Design, synthesis, and antibacterial properties of dual-ligand inhibitors of acetyl-CoA carboxylase

Article abstract of DOI:10.1021/jm501082n

There is an urgent demand for the development of new antibiotics due to the increase in drug-resistant pathogenic bacteria. A novel target is the multifunctional enzyme acetyl-CoA carboxylase (ACC), which catalyzes the first committed step in fatty acid s

Full text of DOI:10.1021/jm501082n

Article
pubs.acs.org/jmc
Design, Synthesis, and Antibacterial Properties of Dual-Ligand
Inhibitors of Acetyl-CoA Carboxylase
†
‡
§
,†
Molly A. Silvers, Gregory T. Robertson, Carol M. Taylor, and Grover L. Waldrop*
†Division of Biochemistry and Molecular Biology, Louisiana State University, Baton Rouge, Louisiana 70803, United States
‡Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, Texas 75390, United States
§Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803, United States
*
S Supporting Information
ABSTRACT: There is an urgent demand for the develop-
ment of new antibiotics due to the increase in drug-resistant
pathogenic bacteria. A novel target is the multifunctional
enzyme acetyl-CoA carboxylase (ACC), which catalyzes the
first committed step in fatty acid synthesis and consists of two
enzymes: biotin carboxylase and carboxyltransferase. Cova-
lently attaching known inhibitors against these enzymes with
saturated hydrocarbon linkers of different lengths generated
dual-ligand inhibitors. Kinetic results revealed that the dual-
ligands inhibited the ACC complex in the nanomolar range.
Microbiology assays showed that the dual-ligand with a 15-
carbon linker did not exhibit any antibacterial activity, while
the dual-ligand with a 7-carbon linker displayed broad-
spectrum antibacterial activity as well as a decreased
susceptibility in the development of bacterial resistance. These results suggest that the properties of the linker are vital for
antibacterial activity and show how inhibiting two different enzymes with the same compound increases the overall potency while
also impeding the development of resistance.
INTRODUCTION
(CT), involves the transfer of the carboxyl group from
BCCP-biotin to acetyl-CoA to form malonyl-CoA. In
eukaryotes, the three proteins that comprise ACC form
■
(
In 2013, the U.S. Centers for Disease Control and Prevention
CDC) issued an alarming report on the dramatic rise in
6
1
domains on a single polypeptide chain, whereas in7
antibiotic-resistant bacteria. The report noted that in the U.S.,
over 2 million people per year are afflicted with bacterial
infections that are resistant to at least one antibiotic. In
addition, there are 23 000 deaths per year directly attributable
to antibiotic-resistant infections. One of the core recommen-
dations in the report for combating this public health crisis is
the development of new antibiotics. The current arsenal of
antibiotics is directed at only 30−40 molecular targets. Thus,
there is a pressing need to develop antibiotics against novel
targets. Fatty acid biosynthesis is an essential metabolic
pathway in both eukaryotes and prokaryotes, and since fatty
acids are required for membrane biogenesis in bacteria, fatty
acid synthesis is an attractive pathway to focus on for antibiotic
prokaryotes, they are encoded as three separate proteins.
The bacterial forms of BC and CT retain their activity when
isolated from the other components and can utilize free biotin
4,5
as a substrate. With respect to antibacterial development, the
different arrangement of the three protein components of ACC
between eukaryotes and prokaryotes provides potential for the
design of inhibitors that selectively target the bacterial
3
,4
system. Moreover, the three components of ACC are highly
conserved among bacterial species, increasing the likelihood of
7
,8
developing broad spectrum antibacterial agents.
Both the BC and the CT components of ACC have served as
targets for antibacterial agents. Miller et al. discovered that the
antibacterial properties of the pyridopyrimidine class of
2
−4
development.
9
molecules was due to the inhibition of BC. The pyridopyr-
The first committed and regulated reaction in fatty acid
biosynthesis is catalyzed by the multifunctional enzyme acetyl-
CoA carboxylase (ACC). The overall reaction catalyzed by
ACC is shown in Scheme 1. ACC requires the vitamin biotin,
which is covalently attached to the biotin carboxyl carrier
protein (BCCP). In the first half-reaction, catalyzed by biotin
carboxylase (BC), BCCP-biotin is carboxylated in an ATP-
dependent manner where bicarbonate is the source of CO2.
The second half-reaction, catalyzed by carboxyltransferase
imidines (e.g., compound 1, Figure 1) bound in the ATP
binding site of bacterial BC but did not inhibit human BC. The
pyridopyrimidines were not readily amenable to synthetic
modifications, so Mochalkin et al. used fragment and virtual
screening to identify a series of aminooxazole derivatives (e.g.,
5
Received: July 17, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Scheme 1. Reactions Catalyzed by Acetyl-CoA Carboxylase
carboxyltransferase is a heterotetramer with α- and β-subunits
encoded by accA and accD, respectively. The accA and accD
15
genes encoding the subunits of CT are not genetically linked.
This genetic arrangement is commonly found in both Gram-
negative and Gram-positive bacteria. Thus, one would predict
that nonessential missense mutations would have to arise in
three different genes in order for the bacteria to become
resistant to a molecule targeting both enzymes simultaneously.
While the use of dual-ligands for lowering the frequency of
Figure 1. Pyridopyrimidine 1 and aminooxazole 2, inhibitors of biotin
carboxylase.
16−19
resistance is well documented,
most examples of dual-
ligands as antibacterial agents involve inhibitors that target two
enzymes that either do not interact and/or are not related
metabolically. Thus, a dual-ligand inhibitor of ACC would
represent a departure from previous practice. In fact, a dual-
ligand inhibitor of ACC is a particularly attractive strategy for
inhibition of the enzyme in light of recent studies that have
demonstrated that all three protein components (BC, BCCP,
and CT) must form a complex in vivo for activity. On this
background, we report the first step toward dual-ligand
inhibition of ACC by covalently linking the BC inhibitory
motif of compound 2 to the core structure of compound 4, that
exhibits CT inhibition, and characterizing the inhibitory and
antibacterial properties.
compound 2 in Figure 1) that also inhibited BC by binding in
10
the ATP binding site. In common with the pyridopyrimidines,
the aminooxazole derivatives were selective inhibitors of the
10
bacterial, but not human, BC.
In contrast to BC, there is only one known class of molecules
that inhibits the CT component of ACC and also possesses
antibacterial activity. While the BC inhibitors are invariably of
synthetic origin, the CT inhibitors are natural products. The
antibacterial activity of andrimid and moiramide B (3 and 4,
20
11
respectively, Figure 2) displayed broad spectrum activity.
RESULTS AND DISCUSSION
■
Design Strategy. In order to generate dual-ligand
inhibitors incorporating these known inhibitory motifs, we
needed to identify loci in each inhibitor where we could attach
a metabolically stable linker without compromising binding and
biological activity. A suitable, stable functional group needed to
be placed on the dibenzylamide moiety of 2 (Figure 1). We
decided to incorporate a hydroxyl group at the para position of
one of the benzene rings with the potential for etherification.
This modification was not expected to affect binding to BC,
since it is the aminooxazole moiety that is responsible for
Figure 2. Andrimid (3) and moiramide B (4), inhibitors of
carboxyltransferase.
Their antibacterial activity had been known for at least 10 years
before Freiberg et al. discovered that the molecular target for
1
1,12
10
andrimid and moiramide B was the CT subunit of ACC.
interaction with the ATP binding site. In the case of
Structure−activity relationship (SAR) studies revealed that the
pyrrolidinedione headgroup was essential for enzyme inhibition
and antibacterial activity, while the fatty acid chain is not
moiramide B (compound 4, Figure 2), it has been shown that
the fatty acid side chain can be significantly modified with
retention of biological activity. We thereby elected to replace
this fatty acid chain with a lipophilic linker at the N-terminus of
the β-phenylalanine residue in the naturally occurring, bio-
logically active pseudopeptide. These design features gave rise
to dual-ligands 5 and 6 (Figure 3).
Finally, we needed to choose linkers to covalently connect
the aminooxazole and moiramide components of the dual-
ligand inhibitor. The most important criterion for the linker
segment was stability. Intracellular degradation of the linker
segment would lead to the false conclusion that dual-ligands
directed against ACC were not feasible as antibacterial agents.
While a variety of chemical constructs could serve as the linker
segment, a saturated hydrocarbon was selected because it was
least susceptible to intracellular degradation and would afford
the best chance for determining the feasibility of the dual-ligand
hypothesis for ACC.
13
1
3
required. Compound 4 was found to have an inhibition
constant of 5 nM and exhibited competitive inhibition versus
1
2,14
malonyl-CoA.
Compound 4 inhibited CT from both Gram-
negative and Gram-positive organisms, which is consistent with
its broad spectrum activity. In common with the BC inhibitors,
12
4
did not inhibit human ACC.
Despite the fact that both BC and CT have been validated as
unique targets for antibacterial development through the
discovery of selective antibacterial inhibitors, both targets suffer
from high potential for spontaneous resistance through
individual target-based mutations. One approach to decrease
the likelihood of developing resistance is to make a single
molecule that incorporates inhibitory motifs for both the BC
and CT enzymatic activities. The potential lower frequency of
resistance against such a dual-ligand inhibitor is explained by
considering the genetic arrangement of ACC. Biotin carbox-
ylase is a homodimer encoded by the accC gene, while
We chose a saturated aliphatic compound with a leaving
group at one end, to form an ether with the phenol at the
B
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 3. Synthesis of Amine 25 and Compound 4
Figure 3. Dual-ligand inhibitors containing aminooxazole and
moiramide pharmacophores with saturated hydrocarbon linkers with
1
5-carbons (compound 5) and 7-carbons (compound 6).
“aminooxazole end”, and a carboxylic acid at the other end to
forge an amide with the N-terminus of the moiramide fragment.
We arbitrarily started with a 15-carbon linker since juniperic
acid (compound 26, see Scheme 4, below) is commercially
available. We later also utilized a 7-carbon linker (from 8-
hydroxyoctanoic acid, see Scheme 4) to more accurately mimic
the fatty acid component of moiramide B. This design strategy
generates dual-ligand inhibitors that incorporate the necessary
pharmacophores for targeting both BC and CT while
covalently attaching them using stable, saturated hydrocarbon
linkers of various lengths (Figure 3).
Chemistry. We generated the requisite unsymmetrical
dibenzylamine via reductive amination of benzaldehyde with
21
p-aminophenol to give compound 10 (Scheme 2). Coupling
of carboxylate 12 with secondary amine 10 afforded compound
a
Reagents and conditions: (i) AcCl, 60 °C, 97%; (ii) 1. Compound
6, THF, Δ; 2. Ac O, Δ, 83%; (iii) Compound 19, CDI, triethylamine,
1
1
3 (Scheme 2), the “aminooxazole” component of the dual-
2
CH Cl , 89%; (iv) 1. N-Boc-L-valine, CDI, dry THF; 2. LiHMDS, −78
2
2
ligand inhibitor, with a handle for attachment to other
1
0
°C, dry THF, 42%; (v) 1. H
, 5% Pd/C (20% w/w), dry methanol; 2.
2
species. However, compound 2 was also synthesized and
used in biological studies as the “aminooxazole only”
representative in order to eliminate any side reactions that
might have been encountered with the phenol functionality of
2
-bromoacetophenone, triethylamine, CH CN, 64%; (vi)
3
TFA:CH Cl , 0 °C, 89−98%; (vii) (S)-N-Boc-3-amino-3-phenyl-
propanoic acid, HATU, Pr NEt, dry DMF, 0 °C to rt, 45%; (viii)
Sorbic acid, TBTU, Pr NEt, dry DMF, 32%.
2
2
i
2
i
2
1
3.
Compounds 3 and 4 have been synthesized previously,
although most reports are in communication format without
1
3,22−24
27
experimental details.
These accounts all alluded to the
the benzyl protected succinimide 17. Indeed, Pohlmann et al.
fragility of the β-ketoamide functionality; the pK for enolate
had prepared a number of N-alkylated derivatives of moiramide
for SAR studies but they did not prepare the benzyl derivative,
and it was not clear why they and others used benzyloxy
protection for the succinimide nitrogen, which requires two
a
25
formation of andrimid is 6.8. We prepared anhydride 15
Scheme 3) from commercially available (4S)-methyl-succinic
(
26
acid according to Midgley and Thomas and converted this to
a
Scheme 2. Synthesis of Aminooxazole 2 and Aminooxazole 13 with a para-Hydroxyl Group
a
Reagents and conditions: (i) triethylamine, dry methanol, 3 Å molecular sieves, 65%; (ii) NaBH , dry methanol, 0 °C to rt, 84%; (iii) 1. 2 M
4
NaOH, 60 °C; 2. HCl, 98%; (iv) HATU, triethylamine, dry DMF, 31−35%.
C
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
steps for removal. We expected to cleave the benzylamine in a
single hydrogenolytic step.
Scheme 4. Synthesis of Modified Saturated Hydrocarbon
Linkers 28 and 31
a
We now faced the challenging acylation of the kinetic enolate
of 17 with an acyl cation equivalent derived from N-tert-
butoxycarbonyl-L-valine (Boc-Val-OH). The discussion by
Davies and Dixon informed us that the enolate was unstable
and must be generated in the presence of an appropriate
2
4
electrophile. While they employed an N-carboxyanhydride,
a
1
3,23
Reagents and conditions: (i) allyl alcohol, H SO , heat, 59−61%; (ii)
others invoked an acyl imidazole.
We found the acyl
2
4
polymer-bound triphenylphosphine, I , imidazole, dry CH Cl , 78−
8
2
2
2
imidazole was more convenient to generate and more stable
than the N-carboxyanhydride. The acyl imidazole generated
from Boc-Val-OH was added to a solution of compound 17.
The solution was added slowly to a solution of LiHMDS in
THF at −78 °C. While TLC analysis of the reaction looked
promising, poor yields of compound 18 were obtained by flash
chromatography. Two-dimensional TLC confirmed that 18 was
unstable to silica gel and was thus isolated by reversed-phase
HPLC in low yield. Furthermore, the product was not
particularly stable, and standard hydrogenolysis procedures
for removing the benzyl protecting group were unsuccessful.
The N-benzyl group was replaced with the N-benzyloxy
protecting group, ultimately leading to compound 21 (Scheme
3%.
derivative was removed by hydrolysis to generate the free
carboxylic acids 34 and 35, by a procedure analogous to one of
Zhang et al. The acids were each coupled to the amine in
compound 25 using standard coupling reagents to generate
compounds 5 and 6 that were purified by RP-HPLC. Using
an Agilent Zorbax Stable Bond-C18 column, compound 6 (7C
linker) eluted as expected, but compound 5 (15C linker) was
retained, most likely due to the increased hydrophobicity of 5.
Fortunately, the use of an Agilent Zorbax Extend-C18 column
with double end-capping allowed for the elution and
purification of 5 (Scheme 5).
31
13
3) that was remarkably stable to silica gel. An earlier paper by
Robin et al., outlining a synthesis of thalidomide, also utilized
benzyloxy protection of the glutarimide ring, where the benzyl
Inhibition of Acetyl-CoA Carboxylase by the Dual-
Ligands. The targets of the antibacterial agents 2 and 4 are BC
and CT, respectively. The question now is how would a
molecule that incorporates the features of both compounds 2
and 4 inhibit the multiprotein complex comprising all three
components of ACC? Recent studies have shown that the three
components of ACCBC, BCCP, and CTform a multi-
28
group had been problematic. They suggested that resonance
donation of electron density from the oxygen reduces the
electrophilicity of the CO groups and the acidity of the
neighboring alpha protons. The acylation of 20 and purification
of 21 was now realized in a better yield of 42%, and the product
was stable, albeit as an equilibrium mixture of predominantly
the trans diastereomer and <10% of the diastereomer with cis-
relative stereochemistry on the succinimide ring that arises via
tautomerization to the enol form, evident in <6%. Earlier
reports did not include NMR spectra, although Davies and
Dixon reported an 87:9:4 (trans:enol:cis) mixture for
compound 21. The benzylic C−O bond in 21 was cleaved by
hydrogenolysis to afford the hydroxylamine that was treated
with 2-bromoacetophenone to generate the phenacyl ester in
20
protein complex both in vitro and in vivo. In order to
compare the dual-ligand inhibitors to the two parent
compounds, the inhibition of ACC by compounds 2 and 4
was also studied because the original characterization of these
inhibitors utilized only the isolated subunits of BC or CT, not
the ACC complex.
The inhibition of ACC by compound 2 was determined by
varying ATP at increasing fixed concentrations of inhibitor,
where acetyl-CoA was held constant at a subsaturating level.
Compound 2 exhibited noncompetitive inhibition versus ATP
with a K of 0.4 ± 0.1 μM and a K of 0.9 ± 0.2 μM (Figure
24
situ which ultimately led to succinimide 22. The deprotected
succinimide 22 existed as an 82:4:14 (trans:enol:cis) mixture;
Davies and Dixon did not report a ratio of isomers for this
compound. Removal of the Boc group at the N-terminus of
compound 22 and elongation with commercially available Boc-
β-Phe-OH gave compound 24. The addition of the Boc-β-Phe
moiety resulted in a single diastereomer, suggesting a stronger
thermodynamic preference for the trans isomer with the larger
acyl side chain. Compound 24 in turn underwent Boc
deprotection to afford the free amine 25 that was coupled to
sorbic acid to generate compound 4 (Scheme 3).
is
ii
4A). In fact, even at 9 times the K value for ATP (56 μM),
m
compound 2 was still able to inhibit the activity of the enzyme.
This pattern was unexpected based on the competitive
inhibition versus ATP that was observed with the isolated
subunit of BC (Supporting Information, Figure S1). However,
the K of 0.8 ± 0.2 μM for isolated BC was close to the K
value observed for ACC. The reason for the noncompetitive
pattern in ACC is not clear at this time and awaits the
determination of the three-dimensional structure of ACC
bound to 2. Nonetheless, the noncompetitive inhibition pattern
observed for ACC has implications for targeting the BC subunit
of the enzyme for pharmaceutical purposes. Namely, the
noncompetitive inhibition pattern means that inhibitor binding
is not precluded by saturating levels of ATP. The K for ATP
in ACC is reported to be 1−2 μM; however, we consistently
measured a K value of 6 μM. Nevertheless, the intracellular
level of ATP in log phase Escherichia coli is 9 mM, therefore,
ACC is very likely saturated with the substrate ATP in vivo.
Thus, at least for the BC inhibitor (compound 2), saturation
with ATP will not be problematic. These results show that it is
imperative that any potential antibacterial agents targeting
either the BC or CT subunits of ACC must be characterized
is
is
Preparation of the 15-carbon linker followed a procedure that
Overman and co-workers had previously reported for the
29
conversion of acid 26 to allyl ester 27 (Scheme 4). The
primary alcohol was converted to alkyl iodide 28 by analogy to
30
a procedure of Jobron et al. We also prepared a significantly
shorter linker, viz. 31, as shown in Scheme 4.
m
20
The convergent assembly of the three fragments to generate
the dual-ligand inhibitors is depicted in Scheme 5. The phenol
of compound 13 was activated as the corresponding cesium salt
using CsOH. This accentuated nucleophile then participated in
m
3
2
an S 2 reaction with the iodide of compounds 28 and 31 to
N
generate an ether linkage, viz. aminooxazole−linker conjugates
3
2 and 33, respectively. The allyl protecting group of each
D
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 5. Synthesis of Compounds 5 and 6
a
Reagents and conditions: (i) 1. CsOH, dry DMF; 2. Compounds 28 or 31, dry THF, Δ, 20−21%; (ii) 2 M NaOH, MeOH:THF, 71−89%; (iii)
i
Compound 34 or 35, HATU, Pr NEt, dry DMF, 0 °C to rt, 17−19%.
2
Figure 4. Inhibition of acetyl-CoA carboxylase. (A) Inhibition by 2 at 0 (●), 0.3 (■), and 0.7 μM (▲). (B) Inhibition by 5 at 0 (●), 6 (■), and 10
nM (▲). (C) Inhibition by 6 at 0 (●), 6 (■), and 10 nM (▲). For all three assays, the substrate ATP was varied from 0.4 to 9.0 times the K value
m
−
for ATP. The other substrates, HCO3 and acetyl-CoA, were held constant at 15 mM and 200 μM, respectively. The points represent the observed
velocities, and the lines represent the best fit of the data to eq 2. (D) Inhibition by 4 at 0 (●), 3 (■), and 6 nM (▲). (E) Inhibition by 5 at 0 (●), 6
(
■), and 10 nM (▲). (F) Inhibition by 6 at 0 (●), 6 (■), and 10 nM (▲). For all three assays, the substrate acetyl-CoA was varied while the other
−
substrates, HCO3 and ATP, were held constant at 15 mM and 5 μM, respectively. The points represent the observed velocities, and the lines
represent the best fit of the data to eq 1.
E
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Antibacterial Activities of Synthesized Compounds against Strains of Bacteria
b
MIC (μg/mL)
a
strain (genotype)
rifampin
gentamicin
chloramphenicol
kanamycin
2
4
5
6
P. aeruginosa (wild-type)
31
31
2
0.5
>64
2
128
64
2
>64
>64
>64
>64
16
>64
8
>64
>64
>64
>64
>64
ND
>64
>64
ND
>64
>64
>64
>64
>64
>64
>64
8
c
P. aeruginosa (ΔRND)
E. coli (wild-type)
E. coli (lpxC101)
E. coli (tolC)
31
0.1
8
>64
4
0.06
8
0.12
0.24
ND
≤0.03
0.5
4
1
1
2
0.5
d
E. coli (tolC, moir-R)
ND
0.5
ND
1
0.5
0.5
4
8
4
16
4
d
E. coli (lpxC101, tolC)
16
0.5
1
S. aureus (wild-type)
S. aureus (moir-R)
S. aureus (norAup)
≤0.03
ND
≤0.03
≤0.03
0.5
8
>64
>64
>64
>64
>64
4
ND
0.5
ND
8
2
8
>64
4
4
4
S. epidermidis (wild-type)
E. faecalis (wild-type)
8
8
>128
31
0.24
8
1
8
8
4
a
The following designate targeted deletion mutants of efflux pumps or subunits of efflux pumps: mexAB, mexCD, mexXY, mexHI, opmH, tolC. The
up
lpxC101 allele enhances permeability of the E. coli outer membrane. The norA genotype results from a spontaneous mutation in S. aureus resulting
in a gain of function of the norA efflux pump. The designation moir-R indicates a spontaneous moiramide resistance selected with moiramide B
containing agar at 4−8 times the MIC. Wild-type P. aeruginosa is PA01; wild-type E. coli is D21; wild-type S. aureus is ATCC #29213; wild-type S.
b
c
epidermidis is ATCC #35984; wild-type E. faecalis is ATCC #29212. Determined by the broth microdilution method. ΔRND = mutation of mexAB,
d
mexCD, mexXY, mexHI, opmH. These strains show a slight growth defect on agar medium. ND = Not determined
with ACC in toto and not just the isolated subunits, because
the multiprotein complex is the active form of the enzyme in
vivo.
The inhibition of ACC by moiramide B was determined by
varying acetyl-CoA at increasing fixed concentrations of
inhibitor, while ATP was held constant at a subsaturating
level. Thus, compound 4 exhibited competitive inhibition
ATP in ACC. Thus, the inhibition patterns for the compounds
5 and 6 are perfectly consistent with the inhibition patterns
observed for the two “parent” compounds, 2 and 4. The
molecular basis for the inhibition patterns will ultimately be
revealed by the determination of the three-dimensional
structure of ACC bound to either of the dual-ligands. Lastly,
given that neither pharmacophore, compounds 2 nor 4, inhibits
1
0,12
versus acetyl-CoA with a K of 7.9 ± 1.1 nM (Figure 4D).
human ACC,
it is unlikely that the dual-ligands would
is
These results are similar to the competitive inhibition pattern of
inhibit the human enzyme.
4
versus malonyl-CoA and the K value of 5 nM for the isolated
Antibacterial Activity of Dual-Ligands. Table 1 lists the
minimum inhibitory concentration (MIC) values for the
“parent” compounds 2 and 4, as well as compounds 5 and 6,
against different bacterial strains. Compound 2 exhibited
antibacterial activity against efflux-pump deficient E. coli and,
in fact, registered the same MIC value of 16 μg/mL as reported
is
1
2
subunit of CT.
For the dual-ligands, inhibition patterns using biochemical
assays were determined versus both ATP and acetyl-CoA. The
inhibition of ACC by compound 5 (15C linker) versus ATP at
increasing fixed concentrations of inhibitor, with acetyl-CoA
held constant at a subsaturating level, exhibited noncompetitive
inhibition with a K of 28.7 ± 12.9 nM and a K of 33.7 ± 7.2
nM (Figure 4B). The inhibition of ACC by compound 5 versus
acetyl-CoA at increasing fixed concentrations of inhibitor, with
ATP held constant at a subsaturating level, exhibited
10
by Mochalkin et al. Also consistent with the findings of
Mochalkin et al., compound 2 was not active against any Gram-
is
ii
10
positive bacteria. Moreover, 2 was effective against an E. coli
tolC derivative that was selected for spontaneous resistance to 4
suggesting that dual-ligands such as 6 would retain activity
against strains bearing pre-existing moiramide resistance.
Compound 4 exhibited antibacterial activity against all strains
tested except the wild-type strains of Pseudomonas aeruginosa
and E. coli. The lack of activity against these strains can be
attributed to efflux-based intrinsic resistance, since efflux-pump-
deficient strains of these organisms were susceptible to 4 (Table
1). Compound 4 also exhibited activity against spontaneous
moiramide-resistant isolates of both E. coli tolC and Staph-
ylococcus aureus, with an observed 8-fold decrease in potency in
each case. The elevated MICs against 4 suggest that some
stable drug resistance was observed, most likely through a
target-based spontaneous mutation. The results reported herein
for 4 are consistent with previously reported antibacterial
competitive inhibition with a K of 5.2 ± 0.7 nM (Figure 4E).
is
The inhibition of ACC by compound 6 (7C linker) versus
ATP at increasing fixed concentrations of inhibitor, with acetyl-
CoA held constant at a subsaturating level, exhibited non-
competitive inhibition with a K of 16.9 ± 6.8 nM and a K of
is
ii
9.6 ± 1.0 nM (Figure 4C). The inhibition of ACC by dual-
ligand 6 versus acetyl-CoA at increasing fixed concentrations of
inhibitor, with ATP held constant at a subsaturating level,
exhibited competitive inhibition with a K of 5.5 ± 0.6 nM
is
(Figure 4F).
The noncompetitive inhibition pattern versus ATP and the
competitive pattern versus acetyl-CoA observed for both 5 and
could be due to the linker not being long enough to allow
6
1
2,13,33
both parts of the dual-ligands to bind to the enzyme
simultaneously. As a consequence, compounds 5 and 6 could
only bind in the moiramide B binding site, by virtue of its
higher affinity, compared to compound 2. While this is a
plausible explanation for the noncompetitive inhibition pattern
versus ATP, it is inconsistent with the fact that the “parent”
compound 2 also exhibited noncompetitive inhibition versus
analyses.
Of the dual-ligands tested, compound 6 bearing the shorter
7C linker was found to exhibit the best overall antibacterial
activity in MIC assays (Table 1). Antibacterial activity was
observed against an efflux compromised E. coli strain lacking
the tolC outer membrane channel and wild-type S. aureus. In
fact, compound 6 was effective against all but one of the Gram-
F
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
positive strains investigated. Of particular note and, in contrast
to 4, MIC values for 6 were unchanged at 4 μg/mL when tested
against wild-type S. aureus and an otherwise isogenic strain
compound 5 did inhibit bacterial growth (Figure 5).
Concentrations of 10, 50, and 100 μM of compound 5
inhibited growth of E. coli when inoculated with at least 5 μg/
mL of PMBN. There was no significant inhibition of growth
with only 1 μM of 5, even with the addition of PMBN. Control
experiments showed no growth inhibition when E. coli was
inoculated with only PMBN at the highest concentration (10
μg/mL) used in the assay (Supporting Information, Table S1).
These results strongly suggest that the lack of antibacterial
activity of compound 5, even at high concentrations, was due to
an inability to diffuse across the bacterial membrane. Thus, the
antibacterial activity of 5 in the presence of PMBN can be
attributed to the ability of 5 to enter the cell and inhibit ACC.
This result implies synergy between PMBN and compound 5
which can be beneficial when developing combination therapies
34
bearing a gain-of-function mutation effecting norA-based
efflux activity (compare MICs for compounds 4 and 6 versus
up
wild-type and norA S. aureus in Table 1). From this, we
conclude that compound 6 is no longer a substrate for NorA-
based efflux in S. aureus; whereas, the parent compound 4 is
appreciably impacted by a mutational gain-of-function of the
NorA efflux pump. Such differential recognition of closely
35
related antimicrobial substrates has been reported previously.
Compound 6 was also still effective against moiramide-resistant
E. coli tolC but ineffective against moiramide-resistant S. aureus.
This result is wholly expected given that the secondary
pharmacophore (the aminooxazole derivative) is not effective
against Gram-positive bacteria due to a threonine residue at
position 437 of BC (which is an isoleucine in Gram-negative
bacteria) causing a loss in necessary hydrophobic contacts
3
8,39
targeting Gram-negative pathogens.
Single-Step Resistance Studies. One of the potential
advantages of a dual-ligand antibacterial agent versus a molecule
directed against a single target is a lower likelihood of
developing resistance. To this end, the spontaneous resistance
frequencies were determined for rifampin, 2, 4, and 6 against
the E. coli tolC mutant (Table 2). Compared to the control
9
between the inhibitor and the binding pocket. Lastly, while the
antibacterial activity of 6 is slightly reduced (i.e., 4−8-fold)
relative to 4 in most strains, there was an increase in potency
for 6 with an MIC of 4 μg/mL against Enterococcus faecalis,
whereas 4 showed an MIC of 8 μg/mL. The absence of any
apparent antibacterial activity of compound 5 is unexpected as
Table 2. Results of Single-Step Resistance Selection Studies
with the E. coli tolC Strain
5
did, in fact inhibit the holoenzyme ACC. We speculated that
compound 5 might be impaired in its ability to diffuse through
the bacterial membrane due to the presence of the longer,
hydrophobic linker.
a
b
MIC
MPC
resistanced
frequency
7 × 10−9
c
compound
(mg/L)
(mg/L)
MSW
rifampin
8
16
0.5
8
>64
>128
16
>8
>8
32
8
Effect of PMBN on the Antibacterial Activity of
Compound 5. Polymyxin B nonapeptide (PMBN) is a
cationic cyclic peptide that is a non-bactericidal version of the
antibiotic polymyxin B (PMB). PMBN binds to the bacterial
lipopolysaccharide found in the outer membrane of Gram-
negative bacteria which leads to increased susceptibility to the
−9
2
4
6
1.75 × 10
<2.5 × 10
<2.5 × 10
−10
−10
64
a
b
Determined from broth microdilution in MHII broth. MPC is the
drug dose at which resistant mutants (using 4 × 10 CFU) were no
longer obtained. MSW is the ratio of the MPC to the MIC.
9
c
3
6,37
d
passage of hydrophobic antibiotics.
Therefore, the effect of
Resistance frequency is the number of resistant colonies observed
PMBN on the antibacterial activity of compound 5 was
examined to determine whether or not an inability to diffuse
across the bacterial membrane was the reason for the lack of
observed antibacterial activity.
over the total number of CFU plated at 8x MIC
The antibacterial activity of 5 when co-incubated in E. coli
with PMBN is shown in Figure 5. As expected, there was no
antibacterial activity observed for 5 when inoculated in E. coli
without PMBN. However, with increasing amounts of PMBN,
rifampin, compound 2 had a lower frequency of resistance and,
in general, was found to be a poor inhibitor in these assays. In
contrast, the mutant prevention concentrations (MPC) for
compounds 4 and 6 were 16 mg/L and 64 mg/L, respectively.
This allowed for the calculation of mutant selection windows
(
MSWs) for these compounds where the MSW is the
40
MPC:MIC ratio, with an ideal drug having an MSW of 1.
The MSW values for 4 and 6 were 32 and 8, respectively. Thus,
even though 6 was somewhat lower in potency than the
moiramide parent compound, the smaller MSW suggests it has
more restriction for selection of resistance. This is a potentially
exciting observation as compounds with a smaller MSW are far
less likely to select for spontaneous resistance in vivo when
41
dosed appropriately above their respective MPC. Lastly, when
dosed at 8 times the MIC value, compounds 4 and 6 both
exhibited an observed resistance frequency of <2.5 × 10−
10
.
This number was too low to determine any significant
difference between the two compounds. Therefore, it could
not be assessed if 6 has a lower frequency of resistance when
compared to the two “parent” compounds. Nonetheless, when
compared to rifampin and 2, compound 6 clearly has a much
lower potential for spontaneous resistance development.
Figure 5. Effect of polymyxin B nonapeptide (PMBN) on the
antibacterial activity of 5. Increasing amounts of PMBN (0, 0.5, 1, 5,
1
5
0 μg/mL) were added to different fixed concentrations of 5 (1, 10,
0, 100 μM). The solutions were inoculated with E. coli and incubated
at 37 °C. After a 4 h incubation, the absorbance at 600 nm was
measured.
G
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
CONCLUSIONS
did retain appreciable antibacterial activity in vitro. Never-
theless, in single-step resistance studies in vitro, compound 6
exhibits low resistance development potential and its MSW was
■
The studies in this report provide “proof-of-concept” that a
dual-ligand inhibitor of the multifunctional enzyme acetyl-CoA
carboxylase can act as an antibacterial agent. First, the synthesis
of the dual-ligand inhibitor led to significant improvements in
the synthetic protocol for compound 4. Second, incorporating
two known inhibitors for the two enzymes comprising ACC,
biotin carboxylase and carboxyltransferase, covalently attached
via a linker into a single chemical entity yielded a potent (nM
inhibition constant) inhibitor of the enzyme. However, based
on our results, we speculate that the more hydrophobic the
linker is, the less likely it is to diffuse across the bacterial
membrane.
The original impetus for using a dual-ligand inhibitor of ACC
as an antibacterial agent as opposed to using each
pharmacophore individually (i.e., combination therapy) was
that previous kinetic studies showed communication between
the two active sites of the enzyme which implies a physical
interaction between the proteins. The design of these dual-
ligand molecules was such that they act as stable, dual-
pharmacophore agents and not as prodrugs. This strategy
should ensure that the pharmacokinetics, pharmacodynamics,
and tissue distribution of the composite pharmacophores are
matched which may impart certain advantages over combina-
tion therapies involving the two parent agents alone. For
example, while the potency of 4 is impacted by efflux-based
exclusion in strains bearing a gain-of-function mutation in the
gene encoding the S. aureus NorA efflux pump, compound 6
does not appear to be impacted by this activity suggesting it is
no longer a substrate for NorA-based efflux. Further studies
comparing the activity of this and other optimized moiramide B
and aminooxazole dual-ligands relative to cocktails of the two
parent agents will be the subject of future studies.
4
times smaller than that of the parent compound 4. This
strongly suggests that the secondary activity of 6 is
antimicrobial, and, as is the case for other dual-targeting classes
of antimicrobials (e.g., fluoroquinolones), may be a better
indicator of the ability of this class of compounds to prevent
spontaneous resistance development, rather than its overall
41
potency.
Nevertheless, while the pharmacophore involving compound
was not as potent as compound 4, many novel aminooxazole
2
derivatives were recently reported as part of a recent virtual
42
screening study which identified several million aminooxazole
derivatives that could potentially be linked to compound 4 to
generate a wide array of dual-ligand hybrids. These new dual-
ligands could, in turn, be screened for potent broad spectrum
antibacterial activity using the tools and techniques described
herein. Future studies will seek to test this proposal directly and
will undoubtedly provide new and important information about
the activity of dual-ligand inhibitors against this novel target.
2
0
EXPERIMENTAL SECTION
■
Materials and Methods. All chemicals and reagents were
purchased from Sigma-Aldrich, Fisher, Acros, NovaBiochem, Matrix,
Combi-Blocks, or Fluka and used without further purification.
Diisopropylethylamine and triethylamine were dried and distilled
from CaH and stored over KOH pellets. Dry methanol was distilled
2
from Mg turnings and stored over 3 Å molecular sieves. All reactions
were performed under a dry nitrogen atmosphere unless otherwise
noted. Flash chromatography was performed using 230−400 mesh
silica gel (40−63 μm) from Sigma-Aldrich. RP-HPLC was performed
with a Waters 616 pump, Water 2707 Autosampler, and 996
Photodiode Assay Detector which are controlled by Waters Empower
2
software. The linear gradient resulted from mixing eluents A (0.1%
The observation that compound 6 exhibited a frequency of
resistance that was less than 10 , whereas the frequency of
TFA in water) and B (0.1% TFA in acetonitrile). Method A: The
separation was performed on an Agilent Zorbax 300 Extend-C18 (5
μm, 150 × 4.6 mm) with Agilent guard column Zorbax 300 SB-C18 (5
μm, 12.5 × 4.6 mm). Method B: The separation was performed on an
Agilent Zorbax 300 SB-C18 (5 μm, 250 × 4.6 mm) with Agilent guard
column Zorbax 300 SB-C18 (5 μm, 12.5 × 4.6 mm). The flow rate was
−10
−6
−9 19
resistance of single targets is between 10 to 10 , bears out
the dual-ligand approach described here. Most importantly,
compound 6 had a smaller MSW than 4 indicating that a lower
dose is needed to prevent the development of resistance
bacteria. The smaller MSW for 6 vs 4 is also consistent with a
role for the aminooxazole pharmacophore in the dual-ligand
since narrower MSW values are expected for two intracellular
targets compared to one target. All in all, the results suggest
that a dual-ligand approach demonstrates the potential for
antibacterial agents directed against ACC.
An unexpected but potentially far-reaching finding was that
compound 2 exhibited noncompetitive inhibition of BC with
respect to ATP in ACC, but competitive inhibition in the
isolated BC subunit. This indicates that inhibitors of BC
discovered using the isolated subunit may react differently with
ACC. Third, the finding that compound 5 had little to no
antibacterial activity, while 6 exhibited antibacterial activity
against Gram-negative and Gram-positive organisms indicates
that the choice of the linking component is critical for
determining whether dual-ligand inhibitors which show activity
in biochemical assays will also show antibacterial activity.
The choice of compound 2 as the BC inhibitor was a
1
.0 mL/min, and detected wavelength was 215 nm. Thin layer
chromatography (TLC) was performed on aluminum-backed 60 F254
silica plates from EMD Chemicals, Inc. Optical rotations were
recorded on a Jasco P-2000 digital polarimeter. Circular dichroism
41
(
CD) spectrum was recorded on a Jasco J-815 circular dichroism
1
13
spectrometer. H and C NMR spectra were recorded at room
temperature on a Bruker AV-400 or AV-500 spectrometer. All NMR
experiments were performed in deuterated solvents, and the chemical
shifts are reported in standard δ notation as parts per million, using
tetramethylsilane (TMS), CDCl , DMSO-d , or acetone-d as an
internal standard for H and C NMR, with coupling constants (J)
reported in Hertz (Hz). High resolution mass spectrometry (HRMS)
was carried out using an Agilent 6210 electrospray ionization-time-of-
flight (ESI-TOF) mass spectrometer. Purity of final compounds was
determined by HPLC analysis as being ≥95%.
(E)-4-[(Benzylideneamino)methyl]phenol (9). Benzaldehyde 7
(413 μL, 431 mg, 4.06 mmol, 1.00 equiv) was added to a suspension of
3 Å molecular sieves (150 mg/100 mg amine), triethylamine (736 μL,
3
6
6
1
13
5
34 mg, 5.28 mmol, 1.30 equiv), and 4-hydroxybenzylamine 8 (500
mg, 4.06 mmol, 1.00 equiv) in dry methanol (1.0 mL/100 mg amine)
at room temperature under N . The mixture was stirred at room
2
compromise between its relatively low affinity (K = 400 nM)
i
temperature for 18 h. The molecular sieves were removed by filtration
through a pad of Celite, washing well with dry methanol, and the
filtrate was concentrated in vacuo to give 9 as a light brown solid (556
and its synthetic tractability. Nonetheless, several lines of
evidence suggest that it played a role in the antibacterial activity
of the dual-ligands. While the pharmacophore involving parent
compound 2 was not as potent as the parent compound 4, it
43
1
mg, 65%). H NMR (DMSO-d , 400 MHz): δ 4.64 (s, 2H), 6.72 (d,
6
J = 8.4 Hz, 2H), 7.11 (d, J = 8.4 Hz, 2H), 7.41−7.49 (m, 3H), 7.70−
H
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
7
.81 (m, 2H), 8.44 (s, 1H), 9.33 (br s, 1H). 13C NMR (DMSO-d , 100
(m, 1H), 2.99−3.17 (m, 2H). C NMR (CDCl , 100 MHz): δ 16.1,
13
6
3
MHz): δ 64.1, 115.6, 128.4, 129.1, 129.6, 130.1, 131.1, 136.6, 156.7,
1
obsd, 212.1070.
35.6, 36.0, 169.9, 174.3. HRMS (ESI, -ve, TOF), calcd for C H O
(M-H) : 113.0244; obsd, 113.0244.
(4S)-Methyl-N-O-benzylsuccinimide (20).
5
5
3
+
+
61.5. HRMS (ESI-TOF): calcd for C H NO (M+H) , 212.1070;
14
14
1
3
Triethylamine
4
-[(Benzylamino)methyl]phenol (10). Sodium borohydride
(1.08 mL, 787 mg, 7.78 mmol, 1.05 equiv) was added to a solution
(
154 mg, 4.06 mmol, 1.80 equiv) was added in portions to a solution
of 15 (845 mg, 7.41 mmol, 1.00 equiv) in CH Cl (5 mL) at 0 °C
2
2
of 9 (556 mg, 2.63 mmol, 1.00 equiv) in dry methanol (5 mL) at 0 °C.
The mixture was stirred for 18 h under N while allowing to warm to
followed by addition of O-benzylhydroxylamine hydrochloride 19
(1.24 g, 7.78 mmol, 1.05 equiv). The mixture was warmed to room
2
room temperature. The mixture was concentrated and then partitioned
between CH Cl (10 mL) and water (10 mL). The aqueous layer was
temperature overnight under N , and then 1,1′-carbonyldiimidazole
2
(1.32 g, 8.15 mmol, 1.10 equiv) was added in portions. The mixture
was stirred at room temperature for 1.5 h and then heated at reflux for
30 min. The mixture was cooled to room temperature, and the organic
layer was washed with 10% hydrochloric acid (2 × 10 mL). The
2
2
further extracted with CH Cl (2 × 10 mL). The combined organic
2
2
layers were washed with brine, dried over MgSO , filtered, and
4
concentrated to give 10 as a light brown solid (470 mg, 84%). R 0.27
f
1
(
3:2 CH Cl :EtOAc). H NMR (acetone-d , 400 MHz): δ 3.69 (s,
organic layer was washed with brine (10 mL), dried over MgSO
4
,
2
2
6
2
8
7
1
H), 3.77 (s, 2H), 5.55 (s, 1H), 6.80 (d, J = 8.4 Hz, 2H), 7.19 (d, J =
filtered, and concentrated to give 20 as a colorless solid (1.44 g, 89%).
23
24
24
.4, 2H), 7.23 (d, J = 7.4 Hz, 1H), 7.30 (t, J = 7.4 Hz, 2H), 7.37 (d, J =
R
f
0.64 (4:1 EtOAc:Hex). [α]
D
−2.0 (c 0.72, CHCl
) [Lit. [α]
3 D
.4 Hz, 2H). 13C NMR (acetone-d , 100 MHz): δ 52.2, 52.5, 115.3,
1
−4.9 (c 0.72, CHCl
)]. H NMR (CDCl , 400 MHz): δ 1.25 (d, J =
6
3
3
26.7, 128.2, 128.3, 129.5, 130.9, 140.7, 156.7. HRMS (ESI-TOF):
7.2 Hz, 3H), 2.20 (dd, J = 17.5, 3.6 Hz, 1H), 2.67−2.80 (m, 1H), 2.82
+
calcd for C H NO (M+H) , 214.1226; obsd, 214.1225.
(dd, J = 17.6, 8.9 Hz, 1H), 5.13 (s, 2H), 7.28−7.40 (m, 3H), 7.43−
14
16
1
0
13
2
-Aminooxazole-5-carboxylic Acid (12).
Ethyl 2-amino-
7.56 (m, 2H). C NMR (CDCl
3
, 100 MHz): δ 16.7, 32.0, 33.7, 78.5,
oxazole-5-carboxylate 11 (2.00 g, 12.8 mmol) was added to 2 M
aqueous NaOH (50 mL) and stirred at room temperature for 2 h. The
mixture was heated to 60 °C and stirred for an additional 2 h.
Concentrated hydrochloric acid was added dropwise to the reaction
mixture until it became acidic by pH paper (pH ∼3−4) and a colorless
precipitate began to form. The mixture was cooled in the freezer for 1
h. The resulting solid was collected by filtration and washed with
128.5, 129.5, 130.1, 133.2, 170.6, 174.7. HRMS (ESI-TOF): calcd for
+
C
H13NaNO
(M+Na) , 242.0788; obsd, 242.0804.
12
3
(3S)-Boc-L-valine-(4S)-methyl-N-O-benzylsuccinimide
1
3,24
(21).
A solution of Boc-L-valine (992 mg, 4.56 mmol, 1.00 equiv)
and 1,1′-carbonyldiimidazole (814 mg, 5.02 mmol, 1.10 equiv) in dry
THF (4 mL) was stirred at room temperature for 2 h under N . A
2
solution of 20 (1.00 g, 4.56 mmol, 1.00 equiv) in dry THF (8 mL) was
added to the acylimidazole mixture at room temperature. This
combined solution was then added dropwise over 45 min to a stirring
solution of lithium hexamethyldisilazide (9.13 mL, 1.00 M, 9.13 mmol,
acetonitrile to give 12 as a colorless solid (1.60 g, 98%). R 0.24 (6:4:1
CHCl :MeOH:H O). H NMR (DMSO-d , 400 MHz): δ 7.38 (s,
f
1
3
2
6
13
2
1
1
H), 7.44 (s, 1H). C NMR (DMSO-d , 100 MHz): δ 136.0, 137.1,
6
+
2
.00 equiv) in dry THF (6 mL) under N at −78 °C. After the
59.1, 164.1. HRMS (ESI-TOF): calcd for C H N O (M+H) ,
29.0295; obsd, 129.0296.
2
4
5
2
3
addition was complete, the mixture was stirred at −78 °C for an
additional 15 min and then quenched with saturated aqueous
ammonium chloride (5 mL). The reaction mixture was allowed to
General Procedure for the Synthesis of Aminooxazole
Derivatives (2 and 13). Amine (1.00 equiv) was added to a solution
of 12 (1.00 equiv) and triethylamine (2.00 equiv) in dry DMF (10
mL). HATU (1.00 equiv) was added to the solution and the mixture
warm to room temperature and then partitioned between Et
mL) and water (20 mL). The organic phase was washed with brine
(20 mL), dried over MgSO , filtered, and concentrated. The resulting
residue was purified by flash chromatography on silica gel, eluting with
3:1 Hex:EtOAc, to give 21 as a colorless solid (800 mg, 42%). R 0.45
−33.9
, 400 MHz): δ 0.78 (d, J = 6.7 Hz,
O (20
2
stirred for 20 h under N . The mixture was partitioned between EtOAc
4
2
(
25 mL) and water (20 mL). The aqueous layer was further extracted
with EtOAc (2 × 20 mL). The combined organic layers were washed
with saturated aqueous NaHCO (15 mL) and brine (15 mL), dried
over MgSO , filtered, and concentrated. The resulting residue was
f
2
3
24
23
(2:1 Hex:EtOAc). [α]
−30.1 (c 0.70, CHCl
) [Lit. [α]
3 D
3
D
1
(c 0.70, CHCl
)]. H NMR (CDCl
4
3
3
dissolved in EtOAc (20 mL) and placed in a freezer overnight. The
resulting precipitate was collected by filtration, washed with EtOAc,
and dried.
3H), 0.99 (d, J = 6.7 Hz, 3H), 1.23 (d, J = 7.4 Hz, 3H), 1.46 (s, 9H),
2.24−2.41 (m, 1H), 3.14−3.30 (m, 1H), 3.71 (d, J = 4.0 Hz, 1H), 4.44
(dd, J = 8.8, 4.8 Hz, 1H), 5.09 (s, 2H), 5.60 (d, J = 8.6 Hz, 1H), 7.30−
1
0
13
2
-Amino-N,N-dibenzyloxazole-5-carboxamide (2). Starting
from dibenzylamine (249 μL, 256 mg, 1.30 mmol), carboxylic acid
2 (166 mg, 1.30 mmol), triethylamine (362 μL, 262 mg, 2.59 mmol),
7.39 (m, 3H), 7.40−7.46 (m, 2H). C NMR (CDCl
3
, 100 MHz): δ
15.6, 17.3, 19.6, 28.3, 29.6, 34.5, 55.0, 65.1, 78.7, 80.2, 128.6, 129.6,
1
130.2, 132.9, 156.0, 166.6, 173.1, 201.9. HRMS (ESI-TOF): calcd for
+
and HATU (493 mg, 1.30 mmol), the general procedure gave 2 as a
C
22
H30NaN
2
O
6
(M+Na) , 441.1996; obsd, 441.1994.
2
4
1
colorless solid (124 mg, 31%). R 0.16 (3:2 CH Cl :EtOAc). H NMR
(3S)-Boc-L-valine-(4S)-methylsuccinimide (22). A catalytic
f
2
2
13
(
DMSO-d , 400 MHz): δ 4.65 (br s, 4H), 7.06−7.41 (m, 13H).
NMR (DMSO-d , 100 MHz): δ 49.9, 52.5, 127.7, 128.5, 128.6, 129.1,
C
amount of 5% Pd/C (127 mg, 20% w/w) was added to a solution of
21 (637 mg, 1.52 mmol, 1.00 equiv) in dry methanol (10 mL), and the
6
6
1
34.9, 137.4, 137.7, 159.0, 163.1. HRMS (ESI-TOF): calcd for
mixture was stirred under H for 1 h. The reaction mixture was filtered
2
+
C H N O (M+H) , 308.1394; obsd, 308.1399.
through a pad of Celite, washed with methanol, and concentrated. The
residue was dissolved in acetonitrile (3 mL) and added dropwise to a
stirred solution of 2-bromoacetophenone (303 mg, 1.52 mmol, 1.00
equiv) in acetonitrile (5 mL) at room temperature. A solution of
triethylamine (318 μL, 231 mg, 2.28 mmol, 1.50 equiv) in acetonitrile
(1 mL) was added dropwise over 2 h, and the reaction mixture was
stirred at room temperature overnight. The mixture was concentrated
1
8
18
3
2
2
-Amino-N-benzyl-N-(4-hydroxybenzyl)oxazole-5-carboxamide
(
13). Starting from amine 10 (1.20 g, 5.64 mmol), carboxylic acid 12
(
721 mg, 5.64 mmol), triethylamine (1.57 mL, 1.14 g, 11.3 mmol),
and HATU (2.14 g, 5.64 mmol), the general procedure gave 13 as a
1
colorless solid (633 mg, 35%). R 0.57 (9:1 CH Cl :MeOH). H NMR
f
2
2
(
DMSO-d , 400 MHz): δ 4.51 (s, 2H), 4.59 (s, 2H), 6.73 (d, J = 8.1
6
13
Hz, 2H), 7.42 (m, 10H), 9.40 (s, 1H). C NMR (DMSO-d , 100
and then partitioned between CH
acid (2 × 20 mL). The organic layer was washed with brine (20 mL),
dried over MgSO , filtered, concentrated, and purified by flash
chromatography on silica gel eluting with 5:1 CH Cl :EtOAc to give
22 as a yellow solid (303 mg, 64%). R 0.63 (3:1 CH Cl :EtOAc).
−41.8 (c 0.66, CHCl )].
, 400 MHz): δ 0.79 (d, J = 6.7 Hz, 3H), 1.02 (d, J =
2 2
Cl (25 mL) and 5% hydrochloric
6
MHz): δ 49.3, 115.8, 127.6, 129.1, 134.7, 137.5, 137.8, 157.1, 158.8,
+
1
63.1. HRMS (ESI-TOF): calcd for C H N O (M+H) , 324.1343;
4
18
18
3
3
obsd, 324.1371.
2
2
(
4S)-Methylsuccinic Anhydride (15). Acetyl chloride (4.53 mL,
f
2
2
2
4
24
23
D
4
1
.75 g, 60.6 mmol, 4.00 equiv) was added to (4S)-methylsuccinic acid
4 (2.00 g, 15.1 mmol, 1.00 equiv) and heated gently at reflux for 6 h.
[α]
D
−26.0 (c 1.005, CHCl
3
) [Lit. [α]
3
1
H NMR (CDCl
3
Excess acetyl chloride and acetic acid were removed in vacuo to give
6.7 Hz, 3H), 1.31 (d, J = 7.4 Hz, 3H), 1.45 (s, 9H), 2.15−2.23 (m,
24
1
[
5 as a colorless solid (1.70 g, 97%). R 0.76 (9:1 CH Cl :MeOH).
1H), 3.14−3.22 (m, 1H), 3.93 (d, J = 5.3 Hz, 1H), 4.61 (dd, J = 9.0,
f
2
2
α]D24 −33.0 (c 1.77, CHCl ) [Lit. [α]D −36.3 (c 1.77, CHCl )].
24
24
13
4.0 Hz, 1H), 5.80 (d, J = 9.2 Hz, 1H), 9.82 (s, 1H). C NMR (CDCl ,
3
3
3
1
H NMR (CDCl , 400 MHz): δ 1.30 (d, J = 7.2 Hz, 3H), 2.45−2.56
100 MHz): δ 15.3, 16.9, 19.7, 28.3, 29.7, 38.4, 59.0, 64.9, 80.3, 156.1,
3
I
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
+
72.1, 179.1, 202.2. HRMS (ESI-TOF): calcd for C H NaN O (M
137.2, 139.9, 143.3, 164.9, 170.3, 174.2, 180.5, 203.8. HRMS (ESI-
15
24
2
5
+
+
Na) , 335.1577; obsd, 335.1582.
General Procedure for the Synthesis of Amines 23 and 25.
TOF): calcd for C H N O (M+H) , 454.2336; obsd, 454.2347.
25
32
3
5
2
4
General Procedure for the Synthesis of Allyl Ester Linker
2
9
Trifluoroacetic acid (2.0 mL) was added dropwise to a solution of Boc-
protected amine (22 or 24) (1.00 equiv) in CH Cl (2.0 mL) at 0 °C,
and stirring continued at 0 °C for 2 h, followed by the addition of
toluene (10 mL) and removal of solvent in vacuo. The resulting
residue was purified by flash chromatography on silica gel, with the less
polar byproducts being flushed with 2−3 columns lengths of 1:1
CH Cl :EtOAc and product being eluted with 9:1 CH Cl :MeOH.
Derivatives (27 and 30). Concentrated sulfuric acid (2.00 equiv)
was added to a solution of acid (1.00 equiv) in allyl alcohol. The
solution was heated at reflux for 6 h. The mixture was cooled to room
2
2
temperature, diluted with Et O (25 mL), and washed with saturated
2
aqueous NaHCO
extracted with Et
(2 × 20 mL). The aqueous phase was back-
3
O (15 mL). The combined organic layers were
2
washed with brine (10 mL), dried over MgSO , filtered, and
2
2
2
2
4
(
3S)-L-Valine-(4S)-methylsuccinimide (23). Starting from Boc-
concentrated. The resulting oil was purified by flash chromatography
on silica gel eluting with 4:1 Hex:EtOAc.
protected amine 22 (200 mg, 0.64 mmol), the general procedure
gave 23 as a light yellow foam (121 mg, 89%). R 0.50 (6:4:1
Allyl 16-Hydroxyhexadecanoate (27). According to the general
procedure, concentrated sulfuric acid (391 μL, 720 mg, 7.34 mmol),
16-hydroxyhexadecanoic acid 26 (1.00 g, 3.67 mmol), and allyl alcohol
f
CHCl :MeOH:H O). HRMS (ESI-TOF): calcd for C H N O (M
3
2
10 16
2
3
+
+
H) , 213.1234; obsd, 213.1238.
(3S)-β-Phenylalanine-L-valine-(4S)-methylsuccinimide (25). Start-
(35 mL) gave 27 as a colorless solid (704 mg, 61%). R 0.61 (1:1
f
1
ing from Boc-protected amine 24 (183 mg, 0.398 mmol), the general
Hex:EtOAc). H NMR (CDCl , 400 MHz): δ 1.20−1.35 (m, 22H),
3
procedure gave 25 as a light pink solid (141 mg, 98%). R 0.64 (6:4:1
1.53−1.65 (m, 4H), 2.33 (t, J = 7.6 Hz, 2H), 3.64 (t, J = 6.6 Hz, 2H),
f
CHCl :MeOH:H O). HRMS (ESI-TOF): calcd for C H N O (M
4.58 (d, J = 5.7 Hz, 2H), 5.24 (d, J = 10.4 Hz, 1H), 5.32 (d, J = 17.2
3
2
19 26
3
4
+
13
+
H) , 360.1918; obsd, 360.1927.
Hz, 1H), 5.92 (ddt, J = 17.2, 10.4, 5.7 Hz, 1H). C NMR (CDCl
3
, 100
(
3S)-N-Boc-3-amino-β-phenylalanine-L-valine-(4S)-methyl-
MHz): δ 24.9, 25.7, 29.1, 29.2, 29.4, 29.6, 32.8, 34.2, 62.9, 64.9, 118.0,
1
3
+
succinimide (24).
(S)-N-Boc-3-amino-3-phenylpropanoic acid
132.3, 173.5. HRMS (ESI-TOF): calcd for C H NaO (M+Na) ,
19
36
3
(
255 mg, 0.96 mmol, 1.20 equiv) was added to a solution of 23
335.2557; obsd, 335.2559.
(
170 mg, 0.80 mmol, 1.00 equiv) in dry DMF (5.0 mL) at 0 °C.
Allyl 8-Hydroxyoctanoate (30). According to the general
procedure, concentrated sulfuric acid (133 μL, 245 mg, 2.50 mmol),
8-hydroxyoctanoic acid 29 (200 mg, 1.25 mmol), and allyl alcohol (8
HATU (366 mg, 0.96 mmol, 1.20 equiv) was added to the solution,
and stirring of the mixture began under N . A solution of N,N-
2
diisopropylethylamine (349 μL, 259 mg, 2.00 mmol, 2.50 equiv) in dry
DMF (1.5 mL) was added dropwise to the reaction over 1 h, while
maintaining the temperature at 0 °C. The reaction was stirred for an
additional 30 min at 0 °C before allowing the reaction mixture to
warm to room temperature where stirring was continued for 20 h
mL) gave 30 as a yellow oil (148 mg, 59%). R 0.54 (1:1 Hex:EtOAc).
f
1
H NMR (CDCl , 400 MHz): δ 1.29−1.38 (m, 7H), 1.55 (p, J = 6.7
3
Hz, 2H), 1.64 (p, J = 6.9 Hz, 2H), 2.34 (t, J = 7.5 Hz, 2H), 3.61 (t, J =
6.6 Hz, 2H), 4.57 (d, J = 5.7 Hz, 2H), 5.23 (d, J = 10.4 Hz, 1H), 5.31
1
3
(d, J = 17.2 Hz, 1H), 5.92 (ddt, J = 17.2, 10.4, 5.7 Hz, 1H). C NMR
under N . The solvent was removed in vacuo, and the mixture was
(CDCl , 100 MHz): δ 24.8, 25.5, 28.9, 29.0, 32.6, 34.2, 62.7, 64.9,
2
3
partitioned between EtOAc (20 mL) and water (20 mL). The organic
118.1, 132.2, 173.5. HRMS (ESI-TOF): calcd for C H NaO (M
11
20
3
+
layer was washed with saturated aqueous NaHCO (20 mL) and brine
+Na) , 223.1305; obsd, 223.1302.
3
(
20 mL), dried over MgSO , filtered, concentrated, and purified by
General Procedure for the Synthesis of Iodoalkanoic Linker
Derivatives 28 and 31. Polymer-bound triphenylphosphine (100−
200 mesh, polystyrene cross-linked with 2% divinylbenzene: ∼3
mmol/g loading; 2.18 equiv) was suspended in dry CH Cl (10 mL).
4
flash chromatography on silica gel eluting with 3:2 Hex:EtOAc to give
4 as a colorless solid (166 mg, 45%). R 0.25 (1:1 Hex:EtOAc).
2
[
f
α]D20 −53.2 (c 1.00, CHCl ). H NMR (DMSO-d , 400 MHz): δ
1
3
6
2
2
0
3
=
.77 (d, J = 6.6 Hz, 3H), 0.85 (d, J = 6.6 Hz, 3H), 1.11 (d, J = 7.3 Hz,
H), 1.34 (s, 9H), 2.23−2.37 (m, 1H), 2.42−2.50 (m, 1H), 2.71 (dd, J
Iodine (2.18 equiv) was added in portions until the dark color
persisted. The mixture was stirred at room temperature under N for
2
13.9, 9.3 Hz, 1H), 2.84−3.03 (m, 1H), 3.99 (d, J = 5.4 Hz, 1H), 4.71
15 min. Imidazole (2.48 equiv) was added, and the reaction was stirred
for an additional 15 min. A solution of compound 27 or 30 (1.00
equiv) in dry CH Cl (5 mL) was added, and the mixture was heated
(
8
dd, J = 7.6, 5.3 Hz, 1H), 4.79−4.95 (m, 1H), 7.12−7.39 (m, 6H),
.01 (d, J = 8.3 Hz, 1H), 11.37 (br s, 1H). 13C NMR (DMSO-d , 100
6
2
2
MHz): δ 15.0, 17.5, 20.0, 28.6, 28.8, 43.0, 52.1, 58.3, 63.4, 78.3, 126.6,
at reflux for 30 min. The mixture was filtered to remove the resin and
1
27.3, 128.7, 144.0, 155.0, 170.4, 174.2, 180.5, 203.7. HRMS (ESI-
resin-bound byproducts, and the resin was washed with CH Cl (15
2
2
+
TOF): calcd for C H N O (M+H) , 460.2442; obsd, 460.2428.
mL). The filtrate was washed with saturated aqueous Na S O (15
24
34
3
6
2 2 3
Moiramide B (4). Sorbic acid (75 mg, 0.67 mmol, 1.20 equiv),
TBTU (215 mg, 0.67 mmol, 1.20 equiv), and N,N-diisopropylethyl-
amine (242 μL, 180 mg, 1.39 mmol, 2.50 equiv) were added to a
solution of 25 (200 mg, 0.56 mmol, 1.00 equiv) in dry DMF (2 mL),
mL) and water (15 mL). The organic layer was dried over MgSO4,
filtered, and concentrated.
Allyl 16-Iodohexadecanoate (28). Starting from polymer-bound
triphenylphosphine (2.19 g, 8.36 mmol), iodine (2.12 g, 8.36 mmol),
imidazole (647 mg, 9.51 mmol), and allyl ester linker 27 (1.20 g, 3.83
mmol), the general procedure gave 28 as a yellow solid (1.26 g, 78%).
and the mixture was stirred for 20 h under N . The mixture was
2
partitioned between EtOAc (10 mL) and water (6 mL). The organic
1
layer was washed with saturated aqueous NaHCO (5 mL) and brine
R 0.74 (2:1 Hex:EtOAc). H NMR (CDCl , 400 MHz): δ 1.21−1.33
3
f
3
(
5 mL), dried over MgSO , filtered, and concentrated. The resulting
(m, 22H), 1.38 (p, J = 7.0 Hz, 2H), 1.62 (p, J = 7.0 Hz, 2H), 1.81 (p, J
= 7.2 Hz, 2H), 2.31 (t, J = 7.5 Hz, 2H), 3.17 (t, J = 7.0 Hz, 2H), 4.56
4
residue was purified by flash chromatography on silica gel eluting with
2
% NH OH in 9:1 EtOAc:MeOH to give 4 as a colorless solid (81 mg,
(d, J = 5.6 Hz, 2H), 5.21 (d, J = 10.6 Hz, 1H), 5.30 (d, J = 17.1 Hz,
4
2
4
24
24
13
3
2%).5 Rf 0.54 (100% EtOAc). [α]D −77.5 (c 1.00, MeOH) [Lit.
1H), 5.91 (ddt, J = 17.1, 10.6, 5.6 Hz, 1H). C NMR (CDCl , 100
3
[
α]D2 −96.6 (c 0.28, MeOH)]. CD (c 1 μM, MeOH) λ (θ) 213.0
MHz): δ 7.0, 24.9, 28.5, 29.1, 29.2, 29.4, 29.5, 29.6, 29.6, 30.5, 33.6,
34.2, 64.8, 117.9, 132.4, 173.2.
max
(
+9.68), 238.0 (−21.08), 292.0 (+0.69), 320.0 (−1.26). CD spectrum
displayed λ values in good agreement with the literature though no
Allyl 8-Iodooctanoate (31). Starting from polymer-bound
triphenylphosphine (394 mg, 1.50 mmol), iodine (382 mg, 1.50
mmol), imidazole (116 mg, 1.71 mmol), and allyl ester linker 30 (138
mg, 0.690 mmol), the general procedure gave 31 as a yellow oil (178
max
11,24
1
concentrations were reported.
H NMR (DMSO-d , 400 MHz): δ
6
0
3
5
.74 (d, J = 6.7 Hz, 3H), 0.79 (d, J = 6.7 Hz, 3H), 1.06 (d, J = 7.4 Hz,
H), 1.77 (d, J = 6.4 Hz, 3H), 2.19−2.34 (m, 1H), 2.62 (dd, J = 14.1,
.9 Hz, 1H), 2.76 (dd, J = 14.2, 8.7, 1H), 2.83−2.97 (m, 1H), 3.92 (d,
1
mg, 83%). R 0.48 (1:1 Hex:EtOAc). H NMR (CDCl , 400 MHz): δ
f
3
J = 5.4 Hz, 1H), 4.62 (dd, J = 8.0, 5.6 Hz, 1H), 5.14−5.29 (m, 1H),
1.22−1.33 (m, 7H), 1.40 (p, J = 6.6 Hz, 2H), 1.64 (p, J = 7.3 Hz, 2H),
1.82 (p, J = 7.2 Hz, 2H), 2.34 (t, J = 7.5 Hz, 2H), 3.18 (t, J = 7.0 Hz,
2H), 4.58 (d, J = 5.7 Hz, 2H), 5.24 (d, J = 10.4 Hz, 1H), 5.32 (d, J =
5
1
5
1
3
.91 (d, J = 15.2, 1H), 6.06 (dq, J = 14.9, 6.8 Hz, 1H), 6.18 (dd, J =
4.9, 11.0 Hz, 1H), 6.94 (dd, J = 14.9, 11.0 Hz, 1H), 7.10−7.37 (m,
H), 8.11 (d, J = 8.4 Hz, 1H), 8.41 (d, J = 8.4 Hz, 1H), 11.35 (br s,
1
3
17.2 Hz, 1H), 5.92 (ddt, J = 17.2, 10.4, 5.7 Hz, 1H). C NMR
H). 13C NMR (DMSO-d , 100 MHz): δ 15.0, 17.7, 18.7, 19.9, 28.6,
(CDCl , 100 MHz): δ 7.2, 24.8, 28.2, 28.9, 30.3, 33.4, 34.1, 65.0,
6
3
9.5, 42.4, 50.3, 58.2, 63.5, 123.3, 126.9, 127.3, 128.6, 128.7, 130.3,
118.1, 132.3, 173.3.
J
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
General Procedure for the Synthesis of Aminooxazole
for 20 h under N . The solvent was removed in vacuo, and the mixture
was partitioned between EtOAc (10 mL) and water (10 mL). The
organic layer was washed with 1 M hydrochloric acid (10 mL),
2
Conjugated to Linker Derivatives 32 and 33. Cesium hydroxide
(
1.00 equiv) was added to a solution of 13 (1.00 equiv) in dry DMF (5
mL). Linker derivative (28 or 31) (1.00 equiv) in dry THF (2 mL)
was added to the reaction, and the mixture was heated at reflux for 24
h. Once cooled to room temperature, the mixture was filtered, and the
precipitate was washed with THF. The solvent was removed in vacuo,
the residue was partitioned between EtOAc (10 mL) and water (10
saturated aqueous NaHCO (10 mL), and brine (10 mL), dried over
3
MgSO , filtered, concentrated.
4
Dual-Ligand Inhibitor with 15C-Linker (5). Starting from amine 25
(21 mg, 0.059 mmol), compound 34 (41 mg, 0.071 mmol), HATU
(27 mg, 0.071 mmol), and N,N-diisopropylethylamine (31 μL, 23 mg,
0.178 mmol), the general procedure gave a residue that was purified by
RP-HPLC (Method A: the gradient started from 30% B to 90% B in
40 min followed by 90% B to 95% B in 5 min and keeping at 95% for
10 min; retention time = 28.7 min) that gave 5 as a yellow solid (21
mL), and the organic layer was dried over MgSO , filtered, and
4
concentrated.
Allyl 16-{4-[(2-Amino-N-benzyloxazole-5-carboxamido)methyl]-
phenoxy}hexadecanoate (32). Starting from cesium hydroxide (224
mg, 1.33 mmol), aminooxazole 13 (430 mg, 1.33 mmol), and linker 28
2
4
mg, 19%). R 0.70 (92:8 CH Cl :MeOH). [α]
−28.5 (c 0.425,
f
2
2
D
1
(
562 mg, 1.33 mmol), the general procedure gave a residue that was
MeOH). H NMR (CD OD, 400 MHz): δ 0.79 (d, J = 6.7 Hz, 3H),
3
purified by flash chromatography on silica gel eluting with 3:1
0.90 (d, J = 6.7 Hz, 3H), 1.16 (d, J = 7.3 Hz, 3H), 1.20−1.41 (m,
22H), 1.42−1.52 (m, 2H), 1.53−1.65 (m, 2H), 1.70−1.82 (m, 2H),
2.11−2.29 (m, 2H), 2.32−2.47 (m, 1H), 2.63−2.93 (m, 3H), 3.06−
3.20 (m, 1H), 3.95 (t, J = 5.9 Hz, 2H), 4.63 (s, 2H), 4.68 (s, 2H), 4.76
EtOAc:Hex that gave 32 as a pale yellow solid (166 mg, 20%). R 0.18
f
1
(
3:1 EtOAc:Hex). H NMR (CDCl , 400 MHz): δ 1.18−1.36 (m,
3
2
7
4
5
2H), 1.44 (p, J = 7.2 Hz, 2H), 1.63 (p, J = 6.9 Hz, 2H), 1.75 (p, J =
.1 Hz, 2H), 2.32 (t, J = 7.5 Hz, 2H), 3.93 (t, J = 6.3 Hz, 2H), 4.53−
.68 (m, 6H), 5.22 (d, J = 10.4 Hz, 1H), 5.30 (d, J = 17.2 Hz, 1H),
.91 (ddt, J = 17.2, 10.4, 5.7 Hz, 1H), 6.00 (br s, 2H), 6.85 (d, J = 7.9
(d, J = 4.5 Hz, 1H), 5.28−5.43 (m, 1H), 6.80−6.91 (m, 2H), 7.05−
+
7.40 (m, 13H). HRMS (ESI-TOF): calcd for C53
H
N
6
O
(M+H) ,
71
8
919.5328; obsd, 919.5344.
13
Hz, 2H), 7.02−7.38 (m, 8H). C NMR (CDCl , 100 MHz): δ 25.0,
Dual-Ligand Inhibitor with 7C-Linker (6). Starting from amine 25
(67 mg, 0.186 mmol), compound 35 (104 mg, 0.224 mmol), HATU
(85 mg, 0.224 mmol), and N,N-diisopropylethylamine (97 μL, 72 mg,
3
2
1
1
6
6.1, 29.1, 29.3, 29.5, 29.6, 29.7, 34.3, 49.1, 50.2, 64.9, 68.1, 114.8,
18.1, 127.6, 127.8, 128.5, 128.9, 132.3, 133.3, 138.1, 158.7, 159.5,
62.0, 173.6. HRMS (ESI-TOF): calcd for C H N O (M+H) ,
18.3901; obsd, 618.3883.
Allyl 8-{4-[(2-Amino-N-benzyloxazole-5-carboxamido)methyl]-
phenoxy}octanoate (33). Starting from cesium hydroxide (257 mg,
.53 mmol), aminooxazole 13 (494 mg, 1.53 mmol), and linker 31
474 mg, 1.53 mmol), the general procedure gave a residue that was
+
0
.559 mmol), the general procedure gave a residue that was purified by
RP-HPLC (Method B: the gradient started from 20% B to 80% B in
0 min followed by 80% B to 95% B in 5 min and keeping at 95% for
10 min; retention time = 20.5 min) that gave 6 as a yellow solid (26
37
52
3
5
3
2
4
1
(
mg, 17%). R
f
0.59 (92:8 CH
2
Cl
2
:MeOH). [α]
D
−22.1 (c 1.17,
1
MeOH). H NMR (CD OD, 500 MHz): δ 0.74 (d, J = 6.7 Hz, 3H),
3
purified by flash chromatography on silica gel eluting with 4:1
0.84 (d, J = 6.7 Hz, 3H), 1.11 (d, J = 7.4 Hz, 3H), 1.21−1.36 (m, 7H),
1.37−1.47 (m, 2H), 1.48−1.60 (m, 2H), 1.62−1.77 (m, 2H), 2.08−
EtOAc:Hex that gave 33 as a yellow oil (160 mg, 21%). R 0.43 (2.5:1
f
1
2
.24 (m, 2H), 2.30−2.41 (m, 1H), 2.71−2.84 (m, 2H), 3.02−3.15 (m,
H), 3.29 (s, 1H), 3.78−3.97 (m, 2H), 4.57 (s, 2H), 4.63 (s, 2H), 4.72
EtOAc:Hex). H NMR (CDCl , 400 MHz): δ 1.25−1.52 (m, 7H),
3
1
1
2
1
6
.65 (p, J = 6.8 Hz, 2H), 1.77 (p, J = 6.9 Hz, 2H), 2.34 (t, J = 7.5 Hz,
H), 3.93 (t, J = 6.3 Hz, 2H), 4.48−4.69 (m, 6H), 5.22 (d, J = 10.4 Hz,
H), 5.31 (d, J = 17.1 Hz, 1H), 5.90 (ddt, J = 17.1, 10.4, 5.4 Hz, 1H),
(
d, J = 4.8 Hz, 1H), 5.24−5.39 (m, 1H), 6.83 (d, J = 7.4 Hz, 2H),
7
.04−7.40 (m, 13H). HRMS (ESI-TOF): calcd for C H N O (M
45 55
6
8
+
13
+H) , 807.4076; obsd, 807.4077.
.00 (br s, 2H), 6.85 (d, J = 8.0 Hz, 2H), 6.96−7.41 (m, 8H).
NMR (CDCl , 100 MHz): δ 24.9, 25.9, 29.0, 29.1, 29.2, 34.2, 49.2,
C
Kinetic Assays. Acetyl-CoA carboxylase was purified according to
3
Broussard et al. from a strain of E. coli that overexpresses the four
6
1
5
5.0, 68.0, 114.8, 118.2, 127.6, 128.9, 132.3, 133.3, 138.4, 158.7, 159.5,
2
0
+
genes that encode the enzyme. The activity of ACC was determined
spectrophotometrically by measuring the production of ADP using
pyruvate kinase and lactate dehydrogenase and following the oxidation
of NADH at 340 nm. Each reaction mixture contained 17.5 units of
lactate dehydrogenase, 10.5 units of pyruvate kinase, 0.5 mM
61.7, 173.5. HRMS (ESI-TOF): calcd for C H N O (M+H) ,
29
36
3
5
06.2649; obsd, 506.2656.
General Procedure for the Synthesis of Allyl Deprotected
Aminooxazole-Linker Derivatives 34 and 35. A 2 M aqueous
NaOH solution (5 mL) was added to a solution of aminooxazole-
linker derivative 32 or 33 (1.00 equiv) in methanol (5 mL) and THF
phosphoenolpyruvate, 0.2 mM NADH, 20 mM MgCl , 15 mM
2
potassium bicarbonate, and 100 mM HEPES (pH 8.0). For reactions
measuring the effects of inhibition versus ATP, each reaction
contained a concentration of 0.2 mM acetyl-CoA, with ATP
concentrations ranging from 0.8 μM to 56 μM. For reactions
measuring the effects of inhibition versus acetyl-CoA, each reaction
contained a concentration of 5 μM ATP, with acetyl-CoA
concentrations ranging from 50 μM to 1000 μM. All reactions were
conducted in a total of 0.5 mL in a 1 cm path length quartz cuvette,
and all reactions were initiated by the addition of enzyme.
Spectrophotometric data were collected using an Agilent Cary 60
UV−vis spectrophotometer interfaced with a personal computer with
a data acquisition program.
(
2 mL), and the mixture was stirred at room temperature for 18 h.
Concentrated hydrochloric acid was added dropwise to the reaction
mixture until it became acidic by pH paper (pH ∼3−4). The aqueous
layer was extracted with EtOAc (3 × 10 mL), and the combined
organic layers were washed with brine (6 mL), dried over MgSO4,
filtered, and concentrated.
1
6-{4-[(2-Amino-N-benzyloxazole-5-carboxamido)methyl]-
phenoxy}hexadecanoic Acid (34). Following the general procedure,
compound 32 (89 mg, 0.144 mmol) gave 34 as a yellow solid (74 mg,
8
0
+
9%) that was taken into the next step without further purification. Rf
.42 (4:1 EtOAc:Hex). HRMS (ESI-TOF): calcd for C H N O (M
3
4
48
3
5
+
H) , 578.3588; obsd, 578.3573.
-{4-[(2-Amino-N-benzyloxazole-5-carboxamido)methyl]-
Kinetic Analysis. Data were analyzed by nonlinear regression
8
4
4
using the computer programs of Cleland. Data describing
phenoxy}octanoic Acid (35). Following the general procedure,
compound 33 (160 mg, 0.317 mmol) gave 35 as a yellow oil (104
mg, 71%) that was taken into the next step without further
purification. R 0.23 (4:1 EtOAc:Hex). HRMS (ESI-TOF): calcd for
C H N O (M+H) , 466.2336; obsd, 466.2324.
competitive inhibition were fitted to eq 1, where v is the initial
velocity, V is the maximal velocity, A is the substrate concentration, Km
is the Michaelis constant, I is the concentration of the inhibitor, and Kis
is the slope inhibition constant.
f
+
2
6
32
3
5
General Procedure for the Synthesis of Dual-Ligand
VA
I
Inhibitors 5 and 6. Amine 25 (1.00 equiv) was added to a solution
of deprotected aminooxazole-linker (34 and 35) (1.20 equiv) and
HATU (1.20 equiv) in dry DMF (3 mL) at 0 °C. N,N-
Diisopropylethylamine (3.00 equiv) was added dropwise, and the
reaction was stirred for an additional 30 min at 0 °C before allowing
reaction to warm to room temperature where stirring was continued
v =
Km 1 + K + A
(
)
is
(1)
Data describing noncompetitive inhibition were fitted to eq 2,
where v is the initial velocity, V is the maximal velocity, A is the
substrate concentration, Km is the Michaelis constant, I is the
K
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
concentration of the inhibitor, and Kis and K are the slope and
intercept inhibition constants, respectively.
ii
ACKNOWLEDGMENTS
■
The authors thank Dr. Jens Pohlmann for invaluable
information on synthetic procedures and Dr. Zhenkun Ma at
TenNor Therapeutics for the use of bacterial strains. The
authors also thank Tyler Broussard for providing the ACC for
kinetics studies and Dr. Ted Gauthier and the LSU AgCenter
Biotechnology Laboratories for the collection of the CD
spectrum.
VA
v =
I
Kis
I
Kii
Km
(1 +
+ A 1 +
)
(
)
(2)
Determination of MICs. Determination of MICs was conducted
in accordance with the Clinical and Laboratory Standards Institute
(
CLSI) methodology using the broth microdilution methods with
45
cation-adjusted Mueller Hinton (MHII) medium. The strains tested
were used with permission from the strain collection of TenNor
Therapeutics. Test organisms were streaked onto agar medium and
incubated overnight at 35 °C. Experimental compounds were tested
over a concentration range prepared by 2-fold serial dilution after
solubilizing compounds in 100% DMSO (or sterile water for
ABBREVIATIONS USED
■
ACC, acetyl-CoA carboxylase; BC, biotin carboxylase; CT,
carboxyltransferase; BCCP, biotin carboxyl carrier protein;
MPC, mutant prevention concentration; MSW, mutant
selection window; PMB(N), polymyxin B (nonapeptide)
gentamicin and kanamycin) to achieve a final stock concentration of
5
1
0
0 mg/L. A standardized inoculum of 5 × 10 CFU/mL in a volume of
.1 mL was incubated with antibiotic for 18−24 h at 35 °C. The MIC
REFERENCES
■
(
1) CDC. Antibiotic Resistance Threats in the United States, 2013; U.S.
was scored visually as the lowest concentration of antibiotic where no
growth is apparent by visual inspection for turbidity.
Department of Health and Human Services, Centers for Disease
Control and Prevention: Atlanta, GA, 2013; http://www.cdc.gov/
drugresistance/threat-report-2013/.
2) Heath, R. J.; Rock, C. O. Fatty acid biosynthesis as a target for
novel antibacterials. Curr. Opin. Investig. Drugs 2004, 5, 146−153.
3) Zhang, Y. M.; White, S. W.; Rock, C. O. Inhibiting bacterial fatty
acid synthesis. J. Biol. Chem. 2006, 281, 17541−17544.
4) Campbell, J. W.; Cronan, J. E., Jr. Bacterial fatty acid biosynthesis:
Sensitization Assay using PMBN. A culture of E. coli strain
JM109 was grown overnight in Luria broth (LB) at 37 °C. Culture
tubes containing 2 mL of fresh LB media were inoculated with 10 μL
of the saturated E. coli culture. Polymyxin B nonapeptide (PMBN) was
dissolved in water to achieve a 1 mM stock and inoculated into the
tubes to generate final concentrations of 0.5, 1, 5, and 10 μg/mL.
Compound 5 was dissolved in 100% DMSO to achieve a 5 mM stock
solution and inoculated into the tubes to generate final concentrations
of 1, 10, 50, and 100 μM. After inoculation, the culture tubes were
incubated for 4 h at 37 °C. Optical density readings were measured at
(
(
(
targets for antibacterial drug discovery. Annu. Rev. Microbiol. 2001, 55,
305−332.
(5) Cronan, J. E., Jr.; Waldrop, G. L. Multi-subunit acetyl-CoA
carboxylases. Prog. Lipid Res. 2002, 41, 407−435.
6
00 nm in 1 mL plastic cuvettes. Spectrophotometric data were
(6) Tanabe, T.; Wada, K.; Okazaki, T.; Numa, S. Acetyl-coenzyme-A
carboxylase from rat liver. Subunit structure and proteolytic
modification. Eur. J. Biochem. 1975, 57, 15−24.
collected using an Agilent Cary 60 UV−vis spectrophotometer.
Generation of Spontaneous Resistant Mutant Strains and
Resistance Studies. Single-step spontaneous resistance selections
were undertaken on agar medium using overnight-saturated cultures
grown in cation-adjusted Mueller Hinton (MHII). Briefly, compounds
were added to molten MHII agar after allowing time to cool to 55 °C.
Compounds were dosed in 2-fold increasing concentrations starting at
× MIC. The test organism was grown overnight to saturation in
MHII broth. The cells were concentrated by centrifugation, and the
viable number of CFU plated was determined by serial dilution and
(7) Rock, C. O.; Cronan, J. E. Escherichia coli as a model for the
regulation of dissociable (type II) fatty acid biosynthesis. Biochim.
Biophys. Acta 1996, 1302, 1−16.
(8) Heath, R. J.; White, S. W.; Rock, C. O. Inhibitors of fatty acid
synthesis as antimicrobial chemotherapeutics. Appl. Microbiol.
Biotechnol. 2002, 58, 695−703.
1
(9) Miller, J. R.; Dunham, S.; Mochalkin, I.; Banotai, C.; Bowman,
M.; Buist, S.; Dunkle, B.; Hanna, D.; Harwood, H. J.; Huband, M. D.;
Karnovsky, A.; Kuhn, M.; Limberakis, C.; Liu, J. Y.; Mehrens, S.;
Mueller, W. T.; Narasimhan, L.; Ogden, A.; Ohren, J.; Prasad, J. V.;
Shelly, J. A.; Skerlos, L.; Sulavik, M.; Thomas, V. H.; VanderRoest, S.;
Wang, L.; Wang, Z.; Whitton, A.; Zhu, T.; Stover, C. K. A class of
selective antibacterials derived from a protein kinase inhibitor
pharmacophore. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 1737−1742.
9
plating. Each drug plate received a total of 4 × 10 CFU/plate (i.e., 0.1
10
mL plated of 4 × 10 CFU/mL in the concentrated culture). Plates
were monitored for the appearance of drug-resistant colonies. For
selection of spontaneous moiramide resistance in the wild-type S.
aureus and E. coli tolC backgrounds, moiramide was dosed at ∼2−8
times their respective MICs, and potential drug-resistant isolates were
evaluated after secondary passage through a drug-free intermediate.
Stable moiramide resistance was verified by broth microdilution MIC
assays, as above. The MPC was defined as the lowest consecutive drug
concentration capable of fully suppressing the emergence of antibiotic-
(
10) Mochalkin, I.; Miller, J. R.; Narasimhan, L.; Thanabal, V.;
Erdman, P.; Cox, P. B.; Prasad, J. V.; Lightle, S.; Huband, M. D.;
Stover, C. K. Discovery of antibacterial biotin carboxylase inhibitors by
virtual screening and fragment-based approaches. ACS Chem. Biol.
2009, 4, 473−483.
41
resistant bacterial subpopulations.
(
11) Needham, J.; Kelly, M. T.; Ishige, M.; Andersen, R. J. Andrimid
and moiramides A-C, metabolites produced in culture by a marine
isolate of the bacterium Pseudomonas fluorescens: Structure elucidation
and biosynthesis. J. Org. Chem. 1994, 59, 2058−2063.
ASSOCIATED CONTENT
■
*
S
Supporting Information
H and 13C NMR spectra for compounds and supplemental
figures and tables. This material is available free of charge via
the Internet at http://pubs.acs.org.
(
12) Freiberg, C.; Brunner, N. A.; Schiffer, G.; Lampe, T.; Pohlmann,
1
J.; Brands, M.; Raabe, M.; Habich, D.; Ziegelbauer, K. Identification
and characterization of the first class of potent bacterial acetyl-CoA
carboxylase inhibitors with antibacterial activity. J. Biol. Chem. 2004,
2
(
79, 26066−26073.
13) Pohlmann, J.; Lampe, T.; Shimada, M.; Nell, P. G.;
AUTHOR INFORMATION
■
*
Pernerstorfer, J.; Svenstrup, N.; Brunner, N. A.; Schiffer, G.;
Freiberg, C. Pyrrolidinedione derivatives as antibacterial agents with
a novel mode of action. Bioorg. Med. Chem. Lett. 2005, 15, 1189−1192.
Corresponding Author
E-mail: gwaldro@lsu.edu. Telephone: (225) 578-5209.
(
14) The activity of carboxyltransferase is measured in the reverse or
Notes
non-physiological direction where malonyl-CoA and biocytin (a biotin
The authors declare no competing financial interest.
analogue) are the substrates. The formation of the product acetyl-CoA
L
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
is detected using the coupling enzymes, citrate synthase and malate
dehydrogenase: Blanchard, C. Z.; Waldrop, G. L. Overexpression and
kinetic characterization of the carboxyltransferase component of
acetyl-CoA carboxylase. J. Biol. Chem. 1998, 273, 19140−19145.
biotic: microbiology profiling studies with staphylococci and
streptococci. Antimicrob. Agents Chemother. 2008, 52, 2324−2334.
(35) Aeschlimann, J. R.; Dresser, L. D.; Kaatz, G. W.; Rybak, M. J.
Effects of NorA inhibitors on in vitro antibacterial activities and
postantibiotic effects of levofloxacin, ciprofloxacin, and norfloxacin in
genetically related strains of Staphylococcus aureus. Antimicrob. Agents
Chemother. 1999, 43, 335−340.
(36) Viljanen, P.; Vaara, M. Susceptibility of Gram-negative bacteria
to polymyxin B nonapeptide. Antimicrob. Agents Chemother. 1984, 25,
701−705.
(
15) Li, S. J.; Cronan, J. E., Jr. Growth rate regulation of Escherichia
coli acetyl coenzyme A carboxylase, which catalyzes the first committed
step of lipid biosynthesis. J. Bacteriol. 1993, 175, 332−340.
(
16) Silver, L. L. Multi-targeting by monotherapeutic antibacterials.
Nat. Rev. Drug Discovery 2007, 6, 41−55.
(
17) Pokrovskaya, V.; Baasov, T. Dual-acting hybrid antibiotics: a
promising strategy to combat bacterial resistance. Expert Opin. Drug
Discovery 2010, 5, 883−902.
(37) Tsubery, H.; Ofek, I.; Cohen, S.; Eisenstein, M.; Fridkin, M.
Modulation of the hydrophobic domain of polymyxin B nonapeptide:
effect on outer-membrane permeabilization and lipopolysaccharide
neutralization. Mol. Pharmacol. 2002, 62, 1036−1042.
(38) Zavascki, A. P.; Goldani, L. Z.; Li, J.; Nation, R. L. Polymyxin B
for the treatment of multidrug-resistant pathogens: a critical review. J.
Antimicrob. Chemother. 2007, 60, 1206−1215.
(39) Velkov, T.; Roberts, K. D.; Nation, R. L.; Wang, J.; Thompson,
P. E.; Li, J. Teaching ’old’ polymyxins new tricks: new-generation
lipopeptides targeting Gram-negative ’superbugs’. ACS Chem. Biol.
(
18) O’Connell, K. M.; Hodgkinson, J. T.; Sore, H. F.; Welch, M.;
Salmond, G. P.; Spring, D. R. Combating multidrug-resistant bacteria:
Current strategies for the discovery of novel antibacterials. Angew.
Chem., Int. Ed. 2013, 10706−10733.
(
19) East, S. P.; Silver, L. L. Multitarget ligands in antibacterial
research: progress and opportunities. Expert Opin. Drug Discovery
013, 8, 143−156.
20) Broussard, T. C.; Price, A. E.; Laborde, S. M.; Waldrop, G. L.
2
(
2
(
014, 9, 1172−1177.
Complex formation and regulation of Escherichia coli acetyl-CoA
carboxylase. Biochemistry 2013, 52, 3346−3357.
40) Zhao, X.; Drlica, K. Restricting the selection of antibiotic-
resistant mutant bacteria: measurement and potential use of the
(
21) Li, H. Q.; Yan, T.; Yang, Y.; Shi, L.; Zhou, C. F.; Zhu, H. L.
mutant selection window. J. Infect. Dis. 2002, 185, 561−565.
Synthesis and structure-activity relationships of N-benzyl-N-(X-2-
hydroxybenzyl)-N’-phenylureas and thioureas as antitumor agents.
Bioorg. Med. Chem. 2010, 18, 305−313.
(41) Drlica, K. The mutant selection window and antimicrobial
resistance. J. Antimicrob. Chemother. 2003, 52, 11−17.
(42) Brylinski, M.; Waldrop, G. L. Computational redesign of
(
22) McWhorter, W.; Fredenhagen, A.; Nakanishi, K.; Komura, H.
bacterial biotin carboxylase inhibitors using structure-based virtual
Stereocontrolled synthesis of andrimid and a structural requirement
screening of combinatorial libraries. Molecules 2014, 19, 4021−4045.
for the activity. J. Chem. Soc., Chem. Commun. 1989, 299−301.
(
43) No R was obtained for imine compound as product hydrolyzed
(
23) Rama Rao, A. V.; Singh, A. K.; Varaprasad, C. V. N. S. A
f
on silica.
convenient diastereoselective total synthesis of andrimid. Tetrahedron
Lett. 1991, 32, 4393−4396.
(44) Cleland, W. W. Statistical analysis of enzyme kinetic data.
Methods Enzymol. 1979, 63, 103−138.
(
24) Davies, S. G.; Dixon, D. J. Asymmetric syntheses of moiramide
(45) Clinical and Laboratory Standards Institute. Methods for
B and andrimid. J. Chem. Soc., Perkin Trans. 1 1998, 2635−2643.
dilution antimicrobial susceptibility tests for bacteria that grow
aerobically, approved standard, 7th ed.; Clinical and Laboratory
Standards Institute: Wayne, PA, 2006; CLSI document M7-A7, Vol.
(
25) Fredenhagen, A.; Tamura, S. Y.; Kenny, P. T. M.; Komura, H.;
Naya, Y.; Nakanishi, K.; Niegyama, K.; Sugiura, M.; Kita, H. Andrimid,
a new peptide antibiotic produced by an intracellular bacterial
symbiont isolated from a brown planthopper. J. Am. Chem. Soc.
2
6, no. 2.
1
(
987, 109, 4409−4411.
26) Midgley, G.; Thomas, C. B. Selectivity of radical formation in
the reaction of carbonyl-compounds with manganese (III) acetate. J.
Chem. Soc., Perkin Trans. 2 1987, 1103−1108.
(
27) Bailey, P. D.; Morgan, K. M.; Smith, D. I.; Vernon, J. M. Spiro
cyclisations of N-acyliminium ions involving an aromatic π-
nucleophile. Tetrahedron Lett. 2003, 59, 3369−3378.
(
28) Robin, S.; Zhu, J.; Galons, H.; Pham-Huy, C.; Claude, J. R.;
Tomas, A.; Viossat, B. A convenient asymmetric synthesis of
thalidomide. Tetrahedron: Asymmetry 1995, 6, 1249−1252.
(
29) Aron, Z. D.; Overman, L. E. Total synthesis and properties of
the crambescidin core zwitterionic acid and crambescidin 359. J. Am.
Chem. Soc. 2005, 127, 3380−3390.
(
30) Jobron, L.; Hummel, G. Solid-phase synthesis of new S-
glycoamino acid building blocks. Org. Lett. 2000, 2, 2265−2267.
(
31) Zhang, Q.; Wu, Y.; Wang, L.; Hu, B.; Li, P.; Liu, F. Effect of
hapten structures on specific and sensitive enzyme-linked immuno-
sorbent assays for N-methylcarbamate insecticide metolcarb. Anal.
Chim. Acta 2008, 625, 87−94.
(
32) Bennett, B. D.; Kimball, E. H.; Gao, M.; Osterhout, R.; Van
Dien, S. J.; Rabinowitz, J. D. Absolute metabolite concentrations and
implied enzyme active site occupancy in Escherichia coli. Nat. Chem.
Biol. 2009, 5, 593−599.
(
33) Freiberg, C.; Pohlmann, J.; Nell, P. G.; Endermann, R.;
Schuhmacher, J.; Newton, B.; Otteneder, M.; Lampe, T.; Habich, D.;
Ziegelbauer, K. Novel bacterial acetyl coenzyme A carboxylase
inhibitors with antibiotic efficacy in vivo. Antimicrob. Agents Chemother.
2
(
006, 50, 2707−2712.
34) Robertson, G. T.; Bonventre, E. J.; Doyle, T. B.; Du, Q.;
Duncan, L.; Morris, T. W.; Roche, E. D.; Yan, D.; Lynch, A. S. In vitro
evaluation of CBR-2092, a novel rifamycin-quinolone hybrid anti-
M
dx.doi.org/10.1021/jm501082n | J. Med. Chem. XXXX, XXX, XXX−XXX