz. Signal multiplicities are characterized as s (singlet), d (doublet),
t (triplet), q (quartet) and m (multiplet). Mass spectra were ob-
tained on Thermo Fisher mass spectrometer with EI ion source.
Elemental analyses were recorded on Thermo Fisher FLASH 2000
organic elemental analyzer. Merck F-254 commercial plates were
used for analytical TLC to follow the course of reactions and check
the product purity. Silica gel 60 (Merck, 300e400 mesh) was used
for flash chromatography (CombiFlash RF 200, Teledyne ISCO).
Structures of all the compounds were found to be consistent with
their analytical and spectral data.
To a stirred solution of the amine (6a) (0.25 g, 1.02 mM) in THF
(10 ml), acetic anhydride (0.12 ml, 1.02 mM) and a few drops of
pyridine were added at 0e5 ꢁC. The reaction mixture was stirred for
2 h at room temperature, poured into ice-cold water (50 ml) and
the excess amount of pyridine was neutralized by addition of a few
drops of dil HCl. The precipitate so formed was filtered, washed
with water and dried. The crude product was purified by column
chromatography using n-hexane/EtOAc as the mobile phase to
obtain the pure product (7a). Yield 65%; mp ¼ 260e262 ꢁC; TLC
Rf
245.20 nm; IR (KBr, cmꢂ1): 3415, 3126, 2962, 1702, 1666, 1615, 1554,
1401, 1288, 845; 1H NMR:
9.78 (s, 1H, NH), 8.88 (s, 1H, NH), 8.09 (s,
¼
0.30 (n-Hexane: Ethyl acetate, 7:3); UVmax (MeOH):
4.2. General procedure for the synthesis of 2-thiazolylurea
d
derivatives (5a-5z’)
1H, NH), 7.83 (s, 1H, Ar-H), 7.48e7.50 (m, 2H, Ar-H), 7.26e7.28 (m,
2H, Ar-H), 7.11e7.13 (m, 2H, Ar-H), 6.96e7.05 (m, 2H, Ar-H),
6.90e6.96 (m, 1H, Ar-H), 6.80e6.84 (m, 1H, Ar-H), 2.24 (s, 3H,
COCH3), 2.04 (s, 3H, Ar-CH3); MS: m/z 478.7 [Mþ]; Anal. Calcd for
To a solution of compounds (4a-4h) (0.82 mM, 1 eq) in dry
toluene (10 ml), the required isocyanates (1.60 mM, 2.0 eq) were
added under nitrogen. The reaction mixture was stirred at room
temperature for 4e5 h and monitored by TLC. The solid precipitate
so obtained was filtered and washed with toluene to remove the
excess isocyanate, dried and collected the products (5a-5z’) as such.
C25H20F2N4O2S: C, 62.75; H, 4.21; N, 11.71. Found: C, 62.88; H, 4.07;
N, 11.90.
4.3.2. 1-[5-(4-Acetamidophenyl)-4-(4-methoxyphenyl)thiazol-2-
yl]-3-(2,4-difluorophenyl)urea (7 m)
4.2.1. 1-Butyl-3-[5-(4-chlorophenyl)-4-(4-methylphenyl)thiazol-2-
yl]urea (5o)
Prepared as per the procedure given under (7a). Yield 62%;
mp ¼ 160e162 ꢁC; TLC Rf ¼ 0.17 (n-Hexane: Ethyl acetate, 3:7);
UVmax (MeOH): 253 nm; IR (KBr, cmꢂ1): 3406, 1675, 1610, 1615,
Prepared as per the general procedure using compound (4e) and
n-butyl isocyanate to obtain the desired compound (5o). Yield 31%;
mp ¼ 181e183 ꢁC; TLC Rf ¼ 0.69 (n-Hexane: Ethyl acetate, 7:3);
UVmax (MeOH): 242 nm; IR (KBr, cmꢂ1): 3410, 2929, 2869, 1698,
1548, 1512, 1403, 1293, 1253, 834; 1H NMR:
d
9.87 (bs, 1H), 8.09 (bs,
1H), 6.74e8.15 (m, 11H), 3.73 (s, 3H), 2.04 (s, 3H); Anal. Calcd for
25H20F2N4O3S: C, 60.72; H, 4.08; N, 11.33. Found: C, 60.85; H, 3.95;
C
1534, 1444, 1295, 1241, 1092, 822; 1H NMR:
d
11.34 (bs, 1H, NH),
6.9e7.4 (m, 8H, Ar-H), 2.8e2.9 (m, 2H, NHCH2), 2.35 (s, 3H, Ar-CH3),
1.16e1.25 (m, 4H, -CH2), 0.82e0.90 (m, 3H, CH3); 13C NMR:
160.10,
N, 11.39.
d
153.03, 142.49, 137.40, 132.38, 130.46, 129.68, 129.41, 128.71, 127.99,
127.89, 122.48, 38.98, 30.73, 20.23, 18.19, 12.68; Anal. Calcd for
4.3.3. 1-[5-(4-Acetamidophenyl)-4-(4-methoxyphenyl)thiazol-2-
yl]-3-butylurea (7q)
C
21H22ClN3OS: C, 63.07; H, 5.54; N, 10.51. Found: C, 63.20; H, 5.35;
Prepared as per the procedure given under (7a). Yield 78%;
mp ¼ 168e170 ꢁC; TLC Rf ¼ 0.17 (n-Hexane: Ethyl acetate, 3:7);
UVmax (MeOH): 251 nm; IR (KBr, cmꢂ1): 3405, 3247, 2953, 1668,
N, 10.59.
4.2.2. 1-[5-(4-Chlorophenyl)-4-(4-fluorophenyl)thiazol-2-yl]-3-
(2,4-difluorophenyl)urea (5p)
Prepared as per the general procedure using compound (4f) and
2,4-difluorophenyl isocyanate to obtain the desired compound
(5p).Yield 37%; mp ¼ 205e207 ꢁC; TLC Rf ¼ 0.51 (n-Hexane: Ethyl
1529, 1402, 1248, 832; 1H NMR:
d 7.5 (bs, 1H), 6.72e7.39 (m, 8H, Ar-
H), 3.71 (s, 3H, OCH3), 2.97e2.99 (m, 2H, NHCH2), 2.10 (s, 3H,
COCH3), 1.41e1.26 (m, 4H, CH2), 0.78e0.81 (t, 3H, CH3); MS: m/z
438.8 [Mþ]; Anal. Calcd for C23H26N4O3S: C, 62.99; H, 5.98; N, 12.78.
Found: C, 63.12; H, 5.77; N, 12.91.

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
119
4.4. General procedure for the synthesis of compounds (11a-11l)
4.5.2. 1-Heptyl-3-(4-(2-methyl-4-(4-fluorophenyl)thiazol-5-yl]
phenyl)urea (12k)
To a solution of compounds (10a-10d) (0.3 g, 0.45 mM) in DMF
(5 ml), DMAP (0.26 g, 2.1 mM) and the desired amines (0.15 g,
0.85 mM) were added. The reaction mixture was stirred for 12 h at
room temperature and monitored by TLC. The reaction mixture was
poured into ice-cold water containing dil HCl. The precipitate so
obtained was filtered off and purified by flash chromatography
using dichloromethane: methanol as eluent to yield the desired
derivatives (11a-11l).
Yield 56%; mp ¼ 154e156 ꢁC; TLC Rf ¼ 0.36 (n-Hexane: Ethyl
acetate, 7:3); UVmax (ACN): 248.80 nm; IR (KBr, cmꢂ1): 3320, 2928,
1633, 1585, 1554, 1403, 1222, 836; 1H NMR:
d 7.98 (bs, 1H, NH), 5.75
(s, 1H, NH), 6.88e7.55 (m, 8H, Ar-H), 3.12e3.32 (m, 2H, NHCH2),
2.72 (s, 3H, thiazole-CH3), 1.21e1.67 (m, 10H, CH2), 0.92e0.95 (t, 3H,
CH3); 13C NMR:
d
163.67 (d, J ¼ 80 Hz), 161.01, 156.22, 147.90, 139.25,
132.20, 131.06, 130.70 (d, J ¼ 32 Hz), 130.11, 125.95, 119.47, 115.33 (d,
J ¼ 84 Hz), 40.29, 31.75, 30.15, 29.01, 26.88, 22.57, 19.10, 14.04; Anal.
Calcd for C24H28FN3OS: C, 67.74; H, 6.63; N, 9.87. Found: C, 67.87; H,
6.49; N, 10.14.
4.4.1. 1-(3,4,5-Trimethoxyphenyl)-3-(4-(2-methyl-4-(4-
fluorophenyl)thiazol-5-yl)phenyl)urea) (11e)
Prepared as per the general procedure using compound (10b)
and 3,4,5-trimethoxyaniline to obtain compound (11e). Yield 82%;
mp ¼ 224e226 ꢁC; TLC Rf ¼ 0.50 (n-Hexane: Ethyl acetate, 7:3); IR
(KBr, cmꢂ1): 3268, 2961, 1703, 1600, 1540, 1507, 1410, 1321, 1207,
4.5.3. 1-(2,6-Diethylphenyl)-3-(4-(2-methyl-4-(4-chlorophenyl)
thiazol-5-yl]phenyl)urea (12o)
Yield 43%; mp ¼ 248e250 ꢁC; TLC Rf ¼ 0.72 (n-Hexane: Ethyl
acetate, 8:2); UVmax (MeOH): 259.20 nm; IR (KBr, cmꢂ1): 3285,
1128, 838; 1H NMR:
d 8.89 (s, 1H, NH), 8.75 (s, 1H, NH), 6.80e7.50
2962, 1637, 1588, 1548, 1464, 1311, 1228, 829; 1H NMR:
d
7.03e7.55
(m,11H, Ar-H), 2.88e2.91 (q, 4H, CH2), 1.19e1.23 (t, 6H, CH3), 2.72 (s,
3H, thiazole-CH3); 13C NMR:
163.85, 155.44, 147.85, 138.97, 133.46,
(m, 10H, Ar-H), 3.78 (s, 6H, OCH3), 3.66 (s, 3H, OCH3), 2.71 (s, 3H,
CH3); Anal. Calcd for C26H24FN3O4S: C, 63.27; H, 4.90; N, 8.51.
Found: C, 63.36; H, 4.82; N, 8.64.
d
133.42, 132.62, 130.31, 130.24, 128.50, 126.39, 119.97, 20.04, 19.18,
13.78; MS: m/z 474.8 [Mþ]; Anal. Calcd for C27H24ClN3OS: C, 68.12;
H, 5.51; N, 8.83. Found: C, 67.98; H, 5.72; N, 8.95.
4.4.2. 1-(2-(4-Morpholino)ethyl)-3-(4-(2-methyl-4-(4-
fluorophenyl)thiazol-5-yl)phenyl)urea) (11f)
Prepared as per the general procedure using compound (10b)
and 4-(2-aminoethyl)morpholine to obtain the desired compound
(11f). Yield 92%; mp ¼ 179e181 ꢁC; TLC Rf ¼ 0.59 (Chloroform:
MeOH, 5:1); IR (KBr, cmꢂ1): 3309, 2925, 1636, 1584, 1515, 1405,
4.5.4. 1-Butyl-3-(4-(2-methyl-4-(4-chlorophenyl)thiazol-5-yl]
phenyl)urea (12p)
Yield 63%; mp ¼ 140e142 ꢁC; TLC Rf ¼ 0.35 (n-Hexane: Ethyl
acetate, 8:2); UVmax (ACN): 252.40 nm; IR (KBr, cmꢂ1): 3326, 1636,
1311, 1219, 1115, 836; 1H NMR:
d 6.17 (s, 1H, NH), 6.93e7.49 (m, 8H,
1585, 1560, 1490, 1349, 1229, 837; 1H NMR:
d 8.31 (s, 1H, NH), 5.93
Ar-H), 3.87e3.89 (m, 4H, CH2), 3.50e3.54 (m, 2H, CH2), 2.81e2.84
(m, 6H, CH2), 2.68 (s, 3H, thiazole-CH3); Anal. Calcd for
C23H25FN4O2S: C 62.71, H 5.72, N 12.72. Found: C, 62.89, H 5.69, N
12.87.
(s, 1H, NH), 7.13e7.66 (m, 8H, Ar-H), 3.17e3.24 (m, 2H, NHCH2), 2.72
(s, 3H, thizole-CH3), 1.45e1.50 (m, 2H, CH2), 1.32e1.41 (m, 2H, CH2),
0.92e0.95 (t, 3H, CH3, J ¼ 7.2 Hz); 13C NMR:
d 162.63, 155.15, 146.49,
140.52, 133.36, 132.81, 132.33, 129.79, 129.41, 127.84, 123.45, 117.51,
31.31, 29.05, 26.42, 22.09, 18.70; Anal. Calcd for C21H22ClN3OS: C,
63.07; H, 5.54; N, 10.51. Found: C, 63.29; H, 5.43; N, 10.67.
4.4.3. 1-(2,6-Diisopropylphenyl)-3-(4-(2-methyl-4-(4-
mehoxyphenyl)thiazol-5-yl)phenyl)urea) (11j)
Prepared as per the general procedure using compound (10d)
and 2,6-diisopropylaniline to obtain compound (11j). Yield 83%;
mp ¼ 200e202 ꢁC; TLC Rf ¼ 0.38 (n-Hexane: Ethyl acetate, 3:2); IR
(KBr, cmꢂ1): 3294, 2962, 1651, 1593, 1549, 1401, 1310, 1248, 1177,
4.6. General procedure for the preparation of urea derivatives (16a-
16o)
The urea derivatives (16a-16o) were prepared by reacting the
amino derivatives (15a-15c) with the desired isocyanates by the
procedure as reported for the preparation of compounds (5a-5z’) to
obtain the urea derivatives (16a-16o).
836; 1H NMR:
d 8.81 (s, 1H, NH), 7.55 (s, 1H, NH), 7.40e7.44 (m, 4H,
Ar-H), 7.25e7.12 (m, 5H, Ar-H), 6.83e6.81 (d, 2H, Ar-H), 3.77 (s, 3H,
OCH3), 3.17e3.24 (m, 2H, diisopropyl-CH), 2.68 (s, 3H, CH3),
1.17e1.27 (m, 12H, diisopropyl-CH3); Anal. Calcd for C30H33N3O2S:
C, 72.11; H, 6.66; N, 8.41. Found: C, 72.19; H, 6.52; N, 8.49.
4.6.1. 1-(2,4-Difluorophenyl)-3-(4-(3-fluorophenyl)-1,2,5-
oxadiazol-3yl)phenyl)urea (16f)
4.5. General procedure for the synthesis of compounds (12a-12x)
Yield 68%; mp ¼ 212e214 ꢁC; TLC Rf ¼ 0.32 (n-Hexane: Ethyl
acetate, 8:2); IR (KBr, cmꢂ1): 3301, 3120, 3085, 1654, 1620, 1453,
Compounds (12a-12x) were synthesized by adopting the pro-
cedure as described under compounds (7a-7t) using compounds
(9a-9d) and the required isocyanates.
1426, 1294, 1229, 1102, 848; 1H NMR:
d 9.3 (bs, 1H, NH), 8.4 (bs, 1H,
NH), 8.12e8.18 (m, 1H, Ar-H), 7.56e7.60 (m, 4H, Ar-H), 7.41e7.43 (m,
2H, Ar-H), 7.23e7.27 (m, 2H, Ar-H), 6.89e7.05 (m, 2H, Ar-H); Anal.
Calcd for C21H13F3N4O2: C, 61.47; H, 3.19; N, 13.65. Found: C, 61.76;
H, 3.03; N, 13.73.
4.5.1. 1-Butyl-3-(4-(2-methyl-4-(4-fluorophenyl)thiazol-5-yl]
phenyl)urea (12j)
Yield 63%; mp ¼ 178e180 ꢁC; TLC Rf ¼ 0.35 (n-Hexane: Ethyl
acetate, 7:3); UVmax (ACN): 248 nm; IR (KBr, cmꢂ1): 3314, 3117,
4.6.2. 1-(2,6-Diethylphenyl)-3-(4-(3-fluorophenyl)-1,2,5-
oxadiazol-3yl)phenyl)urea (16g)
2962, 1634, 1585, 1514, 1403, 1310, 1224, 835; 1H NMR:
d 8.32 (s, 1H,
NH), 7.2e7.75 (m, 8H, Ar-H), 5.95 (bs, 1H, NH), 3.16e3.19 (m, 2H,
NHCH2, J ¼ 6.6 Hz), 2.72 (s, 3H, thiazole-CH3), 1.32e1.52 (m, 4H,
Yield 73%; mp ¼ 210e212 ꢁC; TLC Rf ¼ 0.35 (n-Hexane: Ethyl
acetate, 8:2); IR (KBr, cmꢂ1): 3286, 2966, 1639, 1553, 1453, 1402,
CH2), 0.91e0.95 (t, 3H, CH3, J ¼ 7.28); 13C NMR:
d
163.69, 163.51,
1304, 1235, 841; 1H NMR:
d 8.8 (s, 1H, NH), 7.10e7.51 (m, 11H, Ar-H),
155.40, 138.81, 131.99, 131.05, 130.75, 130.67, 130.31, 126.61, 120.03,
115.36, 40.19, 32.16, 20.03,19.17,13.76; Anal. Calcd for C21H22FN3OS:
C, 65.77; H, 5.78; N, 10.96. Found: C, 65.83; H, 5.63; N, 11.09.
7.0 (s, 1H, NH), 2.61e2.67 (q, 4H, CH2), 1.2e1.4 (t, 6H, CH3); Anal.
Calcd for C25H23FN4O2: C, 69.75; H, 5.39; N, 13.02. Found: C, 69.88;
H, 5.24; N, 13.15.

120
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
4.6.3. 1-(3-Heptyl)-3-(4-(3-fluorophenyl)-1,2,5-oxadiazol-3yl)
phenyl)urea (16i)
as a 10% solution in 0.9% NaCl) was administered intraperitoneally.
On Day 2, blood samples were withdrawn to estimate serum tri-
glyceride levels. Results were analyzed to determine fold-changes
in triglyceride levels over the test period of 24 h.
Yield 69%; mp ¼ 136e138 ꢁC; TLC Rf ¼ 0.35 (n-Hexane: Ethyl
acetate, 8:2); IR (KBr, cmꢂ1): 3365, 3132, 2927, 1647, 1570, 1401,
1310, 1255, 1096, 810; 1H NMR:
d
8.41 (bs, 1H, NH), 7.0e7.5 (m, 8H,
Ar-H), 5.91 (bs, 1H, NH), 3.37e3.45 (m, 2H, NHCH2), 1.37e1.45 (m,
10H, CH2), 0.9e1.01 (t, 3H, CH3); 13C NMR:
165.24, 162.74, 155.53,
5.3. Evaluation of the effect of compound (12d) on lipid profile in
animal model of diet-induced atherogenesis
d
155.52, 152.15, 141.43, 130.95, 130.86, 129.66, 121.98, 121.94, 119.46,
119.26, 116.36, 116.14, 40.46, 31.73, 30.08, 28.98, 26.86, 22.56, 14.03;
Anal. Calcd for C22H25FN4O2: C, 66.65; H, 6.36; N, 14.13. Found: C,
66.82; H, 6.28; N, 14.22.
The studies were performed on adult male Wistar rats (12e16
weeks old; 250e300 g). All the animals received drinking water ad
libitum. Atherogenic diet [with or without compound (12d)] was
made available to all the animals ad libitum. Compound (12d) was
admixed in the desired concentration in the atherogenic diet on the
basis of weekly food intake of the animals. Rats were housed in
polypropylene cages with 12 h light/12 h dark cycles with tem-
4.6.4. 1-(3-Dodecyl)-3-(4-(4-fluorophenyl)-1,2,5-oxadiazol-3yl)
phenyl)urea (16j)
Yield 70%; mp ¼ 82e84 ꢁC; TLC Rf ¼ 0.34 (n-Hexane: Ethyl ac-
etate, 7:3); IR (KBr, cmꢂ1): 3298, 2960, 1638, 1572, 1402, 1305, 1242,
perature and humidity maintained at 23
2
ꢁC and 30e60% RH
990, 821; 1H NMR:
d
8.62 (bs, 1H, NH), 7.9 (bs, 1H, NH), 7.2e7.4 (m,
respectively. All experimental protocols described in the present
study were approved by the Institutional Animal Ethics Committee
(IAEC) of Pharmacy Department, The Maharaja Sayajirao University
of Baroda with permission from Committee for the Purpose of
Control and Supervision of Experiments on Animals (CPCSEA; MSU/
PHARM/IAEC/2011e2013/01), Ministry of Environment, Forests and
Climate Change, Government of India.
8H, Ar-H), 3.1e3.3 (m, 2H, NHCH2), 1.3e1.5 (m, 20H, CH2), 0.9e1.02
(t, 3H, CH3); Anal. Calcd for C27H35FN4O2: C, 69.50; H, 7.56; N, 12.01.
Found: C, 69.37; H, 7.72; N, 12.14.
5. Biological activity
5.1. In vitro screening of the test compounds for SOAT inhibition
5.4. Preparation of atherogenic diet
The biological activity of the test compounds for inhibition of
SOAT was performed by the method described by Gandhi et al. [31]
The SOAT assay mixture consisted of potassium phosphate buffer
The composition of the diet was as follows: cholesterol (2%),
coconut oil (1%), cholic acid (1%), propylthiouracil (0.5%), casein
(16.5%), powdered chow and hydroxypropyl cellulose (HPC) as
binder (q.s. to 100%). After accurately weighing all the ingredients
casein, cholic acid and propylthiouracil were ground in a mortar
pestle until a homogenous solid mixture was formed. This dry
mixture was added to a large vessel and mixed with required
amount of powdered chow. Next, cholesterol was dissolved in
appropriate volume of diethyl ether and this ethereal solution was
properly mixed with the powder mass. The ether was allowed to
evaporate for 2e3 h. Once the solvent got evaporated, required
amount of coconut oil was mixed properly. For every kilogram of
diet to be prepared, 800 ml of 1% HPC solution (prepared by stirring
HPC in distilled water for about half an hour) was added at once to
the dry mass and vigorously kneaded to evenly mix the ingredients.
The consistency achieved was such that the lumps formed were
neither crumbled nor too sticky. The lumps were hand cured to
form suitably shaped masses which looked like pelleted chow. The
final pellets were dried overnight in a hot-air oven at 45e50 ꢁC.
This step improves long term storage and prevents growth of mi-
croorganisms in the lumps. The pellets were stored in a perforated
bag at 2e8 ꢁC for not more than 7 days.
(0.1 M), BSA (5 mg/ml), microsomal protein (200
terol solubilized in 45% w/v hydroxypropyl -cyclodextrin (2 mM).
Final reaction volume was made upto 850 l with the help of 0.1 M
mg) and choles-
b
m
potassium phosphate buffer. Vehicle, test compounds or the stan-
dard SOAT inhibitor (avasimibe, sigma) were added at this point
(final volume NMT 10
binding of the inhibitor with the SOAT enzymes. The reaction was
initiated by the addition of 200 M oleoyl CoA and allowed to
ml) and incubated for 15 min to allow proper
m
proceed for 10 min at 37 ꢁC. The reaction was terminated by the
addition of 6 ml of a solution of chloroform-methanol (2:1, v/v) to
the reaction mixture. The biphasic mixture was shaken in a sepa-
rating funnel and both the phases were allowed to separate. The
lower organic phase was collected and evaporated to dryness under
a stream of N2 gas. The residue was then re-dissolved in 500
ml of
chloroform-methanol (2:1, v/v) and a volume of 25 l from this
m
solution was spotted on TLC plates for quantification. Each sample
was applied to the TLC plates at least in triplicate. Activities of the
compounds were determined by regression analysis of the spot
areas determined by HPTLC (Camag, Switzerland) for cholesteryl
oleate.
5.2. Poloxamer-407 induced lipoprotein lipase inhibition assay
5.5. Experimental protocol
Poloxamer-407 induced lipoprotein lipase inhibition was used
to determine the changes in glyceride levels by the test compounds.
Poloxamer-407 was procured as a generous gift sample from
Wockhardt Ltd., Aurangabad, India. Avasimibe was purchased from
Sigma Aldrich, St. Louis, MO, USA. Triglyceride estimation was
performed by using commercial kits as per the manufacturer's
protocol (GPO-PAP method, Coral Clinical Systems, Vapi, India).
Experiments were performed on healthy adult Wistar rats
(20e24 weeks; 250e350 g). Baseline triglyceride levels were esti-
mated in serum samples of all animals using the commercial kit on
Day 0. Vehicle, test compounds (20 mg/kg) or avasimibe (20 mg/kg)
suspended in 0.5% sodium CMC were administered on Day 1 by oral
gavage in fasted animals. Four hours after the administration, li-
poprotein lipase inhibitor Poloxamer-407 (1000 mg/kg), (prepared
The animals were randomized into 3 groups of 6 animals each.
The groups were as follows: 1. Progression Control group received
atherogenic diet only, 2. Treatment Group I received atherogenic
diet containing compound (12d) in such a way that the final daily
dose of 12d was 10 mg/kg, 3. Treatment Group II received athero-
genic diet containing compound (12d) in such a way that the final
daily dose of 12d was 30 mg/kg. The study protocol is represented
pictorially in Fig. 10.
All the blood samples were allowed to clot at room temperature
for 15 min and centrifuged at 800 ꢃ g for 15 min at 25 ꢁC. The
supernatant serum was collected and subjected to biochemical
estimation for total cholesterol, triglycerides and HDL-C using
commercially available diagnostic test kits. Other parameters like
VLDL, LDL-C and atherogenesis index were derived by formulae on

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
121
Fig. 10. Pictorial representation of the study protocol to see the effect of compound (12d) on lipid profile and related parameters in diet induced atherogenesis in rats.
the basis of these values. Towards the terminal stage of the study,
animals were humanely sacrificed after blood sampling and
thoracic aortae and aortic roots were carefully dissected.
5.8. Statistical analysis
For all the studies, n ¼ 6, except for en face staining, where n ¼ 3.
The data were presented in the form of mean SEM. The results
were compared by one-way ANOVA. Bonferroni's multiple com-
parisons were employed as the post hoc tests wherever required. All
the statistical analyses were performed using Graphpad Prism, San
Diego, CA, USA (Ver. 05). Results were considered to be statistically
significant when P < 0.05.
5.6. En face lipid staining
It is a well known fact that the site of atherogenesis is unpre-
dictable along the length of the thoracic aorta hence the thoracic
aortae were subjected to en face staining with Sudan red IV solution
(5 mg/ml in 70% isopropyl alcohol, vortexed and filtered). Thoracic
aortae were isolated and cleaned of peri-adventitious tissue as it
stains heavily with Sudan red IV. Care was taken to prevent damage
to intima as lipid deposits might be dislodged upon mishandling.
The aortae were split from the lumen to give strips. The strips were
pinned to a white plate and immersed in Sudan red IV solution at
37 ꢁC for 45 min. After this maneuver, the samples were destained
for 10 min in 40% isopropyl alcohol. The stained tissue was pho-
tographed using a digital camera and images were analyzed using
NIH (USA) image analysis software, ImageJ (ver. 1.43u, written by
Wayne Rasband). Gross extent of atherosclerosis within the
thoracic aorta was calculated by estimating the percent lesion area
as compared to the total area. Data obtained were subjected to
statistical analysis.
5.9. Computational studies
5.9.1. Development of 3D-models of SOAT receptors
The MBOAT conserved domain of SOAT-1 was constructed by
using I-TASSER [55e57]. Structurally similar proteins' folds were
first identified from the PDB by multiple threading approach
LOMETS. Continuous fragments excised from the PDB templates
were reassembled into full-length models by replica-exchange
Monte Carlo simulations with the threading of unaligned regions
(mainly loops) built by ab initio modeling. Fragment assembly
simulation was performed again starting from the SPICKER (near
native model selection algorithm) cluster centroids, where the
spatial restrains collected from the LOMETS templates and the PDB
structures by TM-align were used to guide the simulations. The
purpose of the second iteration was to remove the steric clash as
well as to refine the global topology of the cluster centroids. The
decoys generated in the second simulations were then clustered
and the lowest energy structures were selected. The final full-
atomic models were obtained by REMO which built the atomic
details from the selected I-TASSER decoys through optimization of
the hydrogen-bonding network.
5.7. Luminal lipid plaque identification
The aortic roots were immediately washed with ice-cold PBS,
cleared of surrounding adventitious fat, carotid connections were
severed and immersed in 10% buffered formalin for cryosectioning.
At the time of sacrifice, liver lobes were also removed to serve as
support while sectioning the aortic roots. A small cavity was
formed into the pith of the liver and one end of the aortic root was
inserted inside the cavity. This whole lobe was now mounted on a
chuck and put inside the sectioning instrument. The tissue was
allowed to freeze for about 15e20 min and then the chuck was
placed near the blade-rotor assembly. The blade was moved by the
rotor to give a slicing motion and the sections were collected on a
slide. The sections easily stick on the slide as soon as the slides are
brought to room temperature due to the proteins in the section. The
sections were stained with Sudan red IV solutions (5 mg/ml in 70%
isopropyl alcohol, vortexed and filtered). The slides were put inside
a coplin jar filled with Sudan red IV solution. The jar was incubated
in a water bath at 37 ꢁC for 45 min. The slides were then destained
for 10 min in 40% isopropyl alcohol, cleaned after removing,
observed at 10ꢃ magnification and photographed.
5.9.2. Development of pharmacophore based 3D-QSAR model and
docking studies
To develop a pharmacophore based 3D-QSAR model of SOAT
inhibitors, a set of 32 SOAT inhibitors reported in this paper were
used. PHASE was used to develop pharmacophore and 3D-QSAR
models. Activity data and a given hypothesis in combination create
a 3D-QSAR model which is able to recognize significant aspects of
the molecular structures that govern the activity. All the 3D
structures were drawn in MAESTRO and were minimized with
default force field. In the beginning, conformations for all the
molecules were searched through a combination of Monte-Carlo
multiple minimum (MCMM)/Low mode (LMOD) with a maximum
number of 250 conformations per structure and 100 steps of
minimization. A pharmaset was defined by setting a threshold
values of 1/logIC50 ꢀ 5.105 for the ‘actives’ and 1/logIC50 < 4.496 for

122
P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
c) B.A. Golomb, M.A. Evans, Statin adverse effects: a review of the literature
and evidence for a mitochondrial mechanism,, Am. J. Cardiovasc Drugs 8
(2008) 373e418;
d) A. Marcello, P. Giovanni, Treating statin-intolerant patients, Diabetes
Metab. Syndr. Obes. 4 (2011) 155e166.
the ‘inactives’. PHASE presents a standard set of six pharmacophore
features, namely hydrogen bond acceptor (A), hydrogen bond
donor (D), hydrophobic (H) group, negatively ionizable (N) group,
positively ionizable (P) group and aromatic ring (R). Form the data
set many hypotheses were developed, each having five pharma-
cophoric features. The generated common pharmacophore hy-
potheses were checked by scoring alignment of the ‘actives’ against
a reference ligand by using default settings for ‘Score Actives’ to
identify the pharmacophores from each box that resulted into the
best alignment of the active ligands. The hypotheses were ranked
according to the survival values for the ‘active’ and ‘inactive’
compounds.
Further, the developed pharmacophore model was studied with
an atom based 3-D-QSAR model. In the atom-based 3D-QSAR, a
molecule is treated as a set of overlapping van der Waals spheres.
The van der Waals models of the aligned training set molecules
were positioned in a regular grid of cubes, with each cube allotted
zero or more ‘bits’ to account for the diverse types of atoms in the
training set that occupy the cube. This illustration gives rise to
binary-valued occupation patterns that can be used as independent
variables to create partial least-squares 3D-QSAR models.
Molecular docking studies were performed using Glide (Schro-
dinger 2009). Glide performs grid-based ligand docking and ex-
plores favorable interactions between one or more small molecules
i.e. ligands, and a larger receptor/enzyme molecule, a protein. Glide
provides three levels of docking precision: high-throughput virtual
screening (HTVS); standard precision (SP) and extra precision (XP).
The ligand structures were built within Maestro using the Build
module and a single low energy conformation search was per-
formed for molecules under study using OPLS_2005 force field at
[11] A.T. Pearson, I. Laurora, H. Chu, S. Kafonek, The lipid treatment assessment
project (L-TAP): a multicenter survey to evaluate the percentages of dyslipi-
demic patients receiving lipid-lowering therapy and achieving low-density
lipoprotein cholesterol goals, Arch. Intern. Med. 160 (2000) 459e467.
[12] T.C. Rideout, S.V. Harding, P.J. Jones, M.Z. Fan, Guar gum and similar soluble
fibers in the regulation of cholesterol metabolism: current understandings
and future research priorities, Vasc. Health Risk Manag. 4 (2008) 1023e1033.
[13] I. Ikeda, Factors affecting intestinal absorption of cholesterol and plant sterols
and stanols, J. Oleo Sci. 64 (2015) 9e18.
[14] K.R. Norum, A.C. Lilljeqvist, P. Helgerud, E.R. Normann, A. Mo, B. Selbekk,
Esterification of cholesterol in human small intestine: the importance of acyl-
CoA: cholesterol acyltransferase, Eur. J. Clin. Invest. 9 (1979) 55e62.
[15] J.G. Heider, C.E. Pickens, L.A. Kelly, Role of acyl CoA:cholesterol acyltransferase
in cholesterol absorption and its inhibition by 57-118 in the rabbit, J. Lipid.
Res. 24 (1983) 1127e1134.
[16] F.J. Field, Intestinal cholesterol esterase: intracellular enzyme or contamina-
tion of cytosol by pancreatic enzymes, J. Lipid. Res. 25 (1984) 389e399.
[17] K.E. Suckling, E.F. Stange, Role of acyl Co-A: cholesterol acyl transferase in
cellular cholesterol metabolism, J. Lipid Res. 26 (1985) 647e671.
[18] P. Helgerud, K. Saarem, K.R. Nonm, Acyl-CoA: cholesterol acyltransferase in
human small intestine: its activity and some properties of enzymic action,
J. Lipid. Res. 22 (1981) 271e277.
[19] G. Vahouny, V. Falwal, C.R. Treadwell, Factor's facilitating cholesterol ab-
sorption from the intestine via lymphatic pathways, Am. J. Physiol. 188 (1957)
742e746.
[20] N.D. Vaziri, K.H. Liang, ACAT inhibition ameliorates proteinuria, hyperlipid-
emia, lecithin-cholesterol acyltransferase, SRB-1 and low-density lipoprotein
receptor deficiencies in nephritic syndrome, Circulation 110 (2004) 419e425.
[21] C. Ta-Yuan, Li. Bo-Liang, C.Y. Chang, Y. Urano, Acyl-coenzyme A: cholesterol
acyltransferases, Am. J. Physiol. Endocrinol. Metab. 297 (2009) E1eE9.
[22] E.F. Largis, C.H. Wang, V.G. Davries, S.A. Schaffer, CL 277,082 : a novel inhibitor
of ACAT-catalyzed cholesterol esterification and cholesterol absorption,
J. Lipid. Res. 30 (1989) 681e690.
[23] T.A. Bell, J.M. Brown, M.J. Graham, K.M. Lemonidis, R.M. Crooke, L.L. Rudel,
Liver-specific inhibition of acyl-coenzyme a: cholesterol acyltransferase 2
with antisense oligonucleotides limits atherosclerosis development in apoli-
poprotein B100-only low-density lipoprotein receptor-/-mice, Arterioscler.
Thromb. Vasc. Biol. 26 (2006) 1814e1820.
€
physiological pH condition using LigPrep module of Schrodinger.
Docking calculations were executed in XP mode with the active site
of the receptor/enzyme structure.
[24] P. Pal, H. Gandhi, R. Giridhar, M.R. Yadav, ACAT inhibitors: the search for novel
cholesterol lowering agents, Mini. Rev. Med. Chem. 13 (2013) 1195e1219.
[25] T. Ohshiro, T. Hiroshi, Acyltransferase inhibitors:
a
patent review
Appendix A. Supplementary data
(2010epresent), Expert Opin. Ther. Pat. 25 (2015) 145e158.
[26] S.E. Nissen, E.M. Tuzcu, H.B. Brewer, I. Sipahi, S.J. Nicholls, P. Ganz,
L.M. Wisniewski, J.J. Hanyok, L.M. Kassalow, Effect of ACAT inhibition on the
progression of coronary atherosclerosis, N. Engl. J. Med. 354 (2006)
1253e1263.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2017.02.038.
[27] M.C. Meuwese, E. Groot, R. Duivenvoorden, M.D. Trip, L. Ose, F.J. Maritz,
D.C. Basart, J.J. Kastelein, R. Habib, M.H. Davidson, A.H. Zwinderman,
L.R. Schwocho, E.A. Stein, ACAT inhibition and progression of carotid
atherosclerosis in patients with familial hypercholesterolemia: the captivate
randomized trial, JAMA 301 (2009) 1131e1139.
[28] D. Vries, V.G. Largis, E.E. Miner, T.G. Shepherd, R.G. Upeslachis, Potential
antiatherosclerotic agents. 4. [(Functionalized-alkyl)amino]benzoic acid ana-
logs of cetaben, J. Med. Chem. 26 (1983) 1411e1421.
References
[1] W.B. Kannel, W.P. Castelli, T. Gordon, P.M. McNamara, Serum cholesterol, li-
poproteins, and the risk of coronary heart disease. The Framingham study,
Ann. Intern. Med. 74 (1971) 1e12.
[2] S.M. Grundy, G.J. Balady, M.H. Criqui, G. Fletcher, P. Greenland, L.F. Hiratzka,
N. Houston-Miller, P. Kris-Etherton, H.M. Krumholz, J. LaRosa, I.S. Ockene,
T.A. Pearson, J. Reed, R. Washington, S.C. Smith, Primary prevention of coro-
nary heart disease: guidance from framingham, Circulation 97 (1998)
1876e1887.
[29] M.P. Giovannoni, V.P. Piaz, B.M. Kwon, M.K. Kim, Y.K. Kim, L. Toma,
D. Barlocco, F. Bernini, M. Canavesi, 5,6-Diphenylpyridazine derivatives as
ACAT inhibitor, J. Med. Chem. 44 (2001) 4292e4295.
[30] a) S.T. Shirude, P. Patel, R. Giridhar, M.R. Yadav, An efficient and time saving
microwave-assisted selenium dioxide oxidation of 1,2-diarylethanones, In-
dian J. Chem. 45B (2006) 1080e1085;
b) R. Ramajayam, R. Giridhar, M.R. Yadav, R. Balaraman, H. Djaballah, D. Shum,
C. Radu, Synthesis, antileukemic and antiplatelet activities of 2,3-diaryl-6,7-
dihydro-5H-1,4-diazepines, Eur. J. Med. Chem. 43 (2008) 2004e2010.
[31] H. Gandhi, P. Pal, R. Giridhar, M.R. Yadav, An HPTLC method for quantification
of cholesteryl esters from human plasma and rat liver microsomes, Biomed.
Chrom. 28 (2014) 583e588.
[32] T. Kimura, Y. Takase, K. Hayashi, H. Tanaka, I. Ohtsuka, T. Saeki, M. Kogushi,
T. Yamada, T. Fujimori, Structure-activity relationship of N-[2-(dimethyla-
mino)-6-[3-(5-methyl-4-phenyl-1H-imidazol-1-yl)propoxy]phenyl]-N'-pen-
tylurea and analogs. Novel potent inhibitors of acyl-CoA: cholesterol O-
acyltransferase with antiatherosclerotic activity, J. Med. Chem. 36 (1993)
1630e1640.
[33] T. Kumazawa, H. Harakawa, H. Fukui, S. Shirakura, E. Ohishi, K. Yamada, N-(1-
phenyl-2-benzimidazolyl)-N0-phenylurea derivatives as potent inhibitors of
acyl coa: cholesterol acyltransferase (ACAT), Bioorg. Med. Chem. Lett. 5 (1995)
1829e1832.
[3] E. Vassiliadis, N. Barascuk, M.A. Karsdal, Atherofibrosis - a unique and com-
mon process of the disease pathogenesis of atherosclerosis and fibrosis -
lessons for biomarker development, Am. J. Transl. Res. 5 (2013) 1e14.
[4] E. Falk, Pathogenesis of atherosclerosis, J. Am. Coll. Cardiol. 47 (2006) C7eC12.
[5] P.R. Kamstrup, A. Tybjaerg-Hansen, R. Steffensen, B.G. Nordestgaard, Geneti-
cally elevated lipoprotein (a) and increased risk of myocardial infarction,
JAMA 301 (2009) 2331e2339.
[6] T.Y. Chang, C.C. Chang, N. Ohgami, Y. Yamauchi, Cholesterol sensing, traf-
ficking, and esterification, Annu. Rev. Cell Dev. Biol. 22 (2006) 129e157.
[7] I. Tabas, Consequences of cellular cholesterol accumulation: basic concepts
and physiological implications, J. Clin. Invest. 110 (2002) 905e911.
[8] T. Chang, B. Li, C.Y. Catherine, Y. Urano, Acyl-coenzyme A: cholesterol acyl-
transferases, Am. J. Physiol. Endocrinol. Metab. 297 (2009) E1eE9.
[9] D.G. Karalis, Intensive lowering of low-density lipoprotein cholesterol levels
for primary prevention of coronary artery disease, Mayo Clin. Proc. 84 (2009)
345e352.
[10] a) J. Shepherd, S.M. Cobbe, I. Ford, C.G. Isles, A.R. Lorimer, P.W. MacFarlane,
J.H. McKillop, C.J. Packard, Prevention of coronary heart disease with pra-
vastatin in men with hypercholesterolemia. West of Scotland Coronary Pre-
vention Study Group, N. Engl. J. Med. 333 (1995) 1301e1307;
b) M.J. Chapman, J.K. Stock, H.N. Ginsberg, PCSK9 inhibitors and cardiovas-
cular disease: heralding a new therapeutic era, Curr. Opin. Lipidol. 26 (2015)
511e520;
[34] B.K. Trivedi, A. Holmes, T.S. Purchase, A.D. Essenburg, K.L. Hamelehle,
B.R. Krause, M.S. Hes, R.L. Stanfield, A series of conformationally and sterically
constrained analogs of N-phenyl-N0-aralkylurea ACAT inhibitors, Bioorg. Med.
Chem. Lett. 5 (1995) 2229e2234.

P. Pal et al. / European Journal of Medicinal Chemistry 130 (2017) 107e123
123
[35] A. Gelain, I. Bettinelli, D. Barlocco, B. Kwon, T. Jeong, K. Im, L. Legnani, L. Toma,
Ureidopyridazine derivatives as acyl- CoA: cholesterol acyltransferase in-
hibitors, Sci. Pharm. 74 (2006) 85e97, 74.
[36] S. Asano, H. Ban, K. Kino, K. Ioriya, M. Muraoka, Novel 1,4-diarylpiperidine-4-
methylureas as anti-hyperlipidemic agents: dual effectors on acyl-CoA:
cholesterol-O-acyltransferase and low-density lipoprotein receptor expres-
sion, Bioorg. Med. Chem. Lett. 19 (2009) 1062e1065.
[37] M.A. Dominick, W.A. Bobrowski, J.R. Macdonald, A.W. Gough, Morphogenesis
of a zone-specific adrenocortical cytotoxicity in Guinea pigs administered PD
132301-2, an inhibitor of acyl-CoA: cholesterol acyltransferase, Toxicol.
Pathol. 21 (1993) 54e62.
[38] H. Tawada, M. Harcourt, N. Kawamura, Novel acyl-CoA: cholesterol acyl-
transferase inhibitors. Synthesis and biological activity of 3-quinolylurea de-
rivatives, J. Med. Chem. 37 (1994) 2079e2084.
triglycerides to HDL-cholesterol predicts extensive coronary disease, Clinics
63 (2008) 427e432.
[48] T. Oshiro, M. Muraoka, K. Ioriya, Pyripyropene A, an AcyleCoenzyme A:
cholesterol acyltransferase 2eselective inhibitor, attenuates hypercholester-
olemia and atherosclerosis in murine models of hyperlipidemia, Arterioscler.
Thromb. Vasc. Biol. 31 (2011) 1108e1115.
[49] K. Lee, S. Cho, J.H. Lee, Synthesis of a novel series of 2-alkylthio substituted
naphthoquinones as potent acyl-CoA: cholesterol acyltransferase (ACAT) in-
hibitors, Eur. J. Med. Chem. 62 (2013) 515e525.
[50] D.G. Robertson, M.A. Breider, M.A. Milad, Preclinical safety evaluation of
avasimibe in beagle dogs: an acat inhibitor with minimal adrenal effects,
Toxicol. Sci. 59 (2001) 324e334.
[51] B.D. Roth, ACAT inhibitors: evolution from cholesterol-absorption inhibitors
to antiatherosclerotic agents, Drug Discov. Today 3 (1998) 19e25.
[52] H. Yagyu, T. Kitamine, J.I. Osuga, Absence of ACAT-1 attenuates atherosclerosis
but causes dry eye and cutaneous xanthomatosis in mice with congenital
hyperlipidemia, J. Biol. Chem. 275 (2000) 21324e21330.
[53] M. Accad, S.J. Smith, D.L. Newland, D.A. Sanan, L.E. King, M.F. Linton, S. Fazio,
R.J. Farese, Massive xanthomatosis and altered composition of atherosclerotic
lesions in hyperlipidemic mice lacking acyl CoA:cholesterol acyltransferase 1,
J. Clin. Invest. 105 (2000) 711e719.
[54] C.W. Joyce, G.S. Shelness, M.A. Davis, R.G. Lee, K. Skinner, R.A. Anderson,
L.L. Rudel, ACAT1 and ACAT2 membrane topology segregates a serine residue
essential for activity to opposite sides of the endoplasmic reticulum mem-
brane, Mol. Biol. Cell. 11 (2000) 3675e3687.
[55] J. Yang, R. Yan, A. Roy, D. Xu, J. Poisson, Y. Zhang, The I-TASSER Suite: protein
structure and function prediction, Nat. Methods 12 (2015) 7e8.
[56] A. Roy, A. Kucukural, Y. Zhang, I-TASSER: a unified platform for automated
protein structure and function prediction, Nat. Protoc. 5 (2010) 725e738.
[57] Y. Zhang, I-TASSER server for protein 3D structure prediction, BMC Bio-
informa. 9 (2008) 40e47.
[39] G. Llaverias, J.C. Launga, M. Algret, Pharmacology of the ACAT inhibitor ava-
simibe (CI-1011), Cardiovasc. Drug Rev. 21 (2003) 33e50.
[40] T.P. Johnston, W.K. Palmer, Mechanism of poloxamer 407-induced hyper-
triglyceridemia in the rat, Biochem. Pharmacol. 46 (1993) 1037e1042.
[41] T.Y. Chang, C.C. Chang, N. Ohgami, Y. Yamauchi, Cholesterol sensing, traf-
ficking, and esterification, Annu. Rev. Cell. Dev. Boil 22 (2006) 129e157.
[42] R.E. Temel, L. Hou, L.L. Rudel, ACAT2 stimulates cholesteryl ester secretion in
apoB-containing lipoproteins, J. Lip. Res. 48 (2007) 1618e1627.
[43] K. Aragane, J. Kusunoki, T. Kitamine, Effects of F-1394, an acyl-CoA: choles-
terol acyltransferase (ACAT) inhibitor, on ACAT activity in HepG2 cells and on
hepatic secretion of lipids in Triton WR-1339-induced hyperlipidemic rats:
possible role of hepatic ACAT in very low density lipoprotein secretion, Jpn. J.
Pharmacol. 76 (1998) 309e312.
[44] T. Uchida, K. Aoyama, T. Watanbe, Relationship between bioavailability and
hypocholesterolemic activity of YM17E, an inhibitor of ACAT, in cholesterol-
fed rats, Atherosclerosis 137 (1998) 97e106.
[45] T.N. Tulenko, A.E. Summer, The physiology of lipoproteins, J. Nucl. Cardiol. 9
(2002) 638e649.
[46] a) J. Kusunoki, K. Aragane, T. Kitamine, Hypocholesterolemic action and pre-
vention of cholesterol absorption via the gut by f-1394, a potent acyl-coa:
Cholesterol acyltransferase (ACAT) inhibitor, in cholesterol diet-fed rats, Jpn.
J. Pharmacol. 69 (1995) 53e60;
b) R. Arakawa, S. Abe-Dohmae, M. asai, J. I. Ito, S. Yokoyama, Involvement of
caveolin-1 in cholesterol enrichment of high density lipoprotein during its
assembly by apolipoprotein and THP-1 cells, J. Lip. Res. 41 (2000) 1952e1962.
[47] a) W.T. Friedewald, R.I. Levy, D.S. Fredrickson, Estimation of the concentration
of low-density lipoprotein cholesterol in plasma, without use of the prepar-
ative ultracentrifuge, Clin. Chem. 18 (1972) 499e502;
[58] S.L. Dixon, A.M. Smondyrev, S.N. Rao, A novel approach to pharmacophore
modeling and 3D database searching,, Chem. Biol. Drug Des. 67 (2006)
370e372.
[59] A.M. Kanhed, V.P. Zambre, V.A. Pawar, M.K. Sharma, R. Giridhar, M.R. Yadav,
Structural requirements for imidazo [1, 2-a] pyrazine derivatives as Aurora A
kinase inhibitors and validation of the model, Med. Chem. Res. 23 (2014)
5215e5223.
[60] A.M. Kanhed, R.C. Dash, N. Parmar, T. Kumar, R. Giridhar, M.R. Yadav, Benzo[e]
pyrimido[5,4-b][1,4]diazepin-6(11H)-one derivatives as Aurora A kinase in-
hibitors: LQTA-QSAR analysis and detailed systematic validation of the
developed model, Mol. Divers 19 (2015) 965e974.
b) P.A. Luz, D. Favarato, J.R.F. Junior, P. Lemos, A.C.P. Chagas, High ratio of