2-Arylbenzoxazoles as novel cholesteryl ester transfer protein inhibitors: Optimization via array synthesis

Article abstract of DOI:10.1016/j.bmcl.2008.03.030

2-Arylbenzoxazole 5 was identified as a hit from a fluorescence-based high-throughput screen for CETP inhibitors. The synthesis and SAR investigation employing array synthesis of the A- and B-rings are described.

Full text of DOI:10.1016/j.bmcl.2008.03.030

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2640–2644
2-Arylbenzoxazoles as novel cholesteryl ester transfer
protein inhibitors: Optimization via array synthesis
Lalgudi S. Harikrishnan,* Muthoni G. Kamau, Timothy F. Herpin, George C. Morton,
Yalei Liu, Christopher B. Cooper, Mark E. Salvati, Jennifer X. Qiao, Tammy C. Wang,
Leonard P. Adam, David S. Taylor, Alice Ye A. Chen, Xiaohong Yin,
Ramakrishna Seethala, Tara L. Peterson, David S. Nirschl, Arthur V. Miller,
Carolyn A. Weigelt, Kingsley K. Appiah, Jonathan C. O’Connell and R. Michael Lawrence
Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 4000, Princeton, NJ 08543-4000, USA
Received 23 January 2008; revised 6 March 2008; accepted 7 March 2008
Available online 14 March 2008
Abstract—2-Arylbenzoxazole 5 was identified as a hit from a fluorescence-based high-throughput screen for CETP inhibitors. The
synthesis and SAR investigation employing array synthesis of the A- and B-rings are described.
Ó 2008 Elsevier Ltd. All rights reserved.
Lipid modifying therapies for the treatment of coronary
heart disease (CHD) have principally focused on lower-
ing circulating levels of low-density lipoprotein choles-
terol (LDL-C).1 With the recognition that there exists
an inverse relationship between high-density lipoprotein
cholesterol (HDL-C) and the risk of CHD,2 drug discov-
ery efforts now also focus on investigating mechanisms
which have the potential for raising HDL-C in plasma.
Current therapies such as fibrates or niacin exert only
a modest effect on increasing HDL-C levels while dis-
playing considerable side effects.3 The need, therefore,
exists for safer and more efficacious methods of increas-
ing HDL-C levels.
reverse cholesterol transport and therefore be anti-ath-
erogenic.6 Interest in this target by the pharmaceutical
community has been widespread as evidenced by the
number of clinical and/or pre-clinical lead compounds
in various stages of development (Fig. 1).7 In this com-
F
O
CF3
O
MeO
F3C
N
N
F
OH
N
H
CF3
S
Et
O
F3C
N
EtO
O
1
3
2
Cholesteryl ester transfer protein (CETP) is a 74-kDa
plasma glycoprotein secreted principally by the liver.
CETP facilitates the transfer of cholesterol ester from
HDL to LDL and VLDL in exchange for triglycerides.4
Recent data have shown that CETP inhibition leads to
increased HDL-C in humans.5 As HDL particles are
capable of accepting cholesterol from peripheral tissues
including macrophages, CETP inhibitors could promote
Pfizer
Bayer
BAY 194789
Japan Tobacco
JTT 705
Torcetrapib
Cl
O
Me
B
O
O
7
6
O
A
NH Me
5
N
Me
4
N
OCF2CF2H
5
CF3
HO
Bristol-Myers Squibb screening hit
SPA IC50 = 280 nM
Keywords: CETP; HDL; Parallel synthesis; Benzoxazole; Cholesteryl
ester transfer protein; CETP inhibitor; Coronary heart disease; CHD;
SAR.
4
Pharmacia
SC 795
WPA IC50 = 10 μM
*
Corresponding author. Tel.: +1 609 252 3806; fax: +1 609 252
7446; e-mail: lalgudi.harikrishnan@bms.com
Figure 1. CETP inhibitors.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.030

L. S. Harikrishnan et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2640–2644
2641
munication, the identification and optimization of a no-
vel series of 2-arylbenzoxazole-based CETP inhibitors
are described.
shows that the planar 2-arylbenzoxazole moiety (Align-
ment 1) overlays well with the tetracyclic core of choles-
terol, with the 3b-hydroxyl group in the A-ring of
cholesterol being positioned in a similar orientation as
the 5-Me substituent on the benzoxazole of 5. In the
alternative Alignment 2, the 2-aryloxyacetamide group
traverses the tetracyclic cholesterol core overlaying the
A-ring of the steroid core with the phenyl ether. Both
overlays suggest that substitution at the terminal rings
A and B of 5 might be anticipated to modulate CETP
inhibitory activity. The modular structure of this chem-
otype made it well suited for rapid exploration of the
SAR of these terminal rings by applying array synthesis.
The SAR strategy involved developing chemistry that
allowed exploration of substitutions on either the fused
phenyl of the benzoxazole (A) or the phenyl ether (B).
Initial results of these studies are presented in this
communication.
A screening campaign was carried out on the BMS com-
pound collection using a BODIPY fluorescence assay.8
Retest positives in the fluorescence assay were directly
followed by reconfirmation in a dose–response mode
using a scintillation proximity assay (SPA).9 Potent
compounds in SPA were then evaluated for activity in
a human whole plasma assay (WPA).10 The 2-arylbenz-
oxazole 5 (Fig. 1) was one of the most potent analogs
identified in the HTS (SPA IC50 = 0.28 lM) which also
possessed a reasonable level of activity in human plasma
(WPA IC50 = 10 lM).
The 2-arylbenzoxazole hit 5 is structurally dissimilar
from published CETP inhibitors (Fig. 1). The relatively
linear core of the 2-arylbenzoxazole shows a structural
similarity to the molecular framework of cholesterol,
the esters of which are known to bind in equimolar ra-
tios with CETP.11 Molecular modeling of an energy
minimized conformation of 5 with cholesterol (Fig. 2)
The synthetic approach employed to assess various A-
ring substitutions is outlined in Scheme 1. Aryloxyacetic
acid 6 was treated with oxalyl chloride to provide the
corresponding acid chloride 7 in quantitative yield. Cou-
pling of 7 with 4-carbomethoxy aniline, followed by
saponification of the resulting ester gave carboxylic acid
8 in 56% overall yield. Activation of 8 with oxalyl chlo-
ride provided the acid chloride 9 in quantitative yield,
which was subsequently coupled with various commer-
cially available substituted amino phenols in an array
format to afford intermediate amides of the general
structure 10. Conversion of the amides 10 to benzoxaz-
oles 11 was achieved by microwave-mediated cyclization
employing catalytic p-toluenesulfonic acid. The cycliza-
tions typically proceed within 30 min, and the isolated
yields ranged from 1% to 50% (from acid 8).12 The aver-
age purity (by HPLC-UV) of the final products was
94%.
The structure–activity relationships for a variety of A-
ring substitutions are shown in Table 1. Consistent with
Alignment 1, substitution on the A-ring significantly
modulates CETP inhibitory activity. Removal of both
Figure 2. Energy minimized conformation of 5 (blue) and overlays
with cholesterol (green).
O
O
b,c
O
O
a
O
O
HO
O
NH
100%
56%
HO
Cl
6
7
8
O
O
d, c
Cl
a
NH
OH
NH2
100%
O
Z
9
OH
HN
O
O
7
Z
6
Z
O
O
e
O
N
NH
A
4
NH
5
1-50%
(from 9)
O
10
11
Scheme 1. General synthetic route to access A-ring variants. Reagents and conditions: (a) (COCl)2, DCM, cat. DMF; (b) methyl 4-aminobenzoate,
Et3N, DCM; (c) LiOH, THF; (d) DIPEA, DCM; (e) cat. TsOH, toluene, 185 °C, l-wave, 30 min.

2642
L. S. Harikrishnan et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2640–2644
Table 1. SAR for benzoxazole A-ring substitutions against human
CETP
which takes into consideration off-target interactions
in plasma. The 5-cyano compound 30 and the 5-cya-
no-7-chloro compound 45 exhibited the best potencies
across both SPA and WPA.
O
O
7
Z
O
N
6
NH
A
4
The synthetic approach utilized to examine the B-ring in
the 2-arylbenzoxazole series is detailed in Scheme 2.
Condensation of aminophenol 47 with chlorooxime 48
followed by reduction gave the aminobenzoxazole 49
in 94% yield. From intermediate 49, the B-ring was ex-
plored in an array format using one of the two proce-
5
a
Compound
Z
SPA IC50
(lM)
WPA IC50
(lM)
4
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
5
SC 795 (Fig. 1)
0.003
>32
>32
>32
1.1
0.036
NT
NT
NT
31
dures. Acylation with
a range of commercially
available aryloxyacetyl chlorides provided the product
amides 51 directly in 29–81% isolated yield.12 Alterna-
tively, to facilitate more extensive exploration of the B
aryl ring, aniline 49 was acylated with chloroacetyl chlo-
ride to obtain the a-chloro amide 50 in 89% yield. Dis-
placement of the chlorine of 50 with phenols in an
array format yielded the resulting purified phenyl ethers
51 in 10–40% isolated yield.12 The average purity (by
HPLC-UV) of the final products was 98%.
4-Me
4-NO2
5-Me
5-COOMe
5-COOH
5-F
>32
>32
>32
>32
11
NT
NT
NT
NT
71
5-OMe
5-OCF3
5-SO2Et
5-Ph
9.3
1.9
NT
NT
27
5-CH2OH
5-Ac
5-tBu
1.3
1.2
18
Select results from exploration of the substitution on the
2-arylbenzoxazole B-ring are provided in Table 2.
Mono-substitution at the 2, 3, or 4 positions with a vari-
ety of substituents revealed that small alkyl groups and
halogens are generally well-tolerated, although haloge-
nation at the 2 position appears to lead to less active
analogs (55 and 64). Several disubstituted analogs (80–
85) provided further improvements in inhibitory po-
tency both in the SPA and WPA. However, disubstitu-
tion at the 2,5-, 2,6-, or 3,5-positions led to a
significant loss of activity (86–90) suggesting that substi-
tutions were permitted along only one edge of the phe-
nyl ring. A limited set of 2,3,4-trisubstituted analogs
was also examined (e.g., 91–94, 97, 99, and 100). These
proved to be well tolerated, exhibiting comparable SPA
activity to the best di-substituted compounds. Among
them, compounds 99 and 100, which contain the best
A-ring mono-substitution (5-CN) and the best B-ring
trisubstitutions, had low nM SPA IC50s and were the
most potent in WPA (WPA IC50 ca. 1 lM) in the series.
>32
0.59
0.47
0.44
0.12
0.057
13
NT
51
5-CF3
5-Cl
21
28
5-Br
5-NO2
18
5-CN
6-Me
3.3
NT
NT
NT
NT
NT
NT
NT
100
80
6-Cl
6-OMe
6-NO2
>32
>32
>32
>32
>32
>32
1.7
6-COOMe
6-F
6-(Piperidin-1-yl)
7-Me
7-COOMe
7-COOH
5,7-DiMe
5-Me-7-Ac
5-Cl-7-Br
5,7-DiCl
5,7-BisCF3
5-CN-7-Cl
5-Cl-7-NO2
3.5
>32
0.28
0.24
0.097
0.088
0.092
0.052
0.049
NT
8.9
27
41
42
43
44
45
46
26
11
9.1
2.5
18
In summary, high-throughput screening of the BMS
sample collection identified 2-arylbenzoxazole 5 with
submicromolar SPA and modest WPA in vitro activi-
ties. An array synthesis approach was taken to rapidly
a NT, not tested.
Cl
a or b,
OH
A-ring methyl groups from 5 led to a loss of activity.
Reintroduction of a single methyl substitution at the 4
position also led to an inactive analog. However, the
activity was partially recovered by methyl substitution
at the 5, 6, or 7 positions. Incorporation of amino phe-
nols allowed more extensive surveying of the 5, 6, and 7
positions of the benzoxazole. The data indicate that
mono substitution at the 5 position is preferred, with
smaller nitro and cyano groups (29 and 30) exhibiting
enhanced SPA and WPA activities. Disubstitution at
the 5 and 7 positions was also examined further, and
provided a modest improvement in inhibitory potency
in the SPA assay, although no significant improvement
was seen in the physiologically more relevant WPA
O
N
NO2
c
+
NH2
N
OH
94%
NH2
48
49
47
e
89%
O
Cl
O
O Ar
O
N
f
O
N
NH
NH
10-40%
51
50
Scheme 2. Synthetic route to vary B-ring. Reagents and conditions: (a)
EtOH, rt, 1 day; (b) EtOH, 100 °C, l-wave, 5 min; (c) Zn, HOAc,
MeOH; (d) Aryloxyacetylchloride, pyridine, DCM; (e) Chloroacetyl-
chloride, Na2CO3, DCM; (f) ArXH, Na2CO3, cat. KI, acetone,
MeCN, DMSO, 75 °C, 12 h.

L. S. Harikrishnan et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2640–2644
Table 2. SAR for benzoxazole B-ring substitutions against human CETP
2643
2
6
3
5
2
6
3
5
O
O
B
4
Z
O
O
B
4
Z
O
N
O
N
A1
NH
A2
NH
NC
a
A-sub
Compound
2
3
4
5
6
SPA IC50 (lM)
WPA IC50 (lM)
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A1
A2
A2
A2
A2
A2
A2
52
5
0.70
0.28
0.69
0.24
>10
1.1
30
Me
8.9
18
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Me
F
Me
F
26
F
NT
51
7.3
31
CN
OMe
Cl
>10
>10
>10
8.3
NT
NT
NT
NT
NT
NT
NT
11
CN
OMe
Cl
CN
OMe
Cl
>10
>10
>10
1.3
0.38
>10
0.9
20
i-Pr
NT
67
i-Pr
COOMe
CF3
0.22
>10
1.9
NT
NT
91
COOMe
CF3
>10
>10
3.3
32
CF3
Ph
NT
NT
NT
NT
NT
61
Ph
Ph
2.3
>10
0.20
3.6
OCF3
OCF3
Cl
OCF3
Cl
>10
0.20
0.14
0.12
0.096
0.031
0.023
>10
>10
>10
>10
>10
0.085
0.056
0.042
0.040
0.085
0.018
0.015
0.011
0.018
0.010
NT
13
Me
Me
Cl
4.5
4.9
2.7
3.2
4.9
NT
NT
NT
NT
NT
3.5
2.5
2.8
1.7
40
CF3
Me
Me
Me
Me
Cl
Cl
Br
F
Me
Cl
Me
Cl
Cl
Cl
Me
Me
–(CH2)3–
–(CH2)3–
Me
Cl
Cl
Cl
Cl
Cl
Cl
Br
Cl
Cl
Me
Me
Me
COOMe
Me
CF3
4.3
3.2
4.3
1.3
0.91
–(CH2)3–
Me
Me
Me
COOMe
Me
a NT, not tested.
optimize the in vitro CETP inhibition profile for this no-
vel series of compounds. In accordance with the predic-
tions derived from molecular modeling of 5 with
cholesterol, substitutions on either the A- or B-ring
resulted in modulation of CETP inhibitory potency. In
particular, the 5-CN on the A-ring, and di- or trisubsti-
tution with small hydrophobic groups on the B-ring pro-
vided improvements in potency. Finally, combinations

2644
L. S. Harikrishnan et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2640–2644
3. Barter, P.; Kastelein, J.; Nunn, A.; Hobbs, R. Atheroscle-
rosis 2003, 168, 195.
4. Barter, P. J.; Brewer, H. B.; Chapman, M. J.; Hennekens,
C. H.; Rader, D. J.; Tall, A. R. Arterioscler. Thromb.
Vasc. Biol. 2003, 23, 160.
5. Clark, R. W.; Sutfin, T. A.; Ruggeri, R. B.; Willauer, A.
T.; Sugarman, E. D.; Magnus-Aryitey, G.; Cosgrove, P.
G.; Sand, T. M.; Wester, R. T.; Williams, J. A.; Perlman,
M. E.; Bamberger, M. J. Arterioscler. Thromb. Vasc. Biol.
2004, 24, 490.
of preferred substitutions on both the A- and B-rings of
this chemotype resulted in analogs demonstrating fur-
ther improvements in SPA binding activity, with some
trisubstituted B-ring analogs showing improvement in
WPA activity. Further enhancement in WPA activity
for this chemotype may require more extensive explora-
tion of the structure–activity relationship.
Supplementary data
6. Linsel-Nitschke, P.; Tall, A. R. Nat. Rev. Drug Discov.
2005, 4, 193, and references therein.
7. (a) Ruggeri, R. B. Curr. Top. Med. Chem. 2005, 5, 257,
and references cited therein; (b) Sikorski, J. A. J. Med.
Chem. 2006, 49, 1; (c) Krishna, R. et al. Lancet 2007, 370,
1907.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2008.03.030.
8. Based on the following reference using LDL as acceptor:
Bisgaier, C. L.; Minton, L. L.; Essenburg, A. D.; White,
A.; Homan, R. J. Lipid Res. 1993, 34, 1625.
9. Details of scintillation proximity assay will be published in
due course.
10. Details of whole plasma assay will be published in due
course.
11. Clark, R. W.; Ruggeri, R. B.; Cunningham, D.; Bamber-
ger, M. J. J. Lipid Res. 2006, 47, 537.
References and notes
1. Bucher, H. C.; Griffith, L. E.; Guyatt, G. H. Arterioscler.
Thromb. Vasc. Biol. 1999, 19, 187.
2. (a) Rhoads, G. G.; Gulbrandsen, C. L.; Kagan, A. N.
Engl. J. Med. 1976, 294, 293; (b) Castelli, W. P.; Garrison,
R. J.; Wilson, P. W.; Abbott, R. D.; Kalousdian, S.;
Kannel, W. B. J. Am. Med. Assoc. 1986, 256, 2835; (c)
Kannel, W. B. Am. J. Cardiol. 1995, 76, 69c.
12. For a general synthetic procedure, please see Supplemen-
tal information.