Synthesis, anti-proliferative activity, and toxicity of C4(C5) substituted N,N′-bis(arylmethyl)imidazolium salts

Article abstract of DOI:10.1016/j.tet.2016.07.068

The syntheses and characterization of C⁴and C⁵substituted N,N′-bis(arylmethyl)imidazolium salts with hydrophilic or lipophilic substituents on the imidazole ring are reported. A structure-activity relationship study revealed that the

Full text of DOI:10.1016/j.tet.2016.07.068

Accepted Manuscript
4
5
Synthesis, anti-proliferative activity, and toxicity of C (C ) substituted N,N′-
bis(arylmethyl)imidazolium salts
Kerri L. Shelton, Michael A. DeBord, Patrick O. Wagers, Marie R. Southerland,
Alexandra Taraboletti, Nikki K. Robishaw, Daniel P. Jackson, Radisa Tosanovic,
William G. Kofron, Claire A. Tessier, Sailaja Paruchuri, Leah P. Shriver, Matthew J.
Panzner, Wiley J. Youngs
PII:
S0040-4020(16)30726-8
10.1016/j.tet.2016.07.068
TET 27964
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 23 May 2016
Revised Date: 26 July 2016
Accepted Date: 27 July 2016
Please cite this article as: Shelton KL, DeBord MA, Wagers PO, Southerland MR, Taraboletti
A, Robishaw NK, Jackson DP, Tosanovic R, Kofron WG, Tessier CA, Paruchuri S, Shriver LP,
4
5
Panzner MJ, Youngs WJ, Synthesis, anti-proliferative activity, and toxicity of C (C ) substituted N,N′-
bis(arylmethyl)imidazolium salts, Tetrahedron (2016), doi: 10.1016/j.tet.2016.07.068.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract
Synthesis, anti-proliferative activity, and toxicity of C4(C5) substituted
N,N’-bis(arylmethyl)imidazolium salts
Kerri L. Shelton,† Michael A. DeBord,† Patrick O. Wagers,† Marie R. Southerland, Alexandra
Taraboletti, Nikki K. Robishaw, Daniel P. Jackson, Radisa Tosnaovic, William G. Kofron, Claire
A. Tessier, Sailaja Paruchuri, Leah P. Shriver,* Matthew J. Panzner,* Wiley J. Youngs*
Department of Chemistry, The University of Akron, Akron, OH 44325-3601, USA, Tel: 330-
972-5362, youngs@uakron.edu

ACCEPTED MANUSCRIPT
1
Synthesis, anti-proliferative activity, and toxicity of C4(C5) substituted
N,N’-bis(arylmethyl)imidazolium salts
Kerri L. Shelton,† Michael A. DeBord,† Patrick O. Wagers,† Marie R. Southerland, Alexandra
Taraboletti, Nikki K. Robishaw, Daniel P. Jackson, Radisa Tosanovic, William G. Kofron, Claire
A. Tessier, Sailaja Paruchuri, Leah P. Shriver,* Matthew J. Panzner,* Wiley J. Youngs*
Department of Chemistry, The University of Akron, Akron, OH 44325-3601, USA
†These authors contributed equally to the manuscript
Keywords: Lung cancer, Imidazolium salt, Anti-proliferative, Anti-tumor
Abstract
The syntheses and characterization of C4 and C5 substituted N,N’-
bis(arylmethyl)imidazolium salts with hydrophilic or lipophilic substituents on the imidazole
ring are reported. A structure-activity relationship study revealed that the lipophilicity of groups
at the C4 and C5 positions plays a crucial role in modulating the efficacy against select non-small
cell lung cancer cell lines tested. Compounds 11 – 17 were determined to be the most active
against the panel of cell lines studied. Compounds 11 and 12 were examined by the National
Cancer Institute’s Developmental Therapeutic Program where they were tested against the NCI-
60 human cancer cell line panel in a one-dose and five-dose assay. Compound 11 had high
activity against the nine lung cancer lines tested while 12 had cytotoxic effects against 59 of the
60 cell lines. Compound 11 was also studied in a murine model to determine its in vivo toxicity.
1. Introduction
Cancer is the second leading cause of death in the United States, with lung cancer
accounting for 27% of all cancer deaths.1 Despite significant advances in the treatment of cancer,
one of the major drawbacks of many clinically used chemotherapeutics is their substantial toxic

ACCEPTED MANUSCRIPT
2
side effects. An ongoing goal of cancer research is to obtain chemotherapeutics with high
efficacy and reduced side effects.
Imidazolium salts as potential anti-cancer agents are under active investigation by several
research groups.1-6 Research in our group has focused on N,N’-bis(naphthylmethyl)imidazolium
salts, many of which have been shown to exhibit high anti-proliferative activity against non-
small cell lung cancer (NSCLC) cell lines. One notable example is IC23 (Figure 1).6 Despite its
substantial activity, the low water solubility of IC23
–
similar to many
bis(naphthylmethyl)imidazolium salts – presents a severe challenge to clinical applications.
Herein we report the influence of the nature of hydrophilic and lipophilic functional groups in
the C4 and C5 position of N,N’-bis(naphthylmethyl)imidazolium salts, along with the impact of
hydrophilic and lipophilic naphthalene-based solubilizing substituents at the nitrogen (N1 and
N3) atoms of 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium salts as it relates to aqueous
solubility and anti-tumor activity against NSCLC cell lines.
Figure 1. Structure of IC23.
2. Results and Discussion
2.1 Synthesis and Characterization
To investigate the influence of substituents at the C4 and C5 positions on the in vitro anti-
cancer activity of N,N’-bis(naphthylmethyl)imidazolium salts, we first chose to explore the
naturally occurring imidazole derivatives urocanic acid and histamine (Figure 2). Both azoles
consist of an imidazole ring substituted at the C4 position with a two carbon linker connected to a
terminal hydrophilic group.

ACCEPTED MANUSCRIPT
3
Figure 2. Urocanic acid (left) and histamine (right).
The imidazolium salts 4-(3-methoxy-3-oxoprop-1-en-1-yl)-1,3-bis(naphthalen-2-
ylmethyl)imidazolium bromide (2) and 4-(2-carboxyvinyl)-1,3-bis(naphthalen-2-
ylmethyl)imidazolium chloride (3) were synthesized from commercially available urocanic acid
as outlined in Scheme 1. Urocanic acid was esterified to methyl-3-(imidazol-4-yl)acrylate (1)
according to literature procedure.7,8 Compound 1 was deprotonated with potassium carbonate
and reacted with one equivalent of 2-(bromomethyl)naphthalene. Removal of the generated
precipitate and the subsequent addition of a second equivalent of 2-(bromomethyl)naphthalene
1
afforded the imidazolium salt 2 in moderate yield. The formation of 2 was confirmed by H
NMR spectroscopy. The most distinctive resonance in the spectrum was that of the C2-H
imidazolium proton which was observed at 9.54 ppm. This was a significant downfield shift
from the C2 proton resonance of 1 (7.73 ppm) and was indicative of cation formation.
Compound 2 was refluxed for 5 days in 9 M aqueous HCl to form the carboxylic acid
1
derivative, compound 3. This transformation was confirmed by H NMR spectroscopy by the
appearance of the carboxylic acid proton resonance at 12.84 ppm and the loss of the proton
resonance at 3.66 ppm corresponding to the methyl group of 2. Additionally, a shift in the C=O
stretching frequency from 1716 cm-1 (2) to 1693 cm-1 (3) was observed via infrared (IR)
spectroscopy.

ACCEPTED MANUSCRIPT
4
Scheme 1. Synthesis of the urocanic acid-derived N,N’-bis(naphthylmethyl)imidazolium salts 2
and 3.
A series of N,N’-bis(naphthylmethyl)imidazolium salts that contain carbamate
substituents at the C4 position were synthesized from histamine (6a-c, Scheme 2). Because
alkylation of histamine is known to occur at all available nitrogen atoms (Nα, Nπ and Nτ),9,10,11
differentiation of these three positions was achieved by converting the histamine to 5,6,7,8-
tetrahydro-5-oxoimidazo[1,5-c]pyrimidine (4) using N,N’-carbonyldiimidazole (CDI). The
combination of 4 with 2-(bromomethyl)naphthalene resulted in the alkylation of the Nτ atom and
the formation of 2-(naphthalen-2-ylmethyl)-5-oxo-5,6,7,8-tetrahydroimidazo[1,5-c]pyrimidinium
bromide (5). The formation of 5 was confirmed by 1H NMR spectroscopy. Again the downfield
shift of the C2 proton, which was observed at 9.99 ppm, was indicative of the formation of the
cationic species. Crystals of 5 suitable for single crystal X-ray diffraction were obtained by slow
evaporation of a concentrated solution of the compound in acetonitrile (see supplementary
information). Refluxing a mixture of 5 and N,N-diisopropylethylamine (DIPEA) with methanol,
n-butanol, or propargyl alcohol resulted in the deprotection of the Nπ atom and produced the
respective carbamates. These carbamates were substituted with 2-(bromomethyl)naphthalene to
yield the corresponding imidazolium salts 6a-c.

ACCEPTED MANUSCRIPT
5
Br
O
Br
Br
O
τ
π
π
τ
π
τ
α
α
NH
N
N
N
NH
N
N
CDI
NH
∆
CH3CN, reflux
4
5
NH2•2HCl
α
1) DIPEA, ROH
2)
τ
π
N
N
Br
a)
R= CH3
b) R= (CH2)3CH3
c)
R= CH2CCH
CH3CN, reflux
α
NH
O
6a-c
O
R
Scheme 2. Synthesis of the carbamate-substituted N,N’-bis(naphthylmethyl)imidazolium salts
6a-c.
The C2-H imidazolium protons of compounds 6a-c were observed at 9.40 ppm, 9.39 ppm,
1
and 9.35 ppm, respectively, by H NMR spectroscopy. IR spectroscopy indicated a shift in the
C=O stretching frequency from 1758 cm-1 (5) to 1709 cm-1 (6a), 1701 cm-1 (6b), and 1709 cm-1
(6c). The emergence of a C≡C stretching frequency at 2114 cm-1 in the spectrum of 6c indicated
the presence of the propargyl group. Crystals of 6b suitable for single crystal X-ray diffraction
were obtained by slow evaporation of a concentrated solution of the compound in a mixture of
acetonitrile and chloroform (1:1) (Figure 3).
Figure 3. Thermal ellipsoid plot of the cationic portion of 6b, with thermal ellipsoids drawn at
the 50% probability level. Hydrogen atoms and bromide anion omitted for clarity.

ACCEPTED MANUSCRIPT
6
In
a
separate
attempt
to
introduce
aqueous
solubility
to
N,N’-
bis(naphthylmethyl)imidazolium salts, the amphiphilic imidazolium salt 4,5-bis(hydroxymethyl)-
1,3-bis(naphthalen-2-ylmethyl) imidazolium bromide (10) was synthesized (Scheme 3).
Commercially available 4,5-imidazoledicarboxylic acid was used to prepare dimethyl imidazole
dicarboxylate (7).12 The reaction of 7 with an equimolar amount of 2-(bromomethyl)naphthalene
yielded dimethyl 1-(naphthalen-2-ylmethyl)imidazole-4,5-dicarboxylate (8). Reduction of 8 with
lithium aluminum hydride in tetrahydrofuran produced the vicinal diol, (1-(naphthalen-2-
ylmethyl)imidazole-4,5-diyl)dimethanol (9). This functional group transformation was indicated
by the loss of the C=O frequency from the IR spectra of 8 (1713 cm-1) to that of 9, as well as the
1
emergence of the expected doublet (CH2) and triplet (OH) resonances in the H NMR spectrum
of 9. Although 9 theoretically could be synthesized from the reaction of 4,5-
imidazoledicarboxylic acid with 2-(bromomethyl)naphthalene and subsequent reduction of the
carboxylic acid functional groups, this route was determined to be less desirable due to the
difficulties anticipated from the extremely low solubility of 4,5-imidazoledicarboxylic acid in
solvents permissible for the reduction reaction. Compound 9 was reacted with a slight molar
excess of 2-(bromomethyl)naphthalene to yield 10. The formation of 10 was indicated by the
1
characteristic downfield shift of the C2-H imidazolium proton resonance in the H NMR
spectroscopy, observed at 9.45 ppm. The structure of 10 was confirmed from characterization by
X-ray crystallography (Figure 4). Single crystals were grown from slow evaporation of a
concentrated solution of acetonitrile and diethyl ether.

ACCEPTED MANUSCRIPT
7
Scheme 3. Synthesis of the dimethanol-substituted N,N’-bis(naphthylmethyl)imidazolium salt
10.
Figure 4. Thermal ellipsoid plot of 10, with thermal ellipsoids drawn to the 50% probability
level. Hydrogen atoms, with the exception of those of the hydroxyl groups, and carbon labels
have been removed for clarity.
Whereas the synthesis of 2, 3, 6a-c, and 10 was aimed to increase the aqueous solubility
of N,N’-bis(naphthylmethyl)imidazolium salts, the synthesis of 4-methyl-1,3-bis(naphthalen-2-
ylmethyl)imidazolium bromide (11) provided a lipophilic imidazolium salt to test against the

ACCEPTED MANUSCRIPT
8
selected NSCLC cell lines. The methyl substituent in the C4 position of 11 increased the
lipophilicity of the imidazolium salt without significantly adding steric bulk. Compound 11 was
synthesized according to an established procedure in our laboratory for analogous lipophilic
imidazolium salts.6 In refluxing acetonitrile, 4-methylimidazole was deprotonated with
potassium hydroxide and alkylated with one equivalent of 2-(bromomethyl)naphthalene. After
removal of the precipitate, a second equivalent of 2-(bromomethyl)naphthalene was added to the
1
refluxing mixture to generate the imidazolium salt 11 (Equation 1). H NMR spectroscopy
indicated the formation of the imidazolium salt, with the downfield shift of the C2-H
imidazolium proton resonance observed at 9.57 ppm for 11. Additionally, crystals of 11 suitable
for single crystal X-ray diffraction were obtained by slow evaporation of a concentrated solution
of the compound in acetonitrile (Figure 5).
Equation 1. Synthesis of the lipophilic N,N’-bis(naphthylmethyl)imidazolium salt 11.
Figure 5. Thermal ellipsoid plot of the cationic portion of 11, with thermal ellipsoids drawn at
the 50% probability level. Hydrogen atoms and bromide anion omitted for clarity.
Substituents were also introduced at the nitrogen atoms (N1 and N3) of the imidazole ring,
rather than at the C4 and C5 positions, to help understand the potential anti-proliferative activity
and aqueous solubility of these imidazolium salts. 4,5-Diphenylimidazole was functionalized
with quinolinylmethyl and (6-methoxynaphthyl)methyl substituents, in addition to the

ACCEPTED MANUSCRIPT
9
naphthylmethyl group to create a series of six 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium
salts 12 – 17. The synthesis of compounds 12 – 17 were executed by the same route used to
obtain the analogous imidazolium salts.6 Compounds 12 – 17 were prepared by stirring 4,5-
diphenylimidazole with a minimum of 1.2 equivalents of potassium hydroxide and one
equivalent of the appropriate alkyl bromide or chloride in refluxing acetonitrile overnight.
Removal of the generated precipitate and the addition of one equivalent of the naphthalene-based
halide in refluxing acetonitrile produced compounds 12 – 17 (Scheme 4).
Scheme 4. Synthesis of the 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium salts 12 – 17.
All 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium salts 12 – 17 were characterized by
1H NMR, 13C NMR, mass spectrometry, elemental analysis and melting point determination. The
1H NMR spectra of compounds 12 – 17 show a representative resonance for the C2-H
imidazolium proton within the range of 9.77 – 9.88 ppm and was a significant downfield shift
from the C2 proton resonance of the 4,5-diphenylimidazole starting material (7.78 ppm). The
number of chemical resonances in the 13C NMR spectra fully support the series of 4,5-diphenyl-
N,N’-bis(arylmethyl)imidazolium salts 12 – 17. ESI-mass spectrometry in the positive mode was
conducted for the salts 12 – 17 and the corresponding [M-Br]+ and [M-Cl]+ of the salts further
suggested compound identification of the desired products. The structures of compound 16 and
17 were determined by single crystal X-ray diffraction, following the slow evaporation of a

ACCEPTED MANUSCRIPT
10
concentrated solution of the compounds in a water/acetone (1:1) mixture and ethanol,
respectively (Figure 6 and Figure 7). Compound 16 co-crystallized with a water molecule and 17
co-crystalized with an ethanol solvent molecule.
Figure 6. Thermal ellipsoid plot of 16 from solvate 16•(H2O). All thermal ellipsoids have been
drawn at the 50% probability level. Hydrogen atoms, carbon labels, and the co-crystalized water
molecule have been removed for clarity.
Figure 7.Thermal ellipsoid plot of 17 from solvate 17•(C2H6O). All thermal ellipsoids have been
drawn at the 50% probability level. Hydrogen atoms, carbon labels, and the co-crystalized
ethanol molecule have been removed for clarity.
2.2 In vitro Anti-tumor Studies
The in vitro anti-tumor activity of compounds 2, 3, 6a-c, 10 and 11 - 17 was evaluated
against the NSCLC cell lines NCI-H460, NCI-H1975, and HCC827 in order to establish two
structure-activity relationships (SAR). The first SAR study considers compounds 2, 3, 6a-c, 10,

ACCEPTED MANUSCRIPT
11
11, and 12. Each of these compounds has different functional groups at the C4 and/or the C5
position(s). These various functional groups contain a wide range of hydrophobic to hydrophilic
functional groups. The second SAR study involves compounds 12 - 17. Each compound is
functionalized with phenyl substituents at both the C4 and C5 positions with varying arylmethyl
groups at the N1 and N3 positions. Previously we have shown that naphthylmethyl substituents at
the N1 and N3 positions yield compounds with high anti-cancer activity. However, these highly
hydrophobic substituents do not allow for sufficient water solubility for systemic administration.
Therefore, the naphthalene substituents were varied by adding a nitrogen heteroatom to the ring
and/or an ether substituent off the ring in attempt to increase water solubility without lessening
the anti-cancer activity.
The optional MTT assay protocol was used to determine the IC50 values (the
concentration of the compound that inhibits 50% of the growth of cells, relative to controls) of
each compound (Table 1). The well-known chemotherapy agent cisplatin was used as a positive
control in each assay with IC50 values of 3
Although IC23 had activity comparable to that of cisplatin, with IC50 values of 5
M (H1975), and 8 M (HCC827), potential use of the compound as a drug is limited by poor
water solubility. Of the urocanic acid-derived imidazolium salts, compound 2 proved to be the
most active against the NSCLC cell lines, with IC50 values of 13 M (H460), 4 M (H1975), and
16 M (HCC827). The conversion of the ester functionality in 2 to the carboxylic acid group in
ꢀM (H460), 10
ꢀM (H1975), and 4
ꢀM (HCC827).
ꢀ
M (H460), 6
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
3 led to a severe reduction in efficacy, as the IC50 values of 3 were > 30 ꢀM for all three cell
lines.
The imidazolium salts 6a-c derived from histamine all showed comparable activity (IC50
values between 8 -14 M) against all three cell lines, with the exception of 6a against the
HCC827 line (IC50 > 30 M). As expected, the least hydrophobic compound of the three, 6a,
ꢀ
ꢀ
was the only compound to fully dissolve in the 1% DMSO aqueous stock solution.
Compound 10, which was designed to have a relatively high aqueous solubility and the
most hydrophilic compound tested, was ineffective against the H460 and HCC827 lines
(IC50 > 30
ꢀM) and showed only moderate activity against the H1975 line (IC50 = 11 ꢀM).
Compound 10, which fully dissolved in the 1% DMSO aqueous stock solution, demonstrated a
lack of significant anti-proliferative activity against the majority of the cell lines, similar to
compound 3. The data from 3 and 10 suggests that the inclusion of polar hydrophilic groups at

ACCEPTED MANUSCRIPT
12
the C4 and C5 positions of imidazole leads to a substantial decrease in anti-proliferative activity
of those compounds.
Supporting this observation, 11 and 12, which did not contain any polar, hydrophilic
substituents, demonstrated high anti-proliferative activity against all three cell lines. Compound
11 was fully soluble in the 1% DMSO aqueous stock solution and had IC50 values of 3
(H460 and H1975) and 4 M (HCC827). Compound 12 had IC50 values of 3 M (H460), < 1
(H1975), and 2
ꢀ
M
M
ꢀ
ꢀ
ꢀ
ꢀ
M (HCC827), displaying activity comparable to that of 11, despite not being
fully soluble in the stock solution. It is also important to note that 11 and 12 had in vitro activity
comparable to cisplatin and IC23.
Compounds 2, 3, 6a-c, 10, 11, and 12 were all substituted with naphthylmethyl groups at
both the N1 and N3 positions and differences between the IC50 values for these compounds
establishes our first SAR. The most active compounds were 11 and 12 which have added
lipophilicity at the C4 and/or C5 positions. Compound 2 with an ester functional group has higher
activity than 3 with a carboxylic acid group. Compounds 6a-c with a carbamate functional group
all had moderate activity while 10 with two hydroxyl groups had the lowest activity observed.
The conclusion from this series of compounds is that added lipophilicity is beneficial to high
anti-cancer activity while highly hydrophilic groups such as hydroxyls severely inhibits activity.
The second SAR study concerns compounds 12-17 which all have phenyl substituents at
the C4 and C5 positions while the arylmethyl groups at the N1 and N3 positions were varied.
Imidazolium salts 12 – 17 displayed IC50 values in the single-digit micromolar or high
nanomolar range against all three of the NSCLC cell lines (IC50 < 1 – 3 µM). The anti-
proliferative activities of the 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium salts 12 – 17 were
an improvement relative to both IC23 and cisplatin. The variants of the naphthylmethyl-based
substituents at the N1 and N3 atoms appeared to have no effect on the IC50 values of the 4,5-
diphenyl compounds. However, the solubility of 17, which contains two quinolinylmethyl
groups, was increased considering this was the only 4,5-diphenyl derivative to be solubilized by
the 1% DMSO solution, though its solubility in pure water is still too low to be systemically
administered. Summarization of the second SAR study suggests that adding heteroatoms to the
naphthylmethyl rings can increase the aqueous solubility and maintain the potent anti-cancer
activity of the more lipophilic derivatives. The combined data from salts 2, 3, 6a-c, 10 and 11 –
17 further affirmed the direct relationship between the hydrophobicity of the substituents and the

ACCEPTED MANUSCRIPT
13
anti-proliferative activity of the imidazolium salts. Hydrophilic substituents at the C4 and C5
positions provided a more pronounced effect to the imidazolium salt’s anti-tumor activity and
aqueous solubility, as compared to the N1 and N3 positions.
Table 1. IC50 values of synthesized N,N’-bis(arylmethyl)imidazolium salts 2, 3, 6a-c, 10 and 11 –
17.
IC50 values (µM)
Compound
NCI-H460 NCI-H1975 HCC827
3
5
10
6
4
8
cisplatin
IC23
2
13
> 30
9
4
16
> 30
> 30
12
14
> 30
4
> 30
11
8
3
6a
11
11
> 30
3
6b
10
11
3
6c
10
11
3
< 1
2
2
12
2
2
13
2
< 1
< 1
< 1
< 1
< 1
2
14
3
15
2
2
16
3
2
17
The National Cancer Institute’s (NCI) Developmental Therapeutics Program (DTP)
agreed to test 11 and 12 in the NCI-60 human cancer cell line screen using a one-dose assay and
a five-dose assay, in tandem with our own MTT viability assays. The NCI-60 human cancer cell
screen contains nine NSCLC cell lines. Results from the one-dose assay for 11 and 12 are
summarized in Table 2. In this assay the compounds were exposed to the 60 cell lines at one
concentration (10 µM). Results were given as a growth percentage of cells treated with 11 and 12
compared to growth of control cells. Compound 11 was lethal against two cell lines tested (HOP-

ACCEPTED MANUSCRIPT
14
92 and NCI-H522) defined as leaving fewer live cells after treatment with 11 than before
treatment began. Combining the results of 11 from all nine cell lines showed an average growth
of 27% meaning 11 significantly slowed the growth of the nine NSCLC cell lines at a
concentration of 10 µM. Compound 12 proved more potent, as it was lethal against all nine cell
lines and had a negative average growth rate of 70.81%.
Table 2. Results from the one-dose assay (10 µM) performed by the DTP. All cell lines in the
table are NSCLC cell lines. Values are percent growth relative to control cells. Negative values
depict the compound was lethal at the concentration tested.
Growth percentage
Compound
A549/ EKVX HOP- HOP-
NCI-
H226
NCI-
NCI-
MCI-
NCI-
H522
ATCC
60.54
62
92
H23 H322M H460
38.14
33.16 -20.86 31.78 27.66
65.66
21.86 -17.86
11
12
-81.83 -46.67 -86.12 -83.90 -75.79 -74.86 -20.79 -78.57 -85.76
In the five-dose assay, compounds were exposed to all 60 cell lines at concentrations of
10 nM, 100 nM, 1 µM, 10 µM, and 100 µM. Results were given as concentrations of 50%
growth inhibition (GI50), total growth inhibition (TGI), and 50% lethal concentration (LC50)
relative to cell growth with no drug added. Results of the five-dose assay for 11 are shown in
Table 3 and Figure 8. The GI50 concentration ranges from the mid nanomolar range (448 nm) to
the low µM range (5.05 µM), the TGI concentration ranges from 2.02 µM to 18.7 µM, and the
LC50 concentration ranges from 6.49 µM to > 100 µM. These results are consistent with the IC50
values determined in our lab and give further evidence that 11 has potent anti-proliferative
effects against NSCLC and is a quality candidate for further studies.

ACCEPTED MANUSCRIPT
15
Table 3. Results of 5-dose study for compound 11. Values are concentration in molarity for the
growth inhibition 50% (GI50), total growth inhibition (TGI), and lethal concentration 50%
(LC50).
Concentration (µM)
Cell line
A549/ATCC
EKVX
GI50
5.05
2.15
1.61
0.448
2.12
1.49
2.87
1.72
1.41
TGI
18.7
9.71
3.80
2.02
5.87
3.97
10.3
3.65
3.06
LC50
53.5
43.8
8.97
6.49
> 100
12.5
33.7
7.75
6.63
HOP-62
HOP-92
NCI-H226
NCI-H23
NCI-H332M
NCI-H460
NCI-H522
Figure 8. Dose chart for compound 11 against nine NSCLC cell lines.
The five-dose results for 12 are shown in Table 4 and Figure 9. The GI50 concentration
ranges from the low nanomolar range (122 nm) to the low µM range (1.41 µM), the TGI
concentration ranges from 1.26 µM to 3.16 µM, and the only LC50 concentration observed was

ACCEPTED MANUSCRIPT
16
for the HOP-92 cell line at a concentration of 3.98 µM. Strangely, the cell growth percentage
was higher at the 100 µM dose than the 10 µM dose for all NSCLC cell lines (Figure 9). No
explanation was given to us by the DTP and we cannot explain this trend. Totality of these
results give further evidence of high anti-cancer activity previously demonstrated by this
compound qualifying it as a prime candidate for further studies. Experimental procedures for the
one-dose and five-dose assays can be found on the DTP website and were based on the
concentration of protein at the examined time points. Results from all other cell lines tested
including figures and the tables provided by the DTP are included in the supplemental
information.
Table 4. Results of 5-dose study for compound 12. Values are concentration in molarity for the
growth inhibition 50% (GI50), total growth inhibition (TGI), and lethal concentration 50%
(LC50).
Concentration (M)
Cell line
A549/ATCC
EKVX
GI50
1.00
TGI
2.48
2.29
2.97
1.26
3.16
1.88
2.50
1.65
1.78
LC50
N/A
N/A
N/A
3.98
N/A
N/A
N/A
N/A
N/A
0.837
1.41
HOP-62
HOP-92
0.122
1.20
NCI-H226
NCI-H23
NCI-H332M
NCI-H460
NCI-H522
0.445
1.20
0.384
0.325

ACCEPTED MANUSCRIPT
17
Figure 9. Dose chart for compound 12 against nine NSCLC cell lines.
2.3 Annexin V Assay
Compounds 11 and 12 were also studied using the Annexin V assay to determine the
mode of cell death that NCI-H460 cells undergo when exposed to these compounds. In cells that
are undergoing apoptosis, phosphatidylserine is translocated to the outer leaflet of the cell
membrane resulting in a loss of both cell integrity and cell asymmetry. At the outer leaflet, the
phosphatidylserine can interact with Annexin V. The Annexin V is conjugated to FITC which
can be visualized by fluorescence microscopy. A secondary stain, propidium iodide, is also used
which fluoresces red when interacting with DNA. This interaction is only seen in cells whose
membrane integrity is lost, as propidium iodide cannot cross intact cell membranes.13
The Annexin V assay employed in these in vitro studies has the ability to distinguish
between apoptosis and necrosis by use of the propidium iodide. As is well known, apoptosis is
programmed cell death whereas necrosis is a disordered form of cell death. The progression over
time of observed fluorescence in conjunction with changes in cell membrane morphology
indicates the specific mode of cell death. In the case of apoptosis, cell membrane integrity is not
immediately compromised as the phosphatidylserine is translocated to the outer leaflet. This
corresponds to an observation of green fluorescence with the absence of any red fluorescence
from the propidium iodide. At longer treatment time points, apoptosis is observed as both green
and red fluorescence with the presence of a blebbing membrane. In contrast to this is necrosis,

ACCEPTED MANUSCRIPT
18
which exhibits both green and red fluorescence independent of the treatment time. This is due to
the swelling and eventual breakdown of the cell membrane, which is permeable to propidium
iodide even at early time points. The progression in time as well as cell morphology is what can
assist in distinguishing these two modes of cell death in this assay. However, due to ambiguity
between identification of cells in the late stage of apoptosis or those that are necrotic, further
studies can be completed in order to validate the Annexin V assay.
Compound 11 is a highly lipophilic compound and has poor solubility in aqueous
solution. Previously, we have described lipophilic imidazolium salts that could be solubilized by
a chemical excipient, 2-hydroxypropyl-
suggested there was no difference in the anti-cancer activity of compounds solubilized by the 1%
DMSO solution described above or solubilization by the vehicle, 2-HP CD. Therefore,
compound 11 was dissolved in a 20% (w/v) solution of 2-HP CD and exposed to cells at a
β-cyclodextrin (2-HPβ
CD).14 In vitro MTT assays
β
β
concentration of 40 µM for 12, 14, 17, and 20 hour time points. At the end of the treatment
periods, the Annexin V apoptosis detection kit was utilized and cells were visualized by
fluorescence microscopy (Figure 10). Images were compared to cells exposed to only the 2-
HPβCD vehicle solution and cells treated with cisplatin. There was minimal fluorescence from
cells treated with vehicle which is expected. A significant amount of fluorescence and blebbing
was observed from cells treated with cisplatin as this compound is known to induce apoptosis.
Signs of apoptosis including green fluorescence of the cell membrane, blebbing, and staining of
the nucleus from propidium iodide were observed at all the time points for cells treated with 11
(Figure 11). However, not all cells were in the same stage of cell death as indicated by the
progression of fluorescence in individual wells. At the 12 hour time point, a fraction of the cells
were visualized with green fluorescence only while some had green and red fluorescence
together and the rest had no fluorescence. At the 17 and 20 hour time point, a significant
percentage of the cells are both green and red fluorescent, show blebbing and thus suggesting
apoptosis. We predicted 11 to induce an apoptotic mode of cell death considering the structure is
similar to previously reported imidazolium salts that also induce apoptosis in NSCLC cells.6

ACCEPTED MANUSCRIPT
19
Figure 10. Images of the Annexin V assay on H460 cells grown in 6-well plates using compound
11 in 20% (w/v) 2-HP CD aqueous solution as the compound treatment. All images were taken
using a 10x objective. Images are presented as a merged image of the normal transmitted light,
green fluorescence and red fluorescence figures. (A) 2-HP CD control, 20 hours. (B)
Compound 11, 14 hours. (C) Compound 11, 20 hours. Scale bars equal 400 m.
β
β
ꢀ

ACCEPTED MANUSCRIPT
20
Figure 11. Images of the Annexin V assay on H460 cells grown in 6-well plates using compound
11 in 20% (w/v) 2-HP CD aqueous solution as the compound treatment. All images were taken
β
using a 20x objective. Images are presented as a merged image of the green fluorescence and red
fluorescence figures, omitting the normal transmitted light image for blebbing clarity. Images A-
C were taken at the 20 hour time point, D-E were taken at the 17 hour time point, F-G were taken
at the 14 hour time point, and images H-I were taken at the 12 hour time point.
Compound 12 was dissolved in a 1% DMSO solution, because it could not be solubilized
by the 2-HPβCD vehicle solution, and H460 cells were exposed at a concentration of 40 µM for

ACCEPTED MANUSCRIPT
21
1, 3, 6, and 12 hour time points. The vehicle control and cisplatin positive control also contained
the same percentage of DMSO as the cells treated with 12. Again, the Annexin V assay detection
kit was utilized. Minor green fluorescence was observed with cells treated with vehicle, while
early signs of apoptosis are observed in cells treated with cisplatin as would be expected at the
12 hour time point (Figure 12). Strong green fluorescence is observed as early as the 1 hour time
point and continues to be observed at all of the later time points in cells treated with 12.
However, cells treated with 12 were abnormal when compared to the controls, especially in that
the morphology of the membrane is vastly different. No blebbing was observed at any of the
time points and green fluorescence as well as red fluorescence were observed at all time points.
This suggests that compound 12 induces a necrotic cell death pathway versus an apoptotic cell
death pathway as observed with 11 (Figure 13). While observation of membrane morphology
and the progression of fluorescence can give great insight into the mode of cell death, future
work will also incorporate other assays that can confirm the findings of the Annexin V assay.
Figure 12. Images of the Annexin V assay on H460 cells grown in 6-well plates using 12
dissolved in a 1% DMSO aqueous solution as the compound treatment. All images were taken
using a 20x objective. Images are presented as a merged image of the normal transmitted light,
green fluorescence and red fluorescence figures. (A) DMSO control, 12 hours. (B) Compound
12, 12 hours. Scale bars equal 200 ꢀm

ACCEPTED MANUSCRIPT
22
Figure 13. Image depicting the comparison of compound 11 versus compound 12 treated cells.
(Row A) H460 blebbing cells after treatment with compound 11 (normal transmitted light image
omitted for blebbing clarity). (Row B) H460 cells after treatment with compound 12.
The time frame differences of the Annexin V studies on 11 and 12 arise from the use of
different solubilizing agents. We have noted previously that the use of 2-HPβCD does not
change activity by MTT assay analysis. However, the treatment period and compound
concentrations used for the MTT assay are greatly different from those used for the Annexin V
assay. Previous studies have shown that at the same concentration, solubilization in 2-HP
delays the effects of imidazolium salts which we believe comes from the equilibration time
needed for the compound to be released from the lipophilic core of 2-HP CD before it can exert
βCD
β
its effect on the cancer cells.14 This equilibration time leads to each cancer cell being at a
different stage of cell death when imaging occurs. This can be seen where some cells have no
fluorescence while others have either green fluorescence or green paired with red fluorescence
(see Suppmental Information).
2.4 In vivo toxicity study
Considering 11 had high anti-cancer activity and it could be completely solubilized in
aqueous solution by 2-HP
βCD at high concentrations (whereas 12 could not be solubilized by 2-
HP CD), 11 was chosen for a preliminary analysis of the in vivo toxicity using C57BL/6 mice.
β
Six week old C57BL/6 male mice were allowed to acclimate in their cages (n = 3 mice per
group, each group housed in separate cages) for 5 days. Following acclimation, the control group

ACCEPTED MANUSCRIPT
23
received 100 µL of vehicle (20% (w/v) 2-HPβCD) solution by intraperitoneal (IP) injection
while the experimental group received 100 µL of a 20 mg/kg dose of 11 (assuming an average
weight of 20 g) dissolved in the vehicle solution by IP injection on day 0. The weights and health
of all animals were closely monitored for the next 6 days (Figure 14). There was a steady
increase in weight for all control animals. All animals survived the initial injection of 11, but a
sharp decrease in weight occurred for animals treated with 11. However, this was followed by
complete weight recovery and weight gain by 66% of animals by day six. Animals in both
groups were again injected on days 0, 7, 14, 21, 25, and 29 with similar patterns concerning
weight loss and gain. All mice survived for the duration of the study and the last injection, on
day 29, was given one hour before perfusion.
Figure 14. Weight chart for C57BL/6 mice injected with 11 dissolved in vehicle solution. Red
arrows signify days the animals were injected.
Selected reaction monitoring mass spectrometry (SRM-MS) studies were performed on
extraction samples from the brain, heart, lung, liver, kidney, and spleen of all mice. The highest
concentration of 11 was found in the liver, followed closely by the kidney and spleen (Figure
15). Accumulation was also observed in the heart, lung, and brain but at lower concentrations.

ACCEPTED MANUSCRIPT
24
Minimizing side effects is a major goal of potential chemotherapeutics considering current
treatments such as cisplatin have severe side effects, especially renal toxicity. Although 11 did
show accumulation in the kidney, collectively, 11 was less cytotoxic against the panel of renal
cancer cells than the panel of lung cancer cells tested in the NCI-60 cell line assays (see
supplemental information). Results from the in vivo toxicity study of compound 11 were very
promising. All animals survived and although animals lost weight after each injection, they were
able to fully recover during rest periods in-between injections. Compound 11 is a promising
candidate to move on to further in vivo studies with lung cancer models.
1 10- 0 3
×
8 10- 0 4
×
6 10- 0 4
×
4 10- 0 4
×
2 10- 0 4
×
8.0 10- 6
×
4.0 10- 6
×
0
Figure 15. Bar graph depicting the concentration of 11 deposited in the brain, lung, liver, kidney,
heart, and spleen as determined by selective reaction monitoring.
3. Conclusion
In this work, two in vitro SAR studies were performed to establish a route to increase the
solubility of highly active imidazolium salts without reducing the activity. The first SAR study
included a series of N,N’-bis(naphthylmethyl)imidazolium salts with varied hydrophilic and
hydrophobic substituents at the C4 and C5 positions. Compounds 2 and 6a-c, all containing
moderately hydrophobic substituents in the C4(C5) position(s), displayed IC50 values between 8 –
16
ꢀM, with the exception of 6a against the HCC827 line. Although less efficacious than
cisplatin, these values are in a clinically achievable range and compounds 2 and 6a were fully
soluble in the 1% DMSO aqueous stock solution. In addition, the propargyl group of 6c provides

ACCEPTED MANUSCRIPT
25
a site for the potential coupling of targeting and/or imaging moieties bearing an azide group
through ‘click chemistry’.15 The modification would presumably enhance the effectiveness of the
compound in vivo.
The compounds with highly polar, hydrophilic groups, 3 and 10, showed almost no
significant anti-proliferative effects. With the exception of 10 against the H1975 line, all IC50
values for these complexes were > 30
and 12 with highly hydrophobic substituents demonstrated activity comparable to that of
cisplatin, with IC50 values of 4 M or less against all cell lines tested. Results from the in vitro
ꢀM against the three cell lines. Conversely, compounds 11
ꢀ
MTT assays on compounds 2, 3, 6a-c, 10, 11, and 12 strongly suggests that the hydrophobic
substituents are necessary for high anti-proliferative activity against these NSCLC cell lines.
Compound 11 and 12 were also studied by the NCI’s DTP with the NCI-60 human cancer cell
line panel. Both compounds were active against the nine NSCLC lines tested while 12 was
highly cytotoxic against almost all 60 cell lines. Compound 11 was shown to induce cell death
by an apoptotic pathway while 12 induced a necrotic cell death pathway. Considering 11 had
high solubility when dissolved in an aqueous 2-HP
was chosen for an in vivo toxicity study where it showed a favorable side effect profile.
Although compound 11 was solubilized by 2-HP CD, the aqueous solubility of potential
βCD solution and high-anticancer activity, it
β
therapeutics is still of high importance, and future work should focus on increasing the water
solubility of compounds by means other than inclusion of highly polar substituents, while still
maintaining anti-proliferative activity.
The second SAR study involved the 4,5-diphenyl-N,N’-bis(arylmethyl)imidazolium salts
12 – 17. The heteroatoms added to the aryl groups in 13-17 proved to add little solubility to the
highly lipophilic derivative 12. However, 17, with two quinolylmethyl moieties, was solubilized
in the 1% DMSO aqueous solution. Unfortunately, its solubility was limited without the use of
the DMSO vehicle. This second SAR study suggests that heteroatoms can be added to the aryl
groups at the N1 and N3 positions without reducing activity. Therefore, future derivatives will
focus on utilizing the quinolylmethyl and methoxynaphthylmethyl groups to increase solubility.
It may also be possible to increase solubility by protonating the nitrogen of quinoline
substituents.16 However, it is unknown if that will alter the anti-cancer activity of the compound
and resulting pharmacokinetics considering potential differences in pH. Imidazolium salts with a
combination of satisfactory solubility and high anti-proliferative activity are still actively

ACCEPTED MANUSCRIPT
26
investigated and future work should concentrate on substituents with multiple heteroatoms or
other hydrophilic groups.
4. Experimental
4.1 General Considerations
All reactions were carried out under aerobic conditions unless otherwise specified. All acids,
bases and solvents were purchased from Fisher Scientific and used without further purification.
All reagents were used as received without further purification. Histamine dihydrochloride,
urocanic acid, N,N’-carbonyldiimidazole (CDI), N,N–diisopropylethylamine (DIPEA), 4,5-
diphenylimidazole, 4,5-imidazoledicarboxylic acid, 4-methylimidazole and primary alcohols
were purchased from Alfa Aesar. 2-(Bromomethyl)naphthalene was purchased from Waterstone
Technologies. 2-Hydroxypropyl-β-cyclodextrin (average molecular weight: 1400) was purchased
from Tocris (a Biotechne brand). The TACS MTT cell proliferation assay kit was purchased
from Trevigen. All C57BL/6 mice were purchased from Charles River Laboratories and housed
in the biology department at the University of Akron. The FITC annexin V apoptosis detection
kit was purchased from BD Pharmingen. HPLC grade (≥99.9%) acetonitrile, methanol, and
water used for extractions and LC-MS analysis were purchased from Fisher Scientific (Fair
Lawn, NJ, USA). 2-(Chloromethyl)-6-methoxynaphthalene was synthesized according to
literature procedure.17 Melting points were obtained on a MelTemp apparatus. Methyl 3-
(imidazol-4-yl)acrylate (1) was synthesized according to literature procedures and used without
further purification.7,8 Dimethyl imidazole-4,5-dicarboxylate (7) was synthesized according to
literature procedure and used without further purification.12 Column chromatography utilized
silica gel (60 Å, ICN Medicals) embedded with a green 254 nm fluorescent indicator (Fluka
Analytical). Melting points were obtained on a MelTemp apparatus. 1H and 13C nuclear magnetic
resonance (NMR) spectra were collected on a Varian 300 or 500 MHz instrument referenced to
DMSO-d6 (2.50 ppm, 39.51 ppm) or CD3OD-d4 (3.31 ppm, 49.00 ppm, respectively).
Electrospray ionization mass spectrometry (ESI-MS) or high resolution mass spectrometry (ESI-
HRMS) was performed by the University of Akron mass spectrometry laboratory. Elemental
analysis was performed by the microanalysis laboratory at Atlantic Microlabs (Atlanta, GA) and
the University of Akron Department of Geology.
The human non-small cell lung cancer cell lines NCI-H1975 and HCC827 were donated
by Dr. Daniel Lindner from the Cleveland Clinic. The human non-small cell lung cancer cell line

ACCEPTED MANUSCRIPT
27
NCI-H460 was purchased from ATCC (Manassas, VA, USA). All cell lines were grown in
RPMI 1640 media supplemented with 10% fetal bovine serum. All cell lines were grown under
physiological conditions, specifically at 37 °C with 5% CO2 and passed every 2-3 days.
4.2 X-ray Structure Determination Details
Crystals of the compounds were coated in paratone oil, mounted on a CryoLoop and
placed on a goniometer under a stream of nitrogen. Crystal structure data sets were collected on
either a Bruker SMART APEX I CCD diffractometer with graphite-monochromated Mo K
radiation ( = 0.71073 Å) or a Bruker Kappa APEX II Duo CCD system equipped with a Mo
ImuS source and a Cu ImuS micro-focus source equipped with QUAZAR optics ( = 1.54178
α
λ
λ
Å). The unit cells were determined by using reflections from three different orientations. Data
sets were collected using SMART or APEX II software packages. All data sets were processed
using the APEX II software suite.18,19 The data sets were integrated using SAINT.20 An empirical
absorption correction and other corrections were applied to the data sets using multi-scan
SADABS.21 Structure solution, refinement, and modelling were accomplished by using the
Bruker SHELXTL package.22 The structures were determined by full-matrix least-squares
refinement of F2 and the selection of the appropriate atoms from the generated difference map.
Hydrogen atom positions were calculated and Uiso(H) values were fixed according to a riding
model.
4.3 MTT Assay Protocol
Cells were grown to confluency and plated in 96-well plates at 5,000-7,000 cells per well.
Cells were incubated for 24 h prior to adding the compounds. All compounds were dissolved in a
1% DMSO solution and diluted in fresh media to the desired concentrations of 1, 4, 16, and 32
µM. Compounds were added in sextuplet and cells were incubated for 72 h at which time the
optional MTT assay protocol was followed. MTT reagent (10 L) was added to each well and
µ
cells were incubated for 3-4 h depending on the cell line. Medium was removed by vacuum and
DMSO (100 µL) was added to each well. Plates were incubated for 15 min at 37 °C. The optical
density was read at 540 nm on a BioTek Epoch plate reader.
4.4 Annexin V assay protocol

ACCEPTED MANUSCRIPT
28
NCI-H460 cells were grown to confluency and plated in 6-well plates at a concentration
of 100,000 cells per well. Cells were incubated for 24 hours. At this time, medium was removed
by vacuum and fresh medium with vehicle, cisplatin, 11, or 12 was added. Cisplatin, 11, and 12
were dissolved at a concentration of 40 µM. Compound 11 was exposed to cells for 12, 14, 17,
and 20 hours, while the cisplatin and vehicle control were exposed to cells for 20 hours.
Compound 12 was exposed to cells for 1, 3, 6, and 12 hours, while the cisplatin and vehicle
control were exposed to cells for 12 hours. At the end of the treatment period, medium was
removed and cells were washed twice with cold PBS (1.5 mL). 1X binding buffer (400 µL),
FITC annexin V (20 µL), and propidium iodide (20 µL) was added to each well and plates were
incubated in the dark for 20 minutes at room temperature. The binding buffer was removed and
fresh 1X binding buffer (1 mL) was added for imaging. Cells were imaged on an EVOS fl
Digital Inverted Microscope with 10X and 20X objectives.
4.5 In vivo toxicity study protocol
All animal procedures were reviewed and approved by the Institutional Care and Use
Committee at the University of Akron. Eight week old male C57BL/6 mice were obtained from
Charles River laboratories. Animals were housed in a 12h light/dark cycle and food and water
were provided ad libitum. Prior to the toxicity testing, animals were allowed to acclimate for 5
days. Mice received injections on days 0, 7, 14, 21, and 25 that consisted of a 100 µL solution
of a 20% 2HPβCD sterile PBS solution (control) or 100 µL of a 4 mg/mL 20% 2HPβCD sterile
PBS solution (0.4 mg/ 100 µL or ~ 20 mg/kg assuming 20 g mice) of 11 by IP injection. Animals
were closely monitored and weighed on a daily basis. On day 29 animals were sacrificed and the
lung, heart, liver, kidney, spleen, and brain were harvested and frozen for later use.
4.6 Tissue extraction and selective reaction monitoring method
To extract and analyze compound deposited in the collected organs, cold methanol (0.2
mL) was added to a centrifuge tube containing the tissue sample of interest. The sample was
vortexed, frozen in liquid nitrogen, thawed at room temperature, and sonicated. This series of
steps was repeated twice more. The samples were frozen at -20 °C for one hour and centrifuged
at 10,000 rpm for 5 minutes at 4 °C. The supernatant was separated from the protein pellet. The
protein pellet was extracted by the same procedure outlined above and the resulting supernatants