Design and synthesis of emodin derivatives as novel inhibitors of ATP-citrate lyase

Article abstract of DOI:10.1016/j.ejmech.2016.12.018

Aberrant cellular metabolism drives cancer proliferation and metastasis. ATP citrate lyase (ACL) plays a critical role in generating cytosolic acetyl CoA, a key building block for de novo fatty acid and cholesterol biosynthesis. ACL is overexpressed in ca

Full text of DOI:10.1016/j.ejmech.2016.12.018

European Journal of Medicinal Chemistry 126 (2017) 920e928
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Design and synthesis of emodin derivatives as novel inhibitors of ATP-
citrate lyase
,
Steffi K. Koerner a, Jun-ichi Hanai b c, Sha Bai a, Finith E. Jernigan a, Miwa Oki b,
,
c
,
d,
, *
Chieko Komaba b, Emi Shuto b, Vikas P. Sukhatme b
**, Lijun Sun a
a Center for Drug Discovery and Translational Research, Department of Surgery, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA
02215, USA
b Divisions of Interdisciplinary Medicine and Biotechnology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School,
Boston, MA 02215, USA
c Nephrology and Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA
d Hematology-Oncology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Aberrant cellular metabolism drives cancer proliferation and metastasis. ATP citrate lyase (ACL) plays a
critical role in generating cytosolic acetyl CoA, a key building block for de novo fatty acid and cholesterol
biosynthesis. ACL is overexpressed in cancer cells, and siRNA knockdown of ACL limits cancer cell pro-
liferation and reduces cancer stemness. We characterized a new class of ACL inhibitors bearing the key
structural feature of the natural product emodin. Structure-activity relationship (SAR) study led to the
identification of 1d as a potent lead that demonstrated dose-dependent inhibition of proliferation and
cancer stemness of the A549 lung cancer cell line. Computational modeling indicates this class of in-
hibitors occupies an allosteric binding site and blocks the entrance of the substrate citrate to its binding
site.
Received 6 October 2016
Received in revised form
6 December 2016
Accepted 8 December 2016
Available online 9 December 2016
Keywords:
Emodin anthraquinones and anthracenone
ATP citrate lyase inhibitor
Lipogenesis
© 2016 Elsevier Masson SAS. All rights reserved.
Cancer stem cell
1. Introduction
oxidative phosphorylation to glycolysis, even under normal oxygen
concentrations [3]. This effect is regulated by PI3K, hypoxia-
Metabolic changes are a common feature of cancerous tissues.
Mutations in oncogenes and tumor suppressor genes cause alter-
ations to multiple intracellular signaling pathways that rewire tu-
mor cell metabolism and re-engineer it to allow enhanced survival
and growth. The extensive metabolic rewiring of malignant cells
offers a large number of potential drug targets [1]. A proper inter-
vention in a cancer metabolic pathway might provide a therapeutic
advantage that can help overcome resistance to chemotherapy or
radiotherapy [2].
inducible factor (HIF), p53, MYC and AMP-activated protein ki-
nase (AMPK)eliver kinase B1 (LKB1) pathways. Another metabolic
pathway often increased in cancer cells is lipid synthesis. Adeno-
sine triphosphate (ATP) citrate lyase (ACL) is the cytosolic enzyme
that catalyzes the synthesis of acetyl-CoA from citrate. Cytosolic
acetyl-CoA is the requisite building block for endogenous synthesis
of fatty acids, cholesterol and isoprenoids, as well as for post-
translational modification of proteins via acetylation. ACL is up-
stream of the other lipogenic enzymes and connects glucose
metabolism (a way of generating acetyl CoA) and lipogenesis [4].
ACL also affects mitochondrial homeostasis and membrane po-
tential through the increased availability of mitochondrial citrate
for full oxidation and NADH production in the TCA cycle [5].
In tumor cells, de novo fatty acid synthesis occurs at high rates
[6e8]. ACL was identified as a highly expressed protein in many
tumors, including the chemoresistant colorectal cancer cells via
unbiased proteomic profiling [9]. In agreement with reported re-
sults [4,10], our own data demonstrated that ACL knockdown (KD)
by shRNA limits tumor cell proliferation and survival, induces
The best characterized metabolic phenotype observed in tumor
cells is the Warburg effect, which is a shift of ATP generation from
* Corresponding author. Center for Drug Discovery and Translational Research,
Department of Surgery, Beth Israel Deaconess Medical Center, Harvard Medical
School, Boston, MA 02215, USA.
** Corresponding author. Divisions of Interdisciplinary Medicine and Biotech-
nology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard
Medical School, Boston, MA 02215, USA.
E-mail addresses: vsukhatm@bidmc.harvard.edu (V.P. Sukhatme), lsun1@bidmc.
harvard.edu (L. Sun).
http://dx.doi.org/10.1016/j.ejmech.2016.12.018
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
921
differentiation in vitro and in vivo, and reduces tumor growth in a
number of animal models [11]. We have recently shown that ACL
KD also targets cancer stem-like cells (CSCs) in multiple cancer
types, suggesting that ACL inhibitors might be effective in reducing
cancer stemness [12]. The presence of CSCs is believed to be a major
underlying reason for the emergence of resistance to standard
therapy and the occurrence of metastasis, major causes of treat-
ment failure and mortality. Collectively, available data strongly
support the notion that targeting ACL by small molecule inhibitors
is an attractive strategy for developing an innovative cancer ther-
apy to address the highly unmet needs in metastasis and drug
resistance of cancers [13e16].
Nature is a rich source for biologically active compounds,
including the widely used anticancer therapeutics paclitaxel and
doxorubicin. The natural product 2-chloro-1,3,8-trihydroxy-6-
methyl-10H-anthracen-9-one (2a), which is structurally related to
emodin (1a) isolated from the Chinese herbal plant Rheum pal-
matum L. [17], was reported to display inhibitory activity against
the ACL enzyme [18]. Herein we describe the structure-activity
relationship (SAR) study of emodin anthraquinones and anthrace-
none, the characterization of potent ACL inhibitors, and their
in vitro anticancer activities. Further, we show for the first time that
small molecule ACL inhibitors reduce cancer stemness in breast and
lung cancer 3D spheroid assays.
reliable method to regioselectively reduce the 10-carbonyl group of
emodin [28], the reduction resulted exclusively in the de-
chlorinated product 2b. When emodin anthracenone (2b) was
treated with sulfuryl chloride, the novel 5-chloriniated anthrace-
none (2c) was isolated as a pale yellow powder (Fig. 1). The
chemical structure of 2c, in particular the position of the Cl sub-
stituent, was confirmed unequivocally by 13C-HSQC and 13C-HMBC
NMR spectra indicating chlorination at the 4-position. But disap-
pointingly, the treatment of 2b with NCS and zirconium chloride
failed to yield the expected 2-chlorinated anthracenone (2a).
Instead, a dark complex mixture was formed under a variety of
conditions. Attempts to manipulate the reactions by protecting the
free hydroxyl groups, including the known tri-methylated deriva-
tive 4 (Fig. 1), all failed to generate the desired product 2a. In our
hands, 2b and 2c were stable as dry solids. In DMSO, however, they
were only stable when kept at ꢀ20 ꢁC and in the dark, and
decomposed to a brown mixture after standing overnight at room
temperature. It is known that the natural product emodin anthrone
2a undergoes free radical mediated oligomerization to form poly-
aromatics [29,30]. We hypothesize the presence of an electron-
withdrawing Cl group in 2c (as well as in 2a) increases the pro-
duction of free radicals and decreases the stability of the
compounds.
2.2. ACL enzymatic activity
2. Results and discussion
Historically, the in vitro enzymatic activity of ACL, as well as its
inhibition by inhibitors, was measured by the maleate dehydro-
genase catalyzed reduction of oxaloacetate by NADH [31,32]. The
method lacks high sensitivity, is laborious and inadequate for high-
throughput screening. We instead applied the ADP Glo assay to
measure directly the concentration of the enzymatic reaction
product ADP and obtained high assay stability (CV value < 10%, S/B
ratio >25, Z’-factor >0.7). Under our assay conditions, the IC50 of the
2.1. Chemistry
The commercially available natural product emodin (1a) pro-
vides a unique template for the synthesis of structurally diverse
derivatives. Our synthetic efforts were first directed at the synthesis
of halogenated emodins (Fig. 1, Table 1). Mono-chlorinated emodin
1b and 1c are naturally occurring compounds that were extracted
from lichen species and also obtained by biosynthesis utilizing
isolated lichen enzymes [19]. Their regioselective syntheses how-
ever remains unexplored. Chen and Huang described that treat-
ment of emodin 1a with acetic acid and hydrochloric acid followed
by H2O2 yielded the product 1c together with two dichloroemodin
derivatives in an equalmolar ratio [20]. A nonselective multistep
synthesis of 1b has been reported by both Yosioka et al. [19] and
Sargent et al. [21]. We were able for the first time to regioselectively
synthesize 1b and 1c in excellent yield. By applying the procedure
developed by Elban and Hecht [22], we prepared 1b in up to 85%
isolated yield through treatment of 1a with N-chlorosuccinimide
(NCS) and zirconium chloride. The method was also applicable to
the regioselective synthesis of compound 4 from the trimethylated
emodin [23]. Conversely, the regioisomeric 1c was formed exclu-
sively and isolated in 80% yield by treating 1a with sulfuryl chloride.
1H NMR spectra supported the structural assignment, which was
also found to be in agreement with literature values [24]. Regio-
selective iodonization or bromination of 1a produced the 3-
iodoemodin 1e and the 3,5-bisbromoemodin (1d), respectively,
by following literature reported methods [25]. Subsequently, 1d
and 1e underwent Suzuki coupling reactions [26] with a variety of
aryl or heteroaryl boronic reagents to produce a series of aryl
substituted derivatives (1f-o). Finally, using reported literature
procedures [27], we prepared a series of Mannich reaction products
(1p-s and 3a-b) from emodin (1a), which were not explored further
due to their extremely low solubility.
known ACL inhibitor BMS-303141 ranged from 0.3 to 0.5
is within the range of reported values of 0.94
M by a14C radio-
isotope assay and 0.13 M by the traditional maleate dehydroge-
mM, which
m
m
nase catalyzed reduction assay [33,34].
Table 1 summarizes the inhibitory activity of the newly syn-
thesized emodin derivatives. Representative dose-response curves
of ACL inhibition by lead compounds are shown in Fig. 2. As noted,
at highest concentrations tested, the inhibitors were able to
completely abolish the ACL enzymatic activity.
In agreement with previous literature, emodin (1a) and emodin
anthrone (2b) are completely inactive at concentrations as high as
10 m
M.ref.18 The halogenated emodin derivatives 1b-e are among
the more potent inhibitors in our series. The two mono-chlorinated
emodins 1b and 1c are similarly active, with an IC50 of 12.6 and
9.6
mM, respectively. The iodo derivative 1e (IC50: 7.5
mM) is slightly
more active than its chloro counterpart (1b). With an IC50 of 2.9
mM,
the bisbromo derivative 1d is the most potent in this series. Com-
pound 2c, the regioisomer of the reported natural product emodin
anthracenone 2a, potently inhibits ACL activity (IC50: 3.8 mM), a 2.5-
fold increase in potency from its anthraquinone analog 1c.
The 2-aryl and 2-heteroaryl derivatives (1f-o) proved to be
moderately active at best, with the pyrazole compound 1o (IC50
:
13.3 M) being the most active in this series. While the phenyl
m
group in 1f does not render it an active inhibitor, the addition of a
methoxy (1i, 1j) at the 3- or 4-position, or a methyl (1g) at the 3-
position but not at the 4-position (1h), of the phenyl group result
in moderately active inhibitors. In addition, the mono- and bis- 3-
pyridyl derivatives (1l, 1n) are also moderately active, while the
4-pyridyl isomer (1m) is not. The Mannich reaction products 1p-r
are inactive but their conformationally restrained tetracyclic ana-
logs 3a-b are moderately active, which may suggest that the 2-
While substantial effort was devoted to the synthesis of the
anthracenone natural product 2a, we were not able to fully char-
acterize the product, which we believe was primarily due to its
instability as indicated in the literature [18]. Thus, when 1b was
subjected to tin (II) chloride mediated reduction condition - a

922
S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
Table 1
position of emodin occupies a tight hydrophobic pocket on the ACL
protein, that favors rigid substituents over bulky and flexible alkyl
groups. Furthermore, it is worth noting that the moderate activity
by 3a-b indicates, that the 3-OH group of emodin is not absolutely
required for the inhibition of ACL although it likely contributes to
the overall potencies. On the other hand, one or perhaps both of the
two OH groups surrounding the carbonyl group is critical for ac-
tivity as the tri-methylated analog 4 is completely inactive (0% in-
The SAR of emodin derivatives in inhibition of ACL enzymatic activity.
OH
O
OH
OH
O
OH
OH
O
OH
1
1
R
1
R
R
N
O
HO
HO
2
2
2
O
O
R
R
R
Entry
Code
R1
R2
IC50 (mM)
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1f
1g
1h
1i
H
Cl
H
Br
H
H
Cl
Br
H
H
H
H
H
H
H
H
H
0%a
12.6
9.6
2.9
7.5
hibition at 10 mM).
2.3. Docking studies
I
Phenyl
8%a
30.0
0%a
21.2
18.8
0%a
21.1
0%a
23.5
13.3
Sun et al. recently reported the co-crystal structure of a trun-
cated ACL in complex with citrate, which revealed a network of H-
bonds between the substrate and hydrophilic residues including
Arg-379, Asn-346 and Thr-348 [35]. The natural product anthra-
cenone 2a has been previously described as a citrate competitive
ACL inhibitor [18]. To investigate potential key interactions be-
tween 2a and ACL protein, we performed auto docking studies to
analyze low energy binding poses of 2a in the ACL crystal structure
(PDB: 3MWD). The top scoring binding pose (calculated binding
energy: E ¼ ꢀ8.6 kcal/mol) was identified and found to occupy an
allosteric hydrophobic cleft directly adjacent to the substrate
binding domain of ACL protein (Fig. 3a). The identified low energy
binding pose of 2a may have blocked the entrance of citrate to the
binding site (indicated by a red arrow), which we theorize is
responsible for its citrate competitive inhibition of ACL. Key in-
teractions of 2a with ACL are well defined in this binding orienta-
tion, and consist of pi-pi interactions with Phe-347, and H-bonding
interactions of the 9-carbonyl and 8-OH functional groups with
Asp-346 and Gly-664 (Fig. 3b). The 2-chloro, its adjacent 3-OH
3-Methylphenyl
4-Methylphenyl
3-Methoxyphenyl
4-Methoxyphenyl
3,5-Dimethoxyphenyl
3-Pyridyl
4-Pyridyl
3-Pyridyl
1-Methyl-4-pyrazolyl
4-Morpholinylmethyl
1-Piperidinylmethyl
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1j
1k
1l
1m
1n
1o
1p
1q
1r
2a
2b
2c
3a
3b
3-pyridyl
H
H
10%a
6%a
N-Methyl-N-phenylaminomethyl
6%a
Cl
H
H
H
H
Cl
H
H
0.283b
0%a
3.8
Phenyl
25.8
29.7
0.442
5-(1,3-Benzoxazolyl)
BMS-303141
a
% inhibition at 10 M.
m
b
as reported in Ref. [18], and measured by the maleate dehydrogenase catalyzed
reduction of oxaloacetate.
Fig. 1. Synthetic schemes for the preparation of emodin derivatives. Regioselective synthesis of the mono-chlorinated emodin 1b (a) and 1c (b). Synthesis of the bis-brominated
emodin 1d (c) and its Suzuki coupling reaction for compound 1n (d). Synthesis of the mono-iodinated emodin 1e (e) and its Suzuki coupling for compounds 1f-m, 1o (f). Mannich
reaction of emodin for the synthesis of 1o-s and 3a-b (g). Regioselective reduction of emodin for anthracenone 2b (h) and its regioselective chlorination for 2c (i). Synthesis of
compound 4 (j). DCM: dichloromethane; MeCN: acetonitrile; NBS: N-bromosuccinimide; NCS: N-chlorosuccinimide; THF: tetrahydrofuran.

S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
923
Fig. 2. Representative dose-response curves of ACL inhibition in the ADP-Glo enzymatic assay. Inhibitors were tested in duplicates at a dose range of 0.03e30
mM with 3.3-fold
serial dilutions.
Fig. 3. (a) Space filling view of the lowest energy binding pose of 2a. The anthracenone was found to occupy space directly adjacent to the substrate binding domain (red arrow).
Red and blue color indicates negatively or positively charged protein surface, respectively. (b) Key interactions found of 2a with the ACL protein. The anthracenone 2a was found to
participate in pi-pi interactions with Phe-347 and H-bonding interactions with Asp-346 and Gly-664 (indicated by green dash lines). (c) The lowest energy docking pose of 1d
indicates near identical interactions to that of 2a. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
functional groups, and the methylene linker at the 10-position of 2a
were not found to be involved in interactions with the ACL protein
and a significant volume of space surrounding each functional
group is unoccupied. This result corroborates our SAR results
demonstrating that variety of anthraquinone analogs with sub-
stituents at the 2e4 positions retain ACL inhibitory activity. Further,
the lowest energy docking pose of 1d (Fig. 3c) indicates that 1d
essentially forms the same interactions the as that for 2a, with the

924
S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
additional C-10 carbonyl group of 1d extending to the solvents and
lacking interactions with nearby amino acid residues.
stem cells and currently under clinical investigations, and the ACL
inhibitor BMS-303141 were included as positive controls. In this
assay, effective agents for cancer stemness are expected to not only
reduce the spheroid size (area size) but also induce apoptotic cells
as detected by Sytox green. The intensity of Sytox Green correlated
to the level of cellular damages, which was calculated as Intensity
Density (IntDen). IntDen/Area was used as a readout index for an
inhibitor's effectiveness in modulating cancer cell stemness. In both
the A549 lung cancer cell line and the E-snail breast cell line, 1d
dose-dependently reduced the spheroid size and caused dramatic
apoptosis in spheroid, indicating its strong inhibitory function for
cancer stemness (Fig. 5A, B). Similarly but with less efficiency in the
E-snail breast cancer cell line, the ACL inhibitor BMS-303141 also
potently inhibited stemness in the cell based 3D assay.
2.4. Anti-proliferative activity
We tested the anticancer activity of selected ACL inhibitors in
the A549 non-small cell lung cancer (NSCLC) cell line, which we
had shown previously to be sensitive to ACL knockdown (KD) [11].
A549 cell harbors mutations in the KRAS oncogene and does not
respond to molecular therapies that target the epithelial growth
factor receptor (EGFR). We first treated A549 cells with 5 mM of the
active ACL inhibitors 1b-d, 1o, and 2c for 24 h and then measured
by WST-1 reagent for cell proliferation. ACL KD was used as a
positive control and to benchmark the activity of the inhibitors. At
5
mM concentration, our ACL inhibitors inhibited the A549 cell
proliferation by 40e80% vs. 100% denoted to ACL KD (Fig. 4a).
Furthermore, inhibitors 1d and 1o dose-dependently inhibited
3. Conclusions
A549 cell proliferation (Fig. 4bec). At 10
the same level of anticancer activity as achieved by ACL KD (100%),
implying that at 10 M 1d full inhibits the cellular ACL activity in
A549 cells, which correlates to its complete inhibition of ACL
enzymatic activity in the enzymatic assay at similar dose levels
m
M, 1d was able to reach
In summary, we identified a new series of ACL inhibitors based
on the chemical scaffold of the natural product emodin. SAR and
docking analyses indicate that the two OH groups adjacent to the 9-
carbonyl group may be critical for on-target activity. In addition, the
auto docking studies indicated the formation of key interactions
with N346 and G664 in an allosteric site adjacent to the citrate
binding domain. Chemical modifications of the 2e5 positions of
m
(>99% inhibition at 3.3 and 10
It should be noted that emodin, which did not show any inhi-
bition of ACL enzymatic activity at 10 M (Table 1), is known to
mM).
m
emodin lead to ACL inhibitors with IC50 ranging from 30 to 3 mM.
inhibit cancer cell proliferation via diverse mechanisms unrelated
to ACL activity [36e38]. Thus, He et al. reported that treatment of
the A549 lung cancer cell line with high concentrations (40 and
Halogens at the 2- and 4-position of emodin increased activity most
significantly, while aryl substituents at the 2-position of emodin
resulted in moderately active ACL inhibitors. Lead compounds
dose-dependently inhibited the proliferation of the A549 lung
cancer cells, with the best compounds reaching to similar levels of
activity to that of ACL KD. Further, we demonstrated for the first
time that ACL inhibitors significantly reduced cancer stemness in
the 3D spheroid assay. Our data provide further support to the
approach of targeting ACL for treating cancers and will guide the
future lead optimization studies.
70 mM) of emodin led to significant inhibition of cell proliferation,
presumably via down regulation of ERCC1 and Rad51 [37]. It is
unclear whether the emodin derivatives tested in our studies retain
some activities for those known targets and pathways. Therefore,
we postulate that the emodin derivatives reported herein exert
their anti-proliferative activity at least in part, but may not solely,
via their inhibition of ACL. Future studies with more potent ACL
inhibitors include selectivity measurements across a spectrum of
reported pathways that are known to mediate the anticancer ac-
tivity of emodin.
4. Experimental section
4.1. Chemistry
2.5. Reduction of cancer stemness
4.1.1. General
We had previously shown that shRNA KD of ACL in HMLE-Snail
breast cancer cells and in A549 lung adenocarcinoma cells reduced
cancer stemness [12]. We show here for the first time that small
molecule ACL inhibitors are highly effective in altering cancer
stemness. We applied the 3D spheroid assay [39,40] to evaluate the
effect of the ACL inhibitor 1d in modulating cancer stem cells. The
FAK kinase inhibitor VS-6063 [41], which is known to target cancer
All reagents and solvents were purchased from commercially
available sources and used without further purification. All re-
actions were carried out according to the indicated procedures and
conditions. Reactions were monitored by LC/MS analysis and/or
thin-layer chromatography (TLC) on silica-coated glass or
aluminum plates (EMD silica gel 60 F254) with the indicated eluent.
The compounds were visualized by UV light (254 nm). LC/MS
Fig. 4. Inhibition of A549 cancer cell proliferations. A549 cells were treated with (a) a single concentration (5
mM) or escalating doses (1, 5, 10 mM) of ACL inhibitors 1d (b) and 1o
(c) for 24 h and then analyzed for proliferation. ACL knockdown (KD) by siRNA was included as positive control (not shown in (b) and (c)).

S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
925
Fig. 5. Inhibition of cancer stemness in A549 lung and E-snail breast cancer cells. A549 and E-snail cells under the 3D culture condition were treated with the ACL inhibitors 1d
and BMS-303141, or the FAK inhibitor VS6063. Cells were analyzed for spheroid size and intensity density (IntDen). Representative confocal microscopic images of E-snail cells (A)
and evaluation score (IntDen/Area) (B). Spheroid size (Area) was scaled down by inhibitors, which was calculated as an inner area of spheroid borders (blue). Cellular damages
detected as an intensity of Sytox Green, which was calculated as Intensity Density (IntDen). (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
analysis was performed on an Agilent 1200 HPLC/UV (220 nm and/
or 254 nm wavelength) system coupled with a mass spectroscopic
(Applied Biosystems, MDS SCIEX, Q TRAP LC/MS/MS) detector.
Compounds for analysis were dissolved in 100% DMSO and sepa-
rated on C18 cartridge (particle size 2.6 m, dimensions:
100 mm ꢂ 2.1 mm, 0.3 mL/min flow rate, 1 mL injection volume)
using acetonitrile/water mobile phase with 0.1% formic acid as a
modifier. The gradient started at 20% acetonitrile, held for 2min,
and linearly increased to 97% acetonitrile over 10 min, with 3 min
hold at 97% acetonitrile and subsequent re-equilibration to the
original conditions in a total of 17 min. All compounds reported
were obtained in a purity as >95% at 254 nm wavelength. Nuclear
magnetic resonance (1H NMR) spectra were recorded on a Varian
Mercury plus NMR spectrometer operating at 400.13 MHz fre-
quencies for 1H, using a 5 mm ASW PFG probe capable of detecting
1H, 13C, 31P, and 15N nuclei. The proton chemical shifts (ppm) were
referenced to the tetramethylsilane internal standard (0 ppm).
NMR data are reported with these descriptions: s, singlet; d,
doublet; t, triplet; q, quartet; m, multiplet; br, broad peak.
4.1.2. 2-Chloro-1,3,8-trihydroxy-6-methyl-anthracene-9,10-dione
(1b)
A mix of Emodin (34 mg, 0.12 mmol) and ZrCl4 (4.0 mg,
0.017 mmol, 0.14 eq) in dioxane (2.5 mL) was purged with argon for
30 min, heated to 70 ꢁC over 40 min. To the resulting suspension
was added a solution of NCS (22 mg, 0.16 mmol, 1.3 eq) in dioxane
(0.5 mL) dropwise over 5 min at 70 ꢁC. The resulting reaction mix
was stirred at 70 ꢁC for 4 h, cooled to r.t. purified by silica gel
chromatography to give the product as a yellow solid (29 mg, 76%
yield). 1H NMR agrees with literature report [21]. 1H NMR
(400 MHz, DMSO-d6):
d 12.74 (s, 1H), 12.21 (s, br, 1H), 11.81 (s, 1H),
7.50 (s, 1H), 7.31 (s, 1H), 7.18 (d, J ¼ 0.8 Hz, 1H), 2.41 (s, 3H).
4.1.3. 2-Chloro-1,3,8-trimethoxy-6-methyl-anthracene-9,10-dione
(4)
The title compound was obtained analogously to 1b by treating
1,3,8-trimethoxy-6-methyl-anthracene-9,10-dione with ZrCl4 in
dioxane. The 1H NMR corresponds to the literature report.
Emodin derivatives 1d, 1e, 1p-s, and 3a-b were synthesized by
following reported procedures. Analytical data agreed with the
assigned structures and reported values.
1H NMR (400 MHz, DMSO-d6):
d 7.34 (s, 1H), 7.29 (s, 1H), 7.10 (s,
1H), 4.01 (s, 3H), 3.94 (s, 3H), 3.86 (s, 3H), 2.41 (s, 3H). MS (ESIþ): m/
z calc. for [C18H15ClO5] 346.06. Found: 347.3 [MþH]þ

926
S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
4.1.4. 1-Chloro-2,4,5-trihydroxy-7-methyl-anthracene-9,10-dione
(1c)
4.1.11. 1,6,8-Trihydroxy-7-(4-methoxyphenyl)-3-methyl-2,3,9,10-
tetrahydroanthracene-9,10-dione (1j)
To a suspension of Emodin (132 mg, 0.5 mmol) in CHCl3/MeCN
(5:1, 24 mL) was added a solution of SO2Cl2 in DCM (1.0 M, 2.5 mL,
5.0 eq) in one portion at room temperature. The resulting reaction
mix was stirred at room temperature under nitrogen for 2 days,
diluted with hexanes (50 mL), filter to collect the crude product as a
yellow solid (128 mg, 86% yield). NMR agreed with literature report
1H NMR (400 MHz, DMSO-d6):
d 12.54 (s, 1H), 11.91 (s, 1H), 11.22
(s, 1H), 7.48 (s, 1H), 7.32 (s, 1H), 7.25e7.27 (dd, J ¼ 8.4 Hz, 2.0 Hz,
2H), 7.14e7.16 (m, 2H), 6.93 (d, J ¼ 8.4 Hz, 2H), 3.74 (s, 3H), 2.37 (s,
3H). MS (ESIþ): m/z calc. for [C22H18O6] 376.1 Found 377.1[MþH]þ.
4.1.12. 2-(3,5-Dimethoxyphenyl)-1,3,8-trihydroxy-6-methyl-9,10-
[20]. 1H NMR (400 MHz, acetone-d6):
d
12.83 (s, 1H), 11.87 (s, 1H),
dihydroanthracene-9,10-dione (1k)
10.39 (s, br, 1H), 7.57 (s, 1H), 7.17 (s, 1H), 6.86 (s, 1H), 2.49 (s, 3H). MS
1H NMR (400 MHz, acetone-d6):
d 12.64 (s, 1H), 12.09 (s, 1H),
(ESIþ): m/z calc. for [C15H9ClO5] 304.01. Found: 305.5 [MþH]þ
7.62 (s, 1H), 7.49 (s, 1H), 7.18 (d, J ¼ 0.8 Hz, 1H), 6.60 (d, J ¼ 2.4 Hz,
2H), 6.52 (t, J ¼ 2.4 Hz, 1H), 3.84 (s, 6H), 2.49 (s, 3H).
4.1.5. 1,3,8-Trihydroxy-2,4-dibromo-6-methyl-anthraquinone (1d)
To a stirred suspension of emodin (1.08 g, 4.0 mmol) in solvent
of THF (100 mL) was added NBS (Bromosuccinimide 1.7 g,
9.6 mmol) at room temperature. After completion (LCMS) the re-
action mixture was concentrated in vacuum. The crude product
precipitated was filtered and further purified by flash column
chromatography (silica gel; Hexanes: CH2Cl2 ¼ 3:1e100% CH2Cl2)
to yield compound 2d as a red solid. NMR agreed with literature
MS (ESIþ): m/z calc. for [C23H18O7] 406.39 Found 407.6 [Mþ1]þ.
4.1.13. 1,3,8-Trihydroxy-6-methyl-2-(pyridin-3-yl)-9,10-
dihydroanthracene-9,10-dione (1l)
1H NMR (400 MHz, DMSO-d6):
d 12.59 (s, 1H), 11.84 (s, 1H), 11.59
(br s, 1H), 8.68 (br s, 1H), 8.53 (br s, 1H), 7.91 (br s, 1H), 7.68 (br s,
1H), 7.25 (s,1H), 7.16 (s,1H), 6.79 (s,1H), 2.34 (s, 3H). MS (ESIþ): m/z
calc. for [C20H13NO5] 347.08 Found 348 [MþH]þ.
report [25]. 1H NMR (400 MHz, DMSO-d6)
d 13.69 (s, 1H), 11.62 (s,
1H), 7.43 (s, 1H), 7.11 (s, 1H), 2.40 (s, 3H).
4.1.14. 1,3,8-Trihydroxy-6-methyl-2-(4-pyridyl)anthracene-9,10-
dione (1m)
4.1.6. General procedure for the synthesis of 2-aryl emodins 1f-1m,
1o
1H NMR (400 MHz, DMSO-d6):
d 12.69 (s, 1H), 11.92 (s, 1H), 8.70
(d, J ¼ 4.4 Hz, 2H), 7.52e7.57 (m, 3H), 7.40 (s, 1H), 7.23 (s, 1H). 2.45
(s, 3H). MS (ESIþ): m/z calc. for [C20H13NO5] 347.33 Found: 348.5
[MþH]þ.
1,3,8-Trihydroxy-2-iodo-6-methyl-anthracene-9,10-dione (1e)
(79.2 mg, 0.2 mmol) was dissolved in toluene (2 mL), ethanol (2 mL)
and water (0.5 mL). Palladium tetrakis triphenylphosphine
(46.2 mg, 0.04 mmol) was added. Aryl boronic acid (0.8 mmol) and
saturated NaHCO3 (0.5 mL) was added. The reaction mixture was
degassed with argon for 10min. The reaction mixture was heated to
80 ꢁC for 12 h or until complete (indicated by TLC). After cooling to
room temperature Ethyl acetate and water was added (10/10 mL).
The organic phase was separated and dried over Na2SO4. After
filtration the solvent was removed i. vac. The crude residue was
purified by prep. HPLC. Final purity was determined by LCMS. All
reported compounds were obtained in purities of 95% and above.
4.1.15. 1,3,8-Trihydroxy-6-methyl-2-(1-methylpyrazol-4-yl)
anthracene-9,10-dione (1o)
1H NMR (400 MHz, DMSO-d6):
d 13.24 (s, 1H), 11.95 (s, 1H), 11.71
(s, 1H), 8.34 (s, 1H), 8.15 (s, 1H), 7.52 (s, 1H), 7.38 (s, 1H), 7.19 (s, 1H),
3.90 (s, 1H), 2.47 (s, 3H). MS (ESIþ): m/z calc. for [C19H14N2O5]
350.32, Found 351.1 [MþH]þ.
4.1.16. 2,4,5-Trihydroxy-7-methyl-1,3-bis(pyridin-3-yl)-9,10-
dihydroanthracene-9,10-dione (1n)
1,3-dibromo-2,4,5-trihydroxy-7-methyl-anthracene-9,10-dione
1d (85.6 mg, 0.2 mmol) was dissolved in toluene (2 mL), ethanol
(2 mL) and water (0.5 mL). Palladium tetrakis triphenylphosphine
(46.2 mg, 0.04 mmol) was added. 3-Pyridyl boronic acid (123 mg,
1 mmol) and saturated NaHCO3 (0.5 mL) was added. The reaction
mixture was degassed with argon for 10min. The reaction mixture
was heated to 80 ꢁC for 5 h. After cooling to room temperature Ethyl
acetate and water was added (10/10 mL). The organic phase was
separated and dried over Na2SO4. After filtration the solvent was
removed i. vac. The crude residue was purified by prep. HPLC. 1H
4.1.7. 1,3,8-Trihydroxy-6-methyl-2-phenyl-9,10-
dihydroanthracene-9,10-dione (1f)
1H NMR (400 MHz, DMSO-d6):
d 12.56 (s, 1H), 11.95 (s, 1H), 11.33
(s,1H), 7.54 (d, J ¼ 1.6 Hz,1H), 7.34e7.42 (m, 6H), 7.19 (s,1H), 2.42 (s,
3H). MS (ESIþ): m/z calc. for [C21H14O5] 346.08. Found: 347.5
[MþH]þ.
4.1.8. 1,3,8-Trihydroxy-6-methyl-2-(3-methylphenyl)-9,10-
dihydroanthracene-9,10-dione (1g)
1H NMR (400 MHz, DMSO-d6):
d
12.59 (s, 1H), 12.03 (s, 1H), 11.32
NMR (400 MHz, acetone-d6): d 12.31 (s, 1H), 11.91 (s, 1H), 8.75 (s, br,
(s, 1H), 7.58 (s, 1H), 7.42 (s, 1H), 7.34 (t, J ¼ 7.2 Hz, 1H), 7.23 (s, 1H),
7.16e7.19 (m, 3H), 2.47 (s, 3H), 2.35 (s, 3H). MS (ESIþ): m/z calc. for
[C22H16O5] 360.37 Found 361.7 [MþH]þ.
1H), 8.70 (d, J ¼ 4.4 Hz, 1H), 8.63 (d, J ¼ 6 Hz, 1H), 8.61 (s, br, 1H),
7.68e7.65 (dd, J ¼ 8.4 Hz, 5.2 Hz, 1H), 7.59e7.56 (dd, J ¼ 8.4 Hz,
5.2 Hz, 1H), 7.34 (s, 1H), 7.15 (s, br, 1H), 2.44 (s, 3H). MS (ESIþ): m/z
calc. for [C25H16N2O5] 424.4 Found 425.5 [MþH]þ.
4.1.9. 1,3,8-Trihydroxy-6-methyl-2-(4-methylphenyl)-9,10-
dihydroanthracene-9,10-dione (1h)
4.1.17. 4-Chloro-1,3,8-trihydroxy-6-methyl-9,10-dihydroanthracen-
9-one (2c)
1H NMR (400 MHz, DMSO-d6):
d
12.55 (s, 1H), 11.96 (s, 1H), 11.28
(s, 1H), 7.53(d, J ¼ 1.6 Hz, 1H), 7.37 (s, 1H), 7.19e7.26 (m, 5H), 2.42 (s,
3H), 2.34 (s, 3H). MS (ESIþ): m/z calc. for [C22H16O5] 360.37 Found
361.3 [MþH]þ.
1,3,8-trihydroxy-6-methyl-10H-anthracen-9-one
(64
mg,
0.25 mmol) was dissolved in diethyl ether at room temperature.
SO2Cl2 (1 M, 1.25 ml, 1.125 mmol) was added to the solution. The
reaction was stirred at room temperature under exclusion of light
for 25 h. The product was isolated by filtration and stored at ꢀ20 ꢁC
4.1.10. 1,3,8-Trihydroxy-2-(3-methoxyphenyl)-6-methyl-9,10-
dihydroanthracene-9,10-dione (1i)
in the dark. 1H NMR (400 MHz, acetone-d6):
d 12.54 (s, 1H), 12.07 (s,
1H NMR (400 MHz, DMSO-d6):
d
12.55 (s, 1H), 11.96 (s, 1H), 11.32
1H), 11.67 (s, 1H), 6.92 (s, 1H), 6.71 (s, 1H), 6.48 (s, 1H), 4.27 (s, 2H),
2.33 (s, 3H). 13C-HSQC (400 MHz, DMSO-d6):
120.4 (C5), 115.6
(C7), 102.2 (C2), 31.5 (C10), 21.1 (CH3). 13C-HMBC (400 MHz, DMSO-
d6): C5-correlation with CH3 and C10, C7 correlation with CH3, C2
(s, 1H), 7.538 (d, J ¼ 1.6 Hz, 1H), 7.35 (d, J ¼ 15.6 Hz, 2H), 7.19 (s, 1H),
6.89e6.93 (m, 3H), 3.75 (s, 3H), 2.47e2.49 (m, 3H). MS (ESIþ): m/z
calc. for [C22H16O6] 376.36. Found 377.4 [MþH]þ.
d
d

S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
927
no correlation with C10. MS (ESIþ): m/z calc. for [C15H11ClO4]
290.03. Found: [MþH]þ
of the mixed staining solution; Hoechst 33342 (1 mg/mL, Life
Technologies) as counterstain for all nuclei and Sytox Green, as
stain for dead cells (2 mM, Life Technologies) at a final dilution of
1:10,000 each.
4.2. Docking study
The two-dimensional structure and ionization state of the nat-
ural product anthracenone 2a were drawn in two dimensions using
Marvindraw (Marvin 6.2.2, 2014, ChemAxon: http://www.
chemaxon.com). 2a was converted into a low energy three-
dimensional structure using Openbabel (v2.3.2) [42]. The ACL
crystal structure (PDB: 3MWD) and small molecule 2a were pre-
pared for docking using Autodocktools (v1.5.6) [43]. Autodocking
was preformed using Autodock Vina (v1.1.2) and following the
standard published docking protocol [44]. The low energy binding
pose of 2a relative to ACL was visualized using a combination of
Autodocktools (v1.5.6) and USCF chimera (v1.11) [43], [45].
4.5.2. Image acquisition and analysis
3D images were acquired by confocal microscopy (Zeiss LSM 510
Meta), focusing on the spheroid center. Each image was analyzed by
ImageJ, by splitting into Hoechst (blue) image and Sytox Green
image. Spheroid size (Area) was scaled down by effective reagents,
which was calculated as an inner area of spheroid borders (Hoechst
channel). Cellular damages detected as an intensity of Sytox Green,
which was calculated as Intensity Density (IntDen). IntDen/Area
was used as a readout index for inhibitors effectiveness.
Competing interests
4.3. Enzymatic assay
The authors declared no competing interests.
The assay was performed using ADP-Glo luminescence assay
reagents. It measures ACLY activity by quantification of the amount
of ADP generated by the enzymatic reaction. The luminescent
signal from the assay is correlated with the amount of ADP
generated and is proportionally correlated with the amount of ACLY
Author's contributions
SKK and SB synthesized the compounds. FEJ conducted the
modeling study. JH, MO, CK, and SE carried out biological assays.
SKK, FEJ, JH, VS, and LS drafted the manuscript. VS and LS conceived
and directed the studies. All authors read and approved the
manuscript.
activity. The compounds were diluted in 10% DMSO and 2.5
dilution was added to a 25 l reaction so that the final concentra-
tion of DMSO is 1% in all of reactions. All of the enzymatic reactions
were conducted at 30 ꢁC for 60 min. The 25
l reaction mixture
contains 40 mM Tris, pH 8.0, 10 mM MgCl2, 5 mM DTT, ATP, CoA,
and Sodium Citrate and ACLY. After the enzymatic reaction, 25 l of
ADP-Glo reagent was added to each reaction and incubate the plate
for 45 min at room temperature. After then, 50 l of Kinase
ml of the
m
m
Acknowledgement
m
This research is partially supported by a seed fund for the Center
for Drug Discovery and Translational Research (LS) and a research
grant by Fujifilm (VPS and LS). The Sponsors played no roles in
study design, the collection, analysis and interpretation of data, the
writing of the report, and the decision to submit the article for
publication. We thank Charles Sheehan of the NMR Core of Harvard
Medical School for his critical advice on 13C-HMBC/13C-HSQC ex-
periments and structural determinations of compound 2c, and BPS
Bioscience for technical support in the ADP Glo enzymatic assay.
m
Detection reagent was added and incubated for 60 min at room
temperature. The luminescence signal was measured using a Bio-
Tek Synergy 2 microplate reader.
4.4. Proliferation assay (WST-1)
A549 cells were seeded in a 96-well plate at a density of
5 ꢂ 104 cells/well in 100
ml of Ham's F-12 medium supplemented
Appendix A. Supplementary data
with 10% FBS, with or without compounds to be tested, and
cultured for 24, 48 72 h in a CO2 incubator at 37 ꢁC in 5% CO2. At
each time point of measurement, WST-1 reagent (Roche Di-
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2016.12.018.
agnostics, 10 ml) were added to each well, and after 1 h incubation
in a CO2 incubator, the absorbance of each well was measured using
a microplate reader (Beckman Coulter DTX880 multimode detec-
tor) at a wavelength of 450 nm.
References
[1] L. Galluzzi, O. Kepp, M.G. Vander Heiden, G. Kroemer, Metabolic targets for
cancer therapy, Nat. Rev. Drug Disc. 12 (2013) 829e846.
4.5. 3D spheroid assay and evaluation
4.5.1. Spheroid generation
[2] M. Jang, S.S. Kim, J. Lee, Cancer cell metabolism: implications for therapeutic
targets, Exp. Mol. Med. 45 (2013) e45.
[3] R.A. Cairns, I.S. Harris, T.W. Mak, Regulation of cancer cell metabolism, Nat.
Rev. Cancer 11 (2011) 85e95.
50 ml of a heated 1.5% w/v agarose (Sigma A9539; in DMEM
without fetal bovine serum (FBS)) solution was dispended into
sterile 96-well clear bottom imaging plates. For tumor spheroid
seeding, suspended cells in regular medium (100 ml) were added
into agarose-coated well. Cell number for seeding was optimized
for each cell lines to obtain spheroids with an approximate diam-
[4] G. Hatzivassiliou, F. Zhao, D.E. Bauer, C. Andreadis, A.N. Shaw, D. Dhanak,
S.R. Hingorani, D.A. Tuveson, C.B. Thompson, ATP citrate lyase inhibition can
suppress tumor cell growth, Cancer Cell 8 (2005) 311e321.
[5] M. Board, E. Newsholme, Hydroxycitrate causes altered pyruvate metabolism
by tumorigenic cells, Biochem. Mol. Biol. Int. 40 (1996) 1047e1056.
[6] R.J. DeBerardinis, J.J. Lum, G. Hatzivassiliou, C.B. Thompson, The biology of
cancer: metabolic reprogramming fuels cell growth and proliferation, Cell
Metab. 7 (2008) 11e20.
[7] P.F. McAndrew, Fat metabolism and cancer, Surg. Clin. North. Am. 66 (1986)
1003e1012.
[8] J.V. Swinnen, K. Brusselmans, G. Verhoeven, Increased lipogenesis in cancer
cells: new players, novel targets, Curr. Opin. Clin. Nutr. Metab. Care 9 (2006)
358e365.
eter of 400 mm on day 5: 2000 cells per well (c/w) for A549 lung
adenocarcinoma cells and HMLE-Snail breast cancer cells. The
plates were incubated under standard cell culture conditions at
37 ꢁC and 5% CO2 in humidified incubators for 2 days to allow
formation of reproducible spheroids of defined size and
[9] Y. Zhou, L.R. Bollu, F. Tozzi, X. Ye, R. Bhattacharya, G. Gao, E. Dupre, L. Xia, J. Lu,
F. Fan, S. Bellister, L.M. Ellis, Z. Weihua, ATP citrate lyase mediates resistance of
colorectal cancer cells to SN38, Mol. Cancer Ther. 12 (2013) 2782e2791.
[10] K.E. Wellen, G. Hatzivassiliou, U.M. Sachdeva, T.V. Bui, J.R. Cross,
C.B. Thompson, ATP-citrate lyase links cellular metabolism to histone
morphology, followed by adding each reagent (in 100
medium) for additional 3 days.
ml culture
Prior to imaging, spheroids were stained for 24 h by adding 20 ml

928
S.K. Koerner et al. / European Journal of Medicinal Chemistry 126 (2017) 920e928
acetylation, Science 324 (2009) 1076e1080.
and anthraoxazines, Tetrahedron Lett. 54 (2013) 2802e2805.
[11] J. Hanai, N. Doro, A.T. Sasaki, S. Kobayashi, L.C. Cantley, P. Seth, V.P. Sukhatme,
Inhibition of lung cancer growth: ATP citrate lyase knockdown and statin
treatment leads to dual blockade of mitogen-activated protein kinase (MAPK)
and phosphatidylinositol-3-kinase (PI3K)/AKT pathways, J. Cell. Physiol. 227
(2012) 1709e1720.
[12] J.I. Hanai, N. Doro, P. Seth, V.P. Sukhatme, ATP citrate lyase knockdown im-
pacts cancer stem cells in vitro, Cell Death Dis. 4 (2013) e696.
[13] A.D. Khwairakpam, M.S. Shyamananda, B.L. Sailo, S.R. Rathnakaram,
G. Padmavathi, J. Kotoky, A.B. Kunnumakkara, ATP citrate lyase (ACLY): a
promising target for cancer prevention and treatment, Curr. Drug Targets. 16
(2015) 156e163.
[14] X.Y. Zu, Q.H. Zhang, J.H. Liu, R.X. Cao, J. Zhong, G.H. Yi, Z.H. Quan, G. Pizzorno,
ATP citrate lyase inhibitors as novel cancer therapeutic agents, Recent Pat,
Anticancer Drug Disc. 7 (2012) 154e167.
[15] N. Zaidi, I. Royaux, J.V. Swinnen, K. Smans, ATP citrate lyase knockdown in-
duces growth arrest and apoptosis through different cell- and environment-
dependent mechanisms, Mol. Cancer Ther. 11 (2012) 1925e1935.
[16] N. Zaidi, J.V. Swinnen, K. Smans, ATP-citrate lyase: a key player in cancer
metabolism, Cancer Res. 72 (2012) 3709e3714.
[17] A. Liu, H. Chen, W. Wei, S. Ye, W. Liao, J. Gong, Z. Jiang, L. Wang, S. Lin,
Antiproliferative and antimetastatic effects of emodin on human pancreatic
cancer, Oncol. Rep. 26 (2011) 81e89.
[18] J.J. Oleynek, C.J. Barrow, M.P. Burns, D.M. Sedlock, D.J. Murphy, P.V. Kaplita,
H.H. Sun, R. Cooper, A.M. Gillum, C.C. Chadwick, Anthrones, naturally occur-
ring competitive inhibitors of adenosine-triphosphate-citrate lyase, Drug Dev.
Res. 36 (1995) 35e42.
[28] J.L. Liang, H.C. Cha, S.H. Lee, J.-K. Son, H.W. Chang, J.-E. Eom, Y. Kwon, Y. Jahng,
A facile synthesis of emodin derivatives, emodin carbaldehyde, citreorosein,
and their 10-deoxygenated derivatives and their inhibitory activities on m-
calpain, Arch. Pharm. Res. 35 (2012) 447e454.
[29] B. Shroot, C. Brown, Free radicals in skin exposed to dithranol and its de-
rivatives, Arzneimittelforschung 36 (1986) 1253e1255.
[30] J. Fuchs, L. Packer, Investigations of anthralin free radicals in model systems
and in skin of hairless mice, J. Invest. Dermatol. 92 (1989) 677e682.
[31] A.D. Gribble, R.E. Dolle, A. Shaw, D. McNair, R. Novelli, C.E. Novelli,
B.P. Slingsby, V.P. Shah, D. Tew, B.A. Saxty, M. Allen, P.H. Groot, N. Pearce,
J. Yates, ATP-citrate lyase as a target for hypolipidemic intervention. Design
and synthesis of 2-substituted butanedioic acids as novel, potent inhibitors of
the enzyme, J. Med. Chem. 39 (1996) 3569e3584.
[32] F. Fan, H.J. Williams, J.G. Boyer, T.L. Graham, H. Zhao, R. Lehr, H. Qi,
B. Schwartz, F.M. Raushel, T.D. Meek, On the catalytic mechanism of human
ATP citrate lyase, Biochemistry 51 (2012) 5198e5211.
[33] J.J. Li, H. Wang, J.A. Tino, J.A. Robl, T.F. Herpin, R.M. Lawrence, S. Biller, H. Jamil,
R. Ponticiello, L. Chen, C.H. Chu, N. Flynn, D. Cheng, R. Zhao, B. Chen, D. Schnur,
M.T. Obermeier, V. Sasseville, R. Padmanabha, K. Pike, T. Harrity, 2-hydroxy-
N-arylbenzenesulfonamides as ATP-citrate lyase inhibitors, Bioorg. Med.
Chem. Lett. 17 (2007) 3208e3211.
[34] Z. Ma, C.H. Chu, D. Cheng, A novel direct homogeneous assay for ATP citrate
lyase, J. Lipid Res 50 (2009) 2131e2135.
[35] T. Sun, K. Hayakawa, K.S. Bateman, M.E. Fraser, Identification of the citrate-
binding site of human ATP-citrate lyase using X-ray crystallography, J. Biol.
Chem. 285 (2010) 27418e27428.
[19] I. Yosioka, H. Yamauchi, K. Morimoto, I. Kitagawa, Two new chlorine con-
taining anthraquinones from a lichen, anaptychia obscurata (nyl.) vain, Tet-
rahedron Lett. 9 (1968) 1149e1152.
[20] Q.H.Y. Chen, W. Huang, Emodin Derivative and its Application as Antibacterial
Agents, 2012. CN 102775291.
[36] H.Z. Lee, Effects and mechanisms of emodin on cell death in human lung
squamous cell carcinoma, Br. J. Pharmacol. 134 (2001) 11e20.
[37] L. He, J.J. Bi, Q. Guo, Y. Yu, X.F. Ye, Effects of emodin extracted from Chinese
herbs on proliferation of non-small cell lung cancer and underlying mecha-
nisms, Asian Pac. J. Cancer Prev. 13 (2012) 1505e1510.
[21] M.V. Sargent, D.O.N. Smith, J.A. Elix, The minor anthraquinones of Xanthoria
parietina(L.) Beltram, the chlorination of parietin, and the synthesis of fragilin
and 7-chloroemodin ('AO-1'), J. Chem. Soc. C. Org. (1970) 307e311.
[22] M.A. Elban, S.M. Hecht, Total synthesis of the topopyrones: a new class of
topoisomerase I poisons, J. Org. Chem. 73 (2008) 785e793.
[23] M. Koyama, K. Takahashi, T.C. Chou, Z. Darzynkiewicz, J. Kapuscinski,
T.R. Kelly, K.A. Watanabe, Intercalating agents with covalent bond forming
capability. A novel type of potential anticancer agents. 2. Derivatives of
chrysophanol and emodin, J. Med. Chem. 32 (1989) 1594e1599.
[24] S.B. Kalidhar, Structural elucidation in anthraquinones using 1H NMR glyco-
sylation and alkylation shifts, Phytochem 28 (1989) 3459e3463.
[25] F. Duan, X. Li, S. Cai, G. Xin, Y. Wang, D. Du, S. He, B. Huang, X. Guo, H. Zhao,
R. Zhang, L. Ma, Y. Liu, Q. Du, Z. Wei, Z. Xing, Y. Liang, X. Wu, C. Fan, C. Ji,
D. Zeng, Q. Chen, Y. He, X. Liu, W. Huang, Haloemodin as novel antibacterial
agent inhibiting DNA gyrase and bacterial topoisomerase I, J. Med. Chem. 57
(2014) 3707e3714.
[26] J.-M. Lapierre, S. Eathiraj, D. Vensel, Y. Liu, C.O. Bull, S. Cornell-Kennon,
S. Iimura, E.W. Kelleher, D.E. Kizer, S. Koerner, S. Makhija, A. Matsuda,
M. Moussa, N. Namdev, R.E. Savage, J. Szwaya, E. Volckova, N. Westlund,
H. Wu, B. Schwartz, Discovery of 3-(3-(4-(1-aminocyclobutyl)phenyl)-5-
phenyl-3H-imidazo[4,5-b]pyridin-2-yl)pyridin-2-amine (ARQ 092): an orally
bioavailable, selective, and potent allosteric AKT inhibitor, J. Med. Chem. 59
(2016) 6455e6469.
[38] G. Srinivas, S. Babykutty, P.P. Sathiadevan, P. Srinivas, Molecular mechanism
of emodin action: transition from laxative ingredient to an antitumor agent,
Med. Res. Rev. 27 (2007) 591e608.
[39] C. Wenzel, B. Riefke, S. Grundemann, A. Krebs, S. Christian, F. Prinz,
M. Osterland, S. Golfier, S. Rase, N. Ansari, M. Esner, M. Bickle, F. Pampaloni,
C. Mattheyer, E.H. Stelzer, K. Parczyk, S. Prechtl, P. Steigemann, 3D high-
content screening for the identification of compounds that target cells in
dormant tumor spheroid regions, Exp. Cell Res. 323 (2014) 131e143.
[40] J. Friedrich, C. Seidel, R. Ebner, L.A. Kunz-Schughart, Spheroid-based drug
screen: considerations and practical approach, Nat. Protoc. 4 (2009) 309e324.
[41] F.J. Sulzmaier, C. Jean, D.D. Schlaepfer, FAK in cancer: mechanistic findings and
clinical applications, Nat. Rev. Cancer 14 (2014) 598e610.
[42] N.M. O'Boyle, M. Banck, C.A. James, C. Morley, T. Vandermeersch,
G.R. Hutchison, Open Babel: An open chemical toolbox, J. Cheminform 3
(2011) 33.
[43] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A.J. Olson, AutoDock4 and AutoDockTools4: Automated docking with selec-
tive receptor flexibility, J. Comput. Chem. 30 (2009) 2785e2791.
[44] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of
docking with
a new scoring function, efficient optimization, and multi-
threading, J. Comput. Chem. 31 (2010) 455e461.
[45] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt,
E.C. Meng, T.E. Ferrin, UCSF Chimeraea visualization system for exploratory
research and analysis, J. Comput. Chem. 25 (2004) 1605e1612.
[27] L.-M. Zhao, L.-M. Zhang, F.-Y. Ma, X.-S. Wang, H.-S. Jin, Catalyst-free Mannich
reaction of hydroxyanthraquinone: facile access to emodin Mannich bases