4-(3-Aryloxyaryl)quinoline alcohols are liver X receptor agonists

Article abstract of DOI:10.1016/j.bmc.2009.10.001

A series of 4-(3-aryloxyaryl)quinolines with alcohol substituents on the terminal aryl ring was prepared as potential LXR agonists, in which an alcohol group replaced an amide in previously reported amide analogs. High affinity LXR ligands with excellent

Full text of DOI:10.1016/j.bmc.2009.10.001

Bioorganic & Medicinal Chemistry 17 (2009) 8086–8092
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
4-(3-Aryloxyaryl)quinoline alcohols are liver X receptor agonists
a
a
a
a
Ronald C. Bernotas a, , David H. Kaufman , Robert R. Singhaus , John Ullrich , Rayomand Unwalla ,
*
Elaine Quinet b, Ponnal Nambi b, Anna Wilhelmsson c, Annika Goos-Nilsson c, Jay Wrobel a
a Chemical Sciences, Wyeth Pharmaceuticals, 500 Arcola Road, Collegeville, PA 19426, USA
b Cardiovascular and Metabolic Diseases, Wyeth Pharmaceuticals, 500 Arcola Road, Collegeville, PA 19426, USA
c Karo Bio AB, Novum S-141, 57 Huddinge, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 4-(3-aryloxyaryl)quinolines with alcohol substituents on the terminal aryl ring was prepared
as potential LXR agonists, in which an alcohol group replaced an amide in previously reported amide ana-
logs. High affinity LXR ligands with excellent agonist potency and efficacy in a functional model of LXR
activity were identified, demonstrating that alcohols can substitute for amides while retaining LXR activ-
ity. The most potent compound was 5b which had an IC50 = 3.3 nM for LXRb binding and EC50 = 12 nM
(122% efficacy relative to T0901317) in an ABCA1 mRNA induction assay in J774 mouse cells.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 21 August 2009
Revised 29 September 2009
Accepted 1 October 2009
Available online 4 October 2009
Keywords:
Liver X receptor
LXR
Quinoline
Alcohol
Biarylether
ABCA1
1. Introduction
types form obligate heterodimers with retinoid X receptors (RXRs).
Binding of either an RXR agonist (e.g., 9-cis-retinoic acid) to the
The leading cause of death in developed countries is cardiovas-
cular disease.1 One of the most successful drug therapies targeting
cardiovascular disease has been to improve lipid profiles, in partic-
ular, to lower cholesterol and low density lipoproteins (LDL) levels
and possibly increase high density lipoproteins (HDL) levels. Drug
intervention based on this approach has focused on the HMG-CoA
reductase pathway leading to the development of various statins
including LipitorÒ and MevacorÒ.2 More recently, other mecha-
nisms for correcting dislipidemia have been examined. One such
approach is to increase levels of cholesterol transporters including
adenosine-binding cassette transporters (ABCs), a family of lipid
transporters responsible for regulating lipid homeostasis (Fig. 1).
Of particular importance is ABCA1, an important transporter in
macrophages and other cells.3
RXR portion of the heterodimer or of an LXR agonist (e.g., oxyster-
ols such as 24,25-epoxycholesterol) to the LXR portion activates
mRNA transcription encoding various genes including ABCs.7
Nuclear receptors, including LXRs, may also play an important
cholesterol
ABCA1
efflux
cholesterol
oxysterols
(LXR agonists)
Expression of ABCs is regulated by liver X receptors (LXRs),
members of the nuclear hormone receptors superfamily of gene
transcription factors.4 Because of their role in cholesterol efflux,
and their anti-inflammatory role, LXR agonists are an important
therapeutic target for the treatment of dyslipidemia.5 Of the two
9-cis-retinoic acid
(RXR agonists)
ABCA1
LXR
RXR
subtypes of LXR, LXRa and LXRb, LXRb is widely distributed in var-
ious cell types while LXRa is mainly expressed in kidney, intestine,
Macrophage
lung, spleen, and macrophages, and especially in liver.6 Both sub-
* Corresponding author. Tel.: +1 484 865 2168.
E-mail address: bernotr@wyeth.com (R.C. Bernotas).
Figure 1. Oxidized cholesterol activates the LXR-RXR heterodimer, upregulating
the ABCA1 transporter and increasing cholesterol efflux.
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.10.001

R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
8087
three subtypes of the receptor. We embarked on an effort to iden-
tify compounds which maintained the 4-phenyl-quinoline core but
with the benzylacetic acid moiety replaced with other potential
hydrogen bond acceptors. Our first approach replaced the benzyl-
acetic acid with a benzamide (4) which gave several high affinity
LXR ligands with good agonist potency and reduced PPAR activa-
tion.13 X-ray analysis of one analog (4a X = meta-C(O)–morpholine,
Y = benzyl, Z = CF3) cocrystallized in the ligand binding domain of
LXRb indicated strong hydrogen bond interactions with the oxygen
of the amide as well as the quinoline nitrogen. Efforts to identify
other potential acceptable hydrogen bond acceptors replacements
led us to prepare and test quinolines incorporating biarylethers
with alcohols in place of the amides (5).
Ph
Ph
N
CO2H
F3C CF3
HO
O
O
O
S
N
Cl
F3C
1 T0901317 (Tularik)
CF3
2 GW3965 (GSK)
Figure 2. LXR agonists from Tularik and GlaxoSmithKline.
anti-inflammatory role which may help minimize the development
of athlerosclerosis.8
Several pharmaceutical companies have been active in develop-
ing LXR agonists to treat dislipidemia. Several high affinity LXR li-
2. Chemistry
gands with potent agonism at both LXRb and LXR
a subtypes,
Preparation of targets 5 was accomplished in two parts: synthe-
sis of the 4-(3-hydroxyphenyl)quinoline core structure followed by
elaboration of the core into the functionalized biarylethers alcohols
(Scheme 1).
Syntheses of the 4-(3-hydroxyphenyl)quinolines began with the
previously prepared 2-aminophenones 6a and 6b.13 Phenones 6a
and 6b were individually subjected to Friedlander reactions14 with
phenylacetaldehyde to afford 3-phenylquinolines 7a and 7b,
respectively. This approach was analogous to the previously re-
ported synthesis of 7d.13 Using propionaldehyde with 6a gave 3-
methylquinoline 7c.
notably Tularik’s T0901317 (1)9 and GlaxoSmithKline’s GW3965
(2),10 have been identified through these efforts (Fig. 2).
From an initial high throughput screening hit, researchers at
Wyeth developed a series of 4-phenyl-quinolines with an addi-
tional acid side chain that interacts with the Arg 319 residue in
the LXRb subtype while the quinoline nitrogen interacts with the
His435 (Fig. 3).11 These quinolines-acids were generally high affin-
ity LXR ligands with potent agonist activity. Based on X-ray analy-
sis of co-crystals of LXRb and WAY-254011 (3), it was determined
that the carboxylic acid interacts with Arg 319 via a hydrogen
bond.12 Unfortunately, 3 also was a PPAR agonist, activating all
With the requisite 4-(3-hydroxyphenyl)quinoline cores in hand,
installation of the aryl alcohol remained. Two approaches were
used to complete the targets. Copper-mediated coupling of the
phenol to an arylbromide afforded the biarylethers with a methyl
ester either meta or para to the oxygen linker.15 Subsequent reduc-
tion with lithium borohydride gave the primary alcohols while
addition of excess methyl magnesium bromide afforded the ter-
tiary alcohols. Alternatively, biarylether formation using meta or
para HOCH2PhB(OH)2, again mediated by a copper salt, afforded
the primary alcohols directly.16 These reactions were often per-
formed using a microwave reactor. We also prepared a phenol
(9b) and its methoxy precursor (9a) to test the effect of removing
the methylene spacer from the alcohol.
CO2H
O
O
Y
X
N
N
CF3
Z
3
4 X = amide
5 X = -CR1R2OH
Figure 3. Modifications of WAY-254011 (3).
OH
O
OH
Y
CO2R
Y
O
a
b
NH2
N
N
Z
Z
Z
6a Z = CF3
6b Z = Cl
7a Y = Ph, Z = CF3
7b Y = Ph, Z = Cl
7c Y = Me, Z = CF3
7d Y = CH2Ph, Z = CF3
8a Y = CH2Ph, Z = CF3, 3-CO2Me
8b Y = Me, Z = CF3, 3-CO2Me
8c Y = Me, Z = CF3, 4-CO2Me
8d Y = Me, Z = Cl, 3-CO2Me
c or d
O
e, f
X
Y
N
Z
5a-5i, 9a-b
Scheme 1. Reagents and conditions: (a) PhCH2CHO, PhSO3H, toluene (Y = Ph) or MeCH2CHO, cat. H2SO4, AcOH (Y = Me), heat, 3–24 h (63–79%); (b) BrPhCO2R, CuI,
Me2NCH2CO2HÁHCl, Cs2CO3, dioxane, 105 °C (64–78%); (c) R1MgX, THF (82–92%); (d) LiBH4, THF (72–97%); (e) ArB(OH)2, Cu(OAc)2, Et3N, microwave or heat (36–76%); (f)
pyridine hydrochloride, 180–190 °C, 3 h (90%).

8088
R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
Table 1
Biarylether alcohol quinolines 5a
1
3
O
X
4
Y
3
N
8
Z
Compound
X
Y
Z
LXRb IC50 (nM)
LXR
a
IC50 (nM)
LAFb EC50 (nM) (% agonism)
1
3
—
—
—
—
CH2Ph
CH2Ph
CH2Ph
Ph
Ph
Ph
Me
Me
Me
Me
Me
Me
Me
CH2Ph
CH2Ph
—
—
9
2.1
53
11.7
3.3
14
13
16 (100%)
71 (97%)
9.5
231
75
5.2
47
107
36
>1000
10.2
>1000
131
1002
59
6.6
494
115
8a
5a
5b
5c
5d
5e
8b
5f
8c
5g
8d
5h
5i
3-CO2Me
3-CH2OH
3-CMe2OH
3-CH2OH
4-CH2OH
4-CH2OH
3-CO2Me
3-CH2OH
4-CO2Me
4-CH2OH
3-CO2Me
3-CH2OH
3-CMe2OH
4-OMe
CF3
CF3
CF3
CF3
CF3
Cl
CF3
CF3
CF3
CF3
Cl
895 (61%)
1030 (94%)
31 (106%)
210 (79%)
646 (80%)
1010 (104%)
3400 (56%)
233 (69%)
nt
1370 (48%)
1810 (45%)
938 (75%)
108 (84%)
612 (38%)
406 (59%)
10
8.3
435
2.1
>1000
31
231
12
2.5
287
28
Cl
Cl
CF3
CF3
9a
9b
4-OH
a
Results are given as the mean of two independent experiments. The standard deviations for the binding assays were typically 30% of the mean or less. The standard
deviations for the LAF assays were typically 30% of the mean or less. % of efficacy is relative to 1. nt = not tested.
The functional activity of several compounds was further tested
3. Biological assays
in a J774 mouse macrophage cell line, examining upregulation of
mRNA from the transporter ABCA1 (Table 2).19 The ABCA1 protein
removes cholesterol from macrophages and other cells. The J774
cell line is reported to have a preponderance of the LXRb subtype
of the receptor.20 Measuring the mRNA levels, tertiary alcohol 5b
was the most potent with an EC50 value of 12 nM, with another ter-
tiary alcohol, 5i, the second most potent at 37 nM. In contrast, 5c,
with a phenyl directly bonded at the 3-position of the quinoline,
had very weak affinity. The primary alcohols at the meta position
were generally weaker compared to their tertiary counterparts,
for example: 5a versus 5b, and 5i versus 5h. Typically, the primary
alcohols were weaker in the ABCA1 mRNA accumulation assay,
especially 5c, 5g, and 5h. When the 3-substituent on the quinoline
was a phenyl, the para-substituted alcohols were slightly more po-
tent compared to meta-analog 5c. Some differences in potency in
The final targets and their ester intermediates were tested to
determine binding affinity for the two LXR subtypes (Table 1).
The binding assays used recombinant human ligand binding do-
mains (LBDs) of the respective LXRa and LXRb subtypes and mea-
sured displacement of [3H]T0901317 from the LBD.11 3-
Benzylquinolines were examined first. The ester intermediate 8a
had good affinity for the LXRb and was nearly fivefold selective
over the LXRa subtype. The analogous alcohol 5a had even higher
affinity and slightly improved binding selectivity. The tertiary alco-
hol 5b had an IC50 value of 3.3 nM at LXRb but lost essentially all
beta selectivity. To reduce the molecular weight, we examined
the 3-phenyl primary alcohols, which had high LXRb affinity
whether meta (5c) or para (5d) substituted. Replacing the 8-trifluo-
romethyl group with a chlorine atom (5e) retained LXRb affinity
with some erosion in selectivity. To further reduce both clog P
and molecular weight, we had prepared the 3-methyl quinolines.
Both meta ester 8b and para ester 8c had poor affinity for LXR
receptors. However, in alcohols 5, this modification provided some
of the highest affinity compounds including 5f with an IC50 value of
2.1 nM at LXRb receptors. Unlike the 3-phenyl analogs, the meta-
alcohol was preferred over the para-alcohol 5g by 15-fold. 8-
Chloro-3-methyl-quinolines 5h and 5i incorporating the biaryle-
ther alcohol motif had comparable affinity to their higher clog P
and molecular weight comparators 5a and 5b, respectively. Biary-
lether 9b para-substituted with a hydroxyl group had good affinity
for LXRs.
Table 2
ABCA1 activity for 5a
O
3
X
4
Y
N
Z
Compound
X
Y
Z
ABCA1 EC50 (nM) (% ag)
1
3
—
—
—
—
CH2Ph
CH2Ph
Ph
Ph
Ph
Me
Me
Me
Me
—
—
27 (100%)
41 (114%)
111 (99%)
12 (122%)
>1000 (61%)
141 (44%)
168 (136%)
555 (112%)
>1000 (97%)
>1000 (94%)
37 (147%)
The compounds were also tested in a LAFb functional assay.17
The LAFb assay determined the agonist potency of test compounds
and their efficacy relative to 3 in CHO cells stably transfected with
hLXRb using secreted alkaline phosphatase as the reporter gene,
driven by multiple response elements for LXRb. The most potent
in this assay was 3-benzylquinoline 5b, which had an EC50 value
of 37 nM and was as efficacious as the Tularik reference (3). How-
ever, two of the 3-methylquinolines, 5f and 5i, were also relatively
potent with EC50 values below 250 nM and with similar efficacy to
the reference compound.18
5a
5b
5c
5d
5e
5f
5g
5h
5i
3-CH2OH
3-CMe2OH
3-CH2OH
4-CH2OH
4-CH2OH
3-CH2OH
4-CH2OH
3-CH2OH
3-CMe2OH
CF3
CF3
CF3
CF3
Cl
CF3
CF3
Cl
Cl
a
Results are the mean of at least two independent experiments. Standard devi-
ations for the assays were typically 30% of mean or less. % Efficacy is relative to 1.

R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
8089
replaced an amide as the hydrogen bond acceptor. Several of the
alcohols prepared had high affinity for both LXR and LXRb sub-
a
types and were potent agonists in a J774 mouse cell line examining
ABCA1 gene regulation, based on increases in ABCA1 mRNA levels.
The most potent compound for inducing ABCA1 gene expression
was tertiary alcohol 5b, with an EC50 = 12 nM and fully efficacious
relative to the test ligand, T0901317 (1). The affinity and activity of
the alcohols was generally similar to the comparable amides, with
similar trends in SAR. For example, the molecular weight and clog P
could be reduced by changing the 3-substituent on the quinoline
from a benzyl to a methyl with little loss in ABCG1 mRNA induc-
tion. While these compounds had at best modest binding selectiv-
ity for LXRb over LXRa, we have demonstrated that 4-(3-
aryloxyaryl)quinolines with appropriately positioned alcohol sub-
stituents, especially tertiary alcohols, are high affinity LXR ligands
with good agonist potency.
Figure 4. A docked orientation of 5a (yellow) using the previously disclosed X-ray
structure of hLXRb/4a complex. Only key residues and helices are shown for
simplicity. Hydrogen bonds to key residues are shown by dotted lines.
6. Experimental
General experimental: Solvents and chemicals were purchased
from EM Sciences, VWR, Oakwood, and Aldrich Chemical Co. and
used without further purification. High-resolution mass spectra
were obtained on a Waters LC-TOFMS instrument and were mea-
sured to within 5 ppm of calculated values. 1H NMR spectra were
taken on a Bruker DPX300 (300 MHz) or Varian (400 MHz) instru-
ments. NMR data are given as delta values (d) ppm using tetra-
methylsilane as an internal standard (d = 0 ppm). For the NMR
data peak descriptions, app means apparent, fc indicates additional
fine coupling (<1 Hz), and br means broad.
the LAFb assay compared to the ABCA1 assay may be attributable
to differences in cell type, species, and compound penetration into
the cells. Nearly all the compounds were essentially as efficacious
as 1 and the best compounds demonstrated comparable potency to
3. These results indicate that with optimal substitutions, the 4-(3-
aryloxyaryl)quinoline alcohols were high affinity LXR ligands, with
good potency in the ABCA1 induction assay in J774 macrophage
cells, showing that an alcohol group could take the place of an
amide while maintaining activity. Furthermore, none of the five
most potent compounds in the ABCA1 assay (5a, 5b, 5d, 5e, 5i)
6.1. (3-{3-[3-Benzyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)methanol (5a)
had any PPARa, PPARd, or PPARc agonist activity in functional as-
says using transiently transfected cell lines described earlier.13
A stirred mixture of 8a (128 mg, 0.25 mmol) in dry THF (2.5 mL)
was treated with LiBH4 (55 mg, 2.50 mmol) and heated at 40 °C for
6 h. The reaction was treated carefully with 2 M hydrochloric acid
(ca. 1 mL), diluted with water (10 mL), and stirred 15 min. The
solution was extracted with dichloromethane (2 Â 15 mL) and
the combined extracts dried (MgSO4) and concentrated in vacuo
to an oil, which was purified by chromatography eluting with
50:50 ethyl acetate/hexanes to afford 5a as a slightly colored vis-
cous oil (115 mg, 94%). 1H NMR (CDCl3): d 8.98 (1H, s), 8.02 (1H,
d, J = 6.7 Hz), 7.70 (1H, d, J = 6.5 Hz), 7.50–7.45 (2H, m), 7.31 (1H,
app t, J = 7.9 Hz), 7.23–7.09 (6H), 6.99–6.92 (4H, m), 6.84 (1H, m),
4.66 (2H, s), 4.00 (2H, s); MS (ES) m/z 486.2; HRMS: calcd for
C30H22F3NO2+H+, 486.1675; found (ESI, [M+H]+), 486.1691. HPLC
purity: 100%.
4. Molecular modeling studies
Docking studies21 were carried out on compound 5a to under-
stand the ligand binding mode within the LXRb cavity and shed
further light on residues which are important for ligand recogni-
tion. Figure 4 shows the top scoring pose of 5a from docking.
The ligand binding mode was similar to that of the previously
disclosed X-ray structure of hLXRb/4a (4a X = meta-C(O)-
N(CH2CH2)2O, Y = CH2Ph, Z = CF3). Ligand recognition was achieved
by hydrogen bond interaction between the quinoline nitrogen
atom and His435 residue. Additionally, the 7-trifluoromethyl
group was in close proximity to His435 residue, that is, d(N–
F) = 3.0 Å to make favorable electrostatic interaction with this res-
idue. The N2-benzyl group was enclosed in a hydrophobic pocket
surrounded by three Phe residues (329, 340, and 349), while the
benzyloxy linker was able to extend the alcohol group towards
the b-hairpin loop region and make hydrogen bond interaction
with the backbone NH group of Leu330. It is interesting to note
that in the X-ray structure of 4a, a similar hydrogen bond interac-
tion was seen from the carbonyl group of the amide to the NH
backbone of Leu330 residue. In fact, the distance between the
quinoline nitrogen atom and alcohol oxygen atom in 5a, that is,
12.8 Å is almost identical to the distance of 12.7 Å observed be-
tween the nitrogen atom of the quinoline ring and the carbonyl
oxygen atom in the X-ray structure of 4a.
6.2. 2-(3-{3-[3-Benzyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)propan-2-ol (5b)
A stirred solution of 8a (128 mg, 0.25 mmol) in dry THF (2.5 mL)
at 0 °C under nitrogen was treated with 3.0 M MeMgBr in THF
(0.50 mL, 1.50 mmol). The reaction was allowed to warm to ambi-
ent temperature over 2 h, treated with 2 M aqueous HCl (3 mL)
followed by brine (5 mL) and extracted with ethyl acetate
(2 Â 15 mL). The combined extracts were dried (MgSO4), concen-
trated in vacuo, and the resulting oil was chromatographed on sil-
ica gel using 50:50 ethyl acetate/hexane as eluent to afford 5b as a
tan solid (105 mg, 82%). Mp <80 °C. 1H NMR (CDCl3): d 8.98 (1H, s),
8.02 (1H, d, J = 7.3 Hz), 7.70 (1H, d, J = 8.4 Hz), 7.49–7.44 (2H),
7.31–7.13 (7H), 6.98 (2H, d, J = 7.7 Hz), 6.93 (1H, dd, J = 0.9,
7.4 Hz), 6.89 (1H, br d, J = 7.9 Hz), 6.85 (1H, s), 4.00 (2H, s), 1.59
(1H, br s), 1.55 (6H, s); MS (ES) m/z 514.2; HRMS: calcd for
5. Conclusion
To further explore the SAR in a series of quinolines containing a
biarylether, compounds were synthesized in which an alcohol

8090
R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
C32H26F3NO2+H+, 514.1988; found (ESI, [M+H]+), 514.1963. HPLC
purity: 95.1%.
6.7. (4-{3-[3-Methyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)methanol (5g)
6.3. (3-{3-[3-Phenyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)methanol (5c)
Compound 8c (529 mg, 1.21 mmol) stirred in dry THF (20 mL)
at 20 °C under nitrogen was treated with LiBH4 (66 mg, 3.0 mmol).
After 1 d, additional LiBH4 (130 mg, 6.1 mmol) was added and stir-
ring continued for 5 d. The reaction was treated carefully with 2 M
hydrochloric acid (ꢀ10 mL), diluted with water (10 mL), and stir-
red 3 d. The solution was extracted with dichloromethane
(2 Â 20 mL) and the combined extracts dried (MgSO4) and concen-
trated in vacuo to an oil which slowly solidified. Purification by
chromatography eluting with 35:65 ethyl acetate/hexanes affor-
ded 5g as a white solid (355 mg, 72%). 1H NMR (CDCl3): d 8.98
(1H, s), 8.00 (1H, d, J = 7.1 Hz), 7.70 (1H, d, J = 8.4 Hz), 7.53–7.44
(2H), 7.36 (2H, d, J = 8.7 Hz), 7.13 (1H, m), 7.08 (2H, d, J = 8.7 Hz),
6.97 (1H, ddd, J = 1.0, 2.5, 7.7 Hz), 6.89 (1H, m), 4.67 (2H, s), 2.31
(3H, s); MS (ESI) m/z 410; HRMS: calcd for C24H18F3NO2+H+,
410.1362; found (ESI, [M+H]+), 410.1342. HPLC purity: 97.9%.
Compound 7a (0.074 g, 0.21 mmol), 3-(hydroxymethyl)phenyl-
boronic acid (0.068 g, 0.44 mmol), copper acetate (0.033 g,
0.21 mmol), and powdered 4 Å molecular sieves were stirred at
ambient temperature in dichloromethane (26 mL) for 10 min.
Triethylamine (0.10 mL, 0.61 mmol) was added. After 24 h, the reac-
tion was filtered and the crude reaction solution was loaded directly
on a silica gel column. Chromatography eluting with 5:95–50:50
ethyl acetate/hexane gradient gave 5c as a tan solid (66 mg, 67%).
1H NMR (DMSO-d6): d 9.10 (1H, s), 8.24 (1H, d, J = 6.8 Hz), 7.92
(1H, d, J = 7.5 Hz), 7.75 (1H, app t, J = 7.8 Hz), 7.46 (1H, app t,
J = 7.9 Hz), 7.39–7.32 (4H), 7.28–7.26 (2H), 7.24 (1H, d, J = 7.8 Hz),
7.11–7.02 (3H), 6.88–6.86 (2H, 6.55 (1H, m), 5.21 (1H, t, J = 5.7 Hz),
4.46 (2H, d, J = 5.6 Hz); MS (ESI) m/z 472; HRMS: calcd for
C29H20F3NO2+H+, 472.1519; found (ESI-FT/MS, [M+H]+), 472.1517.
6.8. {3-[3-(8-Chloro-3-methylquinolin-4-yl)phenoxy]phenyl}-
methanol (5h)
6.4. (4-{3-[3-Phenyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)methanol (5d)
To a stirred solution of 8d (101 mg, 0.25 mmol) in dry THF
(2.5 mL) was added LiBH4 (55 mg, 2.5 mmol). After heating at
40 °C for 4 h, the reaction was cooled, carefully treated with 2 M
aqueous hydrochloric acid (ca. 1 mL), and diluted with water
(5 mL). The solution was extracted with dichloromethane
(2 Â 5 mL) and the combined extracts dried (MgSO4) and concen-
trated in vacuo. The residue was purified by chromatography elut-
ing with 50:50 ethyl acetate/hexanes to afford 5h as a tacky gum
(87 mg, 92%). 1H NMR (DMSO-d6): d 8.98 (1H, s), 7.90 (1H, dd,
J = 1.3, 7.4 Hz), 7.60 (1H, dd, J = 7.4, 8.4 Hz), 7.50 (1H, dd, J = 7.5,
8.4 Hz), 7.40 (1H, dd, J = 1.3, 8.4 Hz), 7.35 (1H, app t, J = 7.9 Hz),
7.16 (1H, ddd, J = 1.1, 2.6, 8.3 Hz), 7.11–7.07 (3H, m), 6.98–6.96
(2H, m), 5.23 (1H, t, J = 5.8 Hz), 4.49 (2H, d, J = 5.7 Hz), 2.27 (3H,
s); MS (ES) m/z 376.1; HRMS: calcd for C23H18ClNO2+H+,
376.1099; found (ESI, [M+H]+), 376.1090. HPLC purity: 99.3%.
Compound 5d was prepared in essentially the same manner as
5c above except employing 4-(hydroxymethyl)phenylboronic acid
to provide a white solid (42 mg, 44%). 1H NMR (DMSO-d6): d 9.10
(1H, s), 8.24 (1H, d, J = 7.0 Hz), 7.92 (1H, d, J = 7.5 Hz), 7.75 (1H,
app t, J = 8.1 Hz), 7.48 (1H, app t, J = 7.9 Hz), 7.39–7.35 (4H),
7.28–7.23 (3H), 7.14 (1H, m), 7.05 (1H, d with fc, J = 7.4 Hz), 6.80
(1H, app t, J = 2.0 Hz), 6.69 (2H, app d, J = 8.6 Hz), 5.13 (1H, t,
J = 5.7 Hz), 4.44 (2H, d, J = 5.6 Hz); MS (ESI) m/z 472; HRMS: calcd
for C29H20F3NO2+H+, 472.1519; found (ESI-FT/MS, [M+H]+),
472.1513. HPLC purity: 100%.
6.5. {4-[3-(8-Chloro-3-phenylquinolin-4-yl)phenoxy]phenyl}-
methanol (5e)
Compound 5e was prepared in the same manner as 5d above
except using 7b as the phenol reactant giving a colorless, tacky so-
lid (17 mg, 36%). 1H NMR (DMSO-d6): d 9.06 (1H, s), 8.01 (1H, dd,
J = 2.8, 5.8 Hz), 7.62–7.57 (2H), 7.47 (1H, app t, J = 7.9 Hz), 7.39–
7.32 (3H), 7.25–7.22 (3H), 7.12 (1H, finely coupled d, J = 7.5 Hz),
7.05 (2H, m), 6.77 (1H, m), 6.69 (2H, d with fc, J = 8.6 Hz), 5.13
(1H, t, J = 5.7 Hz), 4.44 (2H, d, J = 5.7 Hz); MS (ES) m/z 438.2; HRMS:
calcd for C28H20ClNO2+H+, 438.1255; found (ESI, [M+H]+),
438.1239. HPLC purity: 90%.
6.9. 2-{3-[3-(8-Chloro-3-methylquinolin-4-yl)phenoxy]phenyl}-
propan-2-ol (5i)
A stirred mixture of 8d (101 mg, 0.25 mmol) in dry THF (2.5 mL)
at 0 °C under nitrogen was treated with 3.0 M MeMgBr in THF
(0.50 mL, 1.50 mmol). The reaction was allowed to warm to ambi-
ent temperature after 30 min, then treated with 2 M hydrochloric
acid (1 mL) and additional water, and extracted with dichloro-
methane (2 Â 5 mL). The combined extracts were dried (MgSO4),
concentrated in vacuo, and the resulting oil was chromatographed
on silica gel using 1:1 ethyl acetate/hexanes as eluent to afford 5i
as a white solid (93 mg, 92%). Mp: 163–165 °C; 1H NMR (DMSO-
d6): d 8.98 (1H, s), 7.90 (1H, dd, J = 1.3, 7.4 Hz), 7.61 (1H, app t,
J = 8.0 Hz), 7.49 (1H, dd, J = 7.5, 8.4 Hz), 7.40 (1H, dd, J = 1.3,
8.6 Hz), 7.32 (1H, app t, J = 8.4 Hz), 7.23–7.20 (2H, m), 7.16 (1H,
br d, J = 8.3 Hz), 7.09 (1H, br d, J = 7.9 Hz), 6.94–6.91 (2H, m),
5.05 (1H, s), 2.27 (3H, s), 1.39 (6H, s); MS (ES) m/z 404.2; HRMS:
calcd for C25H22ClNO2+H+, 404.1412; found (ESI, [M+H]+),
404.1414. HPLC purity: 100%.
6.6. 3-{3-[3-Methyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenyl)methanol (5f)
A stirred mixture of 8b (502 mg, 1.125 mmol) in dry THF
(20 mL) was added LiBH4 (62 mg, 2.81 mmol). After 24 h, the reac-
tion was treated with additional LiBH4 (125 mg, 5.63 mmol) and
stirred 6 d. The reaction was carefully treated with 2 M hydrochlo-
ric acid (ca. 15 mL), diluted with water (10 mL), and stirred 3 d. The
solution was extracted with dichloromethane (2 Â 20 mL) and the
combined extracts dried (MgSO4) and concentrated in vacuo. The
resulting oil was purified by chromatography eluting with 35:65
ethyl acetate/hexanes to afford 5f as a colorless viscous oil
(446 mg, 97%). 1H NMR (CDCl3): d 8.99 (1H, s), 8.00 (1H, d,
J = 7.1 Hz), 7.71 (1H, d, J = 8.5 Hz), 7.53–7.45 (2H, m), 7.35 (1H,
app t, J = 7.8 Hz), 7.16–7.11 (2H, m), 6.99 (2H, m), 6.91 (1H, br s),
4.69 (2H, s), 2.31 (3H, s), 1.65 (1H, v br s); MS (ESI) m/z 410; HRMS:
calcd for C24H18F3NO2+H+, 410.1362; found (ESI, [M+H]+),
410.1362. HPLC purity: 100%.
6.10. 3-[3-Phenyl-8-(trifluoromethyl)quinolin-4-yl]phenol (7a)
A stirred mixture of 6a (200 mg, 0.71 mmol) and phenylacetal-
dehyde (285 mg, 2.13 mmol, Aldrich 90% tech grade) in toluene
(3.0 mL) was treated with benzenesulfonic acid (337 mg,
2.13 mmol) and heated at reflux under nitrogen overnight. The
reaction was cooled, concentrated under a nitrogen stream, treated
with saturated aqueous NaHCO3 and the aqueous layer extracted

R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
8091
several times with ethyl acetate. The combined extracts were dried
6.14. Methyl 4-{3-[3-methyl-8-(trifluoromethyl)quinolin-4-
yl]phenoxy}benzoate (8c)
over MgSO4, filtered, and concentrated in vacuo. The residue was
purified by chromatography using a 0:100–20:80 ethyl acetate/
hexanes gradient to afford 7b as a tan solid from a foam
(205 mg, 79%). 1H NMR (DMSO-d6): d 9.57 (1H, s), 9.09 (1H, s),
8.22 (1H, d, J = 7.1 Hz), 7.87 (1H, d, J = 8.3 Hz), 7.72 (1H, app t,
J = 7.9 Hz), 7.34–7.26 (5H), 7.21 (1H, app t, J = 7.9 Hz), 6.78 (1H,
dd, J = 1.6, 7.8 Hz), 6.68 (1H, d, J = 7.6 Hz), 6.62 (1H, m); MS (ES)
m/z 366.1; HRMS: calcd for C22H14F3NO+H+, 366.1100; found
(ESI, [M+H]+), 366.1108.
Compound 8c was prepared in essentially the same manner as
8b except using methyl 4-bromobenzoate to provide a white solid
(78%). 1H NMR (CDCl3): d 9.01 (1H, s), 8.07–8.01 (3H), 7.70 (1H, d,
J = 8.4 Hz), 7.58 (1H, app t, J = 7.9 Hz), 7.49 (1H, app t, J = 7.8 Hz),
7.21 (1H, m), 7.11–7.07 (2H, m), 6.98 (1H, m), 3.91 (3H, s), 2.33
(3H, s); MS (ESI) m/z 438; HRMS: calcd for C25H18F3NO3+H+,
438.1312; found (ESI, [M+H]+), 438.1331. HPLC purity: 98.5%.
6.11. 3-(8-Chloro-3-phenylquinolin-4-yl)phenol (7b)
6.15. 3-Benzyl-4-[3-(4-methoxyphenoxy)phenyl]-8-(trifluoro-
methyl)quinoline (9a)
A stirred mixture of 6b (991 mg, 4.00 mmol) and phenylacetal-
dehyde (624 mg, 5.20 mmol, Aldrich 90% tech grade) in toluene
(20 mL) was treated with benzenesulfonic acid (362 mg) and
heated at 115 °C under nitrogen for 20 h. The reaction was cooled,
treated with saturated aqueous NaHCO3 (20 mL) and the aqueous
layer extracted with dichloromethane (2 Â 50 mL). The combined
extracts were dried over MgSO4, filtered, and concentrated in va-
cuo. The residue was purified by chromatography using a 0:100–
30:70 ethyl acetate/hexanes gradient to afford 7b as an off-white
solid (912 mg, 69%). 1H NMR (DMSO-d6): d 9.55 (1H, s), 9.05 (1H,
s), 8.00 (1H, m), 7.59–7.54 (2H), 7.34–7.26 (5H), 7.20 (1H, app t,
J = 7.9 Hz), 6.78 (1H, d with fc, J = 8.8 Hz), 6.66 (1H, app dt,
J = 1.3, 7.6 Hz), 6.60 (1H, m); MS (ES) m/z 332.1; HRMS: calcd for
C21H14ClNO+H+, 332.0837; found (ESI, [M+H]+), 332.0836. HPLC
purity: 100%.
A mixture 4-methoxyphenylboronic acid (76 mg, 0.50 mmol),
Cu(OAc)2 (45 mg, 0.25 mmol), and powdered 4 Å molecular sieves
(100 mg) was treated with 7d13 (95 mg, 0.25 mmol) in dichlorometh-
ane (2.5 mL). Triethylamine (0.10 mL, 0.75 mmol) was added and the
reaction stirred at 20 °C for 3 d. The reaction was filtered through Cel-
ite, treated with water (5 mL), and extracted with dichloromethane
(2 Â 10 mL). The combined extracts were concentrated in vacuo.
Chromatography afforded compound 9a as a colorless oil (89 mg,
76%). 1H NMR (CDCl3): d 8.98 (1H, s), 8.01 (1H, d, J = 7.2 Hz), 7.69
(1H, d, J = 8.4 Hz), 7.48–7.41 (3H), 7.23–7.17 (3H), 7.07 (1H, m),
6.98–6.95 (3H), 6.89–6.84 (3H), 6.75 (1H, s with fc), 3.99 (2H, s),
3.78 (3H, s); MS (ES) m/z 485.9; HRMS: calcd for C30H22F3NO+H+,
486.1675; found (ESI, [M+H]+), 486.1706. HPLC purity: 98.5%.
6.16. 4-{3-[3-Benzyl-8-(trifluoromethyl)quinolin-4-yl]phenoxy}-
phenol (9b)
6.12. 3-[3-Methyl-8-(trifluoromethyl)quinolin-4-yl]phenol (7c)
A stirred mixture of 6a (1.00 g, 3.56 mmol) and propionalde-
hyde (0.77 mL, 16.7 mmol) in glacial acetic acid (20 mL) was
treated with concd H2SO4 (ca. 0.2 mL) and then heated at 115–
120 °C for 6 h, adding additional propionaldehyde (0.5 mL) at
intervals. The reaction was cooled, diluted with water, and
extracted several times with ethyl acetate. The combined
extracts were washed with twice with water and once with
saturated aqueous NaHCO3. The extracts were dried with MgSO4
and concentrated in vacuo. Chromatography eluting with 20:80
ethyl acetate/hexanes afforded 7c as a yellow solid (0.681 g,
63%). 1H NMR (DMSO-d6): d 9.75 (1H, s), 9.01 (1H, s), 8.12
(1H, d, J = 7.0 Hz), 7.71 (1H, d, J = 7.7 Hz), 7.63 (1H, app t,
J = 7.3 Hz), 7.39 (1H, app t, J = 7.8 Hz), 6.94 (1H, d with fc,
J = 8.2 Hz), 6.72 (1H, d, J = 7.5 Hz), 6.69 (1H, m), 2.27 (3H, s);
MS (ESI) m/z 304; HRMS: calcd for C17H12F3NO+H+, 304.0944;
found (ESI, [M+H]+), 304.0938. HPLC purity: 95%.
A stirred mixture of 9a (0.22 g, 0.45 mmol) and pyridine hydro-
chloride (3.0 g) was heated to 180–190 °C for 3 h, then cooled, trea-
ted with ethyl acetate, and the organic layer washed with 1 M
aqueous hydrochloric acid and then brine. Organic layer was dried
(MgSO4) and concentrated in vacuo. The residue was purified by
chromatography eluting with a 10:90–50:50 ethyl acetate/hexanes
gradient to afford 9b as an off-white solid (190 mg, 90%). Mp 55–
60 °C; 1H NMR (CDCl3): d 8.98 (1H, s), 8.01 (1H, d, J = 7.3 Hz),
7.68 (1H, d, J = 8.4 Hz), 7.48–7.41 (2H), 7.23–7.15 (3H), 7.07 (1H,
d with fc, J = 8.4 Hz), 6.97 (2H, d, J = 8.1 Hz), 6.94–6.88 (3H),
6.80–6.75 (3H), 4.69 (1H, s), 3.99 (3H, s); MS (ES) m/z 469.9; HRMS:
calcd for C29H20F3NO2+H+, 472.1519; found (ESI, [M+H]+),
472.1530. HPLC purity: 100%.
Acknowledgments
We thank Anita Halpern and Dawn Savio for project support.
6.13. Methyl 3-{3-[3-methyl-8-(trifluoromethyl)quinolin-4-
yl]phenoxy}benzoate (8b)
References and notes
A vigorously stirred mixture of 7c (606 mg, 2.00 mmol), methyl
3-bromobenzoate (860 mg, 4.00 mmol), CuO (288 mg, 3.60 mmol),
and K2CO3 (552 mg, 4.00 mmol) in dry pyridine (5.0 mL) was
heated under nitrogen at 120 °C for 65 h. The cooled reaction
was diluted with water (15 mL) and extracted with ether
(2 Â 20 mL). The dried (MgSO4) extracts were concentrated to a
very dark oil which was chromatographed on silica gel using
25:75 ethyl acetate/hexanes as eluent to give compound 8b as a
tacky foam (0.56 g, 64%). 1H NMR (CDCl3): d 8.99 (1H, s), 8.01
(1H, d, J = 7.3 Hz), 7.80 (1H, d, J = 7.8 Hz), 7.75–7.71 (2H, m), 7.54
(1H, app t, J = 7.9 Hz), 7.50–7.42 (2H), 7.30 (1H, br d, J = 8.0 Hz),
7.16 (1H, br d, J = 8.3 Hz), 7.03 (1H, br d, J = 7.7 Hz), 6.92 (1H, br
s), 3.91 (3H, s), 2.33 (3H, s); MS (ESI) m/z 438; HRMS: calcd for
C25H18F3NO3+H+, 438.1312; found (ESI, [M+H]+), 438.1339. HPLC
purity: 98.6%.
1. For a recent review on the impact of cardiovascular disease and stroke, see:
Lloyd-Jones, D. et al Circulation 2009, 119, 480.
2. For an overview of LipitorÒ therapy, see: Bybee, K. A.; Lee, J. H.; O’Keefe, J. H.
Curr. Med. Res. Opin. 2008, 24, 1217.
3. Hozoji, M.; Munehira, Y.; Ikeda, Y.; Makishima, M.; Matsuo, M.; Kioka, N.; Ueda,
K. J. Biol. Chem. 2008, 283, 30057.
4. For a recent review, see: Goodwin, B. J.; Zuercher, W. J.; Collins, J. L. Curr. Top.
Med. Chem. 2008, 8, 781.
5. For a review, see: Bennett, D. J.; Brown, L. D.; Cooke, A. J.; Edwards, A. S. Exp.
Opin. Ther. Pat. 2006, 16, 1673.
6. Auboeuf, D.; Rieusset, J.; Fajas, L.; Vallier, P.; Frering, V.; Riou, J. P.; Staels, B.;
Auwerx, J.; Laville, M.; Vidal, H. Diabetes 1997, 46, 1319.
7. Rastejad, F. Curr. Opin. Struct. Biol. 2001, 11, 33.
8. For a review see: Bensinger, S. J.; Tontonoz, P. Nature 2008, 454, 470.
9. Repa, J. J.; Turley, S. D.; Lobaccaro, J. A.; Medina, J.; Li, L.; Lustig, K.; Shan, B.;
Heyman, R. A.; Dietschy, J. M.; Mangelsdorf, D. J. Science 2000, 289, 1524; Li, L.
et al Bioorg. Med. Chem. Lett. 2006, 16, 1638.
10. Collins, J. L. et al J. Med. Chem. 2002, 45, 1963.

8092
R. C. Bernotas et al. / Bioorg. Med. Chem. 17 (2009) 8086–8092
11. Hu, B.; Collini, M.; Unwalla, R.; Miller, C.; Singhaus, R.; Quinet, E.; Savio, D.;
Halpern, A.; Basso, M.; Keith, J.; Clerin, V.; Chen, L.; Resmini, C.; Liu, Q.-Y.;
Feingold, I.; Huselton, C.; Azam, F.; Farnegardh, M.; Enroth, C.; Bonn, T.; Goos-
Nilsson, A.; Wilhelmsson, A.; Nambi, P.; Wrobel, J. J. Med. Chem. 2006, 49, 6151.
12. Hu, B.; Quinet, E.; Unwalla, R.; Collini, M.; Jetter, J.; Dooley, R.; Andraka, D.;
Nogle, L.; Savio, D.; Halpern, A.; Goos-Nilsson, A.; Wilhelmsson, A.; Nambi, P.;
Wrobel, J. Bioorg. Med. Chem. Lett. 2008, 18, 54.
17. For detailed assay conditions, see Supplementary data in Ref. 12.
18. In a comparable LAF reporter assay using rat LXR , the compounds tested in
this assay (5b, 5d–5f, 5h, 5i, 9a, 9b) had comparable EC50 values for rat LXR
(less than twofold difference), suggesting minimal functional selectivity for
LXRb versus LXR
a
a
a.
19. Quinet, E.; Savio, D. A.; Halpern, A. R.; Chen, L.; Miller, C. P.; Nambi, P. J. Lipid
Res. 2004, 45, 1929.
13. Bernotas, R.; Singhaus, R.; Kaufman, D.; Ullrich, J.; Fletcher, H.; Quinet, E.;
Nambi, P.; Unwalla, R.; Savio, D.; Wilhelmsson, A.; Goos-Nilsson, A.;
Farnegardh, M.; Wrobel, J. Bioorg. Med. Chem. 2009, 17, 1663.
20. Quinet, E. M.; Savio, D. A.; Halpern, A. R.; Chen, L.; Schuster, G. U.; Gustafsson,
J.; Basso, M. D.; Nambi, P. Mol. Pharmacol. 2006, 70, 1340; Joseph, S. B.; Bradley,
M. N.; Castrillo, A.; Bruhn, K. W.; Mak, P. A.; Pei, L.; Hogenesch, J.; O’Connell, R.
M.; Cheng, G.; Saez, E. Cell 2004, 119, 299.
14. Friedlander, P. Chem. Ber. 1882, 15, 2572.
15. Ma, D.; Cai, Q. Org. Lett. 2003, 5, 3799.
16. Evans, D. A.; Katz, J. L.; West, T. R. Tetrahedron Lett. 1998, 39, 2937; Chan, D. M.
T.; Monaco, K. L.; Wang, R.-P.; Winters, M. P. Tetrahedron Lett. 1998, 39, 2933.
21. For methods see: Hu, B.; Unwalla, R.; Collini, M.; Quinet, E.; Feingold, I.; Goos-
Nilsson, A.; Wilhelmsson, A.; Nambi, P.; Wrobel, J. Bioorg. Med. Chem. 2009, 17,
3519.