Phenoxy thiazole derivatives as potent and selective acetyl-CoA carboxylase 2 inhibitors: Modulation of isozyme selectivity by incorporation of phenyl ring substituents

Article abstract of DOI:10.1016/j.bmcl.2007.01.022

A phenyl ring substitution strategy was employed to optimize the ACC2 potency and selectivity profiles of a recently discovered phenoxy thiazolyl series of acetyl-CoA carboxylase inhibitors. Ring substituents were shown to dramatically affect isozyme sele

Full text of DOI:10.1016/j.bmcl.2007.01.022

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1961–1965
Phenoxy thiazole derivatives as potent and selective acetyl-CoA
carboxylase 2 inhibitors: Modulation of isozyme selectivity
by incorporation of phenyl ring substituents
Richard F. Clark,* Tianyuan Zhang, Xiaojun Wang, Rongqi Wang, Xiaolin Zhang,
Heidi S. Camp, Bruce A. Beutel, Hing L. Sham and Yu Gui Gu
Metabolic Disease Research, Global Pharmaceutical Research and Development, Abbott Laboratories,
Abbott Park, IL 60064, USA
Received 30 November 2006; accepted 9 January 2007
Available online 19 January 2007
Abstract—A phenyl ring substitution strategy was employed to optimize the ACC2 potency and selectivity profiles of a recently dis-
covered phenoxy thiazolyl series of acetyl-CoA carboxylase inhibitors. Ring substituents were shown to dramatically affect isozyme
selectivity. Modifications that generally impart high levels of ACC2 selectivity (>3000-fold) while maintaining excellent ACC2
potency (IC50s ꢀ 9–20 nM) were identified.
Ó 2007 Elsevier Ltd. All rights reserved.
As critical regulators of fatty acid metabolism, acetyl-
CoA carboxylases (ACCs) have attracted considerable
attention from the pharmaceutical research community
as targets for the treatment of metabolic syndrome.1
Metabolic syndrome is characterized by a cluster of
obesity-related disorders that collectively increase the
risk of coronary heart disease and type-2 diabetes.
Although the precise mechanisms are unclear, the accu-
mulation of fat in non-lipogenic tissues is recognized as
a primary causal factor in the development of insulin
resistance, a key component of metabolic syndrome.2
ACC catalyzes the carboxylation of acetyl-CoA to pro-
duce malonyl-CoA, a metabolic control signal acting as
both substrate for de novo lipogenesis and an allosteric
inhibitor of carnitine palmitoyltransferase1 (CPT-1)
thereby blocking fatty acid entry into the mitochondria
for b-oxidation. Consequently, the inhibition of ACC is
expected to lower malonyl-CoA levels resulting in
reduced fatty acid synthesis and increased fatty acid
oxidation, leading to weight loss and improved insulin
sensitivity. While several reports have validated the ther-
apeutic potential of inhibiting ACC,3 it remains unclear
whether selective or nonselective inhibition of the
enzyme’s two isoforms will be most advantageous.
Recently, genetic disruption studies in mice revealed
that ACC2-deficient animals were healthy and displayed
a favorable metabolic profile, while ACC1 knockout
resulted in embryonic death.4 A small molecule inhibitor
that selectively inhibits ACC2 may therefore provide
a safer therapeutic approach for chronic treatment of
obesity, diabetes, and other symptoms of metabolic
syndrome.
A primary goal of our research program has been the
optimization of ACC2 potency and selectivity profiles
for a recently discovered series of phenoxy thiazole
inhibitors exemplified by lead compound 1.5,6 Investiga-
tions of the isopropyl ether region of 1 demonstrated
that while sterically compact alkoxy substituents
attached to the C-4 position of the phenyl ring were
required for good ACC2 selectivity, a variety of larger,
more flexible hydrophobic ether groups produced excel-
lent, though nonselective, ACC2 potency (IC50s ꢀ 4–
10 nM).6 These findings encouraged us to further probe
structure–activity relationships (SAR) in the aryloxy re-
gion of the lead template in an effort to identify modifi-
cations that impart selectivity against the ACC1 isoform
yet preserve or enhance established, potent ACC2 activ-
ity. In line with these objectives, the present report
examines the effect of introducing C-2 phenyl ring sub-
stituents into our lead series of analogues (Fig. 1). In
Keywords: Acetyl-CoA carboxylase; Metabolic syndrome; Type-2
diabetes; Obesity.
*
Corresponding author. Tel.: +1 847 938 2509; fax: +1 847 938
3403; e-mail: rick.clark@abbott.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.01.022

1962
R. F. Clark et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1961–1965
2
for bromo, methyl, nitro, and aldehyde 2-substituted
O
S
analogues 7o, 7p, 7q, and 7t, respectively, initial depro-
tection of methyl ethers 3 with boron tribromide as well
as subsequent alkylation sequences were carried out
prior to attachment of alkynyl moiety 5. Reduction of
nitro intermediate 4 (X = NO2) with potassium borohy-
dride and CuCl followed by palladium-catalyzed cou-
pling to 5 produced 2-amino derivative 7r, which was
then acylated to give 2-acetamide product 7s. Similarly,
straightforward modifications of aldehyde 7t enabled
the introduction of a variety of additional position 2
substituents (Scheme 2). Condensation of 7t with
hydroxylamine hydrochloride followed by activation
and dehydration of the resulting oxime (7u) using meth-
anesulfonyl chloride–pyridine provided nitrile analogue
7v. In turn, reductive transformations of 7t provided
hydroxymethyl (7x) and dimethyl amino (7y) targets,
while Peterson olefination yielded vinyl product 7w as
outlined. Phenol precursors required in the synthesis of
compounds 7o and 7p were obtained in a convergent
manner from 4-methoxyphenol as depicted in Scheme 3.
Accordingly, bromination of 4-methoxyphenol pro-
duced 2-bromophenol 2o, which was converted over a
three-step reaction sequence to 2-methylphenol 2p.
Thus, lithiation of the methoxymethyl ether of 2o
with n-butyllithium at À78 °C followed by a methyl
iodide quench and subsequent acidic hydrolysis of the
MOM protecting group afforded 2-methylphenol 2p in
good overall yield. All other phenols 2 represented
in Scheme 1 were obtained from commercial sources.
3-alkoxy compounds 8a and 8b were derived from resor-
cinol and 2-methylresorcinol, respectively, according to
reaction sequences previously described.6
4
NH
N
O
O
1
ACC1 IC50 = >30 μM
ACC2 IC50 = 0.019 μM
X
O
S
NH
R
N
O
O
Figure 1. Lead ACC inhibitor 1 and ring substitution strategy.
the absence of structural information, this approach was
attractive since it enabled the investigation of unex-
plored substituent vectors and physicochemical parame-
ters as potential avenues for improved profiles. An
additional advantage of this course of SAR was the like-
lihood that C-2 ring substituents would alter the confor-
mation of the phenyl moiety to varying degrees relative
to neighboring domains in the parent scaffold allowing
us to assess the consequences of such variations.7
C-2 functionalized derivatives 7g–7y were synthesized
according to methodologies described in Schemes 1–3.8
These protocols represent adaptations of chemistry pre-
viously employed to prepare corresponding unsubstitut-
ed compounds 1 and 7a–7f.6 As shown in Scheme 1,
selective displacement of 2,5-dibromothiazole with read-
ily available phenols 2 provided intermediates 3, which
were elaborated to final products 7 utilizing two related
reaction routes. All 2-chloro compounds 7g–7m as well
as 2-fluoro analogue 7n were prepared from 3 by initial
Sonogashira coupling to propargylic acetamide 59 fol-
lowed by deprotection and alkylation of the resulting
phenols under Mitsunobu conditions. Alternatively,
As a starting point in our phenyl ring substitution study,
2-position chloro groups were incorporated into both
lead structure 1 as well as other series analogues con-
taining a variety of C-4 alkoxy adducts previously deter-
mined to be potent but nonselective ACC inhibitors.6 As
Br
S
X
X
X
Br
O
S
OH
b, c
N
O
S
Br
Br
N
N
O
(X = Br, Me,
CHO, NO2)
O
O
a
2
4
3
X = NO2
X = NH2
X = F, Cl, Br,
Me, CHO, NO2
d
O
(X = F, Cl)
e
N
H
5
5
e
X
O
O
S
R
X
X
NH
N
b, g
O
O
S
S
O
NH
NH
N
R
N
8a: X = H, R =
8b: X = Me, R =
i
-Bu
-Bu
O
O
O
O
i
7g–7t
6
X = NH2
X = NHAc
f
Scheme 1. Reagents and conditions: (a) K2CO3, DMF, 120–140 °C, 56–90%; (b) BBr3, CH2Cl2, À78 °C to rt, 42–85%; (c) Method A (X = Br, Me):
i-BuOH, DEAD, PPh3, THF, rt, 79–91%; Method B (X = CHO, NO2): i-BuBr, KI, K2CO3, DMF, 80 °C, 82–93%; (d) KBH4, CuCl, MeOH, 0 °C to
rt, 39%; (e) Pd(PPh3)2Cl2 (cat.), CuI (cat.), Et3N, THF, 75 °C, 19–83%; (f) acetyl chloride, Et3N, CH2Cl2, 0 °C to rt, 63%; (g) ROH, DEAD, PPh3,
THF, rt, 44–63%.

R. F. Clark et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1961–1965
1963
OH
N
N
O
O
a
O
S
O
S
b
e
S
NH
NH
NH
N
N
d
N
O
O
O
O
O
O
7t
7u
7y
c
HO
CN
O
O
O
S
S
S
NH
NH
NH
N
N
N
O
O
O
O
O
O
7v
7w
7x
Scheme 2. Reagents and conditions: (a) hydroxylamine hydrochloride, EtOH, H2O, 50 °C, 52%; (b) dimethylamine hydrochloride, NaBH(OAc)3,
NaOAc, HOAc, MeOH, rt, 37%; (c) methanesulfonyl chloride, pyridine, rt, 60%; (d) Me3SiCH2MgCl, Et2O, 0 °C to rt, 2 h then 30% aq H2SO4, 21%;
(e) NaBH4, MeOH, À78 to 0 °C, 62%.
OH
Br
a
2-halophenyl analogues was responsible for their diver-
gent selectivities. Rather, the increased steric demands
associated with 2-position chloro and bromo groups rel-
ative to a 2-fluoro substituent, were believed to be
incompatible with the ACC1 isozyme. These conclu-
sions are supported by the activity profile of 2-methyl
analogue 7p, which possesses steric and hydrophobic
features comparable to 2-chloro variant 7h but would
be expected to impart differing electronic effects. The
fact that the 2-chloro (7h) and 2-methyl (7p) compounds
were similarly potent and selective ACC2 inhibitors,
whereas the corresponding 2-fluoro and 2-unsubstituted
derivatives (7n and 7a, respectively) were nonselective,
can therefore reasonably be ascribed to steric features.
Likewise, the activity profiles for analogues incorporat-
ing relatively polar 2-position cyano (7v) and nitro (7q)
groups, respectively, revealed that these substituents
were also strongly disfavored by the ACC1 isoform yet
were reasonably well tolerated by the ACC2 isozyme.
Similarly, double bond-containing substituents exempli-
fied by vinyl (7w), aldehyde (7t), and oxime (7u) C-2
derivatives were devoid of ACC1 activity, although
these analogues also displayed reduced ACC2 potency.
Methyl alcohol entry 7x was also less active against
ACC2. Interestingly, given the relative compactness of
the 2-amino group in compound 7r, the poor overall
activity exhibited by this analogue indicates a lack of
tolerance by the ACC2 isozyme for the basic character
of the nitrogen adduct. More bulky acetamide (7s) and
dimethylamino (7y) moieties in the C-2 position were
also disfavored by both ACC isoforms.
OH
O
O
2o
b
c, d
OH
O
O
O
O
Br
2p
Scheme 3. Reagents and conditions: (a) bromine, CHCl3, 0 °C to rt,
74%; (b) NaH, DMF, rt, 30 min then chloromethyl methyl ether, 0 °C
to rt , 71%; (c) n-butyllithium (2.5 M in hexanes), Et2O, À78 to 0 °C,
30 min then methyl iodide, À78 °C to rt, 94%; (d) 12 N HCl, MeOH,
rt, 89%.
shown in Table 1, a comparison of activities between 1
and it’s 2-chloro counterpart 7g revealed that, other
than a modest reduction in ACC2 potency, the 2-chloro
modification was well tolerated and did not appear to
significantly alter the favorable overall profile of the par-
ent structure. However, while both of these analogues
had ACC1 IC50 values exceeding 30 lM, closer inspec-
tion of the ACC1 dose–response curves showed that
compound 1 and the 2-chloro variant 7g exhibited
35% and 7% inhibition, respectively, at 30 lM, suggest-
ing a selectivity-enhancing effect of 2-chloro substitu-
tion. Consistent with these observations, application of
the 2-chloro modification to potent nonselective deriva-
tives 7a–7f was accompanied by dramatic reductions in
ACC1 activity while the ACC2 potency was uniformly
preserved, within 1- to 3-fold, in all contexts investigated
(7g–7m).
Taken together, these data highlight contrasting toler-
ances for C-2 phenyl ring substitution between the
ACC1 and ACC2 isozymes. Steric properties of position
2 adducts appear to be critical parameters for achieving
ACC2 selectivity in the current series. Smaller, hydro-
phobic methyl and chloro groups were found to be
optimal substituents for retaining excellent ACC2
potency and achieving high degrees of selectivity against
ACC1. It is unclear whether the 2-substituent is directly
involved in an unfavorable steric interaction with
the ACC1 isozyme or if local conformational changes
Following the promise of these initial results we per-
formed a more systematic evaluation of C-2 substitution
in the context of 4-isobutoxy inhibitors 7a and 7h.
2-Fluoro entry 7n displayed nonselective, potent activity
resembling that of the unsubstituted parent analogue
(7a), while 2-bromo derivative 7o exhibited highly
selective ACC2 activity approximating that of 2-chloro
compound 7h. We felt it unlikely that the difference in
hydrophobic and electronic properties between these

1964
R. F. Clark et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1961–1965
Table 1. ACC inhibitory activitya for derivatives 1, 7, and 8
Compound
R
X
ACC1 IC50(lM)
ACC2 IC50(lM)
ACC1/ACC2b
1
i-Pr
i-Bu
H
>30
0.23
0.140
0.94
0.12
0.086
0.76
>30
>30
>30
>30
>30
>30
>30
0.34
>30
>30
>30
>30
>30
>30
>30
>30
>30
>30
>30
0.072
>30
0.019
0.010
0.004
0.007
0.016
0.027
0.067
0.069
0.012
0.013
0.010
0.034
0.029
0.14
>1500
23
35
7a
7b
7c
7d
7e
7f
H
Pr
CH2(cyclopropyl)
H
H
134
8
Cyclohexyl
CH2(cyclohexyl)
CH2(THF-3-yl)c
i-Pr
H
H
3
11
H
Cl
7g
7h
7i
>435
>2500
>2300
>3000
>880
>1030
>210
31
i-Bu
Pr
Cl
Cl
7j
CH2(cyclopropyl)
Cyclohexyl
Cl
Cl
7k
7l
CH2(cyclohexyl)
Cl
Cl
7m
7n
7o
7p
7q
7r
7s
7t
CH2(THF-3-yl)c
i-Bu
i-Bu
F
Br
0.011
0.035
0.009
0.070
15.80
>30
>855
>3300
>425
>1
i-Bu
i-Bu
Me
NO2
NH2
NHCOMe
CHO
CH@NOH
CN
i-Bu
i-Bu
1
i-Bu
i-Bu
0.55
0.57
>55
>50
>710
>115
>30
1
7u
7v
7w
7x
7y
8a
8b
i-Bu
i-Bu
0.042
0.26
1.00
Vinyl
CH2OH
CH2NMe2
H
i-Bu
i-Bu
>30
0.013
0.018
i-Bu
i-Bu
55.00
>1665
Me
a Inhibitory activity was determined using recombinant human ACC1 and ACC2 in an assay measuring ACC-mediated [14C]CO2 incorporation into
malonyl-CoA. Detailed protocols are described in Ref. 5.
b ACC2 selectivity expressed as rounded ratio of ACC1 IC50/ACC2 IC50
c Tetrahydrofuran (THF).
.
imposed on the phenoxy thiazolyl lead scaffold are
responsible for the observed ACC2 selectivity-enhancing
effects. In any case, the ACC1 isoform is more sensitive
to C-2 substitution than is the ACC2 isozyme and the
introduction of groups larger than hydrogen or fluorine
was shown to abolish ACC1 activity.
example, effectively serve as ACC2 selectivity ‘switches’
that impart high selectivity against ACC1 yet conserve
ACC2 potency in otherwise nonselective contexts. The
application of this strategy has resulted in enhanced
overall profiles with respect to both ACC2 selectivity
(>3000-fold) and ACC2 potency (single-digit nanomo-
lar) in the current lead series. Furthermore, the selectiv-
ity-enhancing effects associated with these modifications
were shown to be general across a range of structural
analogues.
Given the remarkable selectivity-modulating effects of
C-2 ring substitution within the 4-alkoxy inhibitor ser-
ies, the 2-methyl analogue (8b) of 3-isobutoxy inhibitor
8a was also prepared. We previously demonstrated that
a series of 3-alkoxy derivatives exemplified by 8a charac-
teristically exhibit potent, though nonselective, ACC
inhibition.6 As shown in Table 1, incorporation of a
2-methyl group (8b) in this context also resulted in a loss
of ACC1 activity and retention of ACC2 potency. These
findings suggest that C-2 substituents may provide
general selectivity-enhancements in our lead series with
the potential for improving ACC2 selectivity profiles
regardless of structural context.
References and notes
1. (a) Harwood, H. J., Jr. Expert Opin. Ther. Targets 2005, 9,
267; (b) Tong, L. Cell Mol. Life Sci. 2005, 62, 1784; (c)
Harwood, H. J., Jr. Curr. Opin. Invest. Drugs 2004, 5, 283;
(d) Burn, P.; Shi, Y. Nat. Rev. Drug Disc. 2004, 3, 695.
2. (a) Unger, R. H. Endocrinology 2005, 144, 5159; (b) Lazar,
M. A. Science 2005, 307, 373; (c) Hulver, M. W.; Berggren,
J. R.; Cortright, R. N.; Dudek, R. W.; Thompson, R. P.;
Pories, W. J.; MacDonald, K. G.; Cline, G. W.; Shulman,
G. I.; Dohm, G. L.; Houmard, J. A. Am. J. Physiol.
Endocrinol. Metab. 2003, 284, E741; (d) Friedman, J.
Nature 2002, 415, 268; (e) Arner, P. Diabetes Metab. Res.
Rev. 2002, 18, S5.
3. (a) Oh, W.; Abu-Elheiga, L.; Kordari, P.; Gu, Z.; Shaike-
nov, T.; Chirala, S. S.; Wakil, S. J. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 1384; (b) Treadway, J. L.; McPherson, R.
K.; Petras, S. F.; Shelly, L. D.; Frederick, K. S.; Sagawa,
K.; Perry, D. A.; Harwood, H. J., Jr. 64th Annual Meeting
In conclusion, we have shown that the introduction of
C-2 phenyl ring substituents is an effective strategy for
modulating isozyme selectivity profiles in the current
ACC inhibitor series. The ACC1 isoform was found to
be generally intolerant to ring substitution, whereas a
variety of relatively hydrophobic, sterically compact
substituents were well tolerated by the ACC2 isoform.
Consequently, 2-position chloro or methyl groups, for

R. F. Clark et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1961–1965
1965
and Scientific Sessions of the American Diabetic Asssocia-
tion, Orlando, FL, June 4–8, 2004; (c) Harwood, H. J., Jr.;
Petras, S. P.; Shelly, L. D.; Zaccaro, L. M.; Perry, D. A.;
Makowski, M. R.; Hargrove, D. M.; Martin, K. A.; Tracey,
W. R.; Chapman, J. G.; Magee, W. P.; Dalvie, D. K.;
Soliman, V. F.; Martin, W. H.; Mularski, C. J.; Eisenbeis, S.
A. J. Biol. Chem. 2003, 278, 37099; (d) Abu-Elheiga, L.; Oh,
W.; Kordari, P.; Wakil, S. J. Proc. Natl. Acad. Sci. U.S.A.
2003, 100, 10207; (e) Abu-Elheiga, L.; Matzuk, M. M.; Abo-
Hashema, K. A. H.; Wakil, S. J. Science 2001, 291, 2613.
4. Abu-Elheiga, L.; Matzuk, M. M.; Kordari, P.; Oh, W.;
Shaikenov, T.; Gu, Z.; Wakil, S. J. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 12011.
6. Clark, R. F.; Zhang, T.; Xin, Z.; Liu, G.; Wang, Y.;
Hansen, T. M.; Wang, X.; Wang, R.; Zhang, X.; Frevert, E.
U.; Camp, H. S.; Beutel, B. A.; Sham, H. L.; Gu, Y. G.
Bioorg. Med. Chem. Lett. 2006, 16, 6078.
7. (a) Rovnyak, G.; Andersen, N.; Gougoutas, J.; Hedberg,
A.; Kimbal, S. D.; Malley, M.; Moreland, S.; Porubcan,
M.; Pudzianowski, A. J. Med. Chem. 1991, 34, 2521; (b)
Green, A. L.; Marshall, I. G. Nature 1970, 228, 1211; (c)
Clark, E. R.; Dawes, P. M.; Williams, S. G. Br. J. Pharma-
col. Chemother. 1968, 32, 113.
8. Final compounds were isolated using either flash chroma-
tography on silica gel or reverse-phase HPLC on a Waters
Sunfire C18 column with a gradient of 5–95% acetonitrile:
0.1% aqueous trifluoroacetic acid. All analogues were in full
5. Gu, Y. G.; Weitzberg, M.; Clark, R. F.; Xu, X.; Li, Q.;
Zhang, T.; Hansen, T. M.; Liu, G.; Xin, Z.; Wang, X.;
Wang, R.; McNally, T.; Camp, H.; Beutel, B. A.; Sham, H.
L. J. Med. Chem. 2006, 49, 3770.
1
agreement with proposed structures by H NMR, HPLC,
and MS (>95% purity).
9. Gardner, J. N. Can. J. Chem. 1973, 51, 1416.