Piperazine oxadiazole inhibitors of acetyl-CoA carboxylase

Article abstract of DOI:10.1021/jm401601s

Acetyl-CoA carboxylase (ACC) is a target of interest for the treatment of metabolic syndrome. Starting from a biphenyloxadiazole screening hit, a series of piperazine oxadiazole ACC inhibitors was developed. Initial pharmacokinetic liabilities of the piperazine oxadiazoles were overcome by blocking predicted sites of metabolism, resulting in compounds with suitable properties for further in vivo studies. Compound 26 was shown to inhibit malonyl-CoA production in an in vivo pharmacodynamic assay and was advanced to a long-term efficacy study. Prolonged dosing with compound 26 resulted in impaired glucose tolerance in diet-induced obese (DIO) C57BL6 mice, an unexpected finding.

Full text of DOI:10.1021/jm401601s

Article
pubs.acs.org/jmc
Piperazine Oxadiazole Inhibitors of Acetyl-CoA Carboxylase
Matthew P. Bourbeau,*,† Aaron Siegmund,† John G. Allen,† Hong Shu,† Christopher Fotsch,†
Michael D. Bartberger,† Ki-Won Kim,‡ Renee Komorowski,‡ Melissa Graham,‡ James Busby,‡
Minghan Wang,‡ James Meyer,§ Yang Xu,§ Kevin Salyers,§ Mark Fielden,∥ Murielle M. Veniant,‡
́
and Wei Gu‡
‡
§
∥
†Therapeutic Discovery, Metabolic Diseases Research, Pharmacokinetics/Drug Metabolism, and Comparative Biology & Safety
Sciences, Amgen Inc., 1 Amgen Center Dr, Thousand Oaks, California 91320, United States
S
* Supporting Information
ABSTRACT: Acetyl-CoA carboxylase (ACC) is a target of
interest for the treatment of metabolic syndrome. Starting
from a biphenyloxadiazole screening hit, a series of piperazine
oxadiazole ACC inhibitors was developed. Initial pharmacoki-
netic liabilities of the piperazine oxadiazoles were overcome by
blocking predicted sites of metabolism, resulting in com-
pounds with suitable properties for further in vivo studies. Compound 26 was shown to inhibit malonyl-CoA production in an in
vivo pharmacodynamic assay and was advanced to a long-term efficacy study. Prolonged dosing with compound 26 resulted in
impaired glucose tolerance in diet-induced obese (DIO) C57BL6 mice, an unexpected finding.
synthesis in a rate-limiting manner. ACC2 is primarily
expressed in muscle tissue, where it is associated with the
mitochondrial membrane. Malonyl-CoA produced by ACC2
serves as a negative regulator of carnityl palmitate transfer
protein 1 (CPT1). CPT1 is the protein primarily responsible
for transport of fatty acyls across the mitochondrial membrane
for subsequent β-oxidation. Thus, malonyl-CoA production by
ACC2 activity serves to directly inhibit mitochondrial fatty acid
oxidation.10 Given the roles of the ACCs in both the synthesis
and metabolism of fatty acids, inhibition of the ACCs would be
expected to decrease fatty acid synthesis while simultaneously
increasing fatty acid oxidation.
ACC activity has been extensively studied in mouse knockout
models. Attempts to generate whole animal ACC1−/− mice
resulted in embryonic lethality.11 However, liver-specific ACC1
knockout mice (LACC1−/−) have been successfully generated
by the groups of Wakil and Kusunoki.12,13 In Wakil’s case, a
75% reduction in liver malonyl-CoA levels was observed along
with a 50% decrease in fatty acid synthesis in the liver.
Conversely, liver samples from the LACC1−/−mice from
Kusunoki’s lab showed no change in malonyl-CoA levels or
fatty acid synthesis but did show an increase in ACC2 protein
expression level and activity, suggesting that ACC2 was being
upregulated to compensate for the ACC1 deletion.
INTRODUCTION
■
Metabolic syndrome is a combination of related conditions that
has been shown to affect >20% of the adult population in the
United States. It is associated with an increased risk of type 2
diabetes and cardiovascular disease.1 Specific complications that
result from metabolic syndrome include insulin resistance, high
blood pressure, central obesity, decreased HDL cholesterol, and
elevated triglycerides. The growing prevalence of metabolic
syndrome in the population has not gone unnoticed, and a
number of different approaches for the treatment of this
condition are being investigated.2,3 Although there is not a clear
understanding of all of the contributing factors to the
pathogenesis of metabolic syndrome, abnormal fatty acid
metabolism is considered to be a key element.4,5 Studies have
demonstrated positive associations between insulin resistance
and intracellular accumulations of triglycerides and fatty acid
metabolites in insulin-responsive tissues such as muscle and
liver. Fatty acid metabolites such as acyl-CoA, diacylglycerol,
and ceramides have been shown to interfere with insulin
signaling.6,7 Accumulation of intracellular fatty acid metabolites
depends on fatty acid uptake and the rates of fatty acid
synthesis and oxidation. As a result, drugs that block fatty acid
uptake, inhibit fatty acid synthesis, and/or increase fatty acid
oxidation might prove useful for the treatment of metabolic
syndrome.
Several groups have developed ACC2 knockouts that are
viable and have normal life spans. The most extensively
characterized ACC2−/− mice were generated by Wakil and co-
workers.14−16 These mice exhibit reduced malonyl-CoA levels
and increased fatty acid oxidation in muscle and heart tissue
relative to wild-type mice as well as increased total energy
The acetyl-CoA carboxylases (ACCs) are a class of enzymes
that catalyze the conversion of acetyl-CoA to malonyl-CoA and,
given their role in fatty acid synthesis and metabolism, have
emerged as attractive targets for the treatment of metabolic
syndrome.8 There are two characterized isoforms of ACC
(known as ACC1 and ACC2) that have divergent functions.9
ACC1 is located in the cytosol and is primarily expressed in
liver and fat tissue. ACC1 catalyzes the initiation of fatty acid
Received: October 15, 2013
Published: December 2, 2013
© 2013 American Chemical Society
10132
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (a) potassium (4-cyanophenyl)trifluoroborate, KOAc, 1,1-bis[(di-tert-butyl-p-methylaminophenyl]palladium(II) chloride,
3:1 MeCN/water, 90 °C, 29%; (b) NH2OH(aq), cat. AcOH, EtOH, 80 °C, propionic anhydride, pyridine, 80 °C, 32%.
a
Scheme 2
a
Reagents and conditions: (a) nPrI, Cs2CO3, 20 °C, DMF, 99%; (b) chloro(2-dicyclohexylphosphino-2′,4′,6′-tri-isopropyl-1,1′-biphenyl)[2-(2-
aminoethyl)phenyl]palladium(II) methyl-tert-butylether adduct, NaOtBu, dioxane, 80 °C, 30−84%; (c) TFA; (d) CNBr, DIEA, CH2Cl2, 0 °C, 25−
88% for two steps; (e) NH2OH-HCl, DIEA, EtOH, 90 °C; (f) 2,2,5-trimethyl-1,3-dioxane-4,6-dione, pyridine, 90 °C, 41% for two steps or
(EtCO)2O, 120 °C, 25−51% for two steps.
expenditure. When placed on a high-fat/high-calorie diet, the
ACC2−/− mice showed improved insulin sensitivity and lipid
profiles as well as reduced weight gain and adiposity relative to
wild-type mice. More recently, the groups of Cooney and
Lowell have published phenotypic analysis of ACC2−/− mice
generated by their respective groups.17,18 The ACC2−/− mice
from Cooney’s lab show reduced malonyl-CoA levels in heart
and muscle tissue as well as increased whole-body fatty acid
oxidation but do not show improvements in any metabolic
parameters relative to wild-type mice when fed a high-fat/high-
calorie diet. In the case of the ACC2−/− mice described by
Lowell’s group, only a modest reduction in malonyl-CoA levels
was observed in heart tissue, and no additional metabolic
changes were apparent. The cause of the observed differences in
the metabolic phenotypes of the three published ACC2−/−
mouse models is not obvious.
Driven largely by the initial ACC2 knockout data published
by Wakil, there have been significant efforts targeting ACC
inhibition as a treatment for metabolic syndrome.19−25 Our
own work in this area began with a high-throughput screening
(HTS) campaign initially targeting human ACC2 (hACC2),
with follow-up counter screening against human ACC1
(hACC1). Our screening campaign identified biphenyl
oxadiazole 1 as a starting point for further investigation
(hACC2 IC50 = 841 nM, hACC1 IC50 = 4280 nM). Given the
lethality of ACC1 embryonic knockout and the conflicting
phenotypes observed between the two conditional ACC1−/−
mouse strains, we were uncertain regarding the extent to which
pharmacological ACC1 inhibition would be necessary and/or
tolerated in vivo. For an initial proof-of-concept tool, we felt
that a compound with dual ACC1/2 inhibitory activity had the
greatest chance of producing the largest metabolic effect.
However, there could potentially be an advantage to having an
ACC2-selective compound if ACC1 inhibition proved to be
unnecessary or a safety concern; therefore, we chose to evaluate
ACC1/ACC2 selectivity to guide efforts toward compounds
with greater ACC2 selectivity for subsequent studies.
CHEMISTRY
■
Synthesis of Biphenyl Oxadiazole (2, Scheme 1). 1-
Bromo-2-methyl-4-proproxylbenzene underwent palladium-cat-
alyzed cross coupling with potassium (4-cyanophenyl)-
trifluoroborate to afford nitrile 27. Nitrile 27 was sequentially
treated with hydroxylamine and propionic anhydride to afford
oxadiazole 2.
Synthesis of Piperazine Oxadiazoles from Mono-Boc-
Protected Piperazines (3−9, Scheme 2). Phenyl piper-
azines 31, 33, 35, 37, 39, 41, and 43 were prepared by
10133
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
a
Scheme 3
a
Reagents and conditions: (a) 29, chloro(2-dicyclohexylphosphino-2′,4′,6′-tri-isopropyl-1,1′-biphenyl)[2-(2-aminoethyl)phenyl]palladium(II)
methyl-tert-butylether adduct, NaOtBu, dioxane, 120 °C, 55%; (b) CNBr, KOAc, MeOH, 72%; (c) NH2OH-HCl, DIEA, EtOH, 90 °C; (d)
(EtCO)2O, 120 °C, 26% for two steps.
a
Scheme 4
a
Reagents and conditions: (a) CNBr, DIEA, CH2Cl2, 0 °C, 99%; (b) NH2OH(aq), EtOH, 20 °C, 99%; (c) (EtCO)2O, pyridine, 120 °C, 47%; (d)
HCl, dioxane, 20 °C, 95%; (e) PhI, N,N,-dimethylglycine, CuI, Cs2CO3, dioxane, 95 °C, 44% or RBr, Cs2CO3, DMF, 20 °C, 65−98%; (f) chloro(2-
dicyclohexylphosphino-2′,4′,6′-tri-isopropyl-1,1′-biphenyl)[2-(2-aminoethyl)phenyl]palladium(II) methyl-tert-butylether adduct, NaOtBu, dioxane,
100 °C, 8−29%.
a
Scheme 5
a
Reagents and conditions: (a) nPrI, Cs2CO3, 20 °C, DMF, 88−93%; (b) S-2-methylpiperazine, Pd2(dba)3, rac-BINAP, NaOtBu, toluene, 110 °C,
65−99%; (c) CNBr, DIEA, CH2Cl2, 0 °C, 99%; (d) NH2OH-HCl, DIEA, EtOH, 90 °C; (e) (EtCO)2O, 120 °C, 47−56%.
palladium-catalyzed cross-coupling of the appropriate aryl
halide with mono-Boc-protected piperazines using Buchwald’s
X-Phos palladacycle.26 Subsequent cleavage of the Boc group
followed by condensation with cyanogen bromide and
diisopropylethylamine afforded amino nitriles 32, 34, 36, 38,
40, 42, and 44. Sequential reaction of the amino nitrile with
10134
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
a
Scheme 6
a
Reagents and conditions: (a) EtI, Cs2CO3, DMF, 20 °C, 32−95%; (b) for 63 and 66: S-2-methylpiperazine, Pd2(dba)3, BINAP, NaOtBu, toluene,
120 °C, 41−51%; for 68: Pd2(dba)3, 2-dicyclohexylphosphino-2,6-diisopropoxy-1,1-biphenyl, NaOtBu, toluene, 100 °C, 55%; (c) CNBr, DIEA,
CH2Cl2, 0 °C, 100%; (d) NH2OH, EtOH, 20 °C, quant; (e) (EtOCO)2O, pyridine, 80 °C, 30−31%; (f) (CH3CF2CO)2O, 70 °C, 21−23%.
a
Scheme 7
a
Reagents and conditions: (a) NH2OH-HCl, DIEA, EtOH, 20 °C, 100%; (b) AcNHCR1R2COOH, HATU, DIEA, 80 °C, 53−67%.
hydroxylamine and 2,2,5-trimethyl-1,3-dioxane-4,6-dione af-
forded desired piperazine oxadiazoles 3−9.
11−14. Analogue 15 was prepared from commercially available
55.
Synthesis of Trifluoromethylpiperazine Oxadiazole
(10, Scheme 3). 2-(Trifluoromethyl)piperazine and 1-bromo-
2-methyl-4-propoxybenzene were coupled using Buchwald’s X-
Phos palladacycle to form trifluoromethylpiperazine 45. The
secondary amine was then treated with cyanogen bromide and
potassium acetate, and resulting amino nitrile 46 was reacted
sequentially with hydroxylamine and propionic anhydride,
affording trifluoromethylpiperazine oxadiazole 10.
Synthesis of Chlorophenyl and Trifluoromethylphen-
yl Analogues (16−17, Scheme 5). 4-Bromo-3-chlorophenol
and 4-bromo-3-(trifluoromethyl)phenol were alkylated with 1-
iodopropane to afford propyl ethers 56 and 59, respectively.
Compounds 56 and 59 were then cross coupled with (S)-2-
methylpiperazine, affording piperazines 57 and 60. Reaction
with cyanogen bromide and diisopropylethylamine afforded
amino nitriles 58 and 61. Sequential reaction of the amino
nitrile with hydroxylamine and propionic anhydride afforded
oxadiazoles 16 and 17.
Synthesis of Pyridine and Difluoro Analogues (18−21,
Scheme 6). Bromopyridines 62 and 65 were cross-coupled
with (S)-2-methylpiperazine mediated by Pd2(dba)3 and rac-
BINAP to afford piperazines 63 and 66. Aryl bromide 52 was
cross-coupled with (S)-2-methylpiperazine mediated by
Pd2(dba)3 and RuPhos to afford piperazine 68. Treatment of
63, 66, and 68 with cyanogen bromide and diisopropylethyl-
Synthesis of Phenoxyether Analogues (11−15,
Scheme 4). (S)-tert-Butyl 3-methylpiperazine-1-carboxylate
was treated with cyanogen bromide and diisopropylethylamine
to afford amino nitrile 47. The amino nitrile was reacted
sequentially with hydroxylamine and propionic anhydride
followed by Boc deprotection with TFA to afford trifluor-
omethylpiperazine oxadiazole 50. Aryl ethers 51−54 were
prepared from 4-bromo-3-methylphenol and were cross-
coupled with trifluoromethylpiperazine oxadiazole 50 to afford
10135
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
a
Scheme 8
a
Reagents and conditions: (a) BocNHCR1R2COOH, HATU, DIEA, DMF, 80 °C; (b) TFA, CH2Cl2, 20 °C; (c) Ac2O, DIEA, CH2Cl2, 20 °C, 42−
44% over three steps.
Figure 1. Initial development from HTS hit. Data are reported as the mean SD, n = 3.
amine afforded amino nitriles 64, 67, and 69. Sequential
treatment of 64 and 67 with hydroxyl amine and propionic
anhydride afforded pyridylpiperazine oxadiazoles 19 and 20,
respectively. Sequential treatment of 69 and 64 with hydroxyl
amine and ethyl-2,2-difluoropropionate afforded difluoroox-
adiazoles 18 and 21, respectively.
Synthesis of Oxadiazole Acetamide Analogues (22−
24, Scheme 7). Amino nitrile 69 was treated with hydroxyl-
amine to form carboximidamide 70. HATU-mediated coupling
of 70 with N-acetylglycine afforded oxadiazole acetamide 22.
HATU-mediated coupling of 70 with (S)- and (R)-2-
acetamidopropanoic acid afforded oxadiazole acetamides 23
and 24, respectively.
Synthesis of Dimethyl and Cyclobutyl Oxadiazole
Acetamide Analogues (25−26, Scheme 8). Carboximida-
mide 70 was subjected to HATU-mediated coupling with Boc-
alpha-methylalanine. Boc cleavage with TFA and treatment of
the resulting free amine with acetic anhydride afforded dimethyl
oxadiazole acetamide 25. Separately, carboximidamide 70 was
subjected to HATU-mediated coupling with 1-(tert-
butoxycarbonylamino)cyclobutanecarboxylic acid. Boc cleavage
with TFA and treatment of the resulting free amine with acetic
anhydride afforded cyclobutyl oxadiazole acetamide 26.
recombinant cell line in which the ACC2 coding sequence
was expressed under the control of an inducible tetracycline
promoter utilizing the CHOK1-TREx platform.27 Compound 1
was found to inhibit the production of malonyl-CoA in the
ACC2 cellular assay and was, in fact, more potent in this cell-
based assay than in the biochemical assay (hACC2 IC50 = 841
nM, cell IC50 = 359 nM). The ACC2 protein expressed in the
CHOK1 cells had a 24 amino acid deletion at the C-terminus of
the protein. These residues are thought to be responsible for
the mitochondrial association of ACC2. The cellular ACC2
protein also possessed a constitutively activating S221A
mutation. It is possible that these changes resulted in the
unexpected cell shift. Nevertheless, we felt that this cellular
assay would serve as a useful tool for assessing the ability of
advanced compounds to inhibit directly malonyl-CoA produc-
tion in cells and thus we used the cellular assay to profile key
compounds as the SAR of the oxadiazole series developed.
Initial efforts to improve potency by modifying the propyl
ether of 1 were unsuccessful. Therefore, a more conservative
approach was undertaken, and analogues with one F or Me at
each unsubstituted position of the two phenyl rings were
prepared. Of these eight analogues, the majority lost activity in
the hACC2 and hACC1 enzymatic assays (data not shown).
Compound 2, however, was more promising, showing
improved activity in the ACC2 and ACC1 enzymatic assays
(IC50 = 126 and 1460 nM, respectively) as well as the ACC2
cellular assay (IC50 = 103 nM) (Figure 1). With this result in
hand, we attempted to replace the central phenyl ring of 2 with
a saturated ring in an effort to break the aromaticity of the lead
molecules and potentially to improve overall physical proper-
ties.28 These efforts resulted in the synthesis of piperazine
oxadiazole 3, which exhibited ACC inhibition comparable to 2.
Encouraged by the results obtained with 3, we next examined
more substituted piperazine analogues (Table 1). Methylation
of the R3 position resulted in a modest boost in ACC2 potency,
with S enantiomer 5 in particular showing inhibition of ACC2
in the enzymatic assay of 71 nM. Removing the R1 methyl
group, as shown in 6, resulted in a loss of potency, indicating
DISCUSSION
■
HTS of the Amgen sample collection identified biphenyl
oxadiazole 1 as a starting point for further investigation (Figure
1). In our primary enzymatic assay, which measured the
conversion of acetyl-CoA to malonyl-CoA by LC/MS/MS, 1
was found to inhibit hACC2 and hACC1 activity with IC50’s of
841 and 4280 nM, respectively, and was found to be
competitive with acetyl-CoA. After determining that we could
conveniently quantify malonyl-CoA levels by LC/MS/MS, we
sought to develop a cellular assay that also measured malonyl-
CoA reduction in a similar manner. Attempts to measure ACC
inhibition in native cell lines was hampered by the relatively low
baseline levels of malonyl-CoA (FAO, L6, and H9C2 cells were
examined). To circumvent this issue, we constructed a
10136
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
a
ACC2 enzymatic ratio. In addition to alkyl ethers, phenyl ether
14 was also examined, but it showed a significant reduction in
potency relative to the propyl and ethyl ethers.
Table 1. SAR Investigations on Piperazine
Given the potency of 12 in our ACC2 enzymatic assay, we
profiled 12 in our ACC2 cellular assay, where it exhibited an
IC50 of 39 nM. (Figure 2) Compound 12 was then evaluated
for in vitro microsomal stability where it had rapid human and
rat microsomal turnover (HLM and RLM). In an in vivo
pharmacokinetic (PK) study in rats, 12 also exhibited a high
clearance of approximately 78% of liver blood flow, indicating
good in vitro prediction of in vivo clearance. Additionally, all
attempts to utilize 12 in in vivo studies measuring changes in
tissue levels of malonyl-CoA were met with no success.
Therefore, our next area of focus was to reduce PK liabilities of
12. To this end, 12 was analyzed using the Metasite metabolism
prediction program.29 Three potential sites of metabolism were
identified: the ethyl ether, the aryl methyl group, and the
oxadiazole ethyl group. A number of different analogues were
made in an attempt to block the predicted sites of metabolism
(Figure 3). Efforts to block the metabolism of the aryl ether by
replacement of the alkyl group with a CF3 led to an analogue
with significantly reduced potency (15). The aryl methyl group
could be replaced with either a chloro (16) or CF3 (17) group
with a modest loss of ACC2 enzymatic potency. In the case of
CF3 analogue 17, there was some improvement seen in the
HLM and RLM values. A similar improvement in the HLM and
RLM values was observed by difluorination of the benzylic
position of the oxadiazole ethyl group (18) as well as the
replacement of the phenyl group with pyridine isomers (19 and
20). In all three cases, however, there was a similar loss of
ACC2 enzymatic potency. Combination of the pyridine A ring
with the gem difluoroethyl oxadiazole did not result in additive
improvement in the HLM and RLM values (21).
entry
R1
R2
R3
hACC2 IC50 (nM) hACC1 IC50 (nM)
3
Me
Me
Me
H
H
H
H
H
Me
H
H
H
H
391 135
178 62
4010 1350
1028 102
1910 322
>10 000
4
Me (R)
Me (S)
Me (S)
H
5
71
9
6
2580 226
>10 000
7
H
>10 000
8
Me
Me
Me
Et (S)
iPr (S)
CF3
231 70
335 30
166 28
3930 240
>10 000
9
10
492 36
a
Data are reported as the mean SD, n ≥ 3.
that substitution on both the phenyl and piperazine rings
improved potency. Shifting the methyl group to the R2 position
of the piperazine ring (7) resulted in a loss of enzyme activity
(IC50 > 10 000 nM). Attempts to place groups larger than
methyl at the R3 position (8, 9, and 10) resulted in reduced
potency relative to 5, indicating that methyl was the optimal
substituent at the R3 position for this set of compounds.
Holding the new methyl piperazine central ring constant, we
examined variations to the aryl ether portion of the molecule.
We found that we were able to vary the length of the alkyl chain
and retain some potency (Table 2). In particular, methyl, ethyl,
a
Table 2. SAR Investigations on Alkyl Ether
Although the previously discussed examples gave us some
encouragement that we could improve upon the poor
microsomal stability of 12, we were much more excited by
the improvements that we saw by substituting the ethyl
oxadiazole group with acetamides (Figure 4). Compound 22
retained reasonable potency in the ACC2 enzyme assay but
exhibited almost no oxidative metabolism in either human or
rat liver microsomes. Methylation at the position alpha to the
acetamide (23 and 24) led to compounds with a slight increase
in microsomal turnover. However, these compounds exhibited
significantly better ACC enzyme activity and, in the case of 23,
potency in the ACC2 cellular assay was in the single-digit
nanomolar range. Dimethyl and cyclobutyl analogues 25 and 26
entry
R1
hACC2 IC50 (nM) hACC1 IC50 (nM) hACC1/hACC2
5
Pr
71 9.5
486 37
37 3.2
273 31
789 72
1910 322
>10 000
27×
11
12
13
14
Me
Et
>20×
123×
>36×
>13×
4570 258
>10 000
iPr
Ph
>10 000
a
Data are reported as the mean SD, n ≥ 3.
and isopropyl ethers were examined (11, 12, and 13). Ethyl
ether 12 not only retained potency on the ACC2 enzymatic
assay but also showed a significant improvement in the ACC1/
Figure 2. Analysis of 12. In vivo PK analysis was performed in male Sprague−Dawley rats, n = 3.
10137
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
Figure 3. Efforts to address metabolic stability.
Figure 4. Oxadiazole amide analogues.
showed improved microsomal stability relative to the
monomethyl analogues with retention of the excellent cellular
potency. It was surprising that these amide analogues showed
such a profound improvement in metabolic stability over 12,
especially considering that difluoro analogue 18, which was also
designed to address metabolism of the oxadiazole ethyl group,
did not show such a significant improvement. Our working
hypothesis for this observation centers on the amide’s effect on
lowering cLogP values. The amide analogues have cLogP values
that are ∼1.5−2.5 log units lower than 12, and it may be that
the increased polarity reduces recognition by P450 enzymes
and thus reduces metabolism.30
stability, these compounds were further evaluated for their
potential as in vivo proof-of-concept tools. In particular, 26 was
extensively profiled (Figure 5). As had been observed with
previous compounds, 26 was more potent in the cell-based
assay than in the in vitro assay (cell IC50 = 8.2 nM). Mouse
ACC1 and ACC2 (mACC1 and mACC2) enzyme activity was
similar to that observed in the corresponding human ACC
assays. Compound 26 was readily formulated at concentrations
needed for in vivo dosing and possessed low clearance and high
exposure in both mouse and rat PK studies. There was no
inhibition of human cytochrome P450 3A4 or 2D6 observed
with 26, and the plasma protein binding of 26 was determined
to be 97.5−98.6% across species. Compound 26 was profiled in
an in vitro pharmacology panel consisting of 116 receptors and
Given the favorable ACC1/2 enzyme potency of the
oxadiazole amides coupled with the improved microsomal
10138
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
Figure 5. PK properties of 26. In vivo PK analysis was performed in male Sprague−Dawley rats, n = 3, and male C57BL6 mice, n = 3, as indicated.
enzymes (Cerep, Inc., Redmond, WA) and an in vitro panel of
100 kinases (Ambit Biosciences, San Diego, CA). The only
additional target on which 26 demonstrated activity was the
benzodiazepine receptor (IC50 = 710 nM), which was thought
to be of little impact on the in vivo end points that we were
going to measure. Compound 26 was also nonmutagenic in a
bacterial mutation (Ames) assay. Thus, we believed 26
possessed the required potency, selectivity, and PK properties
to be a useful in vivo proof-of-concept tool.
was conducted in diet-induced obese (DIO) C57BL6 mice with
10 mice per dose group. The mice were administered daily (po)
for 31 days with vehicle or 15, 35, and 75 mg/kg/day of 26.
Body-weight changes were monitored daily, and food intake
was monitored weekly. After 21 days of dosing, blood samples
were taken to measure glucose, insulin, triglycerides, cholester-
ol, free fatty acids, alanine transaminase (ALT), and aspartate
transaminase (AST) levels. At 28 days of dosing, an oral glucose
tolerance test (GTT) was performed. Three days later, the mice
were euthanized, and tissues were collected to measure terminal
PD and other parameters. The 75 mg/kg dose group exhibited
severe weight loss on the tenth day of administration and
therefore had to be dropped from the study because of animal
welfare concerns (Figure 7). The 10 and the 35 mg/kg doses
were tolerated for the duration of the study, and the 35 mg/kg
dose showed a small but significant decrease in weight gain
relative to the control group, with no concomitant change in
food intake. Surprisingly, however, there was a significant
increase in plasma glucose levels for the animals administered
26 at both 10 and 35 mg/kg relative to control. Additionally,
when the animals were subjected to an oral GTT, the mice in
the 35 mg/kg dose group exhibited increased glucose AUC
relative to the control group. It is not clear whether these
observations resulted directly from ACC inhibition. Terminal
PD conducted 25 h after the last dose of 26 showed significant
inhibition of malonyl-CoA in both liver and heart tissue. Of
particular note, the 35 mg/kg dose showed malonyl-CoA
reductions of 42% in liver and 67% in heart, with a
corresponding unbound drug concentration of 618 nM. This
data is very similar to the 8 h time point from the time-course
PD experiment (44% malonyl-CoA reduction in liver and 68%
malonyl-CoA reduction in heart, with an unbound drug
concentration of 882 nM), suggesting that the PK/PD
relationship of drug concentration to malonyl-CoA lowering
remained relatively constant through the course of the study. In
addition to the previously described findings, there was a slight
but significant (∼10%) elevation of cholesterol levels in the 35
mg/kg group of the study relative to control. There were no
significant differences in insulin, triglyceride, free fatty acid,
ALT, or AST levels in either dose group relative to the control
group.
We next advanced 26 to a mouse pharmacodynamic (PD)
study to examine the effect on malonyl-CoA levels in liver
(ACC1-rich tissue) and heart (ACC2-rich tissue) (Figure 6). At
Figure 6. (a) Dose-response PD study. C57BL6 mice, n = 3 per dose
group. Malonyl-CoA measured by LC/MS/MS 1 h post dose. * = p
value < 0.05. Free drug concentration (plasma concentration ×
unbound fraction in nanomolar): 25 mg/kg: 708 76, 50 mg/kg: 988
89, and 100 mg/kg: 1482
141. (b) Time-course PD study,
C57BL6 mice, n = 3 per dose group. Malonyl-CoA measured by LC/
MS/MS 1 h post dose. * = p value < 0.05. Free drug concentration
(plasma concentration × unbound fraction in nanomolar): 1 h: 1165
68, 4 h: 1142 133, 8 h: 882 43, and 12 h: 498 59.
25, 50, and 100 mg/kg po doses, 26 showed a significant
reduction in malonyl-CoA levels in both the liver and heart 1 h
postdose. There appeared to be an effective maximal effect at
∼85% inhibition, consistent with observations made by
scientists at Pfizer conducting similar experiments with pan-
ACC inhibitors.25 In a 12 h time course study, a 25 mg/kg dose
of 26 showed >50% reduction in malonyl-CoA levels between 4
and 8 h in liver and 8 to 12 h in heart tissue. At the 8 h time
point, the unbound plasma drug concentration was 882 nM,
approximately 10-fold higher than the mACC2 enzyme IC50
and 2-fold higher than the mACC1 enzyme IC50.
Our initial foray into ACC as a target for the treatment of
metabolic syndrome/diabetes was largely inspired by the
pioneering ACC2 knockout work of Wakil’s group.14−16
Subsequent studies by Cooney and Lowell called into question
the metabolic benefits of ACC2 knockout.17,18 We developed a
Having confirmed that dosing 26 resulted in significant
reduction of malonyl-CoA levels in vivo, we next chose to study
26 in a long-term in vivo study to evaluate the effect of ACC
inhibition on a broad set of metabolic parameters. The study
10139
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
Figure 7. Effect of long-term ACC inhibition with 26 on DIO C57BL6 mice. * = p value < 0.05, *** = p value < 0.001. (a) Change in body weight
over 27 days of dosing. (b) Change in fed plasma glucose levels after 21 days of dosing. (c) Glucose tolerance test after 28 days of dosing. (d)
Percent reduction of malonyl-CoA relative to control after 31 days of dosing. Free drug concentration levels (plasma concentration × unbound
fraction in nanomolar): 15 mg/kg: 95 27 and 35 mg/kg: 618 106.
Notes
series of oxadiazole ACC inhibitors with varying degrees of
ACC1/ACC2 selectivity, the most potent members of which
were from an oxadiazole amide subseries. Oxadiazole amide 26
possessed good in vivo PK properties and was shown to lower
malonyl-CoA levels in C57BL6 mice in both liver and heart
tissue, suggesting in vivo inhibition of ACC1 and ACC2. A 31
day efficacy experiment was conducted by dosing 26 in DIO
mice. Terminal PD studies showed a significant reduction in
malonyl-CoA levels in liver and heart tissue, and a decrease in
body-weight gain relative to control animals was observed.
However, the glucose tolerance of the mice dosed with 26
actually worsened relative to control mice, and serum glucose
levels also increased. There were no beneficial changes to any
fatty acid related measurements observed in the animals, which,
given the reportedly direct effect of ACCs on such metabolic
parameters, was surprising. In a publication that appeared
subsequent to the completion of our experiments, scientists at
Sanofi-Aventis observed a similar lack of desired efficacy with
their ACC inhibitors in long-term studies conducted in both
DIO mice as well as in Zucker diabetic fatty (ZDF) rats.31 We
feel that these results, coupled with the ACC2 knockout studies
of Cooney and Lowell, suggest that ACC inhibition may not, in
fact, be a useful strategy for the treatment of metabolic disease/
diabetes, at least as represented in rodent models of disease.
The authors declare the following competing financial
interest(s): The corresponding author is an employee and
stock holder of Amgen, Inc.
DEDICATION
■
This paper is dedicated to the memory of Kevin Salyers, our
colleague and friend.
ABBREVIATIONS USED
■
Ac, acetyl; ACC, acetyl-CoA carboxylase; AcOH, acetic acid;
ALT, alanine transaminase; AST, aspartate transaminase; AUC,
area under the curve; BINAP, 2,2′-bis(diphenylphosphino)-
1,1′-binaphthyl; Boc, tert-butyloxycarbonyl; CHOK1, Chinese
hamster ovary K1 cells; CL, clearance; CL int, intrinsic
clearance; CMax, maximum plasma concentration; CNBr,
cyanogen bromide; CPT1, carnityl palmitate transfer protein
1; Cs2CO3, cesium carbonate; CuI, copper iodide; dba, 1,5-
diphenylpenta-1,4-dien-3-one; DIEA, diisopropylethylamine;
DIO, diet-induced obese; DMF, dimethylformamide;
(EtCO)2O, propionic anhydride; EtI, ethyl iodide; EtOH,
ethyl alcohol; %F, bioavailability; GTT, glucose tolerance test;
hACC, human acetyl-CoA carboxylase; HATU, 1-[bis-
(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]-
pyridinium 3-oxid hexafluorophosphate; HCl, hydrochloric
acid; HDL, high-density lipoprotein; HLM, human liver
microsomes; HTS, high-throughput screen; iv, intravenous
dosing; po, oral dosing; KOAc, potassium acetate; LC/MS/MS,
liquid chromatography/mass spec/mass spec; mACC, mouse
acetyl-CoA carboxylase; MeCN, acetonitrile; MeOH, methyl
alcohol; NaOtBu, sodium tert-butoxide; NH2OH, hydroxyl-
amine; nPrI, 1-iodopropane; PD, pharmacodynamic; PK,
pharmacokinetic; RLM, rat liver microsomes; RuPhos, 2-
dicyclohexyl(2′,6′-diisopropoxybiphenyl-2-yl)phosphine; SAR,
structure−activity relationship; THF, tetrahydrofuran; TFA,
trifluoroacetic acid; t1/2, half life; Vdss, volume of distribution;
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and synthetic procedures for all final
compounds and noncommercially available intermediates. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: 805-447-9430; E-mail: bourbeau@amgen.com.
10140
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141

Journal of Medicinal Chemistry
Article
Weight, Fat Mass, or Food Intake. Proc. Natl. Acad. Sci. U.S.A. 2010,
107, 7598−7603.
X-Phos, 2-dicyclohexylphosphino-2′,4′,6′-triisopropyl-1,1′-bi-
phenyl; ZDF rats, Zucker diabetic fatty rats
(19) Corbett, J. W. Review of Recent Acetyl-CoA Carboxylase
Inhibitor Patents: Mid-2007−2008. Expert Opin. Ther. Pat. 2009, 19,
943−956.
REFERENCES
■
(20) Bourbeau, M. P.; Allen, J. G.; Gu, W. Recent Advances in Acetyl-
CoA Carboxylase Inhibitors. Annu. Rep. Med. Chem. 2010, 45, 95−108.
(21) Bourbeau, M. P. ACC Inhibitors in Development. In New
Therapeutic Strategies for Type 2 Diabetes: Small Molecule Approaches;
Jones, R. M., Ed.; Royal Society of Chemistry, Cambridge, 2012; pp
464−500.
(22) Yamashita, T.; Kamata, M.; Endo, S.; Yamamoto, M.; Kakegawa,
K.; Watanabe, H.; Miwa, K.; Yamano, T.; Funata, M.; Sakamoto, J-i.;
Tani, A.; Mol, C. D.; Zou, H.; Dougan, D. R.; Sang, B.; Snell, G.;
Fukatsu, K. Design, Synthesis, and Structure-Activity Relationships of
Spirolactones Bearing 2-Ureidobenzothiophene as Acetyl-CoA Car-
boxylases Inhibitors. Bioorg. Med. Chem. Lett. 2011, 21, 6314−6318.
(23) Kamata, M.; Yamashita, T.; Kina, A.; Funata, M.; Mizukami, A.;
Sasaki, M.; Tani, A.; Funami, M.; Amano, N.; Fukatsu, K. Design,
Synthesis, and Structure-Activity Relationships of Novel Spiro-
Pyridines as Acetyl-CoA Carboxylase Inhibitors. Bioorg. Med. Chem.
Lett. 2012, 22, 3643−3647.
(24) Kamata, M.; Yamashita, T.; Kina, A.; Tawada, M.; Mizukami, A.;
Sasaki, M.; Tani, A.; Nakano, Y.; Watanabe, Y.; Furuyama, N.; Funami,
M.; Amano, N.; Fukatsu, K. Symmetrical Approach of Spiro-
Pyrazolidinediones as Acetyl-CoA Carboxylase Inhibitors. Bioorg.
Med. Chem. Lett. 2012, 22, 4769−4772.
(25) Freeman-Cook, K. D.; Amor, P.; Bader, S.; Buzon, L. M.;
Coffey, S. B.; Corbett, J. W.; Dirico, K. J.; Doran, S. D.; Elliott, R. L.;
Esler, W.; Guzman-Perez, A.; Henegar, K. E.; Houser, J. A.; Jones, C.
S.; Limberakis, C.; Loomis, K.; McPherson, K.; Murdande, S.; Nelson,
K. L.; Phillion, D.; Pierce, B. S.; Song, W.; Sugarman, E.; Tapley, S.;
Tu, M.; Zhao, Z. Maximizing Lipophilic Efficiency: The Use of Free-
Wilson Analysis in the Design of Inhibitors of Acetyl-CoA
Carboxylase. J. Med. Chem. 2012, 55, 935−942.
(1) Ford, E. S.; Giles, W. H.; Dietz, W. H. Prevalence of the
Metabolic Syndrome Among US Adults. JAMA 2002, 287, 356−359.
(2) Moller, D. E.; Kaufman, K. D. Metabolic Syndrome: A Clinical
and Molecular Perspective. Annu. Rev. Med. 2005, 56, 45−62.
(3) Grundy, S. M. Drug Therapy of the Metabolic Syndrome:
Minimizing the Emerging Crisis in Polypharmacy. Nat. Rev. Drug
Discovery 2006, 5, 295−309.
(4) Bjorntorp, P. Fatty Acids, Hyperinsulinema, and Insulin
Resistance: Which Comes First? Curr. Opin. Lipidol. 1994, 5, 166−174.
(5) McGarry, J. D. What if Minkowski Had Been Ageusic? An
Alternative Angle on Diabetes. Science 1992, 258, 766−770.
(6) Savage, D. B.; Petersen, K. F.; Shulman, G. I. Disordered Lipid
Metabolism and the Pathogenesis of Insulin Resistance. Physiol. Rev.
2007, 87, 507−520.
(7) Shulman, G. I. Cellular Mechanisms of Insulin Resistance. J. Clin.
Invest. 2000, 106, 171−176.
(8) Ploakis, S. E.; Guchhait, R. B.; Zwergel, E. E.; Lane, M. D.;
Cooper, T. G. Acetyl Coenzyme A Carboxylase System of Escherichia
coli: Studies on the Mechanisms of the Biotin Carboxylase- and
Carboxyltransferase-Catalyzed Reactions. J. Biol. Chem. 1974, 249,
6657−6667.
(9) McGarry, J. D.; Foster, D. W. Regulation of Hepatic Fatty Acid
Oxidation and Ketone Body Production. Annu. Rev. Biochem. 1980, 49,
395−420.
(10) Rasmussen, B. B.; Holmback, U. C.; Volpi, E.; Morio-Liondore,
B.; Paddon-Jones, D.; Wolfe, R. R. Malonyl Coenzyme A and the
Regulation of Functional Carnitine Palmitoyltransferase-1 Activity and
Fat Oxidation in Human Skeletal Muscle. J. Clin. Invest. 2002, 110,
1687−1693.
(11) Abu-Elheiga, L.; Matzuk, M. M.; Kordari, P.; Oh, W.; Shaikenov,
T.; Gu, Z.; Wakil, S. J. Mutant Mice Lacking Acetyl-CoA Carboxylase
1 Are Embryonically Lethal. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
12011−12016.
(12) Mao, J.; DeMayo, F. J.; Li, H.; Abu-Elheiga, L.; Gu, Z.;
Shaikenov, T.; Kordari, P.; Chirala, S. S.; Heird, W. C.; Wakil, S. J.
Liver-Specific Deletion of Acetyl-CoA Carboxylase 1 Reduces Hepatic
Triglyceride Accumulation without Affecting Glucose Homeostasis.
Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 8552−8557.
(13) Harada, N.; Oda, Z.; Hara, Y.; Fujinama, K.; Okawa, M.;
Ohbuchi, K.; Yonemoto, M.; Ikeda, Y.; Ohwaki, K.; Aragane, K.;
Tamai, Y.; Kununoki, J. Hepatic De Novo Lipogenesis Is Present in
Liver-Specific ACC1-Deficient Mice. Mol. Cell. Biol. 2007, 27, 1881−
1888.
(14) Abu-Elheiga, L.; Matzuk, M. M.; Abo-Hashema, K. A.; Wakil, S.
J. Continuous Fatty Acid Oxidation and Reduced Fat Storage in Mice
Lacking Acetyl-CoA Carboxylase 2. Science 2001, 291, 2613−2616.
(15) Abu-Elheiga, L.; Oh, W.; Kordari, P.; Wakil, S. J. Acetyl-CoA
Carboxylase 2 Mutant Mice are Protected Against Obesity and
Diabetes Induced by High-Fat/High-Carbohydrate Diets. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 10207−10212.
(16) Choi, C. S.; Savage, D. B.; Abu-Elheiga, L.; Zhu, Z. X.; Kim, S.;
Kulkarni, A.; Distefano, A.; Hwang, Y. J.; Reznick, R. M.; Codella, R.;
Zhang, D.; Cline, G. W.; Wakil, S. J.; Shulman, G. I. Continuous Fat
Oxidation in Acetyl-CoA Carboxylase 2 Knockout Mice Increases
Total Energy Expenditure, Reduces Fat Mass, and Improves Insulin
Sensitivity. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16480−16485.
(17) Hoehn, K. L.; Turner, N.; Swarbrick, M. M.; Wilks, D.; Preston,
E.; Phua, Y.; Joshi, H.; Furler, S. M.; Larance, M.; Hegarty, B. D.;
Leslie, S. J.; Pickford, R.; Hoy, A. J.; Kraegen, E. W.; James, D. E.;
Cooney, G. J. Acute or Chronic Upregulation of Mitochondrial Fatty
Acid Oxidation Has No Net Effect on Whole-Body Energy
Expenditure or Adiposity. Cell Metab. 2010, 11, 70−76.
(18) Olson, D. P.; Pulinilkunnil, T.; Cline, G. W.; Shulman, G. I.;
Lowell, B. B. Gene Knockout of Acc2 Has Little Effect on Body
(26) Biscoe, M. R.; Fors, B. P.; Buchwald, S. L. A New Class of Easily
Activated Palladium Precatalysts for Facile C−N Cross-Coupling
Reactions and the Low Temperature Oxidative Addition of Aryl
Chlorides. J. Am. Chem. Soc. 2008, 130, 6686−6687.
(27) Yao, F.; Svensjo, T.; Winkler, T.; Lu, M.; Eriksson, C.; Eriksson,
̈
E. Tetracycline Repressor, tetR, Rather than the tetR-Mammalian Cell
Transcription Factor Fusion Derivatives, Regulates Inducible Gene
Expression in Mammalian Cells. Hum. Gene Ther. 1998, 9, 1939−
1950.
(28) Lovering, F.; Bikker, J.; Humblet, C. Escape from Flatland:
Increasing Saturation as an Approach to Improving Clinical Success. J.
Med. Chem. 2009, 52, 6752−6756.
(29) Cruciani, G.; Carosati, E.; De Boeck, B.; Ethirajulu, K.; Mackie,
C.; Howe, T.; Vianello, R. MetaSite: Understanding Metabolism in
Human Cytochromes from the Perspective of the Chemist. J. Med.
Chem. 2005, 48, 6970−6979.
(30) Lewis, D. F. V.; Jacobs, M. N.; Dickins, M. Compound
Lipophilicity for Substrate Binding to Human P450s in Drug
Metabolism. Drug Discovery Today 2004, 9, 530−537.
(31) Glien, M.; Haschke, G.; Pfenninger, A.; Zoller, G.; Keil, S.;
Muller, M.; Herling, A. W.; Schmoll, D. Stimulation of Fat Oxidation,
̈
but No Sustained Reduction of Hepatic Lipids by Prolonged
Pharmacological Inhibition of Acetyl CoA Carboxylase. Horm.
Metab. Res. 2011, 43, 601−606.
NOTE ADDED AFTER ASAP PUBLICATION
■
After this paper was published ASAP on December 11, 2013,
corrections were made to Scheme 7 and the footnotes of
Schemes 4, 5, and 6. The corrected version was reposted
December 17, 2013.
10141
dx.doi.org/10.1021/jm401601s | J. Med. Chem. 2013, 56, 10132−10141