From triclosan toward the clinic: Discovery of nonbiocidal, potent FabI inhibitors for the treatment of resistant bacteria

Article abstract of DOI:10.1021/jm301113w

In this paper, we present some elements of our optimization program to decouple triclosan's specific FabI effect from its nonspecific cytotoxic component. The implementation of this strategy delivered highly specific, potent, and nonbiocidal new FabI inhibitors. We also disclose some preclinical data of one of their representatives, 83, a novel antibacterial compound active against resistant staphylococci and some clinically relevant Gram negative bacteria that is currently undergoing clinical trials.

Full text of DOI:10.1021/jm301113w

Article
pubs.acs.org/jmc
From Triclosan toward the Clinic: Discovery of Nonbiocidal, Potent
FabI Inhibitors for the Treatment of Resistant Bacteria
Vincent Gerusz,*,† Alexis Denis,†,§ Fabien Faivre,† Yannick Bonvin,† Mayalen Oxoby,† Sophia Briet,†
Geraldine LeFralliec,† Chrystelle Oliveira,† Nicolas Desroy,†,∥ Cedric Raymond,† Laetitia Peltier,†,∥
́
́
̈
Franco̧
is Moreau,‡ Sonia Escaich,‡,⊥ Vanida Vongsouthi,‡ Stephanie Floquet,‡ Elodie Drocourt,‡
́
Armelle Walton,‡ Laure Prouvensier,‡ Marc Saccomani,‡ Lionel Durant,‡,# Jean-Marie Genevard,‡
Vanessa Sam-Sambo,‡ and Coralie Soulama-Mouze‡,∞
‡
†Medicinal Chemistry and Biology, Mutabilis, 102 Avenue Gaston Roussel, 93230 Romainville, France
S
* Supporting Information
ABSTRACT: In this paper, we present some elements of our
optimization program to decouple triclosan’s specific FabI effect from
its nonspecific cytotoxic component. The implementation of this strategy
delivered highly specific, potent, and nonbiocidal new FabI inhibitors.
We also disclose some preclinical data of one of their representatives, 83,
a novel antibacterial compound active against resistant staphylococci and
some clinically relevant Gram negative bacteria that is currently
undergoing clinical trials.
numerous reports of FabI inhibitors18 involving diazaborines,19
INTRODUCTION
■
4-pyridones,20 naphthyridinones,21 triclosan,22 and ana-
The increase of antimicrobial resistance has become a global
logues23−30 have validated this target as one of the most
healthcare problem, rendering obsolete many antibiotic
attractive of the FASII pathway. Although a couple of these
therapies.1,2 Among Gram-positive pathogens, the emergence
inhibitors have entered clinical trials,31,32 none of them have
and spread of multidrug resistant Staphylococcus aureus
made it yet to market.
constitute a well-known problem.3−5 There is therefore an
Triclosan is a widely used broad-spectrum biocide,33 which
urgent medical need for new antibacterial drugs, especially with
also displays reversible inhibition of E. coli FabI.34 Despite its
novel mechanisms of action that would display minimal cross-
biocidal component and intravenous toxicity warranting against
resistance with currently used treatments.
a systemic use,35 it can still be considered as an attractive lead
The bacterial fatty acid biosynthesis pathway has generated
for a number of reasons. Indeed, its active site entry results in a
much interest for the development of such novel class agents.6,7
reordering of a loop of amino acids, making it a slow, tight-
The organization of the bacterial fatty acid synthase type II
binding inhibitor36 with a long residence time that has been
system based on individual enzymes (FASII system, Figure 1) is
correlated with in vivo activity.37 Its picomolar Ki for binding
different from the multifunctional fatty acid synthase type I
the enzyme−cofactor complex34 and low molecular weight also
system found in eukaryotes, therefore providing good prospects
entail a high ligand efficiency,38 and finally there is a variety of
for selective inhibition. Moreover, all the representative targets
available costructures.34,36,39,40
of the FASII system have already been characterized by X-ray
In this report, we describe our drug design efforts to harvest
crystallography or NMR, which should facilitate the rational
its potency without its biocidal component. We also disclose
design of inhibitors. Although recent communications have
some elements of structure−activity relationships (SARs) that
expressed some concern about the validity of FASII as
led to the identification of a potent antistaphylococcal clinical
antibiotic targets for Gram-positive pathogens,8,9 counterexperi-
candidate.
ments have brought some evidence to the essentiality of this
pathway for S. aureus.10−12 While human clinical efficacy data
CHEMISTRY
■
would be needed to close this debate, significant work
continues by both academic groups and the pharmaceutical
The synthetic routes used to prepare the left-part derivatives of
Table 1 are illustrated in Scheme 1. Guaiacol 1 was substituted
industry to validate the FASII targets and discover relevant
inhibitors.13
by pyridine 2 under basic conditions to provide the
phenoxypyridine 3, which was demethylated with boron
tribromide to afford the final compound 4. This general
The last step of the fatty acid elongation cycle is catalyzed by
enoyl-ACP reductases. Among them, FabI constitutes the single
isoform in major pathogens such as S. aureus,14 Escherichia
coli,15 and Mycobacterium tuberculosis16 (also called InhA for the
Received: July 30, 2012
Published: October 23, 2012
last). The clinical success of the InhA inhibitor isoniazid17 and
© 2012 American Chemical Society
9914
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
Figure 1. Schematic representation of FASII system.
procedure was used to provide the other analogues, with in
some instances additional steps to prepare specific side chains.
The left-part moiety of 56 was obtained by Negishi cross-
coupling reaction of halide 5 with isobutylzinc bromide.
Reaction of 7 with Ruppert−Prakash reagent and tetrabuty-
lammonium fluoride provided the common hydroxyl inter-
mediate 8, which was either dehydroxylated via tosylation and
hydrogenation to afford the trifluoroethyl side chain of 57 or
fluorinated using diethylaminosulfur trifluoride and dehydro-
fluorinated under basic conditions to yield the trifluorovinyl
side chain of 58. Compound 11 was submitted to Suzuki
coupling with 4-methylphenylboronic acid to provide inter-
mediate 12 bearing the left-part of 59. Compound 13 was
chlorinated using mesyl chloride to afford 14, which was in turn
demethylated with boron tribromide and methoxylated by
displacement of the chloride with sodium methoxide to provide
15. Alternatively, the chloro intermediate 14 was substituted
with imidazole under sodium hydride conditions to yield the
side chain of 62 and in similar fashion compounds 63 and 64.
A variety of methods were utilized to prepare the right-part
derivatives of Table 2 as illustrated in Scheme 2. Guaiacol 1 was
coupled to 4-methoxyphenylboronic acid under copper acetate
conditions to provide 17, which was monodemethylated using
boron tribromide to provide 65. Compounds 66, 67, and 68
were obtained in a similar fashion. The side chain of 69 was
elaborated by deprotonating intermediate 18 with s-butyl-
lithium followed by substitution with 2-bromomethyl-1,3-
dioxolane. Compound 70 was obtained by reacting catechol
20 with 2-fluoronitrobenzene under basic conditions followed
by reduction under hydrogenation conditions and the usual
demethylation using boron tribromide. Intermediate 23 and
compounds 71 and 73 were prepared similarly, while
compound 72 was derived from methylation of aniline 21
using formylation with freshly prepared acetic formic anhydride
followed by borane reduction. Chlorosulfonylation via the
diazonium of 23 followed by reaction with ammonia afforded
24, which was demethylated to provide 74 and in a similar
fashion compounds 75, 76, and 77. Reductive amination of
common intermediate 23 yielded 25 bearing the side chain of
78, while its sulfonylation using cyclopropanesulfonyl chloride
afforded 26 to provide 79. Tosylation of catechol 20 followed
by demethylation using boron tribromide led to 27.
Benzylation followed by removal of the tosyl group under
magnesium conditions provided 28, which was reacted with
3,4-difluoronitrobenzene under basic conditions and reduced
using tin chloride to afford 29. Substitution of this aniline with
3-bromo-1-propanol under basic conditions and subsequent
debenzylation using palladium hydroxide and ammonium
formate delivered compound 30. Reacting intermediate 29
with benzyl chloroformate followed by deprotonation with n-
butyllithium and cyclization with racemic glycidyl butyrate
provided oxazolidinone 31. Mesylation of this intermediate,
displacement of the mesyl group with sodium azide, and
reduction of the benzyl and azide groups under hydrogenation
conditions followed by N-acetylation afforded oxazolidinone
32. Compound 33 was obtained by reacting 20 with 3′,4′-
difluoroacetophenone under basic conditions followed by
demethylation, while further addition of O-methylhydroxyl-
amine provided the oxime 34.
The synthetic routes leading to the derivatives of Table 3 are
described in Scheme 3. The acylation of catechol 35 with acetyl
chloride and aluminum trichloride provided 36, which upon
reduction using zinc and reaction with 1,2-difluorophenyl
afforded 38. Demethylation under standard conditions using
9915
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
Table 1. FabI Inhibition and Antibacterial Activity of Left-Part Derivatives
a
b
c
́
SAU: S. aureus CIP54,146 (CRBIP Pasteur Institute). ECO: E. coli C7 (O18:K1:H7, Robert Debre Hospital). EFA: E. faecalis ATCC29212
d
(CRBIP Pasteur Institute). SPN: S. pneumoniae D39 (Pasteur Institute).
boron tribromide provided compound 81. Compounds 39, 41,
and 43 were obtained in a similar way as intermediate 21 of
Scheme 2. Reacting 39 with sodium nitrite and tetrafluoroboric
acid provided 40 as the methylated precursor of 80. Compound
9916
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Table 1 Compounds
a
Reagents and conditions: (a) KOH, DMF, 110 °C (86%); (b) BBr3, DCM, −78 to −20 °C (38%); (c) isobutylzinc bromide, Pd(PPh3)4, dioxane,
105 °C (20%); (d) (i) trifluoromethyltrimethylsilane, TBAF, THF, rt, (ii) aqueous HCl (88%); (e) TsCl, DMAP, TEA, DCM, 0° to rt (41%); (f)
H2, Pd/C, rt (80%); (g) DAST, DCM, −78 °C to rt (37%); (h) LiHMDS, THF, 0 °C to rt (43%); (i) 4-methylphenylboronic acid, K2CO3,
Pd(PPh3)4, DME, water, 105 °C (quant); (j) MsCl, pyridine, DCM, −40 °C to rt (47%); (k) BBr3, DCM, −78 to −20 °C (95%); (l) MeONa, NaI,
MeOH, rt (quant); (m) imidazole, NaH, DMF, 40 °C (93%).
42 was prepared by treating 41 with boron tribromide followed
by sodium borohydride. Hydrolysis of nitrile 43 using
trifluoroacetic acid and sulfuric acid provided amide 44, while
basic treatment with sodium hydroxide led to carboxylic acid
45, both methylated precursors of compounds 83 and 87.
Carboxylic acid 45 was chlorinated with oxalyl chloride and
coupled to methylamine to afford upon phenolic demethylation
84 in a procedure also used to provide compounds 85 and 86.
The 3-pyridyl derivatives of Table 4 were prepared according
to the synthetic routes illustrated in Scheme 4. Substitution of
catechol 47 with 3-fluoro-2-nitropyridine under basic con-
ditions followed by Suzuki coupling of vinylboronic acid
pinacol ester and reduction of the vinyl moiety to the ethyl
group under hydrogenation conditions provided 49. Demethy-
lation of intermediate 49 using boron tribromide led to
compound 94, while a Balz−Schiemann reaction provided 50
and 51, which were demethylated in the usual manner to afford
compounds 95 and 96.
using boron tribromide. In another pathway, tosylation of 1
with tosyl chloride followed by demethylation with boron
tribromide, benzylation with benzyl bromide, and tosyl
deprotection under basic conditions provided intermediate
92. Substitution with 2-chloropyrimidine under basic con-
ditions and benzyl group hydrogenolysis then converted 92
into the same mixture of 89 and 90.
RESULTS AND DISCUSSION
■
FabI is prone to a large degree of induced fit because of a
flexible loop encompassing its active site,36 which may hamper
the success of docking studies.41,42 We therefore decided to
engage at first in a SAR-led rational optimization mainly based
on enzymatic and antibacterial activity. The decoupling of the
nonspecific biocidal action of triclosan was set as one of our
major objectives, which was monitored by the absence of
antibacterial activity on FabI-independent strains such as
Enterococcus faecalis (for whom FabI is complemented by
FabK) and Streptococcus pneumoniae (harboring FabK only).43
Triclosan optimization was initiated by investigating the
replacement of the 5-chloro group on its catechol by other
functionalities. Indeed, superimposing triclosan from an InhA
The intractable mixture of pyrimidines 89 and 90 was
obtained by two different synthetic routes illustrated in Scheme
5. In one pathway, catechol 1 was substituted with 2-
chloropyrimidine under basic condition and demethylated
9917
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
Table 2. FabI Inhibition and Antibacterial Activity of Aromatic Right-Part Derivatives
a
b
c
́
SAU: S. aureus CIP54,146 (CRBIP Pasteur Institute). ECO: E. coli C7 (O18:K1:H7, Robert Debre Hospital). EFA: E. faecalis ATCC29212
d
(CRBIP Pasteur Institute). SPN: S. pneumoniae D39 (Pasteur Institute).
costructure (1P45)44 to the active site of InhA complexed with
NAD+ and a C16-fatty acid substrate (1BVR)45 reveals that the
5-chloro group is overlaying well with the alkyl chain of the
substrate, with space available in this direction to probe other
side chains (Figure 2). Since this area appears lipophilic, most
of our new substituents were selected to match this
characteristic, although a number of more hydrophilic ones
were also evaluated to investigate a potential unexpected
induced fit. To facilitate rapid synthetic access to the final
products, the novel side chain catechols were coupled to 2,6-
difluoropydirine. This right part, although slightly less active
than triclosan’s dichlorophenyl, was found to maintain enough
potency to allow SAR studies on the left-part derivatives and is
interestingly almost devoid of nonspecific action (Table 1,
compound 52).
Substituting the chloride by a bromide leads to better
enzymatic affinity that however does not translate to an
increase of antibacterial activity in E. coli (53 vs 52). Alkyl and
fluoroalkyl substitutions (54, 55, 4, 56, and 57) are all well
tolerated, in agreement with our hypothesis. Ethyl and 2,2,2-
trifluoroethyl are the best substituents in terms of enzymatic,
antibacterial, and specific activity. As active in S. aureus as the
chloride (52) or n-propyl (4) substitutions, they display higher
FabI specificity because they have no effect in S. pneumoniae at
64 μg/mL. Unsaturated groups such as trifluorovinyl (58) or p-
tolyl (59) lose enzymatic and antibacterial activity, which
probably stems from their steric hindrance. Introduction of
hydrophilic moieties in these side chains (60, 15, 61, 62, 63,
and 64) also leads to activity losses, highlighting the lipophilic
character of the left-part pocket. In view of these results, the
ethyl group was further selected as our standard 4-position
substitution on catechol.
In parallel, substitution on the right-part phenyl group was
explored with a variety of short substituents displaying different
stereoelectronic characteristics followed by the subsequent
derivatization of the most promising ones (Table 2). Replacing
the p-chloro group of triclosan with a methoxy group (65)
without the o-chloro substituent is well tolerated, while moving
the methoxy group to the meta position (66) is less favored. A
meta acetamido substituent (67) also leads to a loss of
enzymatic and antibacterial activity. At the para position, the
electron-donating methoxy group (65) was exchanged by an
electron-withdrawing methylsulfonyl group (68) that also
retains S. aureus activity. A dioxolane moiety was then added
to this methylsulfonyl group (69), which maintains antibacterial
activity, indicating that space is available in this direction.
Moving over to the ortho position, replacement of the o-chloro
group of triclosan with an amino group (70) without the p-
chloro substitution proves as potent on the enzyme. Adding a
methyl moiety para to the aniline (71) or directly on the amino
group (72), however, significantly decreases potency. Fluoride
is also demonstrated to be a good ortho-substituting group,
9918
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Table 2 Compounds
a
Reagents and conditions: (a) 4-methoxyphenylboronic acid, Cu(OAc)2, TEA, molecular sieves, DCM, rt (41%); (b) (i) s-BuLi, THF, 0 °C, (ii) 2-
bromomethyl-1,3-dioxolane, THF, 0 °C to rt (quant); (c) 2-fluoronitrobenzene, KOH, ACN, 80 °C (quant); (d) H2, Pd/C, EtOH, rt (quant); (e)
(i) Ac2O, HCO2H, 60 °C, (ii) 21, THF, rt (quant); (f) BH3.Me2S, THF, 0 °C to reflux (quant); (g) NaNO2, AcOH, conc HCl, ACN, H2SO3,
CuCl2, 0−50 °C (quant); (h) conc NH3, THF, 0 °C to rt (33%); (i) (i) propanaldehyde, MeOH, 30 °C, (ii) NaBH4, 30 °C (18%); (j)
cyclopropanesulfonyl chloride, DCM, pyridine 0 °C to rt (quant); (k) TsCl, K2CO3, NaI, ACN, 70 °C (68%); (l) BBr3, DCM, −78 to −20 °C
(41%); (m) BnBr, K2CO3, NaI, acetone, 40 °C (87%); (n) Mg, MeOH, rt (80%); (o) 3,4-difluoronitrobenzene, K2CO3, ACN, 80 °C (quant); (p)
SnCl2, Et2O, HCl, 0 °C to rt (68%); (q) 3-bromo-1-propanol, K2CO3, DMF, microwave 180 °C (42%); (r) Pd(OH)2, ammonium formate, EtOH,
65 °C (45%); (s) benzyl chloroformate, THF, NaHCO3, 0 °C to rt (quant); (t) (i) n-BuLi, THF, −78 °C, (ii) racemic glycidyl butyrate, −78 °C to
rt (quant); (u) MsCl, TEA, DCM, 0 °C to rt (42%); (v) (i) NaN3, DMF, 75 °C, (ii) H2, Pd/C, EtOAc, rt, (iii) H2, Pd/C, MeOH, rt (54%); (w)
Ac2O, DCM, 0 °C to rt (55%); (x) 3′,4′-difluoroacetophenone, KOH, DMF, 110 °C (quant); (y) BBr3, DCM, −78 °C to −20 °C (5%); (z) O-
methylhydroxylamine, TEA, EtOH, rt (28%).
typically increasing E. coli enzymatic potency by a factor 5 (73
vs 68).
success, as the resulting compound (32) does not show such
broadened spectrum and S. aureus antibacterial activity is even
diminished.
The para position was then revisited with an o-fluoride
group. This led to a variety of well-tolerated substitutions,
among them sulfonamides (74, 75, 76, and 77), amines (30
and 78), inverted sulfonamides (79), carbonyl (33), and oxime
(34). These groups generally conserve submicromolar E. coli
enzymatic potency with the exception of 79 and 34 where the
added steric bulk of the cyclopropyl group or the substituted
oxime might prove deleterious. All these substitutions also
display good antibacterial effect in S. aureus, indicating some
degree of structural allowance at this location. This apparent
permissiveness encouraged us to investigate the p-(acetylami-
nomethyl)-2-oxooxazolidin-3-yl substituent in the hope of
restoring streptococci and enterococci activity by introducing
to our FabI-targeted template an oxazolidinone pharmaco-
phore.46 However, this dual-mode attempt met with little
The most active compounds on S. aureus from Table 2
(triclosan, 75, and 33) also turn out to display a biocidal effect,
as testified by their MIC on E. faecalis or S. pneumoniae.
Interestingly, these compounds are rather lipophilic (respective
clogD according to ACDLabs V12.5: 5.2, 3.6, and 3.7), while
more hydrophilic analogues such as 77 (clogD = 0.7) do not
display an antibacterial effect in FabI independent strains. A
similar observation can be made by comparing 78 (clogD =
3.8) that demonstrates the biocidal effect in S. pneumoniae to its
FabI equipotent but more hydrophilic analogue 30 (clogD =
2.2), which does not.
These results prompted us to take a closer look at the clogD
as a physicochemical parameter potentially linked to the
biocidal character of our compounds. Enzymatic inhibition
9919
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
Table 3. clogD and in Vitro Properties of Selected Derivatives
a
b
c
clogD calculated at pH 7.4 with ACDLabs V12.5. SAU: S. aureus CIP54,146 (CRBIP Pasteur Institute). ECO: E. coli K1 (O18:K1:H7, Robert
d
e
f
Debre
mL.
́
Hospital). EFA: E. faecalis ATCC29212 (CRBIP Pasteur Institute). SPN: S. pneumoniae D39 (Pasteur Institute). Cytotoxicity ED50 in μg/
monitoring was also switched to S. aureus FabI to further
increase the odds of optimization on staphylococci, although it
was later realized that this series generally displays parallel
inhibitory activity on both enzymes. During the course of our
study, it soon became apparent that for our FabI-active
compounds, clogD > 3 correlates to nonspecific biocidal effects
monitored by activity on FabI-independent bacterial strains
such as E. faecalis or S. pneumoniae and eukaryotic cells such as
HepG2 (Table 3). As this biocidal effect increases with the
clogD (triclosan with clogD > 5 displays the highest nonspecific
effect of Table 3), so does the degree of lipophilicity
surrounding the hydrophilic catechol group, therefore render-
ing the compounds amphiphilic. Since the nonspecific action of
triclosan has been associated with membrane-destabilizing
effects,47−49 we can surmise that the biocidal effects of its
analogues are directly connected to their amphiphilic character.
Our optimization rationale therefore focused on compounds
displaying suitable polarity distribution to avoid amphiphilic
properties and to bring their clogD below 3.
Since the p-acetyl group on the aromatic right-part represents
a potent antistaphylococcal starting point (Table 2, compound
33), it was accordingly derivatized with hydrophilic moieties.
The 4-hydroxybutyryl group (Table 3, compound 42)
maintains the antibacterial effect in S. aureus but still displays
some cytotoxicity. Further lowering the clogD by switching to
amido or carboxylic moieties successfully phases out the
nonspecific effect (Table 3, compounds 83, 84, 85, 86, and 87).
9920
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
a
Scheme 3. Synthesis of Table 3 Compounds
a
Reagents and conditions: (a) AcCl, AlCl3, 1,2-dichloroethane, 40 °C (90%); (b) Zn, AcOH, 70 °C (97%); (c) 1,2-difluorobenzene, KOH, DMSO,
130 °C (10%); (d) NaNO2, HBF4, AcOH, rt (15%); (e) BBr3, DCM, −78 to −20 °C (17%); (f) NaBH4, MeOH, −78 to −5 °C (63%); (g)
CF3CO2H, H2SO4 reflux (96%); (h) NaOH, MeOH, reflux (87%); (i) (i) (COCl)2, DCM, 0 °C to rt, (ii) MeNH2, EtOH, rt (quant).
Table 4. FabI Inhibition and Antibacterial Activity of 3-Pyridyl Right-Part Derivatives
a
b
c
clogD calculated at pH 7.4 with ACDLabs V12.5. SAU: S. aureus CIP54,146 (CRBIP Pasteur Institute). ECO: E. coli K1 (O18:K1:H7, Robert
d
e
́
Debre Hospital). EFA: E. faecalis ATCC29212 (CRBIP Pasteur Institute). SPN: S. pneumoniae D39 (Pasteur Institute).
Triclosan is known to interact in its active site via strong
hydrogen-bonding of its catechol hydroxyl with a conserved
tyrosine residue and the 2′-hydroxyl of the cofactor (Figure 2
and reported costructures). Since removing this canonical
interaction invariably leads to a loss of activity (44), a 4-fluoride
group was introduced to the catechol with the objective of
increasing this hydrogen-bonding. The reported log KHA for
phenol is 1.66 vs 1.82 for p-fluorophenol, therefore
demonstrating for the latter a slightly higher hydrogen-bond
donation potential.50 This p-fluorine introduction does not,
however, affect much the pKa of the catechol hydroxyl group.
For instance the calculated pKa according to ACDLabs V12.5
for 83 is 8.93 ( 0.48) vs 8.95 ( 0.35) for its nonfluorinated
catechol analogue, implying that in both cases the catechol
hydroxyl group occurs mostly in its neutral form at pH 7.4. The
addition of a p-fluoro group in our series indeed leads to a small
improvement of enzymatic potency (86 vs 85), although one
cannot rule out that this could also be imparted by the
additional hydrophobic contact of this group with Phe 204 in
the active site (Figure 3). Introduction of another fluoro group
at the ortho-position of the right aryl moiety also improves
enzymatic potency against S. aureus by a factor 5, as already
9921
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
a
Scheme 4. Synthesis of Table 4 Compounds
a
Reagents and conditions: (a) 3-fluoro-2-nitropyridine, KOH, ACN, reflux (95%); (b) vinylboronic acid pinacol ester, Pd(PPh3)4, Cs2CO3, toluene
reflux (76%); (c) H2, Pd/C, MeOH, rt (80%); (d) HBF4, NaNO2, AcOH, 0 °C (40% each).
a
Scheme 5. Regioisomerization Observed during the Deprotection of Electron-Deficient Right-Part Heteroaryls
a
Reagents and conditions: (a) 2-chloropyrimidine, K2CO3, ACN, 100 °C (60%); (b) BBr3, DCM, −78 °C to rt (quant); (c) TsCl, NaI, K2CO3,
ACN, 70 °C (66%); (d) BBr3, DCM, −78 to −20 °C (65%); (e) BnBr, K2CO3, NaI, acetone, 40 °C (64%); (f) KOH, EtOH, water, reflux (86%);
(g) 2-chloropyrimidine, K2CO3, ACN, 100 °C (54%); (h) H2, Pd/C, EtOH, rt (quant).
This optimization strategy led to some potent FabI-specific
compounds, as testified by the amides 86 and 83. Docking
studies revealed 83 with possible binding conformation in the
buried active site of S. aureus FabI delineated by key residues
such as Tyr 147 and 157, Met 160, Ala 95 to Leu 102, Pro 192
to Ile 207, and the NADP+ cofactor all along the bottom of the
groove (Figure 3, for clarity Gly 200, Val 201, and Glu 202,
which close the site, have been omitted). The ethyl group is in
hydrophobic contact with the isopropyl group of Val 201 and
the aromatic groups of Tyr 147 and Phe 204. The electron-rich
catechol interacts through π-stacking with the electron-deficient
pyridinium of NADP+, while the phenolic hydroxyl displays
strong hydrogen-bond interactions with Tyr 157 and the 2′-
hydroxyl of NADP’s ribose. The second aryl group of 83 is in
hydrophobic contact with Met 160. Its fluorine inserts in a
small pocket shaped by a cluster of CH groups of NADP+ in
Figure 2. Overlay of triclosan (gray) with the crystal structure of a
C16-fatty acid substrate (yellow) in InhA with its cofactor and Tyr 158
another hydrophobic contact that might explain the potency
(blue).
improvement between 80 and 81 (Table 3). Simultaneously,
Ser 197 pushes this fluoride group away, which helps direct the
observed on E. coli (Table 3, 81 vs 80 and Table 2, 73 vs 68;
vide infra for rationalization).
aryl group toward a nice chelation of the enzyme by hydrogen-
bonding with the carbonyl and amide hydrogen of Ala 97
9922
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
o-amino group with the carbonyl of Ala95. On the contrary,
pyridone 95 loses all potency, presumably because of electronic
repulsion with the carbonyl of Ala95. Introducing the beneficial
hydrophobic o-fluoro group leads to 3-pyridyl 96, which
displays similar levels of activity as the best right-part aromatic
derivatives.
Among the most active compounds, 83 is a specific, potent
FabI inhibitor. 83 synergizes strongly with the cofactor
NADPH against S. aureus FabI (Figure 4), whereas NADP+
Figure 3. Compound 83 (bold) docked in the active site of S. aureus
FabI using Flo+.51 (A) Protein is displayed as orange wire with
detailed key residues and NADP+ in stick style. (B) Protein and
NADP+ are displayed as electrostatic surfaces. For clarity purposes, Val
201 is in stick style.
Figure 4. (a) Influence of NADPH on the IC50 of 83 against FabI of S.
aureus (NADP+ = 0 μM). (b) Influence of NADP+ on the IC50 of 83
against FabI of S. aureus (NADPH = 40 μM).
(Figure 3). Introducing a methyl group to the amide would
result in loss of the hydrogen-bonding to Ala 97 due to the
preferential trans rotamer, therefore providing a possible
explanation for the weaker enzymatic activity of 84 vs 83
(Table 3).
does not up to 1 mM. The postulated binding mode of this
compound involving a π-stacking with the electron-deficient
pyridinium of NADP+ is therefore challenged by this
observation and needs further explanation. As shown in Figure
5, a slow onset of inhibition is observed with 83. Interestingly,
Replacement of the right aryl ring with heteroaryl groups can
constitute another way of lowering the clogD of our
compounds below the threshold of 3. However, it was observed
that for a variety of heterocycles (among which are 2-pyrimidyl,
2-pyridyl, 2-quinolyl) this was not a viable option because of
the self-isomerization encountered with these products. For
instance, independent of the final deprotection procedure,
catechols 89 and 90 are isolated in a ratio close to 50:50
(Scheme 5). Moreover, their separation by HPLC followed by
concentration also leads to reisomerized mixtures. This intrinsic
stability issue could be explained by a Smiles rearrangement at
the electron-deficient heteroaryl carbon bearing the catechol
group.
To avoid this regioisomerization problem, 3-pyridyl hetero-
cycles with less electron-withdrawing character at the carbon
bearing the catechol were synthesized. Satisfyingly, the resulting
compounds turn out to be stable and display clogD < 3 without
any sign of biocidal effect (Table 4). Derivative 94 retains FabI
potency, which might be explained by hydrogen-bonding of its
Figure 5. Influence of the preincubation time on S. aureus FabI
inhibition kinetics of 83.
9923
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
slow inhibition kinetics is not significantly modified by the
preincubation time (0−2 h) between FabI, 83, and NADPH.
These three partners are therefore not sufficient for the onset of
full FabI inhibition: the addition of the fatty acid t-o-NAC-
thioester (trans-2-octenoyl N-acetylcysteamine thioester) sub-
strate, which allows enzyme catalysis to start, was the triggering
event here. As the addition of t-o-NAC-thioester leads to the
oxidation of NADPH to NADP+, it is tempting to speculate the
next events: reduced t-o-NAC-thioester could leave the active
site more rapidly than NADP+ and be replaced by 83, thus
building a more stable ternary complex in agreement with the
postulated π-stacking mode of binding of our inhibitor. The fact
that NADP+ itself does not synergize at all with 83 would imply
that FabI’s native conformation is a very poor binder of
NADP+, since this oxidized cofactor could actually be more
efficiently trapped in a catalytically active conformation. At this
stage, additional investigations would be required to understand
and validate the nature of this inhibition mechanism.
without significant clinical findings on macroscopic or micro-
scopic examinations. In dogs, the NOAEL was decreased at 100
mg kg−1 day−1 because of a significant body weight loss in only
one female. In safety pharmacology studies, 83 administered in
rats as single intravenous injections up to 200 mg/kg did not
display any relevant clinical signs related to the compound, with
no reported potential effects on the central nervous system or
on respiratory functions. Single intravenous infusions up to 100
mg/kg in beagle dogs were also devoid of observable clinical
signs related to the compound, with no reported potential
effects on cardiovascular functions and no QT prolongation.
Overall, the results from these toxicology and safety
pharmacology studies show no findings that should prevent
83 from progressing to human studies.55 In pharmacokinetic
studies in rats, 83 displayed a high plasma clearance of 6.4
L h−1 kg−1, as could be expected from the metabolic liability of
the phenolic moiety.56 Nevertheless high concentrations of
compound were achieved in all tested tissues (Table 7).
The in vivo efficacy of 83 was then evaluated using murine
infection models. In a systemic model of lethal infection of
mice with an intraperitoneal challenge of MRSA, the ED50 of
the compound was determined as the single subcutaneous dose
administered 5 min after the infection that provided 50%
survival of treated animals at 48 h postinfection (Table 8).
These results demonstrate a good protective action against a
variety of multiresistant S. aureus. Additional in vitro and in vivo
data about 83 have already been reported,53 which helps further
validate FabI as a suitable antibiotic target on S. aureus. In light
of these promising results, as well as its good preclinical safety
dossier and low cost-of-goods, 83 has been selected for clinical
development.
When compared to linezolid, a drug currently approved for
nosocomial pneumonia as well as complicated skin and skin
structure infections caused by S. aureus (methicillin-susceptible
and -resistant strains), 83 displays highly potent in vitro
antistaphylococcal activities (Table 5), with a low frequency of
Table 5. MIC90 (μg/mL) of 83 and Linezolid against
Staphylococci
community-
NARSA
a
b
MRSA
f
c
d
e
acquired MRSA
MSSA
f
MRSE
f
MRSH
f
f
compd
83
linezolid
(30)
(48)
(40)
(30)
(30)
≤0.03
2
0.12
0.06
2
2
1
0.5
1
≥16
a
b
To help establish human doses, a PK/PD investigation of 83
was performed against S. aureus in an in vitro one compartment
pharmacodynamic model. Regimens of 600, 900, and 1200 mg
q6h or q8h with a prolonged infusion of 2 h were found to
achieve the greatest efficacy against MRSA USA300, with rapid
and sustainable killing activity. In this study, 83 demonstrated a
dose-dependent response using fAUC/MIC and %T/MIC as
the predictive pharmacodynamic parameters.57 No relevant
clinical or laboratory findings related to 83 were reported in
healthy human volunteers at these doses, which supports
further clinical development.58,59
Methicillin resistant S. aureus. Network on antimicrobial resistance
in S. aureus. Methicillin susceptible S. aureus. Methicillin resistant S.
epidermidis. Methicillin resistant S. hemolyticus. Number of isolates.
c
d
e
f
spontaneous resistance, between 2.5 × 10−9 and 7 × 10−9 at 4
times the MIC.52 In addition to its strong potency against
staphylococci, this compound displays good potential against
some FabI-dependent Gram-negative bacteria (Table 6).
Promising activities have also been reported on Chlamydophila
pneumoniae, Helicobacter pylori, and Acinetobacter baumannii.53
Preclinical studies on 83 showed that it demonstrates
moderate thermodynamic aqueous solubility: around 0.02
mg/mL at pH 7.4. Its solubility could be increased to 11.3
mg/mL in an aqueous solution of 20% hydroxypropyl-β-
cyclodextrin (HPBCD) and 1−5% glucose. This compound
displays high in vitro affinity for human serum albumin (95%)
and weak inhibition of cytochromes P450.54 Toxicology
evaluation of 83 was performed following single intravenous
dosing in rats and mice and twice daily intravenous dosing in
rats and dogs for 14 days. Single dose toxicity studies in mice
and rats showed that the maximum tolerated dose (MTD) was
greater than 200 mg/kg. In the rat 2-week study, the no
observed adverse effect level (NOAEL) was 200 mg kg−1 day−1
CONCLUSION
■
Starting from triclosan as a lead, we have described some
elements of our optimization program to decouple its specific
FabI effect from its nonspecific cytotoxic component. This
strategy was successfully implemented to deliver highly specific
and potent FabI inhibitors, among which is 83, a novel
antibacterial compound active against resistant staphylococci
and some clinically relevant Gram negative bacteria. We also
report murine in vivo efficacy data, which help further validate
FabI as a suitable antibiotic target on S. aureus. In hopes of
exploitation of its untapped mode of action, 83 is currently
Table 6. MIC90 (μg/mL) of 83 and Levofloxacin against Gram-Negative Bacteria
Neisseria gonorrheae
Neisseria meningitidis
Haemophilus influenzae
Moraxella catarrhalis
Escherichia coli
Proteus mirabilis
a
a
a
a
a
a
compd
83
levofloxacin
(10)
(10)
(30)
(20)
(30)
(10)
0.25
0.5
0.25
0.5
2
1
2
≤0.008
0.03
0.03
≥16
0.06
a
Number of isolates.
9924
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
a
Table 7. Rat Pharmacokinetic Data for Compound 83
plasma
liver
kidney
brain
heart
lung
b
Cmax (ng/g)
30436
51700
105067
9.2
63000
20262
8.5
37800
27635
4.2
42100
23920
5
28900
18393
5.9
AUC(0−∞) (ng·h/mL)
13936
4.9
c
T1/2 (h)
a
b
c
Female Sprague−Dawley rats dosed iv, 2 × 50 mg/kg. ng/mL. Terminal elimination half-life.
520 nm) was measured immediately after t-o-NAC addition (T0), and
approximately 50 min later (T50) by a Fluostar Optima (BMG) to
achieve 30% of NADPH conversion. Enzyme activity was calculated
by first subtracting the T0 signal from the T50 and then subtracting
background signal (FabI = 0). Percentages of inhibition were
calculated against untreated samples (inhibitor = 0), and IC50 values
were fitted to a classical Langmuir equilibrium model using XLFIT
(IDBS).
Antibacterial Activity Assay. Whole-cell antimicrobial activity
was determined by broth microdilution method in microtiter plates
according to CLSI guidelines. The compound was assayed in serial 2-
fold dilutions ranging from 0.06 to 64 μg/mL. Test organisms were
selected from the following laboratory strains: Staphylococcus aureus
Table 8. In Vivo Activity of 83 in a Murine Systemic
Infection Model
a
83 MIC
S. aureus strain (μg/mL)
83 mean ED50
vancomycin mean
ED50 (mg/kg) (CI)
b
b
(mg/kg) (CI)
MRSA NRS382
(USA100)
0.06
0.06
0.06
19.3 (12.5−28.5)
6.0 (1.8−12.4)
9.4 (9.4−9.4)
5.0 (1.9−9.8)
MRSA NRS384
(USA300)
45.1 (33.4−50.6)
MRSA NRS385
(USA500)
28.3 (21.0−38.2)
a
b
Female Swiss mice (6 per group). Numbers in parentheses are 95%
confidence ranges.
́
CIP54.146, Escherichia coli K1 Robert Debre Hospital, Paris, France, E.
faecalis ATCC29212, S. pneumoniae D39 Pasteur Institute, Paris,
France. Bacteria were grown in cation-adjusted Mueller−Hinton vroth
(ca-MHB) using an inoculum of 5 × 105 CFU/mL incubated at 35 °C
for 18 h unless otherwise stated. For S. pneumoniae, ca-MHB was
supplemented with 2.5% lysed horse blood, and growth was performed
in 5% CO2. The minimum inhibitory concentration (MIC) was
defined as the lowest concentration of compound at which no visible
bacterial growth was observed.
Docking Method. Compound 83 was docked in the S. aureus FabI
active site using the MCDOCK module with standard parameters
(QXP Flo+)51 and the deposited structure 3GR6. Key residues Tyr
147 and 157, Met 160, Ala 95 to Leu 102, Pro 192 to Ile 207 as well as
the NADP+ cofactor were allowed to be flexible while the rest of the
active site was kept rigid.
General Experimental. All reactions were carried out under an
inert (nitrogen or argon) atmosphere unless indicated otherwise.
Starting materials, reagents, and solvents were obtained from
commercial sources and were used without further purification unless
otherwise specified. Celite is a filter aid composed of diatomaceous
silica and is a registered trademark of Celite Corporation. Analtech
silica gel GF and E. Merck silica gel 60 F-254 thin layer plates were
used for thin layer chromatography. Flash chromatography was carried
out on Flashsmart Pack cartridge irregular silica 40−60 μm or
spherical silica 20−40 μm. Preparative thin layer chromatography was
carried out on Analtech silica gel GF 1000 μm, 20 cm × 20 cm. Yields
refer to purified products and are not optimized. All new compounds
gave satisfactory analytical data. 1H NMR spectra were recorded at 300
undergoing clinical development for the treatment of severe
bacterial infections in human.
EXPERIMENTAL SECTION
■
FabI E. coli Enzyme Assay. Compound inhibitory activity of FabI
enzyme was measured in vitro by IC50 determination using a
fluorescent based assay. The protein FabI from E. coli was prepared
and purified using standard methods for recombinant protein
expression after cloning of the gene in a prokaryotic expression
vector. The biochemical activity of the FabI enzyme was assessed using
the following method. The assay buffer “AB” contained 50 mM Hepes,
pH 7.5, 100 μM dithiothreitol, and 0.006% Triton-X100. The
following components were added in a black polystyrene Costar
plate to a final volume of 55 μL:1.5 μL DMSO, or inhibitor dissolved
in DMSO and 53.5 μL of a FabI/NADH/NAD+ mixture in AB. After
90 min of preincubation at room temperature, the reaction was
initiated by addition of 5 μL of crotonoyl-CoA to a final volume of 60
μL. This reaction mixture was then composed of 40 nM FabI
(produced in house from E. coli, C-terminal 6-His tagged), 20 μM
NADH (Biochemika), 10 μM NAD+ (Biochemika), 50 μM crotonoyl-
CoA (Biochemika), and compound at defined concentration.
Fluorescence intensity of NADH (λex = 360 nm, λem = 520 nm)
was measured immediately after crotonoyl-CoA addition, and 2 h later
by a Fluostar Optima (BMG). Enzyme activity is proportional to the
signal decrease from which inhibition percentages are derived. For IC50
determinations, the inhibitor was tested at 6−10 different concen-
trations, and the related inhibitions were fitted to a classical Langmuir
equilibrium model using XLFIT (IDBS).
or 400 MHz on a Bruker instrument, and chemical shifts are reported
̈
in parts per million (δ) downfield from the internal standard
tetramethylsilane. Mass spectra were obtained using electrospray
(ESI) ionization techniques on an Agilent 1100 series LCMS
instrument. HPLC (analytical and preparative) were performed on
an Agilent 1100 HPLC instrument with diode array detection.
Preparative HPLC was performed at 0.7 mL/min on a Thermo
Electron, Hypersil BDS C-18 column (250 mm × 4.6 mm, 5 μm)
using a gradient of acetonitrile and water with 0.1% TFA (50% in
acetonitrile to 100% and then back to 50%). The tested compounds
were determined to be >95% pure via HPLC.
FabI S. aureus Enzyme Assay. Compound inhibitory activity of
FabI enzyme was measured in vitro by IC50 determination using a
fluorescence based assay. The protein FabI from S. aureus was
prepared and purified using standard methods for recombinant protein
expression after cloning of the gene in a prokaryotic expression vector.
The biochemical activity of the FabI enzyme was assessed using the
following method. The assay buffer “AB” contained 50 mM ADA (N-
(2-acetamido)iminodiacetic acid monosodium salt), pH 6.5, 1 mM
dithiothreitol, 0.006% Triton-X100, and 50 mM NaCl. The following
components were added in a white polystyrene Costar plate (ref 3912)
to a final volume of 55.5 μL:1.5 μL DMSO or inhibitor dissolved in
DMSO and 54 μL of a FabI/NADPH/NADP+ mixture in AB. After 60
min of preincubation at room temperature, the reaction was started by
addition of 5 μL of trans-2-octenoyl N-acetylcysteamine thioester (t-o-
NAC) to a final volume of 60.5 μL. This reaction mixture was then
composed of 2 nM FabI, 40 μM NADPH (Sigma, N7505), 10 μM
NADP+ (Sigma, N5755), 100 μM t-o-NAC, and compound at defined
General Procedure for the Synthesis of (4-Ethyl-5-fluoro-2-
hydroxyphenoxy)-Based Compounds of Table 3 As Illustrated
by Synthesis of 4-(4-Ethyl-5-fluoro-2-hydroxyphenoxy)-3-fluo-
robenzamide (83). 1-(2-Fluoro-4-hydroxy-5-methoxyphenyl)-
ethanone (36). To a suspension of aluminum chloride (1.17g, 8.79
mmol) in 1,2-dichloroethane (2 mL) was added acetyl chloride (0.55g,
7.03 mmol). After the mixture was stirred for 10 min, a solution of 5-
fluoro-2-methoxyphenol 35 (0.50g, 3.52 mmol) in 1,2-dichloroethane
(2 mL) was added dropwise. The reaction mixture was stirred
concentration. Fluorescence intensity of NADPH (λex = 360 nm, λem
=
9925
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
overnight at 40 °C. The mixture was then poured on ice−water and
extracted with diethyl ether. The organic phases were combined,
washed with brine, dried over sodium sulfate, and concentrated to
afford 582 mg (90%) of the title compound as an off-white solid. MS
(ES) m/e 185 (M + H)+ . TLC: eluent cyclohexane/EtOAc, 7/3, Rf =
0.23.
(d, 1H, J = 9.8 Hz), 6.87 (d, 1H, J = 7.9 Hz), 6.78 (t, 1H, J = 8.2 Hz),
2.56 (q, 2H, J = 7.4 Hz), 1.17 (t, 3H, J = 7.3 Hz).
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthetic procedures and spectroscopic details for all final
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
4-Ethyl-5-fluoroguaiacol (37). A solution of 36 (18.0 g, 97.7
mmol) in glacial acetic acid (800 mL) is stirred at 70 °C before adding
added zinc dust (63.9 g, 977 mmol). The resulting gray heterogeneous
mixture is then heated at reflux and stirred overnight using a
mechanical stirrer. After this period, zinc has aggregated and
AUTHOR INFORMATION
Corresponding Author
*Phone: +33-157140522. Fax: +33-157140524. E-mail: vincent.
gerusz@mutabilis.fr.
■
1
conversion rate reaches 90% according to the H NMR analysis of a
crude aliquot. Therefore, the zinc metal is removed by filtration on a
fritted glass, and fresh zinc dust (6.4 g, 98 mmol) is added to the
resulting limpid yellow filtrate. The solution is heated at reflux
overnight until completion of the reaction. The solution is filtered on a
fritted glass and basified until pH 11−12 is reached with a saturated
aqueous solution of potassium carbonate (1.5 L) and with additional
solid potassium carbonate if needed. The resulting aqueous layer is
then extracted with ethyl acetate (1.0 L), dried over sodium sulfate or
by azeotropic toluene distillation, filtered, and concentrated under
vacuum to afford the pure title compound (16.1 g, 94.7 mmol, 97%) as
a pale yellow oil. The title compound is a volatile product and should
be kept in the refrigerator under argon away from light (darkens with
Present Addresses
§GlaxoSmithKline, 25-27 Avenue du Queb
́
ec, 91951 Les Ulis,
France.
∥Galapagos, 102 Avenue Gaston Roussel, 93230 Romainville,
France.
⊥ESE Conseil, 66 Boulevard Senard, 92210 Saint Cloud,
France.
#Alderys, 86 Rue de Paris, 91400 Orsay, France.
∞FAB Pharma, 11 Avenue Myron Herrick, 75008 Paris, France.
1
oxygen and/or UV exposure). No or weak response in MS. H NMR
(DMSO), δ (ppm): 9.20 (bs, 1H), 6.78 (d, J = 3.6 Hz, 1H), 6.54 (d, J
= 9 Hz, 1H), 3.73 (s, 3H), 2.50 (q, J = 7.2 Hz, 2H), 1.12 (t, J = 7.4 Hz,
3H).
Notes
The authors declare no competing financial interest.
4-(4-Ethyl-5-fluoro-2-methoxyphenoxy)-3-fluorobenzonitrile
(43). To a solution of 37 (8 g, 47 mmol) and 3,4-difluorobenzonitrile
(6.53 g, 47 mmol) in 80 mL of anhydrous acetonitrile is added
potassium hydroxide (3.15 g, 56.4 mmol). The reaction mixture under
argon atmosphere is stirred under reflux for 16 h. Concentration,
addition of a saturated aqueous solution of ammonium chloride (100
mL), extraction with ethyl acetate (2 × 25 mL), reunification of the
organic phases, brine wash, drying over sodium sulfate, and final
concentration affords 12.95 g (95%) of the title compound as a brown
solid, which was used as such for the next step. MS (ES) m/e 290 (M
+ H)+. TLC: eluent cyclohexane/EtOAc, 7/3, Rf = 0.74.
4-(4-Ethyl-5-fluoro-2-methoxyphenoxy)-3-fluorobenzamide (44).
To 43 (12.95 g, 7.05 mmol) are added trifluoroacetic acid (52 mL)
and concentrated sulfuric acid (13 mL). After 1 h and 30 min under
reflux the reaction mixture is cooled to room temperature and then
poured into ice−water (400 mL). Dichloromethane extraction (100
mL, then 2 × 25 mL), reunification of the organic phases, saturated
aqueous sodium hydrogenocarbonate wash (250 mL, pH 8−8.5),
drying over sodium sulfate, and final concentration affords 13.31 g
(96%) of the title compound as an off-white solid. MS (ES) m/e 294
(M + H)+. TLC: eluent dichloromethane/methanol, 9/1, Rf = 0.3. 1H
NMR (DMSO) δ (ppm): 7.96 (bs, 1H), 7.82 (d, J = 12.1 Hz, 1H),
7.64 (d, J = 8.3 Hz, 1H), 7.41 (bs, 1H), 7.13 (d, J = 7.2 Hz, 1H), 7.04
(d, J = 10.0 Hz, 1H), 6.80 (t, J = 8.5 Hz, 1H), 3.73 (s, 3H), 2.64 (q, J =
7.6 Hz, 2H), 1.22 (t, J = 7.5 Hz, 3H).
4-(4-Ethyl-5-fluoro-2-hydroxyphenoxy)-3-fluorobenzamide (83).
To 44 (13.31 g, 4.59 mmol) in 130 mL of dichloromethane under
argon at −78 °C under intense stirring is added, over 15−20 min,
boron tribromide (130 mL at 1 M in dichloromethane). The reaction
mixture is warmed to room temperature under stirring and after 3 h is
cooled back to −20 °C for quenching with a saturated aqueous
solution of ammonium chloride (100 mL). Partial concentration is
performed to remove 170 mL of dichloromethane. An amount of 100
mL of ethyl acetate is added. Extraction of the aqueous phase (2 × 25
mL of ethyl acetate), reunification of the organic phases, aqueous
sodium hydrogenocarbonate (200 mL at 1 N) wash, drying over
sodium sulfate, and final concentration afford the crude material which
is purified on silica gel (gradient dichloromethane/methanol, 100/0 →
95/5) to afford the title compound, 8.75g (68%). MS (ES) m/e 294
(M + H)+. TLC: eluent dichloromethane/methanol, 20/1, Rf = 0,4. 1H
NMR (DMSO) δ (ppm): 9.59 (s, 1H, OH), 7.95 (bs, 1H, NH), 7.80
(d, 1H, J = 12.2 Hz), 7.63 (d, 1H, J = 8.3 Hz), 7.40 (bs, 1H, NH), 6.96
ACKNOWLEDGMENTS
The authors gratefully thank Drs. Andre
Fischer for enlightening discussions on these topics, as well as
́
the Robert Debre Hospital and the Pasteur Institute for
providing S. pneumoniae D39 and E. coli strains. This work has
been partly supported by Grants I06-222/R and I09-1739/R
from the Medicen Paris Region Competitiveness Cluster and
the Seine-Saint-Denis County.
■
́
Bryskier and Stefan
REFERENCES
■
(1) Furtado, G. H.; Nicolau, D. P. Overview perspective of bacterial
resistance. Expert Opin. Ther. Pat. 2010, 20, 1273−1276.
(2) Boucher, H. W.; Talbot, G. H.; Bradley, J. S.; Edwards, J. E.;
Gilbert, D.; Rice, L. B.; Scheld, M.; Spellberg, B.; Bartlett, J. Bad bugs,
no drugs: no ESKAPE! An update from the Infectious Diseases Society
of America. Clin. Infect. Dis. 2009, 48, 1−12.
(3) Chambers, H. F.; Deleo, F. R. Waves of resistance: Staphylococcus
aureus in the antibiotic era. Nat. Rev. Microbiol. 2009, 7, 629−641.
(4) Dulon, M.; Haamann, F.; Peters, C.; Schablon, A.; Nienhaus, A.
MRSA prevalence in European healthcare settings: a review. BMC
Infect. Dis. 2011, 11, 138.
(5) Watkins, R. R.; David, M. Z.; Salata, R. A. Current concepts on
the virulence mechanisms of meticillin-resistant Staphylococcus aureus.
J. Med. Microbiol. 2012, 61, 1179−1193.
(6) Zhang, Y. M.; White, S. W.; Rock, C. O. Inhibiting bacterial fatty
acid synthesis. J. Biol. Chem. 2006, 281, 17541−17544.
(7) Wright, H. T.; Reynolds, K. A. Antibacterial targets in fatty acid
biosynthesis. Curr. Opin. Microbiol. 2007, 10, 447−453.
(8) Brinster, S.; Lamberet, G.; Staels, B.; Trieu-Cuot, P.; Gruss, A.;
Poyart, C. Type II fatty acid synthesis is not a suitable antibiotic target
for Gram-positive pathogens. Nature 2009, 458, 83−86.
(9) Brinster, S.; Lamberet, G.; Staels, B.; Trieu-Cuot, P.; Gruss, A.;
Poyart, C. Nature 2010, 463, E4.
(10) Balemans, W.; Lounis, N.; Gilissen, R.; Guillemont, J.; Simmen,
K.; Andries, K.; Koul, A. Nature 2010, 463, E3 ; discussion E4.
(11) Parsons, J. B.; Rock, C. O. Is bacterial fatty acid synthesis a valid
target for antibacterial drug discovery? Curr. Opin. Microbiol. 2011, 14,
544−549.
(12) Parsons, J. B.; Frank, M. W.; Subramanian, C.; Saenkham, P.;
Rock, C. O. Metabolic basis for the differential susceptibility of Gram-
9926
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
positive pathogens to fatty acid synthesis inhibitors. Proc. Natl. Acad.
Sci. U.S.A. 2011, 108, 15378−15383.
Slayden, R. A. Substituted diphenyl ethers as a broad-spectrum
platform for the development of chemotherapeutics for the treatment
of tularaemia. J. Antimicrob. Chemother. 2009, 64, 1052−1061.
(13) Gerusz, V. Recent advances in the inhibition of bacterial fatty
acid biosynthesis. Annu. Rep. Med. Chem. 2010, 45, 295−311.
(14) Heath, R. J.; Li, J.; Roland, G. E.; Rock, C. O. Inhibition of the
Staphylococcus aureus NADPH-dependent enoyl-acyl carrier protein
reductase by triclosan and hexachlorophene. J. Biol. Chem. 2000, 275,
4654−4659.
̀
(30) Freundlich, J. S.; Wang, F.; Vilcheze, C.; Gulten, G.; Langley, R.;
Schiehser, G. A.; Jacobus, D. P.; Jacobs, W. R., Jr.; Sacchettini, J. C.
Triclosan derivatives: towards potent inhibitors of drug-sensitive and
drug-resistant Mycobacterium tuberculosis. ChemMedChem 2009, 4,
241−248.
(15) Heath, R. J.; Rock, C. O. Enoyl-acyl carrier protein reductase
(fabI) plays a determinant role in completing cycles of fatty acid
elongation in Escherichia coli. J. Biol. Chem. 1995, 270, 26538−26542.
(16) Banerjee, A.; Dubnau, E.; Quemard, A.; Balasubramanian, V.;
Um, K. S.; Wilson, T.; Collins, D.; de Lisle, G.; Jacobs, W. R., Jr. InhA,
a gene encoding a target for isoniazid and ethionamide in
Mycobacterium tuberculosis. Science 1994, 263, 227−230.
(31) Kaplan, N.; Flanner, H.; Hafkin, B. Correlation of AFN-1252
Phase 0 Microdosing and Phase 1 Pharmacokinetics. Presented at the
49th Annual ICAAC Meeting, 2009, San Francisco, CA; F1-2006.
Public disclosure from Affinium Pharmaceutical’s Web site www.afnm.
com.
(32) Ro, S.; Son, K. H.; Kim, Y. E.; Chang, H. J.; Park, S. B.; Choi, J.
R.; Cho, J. M. CG400549: Candidate for the Treatment of MRSA
Infection. Presented at the 49th Annual ICAAC Meeting, 2009, San
Francisco, CA; F1-2007. Public disclosure from CrystalGenomics’
Web site www.cgxinc.com.
́
(17) Quemard, A.; Sacchettini, J. C.; Dessen, A.; Vilcheze, C.;
Bittman, R.; Jacobs, W. R., Jr.; Blanchard, J. S. Enzymatic
characterization of the target for isoniazid in Mycobacterium tuber-
culosis. Biochemistry 1995, 34, 8235−8241.
(18) Lu, H.; Tonge, P. J. Inhibitors of FabI, an enzyme drug target in
the bacterial fatty acid biosynthesis pathway. Acc. Chem. Res. 2008, 41,
11−20.
(19) Baldock, C.; Rafferty, J. B.; Sedelnikova, S. E.; Baker, P. J.;
Stuitje, A. R.; Slabas, A. R.; Hawkes, T. R.; Rice, D. W. A mechanism of
drug action revealed by structural studies of enoyl reductase. Science
1996, 274, 2107−2110.
(33) Schweizer, H. P. Triclosan: a widely used biocide and its link to
antibiotics. FEMS Microbiol. Lett. 2001, 7, 1−7.
(34) Ward, W. H.; Holdgate, G. A.; Rowsell, S.; McLean, E. G.;
Pauptit, R. A.; Clayton, E.; Nichols, W. W.; Colls, J. G.; Minshull, C.
A.; Jude, D. A.; Mistry, A.; Timms, D.; Camble, R.; Hales, N. J.;
Britton, C. J.; Taylor, I. W. Kinetic and structural characteristics of the
inhibition of enoyl (acyl carrier protein) reductase by triclosan.
Biochemistry 1999, 38, 12514−12525.
(20) Kitagawa, H.; Kumura, K.; Takahata, S.; Iida, M.; Atsumi, K. 4-
Pyridone derivatives as new inhibitors of bacterial enoyl-ACP
reductase FabI. Bioorg. Med. Chem. 2007, 15, 1106−1116.
(21) Seefeld, M. A.; Miller, W. H.; Newlander, K. A.; Burgess, W. J.;
DeWolf, W. E., Jr.; Elkins, P. A.; Head, M. S.; Jakas, D. R.; Janson, C.
A.; Keller, P. M.; Manley, P. J.; Moore, T. D.; Payne, D. J.; Pearson, S.;
Polizzi, B. J.; Qiu, X.; Rittenhouse, S. F.; Uzinskas, I. N.; Wallis, N. G.;
Huffman, W. F. Indole naphthyridinones as inhibitors of bacterial
enoyl-ACP reductases FabI and FabK. J. Med. Chem. 2003, 46, 1627−
1635.
(35) Lyman, F. L.; Furia, T. Toxicology of 2,4,4′-trichloro-2′-
hydroxy-diphenyl ether. IMS, Ind. Med. Surg. 1969, 38, 64−71.
(36) Stewart, M. J.; Parikh, S.; Xiao, G.; Tonge, P. J.; Kisker, C.
Structural basis and mechanism of enoyl reductase inhibition by
triclosan. J. Mol. Biol. 1999, 290, 859−865.
(37) Lu, H.; England, K.; am Ende, C.; Truglio, J. J.; Luckner, S.;
Reddy, B. G.; Marlenee, N. L.; Knudson, S. E.; Knudson, D. L.;
Bowen, R. A.; Kisker, C.; Slayden, R. A.; Tonge, P. J. Slow-onset
inhibition of the FabI enoyl reductase from Francisella tularensis:
residence time and in vivo activity. ACS Chem. Biol. 2009, 4, 221−231.
(38) Hopkins, A. L.; Groom, C. R.; Alex, A. Ligand efficiency: a
useful metric for lead selection. Drug Discovery Today 2004, 9, 430−
431.
(22) McMurry, L. M.; Oethinger, M.; Levy, S. B. Triclosan targets
lipid synthesis. Nature 1998, 394, 531−532.
(23) Sivaraman, S.; Sullivan, T. J.; Johnson, F; Novichenok, P.; Cui,
G.; Simmerling, C.; Tonge, P. J. Inhibition of the bacterial enoyl
reductase FabI by triclosan: a structure−reactivity analysis of FabI
inhibition by triclosan analogues. J. Med. Chem. 2004, 47, 509−518.
(24) Sullivan, T. J.; Truglio, J. J.; Boyne, M. E.; Novichenok, P.;
Zhang, X.; Stratton, C. F.; Li, H. J.; Kaur, T.; Amin, A.; Johnson, F.;
Slayden, R. A.; Kisker, C.; Tonge, P. J. High affinity InhA inhibitors
with activity against drug-resistant strains of Mycobacterium tuberculosis.
ACS Chem. Biol. 2006, 1, 43−53.
(39) Levy, C. W.; Roujeinikova, A.; Sedelnikova, S.; Baker, P. J.;
Stuitje, A. R.; Slabas, A. R.; Rice, D. W.; Rafferty, J. B. Molecular basis
of triclosan activity. Nature 1999, 398, 383−384.
(40) Priyadarshi, A.; Kim, E. E.; Hwang, K. Y. Structural insights into
Staphylococcus aureus enoyl-ACP reductase (FabI), in complex with
NADP and triclosan. Proteins 2010, 78, 480−486.
(41) Irwin, J. J. Community benchmarks for virtual screening. J.
Comput.-Aided Mol. Des. 2008, 22, 193−199.
(25) Chhibber, M.; Kumar, G.; Parasuraman, P.; Ramya, T. N.;
Surolia, N.; Surolia, A. Novel diphenyl ethers: design, docking studies,
synthesis and inhibition of enoyl ACP reductase of Plasmodium
falciparum and Escherichia coli. Bioorg. Med. Chem. 2006, 14, 8086−
8098.
(42) Hevener, K. E.; Mehboob, S.; Su, P. C.; Truong, K.; Boci, T.;
Deng, J.; Ghassemi, M.; Cook, J. L.; Johnson, M. E. Discovery of a
novel and potent class of F. tularensis enoyl-reductase (FabI) inhibitors
by molecular shape and electrostatic matching. J. Med. Chem. 2012, 55,
268−279.
(26) Park, H. S.; Yoon, Y. M.; Jung, S. J.; Kim, C. M.; Kim, J. M.;
Kwak, J. H. Antistaphylococcal activities of CG400549, a new bacterial
enoyl-acyl carrier protein reductase (FabI) inhibitor. J. Antimicrob.
Chemother. 2007, 60, 568−574.
(27) Tipparaju, S. K.; Mulhearn, D. C.; Klein, G. M.; Chen, Y.;
Tapadar, S.; Bishop, M. H.; Yang, S.; Chen, J.; Ghassemi, M.;
Santarsiero, B. D.; Cook, J. L.; Johlfs, M.; Mesecar, A. D.; Johnson, M.
E.; Kozikowski, A. P. Design and synthesis of aryl ether inhibitors of
the Bacillus anthracis enoyl-ACP reductase. ChemMedChem 2008, 3,
1250−1268.
(43) Payne, D. J.; Warren, P. V.; Holmes, D. J.; Ji, Y.; Lonsdale, J. T.
Bacterial fatty-acid biosynthesis: a genomics-driven target for
antibacterial drug discovery. Drug Discovery Today 2001, 6, 537−544.
(44) Kuo, M. R.; Morbidoni, H. R.; Alland, D.; Sneddon, S. F.;
Gourlie, B. B.; Staveski, M. M.; Leonard, M.; Gregory, J. S.; Janjigian,
A. D.; Yee, C.; Musser, J. M.; Kreiswirth, B.; Iwamoto, H.; Perozzo, R.;
Jacobs, W. R.; Sacchettini, J. C.; Fidock, D. A. Targeting tuberculosis
and malaria through inhibition of enoyl reductase: compound activity
and structural data. J. Biol. Chem. 2003, 278, 20851−20859.
̀
(45) Rozwarski, D. A.; Vilcheze, C.; Sugantino, M.; Bittman, R.;
(28) Tipparaju, S. K.; Joyasawal, S.; Forrester, S.; Mulhearn, D. C.;
Pegan, S.; Johnson, M. E.; Mesecar, A. D.; Kozikowski, A. P. Design
and synthesis of 2-pyridones as novel inhibitors of the Bacillus
anthracis enoyl-ACP reductase. Bioorg. Med. Chem. Lett. 2008, 18,
3565−3569.
Sacchettini, J. C. Crystal structure of the Mycobacterium tuberculosis
enoyl-ACP reductase, InhA, in complex with NAD+ and a C16 fatty
acyl substrate. J. Biol. Chem. 1999, 274, 15582−15589.
(46) Wilson, D. N.; Schluenzen, F.; Harms, J. M.; Starosta, A. L.;
Connell, S. R.; Fucini, P. The oxazolidinone antibiotics perturb the
ribosomal peptidyl-transferase center and effect tRNA positioning.
Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 13339−13344.
(29) England, K.; am Ende, C.; Lu, H.; Sullivan, T. J.; Marlenee, N.
L.; Bowen, R. A.; Knudson, S. E.; Knudson, D. L.; Tonge, P. J.;
9927
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928

Journal of Medicinal Chemistry
Article
(47) Suller, M. T.; Russell, A. D. Triclosan and antibiotic resistance in
Staphylococcus aureus. J. Antimicrob. Chemother. 2000, 46, 11−18.
(48) Villalaín, J.; Mateo, C. R.; Aranda, F. J.; Shapiro, S.; Micol, V.
Membranotropic effects of the antibacterial agent triclosan. Arch.
Biochem. Biophys. 2001, 390, 128−136.
(49) Denyer, S. P.; Maillard, J. Y. Cellular impermeability and uptake
of biocides and antibiotics in Gram-negative bacteria. J. Appl. Microbiol.
2002, 92, 35S−45S.
(50) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Duce, P. P.; Morris,
J. J.; Taylor, P. J. Hydrogen bonding. Part 7. A scale of solute
hydrogen-bond acidity based on log K values for complexation in
tetrachloromethane. J. Chem. Soc., Perkin Trans. 2 1989, 699−711.
(51) Flo+ is a molecular modeling software developed by Colin
McMartin of Thistlesoft: McMartin, C.; Bohacek, R. J. QXP: powerful,
rapid computer algorithms for structure-based drug design. J. Comput.-
Aided Mol. Des. 1997, 11, 333.
(52) Bryskier, A.; Soulama, C.; Escaich, S.; Moreau, F.; Vongsouthi,
V.; Floquet, S. Malacain, E. ; Walton, A.; Prouvensier, L.; Fischer, S.
MUT056399: Mode of Action and Mechanisms of Resistance.
Presented at the 50th Meeting of ICAAC, Boston, MA, 2010; Poster
F1-832.
(53) Escaich, S.; Prouvensier, L.; Saccomani, M.; Durant, L.; Oxoby,
M.; Gerusz, V.; Moreau, F.; Vongsouthi, V.; Maher, K.; Morrissey, I.;
Soulama-Mouze, C. The MUT056399 inhibitor of FabI is a new
antistaphylococcal compound. Antimicrob. Agents Chemother. 2011, 55,
4692−4697.
(54) Observed IC50 > 25 μM in CYP1A2, CYP2C9, CYP2D6, and
CYP3A4.
(55) Escaich, S.; Prouvensier, L.; Saccomani, M.; Durant, L.;
Malacain, E.; Floquet, S.; Walton, A.; Sam-Sambo, V.; Faivre, F.;
Oxoby, M.; Bonvin, Y.; Gerusz, V.; Vongsouthi, V.; Moreau, F.;
Soulama, C. MUT056399 Fab I Inhibitor: A New Antibacterial
Candidate against Staphylococcus aureus. Presented at the 49th Meeting
of ICAAC, San Francisco, CA, 2009; Poster F1-2012.
(56) Identified metabolites were the glucoronidated phenol, the
sulfated phenol, and the acid resulting from the hydrolysis of the
amide. See ref 52 and Sweeney, K.; Henson, C.; McEwen, A.; Read,
H.; Wood, S.; Soulama-Mouze, C. [14C]-MUT056399: Rat ADME
Study with Metabolite Profiling and Identification. Presented at the
51st Meeting of ICAAC, 2011, Chicago, IL; Poster A2-032.
(57) Tsuji, B.; Lanke, S.; Ngo, D.; Soulama-Mouze, C.; Fischer, S.;
Forrest, A. Pharmacokinetic−Pharmacodynamic Investigation of
MUT056399 against Staphylococcus aureus in an in Vitro PD Model.
Presented at the 21st Meeting of ECCMID, Milan, Italy, 2011; Poster
P783.
(58) Soulama-Mouze, C.; Bryskier, A.; Chassard, D.; Fischer, S.
MUT056399: A Single Intravenous Ascending Dose Study in Healthy
Human Volunteers. Presented at the 21st Meeting of ECCMID,
Milan, Italy, 2011; Poster P1510.
(59) Compound 83, referred in the literature as MUT056399, is
currently undergoing clinical development by FAB Pharma.
9928
dx.doi.org/10.1021/jm301113w | J. Med. Chem. 2012, 55, 9914−9928