Acyl-CoA:cholesterol O-acyltransferase (ACAT) inhibitors. 2. 2-(1,3- dioxan-2-yl)-4,5-diphenyl-1H-imidazoles as potent inhibitors of ACAT

Article abstract of DOI:10.1021/jm9505876

The second in this series of papers concerns our further investigations into the search for a potent bioavailable acyl-CoA:cholesterol O- acyltransferase (ACAT) inhibitor suitable for the treatment of atherosclerosis. The design, synthesis, and structure-activity relationship for a series of ACAT inhibitors based on the 2-(1,3-dioxan-2-yl)-4,5- diphenyl-1H-imidazole pharmacophore are described. Compounds such as 13a bearing simple alkyl or hydroxymethyl substituents at the 5-position of the 1,3-dioxane ring are potent bioavailable inhibitors of the rat hepatic microsomal enzyme in vitro (IC₅₀ < 100 nM) but are only weak inhibitors of the human hepatic enzyme. We have found however that 1,3-dioxanes substituted at the 5-cis position with pyrazolylalkyl or aminoalkyl groups are potent inhibitors in vitro of human macrophage ACAT, the potency depending on the nature of the terminal heterocycle and the length of the alkyl chain. An ex vive bioassay described herein demonstrates that potent inhibitors such as 13t (IC₅₀ = 10 nM) which contain lipophilic terminal heterocycles do not appear to be systemically available. Less patent but more water soluble compounds such as 13h (IC₅₀ = 60 nM) and 13n (IC₅₀ = 70 nM) are absorbed following oral dosing and achieve plasma levels significantly in excess of their IC₅₀ for ACAT inhibition. These compounds are therefore possible candidates for further investigation as oral antiatherosclerotic agents.

Full text of DOI:10.1021/jm9505876

J . Med. Chem. 1996, 39, 1423
1423
Acyl-CoA:Ch olester ol O-Acyltr a n sfer a se (ACAT) In h ibitor s. 2.
2-(1,3-Dioxa n -2-yl)-4,5-d ip h en yl-1H-im id a zoles a s P oten t In h ibitor s of ACAT
Peter C. Astles,* Michael J . Ashton, Andrew W. Bridge, Neil V. Harris, Terrance W. Hart, David P. Parrott,
Barry Porter, David Riddell, Christopher Smith, and Robert J . Williams
Rhoˆne-Poulenc Rorer, Dagenham Research Centre, Dagenham, Essex RM10 7XS, U.K.
Received August 7, 1995X
The second in this series of papers concerns our further investigations into the search for a
potent bioavailable acyl-CoA:cholesterol O-acyltransferase (ACAT) inhibitor suitable for the
treatment of atherosclerosis. The design, synthesis, and structure-activity relationship for a
series of ACAT inhibitors based on the 2-(1,3-dioxan-2-yl)-4,5-diphenyl-1H-imidazole pharma-
cophore are described. Compounds such as 13a bearing simple alkyl or hydroxymethyl
substituents at the 5-position of the 1,3-dioxane ring are potent bioavailable inhibitors of the
rat hepatic microsomal enzyme in vitro (IC50 < 100 nM) but are only weak inhibitors of the
human hepatic enzyme. We have found however that 1,3-dioxanes substituted at the 5-cis
position with pyrazolylalkyl or aminoalkyl groups are potent inhibitors in vitro of human
macrophage ACAT, the potency depending on the nature of the terminal heterocycle and the
length of the alkyl chain. An ex vivo bioassay described herein demonstrates that potent
inhibitors such as 13t (IC50 ) 10 nM) which contain lipophilic terminal heterocycles do not
appear to be systemically available. Less potent but more water soluble compounds such as
13h (IC50 ) 60 nM) and 13n (IC50 ) 70 nM) are absorbed following oral dosing and achieve
plasma levels significantly in excess of their IC50 for ACAT inhibition. These compounds are
therefore possible candidates for further investigation as oral antiatherosclerotic agents.
In tr od u ction
also be expected to bring about a beneficial hypocho-
lesterolemic response.7
Part 1 in this series identified the 4,5-diphenyl-1H-
imidazole moiety as a potent pharmacophore for in vitro
inhibition of the rat liver acyl-CoA:cholesterol O-acyl-
transferase (ACAT, EC 2.3.1.26) and described moder-
ately lipophilic 2-(alkylthio)-4,5-diphenyl-1H-imidazoles
which were good inhibitors of ACAT in vitro and reduced
plasma cholesterol levels in the cholesterol-fed rat.1
ACAT is responsible for catalyzing the intracellular
esterification of cholesterol by acyl-CoA’s derived from
long chain fatty acids, principally oleic acid, and as such
plays a key role in the absorption and metabolism of
cholesterol.2 It has been postulated that the inhibition
of ACAT is a rational and efficient approach to the
design of novel hypocholesterolemic agents interfering
with the intestinal absorption of cholesterol.3 In addi-
tion, the consequences of ACAT inhibition in the artery
by a systemically available agent might lead to direct
antiantherosclerotic effects. Arterial macrophages pos-
sess a scavenger receptor for oxidatively modified low-
density lipoprotein (LDL) particles.4 The ingested lipid
particles are metabolized via the cholesteryl ester (CE)
cycle to release their burden of cholesterol.5 After
digestion of the LDL particle, lysosomal hydrolysis by
acidic cholesterol ester hydrolase produces free choles-
terol (FC) which is immediately re-esterified by ACAT.
The scavenger receptor is unregulated, and since cells
do not generally excrete CE,4,5 arterial macrophages can
accumulate considerable quantities of lipid. The result-
ing CE-laden foam cells comprise a major component
of the atherosclerotic plaque,6 so the prevention, or
reduction, of the accumulation of CE in macrophages
by the inhibition of ACAT might be beneficial in the
treatment of atherosclerosis. Recent reports have also
implicated ACAT in the release of VLDL particles from
the liver, and inhibition of this hepatic activity might
A number of structurally diverse compounds showing
potent in vitro and in vivo inhibition of ACAT have been
described.8 Many reported inhibitors are lipophilic,
water-insoluble, and unlikely to exhibit significant
bioavailability following oral dosing. These include, for
example: the amide RP 64477 (1),9 the urea CL 227,082
(2),10 and the diphenylimidazole DuP 128 (3)11 which
has been examined in the clinic.12 These compounds
were designed to limit cholesterol absorption by inhibit-
ing ACAT in the intestinal mucosa without being
systemically available. More recently, ACAT inhibitors
such as CI-976 (4),13 CP-105191 (5),14 and PD-132301
(6a )15 possessing moderate systemic bioavailability have
been reported, and a water-soluble ACAT inhibitor has
been synthesized at Parke-Davis.16 Studies carried out
in the DuPont laboratories to obtain a systemically
available ACAT inhibitor based on 3 have also been
described in a recent series of papers.17 In addition to
a hypocholesterolemic benefit, these systemic inhibitors
have the potential to exert a direct antiatherosclerotic
effect on the artery wall. In confirmation of this
hypothesis, it has been reported that 4 inhibits the
progression of atherosclerotic lesions independently of
its effect on plasma cholesterol concentration.18 The
search for bioavailable ACAT inhibitors has recently
been reviewed by Sliscovic.19
Our objective was to develop a novel, potent, systemi-
cally available ACAT inhibitor that might produce a
direct disease-modifying effect by inhibition of the action
of the arterial enzyme in addition to any beneficial hypo-
cholesterolemic response occasioned by the inhibition
of the intestinal and/or hepatic enzymes. Research in
our laboratories identified the 2-(alkylthio)-substituted
4,5-diphenyl-1H-imidazole RP 70676 (7) as a bioavail-
able ACAT inhibitor which exhibited a hypocholester-
olemic effect in the cholesterol-fed rabbit and reduced
X Abstract published in Advance ACS Abstracts, February 15, 1996.
0022-2623/96/1839-1423$12.00/0 © 1996 American Chemical Society

1424 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7
Astles et al.
Ch a r t 1
Sch em e 1. General Synthesis of 1,3-Dioxanes
13a -d ,f,h ,i,k -y and 14d ,fa
a
Method A: (i) 10, 12, PPTS, toluene, reflux; (ii) Na, NH3, THF.
Method B: 11, 12, PTSA, toluene, 100-110 °C.
was as follows: (1) The 4,5-diphenyl-1H-imidazole phar-
macophore was considered essential for ACAT inhibi-
tion1 and would remain unchanged; (2) the effect of
varying the terminal dimethylpyrazole ring to other
polar/basic heterocycles would be examined; and (3) the
sulfoxide linkage would be replaced with a conforma-
tionally less flexible group, ideally heterocyclic, that
might restrain the side chain in its enzyme-bound con-
formation and increase inhibitory activity. This replace-
ment would also remove a potential site of metabolic
deactivation (oxidation to the sulfone level), previous
studies having shown that conversion of the thioether
to a sulfone significantly reduced in vitro activity.1
Id en tifica tion of th e 1,3-Dioxa n e P h a r m a co-
p h or e. To determine the functionality that could be
tolerated at the 2-position, a number of simple 2-sub-
stituted 4,5-diphenyl-1H-imidazoles 9 were tested for
their ability to inhibit the catalytic activity of rat liver
microsomal ACAT (Table 1). It was noted that the
carbocyclic compounds 9a ,b24 possessed a good level of
inhibitory activity; however, these groups were deemed
to be too lipophilic in nature to facilitate oral absorption.
To improve water solubility, some cyclic ethers of
varying ring size (9c-g25-27) were prepared. The
optimum substituents were found to be the six-ring
tetrahydropyran-2-yl and 1,3-dioxan-2-yl groups (com-
pounds 9c,d , respectively, IC50 ∼ 100 nM). It was
decided to initiate a program of work based around the
2-(1,3-dioxan-2-yl)-4,5-diphenyl-1H-imidazole pharma-
cophore26 due to its potency, its synthetic accessibility,
and its symmetrical, hydrophilic properties.
the accumulation of cholesterol and cholesterol ester in
rabbit aorta and thoracic artery.20 This compound was
found to be a potent in vitro inhibitor of ACAT derived
from a number of species and tissues including the
human hepatic enzyme (IC50 ) 40 nM). It was observed
however that 7 is rapidly metabolized in mammals to
the corresponding enantiomeric sulfoxides. Synthesis
and testing of each enantiomer showed that only the
(S)-(-)-isomer RP 73163 (8) was responsible for ACAT
inhibition. When 8 was tested in vitro against the
human THP-1 macrophage enzyme21 however, the level
of inhibition was only moderate (IC50 ) 160-300 nM).
This was disappointing since the human arterial mac-
rophage enzyme was considered the ultimate target for
an oral antiatherosclerotic ACAT inhibitor. The chem-
istry and hypolipidemic properties of this compound
have recently been reported by these laboratories.22,23
This paper outlines some of our investigations in the
search for a novel series of systemic ACAT inhibitors
to act as a backup to RP 73163 (8). The aim was to
obtain a compound which would possess a 5-10-fold
increase in potency over 8 against the human macro-
phage enzyme in vitro while maintaining a comparable
or improved systemic bioavailability in a suitable animal
model. Our strategy for modifying the structure of 8
Ch em istr y
The general synthesis of the 1,3-dioxanes described
in this paper is illustrated in Scheme 1 and their
characterization summarized in Table 2. The key step
in the synthetic route was the formation of the dioxane
ring from the benzyl-protected imidazole-2-carboxalde-
hyde 1026 and a suitably substituted diol, 12 (method
Ta ble 1. Preparation and in Vitro ACAT Inhibition of 2-Substituted 4,5-Diphenyl-1H-imidazoles 9
no.
R
prep
formulaa
C21H16N2
mp (°C)
rat liver ACAT IC50 (nM)b
9a
9b
9c
9d
9e
9f
phenyl
cyclohexyl
c
d
f
g
h
h
i
272-5
227-9
295-7
193-4
248-50
203-5
233-4
260
910
110
100
4800
180
290
C
C
C
C
C
C
21H22eN2
tetrahydropyran-2-yl
1,3-dioxan-2-yl
1,3-dioxolan-2-yl
tetrahydro-1,3-dioxepin-2-yl
furan-2-yl
20H20N2O
19H18N2O2
18H16N2O2
20H20N2O2
19H14N2O
9g
a
b
d
Satisfactory microanalyses were obtained ((0.4%) for C, H, N. ACAT inhibition in vitro. c Commercially available. See ref 24.
e Calcd, 7.33; found, 6.60. f See ref 25. See ref 26. See the Experimental Section. See ref 27.
g
h
i

Dioxanyldiphenylimidazoles as ACAT Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7 1425
Sch em e 2. Synthesis of Dioxane 13ea
a
(a) (HOCH2)2C(CO2Et)2, PTSA, toluene, reflux; (b) (i) LiAlH4,
F igu r e 1. Stereochemical assignment of cis and trans diox-
anes by H NMR.
THF, (ii) Na, NH3, THF.
1
Sch em e 3. Synthesis of Dioxanes 13g,ia
Sch em e 4. Synthesis of 1,3-Diols 12f,h ,i,k -y (R2 )
CH3)a
a
(a) BrCH2C(CH3)(CH2OH)2, PPTS, toluene, reflux. Method
C: (i) RH, NaH, DMF, reflux; (ii) Na, NH3, THF.
a
(a) NaH, THF, Br(CH2)nBr (n ) 3-5), reflux. Method D: (i)
A). Acetal formation was achieved by reaction with an
acid catalyst (pyridinium p-toluenesulfonate or p-tolu-
enesulfonic acid) in an inert solvent (usually toluene)
with azeotropic removal of water. Cleavage of the
1-benzyl group to give the required 1H-imidazole was
carried out with sodium in liquid ammonia.28 Trans-
acetalization of the unprotected dimethylacetal 1126
with diol 12 (method B) was generally used to prepare
compounds in which substituents on the 5-position of
the dioxane ring were considered to be incompatible
with the reductive deprotection step. The 1,3-dioxanes
13e,g,j were prepared by an extension of the above
general methods (Schemes 2 and 3).
In the majority of the cases, the substituents on the
diol 12 were not identical (R1 * R2 and R1 larger than
R2), leading to the formation of a mixture of cis and
trans isomers (13 and 14, respectively). In general, the
cis isomer predominated over the trans with ratios
varying from 2:1 to 10:1. The cis and trans isomers
could routinely be separated by flash chromatography,
the cis isomer invariably being the more mobile com-
ponent of the mixture. In a few cases it proved possible
to separate the isomers by fractional crystallization.
Isomerically pure compounds could be re-equilibrated
to cis/trans mixtures under acid catalysis, but this
required an inert solvent at elevated temperatures
(usually refluxing toluene). The chemical stability of
the 1,3-dioxane ring with respect to acid-mediated
isomerization and hydrolysis was crucial if oral admin-
istration was to be achieved. We believed that rapid
protonation of the imidazole ring in acidic media
precluded further protonation of the dioxane ring oxygen
atoms. Since this is the first step in the mechanism of
acetal hydrolysis,29 the dioxane is rendered resistant to
ring opening.
NaH, THF, 18 or HOCH2Ph, reflux; (ii) LiAlH4, THF, room
temperature. Method E: (i) 19, THF, reflux; (ii) LiAlH4, THF,
reflux.
oxygen lone pairs, the hydroxymethylene protons reso-
nate at 3.86 ppm in the axial position of isomer 13d
and at 3.41 ppm when equatorial as in 14d . Similarly,
the axial methyl protons in 14d resonate 0.45 ppm
further downfield than the equatorial methyl group in
13d (Figure 1).
The diol starting materials 12a ,b,d were commer-
cially available, and 12c was prepared by a literature
method.30 Scheme 4 outlines the synthesis of the diols
12f,h ,i,k -y for which R2 ) CH3. Diethyl methyl
malonate was alkylated with the appropriate dibro-
moalkane to give a bromoalkyl diester intermediate
(17).31 This was then reacted with either (a) an anion
derived from a suitably substituted pyrazole (18) or
alcohol (method D) or (b) a secondary amine (19) in an
inert refluxing solvent (method E). The resultant
diesters were reduced with lithium aluminum hydride
to the required propane-1,3-diol derivatives 12.
Biologica l Resu lts a n d Discu ssion
Compounds 13a -f were tested initially for in vitro
activity against rat liver microsomal ACAT using a
standard radioassay.32 This assay system has been
adopted by many research groups as a primary screen
for the optimization of ACAT inhibition. The inherent
problems associated with handling human tissue dis-
suaded us from using a human enzyme as a routine
screen. The systemic availability of potent inhibitors
was then determined in the rabbit following an oral dose
of 10 mg/kg, since this was the animal model to be used
for in vivo evaluation of promising compounds. The
results are displayed in Table 3 along with data for RP
70676 (7) and RP 73163 (8) for the purposes of com-
parison.
We began by adding simple substituents to the “lead”
compound 9d , the 5-position of the 1,3-dioxane ring
being the obvious starting point since the complication
of optical isomers would not arise. Substitution by small
alkyl groups significantly improved in vitro activity
against the rat liver enzyme, the 5,5-diethyl analogue
13b being 8 times and the spirane 13c 13 times more
active than the parent 9d . These substitutions however
were accompanied by a concomitant decrease in peak
The assignment of stereochemistry of the dioxane ring
was based on the relative chemical shifts in the 1H-NMR
spectra of the groups at the 5-position. The stereo-
chemistry of 13d and 14d , for example, was assigned
using this argument (see Figure 1) and confirmed when
the X-ray structure of 13d was acquired. It was
assumed that the ring adopts a chair conformation with
the bulky imidazole group equatorial. An axial group
at the 5-position should resonate downfield relative to
an equatorial group on account of the 1,3-diaxial inter-
actions it experiences with the lone pairs on the oxygen
atoms. As a result of this deshielding effect by the

1426 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7
Ta ble 2. Preparation of 1,3-Dioxanes 13 and 14
Astles et al.

Dioxanyldiphenylimidazoles as ACAT Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7 1427
Ta ble 2. (Continued)
a
b
Satisfactory microanalyses ((0.4%) were obtained for C, H, N unless otherwise indicated. Recrystallization solvent: P ) pentane,
C ) cyclohexane, EA ) ethyl acetate, H ) heptane. c Calcd, 75.42; found, 75.90. Hydrochloride salt (C, H, N, Cl). e Calcd, 71.98; found,
d
71.5. f See the Experimental Section. C, H, N, S. Trimaleate salt. Calcd, 78.1; found, 77.5.
g
h
i
Ta ble 3. In Vitro ACAT Inhibition and Systemic Availability
of Compounds 13a -f and 14f
Ta ble 4. Activity against ACAT from Various Species/Tissues
for Compounds 13a ,b,f and 14f
rat liver ACAT plasma levelb
ACAT IC50 (nM)a
no.
R1
R2
IC50 (nM)a
Cmax (nM)c
rat
rabbit
7
8
9d
13a
13b
RP 70676
RP 73163
H
25
85
100
75
no.
R1
RP 70676
RP 73163
CH3
R2
liver
liver/artery
human
2100
500
300
7
8
25
85
75
15
90
40/40
75/150
40b
H
240b/300c
430b
CH3
CH2CH3
CH3
CH2CH3
13a CH3
13b CH2CH3
13d CH2OH
14d CH2OH
13f
3500/3700
15
150
CH2CH3
CH3
CH3
20%d/-
1300b
13c -CH2CH2CH2CH2CH2-
4
40
4300/4450
13d
14d
13e
CH2OH
CH2OH
CH2OH
CH3
CH3
CH2OH
90
110
275
8700
8700
5460
110 >8500/>8500
CH3
55
20/25
75c
a
b
ACAT inhibition in vitro. In the rabbit following 10 mg/kg
dose po. c Maximum concentration of compound in the plasma.
14f
CH3
250
20%d/20%d >500c
plasma level in the rabbit, presumably due to the
increased lipophilicity. In an effort to improve water
solubility, the more hydrophilic hydroxymethyl group
was introduced. The two diastereomeric cis and trans
alcohols 13d and 14d were both equiactive with 9d but
more importantly showed peak levels of parent com-
pound of 8700 nM in rabbit plasma. This was in excess
of the value achieved by RP 73163 (8). Reducing
lipophilicity further (compound 13e) led to a marginal
decrease in potency but without the expected increase
in bioavailability. It was encouraging to observe that
the 1,3-dioxane ring was sufficiently chemically and
metabolically stable to survive passage through the gut
and the liver to allow oral administration.
Despite their moderate in vitro potency, the high
plasma levels achieved by these simple dioxanes made
them interesting compounds for evaluation in a choles-
terol-fed rabbit model of atherosclerosis. Prior to in vivo
testing however, compounds 13a ,b and the isomers 13d /
14d were screened in vitro against rabbit liver and
arterial microsomal ACAT.22 The results are presented
in Table 4 along with values for 7 and 8 for the purposes
of comparison. A significant species selectivity was
observed for the dioxanes with those compounds exam-
ined being at least 50 times less active against the rabbit
enzymes than the rat liver enzyme under comparable
assay conditions. To answer the question as to which,
if either, species enzyme was representative of human
ACAT, the dioxanes 13a ,b were screened in vitro
against the human liver microsomal enzyme22 (Table
4). This highlighted a further species difference with
both of these compounds being sufficiently weaker
inhibitors of the human liver than the rat liver enzyme.
These dioxanes should be contrasted to RP 70676 (7)
which inhibits the rat, rabbit, and human liver enzymes
a
ACAT inhibition in vitro. b Hepatic microsomal enzyme. c THP-1
d
macrophage enzyme. Inhibition at 0.1 µg/mL.
at a similar level with an IC50 of 20-40 nM. We
hypothesized therefore that the rat liver enzyme was
particularly sensitive to inhibition by the dioxanyl-
diphenylimidazole moiety but that the pyrazolylalkyl
group present in 7 was required for good in vitro
inhibition of the rabbit and human liver enzymes. This
prompted the synthesis of the isomeric compounds 13f
and 14f which contained a (3,5-dimethylpyrazol-1-yl)-
butyl substituent on the 5-cis and 5-trans positions of
the 1,3-dioxane ring. The cis dioxane 13f was found to
be a potent inhibitor of the rat liver, rabbit liver, and
rabbit arterial enzymes (20-55 nM); the trans isomer
14f however was only moderately active against rat liver
ACAT and virtually inactive against the rabbit enzymes
(Table 4).
In the light of the unexpected species selectivity
exhibited by the dioxanyl compounds, the human THP-1
macrophage enzyme21 was adopted as the tool for
primary in vitro screening of synthetic molecules since
this enzyme was the desired target of an oral antiath-
erosclerotic ACAT inhibitor. A cellular and a microso-
mal preparation of the THP-1 enzyme was examined
with similar levels of ACAT inhibition being found for
standard compounds in both assay systems. It could
be argued that a cellular in vitro assay would be
desirable since it takes into account the ability of a
compound to enter the macrophage as well as its
inhibitory properties. In general it has been our experi-
ence with the diphenylimidazole series of inhibitors that
cellular and microsomal activity correlates well with few

1428 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7
Astles et al.
Ta ble 5. In Vitro ACAT Inhibition and Systemic Availability
of Compounds 13f-y
exceptions. In this instance, a cellular ACAT assay was
found to be most appropriate for routine screening. The
cis dioxane 13f inhibited the THP-1 macrophage enzyme
with an IC50 of 75 nM, whereas the trans isomer 14f
was at least 10-fold less potent (Table 4). Since RP
73163 (8) was only moderately active against this
enzyme with an IC50 of 160-300, it was felt that 13f
was a good starting point for further optimization.
The pyrazolylbutyl group appeared to be crucial for
activity against human enzymes, and so the nature of
this heteroarylalkyl group on the 5-cis position of 13f
was explored by varying the terminal heterocycle and
the connecting chain. An ex vivo bioassay using rat
plasma samples taken 90 min after dosing with test
compound at 10 mg/kg po (see Biological Methods) was
developed to allow routine estimation of systemic bio-
availability. This method would allow us to compare
the systemic bioavailability of test compounds with
RP73163 (8) and obtain a rank order. A very high
plasma level of 7570 nM (25 times the IC50 for THP-1
enzyme inhibition) was measured for 8 in this assay.
Results from in vitro screening and the ex vivo bioassay
are displayed in Table 5. The trans dioxane isomers,
when they could be obtained, were always found to be
10-20 times less potent than the corresponding cis
compounds (data not shown).
Dioxanes 13g-j were synthesized to vary the con-
necting butyl group. A four- or five-carbon atom chain
appeared to be the optimum for linking the terminal
pyrazole to the dioxane ring, 13f,h being equipotent.
The plasma levels of both of these compounds in the ex
vivo bioassay were estimated to be 20-25 times their
IC50 for enzyme inhibition. Attempts to introduce polar
groups into the chain to improve water solubility as in
the ether 13g were deleterious to potency.
The terminal heterocycle could be replaced by a
variety of basic groups, the most active being the
morpholine 13k , the benzylamine 13l, and the arylpi-
perazine 13m . We decided to focus on this terminal
position as the most likely area for optimization of in
vitro potency and/or bioavailability maintaining the
connecting group as butyl to minimize lipophilicity.
The morpholine 13k was found to achieve a high
plasma level of 1450 nM in the bioassay but was only
moderately active against human macrophage ACAT.
The morpholine ring was therefore systematically varied
to other cyclic amines. The thiomorpholine 13n proved
to be an excellent replacement in terms of potency (IC50
) 70 nM) and systemic availability showing a plasma
level of 2280 nM. Oxidation of this compound to the
more polar sulfoxide and sulfone levels abolished activ-
ity (data not shown). The N-methylpiperazine 13o was
poorly active, although reintroduction of an aromatic
ring as in 13p restored inhibitory potency but not
systemic availability. The lipophilic N-arylpiperazine
13m was similarly very poorly bioavailable although
moderately potent in vitro. Introduction of heteroatoms
into the aryl ring (compounds 13q-s) had little effect
on in vitro activity but unfortunately did not lead to an
improvement in systemic availability despite the de-
crease in lipophilicity.
a
b
THP-1 macrophage enzyme. ACAT inhibition in vitro. c In
d
the rat 90 min following 10 mg/kg dose po. Inhibition at 300 nM.
e Not determined.
most potent dioxane ACAT inhibitor obtained against
the human macrophage enzyme in vitro. The plasma
level achieved by this compound was low (∼400 nM)
when compared to that of 8, but this was still 40 times
the IC50 for inhibition of the THP-1 enzyme.
The effect of substituents in the phenyl ring of 13m
was examined, and it was found that H-bond acceptor
groups (F, NO2, OMe) at the para position generally
improved activity. The methoxy derivative 13t was the

Dioxanyldiphenylimidazoles as ACAT Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7 1429
The pyridyl derivative 13u was 4 times more potent
than the parent benzylamine 13l, but compound was
not detectable in rat plasma following oral dosing. It
was thought that in vivo N-demethylation of these
amines might be occurring, but the potentially meta-
bolically more stable tetrahydroisoquinoline 13v was
still poorly bioavailable in the rat.
Phenyl substitution at the terminal position appeared
to be beneficial to in vitro potency; however, introduction
of aromatic groups into the pyrazole ring (compounds
13w ,x) did not produce any significant improvement in
activity in this series and led to poor oral absorption,
as shown in the bioassay. It was thought unusual that
the terminal heterocyclic pharmacophore could encom-
pass a neutral pyrazole molecule and a series of basic
amine derivatives. It was postulated therefore that the
amines were binding to ACAT in their un-ionized forms.
This hypothesis was partially confirmed by the observa-
tion that the benzyloxy derivative 13y and the benzy-
lamine 13l were equipotent against human macrophage
ACAT.
and hypocholesterolemic agent was halted due to its
adrenal toxicity in the dog.22 This adverse side effect
was also observed by scientists at Parke-Davis for PD-
132301 (6a ). The toxicity of 6a was reported to be
related to inhibition of mitochondrial respiration in
adrenal cells19 and may have been overcome through
the synthesis of the methyleneamino analogue 6b which
is devoid of adrenal toxicity.15,19 The higher basicity and
lower lipophilicity of 6b are the only obvious differences
to 6a since they are equipotent in vitro against the
rabbit intestinal enzyme. Whether the adrenal toxicity
of RP 73163 (8) is mechanism or compound related is
unclear at present. We hopt to address this problem
using the 1,3-dioxane series of ACAT inhibitors 13
detailed above since the structural and overall physi-
cochemical differences to 8 may lead to different side
effect profiles. The results of in vivo and toxicological
studies on these dioxane compounds will be the subject
of furture communications from these laboratories.
Exp er im en ta l Section
Biologica l Met h od s. ACAT Assa ys: (1) Micr osom a l
Assa y. ACAT activity was determined using procedures
previously described in the literature.22,32 Inhibitors were
added as solutions in DMSO (final solvent concentration )
0.5%, v/v). For calculation of IC50 values (the concentration
required to inhibit ACAT activity by 50%), each compound was
examined in triplicate at five concentrations.
Con clu sion
In our search for therapeutically useful bioavailable
ACAT inhibitors for the treatment of atherosclerosis,
we have discovered the 2-(1,3-dioxan-2-yl)-4,5-diphenyl-
1H-imidazole system to be a potent pharmacophore for
enzyme inhibition. A series of these dioxanes (13 and
14) were prepared (Table 2) and their biological activity
and oral bioavailability assayed (Tables 3 and 5). Our
goal was to obtain a compound which was 5-10 times
more potent than the sulfoxide RP 73163 (8) against
the human macrophage enzyme with comparable or
improved systemic availability. Compound 13 substi-
tuted at the 5-position of the dioxane ring with simple
alkyl groups were potent inhibitors in vitro (IC50 < 100
nM) of the rat liver microsomal enzyme and possessed
good oral bioavailability in the rabbit but were signifi-
cantly less active in vitro against the human enzyme.
For potent inhibition of the human macrophage enzyme,
a more complex pyrazolylalkyl or aminobutyl substitu-
ent was required at the 5-cis position. The correspond-
ing trans isomers 14 were generally much less active
in vitro. Human ACAT has only recently been purified
to homogeneity, cloned, and expressed and a sequence
obtained.33 Until this analysis is extended to the rat
and rabbit enzymes, the structural origin of this species
selectivity will remain unknown. By varying the nature
of the terminal heterocyclic group, we were able to
obtain extremely potent in vitro inhibitors of human
macrophage ACAT, such as the (4-methoxyphenyl)-
piperazine 13t (IC50 ) 10 nM). In vitro potency however
could only be increased at the expense of oral bioavail-
ability; we were not able to obtain plasma levels in the
rat for the dioxane series following oral dosing compa-
rable to those obtained for RP 73163 (8). For a number
of moderately potent compounds (the pyrazoles 13f,h
and the thiomorpholine 13n ) however, we were able to
demonstrate plasma levels of biologically active com-
pound after oral dosing in the rat significantly in excess
of their IC50 for enzyme inhibition. We hope to use this
series of compounds to further test the hypothesis that
an orally bioavailable ACAT inhibitor will have a direct
disease-modifying effect at the artery wall on the
formation and progression of atherosclerosis.
(2) Cellu la r Assa y. Human THP-1 cells (purchased from
the ECACC at Porton Down) were routinely cultured in RPMI-
1640 medium supplemented with 50 µM mercaptoethanol and
10% (v/v) fetal calf serum. THP-1 cells (1.5 × 106 cells) were
differentiated by incubation for 72 h in 2 mL of medium (as
above) containing 80 nM phorbol 12-myristate 13-acetate
(PMA). The cells were loaded with cholesterol by incubation
for a further 24 h in 1 mL of fresh medium containing 80 nM
PMA and 100 ng/mL acetylated human LDL. ACAT activity
was determined over the final 5 h of the incubation period
following addition of [14C]oleate/albumin complex to give a final
concentration of 200 µM oleate (with a specific activity of 5
µCi/µmol) and 100 µM albumin. Test compounds or vehicle
control (0.2%, v/v, dimethyl sulfoxide, ethanol, or distilled
water) were added for the final 6 h of the 24 h incubation
period. ACAT activity was measured as the incorporation of
[14C]oleate into [14C]cholesteryl oleate per microgram of cell
protein during the 5 h incubation period. Cellular lipids
(including [3H]cholesterol oleate added as an internal stan-
dard) were extracted using the Folch procedure;34 labeled
cholesteryl oleate was separated by high-performance thin
layer chromatography and quantitated by dual label liquid
scintillation counting. ACAT activity was calculated as na-
nomoles of [14C]oleate incorporated into [14C]cholesteryl oleate
per microgram of cell protein. Results were calculated as the
mean ( SEM of four experiments. IC50 values, the concentra-
tion of compound required to inhibit ACAT activity by 50%,
were calculated from their inhibition curves.
Or a l Bioa va ila bility: (1) Ra bbit. Three male NZW
rabbits were dosed with the test compound at 10 mg/kg po.
Blood samples were obtained predose and at intervals up to
24 h. The corresponding plasma samples were assayed
following solvent extraction by reverse phase HPLC with UV
detection.
(2) Ra t. The microsomal assay was also used to determine
systemic bioavailability by measuring ACAT inhibitory activity
in plasma obtained from treated rats.22 Compounds were
suspended in 1% sodium carboxymethyl cellulose and 0.2%
Tween and administered orally (2 mL/kg) to male Sprague-
Dawley rats at a dose of 10 mg/kg. Animals were sacrificed
after 90 min and blood samples obtained. Aliquots of plasma
(5-20 µL) were preincubated with rat hepatic microsomes in
phosphate buffer (pH 7.4) for 20 min prior to the addition to
radiolabeled oleolyl CoA. The ACAT activity of the plasma
samples was determined as above. Concentration-response
curves were obtained by incubating microsomes with various
During the course of this research, the development
of the sulfoxide RP 73163 (8) as an antiatherosclerotic

1430 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7
Astles et al.
concentrations of test compound (dissolved in DMSO) in the
presence of the appropriate volume of plasma from vehicle-
treated rats using the above conditions. These standard
curves were used to estimate the concentration of compound
in the plasma sample.
and concentrated. The residue was subjected to flash chro-
matography (ethyl acetate-cyclohexane, 1:1) to give 1-benzyl-
(r)-2-(4,5-diphenylimidazol-2-yl)-5-cis-(hydroxymethyl)-5-methyl-
1,3-dioxane as a white solid (15 g, 20%) after recrystallization
from cyclohexane: mp 224-6 °C.
Ch em ica l Meth od s. Reagents, starting materials, and
solvents were purchased from common commercial suppliers
and used as received or distilled from the appropriate drying
agent. All organic solutions were dried over magnesium
sulfate and concentrated at reduced pressure under aspirator
vacuum using a Buchi rotary evaporator. Reaction products
were purified, when necessary, by flash chromatography on
silica gel (40-63 µm) with the solvent system indicated. Yields
are not optimized. Melting points were determined on a
Gallenkamp 595 apparatus and are uncorrected. 1H-NMR
spectra were acquired on Varian VXR-400 and XL-200 spec-
trometers; peak positions are reported in parts per million
relative to internal tetramethylsilane on the δ scale. Elemen-
tal analyses were performed by the Analytical Department at
Rhone-Poulenc Rorer. The structure and purity of all com-
pounds were confirmed by microanalytical and/or spectroscopic
methods. Satisfactory microanalyses ((0.4%) were obtained
for C, H, N unless otherwise stated.
2-(1,3-Dioxola n -2-yl)-4,5-d ip h en yl-1H-im id a zole (9e). A
mixture of 11 (1.24 g, 5 mmol), ethylene glycol (5 mL),
p-toluenesulfonic acid monohydrate (0.12 g, 0.6 mmol), and
toluene (120 mL) was heated at reflux for 18 h with azeotropic
removal of water. The solid which separated on cooling was
collected and purified by flash chromatography (dichlo-
romethane-methanol, 98:2) on silica gel to give 9e as a white
solid (0.4 g, 27%) after recrystallization from acetonitrile: mp
248-50 °C; 1H NMR (200 MHz) (CDCl3) 4.06 (m, 2H), 4.24
(m, 2H), 5.98 (s, 1H), 7.28-7.50 (m, 10H). Anal. (C18H16N2O2)
C, H, N.
2-(4,5,6,7-Tet r a h yd r o-1,3-d ioxep in -2-yl)-4,5-d ip h en yl-
1H-im id a zole (9f). A mixture of 10 (10 g, 29.5 mmol),
2-butene-1,4-diol (17 g), pyridinium p-toluenesulfonate (0.5 g),
and toluene (250 mL) was heated at reflux for 18 h with
azeotropic removal of water. The reaction mixture was cooled
and washed with water (3 × 100 mL) and the organic layer
dried and concentrated. The residue was recrystallized from
ethyl acetate to give 1-benzyl-2-(4,7-dihydro-1,3-dioxepin-2-yl)-
4,5-diphenylimidazole as a white solid (10 g, 83%): mp 168-
70 °C.
A portion (5.0 g, 12 mmol) of this material was stirred at
room temperature in ethyl acetate (300 mL) with 5% palladium
on carbon (0.8 g) under an atmosphere of hydrogen. After 30
min, the reaction mixture was filtered through a pad of
diatomaceous earth and concentrated. The residue was re-
crystallized from ethyl acetate to give 1-benzyl-2-(4,5,6,7-
tetrahydro-1,3-dioxepin-2-yl)-4,5-diphenylimidazole as a white
solid (2.5 g, 50%): mp 149-51 °C. Anal. (C27H26N2O2) C, H,
N.
This material was dissolved in dry tetrahydrofuran (200
mL), and approximately 400 mL of liquid ammonia was
condensed into the flask. Sodium metal (0.63 g, 27 mmol) was
added portionwise over 15 min to the stirred, refluxing reaction
mixture. The reaction mixture was quenched with solid
ammonium chloride (6.0 g) and the ammonia allowed to
evaporate. The residue was partitioned between saturated
aqueous ammonium chloride (200 mL) and ethyl acetate (200
mL) and the aqueous layer extracted with a further 3 × 200
mL of ethyl acetate. The combined organics were washed with
water (4 × 200 mL), dried, and concentrated. The residue was
recrystallized from toluene to give 9f (1.6 g, 94%) as a colorless
solid: mp 203-5 °C; 1H NMR (200 MHz) (CDCl3) 1.84 (m, 4H),
3.96 (m, 4H), 5.84 (s, (1H), 7.26-7.50 (m, 10H). Anal.
(C20H20N2O2) C, H, N.
(r )-2-(4,5-Dip h en yl-1H -im id a zol-2-yl)-5-cis-(h yd r oxy-
m et h yl)-5-m et h yl-1,3-d ioxa n e (13d ) a n d (r )-2-(4,5-Di-
ph en yl-1H-im idazol-2-yl)-5-tr a n s-(h ydr oxym eth yl)-5-m eth -
yl-1,3-d ioxa n e (14d ). Meth od A. A mixture of 10 (67.7 g,
174 mmol), 1,1,1-tris(hydroxymethyl)ethane (120 g), pyri-
dinium p-toluenesulfonate (2.0 g), and toluene (500 mL) was
heated at reflux for 18 h with azeotropic removal of water.
The reaction mixture was cooled and washed with water (4 ×
100 mL) and brine (3 × 50 mL) and the organic layer dried
There was then eluted 1-benzyl-(r)-2-(4,5-diphenylimidazol-
2-yl)-5-trans-(hydroxymethyl)-5-methyl-1,3-dioxane which was
obtained as a fluffy white solid (1 g) after recrystallization from
cyclohexane: mp 198-200 °C.
The cis isomer (3.0 g, 6.8 mmol) was dissolved in dry
tetrahydrofuran (50 mL), and approximately 200 mL of liquid
ammonia was condensed into the flask. Sodium metal (0.62
g, 27 mmol) was added portionwise to the stirred, refluxing
reaction mixture over 40 min. The reaction was quenched with
8.0 g of solid ammonium chloride and the ammonia allowed
to evaporate. The residue was partitioned between dichlo-
romethane (250 mL) and brine (100 mL) and the organic layer
washed with brine (100 mL) and water (100 mL). The organic
layer was dried and concentrated to leave a residue which was
recrystallized from cyclohexane to give 13d (1.8 g, 76%) as a
white solid: mp 205-7 °C; 1H NMR (200 MHz) (CDCl3) 0.8 (s,
3H), 3.62 (d, J ) 12 Hz, 2H), 3.86 (s, 2H), 4.07 (d, J ) 12 Hz,
2H), 5.66 (s, 1H), 7.28-7.56 (m, 10H). Anal. (C21H22N2O3) C,
H, N.
The trans isomer was debenzylated in a similar manner to
give 14 as a white solid after recrystallization from cyclohex-
ane: mp 208-10 °C; 1H NMR (200 MHz) (CDCl3) 1.25 (s, 3H),
3.41 (s, 2H), 3.90 (m, 4H), 5.65 (s, 1H), 7.26-7.50 (m, 10H).
Anal. (C21H22N2O3) C, H. N.
1-Ben zyl-2-[5,5-b is(et h oxyca r b on yl)-1,3-d ioxa n -2-yl]-
4,5-d ip h en ylim id a zole (15). A mixture of 10 (16.8 g, 50
mmol), diethyl bis(hydroxymethyl)malonate (11.0 g, 50 mmol),
and pyridinium p-toluenesulfonate (0.5 g) was heated at reflux
for 18 h with azeotropic removal of water. A further equivalent
of diethyl bis(hydroxymethyl)malonate was added and reflux
continued for 18 h. The reaction mixture was cooled and
washed with water (3 × 100 mL) and brine before being dried
and concentrated. The residue was purified by flash chroma-
tography on silica gel (cyclohexane-ethyl acetate, 2:1) to give
15 as a colorless solid (12 g, 46%) after recrystallization from
cyclohexane: mp 129-31 °C; 1H NMR (400 MHz) (CDCl3) 1.01
(t, J ) 8 Hz, 3H), 1.25 (t, J ) 8 Hz, 3H), 3.64 (d, J ) 12 Hz,
2H), 3.93 (q, J ) 8 Hz, 2H), 4.13 (d, J ) 12 Hz, 2H), 4.18 (q,
J ) 8 Hz, 2H), 5.21 (s, 2H), 5.60 (s, 1H), 6.82-7.43 (m, 15H).
Anal. (C32H32N2O6) C, H. N.
2-[5,5-Bis(h yd r oxym eth yl)-1,3-d ioxa n -2-yl]-4,5-d ip h en -
yl-1H-im id a zole (13e). To a stirred solution of 15 (6.0 g, 10.3
mmol) in dry tetrahydrofuran (200 mL) under an atmosphere
of nitrogen was added a 1.0 M solution of lithium aluminum
hydride in tetrahydrofuran (16 mL, 16 mmol) dropwise over
20 min. The reaction mixture was stirred at ambient tem-
perature for 2 h and the reaction quenched by the dropwise
addition of 3% aqueous sodium hydroxide (8 mL). The
resultant precipitate was removed by filtration through di-
atomaceous earth. The filtrate was concentrated and the
resultant residue recrystallized from ethyl acetate to give
1-benzyl-2-[5,5-bis(hydroxymethyl)-1,3-dioxan-2-yl]-4,5-di-
phenylimidazole (3.45 g, 73%) as a colorless solid: mp 186-8
°C. Anal. (C28H28N2O4) C, H, N.
This material (2.9 g, 6.4 mmol) was dissolved in dry
tetrahydrofuran (20 mL), and approximately 300 mL of liquid
ammonia was condensed into the flask. Sodium metal (0.68
g, 30 mmol) was added portionwise to the stirred, refluxing
reaction mixture over 50 min. The reaction was quenched with
6.0 g of solid ammonium chloride and the ammonia allowed
to evaporate. Saturated aqueous ammonium chloride solution
(20 mL) was added and the mixture extracted with dichlo-
romethane (4 × 100 mL). The combined organic extracts were
washed with water (100 mL) and brine (100 mL) before drying
and concentration. The residue was recrystallized from cy-
clohexane to give 13e (1.5 g, 65%) as a colorless solid: mp 228-
1
30 °C; H NMR (200 MHz) (DMSO) 3.32 (s, 2H), 3.41 (s, 2H),
3.90 (m, 4H), 5.65 (s, 1H), 7.26-7.50 (m, 10H). Anal.
(C21H22N2O4) C, H, N.
(r )-2-(4,5-Dip h en yl-1H-im id a zol-2-yl)-5-m eth yl-5-cis-[4-
(th iom or p h olin -1-yl)bu t-1-yl]-1,3-d ioxa n e (13n ). Meth od
B. A mixture of 11 (2.94 g, 10 mmol), 12n (2.97 g, 12 mmol),

Dioxanyldiphenylimidazoles as ACAT Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7 1431
and p-toluenesulfonic acid (2.28 g, 12 mmol) in toluene (150
mL) was heated at 110 °C for 30 min. The reaction mixture
cooled, diluted with ethyl acetate (100 mL), and treated with
200 mL of 10% aqueous potassium carbonate before stirring
a further 20 min at ambient temperature. The organic layer
was separated, washed with water (250 mL), dried, and
concentrated. The residue was purified by flash chromatog-
raphy on silica gel (dichloromethane-methanol, 19:1) to give
the cis isomer 13n as a white solid (1.5 g, 31%) after
mixture stirred at reflux for 8 h. The reaction mixture was
concentrated and partitioned between diethyl ether (300 mL)
and water (150 mL). The organic layer was washed with water
(2 × 100 mL), dried, and concentrated. The residue was
purified by flash chromatography (diethyl ether-pentane, 1:1)
to give diethyl [4-(3,5-dimethylpyrazol-1-yl)but-1-yl]methyl-
malonate as a colorless oil (28.8 g, 52%). Anal. (C17H24N2O4)
C, H, N.
This material (28.8 g, 90 mmol) was dissolved in dry
tetrahydrofuran (50 mL) and added dropwise to a stirred 1.0
M solution of lithium aluminum hydride (110 mL, 110 mmol).
The reaction mixture was stirred at ambient temperature for
18 h and then cooled to 0 °C. A 3% aqueous sodium hydroxide
solution (22 mL) was added dropwise and the mixture stirred
for 15 min before being filtered through a pad of diatomaceous
earth. The filtrate was dried and concentrated to leave 12f
as an oil (21.2 g, 98%): 1H NMR (400 MHz) (CDCl3) 0.72 (s,
3H), 1.29 (m, 2H), 1.62 (m, 2H), 1.78 (m, 2H), 2.22 (s, 3H),
2.28 (s, 3H), 3.56 (m, 4H), 4.12 (t, J ) 6 Hz, 2H), 5.83 (s, 1H).
2-[4-(Th iom or p h olin -1-yl)b u t -1-yl]-2-m et h ylp r op a n e-
1,3-d iol (12n ). Meth od E. A mixture of thiomorpholine (21
g, 200 mmol), 17 (n ) 4) (32.7 g, 100 mmol), and tetrahydro-
furan (200 mL) was heated at reflux for 18 h before being
cooled and concentrated. The residue was partitioned between
ethyl acetate (250 mL) and 5% aqueous potassium carbonate
(250 mL) and the organic layer dried and concentrated. The
residue was purified by fractional distillation to give diethyl
[4-(thiomorpholin-1-yl)but-1-yl]methylmalonate (24.2 g, 72%)
as a yellow oil: bp 150-2 °C/0.2 Torr.
This material (24.2 g, 73 mmol) was dissolved in dry
tetrahydrofuran (20 mL) and added dropwise to a solution of
lithium aluminum hydride (3.88 g, 102 mmol) in tetrahydro-
furan (350 mL). The reaction mixture was heated to reflux
for 4 h and cooled to 0 °C and the reaction quenched dropwise
with 3% aqueous sodium hydroxide (14 mL). Stirring was
continued for 1 h at ambient temperature and the reaction
mixture filtered through a pad of diatomaceous earth. The
filtrate was dried and concentrated to leave 12n as a white
solid (15.5 g, 86%): mp 54-6 °C; 1H NMR (400 MHz) (CDCl3)
0.78 (s, 3H), 1.60 (br m, 4H), 1.51 (m, 2H), 2.40 (t, J ) 6 Hz,
2H), 2.70 (br s, 8H), 3.51 (s, 4H), 3.65, (br m, 2H).
1
recrystallization from heptane: mp 144-6 °C; H NMR (400
MHz) (CDCl3) 0.73 (s, 3H), 1.30 (m, 2H), 1.55 (m, 2H), 1.72
(m, 2H), 2.44 (br t, J ) 6 Hz, 2H), 2.72 (br m, 8H), 3.63, (d, J
) 12 Hz, 2H), 5.66 (s, 1H), 7.25-7.65 (br m, 10H). Anal.
(C28H35N2O2S) C, H, N, S.
1-Ben zyl-(r )-2-(4,5-d ip h en ylim id a zol-2-yl)-5-cis-(b r o-
m om eth yl)-5-m eth yl-1,3-d ioxa n e (16). A mixture of 10
(33.8 g, 100 mmol), 2-(bromomethyl)-2-methylpropane-1,3-
diol35 (27.5 g, 150 mmol), pyridinium p-toluenesulfonate (2.0
g), and toluene (500 mL) was heated at reflux overnight. The
reaction mixture was cooled, washed with water (3 × 250 mL),
dried, and concentrated. The residue was recrystallized from
2-propanol to give 16 as a white solid (33.3 g, 66%): mp 141
°C; 1H NMR (400 MHz) (CDCl3) 0.82 (s, 3H), 3.57 (s, 2H), 3.70
(d, J ) 12 Hz, 2H), 4.09 (d, J ) 12 Hz, 2H), 5.28 (s, 2H), 5.78
(s, 1H), 6.90-7.47 (m, 15H). Anal. (C28H27BrN2O2) C, H, N.
(r )-2-(4,5-Dip h e n yl-1H -im id a zol-2-yl)-5-cis-[(3,5-d i-
m eth ylp yr a zol-1-yl)m eth yl]-5-m eth yl-1,3-d ioxa n e (13j).
Meth od C. To a stirred suspension of sodium hydride (1.76
g of a 60% dispersion, w/w, in mineral oil, 44 mmol) in dry
dimethylformamide (250 mL) was added 3,5-dimethylpyrazole
(4.23 g, 44 mmol) in small portions. After stirring at ambient
temperature for 30 min, 16 (20.1 g, 40 mmol) was added in
one portion and the reaction mixture stirred at reflux for 45
min. After cooling, the mixture was concentrated and water
(500 mL) added. The resultant solid was collected and
recrystallized from ethanol to give 1-benzyl-(r)-2-(4,5-diphen-
ylimidazol-2-yl)-5-cis-[(3,5-dimethylpyrazol-1-yl)methyl]-5-meth-
yl-1,3-dioxane as white solid (13.0 g, 63%): mp 214-6 °C. Anal.
(C33H34N4O2) C, H, N.
This material (10 g, 19.3 mmol) was dissolved in tetrahy-
drofuran (50 mL) and approximately 250 mL of liquid am-
monia condensed into the flask. Sodium metal was added to
the stirring, refluxing reaction mixture until TLC showed that
no starting material was present. Solid ammonium chloride
(5 g) was added and the ammonia allowed to evaporate. The
residue was partitioned between dichloromethane (100 mL)
and water (100 mL) and the organic layer dried and concen-
trated. The residue was recrystallized from toluene/cyclohex-
ane to give 13j as a white powder (2.88 g, 35%: mp 173-4 °C;
1H NMR (400 MHz) (CDCl3) 0.8 (s, 3H), 2.22 (s, 3H), 2.27 (s,
3H), 3.68 (d, J ) 12 Hz, 2H), 4.05 (d, J ) 12 Hz, 2H), 4.30 (s,
2H), 5.72 (s, 1H), 5.80 (s, 1H), 7.18-7.60 (m, 10H). Anal.
(C26H28N4O2) C, H, N.
1-(Th ia zol-2-yl)p ip er a zin e (19r ). A mixture of 2-bro-
mothiazole (19.9 g, 121 mmol), piperazine (20.9 g, 243 mmol),
and n-butanol (400 mL) was heated at reflux for 18 h. The
reaction mixture was cooled, concentrated, and treated with
150 mL of 10% aqueous potassium carbonate. The mixture
was extracted with ethyl acetate (4 × 250 mL), and the
combined extracts were dried and concentrated to leave 19 as
a brown oil (17.9 g, 88%) which was used without further
purification: 1H NMR (400 MHz) (CDCl3) 2.99 (m, 4H), 3.47
(m, 4H), 6.58 (d, J ) 3 Hz, 1H), 7.19 (d, J ) 3 Hz, 1H).
All other heterocycles 18 and 19 were obtained from
Dieth yl (4-Br om obu tyl)m eth ylm a lon a te (17, n ) 4). A
solution of diethyl methylmalonate (87 g, 500 mmol) in dry
tetrahydrofuran (700 mL) was treated portionwise with so-
dium hydride (19.2 g of a 60% dispersion, w/w, in mineral oil,
500 mmol) with cooling. The reaction mixture was stirred for
30 min at ambient temperature and treated with 1,4-dibro-
mobutane (162 g, 750 mmol). The mixture was heated to
reflux for 3 h and cooled before being concentrated to dryness.
The residue was partitioned between diethyl ether (500 mL)
and water (500 mL) and the organic layer dried and concen-
trated. The residue was purified by fractional distillation to
give 17 (n ) 4) as a colorless oil (98 g, 63%): bp 104-16 °C/
0.2 mbar; 1H NMR (400 MHz) (CDCl3) 1.23 (m, 2H), 1.25 (t, J
) 8 Hz, 6H), 1.43 (s, 3H), 1.88 (m, 4H), 3.41 (t, J ) 6 Hz, 2H),
4.18 (q, J ) 8 Hz, 4H).
The bromo malonates 17 (n ) 3 and 5) were prepared in a
similar manner from the appropriate dibromoalkane.
2-[4-(3,5-Dim et h ylp yr a zol-1-yl)bu t -1-yl]-2-m et h ylp r o-
p a n e-1,3-d iol (12f). Meth od D. To a stirred suspension of
sodium hydride (6 g of a 60% dispersion, w/w, in mineral oil,
150 mmol) in dry tetrahydrofuran (250 mL) was added 3,5-
dimethylpyrazole (14.4 g, 150 mmol) in small portions. After
stirring at ambient temperature for 15 min, 17 (n ) 4) (50.7
g, 170 mmol) was added in one portion and the reaction
commercial sources.
Ack n ow led gm en t. We thank Adnan Al-Shaar, J a-
net Archer, Colin Bright, Brenda Burton, Devandan
Chatterjee, David Neighbour, and Rajesh Patel for
expert technical assistance.
Refer en ces
(1) Harris, N. V.; Smith, C.; Ashton, M. J .; Bridge, A. W.; Bush, R.
C.; Coffee, E. C. J .; Dron, D. I.; Harper, M. F.; Lythgoe, D. J .;
Newton, C. G.; Riddell, D. Acyl-CoA:cholesterol O-Acyl Trans-
ferase (ACAT) Inhibitors. 1. 2-(Alkylthio)-4,5-diphenyl-1H-
imidazoles as Potent inhibitors of ACAT. J . Med. Chem. 1992,
35, 4384-4392.
(2) Billheimer, J . T.; Gillies, P. J . Intracellular Cholesterol Esteri-
fication. In Advances in Cholesterol Research; Esfahani, M.,
Swaney, J . B., Eds.; Telford Press: Caldwell, NJ , 1990; pp 7-45.
(3) Sliskovic, D. R.; White, A. D. Therapeutic Potential of ACAT
Inhibitors as Lipid Lowering and Anti-Athersclerotic Agents.
Trends Pharmacol. Sci. 1991, 12, 194-199. O’Brien, P. M.;
Sliskovic, D. R. ACAT Inhibitors: A Potential New Approach to
the Treatment of Hypercholesterolaemia and Atherosclerosis.
Curr. Opin. Ther. Patents 1992, 2, 507-526.
(4) Brown, M. S.; Goldstein, J . L. Lipoprotein Metabolism in the
Macrophage. Annu. Rev. Biochem. 1983, 52, 223-261.

1432 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 7
Astles et al.
(5) Glick, J . M. Intracellular Cholesteryl Ester Hydrolysis and
Clearance in Cholesterol Research. In Advances in Cholesterol
Research; Esfahani, M., Swaney, J . B., Eds.; Telford Press:
Caldwell, NJ , 1990; pp 171-200.
(6) Gillies, P. J .; Robinson, C. S.; Rathgeb, K. A. Regulation of ACAT
Activity by a Cholesterol Substrate Pool During the Progression
and Regression of Atherosclerosis: Implications for Drug Dis-
covery. Atherosclerosis 1990, 83, 177-185.
(7) Krause, B. R.; Anderson, M.; Bisgaier, C. L.; Bocan, T.; Bousley,
R.; DeHart, P.; Essenburg, A.; Hamelehle, K.; Homan, R.; Kieft,
K.; McNally, W.; Stanfield, R.; Newton, R. S. In vivo Evidence
that the Lipid-Regulating Activity of the ACAT inhibitor CI-
976 in Rats is Due to Inhibition of Both Intestinal and Liver
ACAT. J . Lipid Res. 1993, 34, 279-294.
(8) Roark, W. H.; Roth, B. D. ACAT Inhibitors: Preclinical Profiles
of Clinical Candidates. Exp. Opin. Invest. Drugs 1994, 3, 1143-
1152.
(9) Parrott, D. P.; Bright, C. P.; Burton, B.; Riddell, D.; Sims, A. F.
G. RP 64477: a Potent Inhibitor of ACAT in Human Cells.
Abstracts of the 11th International Symposium on Drugs Af-
fecting Lipid Metabolism, Florence, Italy, 1992; p 89. Ashton,
M. J .; Bridge, A. W.; Bush, R. C.; Dron, D. I.; J oannou, P. P.;
Riddell, D.; Roberts, S.; Stevenson, G. V. W.; Warne, P. J . RP
64477: a Potent Inhibitor of Acyl-CoA:cholesterol O-Acyl Trans-
ferase (ACAT). Abstracts of the 9th International Symposium
on Atherosclerosis, Chicago, IL, 1991; p 183.
(10) Largis, E. E.; Wang, C. H.; DeVries, V. G.; Schaffer, S. A. CL
277,082: A Novel Inhibitor of ACAT-Catalysed Cholesterol
Esterification and Cholesterol Absorption. J . Lipid. Res. 1989,
30, 681-690.
(11) Higley, C. A.; Wilde, R. G.; Maduskuie, T. P.; J ohnson, A. L.;
Pennev, P.; Billheimer, J . T.; Robinson, C. S.; Gillies, P. J .;
Wexler, R. R. Acyl CoA:Cholesterol Acyltransferase Inhibitors:
Synthesis and Structure-Activity Relationship Studies of a New
Series of Trisubstituted Imidazoles. J . Med. Chem. 1994, 37,
3511-3522.
(12) Hainer, J . W.; Terry, J . G.; Connell, J . M.; Zyruk, H.; J enkins,
R. M.; Shand, D. L.; Gillies, P. J .; Livak, K. J .; Hunt, T. L.;
Crouse, J . R. Effect of the Acyl-CoA:Cholesterol Acyltransferase
Inhibitor DuP 128 on Cholesterol Absorption and Cholesterol
in Humans. Clin. Pharmacol. Ther. 1994, 52, 65-74.
(13) Roth, B. D.; Blankley, C. J .; Hoefle, M. L.; Holmes, A.; Roark,
W. H.; Trivedi, B. K.; Essenburg, A. D.; Kieft, K. A.; Krause, B.
R.; Stanfield, R. L. Inhibitors of Acyl-CoA:Cholesterol Acyltrans-
ferase. 1. Identification of a Structure Activity Relationships
of a Novel Series of Fatty Acid Anilide Hypocholesterolaemic
Agents. J . Med. Chem. 1992, 35, 1609-1617. For the descrip-
tion of more recent compounds related to CI-976, see: O’Brien,
P. M.; Sliskovic, D. R.; Anderson, M. K.; Bousley, R. F.; Krause,
B. R.; Stanfield, R. L. Inhibitors of Acyl CoA:Cholesterol Acyl-
transferase (ACAT) as Hypocholesterolemic Agents 12. Syn-
theses and Biological Activity of Structurally Novel Tetrazole
Amides. Bioorg. Med. Chem. Lett. 1995, 5, 289-294 and
subsequent papers in that issue.
ferase (ACAT) Inhibitors: Ureas Bearing Two Heterocyclic Head
Groups. Bioorg. Med. Chem. Lett. 1995, 5, 173-176. Wilde, R.
G.; Klaczkiewicz, J . D.; Billheimer, J . T.; Wexler, R. R.; Gillies,
P. J . Bioorg. Med. Chem. Lett. 1995, 5, 177-180.
(18) Brown, T. M. A.; Mueller, S. B.; Uhlendorf, P. D.; Newton, R.
S.; Krause, B. Comparison of CI-976, an ACAT Inhibitor, and
Selected Lipid Lowering Agents for Antiatherosclerotic Activity
in Iliac-Femoral and Thoracic Lesions. Arteriosclerosis Throm-
bosis 1991, 11, 1830-1843.
(19) Sliskovic, D. R.; Trivedi, B. T. ACAT Inhibitors: Potential Anti-
atherosclerotic Agents. Curr. Med. Chem. 1994, 1, 204-225.
(20) Ashton, M. J .; Bridge, A. W.; Bush, R. C.; Dron, D. I.; Harris, N.
V.; J ones, G. D.; Lythgoe, D. J .; Riddell, D.; Smith, C. RP
70676: A Potent Systemically Available Inhibitor of Acyl-CoA:
Cholesterol O-Acyl Transferase (ACAT). Bioorg. Med. Chem.
Lett. 1992, 2, 375-380.
(21) THP-1 cells are a human monocyte leukemia cell line that can
be stimulated to differentiate in culture in the presence of
phorbol ester. Differentiated THP-1 cells have macrophage-like
characteristic including scavenger receptor activity. Via, D. P.;
Pons, L.; Dennison, D. K.; Fanslow, A. E.; Bernini, F. Induction
of Acetyl-LDL-Receptor Activity by Phorbol Ester in Human
Monocyte Cell Line THP-1. J . Lipid Res. 1989, 30, 1515-1524.
(22) Smith, C.; Ashton, M. J .; Bush, R. C.; Facchini, V.; Harris, N.
V.; Hart, T.; J ordan, R.; MacKenzie, R.; Riddell, D. Bioorg. Med.
Chem. Lett. 1996, 6, 47-50. Riddell, D.; Bright C.; Burton, B.
J .; Bush, R. C.; Harris, N. V.; Hele, D.; Moore, U. M.; Naik, K.;
Parrott, D. P.; Smith, C.; Williams, R. J . Hypolipidaemic
Properties of RP 73163, a Potent and Systemically Bioavailable
Inhibitor of ACAT, in Animals Fed Diets Low in Cholesterol.
Biochem. Pharmacol. 1996, in press.
(23) Pitchen, P. The Asymmetric Synthesis of a Biologically Active
Sulphoxide. Chem. Ind. (London) 1994, 636-639.
(24) Stoeck, V.; Schunack, W. Synthesis of Imidazoles With Liquid
Ammonia. 4. N-Substituted Imidazoles From Aldehydes, 1,2-
Diketones, Primary Amines, and Liquid Ammonia. Arch. Pharm.
1974, 307, 922-925.
(25) Bridge, A. W.; Harris, N. V.; Lythgoe, D. J .; Smith, C. Imidazole
Derivatives. WO 9205148-A, 1991.
(26) Harris, N. V.; Smith, C. Preparation of 2-(4,5-Diphenyl-2-
imidazolyl)-1,3-Dioxanes as Antiatherosclerotics. EP 424195-
A, 1991. Ashton, M. J .; Dron, D. I.; Harris, N. V.; Hart, T. W.;
Riddell, D.; Smith, C.; Walsh, R. J . A. Inhibition of ACAT by
4,5-Diphenylimidazoles Substituted at the 2-Position by an
Acetal Function. Abstracts of the 7th RSC-SCI Medicinal
Chemistry Symposium, Cambridge, U.K., 1993; P46.
(27) Yartseva, S. V.; Tanaseichuk, B. S. Nitrogen Containing Het-
erocyclic Free Radicals. XII. 2-(R-Furyl)-4(5)-(5)-Phenyl-5(4)-
p-X-Phenylimidazoles. J . Org. Chem. USSR (Engl. Transl.)
1971, 7, 1558-1561.
(28) J ones, R. G. Studies on Imidazole Compounds. 1. A Synthesis
of Imidazoles with Functional Groups in the 2-Position. J . Am.
Chem. Soc. 1949, 71, 383-386.
(29) Kreevoy, M. M.; Taft, R. W. Acid-Catalysed Hydrolysis of Acetal
and Chloroacetal. J . Am. Chem. Soc. 1955, 77, 3146-3148.
(30) Searles, S.; Nickerson, R. G.; Witsiepe, W. K. Oxetanes. IX.
Structural and Solvent Effects in the Reaction of ø-Bromoalco-
hols With Base. J . Org. Chem. 1959, 24, 1839-1844.
(31) Crowley, J . I.; Rapoport, H. Unidirectional Dieckmann Cycliza-
tions on a Solid Phase and in Solution. J . Org. Chem. 1980, 45,
3215-3227.
(14) McCarthy, P. A.; Hamanaka, M. S.; Marzetta, C. A.; Bamberger,
M. J .; Gaynor, B. J .; Chang, G.; Kelly, S. E.; Inskeep, P. B.;
Mayne, J . T.; Beyer, T. A.; Walker, F. J .; Goldberg, D. I.; Savoy,
Y. E.; Davis, K. M.; Diaz, C. L.; Freeman, A. M.; J ohnson, D. A.;
LaCour, T. G.; Long, C. A.; Maloney, M. E.; Martingano, R. J .;
Pettini, J . L.; Sand, T. M.; Wint, L. T. Potent, Selective, and
Systemically-Available Inhibitors of Acyl-Coenzyme A:Choles-
terol Acyl Transferase (ACAT). J . Med. Chem. 1994, 37, 1252-
1255.
(15) Trivedi, B. K.; Purchase, T. S.; Holmes, A.; Augelli-Szafran, C.
E.; Essenburg, A. D.; Hamelehle, Stanfield, R. L.; Bousley, R.
F.; Krause, B. R. Inhibitors of Acyl-CoA:Cholesterol Acyltrans-
ferase (ACAT). 7. Development of a Series of Substituted
N-Phenyl-N′-[(1-phenylcyclopentyl)methyl]ureas with Enhanced
Hypocholesterolemic Activity. J . Med. Chem. 1994, 37, 1652-
1659.
(16) Sliskovic, D. R.; Krause, B. R.; Picard, J . A.; Anderson, M.;
Bousley, R. F.; Hamelehle, K. L.; Homan, R.; J ulian, T. N.;
Rashidbaigi, Z. A.; Stanfield, R. L. Inhibitors of Acyl-CoA:
Cholesterol O-Acyl Transferase (ACAT) as Hypocholesterolemic
Agents. 6. The First Water-Soluble ACAT Inhibitor with Lipid
Regulating Activity. J . Med Chem. 1994, 37, 560-562.
(17) Wilde, R. G.; Billheimer, J . T.; Germain, S. J .; Gillies, P. J .;
Higley, C. A.; Kezar, H. S.; Maduskuie, T. P.; Shimshick, E. S.;
Wexler, R. R. Acyl-CoA:Cholesterol Acyltransferase (ACAT)
Inhibitors: Ureas Bearing Heterocyclic Groups Bioisosteric for
an Imidazole. Bioorg. Med. Chem. Lett. 1995, 5, 167-172.
Wilde, R. G.; Billheimer, J . T. Acyl-CoA:Cholesterol Acyltrans-
(32) Lichtenstein, A. H.; Brecher, P. Properties of Acyl-CoA:Choles-
terol Acyltransferase in Rat Liver Microsomes. J . Biol. Chem.
1980, 255, 9098-9104.
(33) Becker, A.; Bottcher, A.; Lackner, K. J .; Fehringer, P.; Notka,
F.; Aslanidis, C.; Schmitz, G. Purification, Cloning and Expres-
sion of a Human Enzyme with Acyl Coenzyme A:Cholesterol
Acyltransferase Activity, Which is Identical to Liver Carboxyl-
esterase. Arteriosclerosis Thrombosis 1994, 14, 1346-1355.
Chang, T. Y.; Chang, C. C. Y.; Cadigan, K. M. The Structure of
Acyl Coenzyme A-Cholesterol Acyltransferase and its Potential
Relevance to Atherosclerosis. Trends Cardiovasc. Med. 1994,
4, 223-230.
(34) Folch, J .; Less, M.; Stanley, G. H. S. A Simple Method for the
Isolation and Purification of Total Lipid from Animal Tissues.
J . Biol. Chem. 1957, 226, 497-509.
(35) Reiff, H.; Dieterich, D.; Braden, R.; Ziemen, H. New 1,3-
Propanediols by Nucleophilic Cleavage of 3-Alkyl-3-Hydroxy-
methyl Oxetanes. J ustus Liebig Ann. Chem. 1973, 43, 365-374.
J M9505876