Synthesis and anticancer activity of novel water soluble benzimidazole carbamates

Article abstract of DOI:10.1016/j.ejmech.2017.11.037

Metastases account for more than 90% of all cancer deaths and respond poorly to most therapies. There remains an urgent need for new therapeutic modalities for the treatment of advanced metastatic cancers. The benzimidazole methylcarbamate drugs, commonly

Full text of DOI:10.1016/j.ejmech.2017.11.037

Accepted Manuscript
Synthesis and anticancer activity of novel water soluble benzimidazole carbamates
Jae Eun Cheong, Michela Zaffagni, Ivy Chung, Yingjie Xu, Yiqiang Wang, Finith E.
Jernigan, Bruce R. Zetter, Lijun Sun
PII:
S0223-5234(17)30940-6
10.1016/j.ejmech.2017.11.037
EJMECH 9912
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 9 August 2017
Revised Date: 6 November 2017
Accepted Date: 17 November 2017
Please cite this article as: J.E. Cheong, M. Zaffagni, I. Chung, Y. Xu, Y. Wang, F.E. Jernigan, B.R.
Zetter, L. Sun, Synthesis and anticancer activity of novel water soluble benzimidazole carbamates,
European Journal of Medicinal Chemistry (2018), doi: 10.1016/j.ejmech.2017.11.037.
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
Graphic abstract
1

ACCEPTED MANUSCRIPT
Synthesis and anticancer activity of novel water soluble benzimidazole carbamates
Authors:
#
1
≠1
≠
#
Jae Eun Cheong , Michela Zaffagni , Ivy Chung^, Yingjie Xu , Yiqiang Wang , Finith E.
#
≠
*
#*
Jernigan , Bruce R. Zetter , Lijun Sun
Affiliations:
#
:
Center for Drug Discovery and Translational Research, Beth Israel Deaconess Medical Center,
Harvard Medical School, Boston, MA 02215, USA
≠
:
Vascular Biology Program, Boston Children’s Hospital, Harvard Medical School, Boston,
MA 02215, USA
^
Department of Pharmacology, Faculty of Medicine, University of Malaya, Kuala Lumpur,
Malaysia
*
: Corresponding authors: bruce.zetter@childrens.harvard.edu (BZ), lsun1@bidmc.harvard.edu
(
LS)
1
These authors contributed equally to this report.
1

ACCEPTED MANUSCRIPT
Abstract
Metastases account for more than 90% of all cancer deaths and respond poorly to most therapies.
There remains an urgent need for new therapeutic modalities for the treatment of advanced
metastatic cancers. The benzimidazole methylcarbamate drugs, commonly used as anti-helmitics,
have been suggested to have anticancer activity, but progress has been stalled by their poor water
solubility and poor suitability for systemic delivery to disseminated cancers. We synthesized and
characterized the anticancer activity of novel benzimidzoles containing an oxetane or an amine
group to enhance solubility. Among them, the novel oxetanyl substituted compound 18
demonstrated significant cytotoxicity toward a variety of cancer cell types including prostate,
lung, and ovarian cancers with strong activity toward highly aggressive cancer lines (IC :
5
0
0
.9-3.8 µM). Compound 18 achieved aqueous solubility of 361 µM. In a mouse xenograft model
of a highly metastatic human prostate cancer, compound 18 (30 mg/kg) significantly inhibited
the growth of established tumors (T/C: 0.36) without noticeable toxicity.
Key words: Anticancer activity; benzimidazole; solubility; oxetane
2

ACCEPTED MANUSCRIPT
1
. Introduction
Despite continued improvement in cancer therapy, there are few effective approaches for the
treatment of metastatic cancers. Greater than 90% of all cancer deaths can be attributed to
metastatic disease yet very few cancer drugs are developed through screens devoted to metastatic
cancer. For many cancers, little progress has been made in extending the lives of patients with
metastatic disease. It is estimated that in the US alone, 150,000 lung cancer patients and 26,000
prostate cancer patients will die from treatment-resistant metastasis in 2017. The
molecular-targeted drugs are initially effective in treating metastatic lung cancers, but their
benefit is limited by the rapid development of resistance and the improvement in lifespan is often
still modest [1-4]. Further, the lack of efficacy of kinase inhibitor therapies in prostate cancer is
likely due to the fact that activating tyrosine kinase mutations or amplifications are uncommon in
castrate-resistant prostate cancer (CRPC) [5, 6]. Chemotherapies such as taxanes and androgen
antagonists are effective in treating CRPC, but their benefit is limited by the emergence of
treatment induced resistance. The development of safe and efficacious novel therapies that can
overcome drug resistance holds great promise for metastatic cancer patients who are refractory to
existing treatments.
Besides molecular targeted therapies that form the cornerstone of modern cancer precision
medicine, repurposing of drugs used in other contexts has identified several novel approaches to
cancer therapy, notably the use of thalidomide and its more recent analogs to treat multiple
myeloma [7, 8]. The aggressiveness of late stage cancers is often driven by numerous orthogonal
and compensatory signaling pathways that collectively promote metastatic tumor growth. The
lack of a defined driver mutation renders many types of metastatic cancers as ‘undruggable’ via
targeted therapies. Conversely, unbiased testing of approved drugs can reveal novel drug families
3

ACCEPTED MANUSCRIPT
that can treat complex, metastatic cancers.
One important premise underlying drug repurposing is that the safety profile of approved
drugs has been comprehensively characterized through many decades of clinical uses. Such
efforts in drug repurposing have led to the identifications of anticancer effects by a number of
commonly prescribed drugs [9]. The in vitro and in vivo anticancer activity of the broad
spectrum anti-fungal drug itraconazole was demonstrated in medulloblastoma, glioblastoma, and
non-small cell lung cancer (NSCLC) [10-12]. Another class of drugs that has emerged as
potential anticancer agents is the anthelmintic benzimidazole drugs, which have been safely used
for decades in the clinic for the treatment of gastrointestinal infections. The anthelmintic
benzimidazole derivatives were found to exhibit potent inhibitory activity in brain tumor, colon
cancer, breast cancer, leukemia, and myeloma [13-21]. One unique property of these
benzimidazole drugs is their extremely low solubility that impedes their preclinical assessment
via dosing regimens appropriate for clinical translation in the context of treatment of
disseminated cancers. Herein we report our phenotypical screening approaches that identified the
unique cytotoxicity profiles of the benzimidazoles in prostate cancer cell lines with high
metastatic potentials as well as in lung cancer cells that are resistant to paclitaxel. We describe
here, our novel structural modifications to improve the systemic anticancer activity of these
compounds, the structure-activity relationship (SAR), and the discovery of a novel, water soluble
oxetane-containing benzimidazole coumpund 18, with potent anticancer activities in highly
aggressive prostate, lung, and ovarian cancers.
2
. Results and discussion
2
.1. Cytotoxicity of anthelmintic benzimidazoles in aggressive prostate cancer cell lines
4

ACCEPTED MANUSCRIPT
We first confirmed the in vitro anticancer activity of anthelmintic benzimidazole drugs using
the prostate cancer cell line PC3M, a metastatic clone of the PC3 cell line, obtained by in vivo
selection; and the PC3MLN4 cell line, an even more aggressive clone of PC3 obtained by
recovering the cells from lymph nodes of mice after orthotropic injection into the mouse prostate
(4 cycles) [22]. The original PC3 cell line was derived from bone metastasis of a prostate cancer
and resistant to androgen depletion therapy [23, 24]. Compounds were analyzed for cytotoxicity
after 48 hours (hr) drug treatment using Cyquant assay [25]. Our results confirmed the in vitro
anticancer activity for several benzimidazoles, including albendazole, fenbendazole,
flubendazole, and mebendazole) which, surprisingly, consistently demonstrated about two-fold
higher anti-proliferative activity in the more highly aggressive PC3MLN4 cells (EC 0.6-0.9 µM)
5
0
than the less aggressive PC3M cells (EC 1.2-2.6 µM) (Table 1). There seemed to be rather
5
0
modest effect on the cytotoxicity by substituents (R-W) on the phenyl ring of the benzimidazole
moiety. Thus, the EC of mebendazole and fenbendazole (which differ from each other by the
5
0
linker -W- group), and that of mebendazole and flubendazole (which differ from each other by
the addition of a fluoro group in the latter), is very similar in both the PC3MLN4 and the PC3M
cancer cell lines. Even with more drastic changes in the chemical structures between albendazole
(R: n-propyl) and fenbendazole (R: phenyl), the cytotoxicity of the two drugs does not differ
much in both cell lines. We therefore reasoned that it would be plausible to achieve
improvements in drug-like properties by modifying the R-W moiety without significantly
compromising the anticancer activity of new benzimidazoles (vide infra).
5

ACCEPTED MANUSCRIPT
Table 1. Cytotoxicity of anthelmintic benzimidazole in prostate cancer cell lines
EC (µM)
50
Drug name
R
W
PC3MLN4
0.67
PC3M
1.43
1.15
2.60
1.21
Flubendazole
Albendazole
Fenbendazole
Mebendazole
4-Fluorophenyl
n-Propyl
Phenyl
C=O
S
0.58
S
0.91
Phenyl
C=O
0.70
2
.2. Drug design and chemistry
The benzimidazole anthelmintic drugs were originally developed for the topical treatment of
parasitic infections in the gastrointestinal track. Because of their extremely poor aqueous
solubility [26, 27] and low bioavailability [28-30], they are poorly suitable for systemic delivery
in the setting of treating metastatic cancers. Increases in drug solubility can have a profound
impact on drug pharmacokinetics/pharmacodynamics (PK/PD) and drug safety profiles [31, 32].
The solubility of organic molecules can be improved via incorporation of polar functional groups
such as an amine. In addition, we were particularly interested in the 4-membered cyclic oxetane
moiety, as it has been shown to dramatically increase water solubility and metabolic stability [33,
3
4]. We therefore pursued multiple structural modification strategies aimed to improve solubility
and maintain anticancer activities of novel compounds containing the benzimidazole carbamate
core. As outlined in Figure 1, one approach was to replace the methyl (CH ) group of the
3
6

ACCEPTED MANUSCRIPT
carbamate moiety with a solubility enhancing polar group (A); a second approach was to
incorporate such a group on the phenyl moiety on the left-hand side of the molecule (B).
Figure 1. The right-hand side methyl group (highlighted by a circle) and the left-hand side moiety (highlighted by a
rectangle) were modified to increase drug like properties. The linker W was determined by synthetic feasibility.
To pursue the approach A, a trans-esterification protocol was applied for the synthesis of new
derivatives of the four active drugs albendazole, fenbendazole, flubendazole, and mebendazole
where the right-hand side methylcarbamate was modified by incorporating a solubility enhancing
moiety. The treatment of the benzimidazole methylcarbamates with an amino-ethanol,
amino-propanol, or oxetanol in the presence of a base, resulted in the isolation of the desired
products 1-10, as crystalline solid after silica gel chromatographic purification (Scheme 1). The
preparation of compounds 5 and 7 was facilitated by using microwave (MW) heating. Although
the reaction was straightforward and successful in producing the desired products, the reaction
condition was not optimized (yield: 10-30%) and confounded by the formation of the
corresponding 2-aminobenzimidazole as a major side product.
7

ACCEPTED MANUSCRIPT
Scheme 1. Conditions: 1 (R
n-propylthio; R = N,N-dimethylaminopropyl): pyridine, 100 °C, 3 hr; 3 (R
0 °C, o/n; 4 (R = n-propylthio; R = (3-methyl-oxetan-3-yl)methyl): TEA, 90 °C, o/n; 5 (R
-oxetanyl): TEA/DMF, MW, 15 min; 6 (R = 4-fluorobenzoyl; R = N,N-dimethylaminopropyl): pyridine, 100 °C,
1
= n-propylthio; R
2
= N,N-dimethylaminoethyl): pyridine, 100 °C, 3 hr; 2 (R
= n-propylthio; R = 3-oxetanyl): TEA,
= benzoyl; R
1
=
2
1
2
9
3
3
1
2
1
2
=
1
2
hr; 7 (R = phenylthio; R = 3-oxetanyl): TEA/DMF, MW, 15 min; 8 (R = benzoyl; R = N,N-dimethylaminoethyl):
1
2
1
2
TEA/DMF, 110 °C, o/n; 9 (R = 4-fluorobenzoyl; R = N,N-dimethylaminoethyl): TEA/DMF, 110 °C, o/n; 10 (R =
1
2
1
phenylthio; R2
=
N,N-dimethylaminoethyl): TEA/DMF, 110 °C, o/n; TEA: triethylamine; DMF:
N,N-dimethylformamide; o/n: overnight; hr: hour; min: minute; MW: microwave.
The modification of the left-hand side of the benzimidazole (approach B), by installing an
amino or oxetanyl group, required de novo synthesis for each designed target molecule. We first
synthesized new compounds containing an amino moiety aimed to improve solubility (Scheme
2
). The intermediate compound 11 was prepared via a 2-step synthesis: palladium catalyzed
amination [35] of 3-bromophenol with N-methylpiperazine was followed by nucleophilic
substitution by 5-chloro-2-nitroaniline to install the ether linkage. The synthesis of intermediate
compound 12 began with the nucleophilic substitution reaction between 4-amino-3-nitrophenol
and 2,6-dichloropyridine to install the ether linkage, which was followed by the nucleophilic
substitution reaction by N-methylpiperazine to incorporate the amino group. Next, the nitro
group in intermediates 11 and 12 was reduced and the resulting 1,2-diamines were treated with
1
,3-bis(methoxycarbonyl)-2-methyl-2-thiopseudourea to form the desired benzimidazoles 13 and
4 [36]. Our attempts to incorporate a carbonyl (-CO-) linker, instead of the ether (-O-) linker as
1
8

ACCEPTED MANUSCRIPT
in compounds 13 and 14, failed to yield the desired compounds due to decomposition during
their multi-step syntheses. This strategy was not further pursued, because the oxetanyl
derivatives afforded not only more potent anticancer activities but also dramatic improvement in
aqueous solubility (vide infra).
Scheme 2. Conditions: (a) N-methylpiperazine, Pd
-chloro-2-nitroaniline, NaH, DMF, 120 °C; (c) 4-amino-3-nitrophenol,
N-methylpiperazine, reflux; (e) zinc, AcOH; (f) 1,3-bis(methoxycarbonyl)-2-methyl-2-thiopseudourea, AcOH,
5 °C.
2
(dba)
3
,
t-BuONa, BINAP, toluene, 120 °C; (b)
5
K
2
CO DMF, 120 °C; (d)
3
,
8
We next developed an eight-step synthetic procedure in order to incorporate an oxetane ring
to the left-hand side of the benzimidazole scaffold (Scheme 3). The starting material
,4-dinitrobenzoic acid was first converted to its acyl chloride and then carried forward for the
Friedel-Crafts acylation of N-methylacetanilide. The syntheses were successfully scaled up to
00 mmol batch size and afforded the para- and ortho- regioisomers as a 4:1 mixture. After the
3
6
acetyl protected amine was liberated by HCl mediated hydrolysis, the para-isomer 15 was
isolated in 26% yield (overall from 3 steps). Attempts to install the oxetane moiety to
intermediate 15 via reductive amination with 3-oxetanone were unsuccessful, likely due to the
9

ACCEPTED MANUSCRIPT
weak basicity of the amine moiety (pK : 3.5) that was deactivated by the electron-withdrawing
a
acyl group. We therefore reduced the ketone functional group in compound 15 to its
corresponding alcohol, leading to increases of basicity of the amine (pK : 4.6), and then
a
subjected it to reductive amination with 3-oxetanone to successfully generate the oxetane
containing intermediate 16. A 62% yield was obtained for the reductive amination step under
microwave irradiation at 100 °C. The ketone functionality was next reinstated via Dess-Martin
oxidation [37], and the nitro groups were reduced via palladium catalyzed hydrogenation to give
the diamine intermediate 17 in 89% yield. The final product methyl
(
5-(4-(methyl(oxetan-3-yl)amino)benzoyl)-1H-benzo[d]imidazol-2-yl)carbamate
obtained in 68% yield via microwave assistant condensation cyclization of
,3-bis(methoxycarbonyl)-2-methyl-2-thiopseudourea and the intermediate 17. Finally, the
(18)
was
1
regioisomeric compound 19, in which the oxetanyl moiety is ortho to the carbonyl group, was
synthesized similarly from the ortho-isomer of intermediate 15.
Scheme 3. Conditions: (a) (COCl) , DCM, DMF (cat.), 0 °C to rt; (b) AlCl , N-methylacetanilide, 100 °C; (c) HCl
2
3
(
conc.), H O; (d) NaBH , MeOH; (e) 3-oxetanone, AcOH, NaBH (CN), MeOH, MW, 30 W; (f) Dess-Martin
2 4 3
periodinane, DCM; (g) Pd/C (10 wt %), H , MeOH; (h) 1,3-bis(methoxycarbonyl)-2-methyl-2-thiopseudourea,
2
MeOH, MW, 30 W.
1
0

ACCEPTED MANUSCRIPT
2
.3. In vitro anticancer activities and structure-activity relationship (SAR)
The newly synthesized compounds were first evaluated for their cytotoxicity at 1 and 10 µM
concentrations in the non-small cell lung cancer (NSCLC) cell line A549, and the prostate cancer
cell lines PC3 and PC3MLN4 (Table 2). Mebendazole, which we have confirmed to have
anticancer activity in the PC3MLN4 cell line, was included as a reference compound for
comparison. Cancer cells were treated with the compounds for 72 hr. Cell viability was measured
by the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, which is
dependent upon the cellular metabolic activity of NAD(P)H flux. Compounds that inhibit the
proliferation of rapidly dividing cancer cells will reduce the rates of MTT reduction. In at least two
of the three cancer cell lines, the novel compounds 1, 10, 13, 18, and 19 induced significant
cytotoxicity at 10 µM concentration (cell viability: < 50%) while the other 9 compounds were
much less active. Further, the compound 18, a para-oxetanyl derivative of mebendazole, was the
most potent compound and showed comparable activity to mebendazole at both 1 and 10 µM
concentrations in all three cell lines. In comparison, its ortho-oxetanyl isomer 19 showed
significant cytotoxicity only at 10 µM concentration. It is worth noting that modifications of the
carbamate moiety lead to compounds (1 - 10) with reduced activity as compared with their parent
methyl carbamates (e.g. mebendazole). Among them, only the N,N-dimethylaminoethyl
derivative of albendazole (1) and that of fenbendazole (10) consistently showed cytotoxicity at 10
µM, but not at 1 µM concentration. The N,N-dimethylaminopropyl derivatives of albendazole (2)
and flubendazole (6) did not show measurable cytotoxic activity. Importantly, and in contrast to
compound 18, incorporating an oxetane moiety onto the right-hand side of the benzimidazole
resulted in compounds with diminished activity (3 - 5, 7). Neither the N-methylpiperazine analogs
1
1

ACCEPTED MANUSCRIPT
1
3 nor 14 were found highly active in the three cancer cell lines, although compound 13 showed
activity at 10 µM concentration. Taken together, our SAR results indicate that the benzimidazole
molecules require the right-hand side methyl carbamate moiety to maintain high anticancer
activity and are sensitive to modifications on the left-hand side of the structure. Significantly,
incorporation of an oxetane moiety on the left-hand phenyl ring and para to the ketone functional
group in mebendazole led to the observed high activity as demonstrated by compound 18.
Table 2. Cytotoxicity of benzimidazole compounds at 1 and 10 µM.
a
No.
R1
R2
% Viable Cells (± SD)
PC3
PC3MLN4
A549
1
0 µM
1 µM
10 µM
1 µM
10 µM
51.81
1 µM
1
2
42.84
92.65
37.85
82.46
75.48
±
6.59 ± 10.13 ± 3.18 ± 3.96 ± 11.86 ± 6.91
88.61
86.65
78.25
95.45
89.87
83.82
±
5.19 ± 0.68
± 7.80 ± 9.41 ± 12.60 ± 5.81
3
4
5
83.82
88.35
84.50 95.15 79.23 73.16
±
11.51 ± 12.99 ± 5.98 ± 8.89 ± 13.33 ± 16.89
89.70
87.63
79.25
87.15
87.63
83.10
±
15.36 ± 14.92 ± 21.75 ± 7.90 ± 21.27 ± 18.90
66.84
70.56
65.55
90.88
62.04
67.97
±
4.62 ± 12.31 ± 6.67 ± 5.09 ± 7.38 ± 13.88
6
90.99
87.30
86.55
104.86
88.07
85.41
±
14.78 ± 10.06 ± 7.88 ± 11.85 ± 17.72 ± 19.14
7
8
79.12
80.22
87.80
107.32
74.77
74.95
±
13.18 ± 6.08
± 7.45 ± 5.14 ± 7.58 ± 12.71
92.64 97.26 89.64 94.37
98.85 100.51
±
1.99 ± 0.88
± 9.67 ± 4.75 ± 14.03 ± 5.94
1
2

ACCEPTED MANUSCRIPT
9
97.22
84.62
90.94
91.21
80.38
89.41
±
3.89 ± 10.38 ± 3.13 ± 7.66 ± 12.33 ± 7.62
1
0
36.65
98.11
35.01
± 5.31 ± 8.62 ± 6.89 ± 10.23
32.16 73.11 47.52 76.96
± 3.27 ± 6.34 ± 12.38 ± 13.28
88.86
39.21
91.81
±
1.76 ± 4.24
1
1
1
3
33.39
86.33
CH
3
±
3.01 ± 8.48
4
8
74.01
82.97
52.08 98.14 72.64 87.85
CH3
±
8.85 ± 1.06 ±14.68 ± 19.22 ± 7.22 ± 13.13
23.19
34.94
33.57
44.50
42.20
50.19
CH
3
±
±
±
4.54 ± 3.72
± 1.82 ± 5.07 ± 5.46 ± 5.16
19
38.43
2.42 ± 2.10
100.44
54.75
100.67
46.79
98.98
CH3
CH3
± 5.12 ± 1.15 ± 9.23 ± 1.77
MBZ
24.94
1.76 ± 6.44
53.93
35.77
63.43
50.77
61.53
± 1.26 ± 9.84 ± 7.71 ± 6.57
a
Data are average of three independent assays; SD: standard deviation; * indicates point for bond connection;
MBZ: mebendazole.
We next conducted dose-response studies (3.2 nM - 50 µM, five-fold serial dilutions) of the
most active new benzimidazole compound 18 and determined its EC50 in six cancer cell lines
(Table 3). In the three lung cancer cell lines (H157, Calu1, and A549), compound 18 showed
EC ranging from 1.4 to 2.4 µM. Consistent with the benzimidazole drugs (Table 1), compound
5
0
1
8 was highly cytotoxic to the aggressive PC3MLN4 prostate cancer cells (EC : 0.9 µM), as
50
well as the parent PC3 cells (EC : 1.6 µM). Similar cytotoxic activity was observed for
50
compound 18 in the SKOV3 ovarian cancer cells (EC : 1.0 µM). Mebendazole, included as a
5
0
positive reference compound in all assays, showed to be about twice as active as compound 18;
but the difference in activity between the two molecules lacks statistical significance (t-test).
Table 3. EC50 of compound 18 in cancer cell viability assay
1
3

ACCEPTED MANUSCRIPT
a
EC ± SD (µM)
5
0
Cancer Type
Lung
Cell Line
H157
1
8
Mebendazole
3.5 ± 1.4
2.4 ± 0.9
1.6 ± 0.6
1.4 ± 0.5
1.6 ± 0.6
0.9 ± 0.4
1.0 ± 0.4
Calu1
A549
1.1 ± 0.3
0.9 ± 0.1
1.1 ± 0.4
0.5 ± 0.2
0.5 ± 0.1
PC3
Prostate
Ovarian
PC3MLN4
SKOV3
a
Data are average of three independent assays; SD: standard deviation.
Taxanes are standard care therapy for the treatment of metastatic prostate and lung cancers.
However, cancers rapidly develop adaptive resistance and become refractory to taxanes. We
asked if taxane-resistant cancer cells would still be sensitive to the treatment by compound 18.
We had shown previously that the anticancer activity of paclitaxel in the paclitaxel-resistant
A549-TR cell line was > 100-fold less than in the parent A549 cells [38]. Encouragingly, we
observed that both mebendazole and compound 18 reduced cell viability of the A549-TR cell
line (Figure 2). As compared to the reduced sensitivity to paclitaxel (EC : 10.3 µM), both
5
0
compound 18 (EC : 3.8 µM) and mebendazole (EC : 1.7 µM) were significantly more potent
5
0
50
than paclitaxel alone (t-test, p-value < 0.05).
Taken together, these data suggest that the new benzimidazole compound 18 is a potent
inhibitor of cancer cell growth, and that installation of the oxetane moiety in the left-hand side of
mebendazole leads to little or no reduction in anticancer activity of compound 18 in most cell
lines tested. If solubility is improved in this composition, it could lead to better promise for
systemic in vivo delivery of the new drug for the treatment of disseminated tumors.
1
4

ACCEPTED MANUSCRIPT
Figure 2. Anticancer activity of paclitaxel, compound 18, and mebendazole (MBZ) in A549-TR. Data shown as the
average ± SD of three independent biological replicates. Statistical analysis was performed by t-test.
2
.4. Cell cycle analyses
We conducted cell cycle analysis to determine the effects of compound 18 on cell cycle
passage in PC3MLN4 prostate cancer cells (Figure 3). Cells were treated with compound 18 (1
µM) or mebendazole (1 µM), and the well characterized mitotic inhibitor nocodazole (1 µM) was
included as a positive control [39]. After 24 hr incubation, cells were harvested, fixed, stained
with Propidium Iodide (PI) and finally counted by fluorescence-activated cell sorting (FACS)
(Figure 3A). Collected data were analyzed with FlowJo software to determine the percentage of
cells in G1, S, and G2 phase (Figure 3B) using the Dean/Jett/Fox Model [40]. Compound 18
caused accumulation of cells at the G2/M phase of the cell cycle. In contrast, mebendazole had a
less significant effect on the accumulation of cells in G2/M. The cyclin B1 is a marker for cell
cycle arrest at the G2/M phase. Indeed, compound 18, as well as the positive control nocodazole,
1
5

ACCEPTED MANUSCRIPT
increased cyclin B1 levels in cells after 24 hr treatment (Figure 3C). Taken together, these data
suggest that compound 18 induces G2M cell cycle arrest.
Figure 3. (a) Cell-cycle phase distribution of PC3MLN4 treated with compound 18, mebendazole (MBZ) or
nocodazole (NCZ) for 24 hr. Cell-cycle phase distribution was assessed by propidium iodide (PI) staining of
permeabilized ethanol-fixed cells. (b) Graph reporting percentile cell-cycle phase distributions. G1-, S-, and
G2-phase distributions were determined by FlowJo analysis. (c) Western blot showing an accumulation of Cyclin
B1, expressed in G2 and M phase, in PC3MLN4 treated with compound 18, MBZ, or NCZ for 24 hr.
2
.5. Microtubule structures
The benzimidazole drugs similar to our lead compounds are known to disrupt microtubule
structures [41-43], although other potential mechanisms have been suggested. These include
blockade of hedgehog pathway signaling and interference with the action of the specific
anti-apoptotic protein X-linked inhibitor of apoptosis (XIAP) [19, 44]. Microtubule structures
perform vital cellular functions including intracellular organelle transportation, cell motion and
1
6

ACCEPTED MANUSCRIPT
movement, as well as mitosis of proliferating cells. Therapeutic targeting of microtubule
dynamics has led to the discovery and clinical success of the broadly used anticancer drugs
taxanes, epothilones, and eribulin [45-47]. To determine whether compound 18 interfered with
microtubule dynamics in cultured cancer cells, we incubated PC3MLN4 prostate cancer cells
with compound 18 (1 µM) for 48 hr and visualized microtubules by staining with anti-tubulin
antibody (Figure 4). Our results demonstrate disruptions of the rearrangement and diminution of
microtubules in cells treated with compound 18 to a degree similar to that of mebendazole (1 µM)
but less dramatic than nocodazole (1 µM), a powerful but extremely toxic microtubule disruptor.
This result indicates that compound 18 retains the activity of the parent compound in disrupting
microtubule dynamics.
Control
18, 1µM
MBZ, 1µM
100X
NCZ, 1µM
100X
100X
100X
200X
200X
200X
200X
Figure 4. Representative immunofluorescence images showing the effect of compound 18, mebendazole (MBZ),
and nocodazole (NCZ) on microtubules in PC3MLN4 prostate cancer cells after treated for 48 hr. Dashed boxes in
top row indicate area that are further magnified and shown in the bottom row. Tubulin structures are in green color;
and DNA in blue.
1
7

ACCEPTED MANUSCRIPT
2
.6. Docking studies
The benzimidazole anthelmintic drugs are postulated to affect tubulin dynamics by their
direct binding to the colchicine binding domain of tubulin [43]. We applied Autodock Vina [48]
and performed docking study to model the interactions between compound 18 and tubulin
utilizing the reported tubulin crystal structure (PDB Code: 3HKC) [49] (Figure 5). The
simulation analyses suggest that the oxetanyl group of compound 18 participate in electrostatic
interaction with Lys352, and together with the phenyl and benzimidazole moieties, occupy a
hydrophobic packet composed of Leu248, Ala250, Leu255, Ala317, and Ile378. Neither the
carbonyl group nor the amino group on the left-hand side of the molecule participated in H-bond
interactions with the protein. Instead, they seem to optimally position the phenyl and the
oxetanyl groups in the narrow hydrophobic packet. More importantly, the carbamate moiety
forms hydrogen-bond (H-bond) interactions with Asn167, Tyr202, and Val238. It is likely that
this network of H-bonds is required for the affinity of the benzimidazole compounds to the
tubulin protein. This tight binding site can only accommodate small methyl group but not
sterically bulky group such as oxetane, which might provide a rational to explain the significant
loss of activity by compounds 1 - 10 (Table 2).
1
8

ACCEPTED MANUSCRIPT
Figure 5. The lowest energy docking pose of 18 in the colchicine binding site of tubulin (PDB Code: 3HKC).
Compound 18 is presented as yellow sticks and shaded in dark grey. Side chains of amino acids of tubulin protein
involved in interactions with 18 are presented as green sticks. Nitrogen and oxygen atoms are colored as blue and
red, respectively.
2
.7. Solubility
Solubility is one most important physiochemical properties that dictates the performance of
small molecule drugs. Drugs with low solubility often require the usage of toxic excipients to
achieve adequate systemic delivery; and their absorption, hence therapeutic effects, is impaired
by significant patient to patient variability [50-52]. In case of the benzimidazoles, as commonly
used as anthelmintic drugs, they are at best only sparsely soluble in aqueous media. We
compared the kinetic and thermodynamic solubility of mebendazole and the oxetanyl compound
1
8. An HPLC method was developed to quantify the amount of solubilized mebendazole and
compound 18, respectively. The analytical method was optimized to achieve a detection limit of
µg/mL for both compounds. In pH 7.4 PBS buffer (Table 4), the kinetic (0.5 hr incubation)
1
1
9

ACCEPTED MANUSCRIPT
and thermodynamic (24 hr incubation) solubility of mebendazole is below the 1 µg/mL detection
limit. Even after 120 hr incubation, there was still no measurable amount of mebendazole in
solution. In contrast, the kinetic and thermodynamic solubility of compound 18 reached as high
as 253 and 361 µM, respectively. Our data clearly indicates dramatic improvement of aqueous
solubility by compound 18, which could be attributed to the incorporation of the oxetanyl moiety
a
in the molecule. Interestingly, the pK of the nitrogen (N) atom that couples the oxetanyl group
to the benzene ring is calculated as -0.51 (ChemAxon Version 14.8.18.0), indicative of a
non-ionizable N atom. Therefore, the N linker is unlikely to contribute significantly to the
enhanced solubility of compound 18. Taken together, our results demonstrate the unique utility
of oxetane for improving aqueous solubility as well as maintaining anticancer activity of
mebendazole.
Table 4. Solubility of mebendazole and compound 18
a
Solubility (µg/mL)
5% DMSO in PBS, pH 7.4
Mebendazole
18
96
3
0 min
4 hr
< 1
< 1
2
137
a
Solubility results are expressed as the mean of three independent
measurements.
2
.8. In vivo efficacy
We used the mouse PC3MLN4 xenograft tumor model to define the in vivo anticancer
activity of systemically delivered compound 18 and compared the effect to mebendazole.
6
PC3MLN4 cells (1 X 10 of cells) were injected subcutaneously between the scapular of nude
2
0

ACCEPTED MANUSCRIPT
3
mice and allowed to grow to an average size of 35 mm . Tumor-bearing mice were randomized
and treated with compound 18 (30 mg/kg), mebendazole (30 mg/kg), and blank formulation
(control group) via intraperitoneal (IP) injections (3 times a week for two weeks). Tumor volume
and body weight (BW) were measured on the same days as the drug administrations as well as
three days after the last dosing. As shown in Figure 6A, compound 18 and mebendazole showed
statistically significant inhibition of tumor growth, and the effect of both drugs was very similar
after day-9 of the study. This finding correlates well with the similar in vitro cytotoxicity of the
two drugs in PC3MLN4 cells (Table 3). At the last measurement of the tumor volume (day-15),
3
the mean (± SEM) size of the compound 18 treated group (T) is 43.2 mm (± 10.36) while that of
3
the control group (C) is 118.8 mm (± 13.35) (Figure 6C). Therefore, the treatment-to-control
(T/C) ratio for compound 18 at the last time point is 0.36 (t-test p value = 0.0481), below the T/C
≤
0.45 threshold that defines active antitumor activity [53]. The T/C for mebendazole is 0.31.
The changes in body weight indicated that both drugs were well tolerated during the course of
the treatment, although mebendazole induced a near 5% reduction in BW, and the changes were
statistically significant on day-3 and day-5 (Figure 6B). This finding is consistent with reports
by others that suggest the benzimidazole drugs could be safely dosed in preclinical models
[19-21].
2
1

ACCEPTED MANUSCRIPT
Figure 6. (a) Inhibition of tumor growth by compound 18 and mebendazole (MBZ) in PC3MLN4 xenograft tumor
model; (b) the change in body weight, normalized as % of day-1 during the course of the study. Black arrows along
the x-axis indicate dates of drug administration. Statistical analysis was performed by t-test. *: p<0.05, **: p<0.01.
2
2

ACCEPTED MANUSCRIPT
Data shown as mean (± SEM). (c) Graph representing the % of tumor volume normalized on control on day 15. Data
shown as mean (± SEM). Statistical analysis was performed by t-test. N.S.: not significant.
3
. Conclusion
One promising strategy for the discovery of new and safe treatments for late stage
progressive cancer is the rational design based on pre-clinical and clinical anticancer activities of
existing drugs that are not prescribed for treating cancers. The benzimidazole anthelmintic drugs,
including mebendazole, have been widely used in the clinic for decades to eliminate parasites in
the gut. Their preclinical anticancer activities have been revealed recently. However, the clinical
translation of those drugs for the treatment of cancer patients is impeded by the lack of adequate
drug exposure [30]. Our results in this report provide intriguing findings that highly aggressive
prostate cancer cells (PC3MLN4) and drug-resistant lung cancer cells (A549-TR) are highly
vulnerable to the treatment by benzimidazoles. Furthermore, we succeeded in the design and
synthesis of the novel lead compound 18 to dramatically improve aqueous solubility while
maintaining potent anticancer activity in a range of cancer cells of the lung, prostate, and ovarian.
In a mouse tumor xenograft model of the highly aggressive PCMLN4 prostate cancer, compound
1
8 significantly inhibited the tumor growth but showed no signs of toxicity. Similar to the
benzimidazole drugs mebendazole and flubendazole, compound 18 induces cancer cell cycle
arrest in the G2/M phase by disrupting tubulin dynamics. It is likely that aggressive cancer cells
are more dependent on tubulin dynamics therefore are also more sensitive to the exposure to
compound 18. As part of our on-going research, we plan to delineate the mechanism of action
and the molecular target of compound 18. Although mebendazole gave equivalent anti-tumor
effects when injected into animals via an intraperitoneal route, we hypothesize that this was due
2
3

ACCEPTED MANUSCRIPT
to a depot effect where the drug remained in the peritoneum and slowly penetrated into the
circulation. We expect that the improvement in aqueous solubility of compound 18 will enable
sufficient systemic exposure to achieve therapeutic efficacy via oral administration, which would
be problematic with the parent compounds. We will report in due course the detailed
characterizations of compound 18 in in vitro and in vivo pharmacokinetics and
pharmacodynamics.
4
4
4
. Experimental Section
.1. Chemistry
.1.1. General
All reagents and solvents were purchased from commercially available sources and used
without further purification. All reactions were carried out according to the indicated procedures
and conditions. Flash Chromatography was performed on an Interchim PufiFlash 430 with
cartridges filled with Silica gel 50 µm (PF-50SIHC). Reactions were monitored by LC/MS
analysis and/or thin-layer chromatography (TLC) on silica-coated glass with the indicated eluent.
The compounds were visualized by UV light (254 nm). LC/MS analysis was performed on an
Agilent 1200 HPLC/UV (λ = 220 nm and/or 254 nm) system coupled with a mass spectroscopic
(Applied Biosystems, MDS SCIEX, Q TRAP LC/MS/MS) detector. High-resolution mass
spectrometry (HRMS) was performed on a Waters Q-TOF Premier™ Mass Spectrometer in the
positive mode that utilizes electrospray ionization (ESI). Compounds for analysis were dissolved
in 100% DMSO and separated on C18 cartridge (particle size 2.6 µm, dimensions: 100 mm X
2
.1 mm, 0.3 mL/min flow rate, 1 µL injection volume) using acetonitrile/water mobile phase
2
4

ACCEPTED MANUSCRIPT
with 0.1% formic acid as a modifier. The gradient started at 0% acetonitrile, held for 1 min, and
linearly increased to 97% acetonitrile over 10 min, with 2 min hold at 97% acetonitrile and
subsequent re-equilibration to the original conditions in a total of 17 min. All compounds
reported were obtained in a purity of > 95% at 254 nm wavelength. Nuclear magnetic resonance
1
(
H NMR) spectra were recorded on a Varian Mercury plus NMR spectrometer operating at
1
13
4
00.13 MHz frequencies for 1H, using a 5 mm ASW PFG probe capable of detecting H and C
nuclei. The proton chemical shifts (ppm) were referenced to the tetramethylsilane (TMS) internal
standard (0 ppm). Chemical shifts (δ) are reported related to CDCl or DMSO-d residual solvent
3
6
peaks and indicated in parts per million (ppm). The splitting patterns are abbreviated as follows:
s (singlet), d (doublet), dd (doublet of doublets), t (triplet), q (quartet), quin (quintet), sxt (sextet),
triplet of doublets (td), br (broad), and m (multiplet).
General Procedure to Synthesis of Bendazole Moiety with Different Side Chains (Compounds
1
- 10)
To a stirred mixture of each bendazole in base/solvent (100 mM) were added 10 equiv of
alcohols indicated in Scheme 1. The reaction proceeded at 90, 100, or 110 °C for 3 hr to o/n. For
compounds 5 and 7, the reaction mixture was exposed to microwave irradiation for 5 - 15 min at
6
0 or 70 °C. Workup: Once the reaction was complete, the solvent was removed under reduced
pressure, diluted with four times the reaction volume of EtOAc and washed with water and brine.
The organic layer was dried with Na SO and concentrated in vacuo, followed by purification
2
4
performed on automated flash chromatography system with the desired method commented
below.
2
-(Dimethylamino)ethyl (5-(propylthio)-1H-benzo[d]imidazol-2-yl)carbamate (1)
2
5

ACCEPTED MANUSCRIPT
Isolated by silica gel chromatography with DCM:MeOH = 3:1 as an off-white solid; R =
f
0
0
.32 (10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with
2
1
.1% formic acid): t = 7.6 min; H NMR (400 MHz, 25 ºC, CDCl ) δ 7.57 (1H, br), 7.45 (1H,
R
3
br), 7.25 (1H, d, J = 8.2 Hz), 4.45, (2H, t, J = 6.1 Hz), 2.88 (2H, t, J = 7.2 Hz), 2.79 (2H, t, J =
6
.1 Hz), 2.37 (6H, s), 1.64 (2H, sxt, J = 7.4 Hz), 1.01 (3H, t, J = 7.2 Hz); MS (ESI) m/z: [M +
+
H] Calcd for C H N O S 323.2; Found 323.6.
1
5
23
4
2
3
-(Dimethylamino)propyl (5-(propylthio)-1H-benzo[d]imidazol-2-yl)carbamate (2)
Isolated by silica gel chromatography with DCM:MeOH = 3:1 as an off-white solid; R =
f
0
0
.07 (10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with
2
1
.1% formic acid): t
R
= 7.6 min; H NMR (400 MHz, 25 ºC, CDCl
3
) δ 7.59 (1H, br), 7.42 (1H,
br), 7.26 (1H, d, J = 7.8 Hz), 4.43 (2H, t, J = 6.7 Hz), 2.87 (2H, t, J = 7.2 Hz), 2.47 (2H, t, J = 6.8
Hz), 2.26 (6H, s), 2.04 (2H, quin, J = 6.8 Hz), 1.64 (2H, sxt, J = 7.3 Hz), 1.01 (3H, t, J = 7.4 Hz);
+
MS (ESI) m/z: [M + H] Calcd for C H N O S 337.2; Found 337.6.
1
6
25
4
2
Oxetan-3-yl (5-(propylthio)-1H-benzo[d]imidazol-2-yl)carbamate (3)
f
Isolated by silica gel chromatography with DCM:MeOH = 40:1 as a white solid; R = 0.38
(
10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with 0.1%
2
1
formic acid): t = 9.2 min; H NMR (400 MHz, 25 ºC, CDCl ) δ 12.02 (1H, br), 7.50 (1H, s),
R
3
7
.41 (1H, d, J = 8.2 Hz), 7.13 (1H, d, J = 8.6 Hz), 5.28 (1H, br), 4.88-4.84 (1H, m), 4.32 (1H, t, J
=
9.4 Hz), 4.02 (1H, dd, J = 9.4, 5.9 Hz), 3.75-3.60 (2H, m), 2.86 (2H, t, J = 7.2 Hz), 1.54 (2H,
+
sxt, J = 7.2 Hz), 0.94 (3H, t, J = 7.2 Hz); MS (ESI) m/z: [M + H] Calcd for C14
H
18
N
3
O
3
S 308.1;
Found 308.6.
2
6

ACCEPTED MANUSCRIPT
(3-Methyloxetan-3-yl)methyl (5-(propylthio)-1H-benzo[d]imidazol-2-yl)carbamate (4)
Isolated by silica gel chromatography with hexanes:EtOAc = 1:1 as a white solid; R = 0.41
f
(
10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with 0.1%
2
1
formic acid): t
R
= 9.1 min; H NMR (400 MHz, 25 ºC, CDCl
3
) δ 7.52 (1H, br), 7.36 (1H, br),
7
.25 (1H, d, J = 7.9 Hz), 4.62 (2H, d, J = 6.3 Hz), 4.54 (2H, d, J = 5.9 Hz), 4.50 (2H, s), 2.88
(2H, t, J = 7.4 Hz), 1.64 (2H, sxt, J = 7.4 Hz), 1.40 (3H, s), 1.01 (3H, t, J = 7.4 Hz) ; MS (ESI)
+
m/z: [M + H] Calcd for C H N O S 336.1; Found 336.5.
1
6
22
3
3
Oxetan-3-yl (5-benzoyl-1H-benzo[d]imidazol-2-yl)carbamate (5)
f
Isolated by silica gel chromatography with DCM:MeOH = 3:1 as a white solid; R = 0.56 (10%
MeOH/DCM); white solid (180 mg, 53%); HPLC analysis (C18 reverse phase, 17 mins, 0-97%
1
H O/ACN with 0.1% formic acid): t = 8.7 min; H NMR (400 MHz, 25 ºC, CDCl ) δ
2
R
3
1
4
2.42-12.31 (1H, m), 7.95-7.54 (8H, m), 5.27 (1H, t, 5.7), 4.92-4.85 (1H, m), 4.30-4.24 (1H, m),
+
.08-4.02 (1H, m), 3.76-3.72 (1H, m), 3.65-3.61 (1H, m); MS (ESI) m/z: [M + H] Calcd for
18 16 3 4
C H N O 338.1; Found 338.5.
3
-(Dimethylamino)propyl (5-(4-fluorobenzoyl)-1H-benzo[d]imidazol-2-yl)carbamate (6)
Isolated by silica gel chromatography with DCM:MeOH = 2:1 as an off-white solid; R =
f
0
0
.08 (10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with
2
1
.1% formic acid): t = 7.9 min; H NMR (400 MHz, 25 ºC, DMSO-d ) δ 7.84 (1H, s), 7.80 (2H,
R
6
dd, J = 8.2, 5.9 Hz), 7.56-7.51 (2H, m), 7.38 (2H, t, J = 8.6 Hz), 4.23 (2H, t, J = 6.5 Hz), 2.57
+
(
2H, t, J = 6.8 Hz), 2.33 (6H, s), 1.88 (2H, quin, J = 6.8 Hz) ; MS (ESI) m/z: [M + H] Calcd for
2
7

ACCEPTED MANUSCRIPT
C H FN O ; Found 385.2; Found 385.6.
20
22
4
3
Oxetan-3-yl (5-(phenylthio)-1H-benzo[d]imidazol-2-yl)carbamate (7)
Isolated by silica gel chromatography with DCM:MeOH = 9:1 as a yellow solid; R = 0.49
f
(
10% MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with 0.1%
2
1
formic acid): t
R
= 9.5 min; H NMR (400 MHz, 25 ºC, CDCl
3
) δ 12.22-12.13 (1H, m), 7.58 (1H,
d, 6.3), 7.52 (1H, t, J = 7.6 Hz), 7.31-7.11 (6H, m), 5.27 (1H, t, J = 5.5 Hz), 4.87 (1H, d, J = 3.5
Hz), 4.24 (1H, td, J = 9.3, 4.5 Hz), 4.05-4.00 (1H, m), 3.76-3.59 (2H, m); MS (ESI) m/z: [M +
+
H] Calcd for C H N O S 342.1; Found 342.4.
1
7
16
3
3
2
-(Dimethylamino)ethyl (5-benzoyl-1H-benzo[d]imidazol-2-yl)carbamate (8)
f
Isolated by silica gel chromatography with DCM:MeOH = 9:1 as a white solid; R = 0.21 (10%
2
MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with 0.1% formic
1
acid): t = 7.9 min; H NMR (400 MHz, 25 ºC, DMSO-d ) δ 7.86 (1H, s), 7.72 (2H, d, J = 7.8
R
6
Hz), 7.67-7.63 (1H, m), 7.59-7.52 (4H, m), 4.28 (2H, t, J = 5.6 Hz), 2.55 (2H, t, J = 5.6 Hz), 2.20
+
(
6H, s); MS (ESI) m/z: [M + H] Calcd for C19
H
21
N
4
O
3
353.2; Found 353.6.
2
-(Dimethylamino)ethyl (5-(4-fluorobenzoyl)-1H-benzo[d]imidazol-2-yl)carbamate (9)
Isolated by silica gel chromatography with DCM:MeOH = 9:1 as a white solid; R = 0.24 (10%
f
MeOH/DCM); HPLC analysis (C18 reverse phase, 17 mins, 0-97% H O/ACN with 0.1% formic
2
1
acid): t = 8.1 min; H NMR (400 MHz, 25 ºC, DMSO-d ) δ 7.84 (1H, s), 7.80 (2H, dd, J = 8.1,
R
6
5
.6 Hz), 7.56-7.51 (2H, m), 7.38 (2H, t, J = 8.8 Hz), 4.28 (2H, t, J = 5.6 Hz), 2.55 (2H, t, J = 5.6
+
Hz), 2.20 (6H, s); MS (ESI) m/z: [M + H] Calcd for C H FN O 371.2; Found 371.5.
1
9
20
4
3
2
8