2-Arylbenzoxazoles as CETP inhibitors: Raising HDL-C in cynoCETP transgenic mice

Article abstract of DOI:10.1016/j.bmcl.2010.10.062

We describe structure-activity studies leading to the discovery of 2-arylbenzoxazole 3, the first in a series to raise serum high-density lipoprotein cholesterol levels in transgenic mice. Replacement of the 4-piperidinyloxy moiety with piperazinyl provid

Full text of DOI:10.1016/j.bmcl.2010.10.062

Bioorganic & Medicinal Chemistry Letters 21 (2011) 558–561
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
2-Arylbenzoxazoles as CETP inhibitors: Raising HDL-C in cynoCETP
transgenic mice
⇑
Florida Kallashi , Dooseop Kim, Jennifer Kowalchick, You Jung Park, Julianne A. Hunt, Amjad Ali,
Cameron J. Smith, Milton L. Hammond, James V. Pivnichny, Xinchun Tong, Suoyu S. Xu, Matt S. Anderson,
Ying Chen, Suzanne S. Eveland, Qiu Guo, Sheryl A. Hyland, Denise P. Milot, Anne-Marie Cumiskey,
Melanie Latham, Laurence B. Peterson, Raymond Rosa, Carl P. Sparrow, Samuel D. Wright, Peter J. Sinclair
Merck Research Laboratories, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe structure–activity studies leading to the discovery of 2-arylbenzoxazole 3, the first in a series
to raise serum high-density lipoprotein cholesterol levels in transgenic mice. Replacement of the 4-pipe-
ridinyloxy moiety with piperazinyl provided a more synthetically tractable lead, which upon optimiza-
tion resulted in compound 4, an excellent inhibitor of cholesteryl ester transfer protein function with
good pharmacokinetic properties and in vivo efficacy.
Received 28 July 2010
Revised 13 October 2010
Accepted 14 October 2010
Available online 20 October 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
CETP
Cholesteryl ester transfer protein
HDL-C
CVD
2-Arylbenzoxazole
It has been established that serum levels of high-density lipo-
protein cholesterol (HDL-C) and the risk of cardiovascular disease
(CVD) are inversely correlated. Niacin, the most effective HDL-C
increasing agent currently available, suffers from lack of compli-
ance. Thus, there is a need for better HDL raising therapies that
combine efficacy with negligible side effects.1,2 One approach to
reducing CVD risk via HDL-C elevation is by inhibition of the cho-
lesteryl ester transfer protein (CETP).3
rescently labeled CE. Aryloxyamides exemplified by 16 showed
good inhibition of CETP-mediated CE transfer and b-alkyloxyamine
27 showed improved pharmacokinetic properties relative to the
aryloxyamides. However, no in vivo efficacy was observed in a cyn-
oCETP transgenic mouse pharmacodynamic (PD) model9 with
these or analogous compounds (Fig. 1).
This communication describes further structure–activity rela-
tionship studies and optimization of the
a-aryloxyamide moiety
CETP is a plasma glycoprotein that facilitates the reshaping of
lipoprotein particles. It mediates the transfer of cholesteryl ester
(CE) from CE-rich HDL particles to triglyceride (TG)-rich particles,
primarily very low-density lipoproteins (VLDL), and the reciprocal
movement of TG from VLDL to HDL. It is a process driven by the
substrate concentration gradient between lipoproteins with a net
effect of reducing serum HDL-C.2 Clinical trials have confirmed that
pharmacological inhibition of CETP causes elevation of serum HDL-
C in humans.4,5
of compound 1 that led to the discovery of 3, the first CETP inhib-
itor in the 2-arylbenzoxazole class reported to show in vivo effi-
cacy. Subsequent optimization efforts to improve potency and
pharmacokinetic properties culminated in 4, which was a potent
inhibitor of CETP in vitro and raised HDL-C in the transgenic mouse
PD assay.
We began by exploring cycloalkyloxy replacements of the aryl-
oxy group (Table 1). Preparation of
a-alkoxyamides 3 and 5–12 has
been previously described.7
Recent reports by our labs6,7 and others8 have described the dis-
covery and ongoing optimization of a series of 2-arylbenzoxazoles
as CETP inhibitors. The potency was determined by measuring a
compound’s ability to inhibit the CETP-mediated transfer of fluo-
As shown in Table 1, the inhibitory potency of compounds
increased with the steric bulk of the cycloalkyl group (5–10). How-
ever, the most potent compound in the series, the methylbicyc-
looctane analog 9, elicited no effect in the in vivo PD assay.
Initially, this lack of in vivo efficacy was attributed to the poor
physicochemical properties of this series of molecules, in particu-
lar, their very low aqueous solubility, and the correspondingly
low drug exposures seen in the pharmacokinetic analyses.
⇑
Corresponding author. Tel.: +1 732 594 6025.
E-mail address: florida_kallashi@merck.com (F. Kallashi).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.062

F. Kallashi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 558–561
559
CF3
CF3
O
O
O
O
N
O
N
NH
NH
NC
NC
CF3
1, CE IC50 = 60 nM
2, CE IC50 = 151 nM
N
O
O
N Boc
O
NH
N
O
O
N
NH
N
NC
NC
3, CE IC50 = 287 nM
4, CE IC50 = 16 nM
Figure 1. CETP inhibitors.
inhibitor of TG transfer. When tested in our PD assay at 20 mpk,
compound 3 raised HDL-C by 13 mg/dL.
Table 1
Effect of cycloalkyloxy substituents on inhibition of CE transfer
We then embarked on an effort to optimize this scaffold. Since
we had previously established that methylation at the C7 position
of the benzoxazole is potency-enhancing,6 we utilized that substi-
tution in subsequent SAR exploration (Table 2). Piperidinyl analogs
16–30 were prepared following the procedure outlined in Scheme
1. Nitration and subsequent reduction of 4-bromo-2-methylphenol
afforded the aminophenol 13. Amide coupling of 13 with 4-nitro-
benzoyl chloride was followed by acid-catalyzed cyclization to
form the 4-nitrophenyl benzoxazole. Reduction of the nitro group
and subsequent installation of the cyano group resulted in com-
pound 14. Acyl chloride 15 was prepared in a two-step fashion
from N-Boc-4-hydroxypiperidine and coupled with aniline 14.
The resulting compound 16 was deprotected and treated with
the appropriate electrophile in the presence of base to afford com-
pounds 17–24 and 26–30, while compound 25 was prepared by
reductive alkylation with 3,3-dimethyl butyraldehyde.
O
O R
7
O
N
6
NH
NC
4
Compd
R
CE IC50 (nM)a
5
6
474
280
7
8
9
146
92
72
Compound 16, the C7-methylated analog of 3, did not show an
increased response in the PD assay, in spite of increased in vitro
10
115
11
12
3
O
1253
Table 2
Effect of substituents on piperidine on HDL-C levels
NH
N
>30,000
492/287b
O
O
N R
Boc
O
N
NH
Assay conditions have been described.10
This data-point was generated by modifying the assay protocol to include a 1 h
a
NC
b
preincubation at 25 °C of inhibitor with CETP/serum premix. This change allowed
for full development of time-dependent inhibition and resulted in better data
reproducibility. Any data reported hereafter was obtained using these modified
assay conditions.
Compd
R
CE IC50
(nM)a
TG IC50
DHDL-C
(nM)a
(mg/dL)b
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
–C(O)–OtBu
91
92
nd
nd
99
130
84
524
252
83
350
256
625
nd
17c
1c
–C(O)–OMe
–C(O)–OEt
–C(O)–OiPr
–C(O)–OCH2-tBu
–C(O)–OC(Me)2–Et
–C(O)–OPh
382
254
160
111
94
221
130
181
228
370
1301
207
336
169
7c
3d
4c
We also investigated introduction of heteroatoms (Table 1).
This approach was generally detrimental to in vitro potency (11–
12), although the Boc-protected piperidinyl 3 was a moderate
inhibitor of CE transfer.
Over the course of the program, we evaluated selected com-
pounds for their ability to inhibit the CETP-mediated transfer of
fluorescently labeled triglyceride (TG).10 We observed that, while
other classes of compounds3,11 inhibited both arms of CETP func-
tion and elicited a PD response, none of the 2-arylbenzoxazoles
that had been tested inhibited TG transfer in this assay. A rescreen
of the remaining benzoxazoles identified a single compound capa-
ble of inhibiting both arms of CETP-mediated neutral lipid transfer.
Thus, Boc-piperidine 3 was a 287 nM inhibitor of CE and a 182 nM
8d
4d
6d
nd
nd
nd
4c
–C(O)–OBn
–CH2–C(O)–OtBu
–(CH2)2–tBu
–C(O)–CH2-tBu
–C(O)–NtBu
–S(O)2–nBu
–S(O)2–Ph
1d
nd
nd
nd
31
–S(O)2–Bn
Assay conditions have been described.10
Assay conditions have been described.9
Dosed at 20 mpk.
a
b
c
d
Dosed at 10 mpk.

560
F. Kallashi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 558–561
OH
OH
O
N
c, d, e, f
a, b
NH2
Br
NH2
g, h
NC
Br
14
13
O
O
N
Boc
i
HO
N Boc
15
Cl
O
O
N Boc
O
j and k, l, m, or n
NH
N
NC
16
O
O
N R
O
N
NH
NC
Scheme 1. Synthesis of 2-arylbenzoxazole
a
-alkyloxyamides 16–30. Reagents and conditions: (a) HNO3 (fuming), AcOH (48%); (b) SnCl2Á2H2O, HCl, MeOH, 0 °C–rt (95%); (c)
4-nitrobenzoyl chloride, 1,4-dioxane, reflux (quant.); (d) p-TsOH, toluene (97%); (e) 10% Pd/C, H2, DCM (74%); (f) CuCN, NMP, 190 °C (86%); (g) bromoacetic acid, NaH, THF,
0 °C–rt (95%); (h) oxalyl chloride, DMF, DCM, 0 °C–rt (quant.); (i) LHMDS, THF, 0 °C–rt (66%); (j) TFA, DCM (quant.); (k) R–X, DIPEA, DCM (63–96%); (l) di-t-amyl dicarbonate,
TEA, DCM (quant.); (m) 3,3-dimethyl butyraldehyde, NaBH(OAc)3, DCE (81%); (n) t-butyl isocyanate, DIPEA, DCM (84%).
potency. However, this was attributed to a suboptimal pharmaco-
kinetic profile.
Furthermore, replacing the piperidine with a piperazine, resulted
in compound 33, which was essentially equipotent to 16.
Conservative modifications of the carbamate moiety of 16 did
not increase potency (Table 2). Replacing the t-butyl substituent
with other alkyl groups or phenyl (17–23), modifications including
incorporation of a methylene spacer (24) and replacement of the
carbamate with alkyl (25), amide (26), urea (27), and sulfonamide
groups (28–30), all resulted in compounds of lesser or equivalent
potency to 16.
A variety of N-substituted piperazine analogs were prepared as
shown in Scheme 2. Compounds 33, 35–44, 46 and 4 were pre-
pared by treating 14 with bromoacetyl bromide and displacing
the resulting
a-bromoacetamide with the appropriate N-substi-
tuted piperazine. Boc-deprotection of compound 33 afforded 34.
Acids 45 and 47 were prepared by treating the corresponding ester
with aqueous LiOH in THF and the appropriate alcohol.
Unable to improve upon the Boc-piperidinyl moiety, we then
The SAR of the piperazine series is shown in Table 4. While the
unsubstituted piperazine 34 did not inhibit CETP, the N-phenyl
analog (35) had a comparable in vitro profile to compound 33. A
variety of substituents on the phenyl ring were then explored.
No improvement over the phenyl (35) was observed with the
regioisomeric methoxy analogs (36–38). When the substituent
was chloro (39–41), the meta position was optimal. While the ester
analogs 44 and 46 had similar potency to the N-phenyl analog (35),
made small modifications to the
a-alkoxyamide ‘linker’ region of
the molecule (Table 3). Compounds 31 and 32 were prepared by
coupling aniline 14 with the corresponding acyl chloride, using
an analogous procedure to the one used to prepare compound
16.
Replacing the oxygen in the linker with a methylene group
caused a 12-fold decrease in CE transfer inhibition (31). However,
when the linker was shortened by one carbon, the resulting com-
pound (32) showed a modest increase in potency relative to 31.
there was
a drastic decrease in in vitro potency with the
corresponding acids (45 and 47). In the case of the trifluoromethyl
Table 3
Effect of ‘linker’ modifications on HDL-C levels
O
R
O
NH
N
NC
Compd
R
CE IC50 (nM)a
91
TG IC50 (nM)a
92
D
HDL-C (mg/dL)b
16
17c
O
N Boc
31
32
33
1094
757
64
951
356
83
nd
nd
8d
N Boc
Boc
N
N
N
Boc
a
b
c
Assay conditions have been described.10
Assay conditions have been described.9
Dosed at 20 mpk.
d
Dosed at 10 mpk.

F. Kallashi et al. / Bioorg. Med. Chem. Lett. 21 (2011) 558–561
561
R
N
O
N
O
N
a, b
NH2
O
N
NH
NC
14
NC
Scheme 2. Synthesis of N-substituted piperazin-1-yl acetamides 33, 35–44, 46 and 4. Reagents and conditions: (a) bromoacetyl bromide, DIPEA, THF, 40 °C; (b) N-substituted
piperazine, DIPEA, THF.
‘dual’ CE/TG transfer inhibitor raised HDL-C by 13 mg/dL in the
transgenic mouse PD model. Further exploration of this series con-
firmed that those 2-arylbenzoxazole compounds which inhibited
both arms (CE and TG) of CETP-promoted lipid transfer showed
in vivo efficacy.
While the underlying cause for this biochemical difference of
the benzoxazole series remains unclear, this observation guided
our subsequent structure activity relationship studies around the
piperidine moiety of compound 3, which led us to compound 4, a
potent ‘dual’ CETP inhibitor with in vivo efficacy. We will report
on the further optimization of this class of CETP inhibitors in due
course.
Table 4
Effect of substituents on piperazine on HDL-C levels
R
N
O
NH
N
O
N
NC
Compd
R
CE IC50 (nM)a TG IC50 (nM)a
D
HDL-C (mg/dL)b
34
35
36
37
38
39
40
41
42
43
4
H
Ph
>30,000
49
111
58
109
69
nd
nd
nd
nd
nd
nd
nd
11
nd
nd
14
24
nd
nd
nd
nd
349
nd
123
99
3720
29
135
3121
15
2-OMe–Ph
3-OMe–Ph
4-OMe–Ph
2-Cl–Ph
3-Cl–Ph
4-Cl–Ph
2-CF3–Ph
3-CF3–Ph
4-CF3–Ph
Acknowledgements
19
53
65
13
F. Kallashi thanks P. J. Sinclair and J. A. Hunt for many useful dis-
cussions and their continuous support.
References and notes
16
18
51
44
45
46
47
3-C(O)OMe–Ph 64
1. Sinclair, P. J. Annu. Rep. Med. Chem. 2005, 40, 71.
2. Singh, I. M.; Shishehbor, M. H.; Ansell, B. J. J. Am. Med. Assoc. 2007, 298, 786.
3. Hunt, J. A.; Lu, Z. Curr. Top. Med. Chem. 2009, 9, 419.
4. Bloomfield, D.; Carlson, G. L.; Sapre, A.; Tribble, D.; McKenney, J. M.; Littlejohn,
T. W.; Sisk, C. M.; Mitchel, Y.; Pasternak, R. C. Am. Heart J. 2009, 157, 352.
5. Barter, P. J.; Caulfield, M.; Eriksson, M.; Grundy, S. M.; Kastelein, J. J. P.;
Komajda, M.; Lopez-Sendon, J.; Mosca, L.; Tardif, J.-C.; Waters, D. D.; Shear, C.
L.; Revkin, J. H.; Buhr, K. A.; Fisher, M. R.; Tall, A. R.; Brewer, B. N. Engl. J. Med.
2007, 357, 2109.
3-C(O)OH–Ph
4-C(O)OEt–Ph
4-C(O)OH–Ph
>30,000
>30,000
40
59
9164
nd
a
Assay conditions have been described.10
Assay conditions have been described. Dosed at 10 mpk.9
b
6. Smith, C. J.; Ali, A.; Chen, L.; Hammond, M. L.; Anderson, M. A.; Chen, Y.;
Eveland, S. S.; Guo, Q.; Hyland, S. A.; Milot, D. P.; Sparrow, C. P.; Wright, S. D.;
Sinclair, P. J. Bioorg. Med. Chem. Lett. 2010, 20, 346.
7. Hunt, J. A.; Gonzalez, S.; Kallashi, F.; Hammond, M. L.; Pivnichny, J. V.; Tong, X.;
Xu, S. S.; Anderson, M. A.; Chen, Y.; Eveland, S. S.; Guo, Q.; Hyland, S. A.; Milot,
D. P.; Sparrow, C. P.; Wright, S. D.; Sinclair, P. J. Bioorg. Med. Chem. Lett. 2010, 20,
1019.
8. Harikrishnan, L. S.; Kamau, M. G.; Herpin, T. F.; Morton, G. C.; Liu, Y.; Cooper, C.
B.; Salvati, M. E.; Qiao, J. X.; Wang, T. C.; Adam, L. P.; Taylor, D. S.; Chen, A. Y. A.;
Yin, X.; Seethala, R.; Peterson, T. L.; Nirschl, D. S.; Miller, A. V.; Weigelt, C. A.;
Appiah, K. K.; O’Connell, J. C.; Lawrence, R. M. Bioorg. Med. Chem. Lett. 2008, 18,
2640.
Table 5
Comparison of mouse PK12 profiles
a
Compd
AUCNpo
Clearance
(ml/min/kg)
Half-life
(h)
Bioavailability
(%)
(l
M h kg/mg)
43
4
0.19
0.72
49
25
2.2
4.1
26
56
a
Normalized area under the curve.
substituent (42–43 and 4), both the meta and the para positions
were potency-enhancing in vitro. However, the para analog 4 out-
performed 43 in the transgenic mouse PD assay, raising HDL-C by
24 mg/dL (relative to 14 mg/dL for 43). This was consistent with
the observed pharmacokinetic profiles of the compounds in the
mouse (Table 5). While the meta analog 43 had a clearance of
49 ml/min/kg and 26% oral bioavailability, the para analog 4 exhib-
ited an improved pharmacokinetic profile with a clearance of
25 ml/min/kg and 56% oral bioavailability.
9. Transgenic mouse pharmacodynamic assay: B6 male mice 12–16 weeks of age
expressing cynomolgus monkey CETP were used (Marotti et al. Arterioscler.
Thromb. Vasc. Biol. 1992, 12, 736.) Pre-dose blood samples were collected by
retro-orbital bleed. Compounds were formulated in DMSO/cremophor/saline at
a 4:4:92 ratio and screened at 10 or 20 mg/kg PO BID. Twenty-four hours after
the first dose, the mice were euthanized and blood was collected by cardiac
puncture. HDL-C was measured on the pre- and post-bleed samples using a
HDL-E biochemical assay kit from Wako Chemicals USA (cat#431-52501),
following manufacturer’s instructions. The
DHDL-C reported is relative to the
baseline. A change in HDL-C of 7 mg/dl or less was generally considered not
significant.
10. Eveland, S. S.; Milot, D. P.; Guo, Q.; Chen, Y.; Hyland, S. A.; Peterson, L. B.;
Jezequel-Sur, S.; O’Donnell, G. T.; Zuck, P. D.; Ferrer, M.; Strulovici, B.; Wagner,
J. A.; Tanaka, W. K.; Hilliard, D. A.; Laterza, O.; Wright, S. D.; Sparrow, C. P.;
Anderson, M. S. Anal. Biochem. 2007, 368, 239. Human recombinant CETP was
used in this assay.
11. Ali, A.; Napolitano, J. M.; Deng, Q.; Lu, Z.; Sinclair, P. J.; Taylor, G. E.; Thompson,
C. F.; Quraishi, N.; Smith, C. J.; Hunt, J. A.; Dowst, A. A.; Chen, Y.; Li, H. U.S.
Patent, US7652049 2010.
In summary, our efforts to optimize 2-arylbenzoxazoles as CETP
inhibitors resulted in a series of
a-cycloalkyloxyamides which
were potent inhibitors of CE transfer; however, none of the
compounds in this class that were tested in a transgenic mouse
pharmacodynamic model proved efficacious.6,7 A screen of our col-
lection of benzoxazoles for inhibitors of CETP-promoted TG trans-
fer allowed us to identify compound 3 as a 182 nM inhibitor. This
12. PK data was generated from wild-type mice.