Diarylacylhydrazones: Clostridium-selective antibacterials with activity against stationary-phase cells

Article abstract of DOI:10.1016/j.bmcl.2013.12.015

Current antibiotics for treating Clostridium difficile infections (CDI), that is, metronidazole, vancomycin and more recently fidaxomicin, are mostly effective but treatment failure and disease relapse remain as significant clinical problems. The shortcomings of these agents are attributed to their low selectivity for C. difficile over normal gut microflora and their ineffectiveness against C. difficile spores. This Letter reports that certain diarylacylhydrazones identified during a high-throughput screening/counter- screening campaign show selective activity against two Clostridium species (C. difficile and Clostridium perfringens) over common gut commensals. Representative examples are shown to possess activity similar to vancomycin against clinical C. difficile strains and to kill stationary-phase C. difficile cells, which are responsible for spore production. Structure-activity relationships with additional synthesised analogues suggested a protonophoric mechanism may play a role in the observed activity/selectivity and this was supported by the well-known protonophore carbonyl cyanide m-chlorophenyl hydrazone (CCCP) showing selective anti-Clostridium effects and activity similar to diarylacylhydrazones against stationary-phase C. difficile cells. Two diarylacylhydrazones were shown to be non-toxic towards human FaDu and Hep G2 cells indicating that further studies with the class are warranted towards new drugs for CDI.

Full text of DOI:10.1016/j.bmcl.2013.12.015

Bioorganic & Medicinal Chemistry Letters 24 (2014) 595–600
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Diarylacylhydrazones: Clostridium-selective antibacterials
with activity against stationary-phase cells
Chao Chen a, , Naveen K. Dolla b, , Gabriele Casadei a,à, John B. Bremner b, Kim Lewis a, Michael J. Kelso b,
⇑
a Antimicrobial Discovery Center, Department of Biology, Northeastern University, Boston, MA 02115, USA
b School of Chemistry, University of Wollongong, NSW 2522, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Current antibiotics for treating Clostridium difficile infections (CDI), that is, metronidazole, vancomycin
and more recently fidaxomicin, are mostly effective but treatment failure and disease relapse remain
as significant clinical problems. The shortcomings of these agents are attributed to their low selectivity
for C. difficile over normal gut microflora and their ineffectiveness against C. difficile spores. This Letter
Received 15 October 2013
Revised 26 November 2013
Accepted 2 December 2013
Available online 10 December 2013
reports that certain diarylacylhydrazones identified during
a high-throughput screening/counter-
screening campaign show selective activity against two Clostridium species (C. difficile and Clostridium
perfringens) over common gut commensals. Representative examples are shown to possess activity
similar to vancomycin against clinical C. difficile strains and to kill stationary-phase C. difficile cells, which
are responsible for spore production. Structure–activity relationships with additional synthesised ana-
logues suggested a protonophoric mechanism may play a role in the observed activity/selectivity and this
was supported by the well-known protonophore carbonyl cyanide m-chlorophenyl hydrazone (CCCP)
showing selective anti-Clostridium effects and activity similar to diarylacylhydrazones against station-
ary-phase C. difficile cells. Two diarylacylhydrazones were shown to be non-toxic towards human FaDu
and Hep G2 cells indicating that further studies with the class are warranted towards new drugs for CDI.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Clostridium difficile
Antibacterial
Protonophore
Diarylacylhydrazone
CCCP
Stationary phase cells
Clostridium difficile-associated diarrhoea (CDAD), also known as
C. difficile infection (CDI), is the leading cause of infectious nosoco-
mial gastrointestinal illness.1,2 Steadily increasing CDI rates in US
hospitals,1 emergence of epidemic and hypervirulent strains (e.g.,
BI/NAP1/027),3 increased incidences of community acquired CDI4
and enormous costs to healthcare systems (estimated at $3.2 bil-
lion/year in the US alone)5 have focused considerable attention
on this disease over the past decade.6,7
environment and they can passage intact through the acidic stom-
ach. C. difficile typically resides asymptomatically in the human
gastrointestinal tract until normal microflora are disrupted, such
as following broad-spectrum antibiotic treatments, after which it
can overgrow producing three toxins; toxin A (TcdA), toxin B
(TcdB) and the binary toxin CDT. Ensuing CDI can range in severity
from mild diarrhoea to life-threatening pseudomembranous colitis
(PMC) and toxic megacolon.6
C. difficile is a Gram-positive, rod-shaped, spore-forming anaer-
obe transmitted via the oral-fecal route. In its vegetative form it is
highly sensitive to oxygen but its spores are heat stable, insensitive
to standard disinfectants, able to survive for long periods in the
Treatments for CDI historically have involved antibiotic with-
drawal followed by oral metronidazole 1 (500 mg t.i.d, 10–14 days)
or vancomycin 2 (125–250 mg q.i.d, 10 days) but treatment failure
remains as a significant and increasing problem (reportedly >35%
for metronidazole and 1–16% for vancomycin).2 For patients who
develop severe CDI (diarrhoea with leucocytosis, PMC or toxic
shock) metronidazole is effective in 76% of cases and vancomycin
97%.2 Upwards of 20–30% of patients can experience recurrent
CDI with intermittent episodes arising over months and sometimes
years.8 A major risk factor for recurrent CDI is failure to re-estab-
lish normal protective gut microflora due to the action of metroni-
dazole and vancomycin on gut commensals and the ineffectiveness
of these antibiotics against C. difficile spores.2
Abbreviations: CCCP, carbonyl cyanide m-chlorophenylhydrazone; CDAD, Clos-
tridium difficile-associated diarrhoea; CDI, Clostridium difficile infection; DACC, Data
Analysis and Coordination Centre; 2,4-DNP, 2,4-dinitrophenol; DMSO, dimethyl
sulfoxide; FDA, US Food and Drug Administration; HTS, high-throughput screening;
MBC, minimum bactericidal concentration; MIC, minimum inhibitory concentra-
tion; NCI, National Cancer Institute; NIH, National Institutes of Health; PBS,
phosphate buffered saline; PCP, pentachlorophenol; PMC, pseudomembranous
colitis.
⇑
Corresponding author. Tel.: +61 2 4221 5085; fax: +61 2 4221 4287.
E-mail address: mkelso@uow.edu.au (M.J. Kelso).
In 2011 the US Food and Drug Administration (FDA) approved
fidaxomicin 3 (DificidÒ, Optimer Pharmaceuticals Inc.),9 an oral
macrocyclic narrow-spectrum antibiotic developed specifically
Authors contributed equally to this work.
Present address: Istituto Zooprofilattico Sperimentale della Lombardia
dell’Emilia Romagna Sede Diagnostica Provinciale di Parma, 43126 Parma, Italy.
à
e
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.12.015

596
C. Chen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 595–600
for CDI (Fig. 1). Fidaxomicin shows very high in vitro potency
against C. difficile (minimum inhibitory concentrations (MICs)
against clinical isolates 0.008–0.25 mg/L) and reduced activity
against gut commensals, in particular Bacteroides species.10 This
and lower post-treatment C. difficile spore counts10 are thought
to contribute to reduced CDI recurrence with fidaxomicin.11,12 Re-
lapse rates for infections caused by the BI/NAP1/027 hypervirulent
strain, however, are the same for vancomycin and fidaxomicin13,14
and many hospitals have been slow to embrace the new drug due
to its high cost (>US$2700 per treatment, cf. metronidazole US$22
and vancomycin $1270).15
While the proper place for fidaxomicin in clinical practice is still
being established the search continues for alternative and, ideally,
more cost-effective agents. Attractive new compound classes
would include those that show high selectivity for C. difficile over
gut commensals along with activity against the stationary-phase
cells responsible for spore formation.16 In a recent high-through-
put screening (HTS) and counter-screening campaign we identified
that certain diarylacylhydrazones are Clostridium-selective
agents.17 This Letter reports the activity and selectivity of diary-
acylhydrazone screening hits, describes structure–activity studies
around the class and reports that a representative member is active
against stationary-phase C. difficile cells. Evidence is presented that
selective anti-Clostridium activity in the class may arise, in part,
through a protonophoric mechanism.
concentrations (MBC) of two examples chosen for further study
(i.e., compounds 5 and 6, see below) against CD196 were equiva-
lent to their MICs, confirming that the compounds are bactericidal.
Quinolinium acetate, chloride and mesylate salts of 7 showed iden-
tical activity to the free base 7. Compound 4 was shown to have
activity similar to vancomycin (<2-fold difference in MIC) against
five C. difficile clinical isolates (Supplementary data, Fig. S1).17
Importantly, compounds 4–7 were essentially inactive (MIC
>50
lg/mL) across the gut commensal panel with only 5 and 6
showingweakactivity(MIC=12.5–25
lg/mL)againstBacteroidesfra-
gilis and B. thetaiotaomicron. In contrast, metronidazole 1, vancomy-
cin 2 and fidaxomicin 3 all showed high potency against C. difficile
accompanied by significant activity against commensals (Fig. 2).
Three structural motifs common in hits 4–7 were: (1) an acyl
hydrazone, (2) an aryl ring at R1 and (3) an aryl ring at R2 carrying
an ortho-hydroxy substituent. Follow-up similarity searches of our
libraries identified six additional diarylacylhydrazones containing
the 2-hydroxynaphthalene group at R2 available for examination
(Supplementary data, Fig. S2, compounds S1–S6). Compounds
S1–S3, which incorporate phenyl, m-bromophenyl and o-hydroxy-
naphthyl groups at R1, respectively, displayed MICs against CD196
in the range 6.25–12.5 lg/mL. Compounds S4–S6, which carry 4-
nitrophenyl, 3-nitrophenyl and 3,5-dinitrophenyl substituents at
R1, respectively, all showed no activity against CD196. These find-
ings indicate that substituents on the aryl group at R1 can dramat-
ically impact potency. A further nine compounds with structures
related to 4–7 but lacking one or more of the above criteria (Sup-
plementary data, Figs. S3, S7–S15) were selected from the libraries
for testing and found to be inactive. Diarylacylhydrazones with
aryl rings at R1 and ortho-hydroxy substituted aryl rings at R2 thus
appear to present the minimal structural requirements for selec-
tive anti-Clostridium activity in the class.
High-throughput screening was carried out to identify hits
against C. difficile CD196.17 Hits were subjected to a counter-
screening panel of ten bacterial species representing the major
taxonomic groups from the human gut environment in order to
identify Clostridium-selective compounds. Counter-screening spe-
cies were chosen from the Data Analysis and Coordination Centre
(DACC) for the Human Microbiome Project (http://www.hmp-
dacc.org/), part of the National Institutes of Health (NIH) Roadmap
for Medical Research, and included abundant members of the gut
flora, for example, Bacteroides thetaiotaomicron, organisms of clini-
cal significance, for example, Escherichia coli, Enterococcus faecalis
and Staphylococcus aureus, along with representatives from each
of the predominant phyla. Clostridium perfringens was included to
identify species-specific anti-C. difficile compounds.
Encouraged by these results, analogues 8–19 (Fig. 3) were syn-
thesised to answer specific structure-activity questions about the
class. The targeted compounds were all prepared by heating the
requisite R1-acylhydrazines (prepared by reacting precursor
methyl esters with hydrazine) and R2-aryl aldehydes overnight in
ethanol (Fig. 3). Yields ranging from 70% to 90% of the pure com-
pounds were obtained after silica-gel column chromatography
and/or recrystallization. Compound 8 (INP0400) carrying a p-chlo-
rophenyl group at R1 and a 2-hydroxynaphthyl group at R2 was tar-
geted because it had previously been reported as an inhibitor of
type III secretion in the common bacterial pathogen Chlamydia tra-
Four acylhydrazones carrying aryl substituents at R1 and R2 (i.e.,
diarylacylhydrazones 4–7, Fig. 2) were initially identified as hits.
Follow-up measurements showed that the minimum inhibitory
concentrations (MIC) of 4–7 against C. difficile CD196 ranged from
1.56–6.25
l
g/mL and that the compounds exhibited similar
chomatis.18 The MIC of 8 against CD196 was found to be 6.25
lg/
MICs against Clostridium perfringens. Minimum bactericidal
mL and it showed higher potency against C. perfringens (MIC 6
Cl
HO
NH2
O
OH
OH
OH
HO
O
Me
HO
OMe
O
Me
O
HO
CH2OH
O
Cl
O
O
O
Cl
O
O
O
O
OH
NH
Cl
O
HO
O
H
H
H
H
N
N
O
N
N
NHMe
H
OH
OH
H
O
H
H
O
O
H
HO2C
O
HO
H
O
N
NH
O
NH2
NO2
N
O
O
OH
HO
OH
Metronidazole 1
HO
Vancomycin 2
Fidaxomicin 3
Figure 1. Current antibiotic treatments for C. difficile infections.

C. Chen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 595–600
597
MIC (μg/mL)
Compound
C. difficile CD196 C. perfringens B. fragilis B. thetaiotaomicron B. longum L. casei L. reuteri S. aureus S. mutans E. coli E. faecalis
1
Metronidazole
Vancomycin
Fidaxomicin
1.56
1.00
0.18
3.13
< 0.78
< 0.0075
< 0.78
12.5 - 25
> 50
< 0.78
12.5 - 25
> 50
< 0.78
< 0.78
< 0.78
> 50
> 50
> 50
> 50
> 50
> 50
3.13
> 50
1.56
3.13
> 50
> 50
> 50
> 50
3.13
12.5
2
3
> 50 3.13 - 6.25
O
N
R2
R2
R1
R1
N
H
HO
4
1.56
3.13
1.56
> 50
12.5
> 50
25
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
Ph
HO
HO
5
6.25 (6.25)*
6
3.13 (3.13)* < 3.13
12.5
> 50
25
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
NH
N
OCH3
Br
HO
7#
6.25
6.25
> 50
N
*Values in parentheses represent minimal bactericidal concentrations(MBC). #Quinolinium acetate, chloride and mesylate salts showed identical activity to
7
.
Figure 2. Activity of metronidazole 1, vancomycin 2, fidaxomicin 3 and diarylacylhydrazones 4–7 against C. difficile CD196 and a panel of gut commensals.
0.78
l
g/mL). The compound’s selectivity for Clostridium species
and thus high structural similarity to 6, was synthesised to explore
the possibility that nitrofurans and diarylacylhydrazones might
share overlapping antibacterial mechanism(s) of action. However,
19 showed broad spectrum activity across the panel, including
over gut commensals was similar to 5 showing only weak activity
against B. fragilis and B. thetaiotaomicron (MIC = 12.5 lg/mL) and
no other activity across the panel.
Compound 9, which substituted the ortho-hydroxynaphthyl
group present at R2 in 5 with the ortho-vanillyl moiety of 7, was
synthesized to probe the effect of interchanging the two different
R2 groups present in diarylacylhydrazone screening hits. Clostrid-
ium-selective activity was observed with 9 which showed MICs
high potency against C. difficile CD196 (MIC = 0.78 lg/mL, Fig. 3),
confirming that diarylacylhydrazones and nitrofurans exert their
antibacterial effects via different mechanisms.
Demonstrating that nitrofurans invoke different antibacterial
mechanisms and that anti-Clostridium activity is lost in diary-
lacylhydrazones upon methylation of the ‘amide’ nitrogen and/or
the ortho-phenolic group present at R2 led to speculation that the
compounds might be exerting their selective effects through a pro-
tonophoric mechanism. To test this hypothesis, MIC measurements
were obtained with three well-known membrane-active protono-
phores; carbonyl cyanide m-chlorophenylhydrazone (CCCP),20
pentachlorophenol (PCP),21 and 2,4-dinitrophenol (2,4-DNP)22
(Fig. 4) against CD196 and the gut panel. It was reasoned that if
Clostridium-selective activity were observed with one or more of
these it would support a protonophoric mechanism for diary-
lacylhydrazones. CCCP was found to be highly active against
of 12.5 and 6.25 lg/mL against C. difficile and C. perfringens, respec-
tively, and no other activity across the panel. Replacing the ortho-
hydroxynaphthyl group at R2 of 6 with the ortho-vanillyl moiety
(i.e., compound 10), however, abolished all activity, including
against both Clostridium species. Total loss of activity across the pa-
nel was similarly observed with 11, where the ortho-vanillyl group
at R2 in 7 was replaced with the ortho-hydroxynaphthyl group.
These results demonstrate that the Clostridium-selective activity
of diarylacylhydrazones is structure-dependent and is affected by
substituents on both aryl groups at R1 and R2.
The importance of the ortho-hydroxy group at R2 was estab-
lished next with compounds 12–15. Transferring the hydroxyl
group of 5 to the naphthyl 4-position (compound 12) led to total
loss of activity, as did removing the ortho-hydroxy group alto-
gether from compounds 5, 6 and 7 (compounds 13, 14, and 15,
respectively). Methylation of the acylhydrazone ‘amide’ nitrogen
of 5 (compound 16), its ortho-phenolic group at R2 (i.e., compound
17) or both of these groups (compound 18) similarly removed all
activity.
It was noted that diarylacylhydrazones structurally resemble
nitrofurans, an older class of broad spectrum antibiotics that have
seen widespread historical use in humans and in veterinary medi-
cine.19 Nitrofurans typically contain an acylhydrazone substituted
with a 5-nitrofuranyl group at the position corresponding to R2 in
diarylacylhydrazones. A variety of substituents, both aryl and
non-aryl, can be present at the position corresponding to R1 (e.g.,
R1 = p-hydroxyphenyl, nifuroxazide; R1 = NH2, nitrofurazone; R1
= hydantoin, nitrofurantoin; R1 = oxazolidinone, furazolidone).
The novel nitrofuran 19, carrying a 3-methylpyrazole group at R1
CD196 (MIC = 1.56
ing only one other panel member (Bifidobacterium longum, MIC =
3.25 g/mL). In contrast to diarylacylhydrazones, CCCP showed
lg/mL) and also showed high selectivity, affect-
l
no activity against C. perfringens. PCP and 2,4-DNP showed only
weak activity across the panel and no selectivity towards Clostrid-
ium, indicating that selective action against Clostridium is not a
general effect of protonophores. The similarity between the activ-
ity/selectivity observed with CCCP and diarylacylhydrazones im-
plies that a protonophoric mechanism probably plays some role
in the mechanism. That CCCP and diarylacylhydrazones both carry
hydrazone moieties may be important but this remains to be
determined.
Recent reports indicate that perturbing membrane function
represents a promising strategy towards novel anti-C. difficile ther-
apeutics,23 especially since some membrane-active agents affect
the quiescent/stationary-phase cells responsible for spore forma-
tion.24 Two membrane-active agents oritavancin (vancomycin ana-
logue)25 and CB-183,315 (daptomycin analogue)26 are in clinical

598
C. Chen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 595–600
O
O
O
EtOH
R2
+
NH2
R1
N
R1
N
R1
N
H
R2
Δ
MIC (μg/mL)
R3
R3
R3
C. difficile CD196 C. perfringens B. fragilis B. thetaiotaomicron B. longum L. casei L. reuteri S. aureus S. mutans E. coli E. faecalis
R2
HO
8
H
6.25
< 0.78
12.5
12.5
> 50
> 50
> 50
> 50
> 50
> 50
> 50
Cl
INP040018
HO
OCH3
9
H
H
12.5
> 50
6.25
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
1.56
> 50
> 50
> 50
> 50
> 50
OCH3
HO
NH
10
N
HO
N
11
12
H
H
> 50
> 50
25
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
Br
OH
> 50
13
14
H
H
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
NH
N
OCH3
N
15
16
17
H
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
> 50
Br
HO
CH3
H
H3CO
H3CO
CH3
H
18
19
> 50
0.78
> 50
> 50
> 50
> 50
> 50
6.25
> 50
12.5
> 50
1.56
> 50
1.56
> 50
12.5
> 50
12.5
NO2
< 12.5
< 3.13
< 3.13
< 0.78
O
NH
N
Figure 3. Activity of synthesized diarylacylhydrazones 8–19 against C. difficile CD196 and a panel of gut commensals.
MIC (μg/mL)
C. difficile CD196 C. perfringens B. fragilis B. thetaiotaomicron B. longum L. casei L. reuteri S. aureus S. mutans E. coli E. faecalis
Protonophore
Cl
N
Carbonyl cyanide
H
N
m-chlorophenyl
1.56
50
> 50
6.25
> 50
> 50
> 50
6.25
25
3.25
12.5
25
> 50
50
> 50
50
> 50
< 0.78
25
> 50
25
> 50
> 50
> 50
> 50
> 50
> 50
N
hydrazone (CCCP)
N
Cl
Cl
Cl
Cl
Pentachlorophenol
6.25/12.5
12.5/25
HO
Cl
O2N
HO
NO2
2,4-dinitrophenol
50
> 50
> 50
> 50
Figure 4. Activity of three protonophores against C. difficile CD196 and a panel of gut commensals.
development for CDI. Several membrane-active compounds
logarithmic- and stationary-phase cultures.27 The activity of diary-
(including CCCP) were recently shown to kill C. difficile cells in both
lacylhydrazones 5 and 6 was thus examined against stationary-

C. Chen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 595–600
599
phase C. difficile CD196 cells alongside CCCP, with metronidazole 1,
Cytotoxic conc. (μg/mL)
vancomycin 2 and fidaxomicin 3 included for comparison.
CD196 cells were grown to stationary phase according to an in-
house procedure (Supplementary data) and treated with 1ꢀ, 5ꢀ,
Compound FaDu HepG2
Miconazole
50
50
10ꢀ and 20ꢀ the previously determined MIC (
lg/mL) concentra-
Metronidazole
tions of each compound. Metronidazole 1, vancomycin 2 and fidax-
omicin 3 showed little or no activity at all concentrations tested
(Fig. 5a–c). Compound 5 showed some ability to kill stationary-
phase cells but the data was not reproducible (not shown) and it
was suspected that the poor solubility of 5 in the assay medium
may have been partially responsible. Switching to the more soluble
6 produced reproducible dose-dependent cell-killing effects that
were remarkably similar to those observed with CCCP (Fig. 5d
and e), supporting the postulate that diarylacylhydrazones and
CCCP may indeed exert their selective anti-Clostridium effects
through overlapping (protonophoric) mechanisms.
1
>200
>200
Vancomycin
2
5
6
>200
>100
>100
50
>200
>100
>100
50
CCCP
DMSO showedno cytotoxicity at 2% (v/v).
Figure 6. Cytotoxicity against human FaDu and HepG2 cells.
similar activity to diarylacylhydrazones against the gut panel and
against stationary phase C. difficile cells suggests that a protono-
phoric mechanism may play a role in the selectivity, although
many questions remain. One avenue we are exploring is the possi-
ble involvement of dynamin-like proteins, which are important
mediators of membrane remodelling in bacteria (called dynamins
in eukaryotes).30 Inhibition of dynamins by diarylacylhydrazones
is well characterised31 and compounds structurally very similar
to those reported here were recently shown to prevent uptake of
C. difficile TcdA into eukaryotic cells.32
Given the known effects of CCCP as a protonophoric uncoupler
of oxidative phosphorylation,28 it was of interest to examine hu-
man cell cytotoxicity as a preliminary indicator of druggability
(or otherwise) in the diarylacylhydrazone class. Compounds 5
and 6 were examined alongside miconazole (positive control),
metronidazole 1, vancomycin 2 and CCCP for cytotoxicity against
human FaDu and Hep G2 cells using a standard 96-well resazu-
rin-based cell viability assay (Supplementary data). The positive
control miconazole showed the expected cytotoxic concentration
(50
cin 2 showed no toxicity at or below 200
line. Compounds 5 and 6 showed no cytotoxicity against either cell
lg/mL) against both cell lines. Metronidazole 1 and vancomy-
lg/mL against either cell
Acknowledgments
line at 100
(Fig. 6).
lg/mLwhile CCCP was found to be toxic at 50 lg/mL
We thank the University of Wollongong (Australia) and North-
eastern University (Boston, USA) for supporting this work. We
gratefully acknowledge the US NIH (R01 AI076372) for funding
and thank the National Cancer Institute (NCI) for generously pro-
viding compounds for screening.
This study demonstrates that certain diarylacylhydrazones are
narrow-spectrum antibacterials with selectivity for C. difficile and
C. perfringens over other gut commensals. The demonstrated struc-
ture-dependence of the selectivity is significant because other
diarylacylhydrazones have previously been shown to have a wider
spectrum of antibacterial activity.29 The activity of 6 against sta-
tionary-phase C. difficile cells and its low human cell cytotoxicity
indicate that further investigations with the class are warranted
towards creating cost-effective antibiotics for CDI that may reduce
treatment failure and relapse rates. Identifying that CCCP shows
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.
12.015.
Figure 5. Activity against stationary-phase C. difficile CD196 cells of: (a) metronidazole 1 (b) vancomycin 2 (c) fidaxomicin 3 (d) 6 (e) CCCP. DMSO was present at a final
concentration of 1% v/v in all assay solutions.

600
C. Chen et al. / Bioorg. Med. Chem. Lett. 24 (2014) 595–600
17. Details of the HTS campaign will be reported elsewhere. The MBC for
compound 4 was 2-fold higher than its MIC (i.e., MBC = 3.13 g/mL).
References and notes
l
0
18. Muschol, S.; Bailey, L.; Gylfe, ÅA.; Sundin, C.; Hultenby, K.; Bergström, S.;
Eloffson, M.; Wolf-Watz, H.; Normark, S.; Henriques-Normark, B. Proc. Natl.
Acad. Sci. U.S.A. 2006, 103, 14566.
19. Vass, M.; Hruska, K.; Franek, M. Vet. Med. 2008, 53, 469.
20. Kasianowicz, J.; Benz, R.; McLaughlin, S. J. Membr. Biol. 1984, 82, 179.
21. Weinbach, E. J. Biol. Chem. 1954, 210, 545.
22. Drysdale, G. R.; Cohn, M. J. Biol. Chem. 1958, 233, 1574.
23. Hurdle, J. G.; O’Neill, A. J.; Chopra, I.; Lee, R. E. Nat. Rev. Microbiol. 2011, 9, 62.
24. Hurdle, J. G.; Heathcott, A.; Yang, L.; Yan, B.; Lee, R. E. J. Antimicrob. Chemother.
2011, 66, 1773.
25. Baines, S. D.; O’Connor, R.; Saxton, K.; Freeman, J.; Wilcox, M. H. J. Antimicrob.
Chemother. 2008, 62, 1078.
26. Mascio, C. T. M.; Mortin, L. I.; Howland, K. T.; Van Praagh, A. D. G.; Zhang, S.;
Arya, A.; Chuong, C. L.; Kang, C.; Li, T.; Silverman, J. A. Antimicrob. Agents
Chemother. 2012, 56, 5023.
27. Wu, X.; Cherian, P. T.; Lee, R. E.; Hurdle, J. G. J. Antimicrob. Chemother. 2013, 68,
806.
28. Kadenbach, B. Biochim. Biophys. Acta Bioenerg. 2003, 1604, 77.
29. Zhang, M.; Xian, D.-M.; Li, H.-H.; Zhang, J.-C.; You, Z.-L. Aust. J. Chem. 2012, 65,
343.
30. Bürmann, F.; Ebert, N.; van Baarle, S.; Bramkamp, M. Mol. Microbiol. 2011, 79,
1294.
1. Hookman, P.; Barkin, J. S. World J. Gastroenterol. 2009, 15, 1554.
2. McFarland, L. V. Exp. Opin. Emerg. Drugs 2011, 16, 425.
3. Pépin, J.; Valiquette, L.; Alary, M.-E.; Villemure, P.; Pelletier, A.; Forget, K.;
Pépin, K.; Chouinard, D. Can. Med. Assoc. J. 2004, 171, 466.
4. Khanna, S.; Pardi, D. S. Clin. Infect. Dis. 2012, 55, 1741.
5. O’Brien, J. A.; Lahue, B. J.; Caro, J. J.; Davidson, D. M. Infect. Control Hosp.
Epidemiol. 2007, 28, 1219.
6. Kachrimanidou, M.; Malisiovas, N. Crit. Rev. Microbiol. 2011, 37, 178.
7. Pacheco, S. M.; Johnson, S. Curr. Opin. Gastroenterol. 2013, 29, 42.
8. McFarland, L. V.; Beneda, H. W.; Clarridge, J. E.; Raugi, G. J. Am. J. Infect. Control
2007, 35, 237.
9. Johnson, A. P.; Wilcox, M. H. J. Antimicrob. Chemother. 2012, 67, 2788.
10. Louie, T. J.; Emery, J.; Krulicki, W.; Byrne, B.; Mah, M. Antimicrob. Agents
Chemother. 2009, 53, 261.
11. Tannock, G. W.; Munro, K.; Taylor, C.; Lawley, B.; Young, W.; Byrne, B.; Emery,
J.; Louie, T. Microbiology 2010, 156, 3354.
12. Louie, T. J.; Miller, M. A.; Mullane, K. M.; Weiss, K.; Lentnek, A.; Golan, Y.;
Gorbach, S.; Sears, P.; Shue, Y.-K. New Engl. J. Med. 2011, 364, 422.
13. Crook, D. W.; Walker, A. S.; Kean, Y.; Weiss, K.; Cornely, O. A.; Miller, M. A.;
Esposito, R.; Louie, T. J.; Stoesser, N. E.; Young, B. C.; Angus, B. J.; Gorbach, S. L.;
Peto, T. E. A. Clin. Infect. Dis. 2012, 55, S93.
14. Petrella, L. A.; Sambol, S. P.; Cheknis, A.; Nagaro, K.; Kean, Y.; Sears, P. S.;
Babakhani, F.; Johnson, S.; Gerding, D. N. Clin. Infect. Dis. 2012, 55, 351.
15. Lancaster, J. W.; Matthews, S. J. Clin. Ther. 2012, 34, 1.
31. Harper, C. B.; Popoff, M. R.; McCluskey, A.; Robinson, P. J.; Meunier, F. A. Trends
Cell Biol. 2013, 23, 90.
32. Gerhard, R.; Frenzel, E.; Goy, S.; Olling, A. J. Med. Microbiol. 2013, 62,
1414.
16. Merrigan, M.; Venugopal, A.; Mallozzi, M.; Roxas, B.; Viswanathan, V. K.;
Johnson, S.; Gerding, D. N.; Vedantam, G. J. Bacteriol. 2010, 192, 4904.