Structure-guided design, synthesis, and biological evaluation of (2-(1 H-Indol-3-yl)-1 H-imidazol-4-yl)(3,4,5-trimethoxyphenyl) methanone (ABI-231) analogues targeting the colchicine binding site in Tubulin

Article abstract of DOI:10.1021/acs.jmedchem.9b00706

ABI-231 is a potent, orally bioavailable tubulin inhibitor that interacts with the colchicine binding site and is currently undergoing clinical trials for prostate cancer. Guided by the crystal structure of ABI-231 in complex with tubulin, we performed structure-activity relationship studies around the 3-indole moiety that led to the discovery of several potent ABI-231 analogues, most notably 10ab and 10bb. The crystal structures of 10ab and 10bb in complex with tubulin confirmed their improved molecular interactions to the colchicine site. In vitro, biological studies showed that new ABI-231 analogues disrupt tubulin polymerization, promote microtubule fragmentation, and inhibit cancer cell migration. In vivo, analogue 10bb not only significantly inhibits primary tumor growth and decreases tumor metastasis in melanoma xenograft models but also shows a significant ability to overcome paclitaxel resistance in a taxane-resistant PC-3/TxR model. In addition, pharmacological screening suggested that 10bb has a low risk of potential off-target function.

Full text of DOI:10.1021/acs.jmedchem.9b00706

Subscriber access provided by Nottingham Trent University
Article
Structure guided design, synthesis, and biological evaluation of 2-
1H-Indol-3-yl)-1H-imidazol-4-yl)(3,4,5-trimethoxyphenyl) Methanone
(
(ABI-231) analogs targeting the colchicine binding site in tubulin
Qinghui Wang, Kinsie Arnst, Yuxi Wang, Gyanendra K Kumar,
Dejian Ma, Stephen W. White, Duane D Miller, Weimin Li, and Wei Li
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00706 • Publication Date (Web): 28 Jun 2019
Downloaded from http://pubs.acs.org on June 29, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure guided design, synthesis, and biological
evaluation of 2-(1H-Indol-3-yl)-1H-imidazol-4-
yl)(3,4,5-trimethoxyphenyl) Methanone (ABI-231)
analogs targeting the colchicine binding site in
tubulin
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Qinghui Wang†,ф,Π, Kinsie E. Arnst†,, Π, Yuxi Wang,‡,Π, Gyanendra Kumar , Dejian Ma , Stephen
§
†
§
†
‡,
†,
W. White , Duane D. Miller , Weimin Li *, Wei Li *
†Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health
‡
Science Center, Memphis, TN 38163, United States. Department of Respiratory and Critical Care
§
Medicine, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, Department of
Structural Biology, St. Jude Children’s Research Hospital, Memphis, TN 38105, United States.
*
Corresponding author:
Wei Li, Ph.D., Department of Pharmaceutical Sciences, University of Tennessee Health Science
Center, Memphis, TN 38163. Email: wli@uthsc.edu, Tel: 901-448-7532, FAX: 901-448-6828.
Weimin Li, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China. Email:
weimin003@scu.edu.cn.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT: ABI-231 is a potent, orally bioavailable tubulin inhibitor that interacts with the
colchicine binding site and is currently undergoing clinical trials for prostate cancer. Guided by
the crystal structure of ABI-231 in complex with tubulin, we performed structure-activity
relationship (SAR) studies around the 3-indole moiety that led to the discovery of several potent
ABI-231 analogs, most notably 10ab and 10bb. The crystal structures of 10ab and 10bb in
complex with tubulin confirmed their improved molecular interactions to the colchicine site. In
vitro, biological studies showed that new ABI-231 analogs disrupt tubulin polymerization,
promote microtubule fragmentation, and inhibit cancer cell migration. In vivo, analog 10bb not
only significantly inhibits primary tumor growth and decreases tumor metastasis in melanoma
xenograft models, but also shows a significant ability to overcome paclitaxel resistance in a taxane-
resistant PC-3/TxR model. In addition, pharmacological screening suggested that 10bb has a low
risk of potential off-target function.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Microtubules are cylindrical polymers that are of great importance for many cellular functions
such as cell division and the formation of the mitotic spindle, cell shape maintenance, and cell
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
motility. Microtubules also play a critical role in the proliferation of cancer cells 1-3, and
microtubule dynamics is therefore one of the most productive therapeutic targets for cancer
3
treatment . Microtubule-targeting agents (MTAs) can interfere with functions of microtubules to
disrupt the formation of the mitotic spindle, eventually leading to mitotic arrest at the
metaphase/anaphase transition. A large number of MTAs have now been discovered and many
have been approved by the FDA for the treatment of different cancer types, such as prostate cancer,
4
breast cancer, lung cancer and ovarian cancer . Based on their effects on microtubule dynamics,
MTAs are generally classified into two categories: microtubule-stabilizing and microtubule-
destabilizing agents. Microtubule-stabilizing agents such as epothilone, ixabepilone, paclitaxel,
docetaxel, and cabazitaxel bind to polymerized tubulin proteins and enhance microtubule
polymerization. On the contrary, microtubule-destabilizing agents such as colchicine and vinca
alkaloids inhibit the polymerization of microtubules 5.
The MTAs have achieved great success in cancer treatment for many types of cancers. However,
drug-resistance is acquired over the course of treatment and becomes a significant clinical
limitation. It is estimated that more than 1.7 million new cancer cases will be diagnosed in United
States in 2019, and developing new anticancer agents to treat resistant phenotypes is therefore
urgently needed 6-7. Mechanisms that mediate drug-resistance against MTAs include, but are not
limited to the overexpression of membrane-bound drug efflux proteins, such as P-glycoprotein (P-
gp); the overexpression of certain β-tubulin isotypes; and tubulin protein mutations 8-9. Taxanes
and vinca alkaloids, are especially susceptible to efflux by ATP-binding cassette (ABC)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transporters and are less effective against cells overexpressing the tubulin β-III isotype 10-12
.
The colchicine binding site is one of the well-documented binding sites in tubulin and located at
1
3
the interface between α and β-tubulin heterodimers . Colchicine binding site inhibitors (CBSIs,
examples are listed in Figure 1) mechanistically decrease cellular motility, alter cell morphology,
impair protein assembly, and arrest mitosis 14-15. Colchicine (Figure 1) is a secondary metabolite
isolated from natural sources and has wide-scope mechanisms of action. Unfortunately, it induces
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
4
systemic toxicities and multi-organ dysfunction in a clinical trial for cancer treatment , and thus
is not currently approved by the FDA for cancer treatment. Other than colchicine and its
derivatives, a number of other CBSIs have also been reported in the past decades, some examples
are listed in Figure 1. Encouragingly, extensive preclinical studies have strongly suggested that
CBSIs are effective in suppressing the overexpression of tubulin isotypes and surmount drug-
resistance mediated by P-gp, MRP1, and MRP2 16-18
.
O
H CO
H3CO
H3CO
N
3
NH
O
HN
O
N
N
NH
H CO
3
N
NH
OCH3
N
H
N
H
O
O
H CO
OH
OCH3
3
O
HN
OCH3
Colchicine
Combretastatins A-4
Lexibulin
Plinabulin
NH2
O
H
HN
N
H N
O
NH2
H CO
S
N
2
3
HN
NH2
O
O
O
H
H
N
N
HN
N
N
NH
O
N
OCH3
S
H CO
OCH3
3
H CO
Br
3
OCH3
OCH3
OH
Crolibulin
ABT-751
Denibulin
ABI-231
Figure 1. Examples of microtubule inhibitors that target colchicine binding sites in tubulin.
ACS Paragon Plus Environment

Page 5 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In our prior reports, we have described the discovery of several scaffolds of potent tubulin
inhibitors, including SMARTs, RABIs, ABIs, and IBIs 19-24. The SMART structure is our initially
discovered scaffold that exhibited nanomolar range antiproliferative activities against a panel of
cancer cell lines and showed promising in vivo activities in both melanoma and prostate xenograft
models 19, 25-27. Subsequent structural modification of the SMART scaffold by replacing the central
thiazole ring with a more metabolically stable imidazole ring resulted in the discovery of 2-aryl-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-benzoyl-imidazole (ABI-I/II) pharmacophores. ABI-I/IIs not only showed more potent activities
but also improved bioavailability in comparison with SMARTs 19-20. The ABI-III’s pharmacophore
differed from ABI-I/IIs in having bicyclic heterocycles 21-22, which are versatile moieties in
2
8
medicinal chemistry . ABI-III exhibited more potent antiproliferative activity (IC values are in
5
0
low single-digit nanomolar range) than ABI-I/IIs. Rotation of the imidazole ring in ABI-I/IIs led
to the establishment of reverse ABIs (RABIs) which maintain comparable antiproliferative
2
3
potency . Substituting the central imidazole ring and the carbonyl linker in ABI-IIIs with bicyclic
24
heterocycles provided the IBIs, which displayed similar potency to that of ABI-IIIs . SMARTs,
ABIs, RABIs, and IBIs share the same mechanism of action: binding to the colchicine binding site
in tubulin and inhibiting tubulin polymerization. Interestingly, they are all effective against
multidrug-resistant cell lines and can circumvent P-gp mediated drug resistance 20-22. Amongst all
the scaffolds, ABI-231, a member of the ABI-III class, displays optimum activity and has an
2
9
average IC value of 5.2 nM against a large panel of cancer cell lines . Additionally, it was
50
revealed that ABI-231 inhibits the expression of tubulin βIII and βIV over other isotypes via
restoring the expression of miR-200c in a panel of pancreatic cancer cell lines, which supports the
3
0
ability of ABI-231 to surmount drug resistance to existing tubulin inhibitors . Recently, we
described a focused SAR investigation of the trimethoxyphenyl (TMP) moiety in ABI-231 that
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
lead to a unique 3- methoxybenzo[4,5]-dioxene analog, which showed more potent
3
1
antiproliferative activity than ABI-231 . In this work, we describe an efficient synthesis of ABI-
231 that facilitated a SAR evaluation of the indole moiety, provide the molecular interactions and
rationales between tubulin proteins and ABI-231 or its two potent derivatives using X-ray
crystallography, investigate the mechanism of action of the new analogs and determine their
biological activities.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CHEMISTRY
To rapidly generate ABI-231 analogs modifying the indole moiety, we designed an efficient
synthetic strategy illustrated in Scheme S1. Creation of the ABI-231 scaffold uses imidazoline
cyclization, which requires a diamine intermediate and indolyl-3-carboxyaldehyde according to a
32
reported methodology . In this method, it is crucial to accumulate a large quantity of the diamine
intermediate, the synthesis of which was furnished in six steps and described in Scheme 1.
Commercially available 3,4,5-trimethoxybenzaldehyde was treated with Grignard reagent
33
vinylmagnesium bromide to generate the racemic mixture 1 , which was subsequently treated
with tert-butyldimethylsilyl chloride in the presence of imidazole to give 2. 2 then underwent the
Upjohn reaction in the presence of a catalytic equivalent of osmium tetroxide to obtain a mixture
of diastereoisomers 3 in 79% yield over three steps. Diazide 5 was produced in two steps from 3
3
4
by following reported procedures : mesylation of the diol 3 in the presence of methansulfonyl
chloride and triethylamine yielded dimesylate 4 in 86% yield; 4 was then refluxed with sodium
azide in DMF to give diazide 5 in 67% yield. Hydrogenation of diazide 5 in the presence of 10%
Pd/C provided diamine 6 in 53% yield.
ACS Paragon Plus Environment

Page 7 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 1. Synthesis of diamine intermediate 6
OCH3
OCH3
TBSCl, Im, DCM H CO
H CO
3
3
H CO
MgBr
o
3
OCH3
OCH3
0 C to r.t.
H CO
H CO
3
3
o
85% for
THF, 0 C to r.t.
OTBS
OH
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
two steps
OHC
2
1
OCH3
OCH3
H CO
H CO
3
3
NaN , DMF
OsO , NMO
3
4
MsCl, TEA
DCM, r.t.
reflux
t-BuOH/H O
2
H CO
H CO
3
3
r.t.
OTBS
OH
OTBS
OMs
58%
93%
71%
OH
3
OMs
4
OCH3
OCH3
H CO
3
H CO
3
H , Pd/C(10%)
2
MeOH-EtOAc, r.t.
H CO
3
H CO
3
OTBS
OTBS
NH2
61%
H N
2
N3
N3
5
6
With 6 in hand, syntheses of ABI-231 and its analogs was pursued as outlined in Scheme 2 and
furnished in 19% yield over four steps. Treatment of indole-3-carboxyaldehyde with di-tert-butyl
dicarbonate gave 7. In the presence of N-bromosuccinimide, 7 then reacted with diamine 6 at ice
temperature to form imidazoline 8, which was subjected to the deprotection of TBS in the presence
of TBAF and the deprotection of Boc in the presence of TFA to provide 9 in two steps. ABI-231
was accomplished through simultaneous oxidation of the imidazoline to imidazole and the
secondary alcohol to carbonyl in 9 in the presence of 3.0-5.0 equivalents of 2-iodoxybenzoic acid
3
5
(
IBX) . By following Scheme 2, we furnished twenty-six new ABI-231 analogs (10aa-10av and
1
0ba-10be) in 10-20% yield over four steps. Hydrogenation of 10ak and 10ao afforded 10au and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
0av.
Scheme 2. Synthesis of ABI-231 analogs modifying the indole moiety
H CO
H CO
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CHO
7aa-7at
Boc
3
OCH3
OCH3
OCH3
R
OCH3
N
HO
TBSO
TBAF, THF
r.t.
NBS, DCM
0oC to r.t.
i) TFA, DCM, r,t,
ii) IBX, DMSO, r.t.
N
N
NH
6
NH
10-20%
R
over 4 steps
R
N
N
Boc
Boc
8aa-8at
9
aa-9at
H CO
ABI-231: R=H
0aa: R=2-Ph-3-indolyl
0ab: R=4-CH -3-indolyl
10ac
: R=4-F-3-indolyl
10ad: R=4-Br-3-indolyl
10ae: R=4-OCH -3-indolyl
10af: R=4-OBn-3-indolyl
10ag: R=5-CH -3-indolyl
10ah: R=5-F-3-indolyl
0ai: R=5-Cl-3-indolyl
10aj: R=5-Br-3-indolyl
0ak: R=5-OBn-3-indolyl
0av: R=5-OH-3-indolyl
0al: R=6-CH -3-indolyl
10am: R=6-F-3-indolyl
10an: R=6-Br-3-indolyl
10ao: R=6-OBn-3-indolyl
10aw: R=6-OH-3-indolyl
10ap
: R=6-OCH -3-indolyl
10aq: R=6-COOCH -3-indolyl
10ar: R=7-CH -3-indolyl
10as
: R=7-OBn-3-indolyl
10at: R=7-CH -3-indolyl
10bb: R=H, 4-indolyl
3
OCH3
1
1
H , Pd/C
2
3
OCH3
O
N
3
3
3
3
NH
3
3
R
1
N
1
0bc: R=H, 5-indolyl
10bd: R=H, 6-indolyl
10be: R=H, 7-indolyl
H , Pd/C
2
H
1
1
1
10aa-10av, 10bb-10be
3
RESULTS AND DISCUSSION
SAR investigation. To compare to the antiproliferative potency of ABI-231 that was reported
previously, the cytotoxicity of all new ABI-231 analogs were evaluated in human melanoma cell
lines A375, WM164, and M14. Colchicine was used as a positive control and ABI-231 was used
as a prototype for comparison. The antiproliferative activities of the compounds were evaluated in
cell viability MTS assay. IC values are reported in nM and were calculated from at least three
5
0
independent experiments, each performed in quadruplicates using 96-well plates. Results from
ACS Paragon Plus Environment

Page 9 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
these in vitro assays are shown in Table 1. ABI-231 has an IC value ranging from 5.6-8.1 nM
5
0
2
2
and is comparable to our previously reported data .
Table 1. In vitro growth inhibitory effects of ABI-231 analogs modifying the indole moiety (nM)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound
ID
ABI-231
R
H
2-Ph
4-CH3
4-F
4-Br
4-OCH3
4-OBn
5-CH3
5-F
5-Cl
5-Br
A375
8.1 ± 1.6
>1000
1.8 ± 0.2
7.2 ± 0.7
13.8 ± 1.7
>1000
M14
5.6 ± 0.9
>1000
1.7 ± 0.2
8.2 ± 0.9
10.4 ± 1.2
>1000
WM164
7.2 ± 0.9
>1000
3.2 ± 0.3
ND
67.1 ±14.5
>1000
>1000
1
0aa
0ab
0ac
10ad
0ae
0af
0ag
0ah
10ai
0aj
1
1
1
1
>1000
>1000
1
1
277.7 ± 36.3 177.4 ± 16.3 552.3 ± 108.3
19.2 ± 1.5
61.8 ± 9.3
164.2 ± 16.7
>1000
18.0 ± 1.3
64.9 ± 4.3
104.3 ± 5.9 441.7 ± 110.0
80.9 ± 16.6
159.5 ± 24.1
1
1
1
0ak
0au
5-OBn
5-OH
>1000
>1000
>1000
>1000
>1000
1
0al
10am
0an
10ao
0av
0ap
10aq
0ar
0as
0at
0bb
6-CH3
6-F
6-Br
6-OBn
6-OH
21.1 ± 3.9
17.5 ± 3.0
67.8 ± 10.5
>1000
21.7 ± 4.7
5.8 ± 0.8
41.2 ± 7.5
>1000
23.7 ± 6.1
10.3 ± 1.2
55.4 ± 7.4
>1000
>1000
>1000
H3CO
OCH3
1
OCH3
O
1
1
>1000
>1000
N
NH
6-OCH3
6-COOCH3
7-CH3
7-OBn
7-OCH3
4-indolyl
5-indolyl
6-indolyl
7-indolyl
793.4 ± 89.9 879.8 ± 88.2
R
>1000
13.1 ± 1.3
>1000
38.3 ± 3.7
3.6 ± 1.0
8.6 ± 3.3
8.0 ± 1.0
>1000
10.8 ± 1.0
>1000
34.1 ± 3.5
3.7 ± 0.8
18.5 ± 3.7
6.1 ± 1.1
>1000
N
H
1
1
1
1
10bc
10bd
61.3 ± 11.0
>1000
115.4 ± 25.0
1.6 ± 0.3
6.1 ± 1.2
3.8 ± 0.5
OCH3
OCH3
O
N
OCH3
NH
1
0be
285.5 ± 78.7 240.6 ± 38.1 435.8 ± 171.3
14.1 ± 2.2 16.6 ± 1.5 10.8 ± 1.9
R
Colchicine
Twenty-two new analogs (10aa to 10at) were synthesized to inspect substituent effects to the 3-
indolyls. Compared to the prototype ABI-231, analog 10aa with a bulky phenyl group on the 2-
position of indole moiety showed essentially a total loss of activity with IC values larger than 1
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
µM. This indicated that bulky substituents on the indole are not tolerated. Five analogs were
synthesized in the 4-position substituted series, including 10ab-10af and 10au. The best inhibitory
effect was observed in 4-methyl analog 10ab, which has IC values ranging from 1.7-3.2 nM,
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
which is ~3-fold more potent than ABI-231. Analog 10ac has a small substituent (4-F) and displays
nearly equipotent antiproliferative activity to that of ABI-231 with its IC values ranging from
5
0
7
.2-8.2 nM. In contrast, 10ad, an analog with a large-size halogen 4-bromo, has IC values
50
ranging from 10.4-67.1 nM and is 2-fold less potent than ABI-231. For analogs 10ae with a 4-
methoxy group and 10af with a 4-benzoxy group, remarkable reductions of activities were
observed, their IC values are more than 1 µM. In general, 5-position substituted analogs show
50
substantially decreased antiproliferative activities in comparison with their corresponding 4-
position substituted counterparts. For example, the 5-fluoro analog 10ah and the 5-bromo analog
1
0aj are 2-fold and 10-fold less potent than 10ac and 10ad, respectively. The 5-methyl analog
10ag has approximately 100-fold decreased antiproliferative activity compared to 10ab. In this
series, the best activity was observed in the 5-fluoro analog 10ah, which has IC values ranging
5
0
from 18.0-80.9 nM. It is surprising to find that hydrogenation of the bulky 5-benzyloxy analog
10ak) to 5-hydroxy analog (10au) does not lead to any improvement of antiproliferative activity.
For the 6-position substituted analogs, similarly, bulky functional groups such as analogs 10ao,
0ap and 10aq lost activities (all have IC values of more than 1 µM) whereas small substituent
(
1
5
0
benefited the cytotoxicity, for example, 6-fluoro analog 10am. The best activity in this series was
observed in 10am, having IC values ranging from 5.8-17.5 nM. Hydrogenation of the bulky
5
0
benzyloxy to hydroxy on the 6-position did not provide any significant improvement of activity
10av) neither. Among the 7-position substituted 3-indolyl analogs, 10ar with a methyl
substitution possesses the most potent inhibitory effect (IC values range from 10.8-61.3 nM).
(
5
0
ACS Paragon Plus Environment

Page 11 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Interestingly, 10at with a 7-methoxy substituent showed significantly stronger activity (IC50
ranges from 34.1-115.4 nM) than its other position substituted counterparts (10ae and 10ap).
Taken together, these SAR data revealed that small substituents (e.g. fluoro and methyl) are
strongly preferred over bulky functional groups on the 3-indolyl moiety, and modification at the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-position generate the most potent analog (10ab in Table 1).
3
6
In addition, we previously reported a series of indolyl-imidazopyridines as CBSIs in 2015 .
SAR studies in those compounds revealed that indolyl rotation significantly affects
antiproliferative activities. We hypothesized that this may also apply to ABI-231, considering its
overall structural similarity to the indolyl-imidazopyridines. To test this hypothesis, four ABI-231
analogs 10bb-10be were synthesized (Scheme 2) and their in vitro growth inhibitory effects are
also shown in Table 1. In this series, 5- and 6-indolyl analogs 10bc and 10bd showed comparable
antiproliferative activities to that of ABI-231and have IC values ranging from 6.1-18.5 and 3.8-
5
0
8.0 nM, respectively. The 7-indolyl analog 10be was significantly less potent than ABI-231,
having IC values more than 100 nM. 10bb, the 4-indolyl analog, showed approximately 2-fold
5
0
improvement of activity compared to ABI-231 and has IC values ranging from 1.6-3.7 nM.
5
0
Collectively, these data suggest that the antiproliferative activity of the ABI-231 scaffold is very
sensitive to the position in the indole moiety connecting to the central imidazole ring (e.g, 3-indole
and 4-indole are the best) and the position/size in this indole ring (e.g., a small, hydrophobic
substitution at the 3- or 4-position are strongly preferred).
X-ray structures of ABI-231, 10ab and 10bb in complex with tubulin. To understand the
molecular interactions between ABI-231 and tubulin protein, and the structural basis of the SAR
observed above, we first determined the crystal structure of the T2R-TTL (consisting of two α/β-
tubulin dimers, the stathmin-like protein RB3, and tubulin tyrosine ligase) in complex with ABI-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
31 at a resolution of 2.6 Å (PDB code: 6O61). The crystallographic data and refinement statistics
are shown in Table 2.
a
Table 2. X-Ray data collection and refinement statistics. Values in parentheses are for highest-
resolution shell.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ABI-231
PDB ID: 6O61)
Compound 10ab
(PDB ID: 6O5N)
Compound 10bb
(PDB ID: 6O5M)
(
Data collection
Space group
P2 2 2
P2 2 2
P2 2 2
1 1 1
1
1
1
1
1
1
Cell dimensions
a, b, c (Å)
105.26 157.09
184.04
105.30 157.76
182.08
104.97 157.72
181.61
α, β, γ ()
Resolution (Å)
90 90 90
90 90 90
50 – 3.00 (3.11 –
3.00)
90 90 90
50 - 2.3 (2.37 -
2.3)
50 - 2.60 (2.69 -
a
2
.60)
Rmeas,
I/σ(I)
Completeness (%)
Redundancy
Refinement
Resolution (Å)
0.116 (0.965)
16.25 (2.0)
99.2 (99.5)
6.8 (6.9)
0.150 (1.00)
12.25 (2.0)
99.9 (100)
6.7 (6.8)
0.113 (1.00)
22.6 (3.0)
99.9 (99.8)
13.0 (13.1)
39.83 - 2.60 (2.69 – 41.78 – 3.00 (3.11 – 49.76 – 2.3 (2.37
2
.60)
3.00)
- 2.3)
No. reflections
Rwork / Rfree
No. atoms
92647 (8453)
0.2049 / 0.2452
17165
58910 (4504)
0.2024 / 0.2452
17476
135160 (12695)
0.1899 / 0.2233
17881
Protein
16695
17228
16860
Ligand/ion
Water
236
234
207
41
253
768
B factors
Protein
Ligand/ion
Water
55.27
55.85
40.72
65.25
48.52
47.30
43.65
42.35
41.74
R.m.s. deviations
Bond lengths (Å)
Bond angles ()
Ramachandra Plot
Favoured (%)
Allowed (%)
Outliers (%)
0.005
1.14
0.004
1.03
0.004
1.05
97.99
1.92
0.09
96.10
3.71
0.19
98.18
1.77
0.05
ACS Paragon Plus Environment

Page 13 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The structure was clearly resolved (Figure 2A) and revealed that ABI-231 is bound at the
colchicine binding site at the interface of α and β-tubulin opposite the GTP that also binds at this
interface (Figure 2B and 2C). It is positioned to make three hydrogen bonds with tubulin; the
carbonyl with the main chain NH of β-Asp249, the imidazole NH with the main chain C=O of α-
Thr179, and the indole NH with the main chain C=O of β-Asn347 and α-Thr178 through a water
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
“
bridge”. The trimethoxyphenyl moiety of ABI-231 occupies a deep pocket lined by β-Tyr200, β-
Val236, β-Cys239 and β-Leu253. The imidazole ring is sandwiched between the side chains of β-
Leu246 on one side and β-Ly252 and β-Asn256 on the other side. The indole ring occupies a
pocket lined by β-Asn256, β-Met257, β-Ala314, β-Lys350, β-Lys350, and α-Val181. Note that
ABI-231 mostly interacts with the β-tubulin, but it also makes two interactions with the α-tubulin
(α-Thr179 and α-Thr178) across the interface. Thus, the compound acts like glue to increase the
α- and β-interaction. ABI-231 shares a similar binding pose to that of colchicine (Figure S1A),
indicating that the two compounds share a similar mode of inhibition.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Complex structures of ABI-231 (6O61) and the new analogs 10ab (6O5N) and 10bb
(6O5M) with tubulin protein. (A, D, G) overall crystal structure of the three compounds in complex
with tubulin proteins. (B, E, H) detailed interactions between ABI-231, 10ab, and 10bb and
tubulin (ABI-231 is in yellow stick representation, 10ab is in green stick representation, and 10bb
is in orange stick representation). (C, F, I) 90-degree counterclockwise rotation view of the
corresponding structures along x axis.
ACS Paragon Plus Environment

Page 15 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To investigate the prominent potencies of 10ab and 10bb, we also solved the crystal structures
of these compounds in complexes with tubulin protein (PDB code: 6O5N for 10ab, 6O5M for
10bb). The crystallographic data and structure refinement statistics for 10ab and 10bb in complex
with tubulin are also shown in Table 2, and the electron density maps are shown in Figure S2B
and S2C, along with the electron density map for ABI-231 (Figure S2A). Similar to ABI-231,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0ab and 10bb both bound to the colchicine binding site of tubulin and overlapped well with
colchicine (Figure S1B and S1C). 10ab shows similar hydrogen bonding interactions to that of
ABI-231 and its increased activity can be attributed to the additional van der Waal interactions
afforded by the extra 4-methyl group that fills a small cavity in the ABI-231 complex bounded by
β-Leu253, β-Met257 and β-Ala314 (Figure 2D to 2F). 10bb as a 4-indole compound shares the
main chain ABI-231 hydrogen bonding interactions with α-Thr179 and β-Asp249. Although the
hydrogen bonding to β-Asp249 is weakened compared with that in ABI-231 or 10ab, the rotated
indole now allows a direct hydrogen bond between the indole NH and the main chain C=O of β-
Asn347 rather than the water mediated hydrogen bond seen in ABI-231 (Figure 2B and 2C). This
direct hydrogen bonding interaction also pull the 10bb slightly up from the pocket where the TMP
moiety occupies, allowing a separate water-bridged hydrogen bonding network between the
middle methoxy oxygen to the backbone NH in β-Cys239 and CO in β-Gly235. This water
molecule has also been seen in a number of previously reported high resolution crystal structures
3
7
with another scaffold of CBSIs . Collectively, the direct hydrogen bond from the 4-indole moiety,
the added water-bridged hydrogen bond at the TMP pocket, and the ability of the rotated indole to
still occupy its binding pocket are likely to explain its increased activity. Both ABI-231 and its
new analogs binding are not compatible with the original straight conformation of tubulin protein.
Specifically, the curve-to-straight transition of tubulin is hindered by steric clashes between ABI-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
31/10ab/10bb and the surrounding secondary elements (Figure S3), which is consistent with
previous reports 13, 38
.
New ABI-231 analogs showed potent activities against NCI-60 panel cell lines. Besides their
excellent antiproliferative activities observed in melanoma cell lines as tested in our lab, 10ab,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0bb, and 10bd were also tested against NCI-60 cell lines in both one-dose and full 5-dose assays.
The compounds effectively inhibited cancer cell growth in a variety of cancer types and were
particularly active against leukemia, colon, and prostate cancer (Figure S4A). They also produced
IC values in the low nanomolar range against the vast majority of the NCI-60 cell lines (Figure
5
0
S4B). It is clear that the new ABI-231 analogs 10ab, 10bb, and 10bd all exhibit strong inhibitory
effects against a wide range of cancers.
Inhibition of tubulin polymerization. We next determined the effects of the most potent new
analogs (10ab and 10bb) for their ability to inhibit tubulin polymerization, the primary
mechanisms of action of ABI-231. In a cell-free purified tubulin polymerization assay, 10 µM
concentrations of 10ab and 10bb were evaluated against vehicle control and docetaxel, a well-
documented inducer of tubulin polymerization. Figure 3A clearly shows that compounds 10ab
and 10bb potently inhibited tubulin polymerization, while the vehicle experienced an initial and
rapid increase in polymerization and docetaxel strongly and consistently promoted polymerization.
The effects on microtubule networks were then observed by immunofluorescence via confocal
microscopy. Microtubule morphology in WM164 cells was visualized after 18 h of treatment of
with 100 nM of 10ab, 10bb, docetaxel, or untreated control cells. As shown in Figure 3B, 10ab
and 10bb dramatically disrupt microtubules, resulting in fragmentation and disassembly.
Docetaxel works in the opposite manner, causing hyperpolymerization and the formation of dense,
aggregated bundles of microtubules. These mechanistic studies confirm a mode of action by which
ACS Paragon Plus Environment

Page 17 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
these compounds induce depolymerization of tubulin and disturb microtubule networks by
interacting with the colchicine binding site.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. New ABI-231 analogs inhibit tubulin polymerization. (A) Polymerization of purified
tubulin in a cell-free assay. Tubulin (3.33 mg/mL) was exposed to vehicle control or 10 µM of the
compounds (n = 2). Absorbance at 340 nm was monitored at 37 °C every minute for 30 min. (B)
Effect on microtubules of WM164 cells following 18 h exposure to 100 nM of 10ab, 10bb,
docetaxel or vehicle control. Confocal microscopy was utilized to visualize the microtubules
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
following immunofluorescent staining. Microtubules are shown in red and DNA is shown in blue.
Scale bar = 20 µm.
In vitro inhibition of cancer growth and migration. The long-term inhibitory growth effects
of 10ab and 10bb were tested in A375 cells in an anchorage-dependent colony formation assay.
The visual result and quantitative data are shown in Figure 4A and 4B. While the vehicle control
treated cells averaged 199.3 ± 10.7 colonies after 12 days, both 10ab and 10bb significantly
inhibited colony number, with an average of 81.7 ± 7.8 and 128.3± 9.1 colonies, respectively. This
represents a 59.0% and 35.6% decrease, respectively, from the control. These compounds out-
performed colchicine, which averaged 150.0 ± 5.3 colonies, corresponding to a 24.7% decrease in
colony number. Tubulin and microtubule function are also implicated in cell migration and
motility. Therefore, we assessed the effect of 10ab and 10bb in a wound healing assay. After
removing a confluent monolayer of adherent A375 or RPMI7951 melanoma cells, the remaining
cells were treated for 18-22 h with 5 or 25 nM of 10ab, 10bb, or colchicine and compared against
untreated control cells. The total closure was calculated as a percentage of the total wound area
after treatment compared to the total area of removed cells immediately prior to treatment. In A375
and RPMI7951 cell lines, the control cells were able to efficiently migrate into the wound channel
apprehended, recovering 60.7% ± 3.5% and 71.8% ± 1.8% of the area, respectively (Figure 4C
and 4D). Following 5 nM treatment, A375 cells only migrated into 48.5% ± 1.6%, 38.5 ± 2.2%,
and 46.0% ± 1.4% of the wound area for 10ab, 10bb and colchicine, respectively (Figure 4C and
4D). At higher concentrations of 25 nM, greater inhibition was observed, where 10ab treated cells
recovered 25.9% ± 1.6%, 10bb recovered 38.5% ± 2.2%, and colchicine treated cells recovered
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0.7% ± 2.2% (P < 0.0001). A significant difference was also observed in RPMI7951 cells treated
at 25 nM of 10ab, 10bb or colchicine, leading to wound closure averaging 35.6% ± 6.6%, 35.0%
ACS Paragon Plus Environment

Page 19 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
±
7.8%, and 42.8% ± 5.8%, respectively (P = 0.0004) (Figure 4C and 4D). Normalized to the
control cell lines, the average inhibition of cell migration for the treated cells was 53.5% for 10ab,
52.8% for 10bb, and 32.8% for colchicine. Taken together, these results demonstrate that 10ab
and 10bb potently inhibit proliferation and migration of cancer cells in vitro at low nM
concentrations.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. 10ab and 10bb inhibit colony growth and cancer cell migration. (A) Representative
images of A375 colony formation for control or treated cells. (B) Quantification of colony numbers
(n = 3). Statistical analysis was performed by ANOVA (P < 0.0001) followed by Dunnett’s
multiple comparison test. (C) Representative images of wound area in a scratch assay. Images were
obtained after 18-22 h for A375 (top) or RPMI7951 (bottom) cells after treatment with 5 or 25 nM
concentrations of 10ab, 10bb, colchicine, or untreated control. (D) Percentage of the wound closed
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for A375 cells (top) or RPMI7951 cells (bottom) (n = 3). Area of the wound channel was calculated
using ImageJ software. Statistical analysis was performed by ANOVA (P < 0.0001 for A375, P =
0.0004 for RPMI7951) followed by Dunnett’s multiple comparison test.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In vitro microsomal stability. To determine the most suitable candidates for in vivo efficacy
studies, we examined the in vitro metabolic stabilities of the most potent analogs by measuring
their half-life upon incubation with mouse, rat and human liver microsomes in the presence of an
NADPH regeneration system. The results are summarized in Table 3. Overall, all three new ABI-
2
31 analogs exhibited very good liver microsomal stabilities in three different species. 10bb,
which showed overall best in vitro potency and stability, was then selected for subsequent in vivo
studies.
Table 3. Microsomal stability study of tested compounds in liver microsome of different species
Compounds
Metabolic Stability (Mouse)
Metabolic Stability (Rat)
Metabolic Stability (Human)
t1/2 (h) Clint (ml/Min/Kg)
t1/2 (h)
Clint
60.66
15.20
23.60
11.34
10.28
t1/2 (h)
1.29±0.09
6.74±0.59
1.80±0.09
5.26±0.35
15.19±2.09
Clint
36.18
6.94
Verapamil
ABI-231
0.94±0.06
3.76±0.24
2.42±0.15
5.04±0.30
5.56±0.35
1.68±0.16
5.13±0.26
6.70±0.82
6.72±0.74
7.61±0.91
12.41
4.05
3.10
3.10
2.73
1
0ab
25.98
8.90
1
1
0bb
0bd
3.10
10bb inhibits melanoma tumor growth and metastasis in vivo. The anticancer efficacy of
0bb was evaluated in vivo in two melanoma mouse models. All animal experiments performed
1
in this report were conducted in compliance with the NIH and institutional guidelines, using
approved animal protocols from the University of Tennessee Health Science Center. First, a
xenograft model of nude mice challenged with A375 melanoma subcutaneously was treated every
ACS Paragon Plus Environment

Page 21 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
other day via i.p. injection with 15 mg/kg or 30 mg/kg doses of 10bb, paclitaxel, or a vehicle
solution only. After 20 days, the group receiving 15 mg/kg treatments of 10bb achieved an average
tumor growth inhibition (TGI) of 82.3% and the group receiving the higher dose of 30 mg/kg
achieved a TGI of 90.6% (Figure 5A). Both treatment groups were more effective against the
xenograft tumors than the paclitaxel group, which had an average TGI of 68.6%. One-way
ANOVA analysis (P = 0.0014) followed by Dunnett’s multiple comparison test also revealed that
there was a significant decrease compared to the control for both concentrations of 10bb treated
groups (P < 0.01) and the paclitaxel treatment group (P < 0.05). Additionally, the final tumor
weights of the 10bb 15 mg/kg and 30 mg/kg treated groups were 67.3% and 70.1% less than the
average final tumor weight for the control group, respectively (Figure 5B). One-way ANOVA (P
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
0.0024) followed by Dunnett’s multiple comparison test confirmed statistical significance
between average tumor weights of the 10bb treatment groups compared to the control (P < 0.01).
To investigate the ability of 10bb to impede cancer metastasis, it was also tested in a lung
metastasis mouse model at 30 mg/kg doses in C57BL mice (Figure 5C). After inoculation with
melanoma cells via the lateral tail vein, mice were treated for two weeks at the same dosing
schedule as before. The average tumor index based on lung metastasis nodule size and the number
was 13.6 ± 2.9 for the control group and 2.6 ± 0.6 for the 10bb treatment group. This was
significantly less than the control based on an unpaired student’s t test (P = 0.0013) and represents
an 80.9% decrease in metastasis. Mouse weight and activity were monitored and for both the
xenograft model and lung metastasis model (Figure 5D and 5E). Pathohistological analysis was
performed on the major organs collected from both in vivo studies, including heart, liver, kidneys,
lungs, and spleen, to assess for acute toxicities. Sections stained with hematoxylin and eosin
revealed no apparent drug-related injury or morphological abnormalities in the cellular structure
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of the various tissues in both the xenograft mouse model (Figure 6A) and the experimental lung
metastasis mouse model (Figure 6B). Additionally, in vitro pharmacological profiling was utilized
to identify undesirable off-target effects and assess the safety profile. 10bb was evaluated at 1 µM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
300~500 of its IC values against cancer cells as determined in Table 1) in the Safety47™ Panel
5
0
which includes the assessment of the functional response of 47 major physiologically important
targets across 78 assays (Figure 6C). Only a value higher than |70%| response is considered
significant. 10bb only showed hit responses in only 2 of the 78 assays, which were norepinephrine
transporter (NET) and serotonin receptor 2B (HTR2B) antagonism. These findings support a good
safety profile for 10bb based on the in vitro and in vivo data.
Tumor volume
final tumor weight
lung metastasis
4
3
2
1
0
40
30
20
10
0
2000
1500
1000
500
0
vehicle control
1
1
0bb 15 mg/kg
0bb 30 mg/kg
pax 15 mg/kg
*
*
*
*
*
*
*
**
**
0
5
10
15
20
vehicle
control
10bb
10bb
pax
vehicle
control
10bb
30 mg/kg
days
15 mg/kg 30 mg/kg 15 mg/kg
A
B
C
mouse body weight
xenograft model
mouse body weight
lung metastasis model
vehicle control
vehicle control
10bb 30 mg/kg
20
0
20
0
1
1
0bb 15 mg/kg
0bb 30 mg/kg
pax 15 mg/kg
-
20
-20
0
5
10
15
20
0
5
10
15
20
days
days
D
E
Figure 5. 10bb inhibits melanoma tumor growth and metastasis in vivo. (A) Tumor volumes in an
A375 xenograft model in nude mice (n = 10). Statistical significance for final tumor volume was
determined by ANOVA (P = 0.0014) followed by Dunnett’s multiple comparison test. (B)
ACS Paragon Plus Environment
2
2

Page 23 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Individual tumor weights. Error bars represent 95% CIs. Statistical significance was determined
by ANOVA (P = 0.0024) analysis followed by Dunnett’s multiple comparison test. (C) B16F10
lung metastasis in C57BL/6 mice. The graph represents the tumor index for mice (n = 10) and
error bars represent 95% CIs. An unpaired student’s t-test was performed to show statistical
significance between the groups (P = 0.0013). Mouse body weights in the xenograft model (D)
and lung metastasis model (E). Graph represents percent change in body weight compared to the
starting weight.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. 10bb does not cause apparent toxicities to major organs. Pathological sections of major
tissues (heart, kidney, liver, lung, and spleen) were obtained from the (A) A375 xenograft and (B)
B16F10 lung metastasis studies. Organs were stained with Hematoxylin and Eosin (H&E) and
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
representative images were captured. (C) In vitro pharmacological profiling of 10bb to assess
potential off-target effects to major targets at 1 µM concentrations using the Safety47™ Panel (n
= 2). Graph represents mean percent response ± range. Values in between -70% and +70%
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(indicated with dashed lines) are considered insignificant.
1
0bb can circumvent paclitaxel resistance in vitro and vivo. We next investigated the
potential for this scaffold of tubulin inhibitor in overcoming resistance mechanisms to which
taxanes are susceptible. We initially determined their activity against PC-3, DU-145, and their
corresponding paclitaxel resistant PC-3/TxR and DU-145/TxR cell lines in vitro. The resistance
index was determined by dividing the average IC for the resistance cells by their parental cell
5
0
lines. We found that paclitaxel had a resistant index of 103.5 and > 869 for PC-3/TxR and DU-
45/TxR, respectively. Remarkably, 10bb demonstrated greater potency against the resistant cells,
1
showing a resistance index of 0.7 and 2.2 (Table 4). Therefore, we wanted to confirm the ability
of 10bb to escape paclitaxel resistance in vivo. First, we confirmed the efficacy of paclitaxel
against tumors that had not developed resistance. Nude mice bearing PC-3 xenograft tumors were
dosed with 15 mg/kg paclitaxel or the vehicle solution every other day for three weeks. As
expected, paclitaxel was effective in inhibiting tumor growth tumor growth, resulting in tumor
growth inhibition (TGI) of 101.2% (P = 0.003) (Figure 7A). Paclitaxel also caused an average
reduction in tumor weight by 79.3% which was significantly less than the vehicle treated group
(P= 0.0003) (Figure 7B). We next tested our compound 10bb against a taxane-resistant PC-3 (PC-
3
9
3/TxR) model that was developed previously by others . At the same dosing concentrations and
frequency as the previous model, paclitaxel showed a marked decline in anticancer capabilities,
yielding a TGI of only 37.8% (Figure 7D) and a modest reduction in tumor weight of 34.5%
(Figure 7E). In contrast, 30 mg/kg treatments of 10bb produced a significant reduction in tumor
ACS Paragon Plus Environment
2
4

Page 25 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
growth by 83.8% and also decreased the average tumor weight by 62.8%. ANOVA analysis (P <
0
.0001) followed by Dunnet’s multiple comparison test confirmed a significant decrease in final
tumor volume and weight (P < 0.0001) for the 10bb treated groups. Animal activity and body
weights were monitored and recorded throughout the experiment, and major deviations in body
weight were not observed (Figure 7C and 7F).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 4. In vitro cytotoxicity of 10ab and 10bb in taxol resistant cancer cell lines.
Cell lines
IC50±SEM(nM)
Colchicine
10.4 ± 0.5
37.4 ± 2.2
3.6
10ab
5.2 ± 0.4
3.9 ± 0.3
0.8
10bb
89.9 ± 6.3
66.9 ± 4.4
0.7
Paclitaxel
1.1 ± 0.2
113.9 ± 4.3
103.5
Docetaxel
0.5 ± 0.1
27.9 ± 1.1
55.8
PC-3
PC-3/TxR
aRI
DU-145
DU-145/TxR
aRI
13.8 ± 1.0
25.2 ± 2.0
1.8
169.1 ± 4.4
379.1 ± 80.0
2.2
35.5 ± 3.5
>1,000
1.15± 0.2
>1,000
0.10 ± 0.01
352.3 ± 118.2
>3,523
>28
>869
a
RI: resistance index is calculated by dividing the IC value of compound in resistant cell lines
5
0
by the IC in the corresponding parental cell lines.
5
0
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 7. 10bb is effective in vivo at inhibiting Taxol resistant tumor growth. Two models were
investigated in nude mice, one with parental PC-3 xenograft tumors and another with Taxol
resistant PC-3/TxR tumors. (A) Tumor volumes for the PC-3 xenograft model (n = 7). Statistical
analysis was performed by an unpaired t test, showing a significant difference for the final volumes
of the paclitaxel treated group compared to the control group (P = 0.003). (B) Final tumor weights
for the PC-3 xenograft model. Error bars represent 95% CIs. An unpaired t test showed a
significant difference between the paclitaxel and vehicle treated groups (P = 0.0003). (C) Mouse
body weights expressed as percent change compared to the starting weight in the PC-3 model. (D)
Tumor volumes for the PC-3/TxR xenograft model (n = 8). Statistical analysis was performed by
ANOVA (P < 0.0001) followed by Dunnett’s multiple comparisons test for the final tumor
volumes. (E) Final tumor weights for PC-3/TxR mice. Error bars represent 95% CIs. Statistical
analysis was performed by ANOVA (P = 0.0001) followed by Dunnet’s multiple comparisons test
(F) Body weights for PC-3/TxR mouse model. Graph represents mean body weight change as a
percentage compared to initial weight.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSION
In this report, a SAR investigation of ABI-231 focusing on the modification of the indole moiety
is reported, revealing that 10ab and 10bb are more potent derivatives than the ABI-231 prototype.
The X-ray crystal structures of ABI-231, 10ab, and 10bb in complex with tubulin were obtained,
which confirmed that they directly bind to the colchicine site in tubulin, and revealed their
molecular mechanisms responsible for their improved interactions to tubulin proteins and the
resulted increased antiproliferative activities. We also demonstrated the abilities of the new
analogs to block tubulin polymerization and to inhibit cancer cell motility and migration. In vivo,
1
0bb effectively inhibited tumor growth and lung metastasis in melanoma xenograft and lung
ACS Paragon Plus Environment
2
6

Page 27 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
metastasis models without showing signs of acute toxicity. More interestingly, 10bb was highly
potent in inhibiting tumor growth of PC-3/TxR prostate xenograft tumors. This suggests, along
with the in vitro cell viability results, that 10bb may circumvent certain types of MDR that taxanes
cannot escape. Taken together, these new ABI-231 analogs represent promising candidates for
further development as a new generation of tubulin inhibitors for cancer treatment.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EXPERIMENTAL SECTION
General chemistry. Tetrahydrofuran was distilled from sodium-benzophenone. All other
solvents and chemical reagents were obtained from commercial sources and directly used without
further purification. Glassware was oven-dried before use. All reactions were performed under an
argon atmosphere. TLC was performed on silica gel 60 GF254 and monitored under UV light or
visualized using phosphomolybdic acid reagent. Flash chromatography was performed on 230-
4
00 mesh silica gel (Fisher Scientific). Melting points were recorded on a MPA100 Automated
Melting Point Apparatus. NMR spectra were obtained on a Bruker Ascend 400 (Billerica, MA)
spectrometer or a Varian Inova-500 spectrometer (Agilent Technologies, Santa Clara, CA). HR-
MS were obtained on Waters Acquity UPLC linked to a Waters Acquity Photodiode Array
Detector and Waters qTof mass detector. All compounds reported herein with biological data had
purities ≥95% as determined by HPLC. The purity of associated compounds was verified by the
HPLC study performed on BEH C18 (2.1ꢀ×ꢀ50ꢀmm, 1.7ꢀμm) column using a mixture of solvent
acetonitrile/water (with 0.1% formic acid) at a flow rate of 0.3ꢀml/min and monitoring by UV
absorption at the appropriate wavelength. Chemical shifts are given in ppm. Tetramethylsilane
(TMS) is used as an internal reference for NMR spectra taken in chloroform-d. All coupling
constants (J) are given in Hertz (Hz).
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 57
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Synthesis of tert-butyldimethyl((1-(3,4,5-trimethoxyphenyl)allyl)oxysilane (2). To a solution
o
of 3,4,5-trimethoxybenzaldehyde (30 g, 0.153 mol) in anhydrous THF (100 mL) at 0 C under
argon was added vinylmagnesium bromide solution (168 mL, 0.168 mol, 1.0 M in THF) dropwise,
the resulting mixture was stirred for 1 h. On completion, reaction was quenched with saturated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH Cl. The mixture was then extracted with ethyl acetate, washed with brine and dried with
4
anhydrous Na SO . The combined extracts were evaporated under vacuum to give 1 as crude oil
2
4
3
3
which was used directly for next step without purification . Imidazole (15.6 g, 0.229 mol) and
tert-butyldimethylsilyl chloride (34.4 g, 0.229 mol) were sequentially added to a solution of the
above crude oil in dichloromethane (150 mL) in a round bottom flask with stirring under argon.
The resulting suspension was stirred for 2 h. Water was then added and the mixture was extracted
with dichloromethane, washed with brine and dried with anhydrous Na SO . The combined
2
4
extracts were evaporated to give crude oil mixture which was directly used for next step without
purification.
Synthesis of 3-((tert-butyldimethylsilyl)oxy)-3-(3,4,5-trimethoxyphenyl)propane-1,2-diol
(
3). To a solution of the compound 2 (44.0 g, 0.13 mol) in acetone (150 mL) was added N-
methylmorpholine N-oxide (22.9 g, 0.195 mol) and osmium tetroxide (40 mg, 0.157 mmol) in tert-
butanol (5 mL) at room temperature with stirring. After 48 h, acetone was removed under vacuum,
water and ethyl acetate were then added and the organic phase was separated, washed with brine
and dried with Na SO . Evaporation under vacuum gave the oily residue that was purified with
2
4
flash chromatography on silica. Elution with hexane/ethyl acetate (2:1-1:1) gave diol 3 as colorless
1
oil (45.1 g, 79% for three steps). H NMR (400 MHz, Chloroform-d) δ 6.55 (d, J = 3.9 Hz, 2H),
4
.65 (dd, J = 35.9, 6.1 Hz, 1H), 3.84 (t, J = 1.9 Hz, 9H), 3.77 – 3.40 (m, 3H), 2.87 – 1.97 (m, 2H),
0.91 (d, J = 0.7 Hz, 9H), 0.08 (d, J = 9.9 Hz, 3H), -0.07 – -0.19 (m, 3H).
ACS Paragon Plus Environment
2
8

Page 29 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis of 3-((tert-butyldimethylsilyl)oxy)-3-(3,4,5-trimethoxyphenyl)propane-1,2-diyl
dimethanesulfonate (4). To a solution of compound 3 (45.0 g, 0.12 mol) and Methanesulfonyl
chloride (23.4 mL, 0.30 mol) in anhydrous dichloromethane (100 mL) in a round bottom flask at
0 oC under argon, trimethylamine (36.9 mL, 0.26 mol) was added dropwise with vigorous stirring.
The resulting mixture was stirred overnight, water was then added and the reaction mixture was
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
extracted with dichloromethane, washed with brine and dried with anhydrous Na SO . The
2
4
combined extracts were evaporated under vacuum to give the oily crude, which was directly used
for next step without purification.
Synthesis of tert-butyl(2,3-diazido-1-(3,4,5-trimethoxyphenyl)propoxy)dimethylsilane (5).
To a solution of mesylate 4 (45.0 g, 0.085 mol) in anhydrous N,N-dimethylformamide (100 mL)
in a round bottom flask under argon, sodium azide (30 g, 0.46 mol) was added in portions with
stirring. The resulting mixture was refluxed for overnight. The precipitate was filtered off and
washed with dichloromethane. The combined filtration was evaporated under vacuum to give the
oily crude which was directly used for next step without purification.
Synthesis of 3-((tert-butyldimethylsilyl)oxy)-3-(3,4,5-trimethoxyphenyl)propane-1,2-
diamine (6). A suspension of azide 5 (20.0 g, 0.047 mol) and 10% Pd/C (0.2 g) in ethyl acetate-
methanol (1:1, 50 mL) was charged with hydrogen and stirred overnight, the reaction mixture was
filtered off and washed with dichloromethane-methanol (1:1). The combined filtration was
evaporated under vacuum to give the oily crude, which was purified with flash chromatography
on silica. Pure diamine 7 was eluted out with dichloromethane/methanol (15:0-15:1) as slightly
1
yellowish oil (10.6 g, 61%). H NMR (400 MHz, Chloroform-d) δ 6.51 (d, J = 13.4 Hz, 2H), 4.45
(dd, J = 34.5, 5.1 Hz, 1H), 3.83 (d, J = 1.2 Hz, 9H), 2.97 – 2.46 (m, 3H), 0.90 (d, J = 10.8 Hz, 9H),
-0.17 (d, J = 8.5 Hz, 3H).
ACS Paragon Plus Environment
2
9