Optimization, pharmacophore modeling and 3D-QSAR studies of sipholanes as breast cancer migration and proliferation inhibitors

Article abstract of DOI:10.1016/j.ejmech.2013.11.039

Sipholenol A, a triterpene isolated from the Red Sea sponge Callyspongia siphonella, was previously shown to reverse multidrug resistance in P-glycoprotein-overexpressing cancer cells. Moreover, sipholanes showed promising in vitro inhibitory effects against the invasion and migration of the metastatic human breast cancer cell line MDA-MB-231. The breast tumor kinase (Brk), a mediator of cancer cell phenotypes important for proliferation, survival, and migration, was proposed as a potential target. This study reports additional semisynthetic optimization of sipholenol A esters to improve the breast cancer antimigratory and antiproliferative activities as well as Brk phosphorylation inhibition. Fifteen new sipholenol A analogs (25-39) were semisynthesized. Sipholenol A 4β-4′,5′-dichlorobenzoate ester (29) was the most potent, with an IC₅₀ value of 1.3 μM in the migration assay. The level of Brk phosphorylation inhibition of 29 was assessed using the Z′-LYTE kinase assay and Western blot analysis. Active analogs showed no toxicity on the non-tumorigenic epithelial breast cell line MCF10A at doses equal to their IC₅₀ values or higher in migration and proliferation assays, suggesting their selectivity towards malignant cells. Pharmacophore modeling and 3D-QSAR studies were conducted to identify important pharmacophoric features and correlate 3D-chemical structure with activity. These studies provided the evidence for future design of novel antimigratory compounds based on a simplified sipholane structure possessing rings A and B (perhydrobenzoxepine) connected to substituted aromatic esters, with the elimination of rings C and D ([5,3,0]bicyclodecane system). This will enable the future synthesis of the new active entities feasibly and cost-effectively. These results demonstrate the potential of marine natural products for the discovery of novel scaffolds for the control and management of metastatic breast cancer.

Full text of DOI:10.1016/j.ejmech.2013.11.039

European Journal of Medicinal Chemistry 73 (2014) 310e324
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Optimization, pharmacophore modeling and 3D-QSAR studies
of sipholanes as breast cancer migration and proliferation inhibitors
*
Ahmed I. Foudah, Asmaa A. Sallam, Mohamed R. Akl, Khalid A. El Sayed
Department of Basic Pharmaceutical Sciences of College of Pharmacy, University of Louisiana at Monroe, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Sipholenol A, a triterpene isolated from the Red Sea sponge Callyspongia siphonella, was previously
shown to reverse multidrug resistance in P-glycoprotein-overexpressing cancer cells. Moreover, sipho-
lanes showed promising in vitro inhibitory effects against the invasion and migration of the metastatic
human breast cancer cell line MDA-MB-231. The breast tumor kinase (Brk), a mediator of cancer cell
phenotypes important for proliferation, survival, and migration, was proposed as a potential target. This
study reports additional semisynthetic optimization of sipholenol A esters to improve the breast cancer
antimigratory and antiproliferative activities as well as Brk phosphorylation inhibition. Fifteen new
Received 8 October 2013
Received in revised form
21 November 2013
Accepted 24 November 2013
Available online 12 December 2013
Keywords:
sipholenol A analogs (25e39) were semisynthesized. Sipholenol A 4
was the most potent, with an IC50 value of 1.3 M in the migration assay. The level of Brk phosphorylation
b
-40,50-dichlorobenzoate ester (29)
Antimigratory
Antiproliferative
Breast cancer
Brk
Pharmacophore modeling
3D-QSAR
m
inhibition of 29 was assessed using the Z0-LYTEÔ kinase assay and Western blot analysis. Active analogs
showed no toxicity on the non-tumorigenic epithelial breast cell line MCF10A at doses equal to their IC50
values or higher in migration and proliferation assays, suggesting their selectivity towards malignant
cells. Pharmacophore modeling and 3D-QSAR studies were conducted to identify important pharma-
cophoric features and correlate 3D-chemical structure with activity. These studies provided the evidence
for future design of novel antimigratory compounds based on a simplified sipholane structure possessing
rings A and B (perhydrobenzoxepine) connected to substituted aromatic esters, with the elimination of
rings C and D ([5,3,0]bicyclodecane system). This will enable the future synthesis of the new active
entities feasibly and cost-effectively. These results demonstrate the potential of marine natural products
for the discovery of novel scaffolds for the control and management of metastatic breast cancer.
Ó 2013 Elsevier Masson SAS. All rights reserved.
Sipholane
1. Introduction
induced multidrug resistance (MDR) in human epidermoid cancer
cells [8e10]. Recently, sipholanes were reported as inhibitors of
The marine environment was shown to be a rich source of
diverse novel chemical structures with promising anticancer
activities [1,2]. Besides the chemical novelty associated with those
compounds, some of them possess novel mechanisms of action
[3,4]. The Red Sea sponge Callyspongia (Siphonochalina) siphonella is
a rich source of triterpenoids. So far, thirty triterpenoids have been
isolated from this sponge, possessing four different skeletons,
namely, sipholane, siphonellane, neviotane, and dahabane [5e10].
Among these, the sipholane triterpenoids, first reported by Kash-
man and co-workers, are the major group and include sipholenol A
(1) and sipholenone A (2) [5]. The sipholane skeleton is composed
of perhydrobenzoxepine (rings A and B) and [5,3,0]bicyclodecane
system (rings C and D), linked together through an ethylene bridge
[6,7]. Sipholenol A and its analogs were able to reverse P-gp-
breast cancer cell migration and invasion [11].
Breast cancer is the most common cancer and the second-
leading cause of cancer-related death among women in the
United States [12,13]. It is estimated that over 230,000 new cases of
invasive breast cancer will be diagnosed in 2013, with up to 20% of
these patients likely to develop metastatic disease. The develop-
ment of chemotherapy resistance continues to be the main problem
in the treatment of metastatic breast cancer [13]. Consequently,
there is an urgent need to discover new entities with new molec-
ular targets and low susceptibility to common drug resistance
mechanisms in order to improve therapeutic outcomes [2,13].
The intracellular protein kinase, Brk (known as breast tumor
kinase, or protein tyrosine kinase 6, PTK6) has been implicated in
the development and progression of a number of different tumor
types [14]. Brk is one of the Src family tyrosine kinases, which has
been cloned from metastatic breast tumor samples and cultured
human melanocytes [15,16]. Brk is overexpressed in up to 86% of
* Corresponding author. Tel.: þ1 318 342 1725; fax: þ1 318 342 1737.
E-mail address: elsayed@ulm.edu (K.A. El Sayed).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.11.039

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
311
invasive human breast tumors, prostate and colon carcinomas [17].
Brk was detected in approximately two-thirds of breast tumors
analyzed, where approximately a third of these showed Brk over-
expression by levels ranging from five- to forty three-fold
compared to normal tissue [17]. Brk is normally expressed in
differentiating epithelial cells of the intestine, skin, prostate, and
oral cavity, where it has been shown to promote cellular differen-
tiation, apoptosis, and more recently to mediate migration/wound-
healing [15]. However, it is neither detected in normal mammary
tissues or fibroadenomas, nor at various stages of mammary
development. Overexpression of Brk in tumor samples relative to
the restricted levels in normal or differentiating tissues suggest that
Brk may have a role in the processes underlying tumorigenesis,
such as promotion of cancer cell proliferation, migration, and sur-
vival [16]. Recent studies suggest that Brk can promote the migra-
tion of breast cancer cells through multiple mechanisms and in
response to a number of different ligands. Its expression levels in-
crease in association with the carcinoma content of breast tumors,
tumor grade, and invasiveness [15e19]. These observations
strongly suggest that high levels of de novo expression of Brk make
it an attractive therapeutic target in breast cancer. Furthermore,
inhibition of Brk kinase activity may provide a potentially novel
approach to sensitize the response of tumor cells to other chemo-
therapeutics and prevent or inhibit metastasis of cancer with
enhanced therapeutic windows.
A previous study from our laboratory reported the potent
antimigratory and anti-invasive activities of several semisynthetic
sipholenol A analogs (1e24) against the metastatic breast cancer
cells, MDA-MB-231 [11]. Aromatic sipholenol A esters have shown
improved antimigratory activity over the ether and carbamate an-
alogs. Accordingly, the main objective of this study was to further
investigate the influence of various aromatic ring substitutions on
the antimigratory and antiproliferative activities of sipholenol
A aromatic ester analogs in breast cancer cells. In addition, the
promising activity of sipholenol A and analogs encouraged the
reinvestigation of the source sponge C. siphonella to test additional
fifteen substituted aromatic and heteroaromatic esters of 1 (25e39)
were semisynthesized to explore the substitution effects around
the aromatic ring and establish a preliminary SAR profile. These
new analogs were prepared using various acid chlorides and N,N-
dimethylaminopyridine (DMAP) as a catalyst (Scheme 1) [11,20].
The ESI-MS spectrum of 25 showed an [M ꢀ H]ꢀ peak at m/z
597.40, suggesting the molecular formula C37H54FO5 and possible
4-O-50-fluorobenzoate ester analog of sipholenol A (1). Esterifica-
tion at C-4 was evident from 1H and 13C NMR data (Tables 1 and 4).
Proton H-4 doublet was downfield shifted to dH 5.19, compared
with its parent 1 (DdH þ1.27). This was associated with the down-
field shift of carbon C-4 (DdC þ3.66), as well as the upfield shift of
carbon C-3 (DdC ꢀ1.37). Esterification was further supported by a
3J-HMBC correlation of H-4 with the ester carbonyl C-10 (dC 164.9).
The aromatic methine doublet (dH 8.15, J ¼ 8.2 Hz) was assigned to
protons H-30 and H-70 based on the 3J-HMBC correlations with C-10
and C-50 (dC 147.2). Therefore, compound 25 was found to be the
4-O-b
-fluorobenzoate ester of 1. Similarly, analysis of the 1H, 13C
NMR, and MS data of compounds 26e39 confirmed their identity.
2.2. Biological evaluations and structureeactivity relationship
Fifty-nine natural and semisynthetic sipholane analogs (1e59)
[8,10,11] were screened in vitro in the MTT wound-healing, and
Western blot assays [21e23]. The human mammary carcinoma cell
lines MDA-MB-231 and MCF-7 were both used in the MTT assay
whereas MDA-MB-231 cell line was used in the wound-healing
assay (WHA) and Western blot analysis. The non-tumorigenic
breast cell line MCF10A was used to evaluate cytotoxicity in the
MTT assay. Additionally, a Z0-LYTE kinase assay was used to
evaluate the ability of the most active analog to inhibit PTK6
phosphorylation.
2.2.1. Antimigratory activity against the highly metastatic MDA-
MB-231 breast cancer cell line
All compounds were evaluated for their antimigratory effect in
the WHA using the metastatic breast cancer cell line, MDA-MB-231.
The WHA is a simple method to study directional cell migration
in vitro [24]. The known antimigratory lead 4-S-ethylphenyl-
methylene hydantoin (S-Ethyl) was used as a positive control at a
related analogs to establish
a preliminary structureeactivity
relationship (SAR). This study reports the antimigratory and
antiproliferative activities of twenty known natural terpenoids
(40e59), possessing two novel sipholane skeletons. Pharmaco-
phore modeling and predictive quantitative structureeactivity
relationship (3D-QSAR) studies were carried out using the most
50
mM dose [25]. The concentrations required to inhibit the
migration of 50% of cells across the wound, IC50 values, were
calculated (Table 8). Fig. 1 shows the effect of the most active an-
alogs 25e27, 29e31 and 39 on cell migration compared with the
vehicle and positive controls. The parent sipholenol A (1) and most
of the natural sipholanes 44e54 and 57e59 showed limited anti-
active sipholanes (IC50 > 25
mM) in order to identify important
pharmacophoric features of the active analogs and correlate 3D-
chemical structure with the antimigratory activity that would be of
value in guiding lead sipholane design.
migratory activity, with IC50 values greater than 40 mM. Generally,
2. Results and discussion
all sipholenol A semisynthetic analogs showed better antimigratory
activity profile than parent compound 1 and natural sipholanes.
This highlights the importance of replacing the hydrogen-bond
donor of the C-4 secondary alcohol group with various sub-
stituents (ester, ether, oxime and carbamate) to the antimigratory
activity (Table 8). The activity profile of natural sipholanes also
2.1. Chemistry
Sipholenol A-4-O-benzoate and related analogs were the most
active in breast cancer migration inhibition assays [11]. Therefore,
Scheme 1. General semisynthetic scheme of sipholenol A (1) analogs.

312
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
25
OR
OR1
RO
N
4
A 4
O
O
4
H
H
H
H
30
20
O
D19
H
OR2
H
8
1
1
B
H
H
H
OH
H
C
10
H
H
15
HO
HO
H
H
HO
H
28
Compound R1 R2
Compound
R
Compound R
H
a
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
s
t
u
v
w
x
H
H
a
H
a
b
c
d
e
f
H
c
g
1
5
6
7
8
9
3
22
23
24
12
13
14
15
16
17
19
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
O
OH
4
4
O
O
H
H
H
H
20
10
11
18
20
21
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
19
OH
1
H
g
H
H
10
O
1'
HO
H
H
CH3
2'
4
2
a
O
1'
O
O
O
1'
7'
7'
O
7'
7'
1'
7'
1'
1'
1'
OH
4'
3'
N
3'
3'
3'
CH3
3'
7'
O
Cl
b
c
d
e
f
g
O
O
S
O
1'
O
7'
O
1'
O
1'
7'
1'
7'
8'
2'
N
1'
O
3'
H
N
H
3'
7'
3'
3'
4'
8'
NO2
CF3
F
h
i
j
k
l
O
1'
O
O
7'
1'
7'
O
1'
O
1'
1'
O
1'
7'
7'
7'
3'
3'
3'
Cl
3'
5'
5'
OCH3
O
NO2
CF3
p
O2N
Cl
Cl
m
n
o
q
r
O
O
1'
O
1'
O
1'
O
7'
7'
7'
1'
1'
3'
O
O
CF3
NO2
5'
5'
5'
F3C
F
CF3
3'
F
F
s
t
u
v
w
x
suggests that variations at rings C and D have little or no impact on
the antimigratory activity. Ester analogs of 1 were more active than
ether, oxime and carbamate analogs. This illustrates the important
role of the ester carbonyl group as well as the distance limit
between the aromatic side chain and C-4 oxygen for optimal anti-
migratory activity. Longer distances, as represented by oximes or
carbamates, significantly decrease the antimigratory activity. Aro-
matic esters showed a better activity profile than aliphatic esters;
for example, the benzoate 8 was twice as active as the acetate 5. The
aromatic ester was tolerant to bioisosteric substitution as evident
from the comparable activities of isonicotinate (9) and furoate (38)
esters to the benzoate ester 8.
In addition to the requirement of an aromatic ester at C-4, ortho-,
meta- and para-substitutions (C-30, C-40 and C-50, respectively) of
the aromatic moiety also appeared to influence the antimigratory
activity. An electron-donating group (such as OCH3 and CH3) at the
para-position (C-50) reduced the antimigratory activity in the
following order: 28 < 10 < 8 (Table 8). On the other hand, an
electron-withdrawing group (EWG) at the para-position, as in 25,
enhanced the antimigratory activity by seven-fold as compared to
electron-donating group (EDG) as in 28, and three-fold enhance-
ment when compared to the unsubstituted aromatic moiety in 8.
However, the influence of the electron-withdrawing group on
activity appeared to have a limit. Using Craig’s plot, several EWGs
were tried and their effect on the antimigratory profile was evalu-
ated. EWG of moderate hydrophobicity was associated with a better
antimigratory activity as can be seen from the activities of
25 > 11 > 27 (Table 8). It was interesting to note that replacing the
50-F group in 25 with a 50-NO2 in 26 did not improve the activity as
would have been expected. This might be due to a steric factor at the
para-position or the involvement of the nitro group in unfavorable
interaction(s) with the target receptor. Moreover, an EWG at the
meta-position (as in 30 and 31) was also associated with a better
antimigratory activity versus the unsubstituted benzoate 8. How-
ever, analogs with ortho-substitution like 32 and 34 were less active
than the unsubstituted benzoate ester (8).
The Topliss Scheme, which designates a special series of sub-
stituents for use in sequence to find suitably substituted aromatic
compounds with biological properties superior to the parent
unsubstituted aromatic compound, was partially applied to the

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
313
Table 1
13C NMR spectroscopic data of compounds 25e30.a
Table 2
13C NMR spectroscopic data of compounds 31e36.a
Position
dC
Position
dC
25
26
27
28
29
30
31
32
33
34
35
36
1
42.7, qC
42.7, qC
42.7, qC
42.7, qC
42.7, qC
42.7, qC
1
42.7, qC
42.7, qC
42.7, qC
42.7, qC
42.7, qC
42.7, qC
2
3
35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2
23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2
2
3
35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2 35.1, CH2
23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2 23.2, CH2
4
5
7
80.8, CH
77.8, qC
77.3, CH
80.8, CH
77.8, qC
77.3, CH
80.8, CH
77.8, qC
77.3, CH
79.7, CH
77.8, qC
77.3, CH
80.8, CH
77.8, qC
77.3, CH
80.9, CH
77.8, qC
77.3, CH
4
5
7
79.7, CH
77.8, qC
77.3, CH
80.8, CH
77.8, qC
77.3, CH
79.7, CH
77.8, qC
77.3, CH
79.7, CH
77.8, qC
77.3, CH
80.8, CH
77.8, qC
77.3, CH
79.7, CH
77.8, qC
77.3, CH
8
9
26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2
39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2
8
9
26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2 26.7, CH2
39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2 39.5, CH2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
10
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
10
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
72.4, qC
56.5, CH
26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2
33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2
57.4, CH
143.0, qC 143.0, qC 143.0, qC 143.1, qC 143.0, qC 143.0, qC
121.4, CH 121.4, CH 121.4, CH 121.3, CH 121.4, CH 121.5, CH
24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2
48.9, CH
82.2, qC
37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2
24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2
52.8, CH
35.4, qC
13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3
29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3
21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3
30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3
30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3
25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3
31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3
29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3
164.9, qC 165.9, qC 164.5, qC 165.5, qC 163.8, qC 163.4, qC
125.7, qC 136.2, qC 133.7, qC 122.9 qC 137.8 qC 132.2, qC
131.5, CH 130.8, CH 129.9, CH 131.5, CH 131.4, CH 124.2, CH
115.4, CH 123.8, CH 125.7, CH 113.8, CH 130.3, qC 148.2, qC
147.2, qC 152.2, qC 134.6, qC 163.5, qC 133.0, qC 127.6, CH
115.4, CH 123.8, CH 125.7, CH 113.8, CH 130.9, CH 130.1, CH
131.5, CH 130.8, CH 129.9, CH 131.5, CH 128.7, CH 135.6, CH
26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2 26.8, CH2
33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2 33.8, CH2
57.4, CH
143.1, qC 143.0, qC 143.1, qC 143.1, qC 143.0, qC 143.1, qC
121.3, CH 121.4, CH 121.3, CH 121.3, CH 121.4, CH 121.3, CH
24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2 24.8, CH2
48.9, CH
82.2, qC
37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2 37.2, CH2
24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2 24.9, CH2
52.8, CH
35.4, qC
13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3 13.2, CH3
29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3 29.1, CH3
21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3 21.8, CH3
30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3
30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3 30.1, CH3
25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3 25.7, CH3
31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3 31.7, CH3
29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3 29.5, CH3
163.4, qC 164.9, qC 163.4, qC 163.4, qC 164.5, qC 165.9, qC
132.2, qC 133.3, qC 125.9, qC 118.6, qC 133.3, qC 121.9, qC
124.2, CH 133.5, qC 146.8, qC 144.2, qC 133.5, qC 160.6, qC
134.6, qC 129.9, CH 123.8, CH 115.4, CH 129.9, CH 112.0, CH
127.6, CH 134.4, CH 133.9, CH 134.6, CH 134.4, CH 136.9, qC
130.1, CH 126.7, CH 131.7, CH 131.4, CH 126.7, CH 120.6, CH
135.6, CH 131.3, CH 127.4, CH 133.5, CH 131.3, CH 131.8, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
57.4, CH
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
48.9, CH
82.2, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
52.8, CH
35.4, qC
20
20
30
30
40
40
50
50
60
60
70
70
80
e
e
124.1, qC 55.6, CH3
e
e
80
124.1, qC
e
e
e
124.1, qC 124.1, qC
a
a
In CDCl3, J in Hz. 100 MHz for 13C NMR. Carbon multiplicities were determined
In CDCl3, J in Hz. 100 MHz for 13C NMR. Carbon multiplicities were determined
by APT or PENDANT experiments, qC
CH2 ¼ methylene, CH3 ¼ methyl carbons.
¼
quaternary, CH
¼
methine,
by APT or PENDANT experiments, qC
CH2 ¼ methylene, CH3 ¼ methyl carbons.
¼
quaternary, CH
¼
methine,
MDA-MB-231 cell line is parallel to the one discussed above for the
antimigratory activity. However, there was an exception related to
design of new sipholenol A analogs in an attempt to improve the
antimigratory activity. The scheme starts with evaluating the
activity of unsubstituted and para-chloro substituted compounds 8
and 11, respectively. Since 11 was more active than 8, we proceeded
to the synthesis of 40,50-dichlorobenzoate analog 29, which proved
to be the most active analog and considered a potential hit
appropriate for further optimization.
the para-substitution at the aromatic ring with (þ
s/þp) sub-
stituents. The less hydrophobic and more electron withdrawing
substituents were associated with better antiproliferative activity as
follows: NO2 group (26) > Cl (11) > F (25) > CF3 (27) (Fig. 2).
Though aromatic esters maintained their higher activity profile
than aliphatic esters as previously discussed, a different pattern was
observed for the effect of C-30, C-40 and C-50 substituents (25e37) on
the antiproliferative activity in MCF-7 cells. Introduction of either
EDG or EWG in meta-(C-40) or para-(C-50) position resulted in a
marginal decrease in the activity, as observed in analogs 25e28 and
30e31. The 40,50-dichlorobenzoate ester, however, showed a two-fold
enhancement in activity as compared to the unsubstituted benzoate
ester 8. Insertion of EWG in C-30 significantlyenhanced the activity, as
can be observed for analogs 32e35. The less hydrophobic and more
electron withdrawing substituents were associated with better
antiproliferative activity in the following order: 33 > 34 > 32 > 35.
The most active of all analogs in this cell line was 33 (Table 8).
2.2.2. Antiproliferative activity against MCF-7 and the highly
metastatic MDA-MB-231 breast cancer cell lines
The MTT assay is sensitive in vitro assay that allows the mea-
surement of cell proliferation in a quantitative colorimetric fashion
by utilizing the ability of metabolically active (viable) cells to reduce
the MTT reagent to insoluble purple formazan crystals [22]. At least
four concentrations per compound were tested and used to calculate
an IC50 value (Table 8). The parent sipholenol A (1) and most of the
natural sipholanes showed moderate to high micromolar anti-
proliferative activity against both cell lines. In general, all sipholenol
A semisynthetic analogs showed antiproliferative activity better
than the parent 1 and most of natural sipholanes, indicating the
importance of introducing bulkier substitutions at the C-4 secondary
alcohol group. Fig. 2 shows the antiproliferative activity of the most
active analogs 25, 26, 29 and 39 compared to the DMSO control. The
antiproliferative structureeactivity relationship (SAR) observed for
2.2.3. Cytotoxic activity against the non-tumorigenic MCF10A
epithelial cell line
The cytotoxicity of sipholanes was assessed in the MTT assay
using the non-tumorigenic human breast cell line MCF10A (Fig. 5).

314
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
Table 3
Table 4
13C NMR spectroscopic data of compounds 37e39.a
1H NMR spectroscopic data of compounds 25e28.a
Position
dC
Position
dH, mult. (J in Hz)
37
38
39
25
26
27
28
1
2
3
4
5
7
8
9
42.7, qC
42.7, qC
35.1, CH2
23.2, CH2
80.1, CH
77.8, qC
77.3, CH
26.7, CH2
39.5, CH2
72.4, qC
56.5, CH
26.8, CH2
33.8, CH2
57.4, CH
143.2, qC
121.4, CH
24.8, CH2
48.9, CH
82.2, qC
37.2, CH2
24.9, CH2
52.8, CH
35.4, qC
13.2, CH3
29.1, CH3
21.8, CH3
30.1, CH3
30.1, CH3
25.7, CH3
31.7, CH3
29.5, CH3
158.0, qC
145.0, qC
117.6, CH
111.9, CH
146.6, CH
e
42.7, qC
35.1, CH2
23.2, CH2
79.7, CH
77.8, qC
77.3, CH
26.7, CH2
39.5, CH2
72.4, qC
56.5, CH
26.8, CH2
33.8, CH2
57.4, CH
143.0, qC
121.5, CH
24.8, CH2
48.9, CH
82.2, qC
37.2, CH2
24.9, CH2
52.8, CH
35.4, qC
13.2, CH3
29.1, CH3
21.8, CH3
30.1, CH3
30.1, CH3
25.7, CH3
31.7, CH3
29.5, CH3
156.2, qC
145.1, qC
118.9, CH
111.6, CH
152.0, qC
e
2
3
4
7
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
35.1, CH2
23.2, CH2
80.8, CH
77.8, qC
77.3, CH
26.7, CH2
39.5, CH2
72.4, qC
56.5, CH
26.8, CH2
33.8, CH2
57.4, CH
143.0, qC
121.4, CH
24.8, CH2
48.9, CH
82.2, qC
37.2, CH2
24.9, CH2
52.8, CH
35.4, qC
13.2, CH3
29.1, CH3
21.8, CH3
30.1, CH3
30.1, CH3
25.7, CH3
31.7, CH3
29.5, CH3
165.9, qC
135.0, qC
115.0, CH
157.5, qC
123.0, qC
126.6, CH
125.8, CH
123.4, qC
8
9
11
12
13
14
16
0.76, m
0.76, m
0.76, m
0.76, m
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
10
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.15, d (8.2)
7.35, d (8.2)
7.35, d (8.2)
8.15, d (8.2)
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.20, d (8.2)
8.33, d (8.2)
8.33, d (8.2)
8.20, d (8.2)
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.15, d (8.2)
7.74, d (8.2)
7.74, d (8.2)
8.15, d (8.2)
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.97, d (8.6)
7.44, d (8.6)
7.44, d (8.6)
7.97, d (8.6)
3.85, 3H, s
17
18
20
21
22
24
25
26
27
28
29
30
31
30
40
60
70
80
20
a
In CDCl3, J in Hz. 400 MHz for 1H NMR.
30
40
50
60
on Brk phosphorylation was only detected at a higher dose (20 mM,
Fig. 4). These findings indicate that 29 inhibits Brk phosphorylation
and its effect is superior to its parent 8.
70
e
e
80
e
e
a
In CDCl3, J in Hz. 100 MHz for 13C NMR. Carbon multiplicities were determined
by APT or PENDANT experiments, qC
CH2 ¼ methylene, CH3 ¼ methyl carbons.
¼
quaternary, CH
¼
methine,
2.3. Pharmacophore modeling and 3D-QSAR
2.3.1. Pharmacophore model
All active analogs were nontoxic at concentrations equal to their
IC50 values or higher in both WHA and proliferation assays, sug-
gesting their good selectivity towards malignant cells.
In order to rationalize chemical structure with the antimigratory
activity of sipholane analogs, a number of pharmacophore models
were generated using PHASE module of the Schrödinger molecular
modeling software [27]. Numerous common pharmacophore hy-
potheses (models) with different combination of variants (features)
were generated [28]. Each of these hypotheses contained a
maximum of 6 of the following features/sites: hydrogen bond
acceptor (A), hydrogen bond donor (D), hydrophobic group (H) and
aromatic ring (R). Hypotheses were evaluated on the basis of sur-
vival, survival-inactive and post-hoc scores [27,29]. Ten pharma-
cophoric hypotheses with three different variant combinations
(AAADHR, AAAHHR and AADHHR) were selected (Table 9). These
represent top scoring hypotheses showing the best alignment of
actives [27,30e32]. In general, a good hypothesis should provide
superior alignment with active compounds, discriminate between
active and inactive compounds and display the lowest possible
relative conformational energy values [27]. Fig. 6 shows the align-
ment of the most active compound (29) with the top scoring hy-
pothesis from each of the three variants. The generated hypotheses
successfully identified the most important structural features
implicated in the antimigratory activity of sipholanes: C-10 carbonyl
oxygen, aromatic ring and C-10 hydroxyl group. The absence of any
of these features compromised the antimigratory activity as pre-
viously discussed. For example, the C-4-O-benzoyl ester 8 is about
2.2.4. Brk phosphorylation inhibition
The effect of compound 29 treatment on Brk phosphorylation
was evaluated using cell-free Z0-LYTEÔ assay and cell-based
Western blotting analysis. Generally, the cell-free kinase assays
use fluorescence-based detection, which are widely used for HTS-
based kinase inhibitors discovery due to their automation
friendly, easy to use, relatively low cost, and wide availability [26].
The Z0-LYTEÔ Kinase Assay-Tyr1 peptide kit (Invitrogen) was
selected. In this experiment, Tyr1 peptide is used as a substrate;
thus, the changes of Z0-LYTEÔ Tyr1 peptide phosphorylation can
directly reflect the Brk kinase activity. The results showed that
analog 29 caused a concentration-dependent suppression of Brk
phosphorylation, with an IC50 value of 5.3
mM, compared to an IC50
value of 1.25 M for the positive control natural product staur-
m
osporine (Fig. 3).
In Western blot analysis, cells were exposed to different doses of
analogs
8 and 29. Results showed a relatively large, dose-
dependent inhibition of Brk phosphorylation after treatment with
29 for 72 h as compared to the vehicle-treated control group,
without affecting total Brk levels (Fig. 4). Moreover, the effect of 8

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
315
Table 5
Table 6
1H NMR spectroscopic data of compounds 29e32.a
1H NMR spectroscopic data of compounds 33e36.a
Position
dH, mult. (J in Hz)
Position
dH, mult. (J in Hz)
29
30
31
32
33
34
35
36
2
3
4
7
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
2
3
4
7
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m 1.38, 1.56, m
1.94, 2.12, m 1.94, 2.12, m
5.19, d (6.6)
3.55, dd
5.19, d (6.6)
3.55, dd
(11.3, 3.6)
(11.3, 3.6)
8
9
8
9
1.45, 1.82, m 1.45, 1.82, m
1.64, m
0.76, m
1.64, m
0.76, m
11
12
13
14
16
0.76, m
0.76, m
0.76, m
0.76, m
11
12
13
14
16
0.76, m
0.76, m
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.11, d (2.2)
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.91, s
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.91, s
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd (9.2, 5.1) 5.44, dd (9.2, 5.1) 5.44, dd
(9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.16, 1.52, m
1.66, 1.95, m
1.57, m
1.16, 1.52, m 1.16, 1.52, m
1.66, 1.95, m 1.66, 1.95, m
1.57, m
1.57, m
5.44, dd
(9.2, 5.1)
17
18
20
21
22
24
25
26
27
28
29
30
31
30
17
18
20
21
22
24
25
26
27
28
29
30
31
40
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.35, d (8.6)
7.95, dd
1.78, 1.98, m 1.78, 1.98, m
1.77, m
1.60, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.68, 1.87, m 1.68, 1.87, m
2.41, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.49, d (8.6)
7.89, dd
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.57, d (8.6)
7.61, dd
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.27, d (8.6)
e
40
50
e
e
7.57, d (8.6)
7.61, dd
(8.6, 8.6)
7.63, dd
(8.6, 8.6)
50
e
8.45, d (8.2)
8.45, d (8.2)
(8.6, 8.6)
7.92, dd
(8.6, 8.6)
8.47, d (8.6)
(8.6, 8.6)
7.33, dd
(8.6, 8.6)
8.15, d (8.6)
(8.6, 8.6)
60
7.63, dd
(8.6, 8.6)
8.20, d (8.6)
7.50, d (8.6)
7.98, d (8.6)
60
70
7.56, d (8.2)
7.71, dd
(8.2, 7.8)
8.41, d (7.8)
7.71, dd
(8.2, 7.8)
8.41, d (7.8)
70
7.87, dd
(8.2, 1.8)
8.20, d (8.6)
a
In CDCl3, J in Hz. 400 MHz for 1H NMR.
a
In CDCl3, J in Hz. 400 MHz for 1H NMR.
Table 7
1H NMR spectroscopic data of compounds 37e39.a
twice as active as both the C-4-O-phenyl ether 18 and the C-4-O-
acetate ester 5. Moreover, sipholenol A (1) is almost as active as its
anhydro analog 3 but is significantly more active than the dia-
nhydro analog 4, indicating the importance of the C-10 hydroxyl,
but not the C-19 hydroxyl, to the antimigratory activity of sipho-
lenol A and analogs.
Position
dH, mult. (J in Hz)
37
38
39
2
3
4
7
8
9
1.38, 1.56, m
1.94, 2.12, m
5.19, d (6.6)
3.55, dd (11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.14, d (6.8)
3.55, dd (11.3, 3.6)
1.45, 1.82, m
1.64, m
1.38, 1.56, m
1.94, 2.12, m
5.20, d (6.8)
3.55, dd (11.3, 3.6)
1.45, 1.82, m
1.64, m
2.3.2. 3D-QSAR
In order to statistically establish a relationship between the 3D-
spatial arrangement of the pharmacophoric features and the anti-
migratory activity of sipholanes, a 3D-QSAR model was generated.
Pharmacophore hypotheses (Table 9) were used to align com-
pounds for building atom-based 3D-QSAR models by partial least
square (PLS) analysis [27,30]. Analogs were divided into training
and test sets (see Experimental section 4.4 for details). The test set
was used to validate the generated models and was composed of
analogs with IC50 values covering the entire range of antimigratory
activity [31]. Several models containing up to five PLS factors were
generated. The best 3D-QSAR model generated was based on the
AADHHR.3334 pharmacophore hypothesis (Fig. 7). Table 10 shows
the details of the best 3D-QSAR model, which is a five-PLS factor
model with good statistics and predictive ability as reflected from r2
(0.910), q2 (0.853) and Pearson-R (0.943) values. These features
together with the large F-value (48.4) and the small variance ratio p
(9.229e-012) supported the significance of the selected model. This
provided a good correlation between predicted and actual anti-
migratory activity (Table 11). Scatter plots of actual versus predicted
activities showed that pIC50 values are effectively predicted for both
training and test set compounds (Fig. 8a and b, respectively). These
11
12
13
14
16
17
18
20
21
22
24
25
26
27
28
29
30
31
30
0.76, m
0.76, m
0.76, m
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd (9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
8.09, s
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd (9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.15, d (3.7)
6.53, dd (3.7, 1.8)
7.61, d (1.8)
e
1.16, 1.52, m
1.66, 1.95, m
1.57, m
5.44, dd (9.2, 5.1)
1.78, 1.98, m
1.77, m
1.60, m
1.68, 1.87, m
2.41, m
1.04, 3H, s
1.26, 3H, s
1.24, 3H, s
1.12, 3H, s
1.73, 3H, s
1.24, 3H, s
1.04, 3H, s
0.90, 3H, s
7.32, d (3.6)
7.35, d (3.6)
e
40
e
50
e
60
7.86, d (8.2)
8.15, d (8.2)
e
70
e
e
a
In CDCl3, J in Hz. 400 MHz for 1H NMR.

316
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
Table 8
mainly occupies blue regions (Fig. 9a), whereas the least active
analog 54 occupies none of the blue regions observed for 29
(Fig. 9b).
Antiproliferative and antimigratory activities of compounds 1e59 against human
breast cancer cell lines.
Compound
Antiproliferative
Antimigratory
activity (IC50, mM)
activity (IC50
,
m
M)
3. Conclusion
MCF-7
MDA-MB-231
MDA-MB-231
1
2
3
4
5
6
7
8
44.5
>50
>50
>50
47.9
39.2
>50
21.0
25.3
>50
24.1
43.5
>50
23.1
>50
>50
22.0
>50
>50
24.5
>50
46.8
25.6
>50
33.5
21.6
23.4
41.7
11.9
36.0
37.9
12.1
5.2
11.3
19.2
>50
>50
25.7
21.3
>50
>50
>50
31.2
>50
39.8
>50
>50
27.4
21.8
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
37.5
28.1
35.3
46.7
16.8
20.3
18.0
7.9
10.6
11.8
5.3
16.0
21.0
7.2
Substituted aromatic sipholenol A esters showed improved
antimigratory activity versus the previously reported unsubstituted
aromatic ester 8. The 40,50-dichlorobenzoate ester afforded the
most active analog 29. The cell-based in vitro activity of 29 was
well-correlated with its cell-free Brk phosphorylation inhibitory
activity coupled with the absence of any cytotoxicity to MCF10A
non-tumorigenic cells. Pharmacophore model and 3D-QSAR ana-
lyses characterized the key structural elements of sipholanes
required for efficient migration and Brk phosphorylation inhibition.
These studies provide the basis for future design of novel anti-
migratory entities based on a simplified sipholane structure, pos-
38.0
26.2
27.9
16.1
18.0
33.4
10.3
39.2
29.6
18.4
19.1
32.7
12.7
29.1
35.7
40.3
47.1
42.7
27.4
38.2
11.3
8.9
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
sessing rings
A and B (perhydrobenzoxepine) connected to
9.5
12.3
5.9
substituted aromatic esters with the elimination of rings C and D,
the [5,3,0]bicyclodecane system. This will make future synthesis of
the new active entity more feasible and cost-effective. Results
demonstrate the potential of the marine natural products to inspire
the discovery of novel Brk inhibitory scaffolds appropriate for use to
control metastatic breast cancer.
12.2
13.4
13.9
18.0
>50
28.5
31.4
2.4
3.4
7.8
17.9
1.3
5.9
4. Materials and methods
19.5
30.6
8.1
4.1. General experimental procedures
34.7
36.3
24.3
38.7
28.7
31.5
36.3
22.2
14.0
11.1
IR spectra were recorded on a Varian 800 FT-IR spectropho-
tometer. TLC analysis was carried on precoated Si gel 60 F254
6.1
14.1
18.6
22.7
34.5
26.4
11.9
9.3
500
a developing system. For column chromatography, Si gel 60 (Nat-
land International Corporation, 230e400 m) was used, and
mm TLC plates (EMD Chemicals), using n-hexane-EtOAc (7:3) as
m
gradient n-hexaneeEtOAc solvent system was used as a mobile
phase. 1H and 13C NMR spectra were recorded in CDCl3, using tet-
ramethylsilane (TMS) as an internal standard, on a JEOL Eclipse-ECS
NMR spectrometer operating at 400 MHz for 1H NMR and 100 MHz
for 13C NMR. The ESIMS experiments were conducted with a 3200
Q-trap LC/MS/MS system (Applied Biosystems, Foster City, CA) us-
ing Analyst version 1.4.1 software (MDS Sciex; Toronto, Canada)
where analytes were ionized using electrospray ionization (ESI)
interface operated in the negative mode.
3.5
>50
28.0
33.0
38.3
39.0
42.0
45.0
47.5
48.0
>50
>50
>50
>50
>50
>50
>50
>50
>50
42.6
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
4.2. Preparation of aromatic ester analogs of sipholenol A
Esterification of 1. To a solution of 1 (0.052 mmol) in CH2Cl2
(6.0 mL), N,N-dimethylaminopyridine (2 equivalents) and aryl
chloride (2 equiv) were added and refluxed overnight (Scheme 1)
[11,20]. Reaction mixture was worked out by dilution with CH2Cl2,
followed by the addition of water (10.0 mL). The organic layer was
then collected and the aqueous layer further extracted with CH2Cl2.
Organic layers were then dried over anhydrous Na2SO4 and the
solvent was removed in vacuo. Residue was collected and purified
on Si gel 60 using gradient n-hexaneeEtOAc system to afford the
product.
results can be visualized using 3D-plots of crucial volume elements
occupied by ligands [27]. In such representations, blue cubes
represent characteristics of the ligand structure that have moderate
to strong positive effect on calculated activity (positive co-
efficients). By contrast, red cubes represent moderate to strong
negative effects on calculated activity (negative coefficients)
[27,33]. Fig. 9 shows the combined effects of all features on the
antimigratory activity of sipholanes. The most active analog 29
4.2.1. 4b
-O-50-Fluorobenzoylsipholenol A (25)
Compound 25 was prepared according to the abovementioned
procedure to afford 25 mg, white powder (86% yield); IR nmax
(CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282, 1114, 910 cmꢀ1
;
1H and 13C NMR, see Tables 1 and 4; ESI-MS m/z 597.40 [M ꢀ H]ꢀ
(calcd for C37H54FO5, 597.40).

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
317
Fig. 1. Dose-response antimigratory activity of 25e27, 29e31, and 39 against the highly metastatic MDA-MB-231 human breast cancer cells in wound-healing assay. Error bars
indicate the SEM of n ¼ 3/dose. DMSO was used as a vehicle control, while 4-S-ethylphenylmethylene hydantoin (S-Ethyl) was used as a positive control at 50 M.
m
4.2.2. 4
b
-O-50-Nitrobenzoylsipholenol A (26)
4.2.4. 4b
-O-50-Methoxybenzoylsipholenol A (28)
Compound 26 was prepared to afford 20 mg, amorphous solid
(45% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 1 and 4; ESI-MS
m/z 624.30 [M ꢀ H]ꢀ (calcd for C37H54NO7, 624.30).
Compound 28 was prepared to afford 31 mg, colorless oil (78%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 1 and 4; ESI-MS m/z
;
609.40 [M ꢀ H]ꢀ (calcd for C38H57O6, 609.40).
4.2.3. 4
b
-O-50-(Trifluoromethyl)benzoylsipholenol A (27)
4.2.5. 4b
-O-40,50-Dichlorobenzoylsipholenol A (29)
Compound 27 was prepared to afford 27 mg, white powder (66%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
Compound 29 was prepared to afford 28.5 mg, colorless oil (82%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1
;
1H and 13C NMR, see Tables 1 and 4; ESI-MS m/z
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 1 and 5; ESI-MS m/z
;
647.40 [M ꢀ H]ꢀ (calcd for C38H54F3O5, 647.40).
647.30 [M ꢀ H]ꢀ (calcd for C37H53Cl2O5, 647.30).
Fig. 2. Effect of sipholenol A analogs 25, 26, 29, and 39 on the viability of the highly metastatic MDA-MB-231 cells. Error bars indicate the SEM of n ¼ 3/dose.

318
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 2 and 5; ESI-MS
m/z 613.30 [M ꢀ H]ꢀ (calcd for C37H54ClO5, 613.30).
4.2.9. 4b
-O-30-Nitrobenzoylsipholenol A (33)
Compound 33 was prepared to afford 15 mg, white powder (38%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 2 and 6; ESI-MS m/z
;
624.30 [M ꢀ H]ꢀ (calcd for C37H54NO7, 624.30).
4.2.10. 4b
-O-30-Fluorobenzoylsipholenol A (34)
Compound 34 was prepared to afford 20 mg, white powder (65%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 2 and 6; ESI-MS m/z
;
597.40 [M ꢀ H]ꢀ (calcd for C37H54FO5, 597.40).
4.2.11. 4b
-O-30-(Trifluoromethyl)benzoylsipholenol A (35)
Fig. 3. PTK6 phosphorylation inhibition of the most active analog 29 at various con-
centrations. Results were obtained using a Z0-LYTE assay kit; error bars indicate the
SEM of n ¼ 3/dose; staurosporine was used as a positive control.
Compound 35 was prepared to afford 19 mg, white powder (62%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 2 and 6; ESI-MS m/z
;
647.40 [M ꢀ H]ꢀ (calcd for C38H54F3O5, 647.40).
4.2.6. 4b
-O-40-Nitrobenzoylsipholenol A (30)
Compound 30 was prepared to afford 22 mg, colorless oil (55%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
4.2.12. 4b
-O-30-Fluoro-50-(trifluoromethyl)benzoylsipholenol A
(36)
Compound 36 was prepared to afford 24.6 mg, white powder
(56% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 2 and 6; ESI-MS
m/z 665.30 [M ꢀ H]ꢀ (calcd for C38H53F4O5, 665.30).
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 1 and 5; ESI-MS m/z
;
624.30 [M ꢀ H]ꢀ (calcd for C37H54NO7, 624.30).
4.2.7. 4b
-O-40-(Trifluoromethyl)benzoylsipholenol A (31)
Compound 31 was prepared to afford 26 mg, white powder
(72.5% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 2 and 5; ESI-MS
m/z 647.40 [M ꢀ H]ꢀ (calcd for C38H54F3O5, 647.40).
4.2.13. 4b
-O-40-Fluoro-50-(trifluoromethyl)benzoylsipholenol A
(37)
Compound 37 was prepared to afford 28 mg, amorphous solid
(76% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 3 and 7; ESI-MS
m/z 665.30 [M ꢀ H]ꢀ (calcd for C38H53F4O5, 665.30).
4.2.8. 4b
-O-30-Chlorobenzoylsipholenol A (32)
Compound 32 was prepared to afford 21 mg, amorphous solid
(67% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
Fig. 4. Western blot and densitometric analysis of Brk and p-Brk after exposure of MDA-MD-231 cells to 1, 10, and 20 mM treatments of analogs 8 and 29 for 72 h. b-Tubulin was used
as a loading control.

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
319
4.2.14. 4b
-O-20-Furoylsipholenol A (38)
Table 9
Various pharmacophore hypotheses generated by PHASE.
Compound 38 was prepared to afford 30.5 mg, white powder
(80% yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713,
1282, 1114, 910 cmꢀ1; 1H and 13C NMR, see Tables 3 and 7; ESI-MS
m/z 569.30 [M ꢀ H]ꢀ (calcd for C35H53O6, 569.30).
Hypothesis
Survival Survival- Post-hoc
#
Energyb Activityc
score
inactive
score
score
Matchesa
AAADHR.348
AAADHR.389
3.500
3.497
2.181
1.992
2.095
1.907
2.328
2.214
1.991
2.437
2.337
2.129
6.430
6.428
6.402
6.401
6.413
6.411
6.405
6.101
6.094
6.090
19
19
19
19
19
19
19
19
19
19
0.012
0.012
0.012
0.012
0.012
0.012
0.012
0.313
0.313
0.313
5.886
5.886
5.886
5.886
5.886
5.886
5.886
5.886
5.886
5.886
4.2.15. 4b
-O-50-Nitro-20-furoylsipholenol A (39)
AAAHHR.7687 3.472
AAAHHR.5876 3.470
AADHHR.3334 3.483
AADHHR.7559 3.481
AADHHR.3400 3.475
AADHHR.308
AADHHR.290
AADHHR.305
Compound 39 was prepared to afford 27 mg, white powder (70%
yield); IR nmax (CH2Cl2) 3454, 2987, 2950, 2930, 2858, 1713, 1282,
1114, 910 cmꢀ1 1H and 13C NMR, see Tables 3 and 7; ESI-MS m/z
;
614.30 [M ꢀ H]ꢀ (calcd for C35H52NO8, 614.30).
3.471
3.464
3.460
4.3. In vitro assays
a
The number of active ligands that matched the hypothesis.
Breast cancer cell lines, MCF-7 and MDA-MB-231, were pur-
chased from ATCC (Manassas, VA). Both cell lines were grown in
RPMI 1640 medium (GIBCO-Invitrogen, NY) supplemented with
b
c
Energy of the reference conformer relative to the lowest-energy conformer.
Reference ligand activity expressed as pIC50
.
Fig. 5. Cytotoxic activity of 25e39 against the non-tumorigenic human mammary epithelial cells MCF10A. Error bars indicate the SD of n ¼ 3/dose.
Fig. 6. Most active compound 29 aligned with the top scoring hypothesis of variant (a) AAADHR; (b) AADHHR; and (c) AAAHHR. Red spheres represent hydrogen bond acceptors
(A), green spheres represent hydrophobic groups (H), cyan spheres represent hydrogen bond donors (D) and brown ring represents aromatic (R) groups. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)

320
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
Fig. 7. Alignment based on the pharmacophore hypothesis AADHHR.3334 generated the best 3D-QSAR model. (a) Common pharmacophoric features identified by this hypothesis.
ꢀ
All distances are in A unit. Alignment of some of the most active (b) and least active (c) sipholane A analogs to this pharmacophore. The most active analog 29 is shown in cyan. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
10% fetal bovine serum (FBS) and glutamine (2 mmol/L), penicillin
G (100 U/mL), and streptomycin (100
g/mL) and incubated at 37 ꢁC
in a humidified incubator under 5% CO2.
4.3.2. MTT (proliferation assay)
m
The antiproliferative effect in this study was tested on the
human breast cancer cell lines, MCF-7 and MDA-MB-231 following
the procedure described previously [22]. Briefly, cells in expo-
nential growth were plated in a 96-well plate at a density of
10 ꢂ 103 cells per well, and allowed to attach overnight at 37 ꢁC
under 5% CO2 in a humidified incubator. Complete growth me-
A stock solution of each compound was prepared in DMSO at
a concentration of 25 mM for all assays. Appropriate media
(serum-free, 0.5% FBS or 5% FBS) was used to prepare com-
pounds at their final concentrations for each assay. The vehicle
(DMSO) control was prepared by adding the maximum volume
of DMSO used in preparing test compounds to the appropriate
media type such that the final DMSO concentration never
exceeded 0.2%.
dium was then replaced with 100 mL of either RPMI serum-free
medium (GIBCO-Invitrogen, NY) for MCF-7 cells or RPMI media
supplemented with 5% FBS for MDA-MB-231 cells, containing
various doses of the specific test compound and incubation
resumed at 37 ꢁC under 5% CO2 for 72 h. Control and treatment
media were then removed, replaced with fresh media, and 50 mL
4.3.1. Wound-healing assay (WHA)
MTT solution (at 1 mg mLꢀ1) were added to each well and plates
were re-incubated for 4 h. At the end of the incubation period, the
color reaction was stopped by removing the media and adding
The WHA is a simple method for evaluating directional cell
migration in vitro [23,24]. MDA-MB-231 cells were plated in sterile
24-well plates and allowed to form a confluent cell monolayer per
well (>95% confluence) overnight. Wounds were then inflicted in
100
at 37 ꢁC was resumed for up to 20 min to ensure complete
dissolution of crystals. Absorbance was determined at 570 nm
mL DMSO to dissolve the formazan crystals formed. Incubation
each cell monolayer using a sterile 200 mL pipette tip. Media was
l
removed and cells were washed twice with PBS and once with fresh
RPMI media. Test compounds at the desired concentrations were
prepared in fresh media (0.5% FBS) and were added to wells in
triplicate. The incubation was carried out for 24 h, after which
media was removed and cells were washed, fixed and stained using
Diff-QuickÔ staining (Dade Behring Diagnostics, Aguada, Puerto
Rico). Cells which migrated across the inflicted wound were
counted under the microscope in at least five randomly selected
fields (magnification: 400ꢂ).
using an ELISA plate reader (BioTek, VT, USA). The number of cells
per well was calculated against a standard curve prepared at the
start of each experiment by plating various numbers of cells (in
the range 1000e60,000 cells per well), as counted by a hemocy-
tometer. The IC50 value for each compound was calculated by
nonlinear regression (curve-fit) of log (concentration) versus the
number of cells, implemented in GraphPad Prism version 5.0
(GraphPad Software, La Jolla, CA, USA) [34].
4.3.3. MTT (cytotoxicity assay)
The cytotoxic effect of the sipholenol A and its analogs was
tested in culture on the normal human non-tumorigenic mammary
epithelial cell line MCF10A. MCF10A (ATCC cat # CRL-10317) cells
were maintained in defined medium consisting of Dulbecco’s
modified Eagle’s medium (DMEM)/F12 containing 5% horse serum,
Table 10
Statistical parameters of the best atom-based 3D-QSAR model generated by PHASE.
Hypothesis
PLSa SDb r2c
Fd
Pe
RMSEf q2g
Pearson-Rh
1% penicillin/streptomycin, 0.5
mg/mL hydrocortisone, 100 ng/mL
AADHHR.3334
1
2
3
4
5
0.302 0.650 51.9 7.711e-008 0.194 0.664 0.847
0.253 0.763 43.3 3.673e-009 0.135 0.836 0.953
0.237 0.799 34.5 3.264e-009 0.151 0.795 0.940
0.211 0.847 34.7 7.219e-010 0.134 0.840 0.938
0.166 0.910 48.4 9.229e-012 0.128 0.853 0.943
cholera toxin, 10 g/mL insulin, and 20 ng/mL epidermal growth
m
factor (rhEGF). For subculturing, cells were rinsed twice with sterile
Ca2þ and Mg2þ free phosphate buffered saline (PBS) and incubated
in 0.25% trypsin containing 0.025% EDTA in PBS for 10 min at 37 ꢁC.
The released cells were centrifuged, resuspended in fresh media
and counted using hemocytometer. Cells in exponential growth
were plated in a 96-well plate at a density of 20 ꢂ 103 cells/well,
and allowed to attach for 24 h at 37 ꢁC under 5% CO2. Complete
a
Number of factors in the partial least squares (PLS) regression model.
Standard deviation (SD) of the regression.
b
c
Value of r2 for the regression.
d
Variance ratio. Large values of
F
indicate
a more statistically significant
regression.
e
Significance level of variance ratio. Smaller values indicate a greater degree of
growth medium was then replaced with 100 mL of DMEM serum-
confidence.
free medium (GIBCO-Invitrogen, NY) containing various doses of
the specific test compound and incubation resumed at 37 ꢁC under
5% CO2 for 24 h. The cells were then treated with MTT solution
f
g
h
Root-mean-square error of the test set.
Value of q2 for the predicted activity of the test set.
Value of Pearson-R for the predicted activities of the test set.

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
321
Table 11
Experimental and predicted activity data for training and test set compounds ob-
tained from the best generated atom-based 3D-QSAR model (PLS factors ¼ 5).
Compound
QSAR set
Experimental
activity (pIC50
Predicted activity
(pIC50)
Pharm set
)
1
2
3
4
6
7
8
9
Training
Training
Training
Training
Training
Test
Training
Training
Test
4.426
4.551
4.452
4.331
4.693
4.745
5.102
4.975
4.928
5.276
4.796
5.143
5.022
4.910
4.914
4.873
4.857
4.278
4.545
4.503
5.620
5.475
5.108
4.747
5.886
5.229
5.226
5.215
4.730
4.644
4.462
4.578
4.924
5.032
5.458
4.553
4.481
4.417
4.409
4.377
4.347
4.323
4.319
4.268
4.229
4.172
4.109
4.071
4.058
4.032
4.650
4.441
4.371
4.291
4.161
4.257
4.556
4.325
4.454
4.766
4.779
5.086
4.996
5.112
5.197
4.683
4.977
4.953
4.915
4.963
4.768
4.919
4.234
4.554
4.548
5.213
5.564
5.033
4.858
5.445
5.157
5.067
5.082
4.805
4.891
4.456
4.801
5.136
5.147
5.623
4.246
4.395
4.192
4.348
4.479
4.504
4.379
4.256
4.168
4.119
4.131
4.083
4.013
4.001
4.018
4.371
4.339
4.182
4.366
4.139
Inactive
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Inactive
10
11
12
14
15
16
18
19
20
22
23
24
25
26
27
28
29
30
30
31
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Test
Training
Training
Training
Training
Training
Training
Test
Training
Test
Training
Training
Training
Test
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Training
Test
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Active
Fig. 8. Scatter plots for the predicted and experimental pIC50 values as calculated by
the 3D-QSAR model for the (a) training set, and (b) test set compounds.
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Active
buffer, 150
mL PTK6, and test compound (inhibitor). After 1 h of incubation at
RT, 5 L development solution containing site-specific protease
was added to each well. Incubation was continued for 1 h. The
reaction was then stopped, and the fluorescent signal ratio of
445 nm (coumarin)/520 nm (fluorescein) was determined on an
FLx800 plate reader (BioTek), which reflects the peptide substrate
cleavage status and/or the kinase inhibitory activity in the
reaction.
mm ATP, 2 m
m Z0-LYTE-Tyr 1 peptide substrate, 5000 ng/
m
Test
Test
Test
Test
Inactive
Inactive
Inactive
4.3.5. Western blot analysis
Western blot analysis was performed according to the method
described previously [35]. Briefly, MDA-MD-231 cells were
initially plated at a density of 1 ꢂ 106 cells/100 mm culture plate,
allowed to attach overnight in RPMI-1640 media containing 10%
FBS. Cells were then washed with PBS and incubated with vehicle
control or treatment in serum-free media for 3 days in culture.
At the end of treatment period, cells were lysed in RIPA buffer
(Qiagen Sciences Inc., Valencia, CA). Protein concentration was-
determined by the BCA assay (Bio-Rad Laboratories, Hercu-
les,CA). Equivalent amounts of protein were electrophoresed on
SDSepolyacrylamide gels. The gels were then electroblotted onto
PVDF membranes. These PVDF membranes were then blocked
with 2% BSA in 10 mM TriseHCl containing 50 mM NaCl and 0.1%
Tween 20, pH 7.4 (TBST) and then, incubated with specific pri-
mary antibodies against Brk (Abnova, CA) and p-Brk (Santa Cruz,
(50
stopped by the addition of solubilization/stop solution (DMSO)
(100
L/well), and incubation at 37 ꢁC was continued to ensure
complete dissolution of the formazan product. Absorbance of the
mL/well) and re-incubated for 4 h. The color reaction was
m
samples was determined at
(BioTek, VT, USA).
l 570 nm with an ELISA plate reader
4.3.4. Phosphorylation inhibition assay
A Z0-LYTE kinase assay-Tyr1 peptide kit (Invitrogen) was used
to assess the ability of sipholenol A and its analogs to inhibit
protein tyrosine kinase 6 (PTK6) phosphorylation. Briefly, 10
mL/
well reactions were set up in 384-well plates containing kinase

322
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
Fig. 9. 3D-QSAR visualization showing the combined effects of all features on the antimigratory activity for (a) the most active analog 29, and (b) the least active sipholane 54. Blue
cubes represent positive coefficients whereas red cubes represent negative ones. Analog 29 mostly occupies positive areas that are not occupied by 54. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)
CA) and incubated overnight at 4 ꢁC. At the end of incubation
period, membranes were washed 5 times with TBST and then
incubated with respective horseradish peroxide-conjugated anti-
rabbit or anti-mouse secondary antibodies (PerkinElmer Bio-
sciences, MA) in 2% BSA in TBST for 1-h at room temperature
followed by rinsing with TBST 5 times. Blots were then visualized
by chemiluminescence according to the manufacturer’s in-
structions (Pierce, Rockford, IL, USA). Images of protein bands
from all treatment groups within a given experiment were ac-
quired using Kodak Gel Logic 1500 Imaging System (Carestream
in at least 19 of the 23 active compounds with a final box size of
ꢀ
ꢀ
1 A and minimum intersite distance of 2 A. Resulting pharmaco-
phore hypotheses were then scored using default weights of
scoring parameters for both active (survival score) and inactive
ones (survival-inactive scores). To further assist in ranking hy-
potheses, post-hoc score was generated in which an activity
reward was added for hypotheses that utilize the most active
compound as the reference and a penalty assigned for hypotheses
in which the reference ligand shows a high relative conforma-
tional energy. Hypotheses were clustered and only those in which
actives showed good fitness scores and reasonable relative
conformational energies were selected and considered for 3D-
QSAR model generation. The quality of each pharmacophore is
measured in three ways based on the alignments to the input
structures: (1) the alignment score, which is the root-mean-
squared deviation (RMSD) in the site-point positions; (2) the
vector score, which is the average cosine of the angles formed by
corresponding pairs of vector features (acceptors, donors, and ar-
omatic rings) in the aligned structures; (3) a volume score which
measures how well each ligand overlays with the reference ligand
and is based on the overlap of van der Waals models of the non-
hydrogen atoms in each pair of structures.
Health Inc, New Haven, CT, USA). The visualization of b-tubulin
(Cell Signaling Technology, MA) was used to ensure equal sample
loading in each lane. All experiments were repeated at least three
times and
a representative Western blot image from each
experiment is shown in Fig. 4.
4.4. Computational details
4.4.1. Pharmacophore model
Pharmacophore modeling and 3D-QSAR generation were car-
ried out using PHASE (version 3.5, 2013) module of the Schrö-
dinger suite implemented in Maestro (Maestro 9.3.5, 2012)
molecular modeling package. PHASE identifies the spatial
arrangement of functional groups that are common and essential
for the biological activity of the compounds [27]. The structures of
compounds were built using the structure drawing tool and pre-
pared using LigPrep (version 2.5, 2012). Conformers were gener-
ated using ConfGen by applying OPLS-2005 force field method
with implicit distance-dependent dielectric solvent model at cut-
4.4.2. 3D-QSAR data setting
A total of 10 pharmacophore models were chosen, all of which
mapped well onto the most active compound (29). Molecules were
divided into training and test sets as described earlier. However,
actives within the training set with fitness scores <2.00 were
eliminated. Phase presents two options for aligning the 3D struc-
ture of compounds: atom-based alignment and pharmacophore-
based alignment [27]. In atom-based QSAR models, a molecule is
ꢀ
off root mean square deviation (RMSD) of 1 A. Conformers not
within 10 kJ/mol of the global minimum were automatically dis-
carded. A maximum of 250 conformers and 100 minimization
steps were set. Compounds were divided into a training set of 48
compounds and a test set of 11 compounds using the leave-n-out
method. The test set was used for the validation of the generated
models. An arbitrary activity threshold value was then assigned to
treated as
a set of overlapping van der Waals spheres.
Pharmacophore-based models are more useful if the compounds
are highly flexible or if they exhibit significant chemical diversity.
ꢀ
Atom-based QSAR models were built using a grid spacing of 0.5 A
and 5 partial least-square (PLS) factors. It is recommended that the
number of PLS factors does not exceed 1/5 of the total number of
compounds in the training set [27].
divide training set compounds into 23 actives (IC50 < 25
mM), 10
intermediates (25 < IC50 < 40 M) and 13 inactives (IC50 > 40
m
mM).
Compounds 18 and 19 were excluded because they lack the
carbonyl group found in all other actives and their inclusion may
jeopardize the correct selection of features. Four features/sites
were considered in generating pharmacophore variants: hydrogen
bond acceptor (A), hydrogen bond donor (D), hydrophobic group
(H) and aromatic ring (R). The maximum number of sites was set
to 6 and minimum to 4. Common pharmacophores were searched
Acknowledgments
The Louisiana Board of Regents is acknowledged for supporting
the new MS facility (LEQSF(2013e14)-ENH-TR-26). The University
of Prince Salman Bin Abdulaziz, Saudi Arabia, is acknowledged for
the fellowship support of A.F.

A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
323
Appendix A
O
O
O
OH
O
O
H
HO
H
O
OH
O
HOO
H
HO
H
H
H
H
H
OH
H
H
OH
H
H
HO
OH
H
HO
O
H
HO
HO
O
40
41
42
43
OH
OH
OH
OH
O
O
O
O
O
H
H
H
H
O
O
O
H
H
OH
H
H
H
H
H
H
OH
H
H
OH
OH
H
H
HO
H
O
HO
H
HO
H
HO
O
OOH
44
45
46
47
OH
OH
OH
O
H
O
H
H
OOH
OH
H
H
H
OH
H
H
OH
H
H
OH
OH
HO
HO
HO
H
HO
HO
48
49
50
51
OH
OH
OH
OH
O
H
H
O
H
H
H
H
OH
H
H
H
H
OH
H
H
H
H
OH
OH
OH
HO
O
H
H
HO
HO
H
HO
H
HO
HO
O
OH
52
53
54
55
OH
O
OH
OH
O
O
O
H
H
H
H
H
O
H
OH
HO
H
H
H
H
H
H
OH
H
H
OH
H
HO
H
H
H
H
HO
HO
H
HO
O
O
O
O
56
57
58
59
[10] (a) S. Jain, S. Laphookhieo, Z. Shi, L. Fu, S. Akiyama, Z.S. Chen, D.T. Youssef,
R.W.M. van Soest, K.A. El Sayed, J. Nat. Prod. 70 (2007) 928e931;
(b) S. Jain, A. Shirode, S. Yacoub, A. Barbo, P.W. Sylvester, E. Huntimer,
F. Halaweish, K.A. El Sayed, Planta Med. 70 (2007) 591e596.
[11] A.I.Foudah, S. Jain, B.A.Busnena, K.A. ElSayed,ChemMedChem 8 (2013)497e510.
[12] American Cancer Society, http://www.cancer.org/Cancer/BreastCancer/
OverviewGuide/breast-cancer-overview-key-statistics (accessed September
2013).
Appendix B. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2013.11.039.
References
[13] N.S. Wind, I. Holen, Int. J. Breast Cancer 2011 (2011), 12, article 967419, http://
dx.doi.org/10.4061/2011/967419.
[14] K.T. Barker, L.E. Jackson, Oncogene 15 (1997) 799e805.
[15] P.J. Mitchell, K.T. Barker, J.E. Martindale, T. Kamalati, P.N. Lowe, M.J. Page,
Oncogene 9 (1994) 2383e2390.
[1] T.L. Simmons, E. Andrianasolo, K. McPhail, P. Flatt, W.H. Gerwick, Mol. Cancer
Ther. 4 (2005) 333e342.
[2] R. Singh, M. Sharma, P. Joshi, D.S. Rawat, Anti-Cancer Agents Med. Chem. 8
(2008) 603e617.
[16] M. Aubele, G. Auer, A.K. Walch, A. Munro, M.J. Atkinson, H. Braselmann, Br. J.
Cancer 96 (2007) 801e807.
[3] D.J. Newman, G.M. Cragg, J. Nat. Prod. 75 (2012) 311e335.
[4] I. Ojima, J. Med. Chem. 51 (2008) 2587e2588.
[17] P.J. Mitchell, K.T. Barker, J. Shipley, M.R. Crompton, Oncogene 15 (1997)
1497e1502.
[18] A.J. Harvey, C.J. Pennington, S. Porter, R.S. Burmi, D.R. Edwards, W. Court, Am.
J. Pathol. 175 (2009) 1226e1234.
[19] P.M. Brauer, A.L. Tyner, Biochim. Biophys. Acta 1806 (2010) 66e73.
[20] A. Miller, P.H. Solomon, Reactions of nucleophiles and bases, in: Writing Re-
action Mechanisms in Organic Chemistry, second ed., Academic Press, San
Diego, 2000, pp. 105e193.
[5] S. Carmely, Y. Kashman, J. Org. Chem. 48 (1983) 3517e3525.
[6] S. Carmely, Y. Kashman, J. Org. Chem. 51 (1986) 784e788.
[7] Y. Kashman, T. Yosief, S. Carmely, J. Nat. Prod. 64 (2001) 175e180.
[8] S. Jain, I. Abraham, P. Carvalho, Y.H. Kuang, D.T. Youssef, M.A. Avery, Z.S. Chen,
K.A. El Sayed, J. Nat. Prod. 72 (2009) 1291e1298.
[9] I. Abraham, S. Jain, C.P. Wu, Y. Kuanga, Z. Shia, X. Chen, M. Khanfar, L. Fu,
S.V. Ambudkar, K.A. El Sayed, Z.S. Chen, Biochem. Pharmacol. 80 (2010) 1497e
1506.

324
A.I. Foudah et al. / European Journal of Medicinal Chemistry 73 (2014) 310e324
[21] B.A. Busnena, A.I. Foudah, T. Melancon, K.A. El Sayed, Bioorg. Med. Chem. 21
(2013) 2117e2127.
[22] Trevigen, Inc., TACS MTT Assays: Cell Proliferation and Viability Assays, 2003,
pp. 2e7.
[23] L.G. Rodriguez, X. Wu, J.L. Guan, Wound-healing assay, Methods Mol. Biol. 294
(2005) 23e29.
[24] K.A. El Sayed, A.I. Foudah, A.M. Mayer, A.M. Crider, D. Song, MedChemComm 4
(2013) 1231e1238.
[25] L.A. Shaala, D.T. Youssef, M. Sulaiman, F.A. Behery, A.I. Foudah, K.A. El Sayed,
Mar. Drugs 10 (2012) 2492e2508.
[26] S.M. Rodems, B.D. Hamman, C. Lin, J. Zhao, S. Shah, D. Heidary, L. Makings,
J.H. Stack, B.A. Pollok, Assay Drug Dev. Technol. 1 (2002) 9e19.
[27] S.L. Dixon, A.M. Smondyrev, E.H. Knoll, S.N. Rao, D.E. Shaw, R.A. Friesner,
J. Comput. Aided Mol. Des. 20 (2006) 647e671.
[28] Phase 3.5 Quick Start Guide, Schrödinger, LLC, May 2013.
[29] A.A. Sallam, W.E. Houssen, C.R. Gissendanner, K.Y. Orabi, A.I. Foudah, K.A. El
Sayed, MedChemComm 4 (2013) 1360e1369.
[30] M.D. Hall, N.K. Salam, J.L. Hellawell, H.M. Fales, C.B. Kensler, J.A. Ludwig,
G. Szakáes, D.E. Hibbs, M. Gottesman, J. Med. Chem. 52 (2009) 3191e3204.
[31] M.K. Teli, G.K. Rajanikant, Org. Med. Chem. Lett. 2 (2012) 1e10.
[32] P. Kamaria, N. Kawathekar, Med. Chem. Res. 22 (2013) 1054e2523.
[33] S. Brogi, P. Papazafiri, V. Roussis, A. Tafi, Eur. J. Med. Chem. 67 (2013)
344e351.
[34] F.M.A. Bar, M.A. Khanfar, A.Y. Elnagar, H. Liu, A.M. Zaghloul, F.A. Badria,
P.W. Sylvester, K.F. Ahmad, K.P. Raisch, K.A. El Sayed, J. Nat. Prod. 72 (2009)
1643e1650.
[35] M.R. Akl, N.M. Ayoub, B.S. Abuasal, A. Kaddoumi, P.W. Sylvester, Fitoterapia 84
(2013) 347e359.