Design of small molecule inhibitors of acetyl-CoA carboxylase 1 and 2 showing reduction of hepatic malonyl-CoA levels in vivo in obese Zucker rats

Article abstract of DOI:10.1016/j.bmc.2011.04.014

Inhibition of acetyl-CoA carboxylases has the potential for modulating long chain fatty acid biosynthesis and mitochondrial fatty acid oxidation. Hybridization of weak inhibitors of ACC2 provided a novel, moderately potent but lipophilic series. Optimization led to compounds 33 and 37, which exhibit potent inhibition of human ACC2, 10-fold selectivity over inhibition of human ACC1, good physical and in vitro ADME properties and good bioavailability. X-ray crystallography has shown this series binding in the CT-domain of ACC2 and revealed two key hydrogen bonding interactions. Both 33 and 37 lower levels of hepatic malonyl-CoA in vivo in obese Zucker rats.

Full text of DOI:10.1016/j.bmc.2011.04.014

Bioorganic & Medicinal Chemistry 19 (2011) 3039–3053
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Design of small molecule inhibitors of acetyl-CoA carboxylase 1 and 2 showing
reduction of hepatic malonyl-CoA levels in vivo in obese Zucker rats
Christoffer Bengtsson, Stefan Blaho, David Blomberg Saitton, Kay Brickmann, Johan Broddefalk,
Öjvind Davidsson, Tomas Drmota, Rutger Folmer, Kenth Hallberg, Stefan Hallén, Ragnar Hovland,
Emre Isin, Petra Johannesson, Bengt Kull, Lars-Olof Larsson, Lars Löfgren, Kristina E. Nilsson,
⇑
Tobias Noeske, Nick Oakes, Alleyn T. Plowright , Volker Schnecke, Pernilla Ståhlberg,
Pernilla Sörme, Hong Wan, Eric Wellner, Linda Öster
AstraZeneca Research and Development, Pepparedsleden 1, Mölndal 43183, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Inhibition of acetyl-CoA carboxylases has the potential for modulating long chain fatty acid biosynthesis
and mitochondrial fatty acid oxidation. Hybridization of weak inhibitors of ACC2 provided a novel, mod-
erately potent but lipophilic series. Optimization led to compounds 33 and 37, which exhibit potent inhi-
bition of human ACC2, 10-fold selectivity over inhibition of human ACC1, good physical and in vitro
ADME properties and good bioavailability. X-ray crystallography has shown this series binding in the
CT-domain of ACC2 and revealed two key hydrogen bonding interactions. Both 33 and 37 lower levels
of hepatic malonyl-CoA in vivo in obese Zucker rats.
Received 3 February 2011
Revised 6 April 2011
Accepted 9 April 2011
Available online 13 April 2011
Keywords:
Acetyl-CoA carboxylase
Malonyl-CoA
Ó 2011 Elsevier Ltd. All rights reserved.
Ligand lipophilicity efficiency
Quinoline
1. Introduction
The carboxylated biotin then enters the CT-domain via a narrow
channel where carbon dioxide is transferred from the biotin carbox-
Obesity and type 2 diabetes mellitus are characterized as world-
wide epidemics with associated global deaths of up to 5%.1 The rate
of developing obesity is alarming considering the connection be-
tween a body mass index above 30 and diseases such as diabetes,
hypertension, artery cardiovascular disease, stroke, and even some
forms of cancer. There is a strong demand for drugs capable of
treating both obesity and type 2 diabetes mellitus. One biological
principle that has been considered for treating these diseases are
the enzymes acetyl-CoA carboxylases (ACCs.a: EC_6.4.1.2; ACC1
(ACACA); ACC2 (ACACB)).2 ACCs have pivotal functions in fatty acid
metabolism and inhibition of ACCs gives the potential for modulat-
ing both long chain fatty acid biosynthesis and mitochondrial fatty
acid oxidation. This modulation presents opportunities for treat-
ment of both diabetes and obesity.
ACC is a biotin-dependent protein, composed of a carboxyltrans-
ferase (CT), biotin carboxy carrier protein, and biotin carboxylase
(BC) domains, which synthesises malonyl-CoA from acetyl-CoA in
an ATP-dependent reaction via the fixation of bicarbonate.3 The
synthesis of malonyl-CoA is a two step process where the first reac-
tion occurs in the BC-domain, upon which the biotin is carboxylated.
ylate to acetyl-CoA to form malonyl-CoA.
Mammalian ACC is known to exist in two isoforms, ACC1 and
ACC2. At energy surplus, ACC1 converts acetyl-CoA into malonyl-
CoA for de novo lipogenesis in the cytosol while the ACC2 isoform,
present on the mitochondrial surface, carries out the same reaction
to generate malonyl-CoA for the inhibition of carnitine palmitoyl
transferase 1 (CPT-1). The latter reaction constitutes the main reg-
ulation of mitochondrial import of fatty acids for b-oxidation. Inhi-
bition of ACCs will cause a switch in cellular substrate handling
and favour lipid oxidation, thus preventing deleterious lipids from
accumulating in oxidative tissue such as muscle, heart and liver.
This reduction in intracellular saturated fatty acids and down-
stream lipids such as diacylglycerols and ceramide is expected to
improve insulin sensitivity. Additionally, the achieved change of li-
pid flux has the potential to influence whole body energy balance
towards an increase of energy expenditure and also correction of
dyslipidaemia. In the clinic, failure to maintain weight loss is asso-
ciated with an impairment of fatty acid oxidation and reduced en-
ergy expenditure. These metabolic changes represent a major
hurdle in the pharmacotherapy of obesity. Moreover, intramyocel-
lular accumulation of lipids and elevated malonyl-CoA levels are
events associated with obesity, leading to reduced oxidative capac-
ity and attenuated insulin sensitivity.4–6
⇑
Corresponding author.
E-mail address: alleyn.plowright@astrazeneca.com (A.T. Plowright).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.04.014

3040
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
It has been reported that mice with a global deletion in the
sequentially: 1/intravenous rat PK studies, to obtain in vivo clear-
ance (CL) data; 2/intravenous target engagement studies in obese
Zucker rats; 3/oral PK experiments in mice to assess oral absorp-
tion/exposure prior to testing in a chronic efficacy study (results
not included in the present study). This DMPK data is described
for key compounds and the best compounds, 33 and 37, are shown
to provide a good reduction in levels of hepatic malonyl-CoA
in vivo in obese Zucker rats.
ACC2 gene have lower body weight, less fat tissue, lower levels
of insulin and improved glucose tolerance after long-term expo-
sure to a high fat/high carbohydrate diet, as well as higher rates
of fat and carbohydrate oxidation resulting in an increased energy
expenditure.7 Other published data supporting the benefits of ACC
suppression derive from a gene silencing approach using specific
antisense oligonucleotides to knock-down hepatic ACC1 and/or
ACC2. The best result was achieved with a combined ACC1/ACC2
knock-down which caused a significant reduction of hepatic malo-
nyl-CoA levels accompanied by lower liver triglycerides and im-
proved hepatic insulin sensitivity.8 These results would support
the potential for an ACC1/2 isozyme non-selective inhibitor to se-
cure a desirable metabolic profile with increased insulin sensitiv-
ity, correction of dyslipidaemia and weight maintenance, as seen
in the published knock-out mice. However, there has also been
additional research highlighting the complexity and uncertainty
regarding the level of selectivity required between the two iso-
forms. It has been reported that ACC1 knock-out mice show
embryonic lethality suggesting that a selective ACC2 inhibitor
would be preferred.9 On the other hand, in 2010 Olson et al. pub-
lished, contrary to previous results, that mice with either a global
deletion of ACC2 or selective deletion of ACC2 from skeletal muscle
has no effect on body weight, fat mass, food intake or glucose
homeostasis suggesting that selective ACC2 inhibition may be inef-
fective.10 Inspired by the potential of the target for treating obesity
and type 2 diabetes mellitus a search for potent inhibitors of ACC1
and 2 was initiated. Due to the uncertainty in the literature regard-
ing the impact of inhibiting ACC1 a project was set up to drive on
both ACC1 and 2 but ACC2 was the primary data.
The interest in ACC as a target is clearly seen in the increase of
number of patent applications and publications in recent years.2,11
Inhibitors have been reported for both the BC-domain, including
the natural product Soraphen,12, and the CT-domain. Most have fo-
cused on the CT-domain (Fig. 1), including the piperidinylpiperi-
dines as exemplified by 1 (CP-640186),4 the spirochromanone
class of compounds such as 2,11e and the (4-piperidinyl)-pipera-
zines such as 3.11f To date, these compounds reported to bind in
the CT-domain are isozyme ACC1/2 non-selective inhibitors. Selec-
tive inhibitors of ACC2 have been reported with very different
chemotypes.11b,d However, it has not been reported where these
compounds bind.
2. Results and discussion
2.1. Chemistry
The synthesis of the trans-1,4-disubstituted cyclohexyl
derivatives 17 and 19–41 started from commercially available
tert-butyl {[trans-4-(aminomethyl)cyclohexyl]methyl}carbamate 4
(Scheme 1). Coupling of the amine 4 to the carboxylic acids 5 using
amide coupling conditions such as EDC/HOBT or TBTU gave the
chloro- or bromoamides 6. Suzuki cross coupling of 6 with the re-
quired boronic acid or boronic ester catalysed by either Pd(PPh3)4
or 1,3-bis(2,6-diisopropylphenyl)imidazol-2-ylidene]-(3-chloro-
pyridyl)-palladium(II)dichloride (PEPPSI) gave the derivatives 7.
One derivative, 25, was alkylated further on the pyrazole with 2-
dimethylaminoethylchloride hydrochloride to give 30. Alterna-
tively, the 2-chloroheteroaromatic derivatives 6 underwent a nucle-
ophilic aromatic substitution reaction with a range of amines to give
derivatives 8. One derivative, 32, acted as a useful intermediate and
could be modified by functional group interconversions to give
derivatives 33, 34 and 37 (Scheme 1). For example, oxidation of 32
under Swern conditions gave an aldehyde that underwent a reduc-
tive amination upon treatment with 3-hydroxyazetidine and
sodium triacetoxyborohydride to give 37. Variation of the carba-
mate was carried out by removing the tert-butylcarbamate group
in 8 under acidic conditions to provide a primary amine that, upon
treatment with either chloroformates or mixed 4-nitrophenylcar-
bonates,13 gave the alternative carbamate derivatives 9.
Alternatively, the required R3 substituent in 7 or the required
amine side chain in 8 was introduced to the carboxylic acid coupling
partner prior to reaction with the amine 4. In some cases the re-
quired compounds were commercially available whereas others
were synthesised by different methods (Scheme 2). Suzuki cross
coupling of the substituted 2-chloro- or 2-bromoheteroaromatic
compounds 5 with the required boronic acid or boronic ester cata-
lysed by either Pd(PPh3)4 or PEPPSI gave the coupling partners 10.
For example, compound 28, where R3 = 4-benzylamine was synthe-
sized via this method. Alternatively, 5 could undergo a nucleophilic
substitution reaction upon treatment with amines to give 10 where
R3 is an amino substituent. The 2-pyridin-4-ylquinoline-4-carbox-
ylic acid 10 (where R3 = 4-pyridyl) required to synthesise 24 was
prepared by treating 2,3-dihydro-indole-2,3-dione 11 with 4-ace-
tylpyridine and KOH.14 Finally, 6-phenyl-1H-pyrrolo[2,3-b]pyri-
dine-4-carboxylic acid required to prepare 21 was synthesised in a
Herein, we describe the structure–activity relationship and
optimization of a series of 1,4-disubstituted cyclohexanes to pro-
vide compounds with good potency against the human ACC2
(hACC2) enzyme, moderate 10-fold selectivity against human
ACC1 (hACC1) and good physical and ADME properties. To test
the concept of ACC inhibition in vivo, we chose to build overall
in vivo PK-PD understanding in a stepwise fashion, first establish-
ing target engagement in an acute experiment (lowering of
malonyl-CoA in obese Zucker rats) and then examining repeat dos-
ing effects in a disease model. To achieve this, we performed
O
O
O
O
N
O
N
N
N
N
N
O
N
O
O
N
O
O
N
CN
N
N
NH
N
H
2 (Pfizer, Compound 59)11(e)
3 (Taisho, Compound 13(i))11(f)
1 (CP-640186)4
Figure 1. Structures of ACC inhibitors binding to the CT-domain.

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3041
O
O
O
NH
O
NH
O
HO
R2
R1
O
(c) or (d)
(a) or (b)
O
NH
+
HN
R2
O
HN
R2
O
N
X
R1
N
R3
NH2
R1
N
X
X = Br, Cl
4
6
7,
5
Where R3 = 3-pyrazole (25)
Where
X = Cl
(e)
O
(f)
30
O
O
NH
Where R4 = CH2CH2OH, Y = C (32)
R4 = CH2CH2N(CH3)2, (33)
R4 = CH2CH2NH2, (34)
(g),(h)
(g),(i),(j)
(k),(l)
O
R
O
NH
(n)
HN
R2
R1
(o) or (p)
HN
R2
O
R4 = CH2CH2azetidin-3-ol (37)
Where R4 = CH2NHCbz, Y = C
R4 = CH2NH2 (35)
N
N
4
Y
R
R1
N
N
8, Y = C, N
(m)
4
Y
R
9
Scheme 1. Reagents and conditions: (a) N-methylmorpholine, EDC, HOBT, 2-MeTHF, rt; (b) TBTU, Et3N, CH2Cl2, rt; (c) R3B(OH)2, Pd(PPh3)4 K2CO3, dioxane, H2O, 60 °C; (d)
R3B(OC(CH3)2C(CH3)2O), PEPPSI, K2CO3, dioxane, H2O, 140 °C; (e) RxRyNH, pyridine, 135 °C; (f) Cs2CO3, (CH3)2NCH2CH2Cl, EtOH 140 °C; (g) CH3SO2Cl, DIPEA, CH2Cl2, rt; (h)
dimethylamine, DMF, rt; (i) NaN3, DMF, rt; (j) SnCl2, Et3N, PhSH, THF, rt; (k) oxalyl chloride, DMSO, Et3N, CH2Cl2, À78 °C to rt; (l) 3-hydroxyazetidine hydrochloride, sodium
triacetoxyborohydride, AcOH, EtOH, 120 °C; (m) 5% Pd/C, H2, rt; (n) (i) HCl, 1,4-dioxane, or (ii) TFA, CH2Cl2, rt; (o) EtOC(O)Cl, Et3N, THF, 0 °C; (p) 4-nitrophenyl tetrahydro-2H-
pyran-4-yl carbonate (42), Et3N, THF, 40 °C to rt.
O
O
(d)
N
H
HO
R2
R1
O
HO
R2
R1
O
11
(a) or (b)
or (c)
N
R3
N
X
HO
O
MeO
O
10
5, X = Br, Cl
(e), (f), (g)
(h), (i)
N
N
H
N
N
Cl
H
13
12
Scheme 2. Reagents and conditions: (a) R3B(OH)2, Pd(PPh3)4ÁK2CO3, dioxane, H2O, 60 °C; (b) R3B(OC(CH3)2C(CH3)2O), PEPPSI, K2CO3, dioxane, H2O, 140 °C; (c) RxRyNH,
pyridine, 135 °C; (d) 4-acetylpyridine, KOH, EtOH, 80 °C; (e) TMS–diazomethane, CH2Cl2/MeOH (3:1), rt; (f) m-CPBA, DME, rt; (g) ethyl carbonochloridate, HMDS, THF, 50 °C;
(h) R3B(OH)2, bis(di-tert-butylphosphino)ferrocene palladium dichloride, K2CO3, toluene, MeOH, 80 °C; (i) LiOH, 1,4-dioxane, H2O, 50 °C.
five step sequence starting from 1H-pyrrolo[2,3-b]pyridine-4-car-
boxylic acid 12. Esterification of 12, followed by oxidation with
meta-chloroperoxybenzoic acid (m-CPBA) and then treatment with
ethyl carbonochloridate and hexamethyldisilazane gave methyl
6-chloro-1H-pyrrolo[2,3-b]-pyridine-4-carboxylate 13.15 Cross
coupling of 13 with phenylboronic acid using bis(di-tert-buty-
lphosphino)-ferrocene palladium dichloride as catalyst followed
by hydrolysis of the methyl ester with lithium hydroxide gave the
required 6-phenyl-1H-pyrrolo[2,3-b]pyridine-4-carboxylic acid 10
(where R1 and R2 make up a pyrrole ring and R3 = phenyl). Once
the required carboxylic acids 10 were in hand they were coupled
to the amine 4 under the conditions described in Scheme 1. The
cis-1,4-disubstituted cyclohexane derivative 18 was synthesized
from the same sequence as shown in Scheme 1 but starting from
tert-butyl {[cis-4-(aminomethyl)cyclohexyl]methyl}carbamate, the
synthesis of which is described in Supplementary data.16
2.2. Discovery of lead compounds
A high throughput screen of the AstraZeneca compound collec-
tion provided several hit series feasible for further exploration.
Two of the series, represented here by compounds 14 and 15, were
chosen for further work (Fig. 2). Both series had moderate potency
versus the hACC2 enzyme and high lipophilicity as measured by
logD.17 Hybridizing compounds 14 and 15 provided either the
sulfonamide 16 or the amide 17. Both compounds showed an
improvement in inhibition of hACC2, particularly the amide
derivative 17 with a 10-fold improvement in enzyme inhibition.

3042
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
O
O
O
NH
O
NH
O
O
N
N
O
O
O
HN
HN
X
N
S
N
15: hACC2 IC50 = 3.6 µM
16: X = SO2; hACC2 IC50 = 0.69 µM
17: X = CO; hACC2 IC50 = 0.21 µM
14: hACC2 IC50 = 2.5 µM
18: X = CO, cis-1,4-cyclohexyl core;
hACC2 IC50 = 1.0 µM
Figure 2. High throughput screening hits and hybridized cis- and trans-cyclohexyl derivatives.
However, this compound suffered from high lipophilicity
(log D = 5.9) leading to poor physical properties such as low solu-
bility (<1 lM) and high human plasma protein binding (fraction
shown in Table 1. Replacement of the quinoline with a pyridine
ring as in 19 led to a large drop in potency versus hACC2. Adding
an additional methyl group to the pyridyl ring (20) failed to bring
back any potency and converting the pyridine in 20 to a 1,3-pyrim-
idine resulted in significantly weaker inhibition of hACC2 (data not
unbound = 0.1%). It was considered critical to improve these
parameters at this early stage to enable effective testing of com-
pounds throughout the project screening cascade and to achieve
the required free compound concentrations to give an opportunity
to provide significant in vivo effects. Amide 17 had good perme-
ability as measured in the caco-2 assay (Papp A–B 22 Â 10À6 cm s
À1) but underwent rapid oxidative metabolism leading to high
intrinsic CL in vitro in human liver microsomes (HLM,
Table 1
Quinoline replacements and substitutions
O
O
NH
CLint = 124 l
L/min/mg).18 It is known that high plasma protein
binding can mask a compound from CL lowering its in vivo CL.
However, dosing of 17 intraveneously in rat in vivo DMPK experi-
ments gave rapid CL (60 mL/min/Kg). When the plasma protein
binding in rat (fraction unbound = 0.5%) was taken into account
the unbound CL (CLu) in vivo was very high (12,000).19 Hence,
the focus was to improve physical properties including solubility
and plasma protein binding and reduce both the intrinsic meta-
bolic instability and CLu of the compounds in vivo whilst maintain-
ing or improving potency versus hACC2 to enable effective testing
of ACC inhibition in vivo.
The role of lipophilicity has been widely reported in the litera-
ture as an important parameter for designing compounds to im-
prove both physical and ADME properties.20 Clearly it is of
paramount importance to also take into account the potency of
the compounds and therefore the challenge is to find ligands that
are sufficiently potent in the optimal lipophilicity range. As a result
the use of the composite parameter ligand lipophilicity efficiency
(LLE), calculated as pIC50–log D was used.21 LLE is a measure of
how efficiently a compound uses its lipophilicity to drive potency
and is used as a means to assess compound efficiency and help
analyse SAR within a series. Given the high lipophilicity of this ser-
ies and the associated issues the project strategy was to focus on
reducing lipophilicity whilst maintaining or improving potency
versus hACC2 and therefore LLE was used to assess progress within
the series. Given the improved potency and lower lipophilicity of
17 (LLE = 0.8) compared to 16 (LLE = À0.5), chemistry was focused
on the amide sub-series. The preferred configuration of the 1,4-
disubstituted cyclohexyl ring was also tested. Compound 17 was
in the trans-configuration, whereas the cyclohexyl derivative in
the cis-configuration, 18, showed a fivefold drop in potency versus
the hACC2 enzyme (Fig. 2). Therefore all further work was focused
on the trans-1,4-disubstituted cyclohexyl core.
HN
O
R
Compd
R
hACC2a
4.5
(l
M)
hACC1a
6.6
(lM)
Log D
LLE
0.6
19
4.7
N
20
5.8
8.3
5.3
0
N
21
22
23
0.78
3.8
ND
10
4.7
4.9
5.0
1.5
0.5
0.5
N
N
H
N
Cl
2.9
ND
N
O
N
24
25
26
0.21
0.57
1.3
1.1
2.3
13
4.4
3.5
2.9
2.3
2.7
3.0
N
N
N
NH
N
N
N
The phenylquinoline of compound 17 contributes significantly
to the overall lipophilicity of the compounds. Hence, the initial fo-
cus was to vary this part of the molecule to understand its impact
on the inhibition of hACC2. Some of the variations attempted are
NH
a
The IC50 values are reported as means of at least two separate determinations
with a typical standard deviation of less than 30%. ND means not determined.

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3043
Table 2
shown). Attempts to replace the quinoline ring with alternative,
Substitution of phenyl and heterocyclic ring side chains
less lipophilic bicyclic heterocyclic systems such as pyrrolopyri-
dine 21 also led to a reduction of inhibition of hACC2. With no
success in removing or replacing the quinoline ring system the
strategy turned to replacing the phenyl ring attached in the
2-position. Initial attempts to replace the phenyl ring with smaller,
less lipophilic substituents such as chloro (22) or methoxymethyl
(data not shown) also led to a significant drop in potency. An alter-
native way to reduce the overall lipophilicity was to replace the
phenyl ring with less lipophilic heterocycles. However, many of
the heterocyclic phenyl replacements tested gave a significant
reduction in potency (e.g., isoxazole 23). Only two replacements
were more interesting; the pyridyl derivative 24 was equipotent
with 17 and the pyrazole derivative 25 gave a ꢀ3-fold reduction
in potency compared to 17. Whilst potency was not improved,
both 24 and 25 were much less lipophilic compared to 17. Hence,
despite the knowledge that the potency would need to be
improved, the LLE of these derivatives was improved. Finally,
replacing the phenyl ring with piperazine to give 26, also gave a
reduction in potency but also a significant reduction in log D lead-
ing to an improvement in LLE. Despite the reduction in lipophilicity
of many of these compounds they still had poor physical properties
and low metabolic stability. For example, the pyridyl derivative 24
O
O
NH
HN
O
N
R
Compd
R
hACC2a
0.052
(
lM)
hACC1a
(l
M)
Log D
LLE
NH2
27
0.24
3.4
3.9
3.3
28
29
0.33
0.75
1.7
2.9
3.2
2.5
NH2
OH
3.6
3.0
O
N
N
30
0.24
1.3
3.6
N
N
had low solubility (<1
146 L/min/mg).
lM) and low stability in HLM (CLint =
N
31
32
33
0.73
0.23
0.11
5.3
2.3
1.2
3.4
4.5
3.6
2.8
2.1
3.3
l
N
All attempts to replace the phenylquinoline moiety, with the
exception of 24, resulted in a loss of potency. Despite an increase
in LLE for some derivatives, a boost in potency and a significant
improvement in both physicochemical and CL properties was re-
quired. Additional studies, including NMR studies of ACC2 in a
competitive binding study, revealed that these compounds bind
reversibly to the CT-domain in the same site reported in the pub-
lished X-ray structures of 1 (data not shown).22,23 It was hypothe-
sized that the phenylquinoline moiety of 17 occupied the same
area of space as the anthracene ring system in 1. Further analysis
of the X-ray structure of 1 revealed that the anthracene group is
oriented close to the hydrophilic surface of the protein. This pre-
sented an opportunity to substitute the phenyl ring of 17 with
small polar substituents in an attempt to pick up interactions with
the more hydrophilic exterior of the protein to increase potency
whilst simultaneously reducing the lipophilicity of the compounds
to improve the overall properties. Introduction of a small, basic
side chain by replacing the phenyl ring with a benzylamine group
showed positive results, especially with the 3-substituted deriva-
tive 27 giving a good improvement in potency (Table 2). Com-
pound 27 showed a major improvement in stability in HLM
(Table 3). However, the unbound clearance in vivo was still very
high. Also, despite the significant decrease in log D, the physico-
chemical properties of 27 remained poor. Compound 27 also
showed significant inhibition of the cytochrome P450 isoforms
N
N
OH
N
N
34
35
0.089
0.69
0.78
4.3
3.2
3.2
3.8
3.0
NH2
N
N
NH2
O
36
37
0.63
4.5
1.0
3.0
3.6
3.2
3.5
N
N
0.099
N
OH
a
The IC50 values are reported as means of at least two separate determinations
2C9, 3A4 and 1A2 (IC50 values of 0.8, 2.4 and 4 lM, respectively).
with a typical standard deviation of less than 30%. ND means not determined.
Attempts were also made to introduce acidic side chains such as
in 29. However, this led to a reduction in potency versus hACC2.
As mentioned, replacement of the phenyl ring in 17 with a
pyrazole or piperazine led to a decrease in potency but an improve-
ment in the LLE due to the significant reduction in log D compared
with the starting phenyl derivative 17. Based on the increase in
potency achieved with 27 it was hypothesized that introducing
small basic side chains to both the pyrazole 25 and the piperazine
derivative 26 could improve the inhibition of hACC2. Hence, the
pyrazole and piperazine derivatives, 30 and 31, were prepared
but only gave a ꢀ2-fold improvement in potency against hACC2
compared with the unsubstituted compounds 25 and 26, respec-
tively (Table 2). However, the properties of 31 were considerably
improved compared with both 17 and 27 (good solubility
ity against a panel of cytochrome P450 isoforms (1A2, 2C9, 2C19,
2D6, 3A4; <50% inhibition at 13 lM)). Replacement of the pipera-
zine ring in 31 with a piperidine led to 33, which gave an improve-
ment in potency whilst maintaining the log D. Furthermore, 33
again displayed high solubility, good stability in HLM and also high
permeability in the caco-2 assay (Table 3). Compound 33 also had
significantly reduced human plasma protein binding and a much
reduced CLu in vivo in rat so overall a significant improvement
compared with both 17 and 27. Attempts were made to reduce
the molecular weight by preparing the primary amine derivative
34 and this compound again displayed good inhibition of hACC2
(IC50 = 0.089
lM). However, the introduction of two additional
(73 lM), excellent stability in HLM (<12 lL/min/mg) and no activ-
hydrogen bond donors significantly reduced the permeability

3044
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
Table 3
Physical property, in vitro ADME and in vivo rat clearance data for 17, 27, 33 and 37
Compd
Solubilitya
(
lM)
Fraction unbounda (Hu, (%))
HLMa
(
lL/min/mg)
Caco-2 Papp A–Ba (cm sÀ1
)
Rat CLb (mL/min/Kg)
Rat CLuc
17
27
33
37
<1
8
79
80
0.1
1.4
3.4
3.3
124
14
14
22
13
9.4
1.4
60
104
32
12,000
6120
800
16
70
2120
a
b
c
See Supplementary data for assay details.
Compounds dosed intravenously at 3–5
Rat unbound clearance defined as rat CL in vivo/fraction unbound (rat).19
l
mol/Kg to Sprague–Dawley rats.
compared to 33 (Papp A–B 0.5 Â 10À6 cm sÀ1).24 Attempts to reduce
the length of the side chain, for example, 35, or vary the heterocy-
clic ring system further, as in 36, also reduced the inhibition of
hACC2. As the permeability of 33 was good, the side chain was
lengthened to try and achieve a further increase in potency. This
is exemplified with 37, which maintained potency and had similar
physical and in vitro ADME properties to 33 (Table 3). However, 37
had a different in vivo rat DMPK profile compared to 33 with high-
er CL and therefore higher CLu and also a higher volume of distri-
bution. Further attempts to modify the properties of the side chain
(pKa, size, substitution) led to no further improvement in potency
or ADME properties. Introducing a non-basic side chain such as
the hydroxyethyl derivative 32 also gave no further enhancement
in the inhibition of hACC2.
In order to use protein structure information in our design ef-
forts, we embarked on the crystal structure elucidation of the CT-
domain of hACC2. However, despite extensive efforts, we were
not able to grow protein crystals for hACC2 suitable for structure
determination. As a backup strategy alternative mammalian ana-
logs were tested and we found that bovine ACC2 (bACC2) crystal-
lized well. Subsequently a crystal structure was obtained for the
CT-domain of bACC2 in complex with the benzylamine derivative
28 (Fig. 3) (X-ray coordinates deposited in the Protein Data Bank
with accession number 2X24). The CT-domain of bovine and hu-
man ACC2 have 83% sequence identity and only one amino acid
differs within a 6 Å distance from any atom of 28.25 The X-ray
structure revealed useful information such as protein–ligand
interactions and ligand orientation that can be used to generate
ideas how to influence inhibition of hACC2. The X-ray structure
and, in particular, the orientation of the bound ligand shows large
similarities with previously published structures of yeast and hACC
in complex with 1 as shown in Figure 4.22,23 Compound 28 showed
the same two pivotal anchoring points: 1. Gly461 making a hydro-
gen bonding interaction to the carbonyl oxygen atom of the tert-
butylcarbamate group and 2. Glu529 establishing a hydrogen bond
with the carbonyl oxygen atom of the amide bond linking the
quinoline ring to the cyclohexyl core. Additional hACC2 data on
derivatives within this series suggests the distance between these
two carbonyl oxygen atoms, both functioning as hydrogen bond
acceptors, is important for inhibition of the enzyme (data not
shown). The binding site in the structure is located at the interface
of the two monomers making up the CT-domain (Fig. 5) and, as
hypothesized earlier, the X-ray structure shows the basic benzyl-
amine side chain pointing towards the hydrophilic exterior of the
protein.
Another important observation from the X-ray structure is the
out-of-plane twist of the quinoline system resulting in a 77°
dihedral angle between the quinoline ring and the amide group.
Noteworthy, there is no favorable p–p interaction of the quinoline
system in the binding site contributing to potency. Therefore this
does not explain the loss of potency upon replacement of the quin-
oline ring with pyridine when comparing 17 and 19. However, a
torsional scan analysis provides a rationale for the observed po-
tency drop-off (Fig. 6).26 A high energy barrier was observed for
the amide group being in the plane of the quinoline ring whereas
the out-of-plane orientation is energetically favored. The opposite
applies to the pyridine ring. In this case an energy of ꢀ2.5 kcal/mol
Figure 3. X-ray structure of the CT-domain of bACC2 in complex with the
benzylamine derivative 28 (shown in green), highlighting the hydrogen bond
coordination between the carbonyl oxygens in 28 and the amino acid residues
Glu529 and Gly461.
Figure 4. Overlay of the CT-domain of bACC2 (grey) and hACC2 (orange) with the
corresponding ligands: 28 (green) and 1 (yellow), respectively. Dashed lines denote
the hydrogen bond coordination between the carbonyl oxygens in 1 and the amino
acid residues Glu529 and Gly461 in hACC2.

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3045
Table 4
tert-Butyl carbamate replacements
R
HN
O
N
N
N
Compd
R
hACC2a
1.4
(l
M) hACC1a
(lM) Log D LLE
O
O
NH
O
38
8.0
6.3
1.0
9.9
2.6
2.3
4
3.3
O
O
NH
39
40
41
0.68
0.13
1.7
3.9
2.9
2.2
O
O
O
O
NH
Figure 5. The surface of bACC2 showing the structure of 28 (in green) binding at
the interface of the two chains making up the CT-domain. The red and grey colours
of the surface show the separate monomers with the benzylamine side chain of 28
pointing out of the exterior of the protein (the blue atom is the amine functionality
of 28).
N
H
3.6
a
is required to force the amide group out of the pyridine plane into
the ꢀ77° angle it adopts when bound to the protein (red line),
which, in turn, negatively affects the binding energy of the whole
molecule. Hence the quinoline displays greater potency versus
hACC2.
The tert-butylcarbamate group contributes a large degree to the
overall lipophilicity of the compounds. The potential chemical
instability of a tert-butylcarbamate group at low pH is also well
known as it is commonly used as a protective group for amines.
Hence, alternative carbamate replacements for the tert-butylcarba-
mate group were sought. Chemistry was carried out rapidly as de-
scribed in Scheme 1 to synthesize a broad range of carbamate
derivatives, four of which are shown in Table 4. Replacing the
tert-butylcarbamate with an ethyl carbamate (38) to lower both
molecular weight and lipophilicity gave a 10-fold drop in potency.
The IC50 values are reported as means of at least two separate determinations
with a typical standard deviation of less than 30%. ND means not determined.
Introduction of the tetrahydropyranylcarbamate (39) to reduce
lipophilicity gave a fivefold drop in potency versus hACC2. Overall
LLE was improved but this did not provide the required level of po-
tency to allow further progress of these compounds. Increasing the
size of the tert-butyl group to give the neo-pentyl derivative 40 to
reduce the risk of acid instability showed similar potency versus
hACC2 compared to 33. Finally, reversing the carbamate as in 41
led to a significant reduction of inhibition of hACC2.16 In summary,
all attempts to replace to tert-butyl group, with the exception of
the neo-pentyl lipophilic substituent, led to a drop in potency indi-
cating the need for a lipophilic group in this part of the molecule.
Figure 6. Torsional scan analysis of the amide bond relative to the ring system in 10° increments, starting from the depicted orientation. The red vertical line at 77° denotes
the dihedral angle of the carbonyl-quinoline system in the X-ray crystal structure of 28.

3046
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
Table 5
time curve out to infinity. Instead we used the plasma concentra-
Pharmacokinetic data for 33, 37 and 40 in micea
tion-time curve to support the progression of the compounds into
further studies. Additionally all three compounds showed low inhi-
bition of the hERG encoded potassium channel28 (33 and 37; IC50
Compd
CL (mL/
min/Kg)
Volume of
distribution (L/Kg)
Oral half
life, t1/2 (h)
Oral bioavailability
(%)
>33
panel of cytochrome P450 isoforms (1A2, 2C9, 2C19, 2D6, 3A4;
<50% inhibition at 13 M).
lM and 40; IC50 14 lM) and 33 showed no activity against a
33
37
40
46
33
52
6.7
3.6
9.6
4.5
4.3
2.9
131
135
115
l
a
The ability of compounds to inhibit ACCs in vivo was tested by
examining the lowering of malonyl-CoA in obese Zucker rats. This
animal model was selected as it shares many key metabolic
disorders with the intended target patient group, obese insulin
resistant humans and it has substantially elevated hepatic malo-
nyl-CoA levels compared to lean Zucker rat controls offering a
larger treatment window model for detection of inhibition of
ACCs. There is a sequence homology between hACC2 and rat
ACC2 of 87% for the whole enzyme sequence but to ensure that
these compounds were active in the rat they were tested in an
in vitro rat ACC2 and ACC1 assay.29 Compounds 33, 37 and 1
showed good levels of inhibition of rat ACC2 (IC50 values of
Compounds 33, 37 and 40 dosed at 20, 5 and 33
lmol/Kg (po), respectively, to C57BL/6 mice.
l
mol/Kg (iv) and 20, 10 and
33
This may be rationalized from the X-ray structure, which shows
the tert-butylcarbamate occupying a lipophilic part of the binding
pocket formed by amino acids with lipophilic side chains such as
Val267, Leu268, Ala424 and Val427.
Throughout this work, inhibition of hACC2 has been used as the
primary driver whilst inhibition of hACC1 was monitored. All com-
pounds tested showed greater inhibition of hACC2 than hACC1 and
limited selectivity over hACC1 was observed (Tables 1, 2 and 4). In
general, most compounds showed less than five-fold selectivity be-
tween IC50 values for hACC2 and hACC1. For example, the most po-
tent compound versus hACC2, 27, also had good potency against
hACC1 (Table 2). This is perhaps not surprising when looking at
the amino acid sequence identity between hACC1 and 2. The two
enzymes show 74% identity when comparing the whole enzyme
sequence which is in good agreement with earlier findings.27 In
the CT-domain, where these derivatives bind, there is only one
residue that differs between the two hACC isoforms that lies less
than 6 Å away from the bound ligand 28; hACC2 has a serine res-
idue at position 262 compared with a glycine residue in hACC1.25
The highest selectivity that has been observed over hACC1 was
ꢀ10-fold. Interestingly this selectivity is fairly general for com-
pounds where a piperidine ring is joined directly in the 2-position
of the quinoline ring. Examples include 32, 33 and 37 which all
display modest 10-fold selectivity versus hACC1.
0.19, 0.15 and 0.17
inhibition of rat ACC2 and ACC1 (rat ACC1 IC50 values of 0.46,
0.25 and 0.23 M, respectively). Compounds 33, 37 and 1 were
lM, respectively) with similar potencies for
l
administered as a constant intravenous infusion for 2 h. At the
end of this period, compound infusion was stopped and hepatic
malonyl-CoA levels and compound plasma levels were measured.
Key results for these studies are shown in Figure 7 where hepatic
malonyl-CoA level is plotted as a function of unbound compound
plasma levels normalized to in vitro potency for rat ACC2 inhibi-
tion. As can be seen, untreated obese Zucker rats (obese vehicle)
exhibited
a substantial elevation in levels of malonyl-CoA
(ꢀ10 nmol/g) compared to the metabolically healthy lean Zucker
controls (lean vehicle) (ꢀ1 nmol/g). The three compounds suc-
cessfully reduced hepatic malonyl-CoA levels in an exposure
dependent fashion. All three compounds show ꢀ50% reduction
in hepatic malonyl-CoA levels with unbound plasma concentra-
tions 2–3-fold above the in vitro IC50 for rat ACC2 inhibition. Com-
plete normalization of malonyl-CoA to levels seen in the healthy
animals can be achieved at unbound plasma levels between
approximately 10-fold and 100-fold the in vitro IC50 for rat
ACC2 inhibition as shown with higher concentrations of 1. These
data confirm the efficacy of these substances for ACC inhibition
in vivo. Note that in Figure 7 substance exposure has been nor-
malized to the IC50 for rat ACC2. Since the potencies for inhibition
of rat ACC2 and ACC1 are similar for this chemical series the data
could equally well be normalized to the corresponding value for
rat ACC1. The lack of selectivity in the rat, as well as the relatively
high abundance of the two ACC isoforms in the liver means that
we are not able to assign relative contributions of the individual
isoforms to the observed hepatic malonyl-CoA lowering.
2.3. In vivo evaluation
Compounds 33, 37 and 40 displayed good potency versus
hACC2 and good physical properties, in vitro ADME and in vivo
CL profiles to continue progression. To ensure that the measured
caco-2 permeability would translate to good oral absorption in
mice to support further testing these compounds were progressed
into mouse in vivo DMPK studies and the data is shown in Table 5.
All three compounds show medium CL, high volumes of distribu-
tion, good half life’s after oral dosing and excellent oral bioavail-
ability. The fact that the compounds have a bioavailability of
greater than 100% could be due to a number of reasons, including
enterohepatic recirculation or extrapolation of the concentration-
Figure 7. Relationship between hepatic malonyl-CoA levels and unbound compound plasma levels normalized to in vitro potency for rat ACC2 inhibition for 33, 37 and 1.

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3047
DMSO/DMSO-d6: The solutions are taken from a concentrated
sample dissolved in (CH3)2SO and are diluted with (CD3)2SO. Since
a substantial amount of (CH3)2SO is present in the sample, first a
pre-scan is run and analysed to automatically suppress the (CH3)2SO
(2.54 ppm) and H2O (3.3 ppm) peaks. This means that in this so-
called wet 1D experiment the intensity of peaks that reside in these
areas around 3.3 and 2.54 ppm are reduced.
3. Conclusion
A novel series of inhibitors of hACC1 and 2 is described. Hybri-
dising two hit compounds identified from a high throughput
screen with IC50 values of 2–4 lM versus hACC2 led to a lead series
with improved potency but high lipophilicity contributing to poor
physical properties, low metabolic stability in vitro and high CLu in
vivo. Subsequent optimization led to further improvements in po-
tency, a good reduction in lipophilicity as measured by log D and
hence a good increase in LLE and resulting improvements in phys-
ical properties, stability in HLM and CLu in vivo. Principal to this
was the replacement of the phenyl substituent on the quinoline
ring with substituted piperidines. X-ray crystallography studies re-
vealed these compounds binding in the CT-domain of the homolo-
gous bACC2. The crystal structure of 28 revealed the orientation of
the compounds with the small basic side chain extending to the
hydrophilic exterior of the enzyme and two key hydrogen bonding
interactions that appear crucial in this series of ACC2 inhibitors.
Ultimately this work led to 33 and 37, which exhibit a good bal-
ance of properties including potent inhibition of hACC2 in vitro,
10-fold selectivity over inhibition of hACC1, good physical and
in vitro ADME properties, significantly improved CLu in rat and
good bioavailability in mouse. Both compounds have been shown
to lower levels of hepatic malonyl-CoA in vivo in obese Zucker rats.
With in vivo target engagement now verified, additional work is
required to develop the series further to enhance potency and fur-
ther improve the DMPK profile. Importantly these compounds, as
well as those already reported in the literature, can now be used
to further test the concept of ACC2 and ACC1 inhibition including
the effects of chronic inhibition and the extent of selectivity re-
quired for successful therapeutic intervention.
4.1.1. tert-Butyl {[trans-4-({[(2-phenylquinolin-4-yl)carbonyl]-
amino}methyl)cyclohexyl]methyl}carbamate (17)
2-Phenyl-quinoline-4-carboxylic acid (200 mg, 0.80 mmol) was
dissolved in CH2Cl2 (8 mL) and tert-butyl {[trans-4-(aminomethyl)-
cyclohexyl]methyl}carbamate (4, 204 mg, 0.84 mmol), TBTU
(283 mg, 0.88 mmol) and Et3N (0.45 mL, 3.2 mmol) were added
and the reaction mixture was stirred at rt for 15 h. The reaction
mixture was concentrated in vacuo to leave a residue, which was
dissolved in DMSO and purified by HPLC, using a gradient of
40–100% mobile phase A (100% CH3CN) over 30 min (mobile phase
B = 5% CH3CN + 95% 0.1 M NH4OAc) to give the title compound as a
white solid (247 mg, 65%). 1H NMR (400 MHz, DMSO-d6) d 8.77 (t,
J = 5.6 Hz, 1H), 8.29–8.23 (m, 2H), 8.14–8.04 (m, 3H), 7.81–7.75 (m,
1H), 7.64–7.46 (m, 5H), 3.19 (t, J = 6.1 Hz, 2H), 2.74 (t, J = 6.2 Hz,
2H), 1.80 (d, J = 11.1 Hz, 2H), 1.68 (d, J = 11.5 Hz, 2H), 1.57–1.45
(m, 1H), 1.33 (s, 9H), 1.32–1.22 (m, 1H), 1.00–0.76 (m, 4H); HRMS
(ESI) m/z calcd for C29H36 N3O3 [M+H]+ 474.2757; found 474.2773.
4.1.2. tert-Butyl {[cis-4-({[(2-phenylquinolin-4-yl)carbonyl]-
amino}methyl)cyclohexyl]methyl}-carbamate (18)
Prepared as a solid using essentially the same procedure as
described for 17 starting from tert-butyl {[cis-4-(aminomethyl)-
cyclohexyl]-methyl}carbamate16 in place of 4. 1H NMR (500 MHz,
CDCl3) d 8.15 (d, J = 8.3 Hz, 1H), 8.11 (dd, J = 8.2, 1.3 Hz, 3H), 7.82
(s, 1H), 7.76–7.70 (m, 1H), 7.55–7.45 (m, 4H), 6.33 (t, J = 5.8 Hz,
1H), 4.60 (s, 1H), 3.47 (t, J = 6.7 Hz, 2H), 3.07 (t, J = 6.4 Hz, 2H),
1.85 (s, 1H), 1.71–1.38 (m, 18H); HRMS (ESI) m/z calcd for
4. Experimental section
4.1. Synthetic methods
C
29H36N3O3 [M+H]+ 474.2757; found 474.2764.
All solvents and reagents were obtained from commercially
available sources and used without further purification. Reactions
were carried out under nitrogen atmosphere unless otherwise sta-
ted. Reactions carried out using a microwave reactor were per-
formed using a Biotage Initiator or Personal Chemistry [Biotage]
Emrys Optimizer. Flash chromatography was carried out on pre-
packed silica gel columns supplied by Biotage and using Horizon/
Biotage systems. Analytical HPLC/MS was conducted on a Waters
Zevo QTof or Waters LCT Premiere mass spectrometer using an
Acquity PDA (Waters) UV detector monitoring either at (a) 210 nm
4.1.3. tert-Butyl [(trans-4-{[(2-phenylisonicotinoyl)amino]-
methyl}cyclohexyl)methyl]carbamate (19)
2-Bromo-isonico-tinic acid (0.92 g, 4.5 mmol), TBTU (1.6 g,
5.0 mmol) and DIPEA (1.1 g, 8.3 mmol) were added to a solution
of 4 (1.0 g, 4.1 mmol) in DMF (10 mL) and the reaction mixture
was stirred at rt for 22 h. Water was added to give a precipitate
which was filtered to give the intermediate amide (1.7 g, 96%).
The amide (62 mg, 0.15 mmol) was dissolved in CH3CN (3 mL)
and treated with phenylboronic acid (32 mg, 0.26 mmol), Pd(OAc)2
(14 mg, 0.06 mmol), and a 1 M aqueous solution of NaHCO3 (1 mL).
The reaction mixture was sealed, degassed with nitrogen for
15 min, and heated in a microwave reactor at 150 °C for 10 min.
The solvent was concentrated in vacuo to leave a residue. The
residue was dissolved in CH2Cl2 and washed with a saturated
aqueous solution of NaHCO3 and then dried using a phase separa-
tor. The solvent was concentrated in vacuo to leave a residue which
was purified by flash chromatography, using EtOAc/heptane
(2:1) + 1% Et3N as eluent, to give the title compound as a white
solid (33 mg, 53%). 1H NMR (500 MHz, CDCl3) d 8.80 (d, J = 5.0 Hz,
1H), 8.09–8.03 (m, 3H), 7.54–7.43 (m, 4H), 6.43–6.32 (m, 1H),
4.65–4.57 (m, 1H), 3.36 (t, J = 6.4 Hz, 2H), 2.99 (t, J = 6.2 Hz, 2H),
1.91–1.78 (m, 4H), 1.67–1.55 (m, 2H), 1.45 (s, 9H), 1.10–0.90
(m, 4H); HRMS (ESI) m/z calcd for C25H34N3O3 [M+H]+ 424.2600;
found 424.2610.
with an Acquity BEH C18 column (2.1 Â 100 mm, 1.7
lm, 0.7 mL/
min flow rate), using a gradient of 2% v/v CH3CN in H2O (ammonium
carbonate buffer pH 10) to 98% v/v CH3CN in H2O or (b) 230 nm with
an Acquity HSS C18 column (2.1 Â 100 mm, 1.8
lm, 0.7 mL/min
flow rate), using a gradient of 2% v/v CH3CN in H2O (ammonium for-
mate buffer pH 3) to 98% v/v CH3CN in H2O. Preparative HPLC was
conducted using a Waters Fraction Lynx Purification System using
either (i) Xbridge Prep C18 5
(ii) Kromasil, C8, 10
m, 50.8 Â 300 mm columns, or (iii) Kromasil,
C8, 10
m, 20 Â 50 mm columns. MS-triggered fraction collection
lm OBD 19 Â 150 mm columns, or
l
l
was used. All compounds were determined to be >95% pure using
the analytical methods (a) or (b) above based on the peak area
percentage. 1H NMR spectra were generated on a Varian 300 MHz,
Varian 400 MHz, Varian 500 MHz or Varian 600 MHz instrument
as indicated. Chemical shifts (d) are given in parts per million
(ppm), with the residual solvent signal used as a reference. Coupling
constants (J) are reported as Hertz. NMR abbreviations are used as
follows: s = singlet, d⁄= doublet, t = triplet, q = quartet, m = multi-
plet. DMSO/DMSO-d6 as NMR solvent denotes: wet 1H NMR in
4.1.4. tert-Butyl [(trans-4-{[(2-methyl-6-phenylisonicotinoyl)-
amino]methyl}cyclohexyl)methyl]carbamate (20)
Prepared as a solid using essentially the same procedure as
described for 19 starting from 2-chloro-6-methylisonicotinic acid

3048
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
in place of 2-bromoisonicotinic acid. 1H NMR (500 MHz, CDCl3) d
8.03–7.99 (m, 2H), 7.81 (s, 1H), 7.50–7.45 (m, 2H), 7.45–7.40
(m, 1H), 7.36 (s, 1H), 6.25 (s, 1H), 4.58 (s, 1H), 3.34 (t, J = 6.4 Hz,
2H), 2.98 (t, J = 6.2 Hz, 2H), 2.68 (s, 3H), 1.90–1.76 (m, 4H), 1.44
(s, 9H), 1.37–1.23 (m, 2H), 1.10–0.90 (m, 4H); HRMS (ESI) m/z calcd
for C26H36N3O3 [M+H]+ 438.2757; found 438.2751.
4-acetylpyridine (0.50 g, 4.1 mmol) in EtOH (4 mL) and the reac-
tion mixture was heated at 80 °C for 20 h. The reaction mixture
was diluted with water and purified by HPLC using a gradient of
0–100% mobile phase A (100% CH3CN) over 30 min (mobile phase
B = 5% CH3CN + 95% 0.1 M NH4OAc) to give the intermediate
2-pyridin-4-ylquinoline-4-carboxylic acid (0.20 g, 17%). The inter-
mediate 2-pyridin-4-ylquinoline-4-carboxylic acid was then cou-
pled to 4 in a similar way to that described for 17 to give the
title compound as a solid. 1H NMR (400 MHz, DMSO-d6) d 8.83 (t,
J = 5.6 Hz, 1H), 8.81–8.77 (m, 2H), 8.29–8.25 (m, 2H), 8.22 (s, 1H),
8.21–8.15 (m, 2H), 7.90–7.84 (m, 1H), 7.75–7.68 (m, 1H), 6.79 (t,
J = 5.6 Hz, 1H), 3.24 (t, J = 6.1 Hz, 2H), 2.78 (t, J = 6.2 Hz, 2H), 1.84
(d, J = 11.0 Hz, 2H), 1.73 (d, J = 11.7 Hz, 2H), 1.60–1.49 (m, 1H),
1.37 (s, 9H), 1.34–1.27 (m, 1H), 1.05–0.75 (m, 4H); HRMS (ESI)
m/z calcd for C28H35N4O3 [M+H]+ 475.2709; found 475.2704.
4.1.5. tert-Butyl ((trans-4-((6-phenyl-1H-pyrrolo[2,3-b]-
pyrid-ine-4-carboxamido)methyl)cyclohexyl)methyl-
carbamate (21)
Prepared as a solid using essentially the same procedure as de-
scribed for 17 starting from 6-phenyl-1H-pyrrolo[2,3-b]pyridine-
4-carboxylic acid16 in place of 2-phenyl-quinoline-4-carboxylic
acid. 1H NMR (600 MHz, DMSO/DMSO-d6⁄) d 11.85 (s, 1H), 8.63
(t, J = 5.8 Hz, 1H), 8.16–8.10 (m, 2H), 7.92 (s, 1H), 7.58–7.54 (m,
1H), 7.51–7.46 (m, 2H), 7.42–7.36 (m, 1H), 6.78–6.73 (m, 2H),
3.16 (t, J = 6.3 Hz, 2H), 2.75–2.72 (m, 2H), 1.78 (d, J = 11.1 Hz,
2H), 1.68 (d, J = 12.6 Hz, 2H), 1.56–1.48 (m, 1H), 1.33 (s, 9H),
1.32–1.25 (m, 1H), 0.97–0.73 (m, 4H); HRMS (ESI) m/z calcd for
4.1.9. tert-Butyl ({trans-4-[({[2-(1H-pyrazol-4-yl)quinolin-4-yl]-
carbonyl}amino)methyl]cyclohexylmethyl)carbamate (25)
4-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)-1H-pyrazole
(194 mg, 1.00 mmol) and 22 (290 mg, 0.67 mmol) were dissolved
in dioxane (3 mL) and PEPPSI (11 mg, 0.017 mmol) and a solution
of K2CO3 (176 mg, 1.27 mmol) in water (3 mL) were added. The
reaction mixture was heated in a microwave reactor at 140 °C for
30 min and then a saturated aqueous solution of NaHCO3 was
added. The layers were separated and the aqueous layer was ex-
tracted with EtOAc. The combined organic layers were concen-
trated in vacuo and purified by flash chromatography using a
gradient of 0 to 15% EtOAc in hexane as eluent. Further purification
was achieved by dissolving the title compound in DMSO and puri-
fying by HPLC, using varying gradients of CH3CN and 0.2% NH3 buf-
fer to give the title compound as a solid (20 mg, 9%). 1H NMR
(400 MHz, DMSO-d6) d 8.75 (t, J = 5.6 Hz, 1H), 8.58 (s, 1H), 8.25
(s, 1H), 8.02–7.96 (m, 2H), 7.86 (s, 1H), 7.77–7.71 (m, 1H),
7.57–7.51 (m, 1H), 6.79 (t, J = 5.6 Hz, 1H), 6.25 (t, J = 1.9 Hz, 1H),
3.21 (t, J = 6.2 Hz, 2H), 2.78 (t, J = 6.3 Hz, 2H), 1.83 (d, J = 11.0 Hz,
2H), 1.72 (d, J = 11.0 Hz, 2H), 1.59–1.47 (m, 1H), 1.37 (s, 9H),
1.36–1.26 (m, 1H), 1.02–0.80 (m, 4H); HRMS (ESI) m/z calcd for
C
27H35N4O3 [M+H]+ 463.2709; found 463.2674.
4.1.6. tert-Butyl{[trans-4-({[(2-chloroquinolin-4-yl)carbonyl]-
amino}methyl)cyclohexyl]methyl}-carbamate (22)
N-Methyl morpholine (5.9 mL, 54 mmol) and 4 (6.7 g, 28 mmol)
were added to a solution of 2-chloroquinoline-4-carboxylic acid
(5.6 g, 27 mmol) in a mixture of 2-MeTHF (50 mL) and H2O
(34 mL) at rt. An aqueous solution (9 mL) of HOBt (20% w/w) and
N-methyl morpholine (15% w/w) was added to the stirred solution
followed by the addition of EDC (6.7 g, 35 mmol). The reaction
mixture was stirred vigorously at rt for 4 days. The mixture was fil-
tered and the collected solid was washed with water containing
10% MeOH to leave the title compound as an off-white solid
(6.6 g, 57%): 1H NMR (400 MHz, DMSO-d6) d 8.81 (t, J = 5.7 Hz,
1H), 8.07 (d, J = 8.4 Hz, 1H), 7.99 (d, J = 8.4 Hz, 1H), 7.89–7.81
(m, 1H), 7.73–7.66 (m, 1H), 7.59 (s, 1H), 6.77 (t, J = 5.6 Hz, 1H),
3.21–3.12 (m, 2H), 2.79–2.72 (m, 2H), 1.83–1.74 (m, 2H),
1.72–1.64 (m, 2H), 1.56–1.43 (m, 1H), 1.35 (s, 9H), 1.33–1.24 (m,
1H), 0.90–0.75 (m, 4H); HRMS (ESI) m/z calcd for C23H31ClN3O3
[M+H]+ 432.2054; found 432.2072.
C
26H34N5O3 [M+H]+ 464.2661; found 464.2671.
4.1.10. tert-Butyl {[trans-4-({[(2-piperazin-1-ylquinolin-4-yl)-
carbonyl]amino}methyl)cyclohexyl]methyl}carbamate (26)
Piperazine (239 mg, 2.78 mmol) was added to a solution of 22
(100 mg, 0.23 mmol) in pyridine (2 mL) and the reaction mixture
was heated to 130 °C for 30 min in a microwave reactor. The reac-
tion mixture was concentrated in vacuo to leave a residue which
was purified by flash chromatography, using a gradient of 5–70%
MeOH (2% Et3N) in CH2Cl2 as eluent, and then HPLC, using a gradi-
ent of 30–100% CH3CN and 0.2% NH3 buffer as eluent to give the ti-
tle compound as a solid (43 mg, 39%). 1H NMR (300 MHz, CDCl3) d
8.61–8.56 (m, 1H), 7.80–7.77 (m, 1H), 7.57–7.48 (m, 3H), 7.23–7.18
(m, 1H), 6.79–6.74 (m, 1H), 3.63–3.58 (m, 4H), 3.16–3.10 (m, 2H),
2.81–2.73 (m, 6H), 1.81–1.75 (m, 2H), 1.72–1.66 (m, 2H), 1.52–1.44
(m, 1H), 1.36 (s, 9H), 1.32–1.26 (m, 1H), 0.96–0.80 (m, 4H); HRMS
(ESI) m/z calcd for C27H40N5O3 [M+H]+ 482.3131; found 482.3124.
4.1.7. tert-Butyl ({trans-4-[({[2-(3,5-dimethylisoxazol-4-yl)-
quinolin-4-yl]carbonyl}amino)-methyl]cyclohexyl}methyl)-
carbamate (23)
2-Chloroquinoline-4-carboxylic acid (0.15 g, 0.72 mmol), 3,5-
dimethylisoxazole-4-boronic acid (0.12 g, 0.87 mmol) and
Pd(PPh3)4 (42 mg, 0.04 mmol) were added to a mixture of dioxane
(2 mL) and a 1 M aq solution of K2CO3 (2 mL). The reaction mixture
was degassed, sealed, and heated in a microwave reactor at 140 °C
for 15 min. The reaction mixture was concentrated in vacuo to
leave a residue which was purified by HPLC using a gradient of
30–100% mobile phase A (100% CH3CN) over 30 min (mobile phase
B = 5% CH3CN + 95% 0.1 M NH4OAc) to give the intermediate car-
boxylic acid (148 mg, 75%). m/z 269.6 (M+H)+. The intermediate
carboxylic acid was then coupled to 4 using essentially the same
method as for 17 to give the title compound as a solid. 1H NMR
(400 MHz, DMSO-d6) d 8.76 (t, J = 5.5 Hz, 1H), 8.12–8.04 (m, 2H),
7.83 (t, J = 7.3 Hz, 1H), 7.71–7.60 (m, 2H), 6.79 (s, 1H), 3.22
(t, J = 6.0 Hz, 2H), 2.77 (t, J = 6.1 Hz, 2H), 2.69 (s, 3H), 2.50 (s, 3H),
1.81 (d, J = 10.9 Hz, 2H), 1.72 (d, J = 11.6 Hz, 2H), 1.60–1.47
(m, 1H), 1.37 (s, 9H), 1.35–1.26 (m, 1H), 1.05–0.79 (m, 4H); HRMS
(ESI) m/z calcd for C28H37N4O4 [M+H]+ 493.2815; found 493.2821.
4.1.11. tert-Butyl [(trans-4-{[({2-[3-(aminomethyl)phenyl]-
quinolin-4-yl}carbonyl)amino]methyl}cyclohexyl)methyl]-
carbamate (27)
Prepared using essentially the same procedure as described for
23 starting from 22 (200 mg, 0.49 mmol) and 3-(amino-methyl)-
phenyl]boronic acid hydrochloride to give the title compound as
a solid (109 mg, 0.58 mmol). 1H NMR (400 MHz, MeOH-d4) d
8.21–8.11 (m, 3H), 8.08 (dt, J = 7.4, 1.6 Hz, 1H), 8.02 (s, 1H),
7.84–7.77 (m, 1H), 7.67–7.60 (m, 1H), 7.57–7.48 (m, 2H), 3.96 (s,
2H), 3.35 (d, J = 6.8 Hz, 2H), 2.90 (d, J = 6.7 Hz, 2H), 1.94 (d,
J = 12.4 Hz, 2H), 1.83 (d, J = 12.2 Hz, 2H), 1.71–1.58 (m, 1H), 1.43
4.1.8. tert-Butyl {[trans-4-({[(2-pyridin-4-ylquinolin-4-yl)car-
bonyl]amino}methyl)-cyclohexyl]methyl}-carbamate (24)
Potassium hydroxide (0.57 g, 10 mmol) was added to a mixture
of 2,3-dihydro-indole-2,3-dione (isatin) (0.50 g, 3.4 mmol) and

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3049
(s, 10H), 1.15–0.91 (m, 4H); HRMS (ESI) m/z calcd for C30H39N4O3
(t, J = 5.6 Hz, 1H), 8.55 (s, 1H), 8.20 (s, 1H), 8.03–7.95 (m, J = 11.4,
7.8 Hz, 2H), 7.80 (s, 1H), 7.77–7.71 (m, 1H), 7.58–7.51 (m, 1H),
6.79 (t, J = 5.6 Hz, 1H), 4.27 (t, J = 6.3 Hz, 2H), 3.21 (t, J = 6.2 Hz,
2H), 2.78 (t, J = 6.3 Hz, 2H), 2.71 (t, J = 6.4 Hz, 2H), 2.19 (s, 6H),
1.83 (d, J = 10.8 Hz, 2H), 1.72 (d, J = 10.6 Hz, 2H), 1.58–1.47 (m,
1H), 1.37 (s, 9H), 1.35–1.26 (m, 1H), 1.03–0.80 (m, 4H); HRMS
(ESI) m/z calcd for C30H43N6O3 [M+H]+ 535.3397; found 535.3371.
[M+H]+ 503.3022; found 503.2985.
4.1.12. tert-Butyl [(trans-4-{[({2-[4-(aminomethyl)phenyl]quin-
olin-4-yl}carbonyl)amino]methyl}cyclo-hexyl)methyl]carba-
mate acetate (28)
2-Chloroquinoline-4-carboxylic acid (0.15 g, 0.72 mmol) and
4-(aminomethylphenyl)boronic acid (0.16 g, 0.87 mmol) were
reacted under similar conditions to that described for 23. The
reaction mixture was then cooled to 0 °C and a solution of (9-flu-
orenylmethyl)chloroformate (0.46 g, 1.8 mmol) in dioxane (1 mL)
was added dropwise. The reaction mixture was warmed to rt and
stirred for 36 h. The reaction mixture was filtered and the solid
was washed with DMSO, water and MeOH to give the carboxylic
acid (0.22 g, 60% over two steps) as a crude solid, which was used
directly in the next step without further purification. The crude
acid was reacted with 4 using essentially the same procedure as
described for 17 to give the crude amide, which was dissolved in
CH2Cl2 (4 mL) and treated with piperidine (1 mL). The reaction
mixture was stirred at rt for 5 min and then concentrated in vacuo
to leave a residue, which was dissolved in DMSO and purified by
HPLC, using a gradient of 0–100% mobile phase A (100% CH3CN)
over 30 min (mobile phase B = 5% CH3CN + 95% 0.1 M NH4OAc) to
4.1.15. tert-Butyl [(trans-4-{[({2-[4-(2-hydroxyethyl)piperazin-
1-yl]quinolin-4-yl}carbonyl)amino]-methyl}cyclohexyl)meth-
yl]-carbamate (31)
Prepared using essentially the same procedure as 26 using
1-(2-dimethylaminoethyl)-piperazine (0.52 g, 3.3 mmol) in place
of piperazine to give the title compound as a solid. 1H NMR
(600 MHz, DMSO/DMSO-d6)⁄ d 8.58 (t, J = 5.7 Hz, 1H), 7.77 (d,
J = 7.4 Hz, 1H), 7.55 (d, J = 7.8 Hz, 1H), 7.53–7.49 (m, 1H),
7.24–7.18 (m, 1H), 7.14 (s, 1H), 6.76 (t, J = 5.9 Hz, 1H), 3.67–3.64
(m, 4H), 3.12 (t, J = 6.2 Hz, 2H), 2.74 (t, J = 6.4 Hz, 2H), 2.43–2.38
(m, 4H), 2.14 (s, 6H), 1.77 (d, J = 12.7 Hz, 2H), 1.68 (d, J = 12.2 Hz,
2H), 1.51–1.43 (m, 1H), 1.34 (s, 9H), 1.30–1.24 (m, 1H), 0.95–0.78
(m, 4H); HRMS (ESI) m/z calcd for C31H49N6O3 [M+H]+ 553.3866;
found 553.3854.
give the title compound as
a
solid (35 mg, 32%). 1H NMR
4.1.16. tert-Butyl [(trans-4-{[({2-[4-(2-hydroxyethyl)piperidin-
1-yl]quinolin-4-yl}carbonyl)amino]-methyl}cyclohexyl)-meth-
yl]carbamate (32)
(400 MHz, DMSO-d6) d 8.77 (t, J = 5.7 Hz, 1H), 8.21 (d, J = 8.2 Hz,
2H), 8.07 (dd, J = 8.5, 2.7 Hz, 1H), 8.03 (s, 1H), 7.77 (t, J = 7.7 Hz,
1H), 7.61–7.46 (m, 4H), 6.74 (t, J = 5.4 Hz, 1H), 3.77 (s, 2H), 3.19
(t, J = 6.1 Hz, 2H), 2.74 (t, J = 6.2 Hz, 2H), 1.83 (s, 2H), 1.79
(d, J = 12.6 Hz, 2H), 1.68 (d, J = 12.4 Hz, 2H), 1.55–1.45 (m, 1H),
1.33 (s, 9H), 1.30–1.23 (m, 1H), 1.00–0.76 (m, 4H); HRMS (ESI)
m/z calcd for C30H39N4O3 [M+H]+ 503.3022; found 503.3018.
Prepared using essentially the same procedure as described for
26 using 4-piperidineethanol in place of piperazine to give the title
compound as a solid (50 mg, 51%). 1H NMR (600 MHz, DMSO/
DMSO-d6)⁄ d 8.57 (t, J = 5.7 Hz, 1H), 7.74 (d, J = 8.2 Hz, 1H),
7.54–7.51 (m, 1H), 7.50–7.46 (m, 1H), 7.19–7.15 (m, 1H), 7.11 (s,
1H), 6.77–6.73 (m, 1H), 4.52–4.46 (m, 2H), 4.43–4.36 (m, 1H),
3.12 (t, J = 6.2 Hz, 2H), 2.90–2.83 (m, 2H), 2.74 (t, J = 6.3 Hz, 2H),
1.79–1.71 (m, 4H), 1.69–1.62 (m, 3H), 1.50–1.42 (m, 1H),
1.38–1.33 (m, 2H), 1.33 (s, 9H), 1.30–1.24 (m, 1H), 1.14–1.05
(m, 2H), 0.94–0.76 (m, 4H); HRMS (ESI) m/z calcd for C30H45N4O4
[M+H]+ 525.3441; found 525.3452.
4.1.13. 4-(4-{[(trans-4-{[(tert-Butoxycarbonyl)amino]methyl}-
cyclohexyl)methyl]carbamoyl}quinolin-2-yl)benzoic acid (29)
Prepared using essentially the same procedure as described for
28 using (4-methoxycarbonylphenyl)boronic acid in place of
4-(aminomethylphenyl)boronic acid and then similar acid cou-
pling conditions as described for 17 to give the intermediate
methyl ester. The methyl ester (13 mg, 0.024 mmol) was dissolved
in THF (1 mL) and treated with a 1 M aqueous solution of lithium
hydroxide (0.5 mL). The reaction mixture was stirred at rt for 3 h
and then concentrated in vacuo to leave a residue. The residue
was dissolved in DMSO and purified by HPLC using a gradient of
30–100% mobile phase A (100% CH3CN) over 30 min (mobile phase
B = 5% CH3CN + 95% 0.1 M NH4OAc) to give the title compound as a
colourless oil (4.7 mg, 39%). 1H NMR (400 MHz, DMSO-d6) d 13.05
(s, 1H), 8.87–8.79 (m, 1H), 8.49–8.40 (m, 2H), 8.21–8.09 (m, 5H),
7.85 (t, J = 8.3 Hz, 1H), 7.72–7.65 (m, 1H), 6.80 (t, J = 6.0 Hz, 1H),
3.24 (t, J = 6.2 Hz, 2H), 2.78 (t, J = 6.2 Hz, 2H), 1.88–1.80 (m, 2H),
1.78–1.68 (m, 2H), 1.60–1.49 (m, 1H), 1.37 (s, 9H), 1.36–1.27
(m, 1H), 1.07–0.80 (m, 4H); HRMS (ESI) m/z calcd for C30H36N3O5
[M+H]+ 518.2655; found 518.2651.
4.1.17. tert-Butyl {[trans-4-({[(2-{4-[2-(dimethylamino)ethyl]-
piperidin-1-yl}quinolin-4-yl)carbonyl]-amino}methyl)cyclo-
hexyl]-methyl}carbamate (33)
DIPEA (0.3 mL, 1.73 mmol) and methane sulfonyl chloride
(158 mg, 1.40 mmol) were added to a solution of 32 (360 mg,
0.69 mmol) in CH2Cl2 (10 mL) at 0 °C and the reaction mixture
was stirred for 1 h at rt. The reaction mixture was concentrated
in vacuo to leave the crude mesylate which was used in the next
step with no further purification. Dimethylamine (2 M in MeOH,
1.66 mL, 3.3 mmol) was added to the mesylate (600 mg, 0.17
mmol) in DMF (1 mL) and the reaction mixture was stirred for
20 h at rt. The reaction mixture was then purified by flash chroma-
tography, using a gradient of 5–70% MeOH (2% Et3N) in CH2Cl2 as
eluent, to give the title compound as a white solid (30 mg, 33%).
1H NMR (500 MHz, CDCl3) d 7.83–7.80 (m, 1H), 7.68–7.64 (m,
1H), 7.54–7.49 (m, 1H), 7.24–7.14 (m, 2H), 6.96 (s, 1H),
6.28–6.24 (m, 1H), 4.65–4.60 (m, 1H), 4.49–4.44 (m, 2H), 3.35 (t,
J = 6.4 Hz, 2H), 2.99–2.87 (m, 5H), 2.43–2.39 (m, 2H), 2.30 (s, 6H),
1.89–1.76 (m, 6H), 1.64–1.54 (m, 1H), 1.49–1.40 (m, 2H), 1.43 (s,
9H), 1.31–1.19 (m, 2H), 1.09–0.89 (m, 4H); HRMS (ESI) m/z calcd
for C32H50N5O3 [M+H]+ 552.3914; found 552.3864.
4.1.14. tert-Butyl {[trans-4-({[(2-{1-[2-(dimethylamino)ethyl]-
1H-pyrazol-4-yl}quinolin-4-yl)carbonyl]-amino}methyl)cyclo-
hexyl]methyl}carbamate (30)
Cesium carbonate (90 mg, 0.28 mmol) and 2-dimethylamino-
ethylchloride hydrochloride (24 mg, 0.17 mmol) were added to a
solution of 25 (64 mg, 0.14 mmol) in EtOH (1 mL) and the reaction
mixture was heated in a microwave reactor at 140 °C for 20 min.
The reaction mixture was concentrated in vacuo to leave a residue
which was purified by HPLC using a gradient of 20–100% mobile
4.1.18. tert-Butyl [(trans-4-{[({2-[4-(2-aminoethyl)piperidin-1-
yl]quinolin-4-yl}carbonyl)amino]methyl}cyclohexyl)methyl]-
carbamate (34)
phase
A (100% CH3CN) over 25 min (mobile phase B = 5%
CH3CN + 95% 0.1 M HCO2H) as eluent to give the title compound
Sodium azide (97 mg, 1.5 mmol) was added to a solution of the
intermediate mesylate used on route to 33 (300 mg, 0.5 mmol) in
as a solid (37 mg, 50%). 1H NMR (400 MHz, DMSO-d6) d 8.76

3050
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
DMF (3 ml) and the reaction mixture was stirred for 20 h at rt. The
reaction mixture was purified by flash chromatography, using a
gradient of 50–100% EtOAc in heptane as eluent, to give the azide
(197 mg, 72%). Tin(II) chloride (118 mg, 0.62 mmol), Et3N (0.26 mL,
1.86 mmol) and thiophenol (0.26 mL, 2.5 mmol) were added to a
solution of the azide (93 mg, 0.17 mmol) in THF (4 mL) and the
reaction mixture was stirred for 10 min at rt. The reaction mixture
was concentrated in vacuo to leave a residue, which was purified
by flash chromatography, using a gradient of 0–10% MeOH/CH2Cl2
and then a gradient of 10–75% MeOH (2% Et3N) in CH2Cl2 as eluent,
to give the title compound as a solid (17 mg, 19%). 1H NMR
4.1.21. tert-Butyl ({trans-4-[({[2-(4-{2-[(2-methoxyethyl)-
amino]ethyl}piperidin-1-yl)quinolin-4-yl]carbonyl}amino)-
methyl]cyclohexyl}methyl)carbamate (37)
Oxalyl chloride (0.26 mL, 3.1 mmol) was added dropwise to a
solution of DMSO (0.44 mL, 6.2 mmol) in CH2Cl2 (15 mL) at
À78 °C. A solution of 32 (0.54 g, 1.0 mmol) in CH2Cl2 (10 mL)
was added in portions at À78 °C. The mixture was stirred for
30 min and then Et3N (2 mL, 14.5 mmol) was added. The reaction
mixture was warmed to rt and stirred for 30 min and then diluted
with CH2Cl2 (100 mL) and water was added. The layers were sep-
arated and the aq layer was extracted with CH2Cl2 (50 mL). The
combined organic layers were dried (Na2SO4), filtered and con-
centrated in vacuo to leave a residue, which was purified by HPLC
using a gradient of 10–95% mobile phase A (100% CH3CN) over
25 min (mobile phase B = 5% CH3CN + 95% 0.1 M HCO2H) as elu-
ent. The required fractions were concentrated in vacuo to leave
the water phase, which was diluted further with water and pH
of the solution was raised to 11 by adding saturated aq sodium
carbonate solution and extracted with EtOAc. The combined or-
ganic layers were dried (Na2SO4), filtered and concentrated in va-
cuo to leave the intermediate aldehyde (0.52 g, 96%). A mixture of
the aldehyde (100 mg, 0.19 mmol), 3-hydroxy-azetidine hydro-
chloride (31 mg, 0.29 mmol), and Et3N (0.053 mL, 0.38 mmol) in
EtOH (3 mL) was stirred at rt for 5 min. Sodium triacetoxyborohy-
dride (73 mg, 0.34 mmol) was then added and the reaction mix-
ture was heated in a microwave reactor at 120 °C for 10 min.
The reaction mixture was concentrated in vacuo to leave a resi-
due, which was purified by HPLC, using a gradient of 30–100%
CH3CN and 0.2% NH3 buffer as eluent to give the title compound
(300 MHz, CDCl3)
d 8.27–8.24 (m, 1H), 8.10–8.07 (m, 1H),
7.97–7.93 (m, 1H), 7.83 (s, 1H), 7.67–7.63 (m, 2H), 7.59–7.56 (m,
1H), 4.94–4.88 (m, 2H), 3.78–3.71 (m, 3H), 3.61–3.57 (m, 1H),
3.41–3.34 (m, 4H), 3.23–3.19 (m, 2H), 2.76 (s, 1H), 2.32–2.20 (m,
6H), 2.11–1.95 (m, 1H), 1.94–1.79 (m, 2H), 1.85 (s, 9H),
1.74–1.66 (m, 3H), 1.51–1.33 (m, 4H); HRMS (ESI) m/z calcd for
C
30H46N5O3 [M+H]+ 524.3600; found 524.3592.
4.1.19. tert-Butyl [(trans-4-{[({2-[4-(aminomethyl)piperidin-1-
yl]quinolin-4-yl}carbonyl)amino]methyl}-cyclohexyl)methyl]-
carbamate (35)
Prepared using essentially the same procedure as described for
26 using benzyl (piperidin-4-ylmethyl)carbamate hydrochloride
together with sodium hydrogen carbonate in place of piperazine
to give the intermediate piperidine. A mixture of the intermediate
piperidine (0.087 g, 0.14 mmol) and 5% palladium on carbon
(0.080 g, 0.038 mmol) in MeOH (10 mL) was stirred under a hydro-
gen atmosphere at rt for 1 h. The reaction mixture was filtered and
the filtrate was concentrated in vacuo to leave a residue, which
was purified by HPLC using a gradient of 10–90% mobile phase A
(100% CH3CN) over 20 min (mobile phase B = water/CH3CN/HCO2H
95:5:0.2) as eluent to give the title compound as a solid (34 mg,
49%). 1H NMR (600 MHz, DMSO/DMSO-d6⁄) d 8.58 (t, J = 5.8 Hz,
1H), 8.32 (s, 2H), 7.76 (d, J = 8.3 Hz, 1H), 7.54 (d, J = 8.2 Hz, 1H),
7.49 (t, J = 7.6 Hz, 1H), 7.19 (t, J = 7.5 Hz, 1H), 7.15 (s, 1H), 6.76
(t, J = 5.7 Hz, 1H), 4.54 (d, J = 13.5 Hz, 2H), 3.31 (s, 5H), 3.13
(t, J = 6.1 Hz, 2H), 2.89 (t, J = 11.8 Hz, 2H), 2.74 (t, J = 6.3 Hz, 2H),
2.62–2.59 (m, 2H), 2.05–2.04 (m, J = 0.6 Hz, 2H), 1.82–1.65
(m, 6H), 1.51–1.43 (m, 1H), 1.34 (s, 9H), 1.31–1.24 (m, 1H),
1.18–1.09 (m, 2H), 0.96–0.78 (m, 4H); HRMS (ESI) m/z calcd for
as
a
white solid (71 mg, 64%). 1H NMR (600 MHz, DMSO/
⁄
DMSO-d6
)
d 8.57 (t, J = 5.8 Hz, 1H), 7.76–7.74 (m, 1H),
7.54–7.51 (m, 1H), 7.50–7.46 (m, 1H), 7.19–7.16 (m, 1H), 7.11
(s, 1H), 6.76 (t, J = 5.8 Hz, 1H), 4.51–4.47 (m, 2H), 4.13–4.08
(m, 1H), 3.47–3.44 (m, 2H), 3.12 (t, J = 6.2 Hz, 2H), 2.89–2.83
(m, 2H), 2.74 (t, J = 6.3 Hz, 2H), 2.61–2.58 (m, 2H), 2.39–2.35
(m, 2H), 1.79–1.65 (m, 6H), 1.60–1.52 (m, 1H), 1.50–1.43
(m, 1H), 1.34 (s, 9H), 1.30–1.24 (m, 1H), 1.20–1.16 (m, 2H),
1.11–1.03 (m, 2H), 0.94–0.78 (m, 4H); HRMS (ESI) m/z calcd for
C
33H50N5O4 [M+H]+ 580.3863; found 580.3870.
4.1.22. Ethyl {[trans-4-({[(2-{4-[2-(dimethylamino)ethyl]-piper-
idin-1-yl}quinolin-4-yl)carbonyl]-amino}methyl)cyclo-hexyl]-
methyl}carbamate (38)
C
29H44N5O3 [M+H]+ 510.3444; found 510.3450.
4.1.20. tert-Butyl {[trans-4-({[(2-{3-[2-(dimethylamino)ethoxy]-
azetidin-1-yl}-quinolin-4-yl)carbonyl]-amino}methyl)cyclo-
hexyl]methyl}carbamate (36)
TFA (3 mL) was added to 33 (106 mg, 0.19 mmol) in CH2Cl2 and
the reaction mixture was stirred at rt for 3 h. Toluene was added
and the reaction mixture was concentrated in vacuo to leave the
crude amine, which was used directly in the next step with no fur-
ther purification. Ethyl chloroformate (25 mg, 0.23 mmol) was
added to a solution of the crude amine and Et3N (175 mg,
1.73 mmol) in THF at 0 °C and the reaction mixture was stirred
at rt for 16 h. MeOH was added and the reaction mixture was con-
centrated in vacuo to leave a residue, which was purified by HPLC
using a gradient of 30–100% mobile phase A (100% CH3CN) over
30 min (mobile phase B = 5% CH3CN + 95% 0.1 M NH4OAc) to give
the title compound as a solid (65 mg, 65% yield over 2 steps). 1H
NMR (600 MHz, DMSO/DMSO-d6⁄) d 8.57 (t, J = 5.7 Hz, 1H), 7.75
(d, J = 8.1 Hz, 1H), 7.53 (d, J = 8.2 Hz, 1H), 7.51–7.46 (m, 1H), 7.17
(t, J = 7.5 Hz, 1H), 7.12 (s, 1H), 7.03 (t, J = 5.7 Hz, 1H), 4.50 (d,
J = 13.1 Hz, 2H), 3.93 (dd, J = 14.1, 7.1 Hz, 2H), 3.12 (t, J = 6.2 Hz,
2H), 2.87 (t, J = 11.8 Hz, 2H), 2.79 (t, J = 6.3 Hz, 2H), 2.29–2.23 (m,
2H), 2.12 (s, 6H), 1.80–1.65 (m, 6H), 1.62–1.54 (m, 1H), 1.51–
1.42 (m, 1H), 1.36–1.26 (m, 3H), 1.15–1.06 (m, 5H), 0.95–0.78
(m, 4H); HRMS (ESI) m/z calcd for C30H46N5O3 [M+H]+ 524.3600;
found 524.3600.
The first step to the intermediate 3-hydroxyazetidine was
prepared by the method described in 26 using azetidin-3-ol in
place of piperazine. Sodium hydride (60% in mineral oil, 21 mg,
0.53 mmol) was then added to the intermediate 3-hydroxyazeti-
dine (0.10 g, 0.21 mmol) in DMF (2 mL) at 0 °C. After 30 min
2-chloro-N,N-dimethylethanamine hydrochloride (34 mg, 0.23
mmol) was added and the reaction mixture was stirred at 50 °C
for 16 h. The reaction mixture was filtered and purified by HPLC,
using a gradient of 30–100% CH3CN and 0.2% NH3 buffer as eluent
to give the title compound as a solid (10 mg, 9%). 1H NMR
(600 MHz, CDCl3) d 7.87 (d, J = 8.6 Hz, 1H), 7.72 (d, J = 8.4 Hz, 1H),
7.58–7.53 (m, 1H), 7.26–7.23 (m, 1H), 6.61 (s, 1H), 5.97
(t, J = 5.8 Hz, 1H), 4.60–4.54 (m, 1H), 4.51–4.46 (m, 1H),
4.38–4.33 (m, 2H), 4.09 (dd, J = 9.0, 4.3 Hz, 2H), 3.54 (t, J = 5.6 Hz,
2H), 3.38 (t, J = 6.4 Hz, 2H), 2.98 (t, J = 6.0 Hz, 2H), 2.53 (t, J =
5.6 Hz, 2H), 2.27 (s, 6H), 1.89–1.79 (m, 4H), 1.61–1.52 (m, 1H),
1.45–1.38 (m, 1H), 1.43 (s, 9H), 1.09–0.93 (m, 4H); HRMS (ESI)
m/z calcd for C30H46N5O4 [M+H]+ 540.3550; found 540.3500.

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3051
4.1.23. Tetrahydro-2H-pyran-4-yl {[trans-4-({[(2-{4-[2-(di-met-
hylamino)ethyl]piperidin-1-yl}quinolin-4-yl)carbonyl]-amino}-
methyl)cyclohexyl]methyl}carbamate (39)
vacuo to leave a residue which was purified by preparative HPLC
(20?95% CH3CN in 0.2% aqueous NH3 as eluent) to give 2-{4-[2-
(dimethylamino)ethyl]piperidin-1-yl}-N-{[trans-4-(hydroxyl-met-
hyl)cyclohexyl]methyl}-quinoline-4-carboxamide (0.017 g, 14%).
2-Isocyanato-2-methylpropane (0.017 mL, 0.15 mmol) was added
to the carboxamide (0.017 g, 0.04 mmol) in pyridine (0.5 mL) and
the reaction mixture was stirred at 70 °C for 16 h. Additional 2-iso-
cyanato-2-methylpropane (0.1 mL, 0.9 mmol) was added and the
reaction mixture was stirred at 70 °C for 32 h and then concen-
trated in vacuo and the residue was purified by reversed phase
HPLC, using a gradient of 30?100% CH3CN and 0.2% NH3 buffer
as eluent to give the title compound as a solid (6.5 mg, 31%).
1H NMR (600 MHz, DMSO/DMSO-d6⁄) d 8.64 (t, J = 5.8 Hz, 1H),
7.82–7.76 (m, 1H), 7.57 (d, J = 7.6 Hz, 1H), 7.55–7.51 (m, 1H),
7.24–7.20 (m, 1H), 7.17 (s, 1H), 4.55 (d, J = 13.1 Hz, 2H),
3.77–3.67 (m, 2H), 3.17 (t, J = 6.3 Hz, 2H), 2.96–2.87 (m, 2H), 2.55
(s, 1H), 2.38–2.31 (m, 2H), 2.27–2.14 (m, 6H), 1.90–1.72 (m, 6H),
1.68–1.58 (m, 1H), 1.57–1.47 (m, 2H), 1.43–1.35 (m, 2H), 1.21
(s, 9H), 1.19–1.09 (m, 2H), 1.03–0.92 (m, 4H); HRMS (ESI) m/z calcd
for C32H50N5O3 [M+H]+ 552.3914; found 552.3942.
HCl (4 M in 1,4-dioxane, 6 mL, 24 mmol) was added to a solution
of 33 (0.2 g, 0.36 mmol) in a mixture of 1,4-dioxane and water 1:1
(6 mL) and the reaction mixture was stirred at rt for 15 min. The
reaction mixture was concentrated in vacuo to leave the crude
amine (0.36 g, 98%) that was used with no further purification. A
solution of 4-nitrophenyl tetrahydro-2H-pyran-4-yl carbonate (42,
75 mg, 0.28 mmol) in THF (3 mL) and Et3N (0.27 mL, 1.94 mmol)
was added to a solution of the amine (99 mg, 0.22 mmol) in THF
(2 mL) and the reaction mixture was stirred at 40 °C for 3 h and then
at rt for 16 h. The reaction mixture was concentrated in vacuo to
leave a residue which was purified by flash chromatography, using
a mixture of 2–25% MeOH in CH2Cl2 as eluent, to give the title com-
pound as a solid (16 mg, 13%). 1H NMR (600 MHz, DMSO/DMSO-d6
)
⁄
d 8.57 (t, J = 5.8 Hz, 1H), 7.75 (d, J = 8.2 Hz, 1H), 7.54–7.51 (m, 1H),
7.50–7.46 (m, 1H), 7.19–7.15 (m, 1H), 7.12 (s, 1H), 7.08 (t,
J = 5.8 Hz, 1H), 4.66–4.59 (m, 1H), 4.53–4.46 (m, 2H), 3.80–3.72
(m, 2H), 3.40–3.33 (m, 2H), 3.12 (t, J = 6.2 Hz, 2H), 2.90–2.83 (m,
2H), 2.80 (t, J = 6.3 Hz, 2H), 2.26–2.19 (m, 2H), 2.09 (s, 6H),
1.82–1.64 (m, 8H), 1.62–1.53 (m, 1H), 1.52–1.39 (m, 3H),
1.37–1.26 (m, 3H), 1.16–1.05 (m, 2H), 0.95–0.78 (m, 4H); HRMS
(ESI) m/z calcd for C33H50N5O4 [M+H]+ 580.3863; found 580.3875.
4.2. hACC2 and hACC1 in vitro assay description
Human recombinant ACC1 and ACC2 were assayed by
determining the amount of inorganic phosphate (Pi) generated.
4.1.24. 2,2-Dimethylpropyl {[trans-4-({[(2-{4-[2-(dimethyl-
amino)ethyl]piperidin-1-yl}quinolin-4-yl)-carbonyl]amino}-
methyl)cyclohexyl]methyl}carbamate (40)
Pre-incubation of 15
compound for 5 min was done before the reaction was initiated
by the addition of 10 L substrate. The reaction was carried out
lL enzyme together with 0.75 lL DMSO
l
2,2-Dimethylpropyl chlorocarbonate (32
l
L, 0.21 mmol) was
in 384-well plates at 37 °C containing the following constituents
(n = 2); 1 mM ATP, 0.4 mM acetyl-CoA, 20 mM citrate, 50 mM
Hepes (pH 7.5), 12.5 mM NaHCO3, 20 mM MgAc2 Â 4H2O, 1.5 mM
MgSO4 Â 7H2O, 1 mM TCEP, 0.025% BSA, compound at different
concentrations, ACC1 and ACC2 diluted to generate signals result-
ing in assay Z’ >0.5. After 1.5 h the reaction was stopped by addi-
added to a solution of the intermediate amine, from on route to
39 (100 mg, 0.18 mmol), and DIPEA (0.31 mL, 1.8 mmol) in CH2Cl2
(5 mL) and the reaction mixture was stirred at rt for 10 min. The
reaction mixture was concentrated in vacuo to leave a residue,
which was purified by HPLC using a gradient of 10–90% mobile
phase A (100% CH3CN) over 20 min (mobile phase B = water/
CH3CN/HCO2H 95:5:0.2) as eluent to give the title compound as a
white solid (75 mg, 74%). 1H NMR (500 MHz, CDCl3) d 7.88–7.80
(m, 1H), 7.71–7.64 (m, 1H), 7.57–7.49 (m, 1H), 7.28–7.19 (m,
1H), 7.01 (s, 1H), 6.20–6.12 (m, 1H), 4.80–4.70 (m, 1H), 4.55–4.46
(m, 2H), 3.79–3.70 (m, 2H), 3.42–3.33 (m, 2H), 3.08–2.99
(m, 2H), 2.98–2.86 (m, 2H), 2.72–2.62 (m, 2H), 2.48 (s, 6H),
1.93–1.75 (m, 6H), 1.68–1.53 (m, 4H), 1.50–1.39 (m, 1H),
1.34–1.21 (m, 2H), 1.11–0.88 (m, 4H), 0.92 (s, 9H); HRMS (ESI)
m/z calcd for C33H52N5O3 [M+H]+ 566.4070; found 566.4029.
tion of 74
transfer of 2 Â 40
well (n = 2?n = 4). Colour reaction was initiated by addition of
30 L malachite green reagent per well and allowed 15 min before
lL of water followed by extensive mixing, split and
l
L to a read-plate encompassing 30 L H2O per
l
l
absorbance reading of the plates at 660 nm. The malachite green
reagent was prepared by mixing 16.4 mL ammonium molybdate
solution (7.5%), 1.64 mL Tween (11%) with 82 mL Malachite
Green/H2SO4-solution (0.7 g Malachite Green to 1000 mL of 10%
(v/v) H2SO4). All compounds were tested in triplicate at 10 differ-
ent concentrations. Control compound (0%, 50% and 100% effect)
was included on each plate. The inhibitory effect of a compound
was calculated by %inhibition = 100 Â [(X À Max)/(Min À Max)],
where X is the readout value in the compound well. Min is the
average signal for 100% inhibition and Max is the average signal
for 0% inhibition in the control wells. Curve fitting was done using
the following equation: y = (A + ((B À A)/(1 + ((x/C)^D)))) where A
and B represents the curve bottom (% inhibition) and the total
amplitude (% inhibition), respectively, whereas C and D are the
apparent IC50 and curve slope, respectively. Some compounds were
tested using a slightly modified protocol at rt instead of 37 °C. The
data transferred well between assays and no temperature effect
was observed with these compounds. See Supplementary data for
full protocol.
4.1.25. [trans-4-({[(2-{4-[2-(Dimethylamino)ethyl]piperidin-1-
yl}quinolin-4-yl)carbonyl]amino}methyl)-cyclohexyl]methyl
tert-butylcarbamate (41)
A 0.5 M aqueous solution of lithium hydroxide (0.639 mL,
0.32 mmol) was added to a solution of methyl trans-4-({[(2-{4-
[2-(dimethylamino)ethyl]piperidin-1-yl-}quinolin-4-yl)carbonyl]-
amino}methyl)cyclohexanecarboxylate 4416 (0.13 g, 0.27 mmol) in
THF (6 mL). The reaction mixture was stirred at rt for 16 h. Addi-
tional 0.5 M aqueous solution of lithium hydroxide (0.639 mL,
0.32 mmol) was added and the reaction mixture was stirred at rt
for 16 h. Toluene was added and the reaction mixture was concen-
trated in vacuo to give the crude carboxylic acid that was used with
no further purification. 4-Methylmorpholine (0.030 mL, 0.27
mmol) and butyl carbonochloridate (0.034 mL, 0.27 mmol) were
added to the acid (0.27 mmol) in THF (10 mL) and the reaction
mixture was stirred for 30 min. Sodium borohydride (0.014 g,
0.38 mmol) followed by MeOH (20 mL) were added and the reac-
tion mixture was stirred at rt for 16 h. The reaction mixture was
then partitioned between CH2Cl2 and saturated aqueous NaCl.
The aqueous phase was extracted with CH2Cl2 and the combined
organic phases were dried (phase separator) and concentrated in
4.3. X-ray structure determination
Hexa-histidine tagged ACC2 CT-domain (aa1685–2431) from
bovine was over-expressed in Escherichia coli BL21 gold (DE3) cells
and purified using Ni2+-NTA affinity chromatography. The histi-
dine-tag was removed by thrombin and separated by another
Ni2+-NTA affinity chromatography step followed by size exclusion
chromatography in a buffer containing 20 mM Tris, 200 mM NaCl,

3052
C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
5% glycerol and 1 mM TCEP at pH 7. The protein was concentrated
to around 20 mg/mL before crystallization at 20 °C using the hang-
ing-drop method. The precipitating solution consisted of 20–25%
PEG4000, 400–800 mM LiCl, 100 mM Tris pH 8–8.5. The ligand
complex was prepared by soaking crystals in 15 mM 28 for 3 days
before freezing the crystals in liquid nitrogen using 20% glycerol in
mother liquor solution as cryo-protectant. X-ray data was collected
at 100 K on beam-line ID29 at the ESRF. The structure was solved
with the molecular replacement using the yeast ACC2 (PDB entry
1OD2) as search model. The final resolution was 2.4 Å, with
R = 23.1, Rfree = 28.3. The structure has been deposited in the PDB
where further refinement details can be found (2X24).
13C3–malonyl-CoA) and 100
lL 30% bovine serum albumin in phos-
phate buffered saline (Sigma, product A9576–50 mL). Complete
extraction and sample purification (removal of lipids) at À20 °C
with the Mixer Mill 300 (Retsch, Germany) was performed for
5 min at 25 Hz after addition of 0.5 mL CHCl3. The upper aqueous
sample extract was evaporated to dryness in HPLC glass vials at
50 °C under a stream of nitrogen, after centrifugation of samples
at 3000 G for 10 min, and analytes were dissolved in 100 lL injec-
tion solvent (0.5% TFA and 2% CH3CN). Separation was achieved in
only 8 min using a binary solvent system and a gradient run (0% B
from 0.0–0.2 min then 100% B from 3.5–5.5 min and 0% B from 5.6
to 7.6 min) where phase A was 10 mM NH4OAc, 5 mM AcOH,
10 mM DIPEA and 2% CH3CN and phase B was 25% phase A in
CH3CN. The column was a Hypersil Gold C18, 30 Â 2.1 mm,
4.4. Inhibition of ACCs in vivo
1.9
l
m (Thermo, product 25002–032130) connected to a precol-
m filter. Malo-
Experimental procedures were in accordance with Swedish
laws governing the use of animals for experimental purposes and
were approved by the Local Ethics Review on Animal Experiments
in Gothenburg region. In vivo inhibition of ACCs was assessed indi-
rectly based on the extent of substance induced malonyl-CoA low-
ering in the liver of male obese Zucker rats. On the day of the
experiment male obese Zucker rats, weighing 370–540 g, were
individually housed in a clean cage without food but with access
to water at 07:00. The animals were weighed and anesthetized
with an intraperitoneal injection of Na-thiobutabarbital (InactinÒ;
RBI, Natick, MA), 180 mg/kg BW at 10:00. They were tracheotom-
ized with PE 240 tubing and breathed spontaneously. One catheter
(PE 50 tubing) was placed in a carotid artery for blood sampling.
Two catheters (PE 50) were placed in one jugular vein for sub-
stance/vehicle infusion and thiobutabarbitol top-up dosing,
respectively. The arterial catheter was maintained patent through-
umn (Thermo, product 25,003–012101) and a 0.5
l
nyl-CoA and the internal standard were then determined
quantitatively by negative electrospray tandem-MS using an Ulti-
ma PT mass spectrometer (Micromass, England) The electrospray
probe was operated at 120 °C, with the capillary at 2.8 kV, desolv-
ation gas at 350 °C, cone gas flow at 50 L/hour and the desolvation
gas flow at 800 L/hour. The waste valve solvent delay was on from
0–2 min and 4–5 min. The mass spectrometer was operated in
multiple reactions monitoring mode with a cone voltage of 50 kV
and dwell time of 0.050 s for all analytes. The mass transitions
were 852 > 808 (malonyl-CoA), 855 > 810
(
13C3–malonyl-CoA),
and optionally 808 > 408 for acetyl-CoA. The collision energy was
25 eV (malonyl-CoA and 13C3–malonyl-CoA) or 35 eV (acetyl-
CoA). For quantitative analysis, tissue samples were calibrated
against 5-point calibration curves ranging from 160 nM to
100 lM prepared in extraction buffer diluted 10 times with 40%
out the experiment by continuous infusion (10
l
L/min) of a sterile
methanol and stored at À80 °C. Calibration samples were pro-
cessed as tissue samples. The detection limits were less than
1 pmol on-column.
saline solution containing sodium citrate (20.6 mM). The acute
experimental phase commenced 13:00 following a 1.5 h post-
surgery stabilization period. Body temperature was monitored
using a rectal thermocouple and maintained at 37.5 °C by means
of external heating. Substances were administered as constant
intravenous infusions for 2 h. At the end of this period, substance
infusion was stopped and the deeply anaesthetized rat was moved
to the dissection table where the liver tissue for malonyl-CoA con-
tent was excised and frozen in liquid N2 as quickly as possible.
Blood was then quickly collected from the carotid cannula for sub-
stance analysis in plasma. The frozen liver samples, immersed in
liquid N2 were broken into smaller pieces (50–100 mg) and
weighed into 2 mL Micro tubes (PP, Sarstedt, Numbrecht,
Germany) and stored in a À80 °C freezer awaiting malonyl-CoA.
4.6. Substance exposures
Normalized substance exposures were calculated as, Expo-
sure = Cpfree/IC50, where IC50 is the estimated free in vitro concen-
tration required to half maximally inhibit rat ACC2 and Cpfree is the
estimated free plasma concentration obtained from the measured
compound concentration Cp as Cpfree = fu  Cp, where fu is the un-
bound fraction of substance in plasma measured in separate exper-
iments under standard conditions (see Supplementary data for
measurement of plasma protein binding).
Blood for substance analysis was collected into 500 l
L MicrovetteÒ
500 K3E tubes (Sarstedt), plasma separated by centrifugation
(10,000 g, 1 min, 4 °C) and plasma stored at À20 °C awaiting
analysis.
Acknowledgments
We thank Lisa Alderin, Åsa Einarsson, Henrik Gradén and Åsa
Månsson for synthesis of some compounds and all involved in
purification and analysis and Anna-Karin Asztély, Anita Dellsén,
Pia Hansson, Pia Thalén, Ann Kjellstedt, Zulma Santisteban and
Helena von Bahr for help with in vitro and in vivo experiments.
We also thank Sara Moses, Sven Sjögren and Carl-Gustav Sigfrids-
son, Johan Wernevik and Christer Westerlund for helpful
discussions.
4.5. Analysis of malonyl-CoA in tissue by ion-pair-HPLC-
tandem-mass spectrometry
Malonyl-CoA was analysed in small tissue samples by a cold,
semi-automated homogenization and extraction method followed
by separation by ion-pair HPLC and detection by negative electro-
spray tandem mass spectrometry. The method is based on
cryogenic sampling of 50–100 mg tissue in 2 mL Sarstedt polypro-
pylene screw cap micro tubes (Sarstedt, product 72.694.007) pre-
filled with six yttrium stabilized (3 mm) zirconium oxide beads
(Retsch, product 05.368.0090). Simultaneous and automated batch
homogenization and initial extraction with the Precellys-24 (Ber-
tin, France) was performed for 2 Â 15 s at power 5000 after addi-
tion of 0.5 mL –20 °C cold extraction buffer (40% MeOH, 5% TFA
Supplementary data
Supplementary data (experimental details for the synthesis of
intermediates on route to 18, 21, 29 and 41 as well as 42 and 44.
Further details on the enzyme assay, details of enzyme preparation
and activity measurement and procedures for plasma protein
binding, solubility, stability in HLM and caco-2 permeability
measurements. Explanation of torsional scan analysis, sequence
and 0.01% dithioerythritol (DTE)), 50 lL internal standard (15 lM

C. Bengtsson et al. / Bioorg. Med. Chem. 19 (2011) 3039–3053
3053
R.; Vajdos, F.; Rajamohan, F.; Kohls, D.; Marr, E.; Zhang, H.; Tong, L.; Tu, M.;
Murdande, S.; Doran, S. D.; Houser, J. A.; Song, W.; Jones, C. J.; Coffey, S. B.;
Buzon, L.; Minich, M. L.; Dirico, K. J.; Tapley, S.; McPherson, R. K.; Sugarman, E.;
Harwood, H. J., Jr.; Esler, W. Bioorg. Med. Chem. Lett. 2010, 20, 2383; (f) Chonan,
T.; Tanaka, H.; Yamamoto, D.; Yashiro, M.; Oi, T.; Wakasugi, D.; Ohoka-Sugita,
A.; Io, F.; Koretsune, H.; Hiratate, A. Bioorg. Med. Chem. Lett. 2010, 20, 3965.
12. Shen, Y.; Volrath, S. L.; Weatherley, S. C.; Elich, T. D.; Anderson, R.; Tong, L.
Patent WO06017515, 2006.
alignments for hACC1, hACC2, rat ACC2 and bACC2, test substance
concentrations in terminal plasma samples and substance parti-
tioning in plasma: bound versus free fractions) associated with this
article can be found, in the online version, at doi:10.1016/
j.bmc.2011.04.014.
13. Anderson, G. W.; McGregor, A. C. J. Am. Chem. Soc. 1957, 79, 6180.
14. Risaliti, A. Ric. Scient. 1958, 28, 561.
References and notes
15. Minakata, S.; Komatsu, M.; Ohshiro, Y. Synthesis 1992, 661.
16. See Supplementary data for synthesis description.
17. Log D measurement based on chromatographic retention times (log k’) from a
method similar to that described in Valkó, K.; Bevan, C.; Reynolds, D. Anal.
Chem. 1997, 69, 2022.
18. Yee, S. Pharm. Res. 1997, 14, 763.
19. Chiou, W. L.; Robbiel, G.; Chung, S. M.; Wu, T.-C.; Ma, C. Pharm. Res. 1998, 15,
1474.
20. (a) Gleeson, M. P. J. Med. Chem. 2008, 51, 817; (b) Waring, M. J. Expert Opin. Drug
Disc. 2010, 5, 235.
21. (a) Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881; (b)
Ryckmans, T.; Edwards, M. P.; Horne, V. A.; Correia, A. M.; Owen, D. R.;
Thompson, L. R.; Tran, I.; Tutt, M. F.; Young, T. Bioorg. Med. Chem. Lett. 2009, 19,
4406.
22. Zhang, H.; Tweel, B.; Li, J.; Tong, L. Structure 2004, 12, 1683.
23. Madauss, K. P.; Burkhart, W. A.; Consler, T. G.; Cowan, D. J.; Gottschalk, W. K.;
Miller, A. B.; Short, S. A.; Tran, T. B.; Williams, S. P. Acta Crystallogr., Sect. D 2009,
65, 449.
24. Waring, M. J. Bioorg. Med. Chem. Lett. 2009, 19, 2844.
25. Sequence Alignments for hACC1, hACC2, rat ACC2 and bACC2 are shown in
Supplementary data.
26. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648; (b) Lee, C.; Yang, W.; Parr, R. G.
Phys. Rev. B 1988, 37, 785; (c) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.;
Frisch, M. J. J. Phys. Chem. 1994, 98, 11623.
27. Abu-Elheiga, L.; Almarza-Ortega, D. B.; Baldini, A.; Wakil, S. J. J. Biol. Chem.
1997, 272, 10669.
28. Redfern, W.; Carlsson, L.; Davis, A. S.; Lynch, W. G.; MacKenzie, I.; Palethorpe,
S.; Siegl, P. K. S.; Strang, I.; Sullivan, A. T.; Wallis, R.; Camm, A. J.; Hammond, T.
G. Cardiovasc. Res. 2003, 58, 32.
29. Rat ACC2 and ACC1 in vitro assay used to test compounds 1, 33 and 37 was
performed in an analogous manner to the hACC2 and hACC1 in vitro assays
described in the Section 4 using rat ACC2 and ACC1 in place of hACC2 and
hACC1, respectively.
1. Data was obtained from the World Health Organization website: http://
www.who.int/diabetes/en/.
2. Corbett, J. W. Expert Opin. 2009, 19, 943.
3. Polakis, S. E.; Guchhait, R. B.; Zwergel, E. S.; Lane, M. D.; Cooper, T. G. J. Biol.
Chem. 1974, 249, 6657.
4. Harwood, H. J., Jr.; Petras, S. F.; 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.
5. (a) Harwood, H. J., Jr. Expert Opin. Ther. Targets 2005, 9, 267; (b) Tong, L.;
Harwood, H. J., Jr. J. Cell. Biochem. 2006, 99, 1476.
6. An, J.; Muoio, D. M.; Shiota, M.; Fujimoto, Y.; Cline, G. W.; Shulman, G. I.; Koves,
T. R.; Stevens, R.; Millington, D.; Newgardet, C. B. Nat. Med. 2004, 10, 268.
7. (a) Abu-Elheiga, L.; Matzuk, M. M.; Abo-Hashema, K. A. H.; Wakil, S. J. Science
2001, 291, 2613; (b) Abu-Elheiga, L.; Oh, W.; Kordari, P.; Wakil, S. J. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 10207; (c) Oh, W.; Abu-Elheiga, L.; Kordari, P.; Gu, Z.;
Shaikenov, T.; Chirala, S. S.; Wakil, S. J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
1384.
8. Savage, D. B.; Choi, C. S.; Samuel, V. T.; Lui, Z. X.; Zhang, D.; Wang, A.; Zhang, X.
M.; Cline, G. W.; Yu, X. X.; Geisler, J. G.; Bhanot, S.; Monia, B. M.; Shulman, G. I. J.
Clin. Invest. 2006, 116, 817.
9. 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.
10. Olson, D. P.; Pulinilkunnil, T.; Cline, G. W.; Shulman, G. I.; Lowell, B. B. Proc.
Natl. Acad. Sci. U.S.A. 2010, 107, 7598.
11. (a) Corbett, J. W.; Harwood, H. J., Jr. Recent Patents Cardiovasc. Drug Disc. 2007,
2, 162; (b) Clark, R. F.; Zhang, T.; Wang, X.; Wang, R.; Zhang, X.; Camp, H. S.;
Beutel, B. A.; Sham, H. L.; Gu, Y. G. Bioorg. Med. Chem. Lett. 2007, 17, 1961; (c)
Shinde, P.; Srivastava, S. K.; Odedara, R.; Tuli, D.; Munshi, S.; Patel, J.; Zambad,
S. P.; Sonawane, R.; Gupta, R. C.; Chauthaiwale, V.; Dutt, C. Bioorg. Med. Chem.
Lett. 2009, 19, 949; (d) Haque, T. S.; Liang, N.; Golla, R.; Seethala, R.; Ma, Z.;
Ewing, W. R.; Cooper, C. B.; Pelleymounter, M.-A.; Poss, M. A.; Cheng, D. Bioorg.
Med. Chem. Lett. 2009, 19, 5872; (e) Corbett, J. W.; Freeman-Cook, K. D.; Elliott,