Synthesis and Antibacterial Evaluation of Bis-thiazolium, Bis-imidazolium, and Bis-triazolium Derivatives

Article abstract of DOI:10.1002/cmdc.201900151

Given the worldwide spread of bacterial drug resistance, there is an urgent need to develop new compounds that exhibit potent antibacterial activity and that are unimpaired by this phenomenon. Quaternary ammonium compounds have been used for many years as disinfectants, but recent advances have shown that polycationic derivatives exhibit much stronger activity and are less prone to bacterial resistance than commonly used monocationic compounds. In this sense, we prepared three series of new bis-cationic compounds: bis-thiazoliums, bis-imidazoliums, and bis-1,2,4-triazoliums. If some compounds of the first series showed fair antibacterial activity, most of those belonging to the two other series were highly potent, with minimum inhibitory concentrations close to 1 μg mL^?1. Some of them also exhibited low toxicity toward eukaryotic MRC-5 lung fibroblasts, and we showed that this toxicity is clearly correlated with clogP. Finally, four selected compounds were found to exhibit a clear bactericidal effect.

Full text of DOI:10.1002/cmdc.201900151

DOI: 10.1002/cmdc.201900151
Communications
Synthesis and Antibacterial Evaluation of Bis-thiazolium,
Bis-imidazolium, and Bis-triazolium Derivatives
Benoꢀt Thomas,[a, c] Raphaꢁl E. Duval,[a, b] Stꢂphane Fontanay,[a, b] Mihayl Varbanov,[a] and
Michel Boisbrun*[a]
Given the worldwide spread of bacterial drug resistance, there
is an urgent need to develop new compounds that exhibit
potent antibacterial activity and that are unimpaired by this
phenomenon. Quaternary ammonium compounds have been
used for many years as disinfectants, but recent advances have
shown that polycationic derivatives exhibit much stronger ac-
tivity and are less prone to bacterial resistance than commonly
used monocationic compounds. In this sense, we prepared
three series of new bis-cationic compounds: bis-thiazoliums,
bis-imidazoliums, and bis-1,2,4-triazoliums. If some compounds
of the first series showed fair antibacterial activity, most of
those belonging to the two other series were highly potent,
with minimum inhibitory concentrations close to 1 mgmLÀ1.
Some of them also exhibited low toxicity toward eukaryotic
MRC-5 lung fibroblasts, and we showed that this toxicity is
clearly correlated with clogP. Finally, four selected compounds
were found to exhibit a clear bactericidal effect.
cationic compounds. They also suggest that some derivatives
of this class could be used for other purposes than simple top-
ical administration.[4–6] In particular, polycationic compounds
are less prone to induce bacterial resistance than the classical
monocationic compounds, which are still in commercial use.[7,8]
Furthermore, one general drawback of this class of compounds
comes from the low selectivity toward eukaryotic cells. Thus,
some efforts have been made to correlate the structural fea-
tures of QACs and both their antibacterial activity and eukary-
otic toxicity.[9,10] In addition, other recent efforts led to the
design of powerful molecules that show very low toxicity
toward mammalian cells.[11] Besides, QACs may exhibit antima-
larial activity, and some bis-cationic derivatives proved to be
very potent both in vitro and in vivo.[12] Later, analogous com-
pounds were prepared, in which the cationic charges were har-
bored by thiazolium moieties.[4,13] The latter compounds were
found to be as potent as their classical ammonium counter-
parts, but were much less toxic on a rodent model. We won-
dered whether these positively charged heterocycles could
also exhibit attractive antibacterial activity. However, some
mono-thiazolium compounds showed low potency on refer-
ence bacterial strains,[14] but we envisioned that dimeric ana-
logues could be much more active. Thus, we prepared numer-
ous derivatives and tested them against reference bacterial
strains. The most active compound was selected for toxicity
evaluation on eukaryotic MRC-5 cells. Moreover, we intended
to extend this study to related five-membered heterocycles
such as imidazole derivatives. Some bis-imidazolium com-
pounds were already prepared by others[15,16] and showed in-
teresting, yet moderate, antibacterial activity. We envisioned
replacing some alkyl moieties with aromatic groups to deter-
mine their influence on both antibacterial activity and toxicity
versus eukaryotic cells. Finally, 1,2,4-triazole analogues were
also prepared, as in the context of antifungal therapy, they
have led to more active compounds,[17] and interesting anti-
bacterial activities have recently been reported.[18] All these
compounds were tested against reference bacterial strains and
also against clinical isolates harboring various and different
mechanisms of resistance toward antibiotics. Furthermore, the
cytotoxicity on eukaryotic cells was determined for each com-
pound.
The World Health Organization recently published a list[1] of 12
bacteria whose capacity for resistance against antibiotics is so
high that they constitute a real threat to human health; this
highlights that the time to act is now. A recent review chaired
by Jim O’Neill[2] also listed major recommendations to fight an-
timicrobial resistance. These include improving hygiene, mini-
mizing the unnecessary use of antibiotics, improving global
surveillance of drug resistance, and implementing better incen-
tives to promote investment for new drugs. Developing new
chemical entities is thus urgently required. Quaternary ammo-
nium compounds (QACs) were first discovered by Domagk in
1935[3] and were used for many years as skin disinfectants.
Mono-ammonium compounds such as benzalkonium chloride
or cetylpyridinium chloride were followed by bis-ammonium
compounds such as dequalinium chloride and chlorhexidine.
Recent advances in this area indeed showed that bis- or poly-
cationic compounds exhibit much stronger activity than mono-
[a] B. Thomas, Prof. Dr. R. E. Duval, Dr. S. Fontanay, Dr. M. Varbanov,
Dr. M. Boisbrun
Universitꢀ de Lorraine, CNRS, L2CM, 54000 Nancy (France)
E-mail: michel.boisbrun@univ-lorraine.fr
[b] Prof. Dr. R. E. Duval, Dr. S. Fontanay
Bis-thiazolium derivatives were easily prepared as previously
described[13] (Scheme 1) by N-alkylation of 4-methyl-5-(2-hy-
droxyethyl)thiazole with various a,w-diiodoalkanes, the latter
being prepared from commercial a,w-dibromoalkanes.[13] In
contrast to this previous study, each even and odd linkers from
n=4 to n=12 were prepared. As activity increased with n,
ABC Platform, Facultꢀ de Pharmacie, 54000 Nancy (France)
[c] B. Thomas
Laboratoire de Biologie Mꢀdicale, Hꢁpitaux Privꢀs de Metz, Metz (France)
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/cmdc.201900151.
ChemMedChem 2019, 14, 1 – 7
1
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
Compound 12 was the most active compound of this series,
with MICs reaching 16 mgmLÀ1 against these strains. No activi-
ty was observed against E. faecalis and P. aeruginosa. Increasing
the linker length did not bring significantly better activities.
Compound 14 (n=16), however, showed better MIC than 12
against E. faecalis (64 vs. 256 mgmLÀ1), but it was less potent
against E. coli (128 vs. 16 mgmLÀ1). Thus, we selected the
lower-molecular-weight compound 12 for further consider-
ation. In vitro cytotoxicity (IC50) on eukaryotic lung fibroblasts
(MRC-5 cells, ATCC CCL-171) was determined for this molecule
and was equal to 256 mgmLÀ1, leading to a selectivity index
(SI=IC50/MIC) of 16 toward the sensitive strains. Surprisingly,
compounds 15–17, O-methylated analogues of compounds
10–12, were much less active than their counterparts.
With these considerations in mind, we intended to prepare
bis-imidazolium and bis-1,2,4-triazolium analogues. Prior to di-
merization, we first alkylated the heterocycles with various
agents. In addition to a simple alkyl (i.e., decyl) chain, we also
introduced aromatic substituents such as 4-tert-butylbenzyl or
2-methylnaphtalenyl moieties. This was achieved by using
commercially available alkyl bromides under basic conditions
(Scheme 2). To avoid any trace of over-alkylation, the heterocy-
cles were introduced in excess relative to the alkylating agents
(3:1 ratio). The desired compounds were obtained in 41–74%
Scheme 1. Synthesis of bis-thiazolium compounds. a) NaH/THF, CH3I;
b) CH3CN, reflux, 72 h.
1,14-dibromotetradecane and 1,16-dibromohexadecane were
prepared according to a reported procedure.[19] Conversion
into the diiodo counterparts allowed the preparation of longer
analogues. Besides, O-methylation[13] prior to dimerization af-
forded three more compounds.
The minimum inhibitory concentrations (MICs) were evaluat-
ed against four reference strains: two Gram-positive bacteria
(Staphylococcus aureus and Enterococcus faecalis) and two
Gram-negative bacteria (Escherichia coli and Pseudomonas aer-
uginosa) (Table 1).
As a result, short-linker compounds exhibited no activity,
which overall increased with the linker length. Surprisingly,
compound 11 (n=11) showed no potency, whereas 10 (n=10)
and 12 (n=12) were active against E. coli and S. aureus strains.
Table 1. Minimum inhibitory concentration values of prepared bis-thiazo-
liums.
Scheme 2. Synthesis of bis-imidazolium and bis-1,2,4-triazolium compounds.
a) K2CO3, DMF, D; b) CH3CN, D.
[a]
Compd
n
MIC [mgmLÀ1
]
S.a.
E.f.
E.c.
P.a.
yields after column chromatography. Under the conditions
used (DMF, K2CO3, 808C), a highly regioselective 1-alkylation
was observed in the case of the 1,2,4-triazole compounds 26–
Chx[b]
4
1
2
>256
>256
128
>256
>256
>256
128
0.5
8
4
5
>256
>256
128
>256
>256
>256
64
256
16
32
16
>256
>256
128
>256
>256
>256
64
>256
16
256
>256
>256
>256
>256
>256
>256
128
>256
>256
256
5
1
6
6
28. This was easily determined by H NMR spectroscopy, which
7
7
showed unsymmetrical patterns. Only traces of the other re-
gioisomer were observed in the case of compound 28 and
they were easily removed by chromatography (see Supporting
Information). Dimerization with a,w-dibromoalkanes afforded
the desired bis-imidazolium or bis-1,2,4-triazolium compounds
in excellent (75–98%) yields after crystallization (Scheme 2).
The above-mentioned results led us to choose here linkers
with n=10–12 methylene units.
8
8
9
9
10
11
12
13
14
15
16
17
10
11
12
14
16
10
11
12
>256
>256
256
64
128
>256
256
256
>256
256
>256
>256
>256
>256
>256
>256
64
128
The MICs of the resulting 14 compounds were evaluated
against the same four reference strains, but also against four
clinical isolates harboring various resistance mechanisms
(Table 2). Unfortunately, compound 41 was too highly lipophil-
ic to be soluble under the conditions used, and thus its anti-
bacterial activity could not be evaluated. Globally, all the pre-
pared bis-imidazolium and bis-1,2,4-triazolium compounds
showed strong to very strong activities toward most of the
[a] All MIC data were acquired by compilation of the highest value of at
least two independent trials. Twofold serial dilutions of the compounds
were prepared from a stock solution at 1024 mgmLÀ1. Each condition was
repeated in eight wells. See the Supporting Information for further de-
tails. Organisms used in this study: S.a. (Staphylococcus aureus ATCC
29213), E.f. (Enterococcus faecalis ATCC 29212), E.c. (Escherichia coli ATCC
25922), P.a. (Pseudomonas aeruginosa ATCC 27853). [b] Reference com-
pound: chlorhexidine.
&
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
2
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
Table 2. Minimum inhibitory concentration and IC50 values of prepared bis-imidazolium and bis-triazolium compounds.
[a]
[e]
Compd
n
clogP
R[c]
MIC [mgmLÀ1
]
IC50 [mgmLÀ1
]
S.a.
MRSA
E.f.
VRE
E.c.
ESBLE
P.a.
PAR
Chx[b]
32
33
34
35
36
37
38
39
40
41
42
43
44
45
1
1
0.5
0.25
1
1
1
2
1
1
0.5
0.5
1
0.5
1
2
2
1
0.5
4
0.5
1
1
0.5
1
1
1
1
1
1
n.d.
1
2
1
0.5
1
2
1
1
4
1
1
1
4
2
n.d.
2
4
1
8
16
16
64
16
8
22.3
15.2
9.6
10
11
12
10
11
12
10
10
11
12
10
11
12
10
2.68
3.08
3.48
1.43
1.83
2.23
3.05
3.35
3.74
4.14
2.1
4-tbb
4-tbb
4-tbb
2-mn
2-mn
2-mn
decyl
4-tbb
4-tbb
4-tbb
2-mn
2-mn
2-mn
decyl
0.5
0.5
0.5
1
0.5
0.5
1
0.5
n.d.
1
0.5
1
16
4
64
32
16
2
16
16
n.d.
6.6
128
25.2
13.3
3.9
1.7
9.8
1
1
64
16
2
32
16
0.5
0.5
1
0.5
n.d.[d]
1
0.5
1
0.5
0.5
0.5
4
2
n.d.
2
2
1
0.5
2
0.5
n.d.
2
2
1
8.4
n.d.
n.d.
37.6
20.7
8.5
64
16
8
64
32
16
2
2.49
2.89
3.71
0.5
0.5
1
2
1.8
[a] All MIC data were acquired by compilation of the highest value of at least two independent trials. Twofold serial dilutions of the compounds were pre-
pared from a stock solution at 1024 mgmLÀ1. Each condition was repeated in eight wells. See the Supporting Information for further details. Organisms
used in this study: S.a. (Staphylococcus aureus ATCC 29213), MRSA (methicillin-resistant Staphylococcus aureus, mecA), E.f. (Enterococcus faecalis ATCC
29212), VRE (vancomycin-resistant Enterococcus faecalis, vanA), E.c. (Escherichia coli ATCC 25922), ESBLE (extended-spectrum b-lactamase-producing Escheri-
chia coli), P.a. (Pseudomonas aeruginosa ATCC 27853), PAR (Pseudomonas aeruginosa resistant, efflux pump). [b] Reference compound: chlorhexidine. [c] 4-
tbb=4-tert-butylbenzyl; 2-mn=2-methylnaphthalenyl. [d] Not determined (insoluble compound). [e] IC50 values were determined on MRC-5 cells according
to published procedures.[20, 21]
tested strains, with MICs around 1 mgmLÀ1. Besides, both
strains of P. aeruginosa appeared to be less sensitive; however,
some compounds exhibited low to very low MICs. Such strong
activity of our compounds against Gram-negative bacteria is
rare in this class of antibacterials and has to be highlighted.
The majority of these compounds proved to be as active as
the reference compound (chlorhexidine), and a few were
found to be more active. This is even more evident when
taking into account the lower molecular weight of chlorhexi-
dine (see the Supporting Information for MICs in mm units). In
addition, for most of the compounds and for each bacterial
species, there is almost no difference in activity toward the ref-
erence strain compared with the resistant clinical isolate. This
is highly important considering the need for treatments devot-
ed to drug-resistant bacteria. There is no clear difference in an-
timicrobial activity between the bis-imidazolium and bis-1,2,4-
triazolium series. In either case, the positive charge is spread
over the heterocyclic structure, as in QACs bearing a pyridini-
um structure.[9] Considering compounds with the same linker
length, derivatives bearing a 4-tert-butylbenzyl moiety are
slightly more active than those with a 2-methylnaphthalenyl
group (for example, compare 33 with 36, 34 with 37, and 40
with 43). Nevertheless, it is clear that independent of these
two substituents, a longer linker is associated with higher po-
tency. For the compounds bearing a decyl substituent (com-
pounds 38 and 45), only one linker (n=10) was envisioned; re-
markably, these compounds were found to be the most active,
especially against both strains of P. aeruginosa.
To evaluate the potential safety of our compounds, they
were tested against MRC-5 eukaryotic lung fibroblasts, and the
results are listed as IC50 values (Table 2). The obtained values
are very different from each other. Clearly, compounds 38 and
45, which showed the strongest antibacterial activity, also ex-
hibited the highest toxicity toward these eukaryotic cells. Com-
pounds with aromatic substituents are less toxic than those
bearing an alkyl (decyl) group. A longer linker increases the
toxicity, and for the same linker length, 2-methylnaphtalenyl
derivatives are less toxic than 4-tert-butylbenzyl compounds.
No clear difference in toxicity was observed between imidazoli-
um and 1,2,4-triazolium compounds, as 4-tert-butylbenzyl bis-
imidazoliums are less toxic than their 1,2,4-bis-triazolium coun-
terparts (for example, compare 32 with 39), but 2-methylnaph-
talenyl bis-imidazoliums are more toxic than their 1,2,4-bis-tria-
zolium counterparts (compare 35 with 42). The latter are the
least toxic derivatives of the series, with compound 42 exhibit-
ing an IC50 value of 37.6 mgmLÀ1. To confirm the low toxicity of
this compound, it was further evaluated in a red blood cell
lysis test. At roughly its IC50 value (i.e., 40.0 mgmLÀ1), no signifi-
cant lysis was observed (see Supporting Information), confirm-
ing the high potential of this compound.
To rationalize the above results, we intended to search for a
possible link between the antibacterial activities and eukaryotic
toxicities on the one hand, and the lipophilicities of the mole-
cules on the other. For this purpose, we used calculated logP
(clogP) as reported in Table 2. As mentioned above, the more
lipophilic 4-tert-butylbenzyl moiety is associated with slightly
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
3
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
higher activity than the less lipophilic 2-methylnaphthalenyl
derivatives. In addition, the better activity observed with the
compounds with longer linkers may be attributed to their
higher lipophilicity. Finally, the two compounds exhibiting the
highest activities—compounds 38 and 45—are highly lipophil-
ic, but not the most so. In addition, as reported above, no
clear difference in activity was observed between the bis-imi-
dazolium and bis-1,2,4-triazolium series, even though the
former is less lipophilic than the latter. To clarify these facts
and with the aim of linking them with the eukaryotic toxicities,
we generated some plots (Figure 1) showing the relationships
between IC50 and clogP, and between SI (SI=IC50/MIC) and
clogP for three bacterial strains. In the case of compounds
bearing the less lipophilic 2-methylnaphthalenyl moiety, there
is a clear linear correlation between IC50 and clogP in the bis-
imidazolium series (compounds 35, 36, and 37) and also in the
bis-1,2,4-triazolium series (compounds 42, 43, and 44). In both
cases, the slope is negative, as less lipophilic compounds (i.e.,
with shorter linkers) show higher IC50 values. In contrast, for
these six compounds and for the same linker length, each bis-
1,2,4-triazolium analogue is more lipophilic and shows a higher
IC50 value than its bis-imidazolium counterpart, with com-
pound 42 being the least toxic. Besides, considering com-
pounds bearing the more lipophilic 4-tert-butylbenzyl moiety,
all the compounds are more toxic than their 2-methylnaphtha-
lenyl counterpart. Once again, a linear correlation is observed
between IC50 and clogP in the bis-imidazolium series, but with
a lower (negative) slope than previously, and only two bis-
1,2,4-triazolium compounds could be tested.
toxic than their bis-1,2,4-triazolium counterparts. Finally, both
compounds harboring a linear alkyl chain as substituent (com-
pounds 38 and 45) are highly lipophilic and exhibit high toxici-
ty. To study the potential therapeutic interest of our molecules,
the selectivity indexes (SI) toward three bacterial strains were
evaluated (Figure 1). Concerning E. coli, as most of the MIC
values are equal to 1, SI values reflect the IC50 values, except
for compounds 34 (MIC=0.5) and 43 (MIC=2), with com-
pound 42 showing the best SI of 37.6. Concerning S. aureus, it
is interesting to take a closer look at bis-imidazolium com-
pounds. The one bearing a 2-methylnaphthalenyl moiety
shows the usual negative slope: a lower clogP correlates with
a higher SI. In contrast, the compounds bearing a 4-tert-butyl-
benzyl moiety show a positive slope: a higher clogP correlates
with a higher SI. As a result, compounds 34 (4-tbb, n=12) and
35 (2-mn, n=10) exhibit nearly the same SI (26.4 vs. 25.2, re-
spectively). This reflects the very low MIC of 34. A similar ten-
dency is observed for 1,2,4-bis-triazolium bearing a 4-tert-butyl-
benzyl moiety. For those bearing a 2-methylnaphthalenyl sub-
stituent, the MIC value (0.5) of 43 compensates for its relatively
low IC50 (20.7), yielding an interesting SI of 41.4, to be com-
pared with the SI of the least toxic compound 42 (37.6). Con-
cerning E. faecalis, most of the compounds show low SI,
except compound 35, which is a bis-imidazolium compound
bearing a 2-methylnaphthalenyl substituent. With a MIC of 0.5
and an IC50 value of 25.2, it combines high activity with rela-
tively low toxicity toward our eukaryotic cell model; thus its SI
reaches 50.4, which is the best of the series.
Based on these data, we decided to further characterize the
antibacterial properties of some compounds. The bis-imidazoli-
um and bis-1,2,4-triazolium derivatives showing the best anti-
In contrast with the previously evoked 2-methylnaphthalenyl
series, 4-tert-butylbenzyl bis-imidazolium derivatives are less
Figure 1. A) Toxicity of the various molecules on MRC-5 strains as a function of clogP. B–D) Selectivity indexes (SI) of the various molecules toward E. coli (B),
S. aureus (C), and E. faecalis (D) as a function of clogP.
&
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
4
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
microbial activities, lowest cytotoxicities, or interesting SI were
selected. Thus, we evaluated the bactericidal activity of com-
pounds 35, 42, 43, and 45 on two reference bacterial strains,
E. coli (Gram-negative) and S. aureus (Gram-positive). The bac-
tericidal activity was defined as a ꢀ3-log-unit decrease in the
initial inoculum with treatment compared with untreated con-
trol, for a compound concentration not greater than 4ꢄMIC,
after 24 h of incubation at 358C.[22] In the case of compounds
43 and 45, analysis of the minimal bactericidal concentration
(MBC) showed clear bactericidal activity on both tested bacteri-
al strains, with >3log decrease in the initial inoculum, after
24 h of incubation (Table 3). In that instance, the MBC was at-
tained independently of the compound concentration used in
the assay, 1ꢄMIC and 4ꢄMIC. Regarding compound 35 at 1ꢄ
MIC, we observed a killing effect for E. coli (starting inoculum
dose bactericidal activity toward two representative bacterial
strains. Interestingly, some of our compounds showed low tox-
icity toward eukaryotic MRC-5 lung fibroblasts. In addition, we
could demonstrate very interesting linear relationships be-
tween eukaryotic cytotoxicity and clogP of our compounds,
and between selectivity index and clogP. This study opens new
perspectives in the design of new bis-cationic features with
high antibacterial potency while maintaining low eukaryotic
toxicity.
Experimental Section
Full experimental details are provided in the Supporting Informa-
tion.
Acknowledgements
Table 3. Reduction of initial bacterial inoculum after 24 h of incubation
with the test compound at 358C.
The authors are grateful to the MRES, CNRS, and Rꢀgion Grand-
Est for their financial support. They also thank Fabien Lachaud
for spectroscopic measurements, and Sandrine Adach for elemen-
tal analysis. Special thanks are due to Arnaud Risler for technical
assistance, technical editing, and proofreading of the manuscript.
Technical editing help from Dorian Dupommier was also greatly
appreciated.
Compd
S. aureus
E. coli
ATCC 29213
ATCC 25922
1ꢄMIC
4ꢄMIC
1ꢄMIC
4ꢄMIC
35
42
43
45
>3log10
<3log10
>3log10
>3log10
>3log10
<3log10
>3log10
>3log10
<3log10
>3log10
>3log10
>3log10
>3log10
>3log10
>3log10
>3log10
Conflict of interest
of 4.2ꢄ105 Æ0.2ꢄ105 CFUmLÀ1, breakpoint at 420 CFUmLÀ1,
mean of 770 CFUmLÀ1 obtained), although the bactericidal
breakpoint is not reached. Nevertheless, the bactericidal break-
point for E. coli is reached at 4ꢄMIC (<10 CFUmLÀ1), while for
S. aureus, both tested concentrations proved bactericidal. Simi-
larly, for compound 42, at 4ꢄMIC the bactericidal breakpoint
for S. aureus was not reached, but a killing effect was also ob-
tained (starting inoculum of 6.7ꢄ105 Æ0.5ꢄ105 CFUmLÀ1,
breakpoint at 670 CFUmLÀ1, mean of 1600 CFUmLÀ1 obtained).
On the opposite, the impact of the compound on E. coli led to
>3log reduction of the initial inoculum. As compound 42 was
particularly interesting in terms of SI, we went further and
made a preliminary evaluation of the reduction of initial inocu-
lum of E. coli. At the MIC, the inoculum reduction was superior
to 5log, which emphasizes the high activity of this molecule.
In this study, we addressed the issue of the increase in anti-
biotic-resistant bacteria by designing new antibacterial com-
pounds designed to maintain their activity toward clinically oc-
curring drug-resistant bacteria. With this aim, we prepared nu-
merous bis-cationic compounds, among which bis-imidazolium
and bis-1,2,4-triazolium derivatives appeared to show high an-
tibacterial activity against two Gram-positive (S. aureus and
E. faecalis) and two Gram-negative (E. coli and P. aeruginosa)
reference bacterial strains. For most of these compounds, the
MICs were around 1 mgmLÀ1, except against P. aeruginosa. All
these compounds were at least as potent as chlorhexidine, a
molecule commonly used as a disinfectant. Notably, the anti-
bacterial activity was maintained against resistant clinical iso-
lates. Furthermore, four selected compounds exhibited low-
The authors declare no conflict of interest.
Keywords: 1,2,4-triazolium · antibacterial agents · clogP ·
cytotoxicity · imidazolium
[1] “WHO publishes list of bacteria for which new antibiotics are urgently
needed”, World Health Organization, http://www.who.int/mediacentre/
news/releases/2017/bacteria-antibiotics-needed/en/, February 27, 2017.
[2] “Tackling Drug-Resistant Infections Globally: Final Report and Recom-
mendations”, J. O’Neill, The Review on Antimicrobial Resistance, https://
amr-review.org/sites/default/files/160518_Final%20paper_with%20co-
ver.pdf, May 2016.
[3] G. Domagk, Dtsch. Med. Wochenschr. 1935, 61, 250–253.
[4] M. Tischer, G. Pradel, K. Ohlsen, U. Holzgrabe, ChemMedChem 2012, 7,
22–31.
[5] K. P. C. Minbiole, M. C. Jennings, L. E. Ator, J. W. Black, M. C. Grenier, J. E.
LaDow, K. L. Caran, K. Seifert, W. M. Wuest, Tetrahedron 2016, 72, 3559–
3566.
[6] M. E. Forman, M. C. Jennings, W. M. Wuest, K. P. C. Minbiole, ChemMed-
Chem 2016, 11, 1401–1405.
[7] M. C. Jennings, B. A. Buttaro, K. P. C. Minbiole, W. M. Wuest, ACS Infect.
Dis. 2015, 1, 304–309.
[8] M. C. Jennings, M. E. Forman, S. M. Duggan, K. P. C. Minbiole, W. M.
Wuest, ChemBioChem 2017, 18, 1573–1577.
[9] S. E. Al-Khalifa, M. C. Jennings, W. M. Wuest, K. P. C. Minbiole, ChemMed-
Chem 2017, 12, 280–283.
[10] R. C. Kontos, S. A. Schallenhammer, B. S. Bentley, K. R. Morrison, J. A. Fe-
liciano, J. A. Tasca, A. R. Kaplan, M. W. Bezpalko, W. S. Kassel, W. M.
Wuest, K. P. C. Minbiole, ChemMedChem 2019, 14, 83–87.
[11] J. Hoque, M. M. Konai, S. S. Sequeira, S. Samaddar, J. Haldar, J. Med.
Chem. 2016, 59, 10750–10762.
[12] K. Wengelnik, V. Vidal, M. L. Ancelin, A.-M. Cathiard, J. L. Morgat, C. H.
Kocken, M. Calas, S. Herrera, A. W. Thomas, H. J. Vial, Science 2002, 295,
1311–1314.
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
5
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
[13] A. Hamzꢂ, E. Rubi, P. Arnal, M. Boisbrun, C. Carcel, X. Salom-Roig, M.
Maynadier, S. Wein, H. Vial, M. Calas, J. Med. Chem. 2005, 48, 3639–
3643.
[20] N. K. Tittikpina, F. Nana, S. Fontanay, S. Philippot, K. Batawila, K. Akpaga-
na, G. Kirsch, P. Chaimbault, C. Jacob, R. E. Duval, J. Ethnopharmacol.
2018, 212, 200–207.
[21] Assay Guidance Manual (Eds.: G. S. Sittampalam, N. P. Coussens, K. Bri-
macombe, A. Grossman, M. Arkin, D. Auld, C. Austin, J. Baell, B. Bejcek,
J. M. M. Caaveiro, et al.) Eli Lilly & Company and The National Center for
Advancing Translational Sciences, Bethesda, MD (USA), 2004.
[22] R. K. Flamm, D. J. Farrell, P. R. Rhomberg, N. E. Scangarella-Oman, H. S.
Sader, Antimicrob. Agents Chemother. 2017, 61, e00468.
[14] A. Zablotskaya, I. Segal, A. Geronikaki, G. Kazachonokh, Y. Popelis, I.
Shestakova, V. Nikolajeva, D. Eze, MedChemComm 2015, 6, 1464–1470.
´
[15] Ł. Pałkowski, J. Błaszczynski, A. Skrzypczak, J. Błaszczak, K. Kozakowska,
´
´
J. Wrꢅblewska, S. Koz˙uszko, E. Gospodarek, J. Krysinski, R. Słowinski,
Chem. Biol. Drug Des. 2014, 83, 278–288.
[16] N. N. Al-Mohammed, Y. Alias, Z. Abdullah, RSC Adv. 2015, 5, 92602–
92617.
[17] J. Heeres, L. Meerpoel, P. Lewi, Molecules 2010, 15, 4129–4188.
[18] J. P. Shrestha, C. Baker, Y. Kawasaki, Y. P. Subedi, N. N. Vincent de Paul,
J. Y. Takemoto, C.-W. T. Chang, Eur. J. Med. Chem. 2017, 126, 696–704.
[19] D. Taffa, M. Kathiresan, L. Walder, Langmuir 2009, 25, 5371–5379.
Manuscript received: March 6, 2019
Revised manuscript received: April 27, 2019
Accepted manuscript online: && &&, 0000
Version of record online: && &&, 0000
&
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
6
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

COMMUNICATIONS
Highly potent antibacterials: New bis-
cationic compounds were prepared,
among which bis-imidazolium and bis-
1,2,4-triazolium compounds were found
to exhibit low to very low MICs toward
a panel of bacterial strains. These activi-
ties were also observed toward resistant
clinical isolates. Some derivatives dis-
played low toxicity toward a model of
eukaryotic cells. Very interesting correla-
tions were found between these activi-
ties and the lipophilicity of the com-
pounds.
B. Thomas, R. E. Duval, S. Fontanay,
M. Varbanov, M. Boisbrun*
&& – &&
Synthesis and Antibacterial Evaluation
of Bis-thiazolium, Bis-imidazolium,
and Bis-triazolium Derivatives
ChemMedChem 2019, 14, 1 – 7
www.chemmedchem.org
7
ꢃ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ