Indirubin core structure of glycogen synthase kinase-3 inhibitors as novel chemotype for intervention with 5-lipoxygenase

Article abstract of DOI:10.1021/jm401740w

The enzymes 5-lipoxygenase (5-LO) and glycogen synthase kinase (GSK)-3 represent promising drug targets in inflammation. We made use of the bisindole core of indirubin, present in GSK-3 inhibitors, to innovatively target 5-LO at the ATP-binding site for the design of dual 5-LO/GSK-3 inhibitors. Evaluation of substituted indirubin derivatives led to the identification of (3Z)-6-bromo-3-[(3E)-3-hydroxyiminoindolin-2-ylidene]indolin-2-one (15) as a potent, direct, and reversible 5-LO inhibitor (IC₅₀ = 1.5 μM), with comparable cellular effectiveness on 5-LO and GSK-3. Together, we present indirubins as novel chemotypes for the development of 5-LO inhibitors, the interference with the ATP-binding site as a novel strategy for 5-LO targeting, and dual 5-LO/GSK-3 inhibition as an unconventional and promising concept for anti-inflammatory intervention.

Full text of DOI:10.1021/jm401740w

Article
pubs.acs.org/jmc
Indirubin Core Structure of Glycogen Synthase Kinase‑3 Inhibitors as
Novel Chemotype for Intervention with 5‑Lipoxygenase
Carlo Pergola,*,† Nicolas Gaboriaud-Kolar,∥ Nadine Jestadt,† Stefanie Konig,† Marina Kritsanida,∥
̈
̈
Anja M. Schaible,† Haokun Li,† Ulrike Garscha,† Christina Weinigel,‡ Dagmar Barz,‡ Kai F. Albring,§
Otmar Huber,§ Alexios L. Skaltsounis,∥ and Oliver Werz*,†
†Pharmaceutical/Medicinal Chemistry, Institute of Pharmacy, Friedrich-Schiller-University, Jena 07743, Germany
§
‡Institute of Transfusion Medicine and Institute of Biochemistry II, Jena University Hospital, Jena 07743, Germany
∥Department of Pharmacognosy and Chemistry of Natural Products, School of Pharmacy, University of Athens, Athens 15771,
Greece
S
* Supporting Information
ABSTRACT: The enzymes 5-lipoxygenase (5-LO) and
glycogen synthase kinase (GSK)-3 represent promising drug
targets in inflammation. We made use of the bisindole core of
indirubin, present in GSK-3 inhibitors, to innovatively target 5-
LO at the ATP-binding site for the design of dual 5-LO/GSK-
3 inhibitors. Evaluation of substituted indirubin derivatives led
to the identification of (3Z)-6-bromo-3-[(3E)-3-hydroxyimi-
noindolin-2-ylidene]indolin-2-one (15) as a potent, direct, and
reversible 5-LO inhibitor (IC50 = 1.5 μM), with comparable
cellular effectiveness on 5-LO and GSK-3. Together, we
present indirubins as novel chemotypes for the development of
5-LO inhibitors, the interference with the ATP-binding site as
a novel strategy for 5-LO targeting, and dual 5-LO/GSK-3
inhibition as an unconventional and promising concept for anti-inflammatory intervention.
to identify potent 5-LO inhibitors.4 These compounds interfere
with 5-LO mainly in three different ways: (I) active-site iron
INTRODUCTION
■
5-Lipoxygenase (5-LO) is an iron-containing dioxygenase that
chelation, (II) uncoupling of the redox cycle of the active site
initiates the biosynthesis of leukotrienes (LT) from arachidonic
iron, and (III) substrate/fatty acid competition at the active
acid (AA), by the incorporation of molecular oxygen to form
site. Accordingly, most 5-LO inhibitors are metal ion chelators,
5(S)-hydroperoxy-6-trans-8,11,14-cis-eicosatetraenoic acid (5-
HpETE) and the subsequent dehydration to leukotriene
reducing agents, or fatty acid mimetics.4 Despite intensive
(LT)A4.1 In the cells, AA is liberated by the cytosolic
research, however, only the iron ligand-type 5-LO inhibitor
compound 25 (zileuton, 1-(1-benzothiophen-2-ylethyl)-1-
phospholipase (cPL)A2 and delivered to 5-LO by the 5-LO-
hydroxy-urea, Table 1) has been approved so far for asthma
activating protein (FLAP). The unstable epoxide LTA4 is then
treatment, though its use is limited due to liver toxicity. Other
metabolized to LTB4 or to cysteinyl-containing LTS (C4, D4,
and E4; cysLTs), which act via specific receptors (BLT and
lead compounds or concepts have instead failed during
development, as a result of severe side effects (e.g.,
CysLT receptors 1 and 2) to promote inflammation and
methemoglobinemia for redox active compounds) or lack of
regulate immunity (LTB4) or to induce smooth muscle
contraction and plasma extravasation (cys-LTs).2 Thus,
CysLT1 antagonists (e.g., montelukast) are effectively used to
treat asthma and allergy, while clinical trials for BLT1
efficacy (e.g., ineffectiveness under oxidative stress for non-
redox type inhibitors),4 thereby generating demand for novel
chemotypes and mechanisms of intervention with 5-LO, as well
as for novel therapeutical strategies. In fact, considerable efforts
have been directed toward the identification of dual-inhibition
approaches, mainly targeting 5-LO and another enzyme of the
antagonists have not been successful so far, despite excellent
preclinical results.3 As compared to LT-receptor antagonism,
however, targeting LT biosynthesis offers the obvious
AA cascade, in order to improve therapeutic efficacy and to
advantage of suppressing all LTs and appeals for its potentially
higher efficacy as anti-inflammatory therapy.4
reduce side effects. Thus, dual-inhibition of 5-LO/cyclo-
oxygenase (COX),10 5-LO/microsomal prostaglandin E2
Thus, different sources (e.g., nature, synthesis) and diverse
medicinal chemistry approaches (e.g., lead identification,5
scaffold-hopping,6 structure-based drug design,7 dynamic
Received: November 12, 2013
Published: April 3, 2014
modeling,8 structural optimization9) have long been explored
© 2014 American Chemical Society
3715
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
Table 1. Inhibition of 5-LO Product Formation by IO and 7-Bromo Indirubin Derivatives
a
b
Values are means SE, n = 3 or 4. **p < 0.01; ***p < 0.001 vs control; ANOVA + Bonferroni. Cell-free: isolated human recombinant 5-LO.
c
Intact cells: intact human neutrophils stimulated with 2.5 μM A23187 ionophore.
synthase-1 (mPGES-1),11 or 5-LO/soluble epoxide hydrolase
(sEH)12 has been examined with promising results.
ATP-binding to 5-LO, apparently via the C2-like domain, is
exploited for purification of 5-LO using an ATP affinity
column.20 However, the exact position of the ATP-binding site
of 5-LO is unknown, thus precluding molecular modeling
approaches, and this site has only been analyzed with probes for
photoaffinity labeling.20,21 Consequently, we here aimed to
analyze (i) whether indirubin analogues may serve as novel
chemotypes for the development of 5-LO inhibitors with
improved potency as compared to that of 1, (ii) the possible
interference with the ATP-binding site of 5-LO as a new
molecular strategy for 5-LO inhibition, and (iii) whether dual
inhibition of 5-LO and GSK-3, which are both attractive targets
to control inflammation, may take place at similar compound
concentrations in relevant assays.
A comprehensive evaluation of 5-LO inhibitors requires both
cell-free and cellular test systems.4 We thus performed assays
with human recombinant 5-LO and with Ca2+-ionophore-
stimulated human neutrophils, respectively. Compound 1
inhibited 5-LO activity with an IC50 of 15.0 (cell-free) and
11.1 (cellular) μM. To identify 5-LO inhibitors more potent
than 1, we thus performed an initial screening of synthetic
indirubin derivatives at a concentration of 3 μM. At this
concentration, no significant inhibition was observed for 1
(Table 1), while the reference compound 25 worked as
expected (60−70% inhibition). To fulfill the objective, a set of
indirubins from our in-house library was rationally selected.
Thus, our selection included compounds known to interact
with the kinase ATP cavity in different fashions and with direct
or inverted binding modes, as previously shown for effects on
GSK-3 (14, 15),22 Aurora kinases (2, 6, 11, 12),23 and dual-
specificity tyrosine-regulated kinases (DYRKs; 7, 8, 9, 10).24
Overall, all known types of interaction of indirubins with the
ATP cavity were covered in a small set of compounds.
Here, we present the bisindole core of the natural alkaloid
indirubin as a novel chemotype for the development of
synthetic 5-LO inhibitors with a new molecular mode of action
of 5-LO interference, i.e., targeting the ATP-binding site. Since
these compounds have been revealed as potent inhibitors of
glycogen synthase kinase (GSK)-3,13,14 a central kinase in
immunity and inflammation, an unconventional 5-LO/GSK-3
dual-inhibition strategy is proposed.
RESULTS AND DISCUSSION
■
Several natural compounds suppress 5-LO product formation,15
but most of them act by unselective antioxidant activity, possess
metabolically labile moieties, and exert severe side effects.
Certain natural substances (e.g., hyperforin, tryptanthrin),
however, allowed the identification of novel 5-LO inhibitory
mechanisms,16,17 but their synthesis is often hardly accessible
and/or presents limitation for structural optimization. In a
recent study aimed to investigate the effect of the natural
indirubin derivative 1 (also known as indirubin-3′-oxime, IO)
on vascular inflammation as a promising vasoprotective
compound, we have identified a direct, though moderate, effect
of 1 on 5-LO activity, causing suppression of LT biosynthesis in
inflammatory cells.18 Interestingly, the 3,2′-bisindol core of 1
represents a novel structure as compared to known natural
and/or synthetic inhibitors of 5-LO and may prevent classical
limitations of natural 5-LO inhibitors, since it lacks obvious
antioxidant properties, while being prone to structural
modifications. Importantly, compound 1 has been a lead
compound for the development of inhibitors of kinases (such as
GSK-3; IC50, 22 nM in a cell-free assay),13,14,19 due to
interference with the ATP-binding site. Remarkably, 5-LO also
binds ATP,1 though this interaction has remained thus far
unexplored in terms of development of 5-LO inhibitors. In fact,
The presence of bromine in 7-position (2, also known as 7-
bromo-indirubin-3′-oxime, 7BIO), which increased the anti-
3716
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
Table 2. Inhibition of 5-LO Product Formation by IO Derivatives
a
b
Values are means SE, n = 3 or 4. *p < 0.05; **p < 0.01; ***p < 0.001 vs control; ANOVA + Bonferroni. Cell-free: isolated human recombinant
c
5-LO. Intact cells: intact human neutrophils stimulated with 2.5 μM A23187 ionophore.
tumoral effect of indirubins,25 improved 5-LO inhibition in
comparison to 1 in both cell-free and cellular test systems
(Table 1), which let us to further consider 7-bromo-indirubin
derivatives. N-Methylation in the 1-position is used as negative
control in kinase inhibition assays, because of a steric clash
between the N-methyl group and amino acid residues of the
ATP pocket.22 Importantly, the N-methylated compound 3 was
also inactive for 5-LO, highlighting a possible similar steric
hindrance on 5-LO. Replacement of the 3′-oxime by a keto
group (4), 3′-methoxyme (5), or acetoxyme (6) as well as
moving the oxime from the 3′- to 5′-position (7) was
detrimental, essentially indicating the requirement for the free
oxime moiety in the 3′-position and for the hydrogen in
position N1 for an improved inhibitory activity on 5-LO. In
addition, incorporation of a carboxylic acid in 6′-position (8, 9)
or of a methyl ester in the 5′-position (10) also did not
improve 5-LO inhibition in both cell-free and cellular assays.
Note that several known LT synthesis inhibitors of different
classes are lipophilic compounds with an acidic moiety,4 which
has been correlated to high plasma protein binding and, as
observed for example for boswellic acids,26 to ineffectiveness in
vivo. Moreover, those findings rather exclude an inverted mode
of interaction between indirubins and 5-LO, as 8, 9, and 10
were initially designed to interfere with the ATP cavity of
DYRKs via inverted binding.
5-LO has been discussed as possible drug target in cancer,
and therefore the compound 2 (7BIO) may be an interesting
lead in this respect. However, for further SAR studies, we aimed
to test whether the bromine in the 7-position was required or if
a different substitution pattern leading to less cytotoxic
compounds may be tolerated or even advantageous in terms
of 5-LO inhibition. Indeed, replacement of the bromine by
chlorine (11) or fluorine (12) was tolerated, while replacement
with trifluoromethyl (13) was detrimental in intact cells (Table
2). Note that incorporation of the bromine in the 5-position
(14, 5BIO) and especially in the 6-position (15, 6BIO) instead
of the 7-position increased the 5-LO inhibitory potency. These
data indicate the need of a halogen for high efficiency and
position 6 as preferred. Interestingly, insertion of the bromine
in the 6′-position (compound 16) instead of the 6-position was
detrimental for inhibition of 5-LO, but only in intact cells. Note
that compound 15 has been previously reported to reduce SH-
SY5Y cancer cells viability only at high concentrations (EC50
>
30 μM for compound 15 vs 8 μM for compound 2).25 Thus, 15
is more potent as 5-LO inhibitor than 2, but less cytotoxic.
As 15 (6BIO) has been found to be a potent GSK-3 inhibitor
(IC50 in cell-free assay, 5 nM),22 we then explored the
inhibitory potential on 5-LO of 6BIO analogues with improved
solubility, possessing hydroxy- (17−19) and/or amine-moieties
(e.g., pyrrolidine-, piperazine-, morpholino-, 20−24) (Table 3)
and known to be more potent on GSK-3 than the parent 15.14
However, these structural modifications resulted in a lower
inhibition as compared to compound 15, and significant effects
were observed only for the 1-glycerol oximether 18 and the
morpholino-N-ethyl oximether 24, suggesting a possible
difference in the size and shape of 5-LO binding site as
compared to the ATP cavity of GSK-3. Table 4 summarizes the
IC50 values of the most relevant derivatives for inhibition of
human recombinant 5-LO and 5-LO product formation in
human neutrophils. Compound 15 showed the same potency
in cell-free and cellular assays (IC50 = 1.5 μM), indicating an
improved potency of about 10-fold over that of parental
compound 1, and was equipotent to the reference compound
25 (IC50 = 0.6−3 μM, Table 4). The high potency of 15 was
also confirmed after addition of exogenous AA to the cellular
test system, which bypasses the need for cPLA2. Note that a
reduced potency after addition of exogenous AA was instead
observed for FLAP inhibitors.27 Hence, our findings suggest
that 15 is likely not targeting cPLA2 or FLAP (in the low
micromolar range) but instead shows features of a direct and
cell-permeable 5-LO inhibitor.
To evaluate interference with the putative ATP-binding site,
we first assessed whether 15 interferes with the binding of 5-LO
to ATP-agarose. As shown in Figure 1A, human recombinant 5-
LO was eluted from ATP-agarose column by incubation with
15, while the inactive N-methylated indirubin-derivative 3 was
ineffective. As expected, no elution was observed by 25, while
ATP eluted 5-LO. Next, we evaluated whether 5-LO inhibition
by 15 was affected by ATP. We observed a significant reduction
of both potency (right shift of inhibitory curve, higher IC50
3717
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
and its ineffectiveness as radical scavenger (Figure 2A)
essentially excluded classical non-redox and redox mechanisms
of 5-LO inhibition. Interestingly, 15 inhibited 5-LO in a
reversible fashion (Figure 2B) and showed efficiency also in
human monocytes (Figure 3), indicating that its suppressive
effect on LT synthesis is not restricted to only a certain cell
type. Human monocytes also express 15-LO, which is closely
related to the murine leukocyte-type 12-LO (also known as 12/
15-LO), and allow assessing inhibitory effects on 12- and 15-
HETE formation. However, 15 was not active in this respect
(Figure 3), indicating a selective effect on 5-LO. Note that 12/
15-LO fails to bind ATP.28 Importantly, 15 also inhibited 5-LO
product formation when neutrophils or monocytes were
stimulated with the pathophysiologically relevant stimuli
lipopolysaccharide (LPS) and N-formyl-methionyl-leucyl-phe-
nylalanine (fMLP) (IC50 = 1.5 0.7 and 0.5 0.2 μM for 5-
HETE reduction in neutrophils and monocytes, respectively)
(Supplementary Figure S2A). It should be noted that no
reduction in cell viability was observed after incubation with 15
(Supplementary Figure S2B), excluding possible toxic effects at
this level. We also tested whether 15 simultaneously targets
other enzymes of the AA cascade that have been previously
considered for dual inhibitory strategies. However, in cell-free
assays, 15 at 10 μM did not significantly inhibit COX-1 or
COX-2, and only partial inhibition was observed for mPGES-1
and sEH, while the positive controls indomethacin (indo, for
COXs), MK886 (for mPGES-1), and 12-(3-adamantane-1-yl-
ureido)-dodecanoic acid (AUDA, for sEH) inhibited as
anticipated (Figure 4).
Table 3. Inhibition of 5-LO Product Formation by 6-Bromo
Indirubin Derivatives
Together, 15 shows a novel and reversible mechanism of
interference with 5-LO, seemingly via the ATP-binding site,
with relevance in different cells and assay conditions, and
selectivity over other LOs. We propose that binding to the ATP
site may cause structural alteration of 5-LO that results in
reduced catalytic activity. For instance, changes in the relative
location/orientation of the regulatory C2 domain of 5-LO
toward the catalytic domain has been also proposed as
mechanism for the novel type 5-LO inhibitor hyperforin.16
Targeting the ATP-binding site represents a novel molecular
strategy for 5-LO inhibition without affecting other iron-
containing enzymes and proteins or fatty acid metabolizing
enzymes (including 12/15-LOs that do not bind ATP). In fact,
the development of 5-LO inhibitors acting by classical
mechanism of 5-LO inhibition (i.e., iron chelation, redox
interference, AA competition) has essentially been hampered
by severe side effects due to the reducing properties or to
unspecific iron-chelating features of the compounds. Also, in
vivo ineffectiveness because of rapid oxidation of reducing
agents or because of competition with fatty acids (or their
hydroperoxides, present in high amounts in plasma) for non-
redox “substrate analogues” have been observed.4 Compounds
acting at the putative ATP-binding site may present the
potential advantages to offer novel chemotypes with favorable
pharmacokinetic profile and may also be useful to gain insights
into structural information on the 5-LO ATP binding site,
which, as mentioned remains still to be characterized.
a
Values are means SE, n = 3 or 4. *p < 0.05; **p < 0.01; ***p <
b
0.001 vs control; ANOVA + Bonferroni. Cell-free: isolated human
recombinant 5-LO. Intact cells: intact human PMNL stimulated with
2.5 μM A23187 ionophore.
c
Table 4. IC50 Values for Inhibition of 5-LO Product
Formation by Indirubin Derivatives and Positive Control 25
a
5-LO activity (IC50 values)
c
intact cells
b
compd
cell-free
A23187
A23187/AA
1
15.0 2.1
5.8 0.8
4.9 0.6
3.7 0.4
5.6 0.4
1.5 0.1
0.6 0.1
11.1 1.3
1.7 0.5
4.9 0.7
5.0 0.8
2.6 0.6
1.5 0.5
1.7 0.7
8.2 1.5
3.6 0.9
3.1 0.3
3.3 1.7
3.3 0.8
1.5 0.3
2
11
12
14
15
25
3
1
a
The IC50 values were calculated after sigmoidal concentration−
response fitting (variable slope) by GraphPad Prism software. Values
b
are means SE, n = 3. Cell-free: isolated human recombinant 5-LO.
c
Intact cells: intact human neutrophils stimulated with 2.5 μM A23187
ionophore in the absence or presence of 20 μM AA.
values) and efficacy (lower magnitude of inhibition) by
increasing ATP concentrations in both the S40 fraction of E.
coli transformed with pT3−5LO (Figure 1B) and isolated 5-LO
(Figure 1C), while no significant differences were observed for
the iron ligand-type inhibitor 25 (Supplementary Figure S1).
Notably, the observation that the potency of 15 was not
impaired in the cell-free assay under non-reducing conditions
(Table 4), which is instead typical for non-redox inhibitors,4
IO derivatives have been previously reported to inhibit GSK-
3. Interestingly, 6-bromo substitution also resulted in an
improved inhibitory potency on GSK-3 in vitro (cell-free), with
IC50 values of 5 nM for compound 15 vs 22 nM for compound
1, while N-methylation inactivated the compounds.13 Since 5-
LO and GSK-3 are both relevant targets for an anti-
inflammatory therapy, we aimed to analyze whether 15 affects
3718
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
Figure 1. (A) Compound 15 elutes 5-LO bound to ATP. Results are representative for three comparable experiments. (B,C) 5-LO inhibition by 15
in (B) the S40 fraction of E. coli transformed with pT3−5LO or (C) isolated 5-LO is impaired by increasing ATP concentrations. Values are means +
SE, n = 3. 5-LO products in 100% controls correspond to (ng/mL): (B) S40 fraction, 538.1 90.3, 690.1 87.5, 722.1 108.3, and 27.9 8.8, for
incubations without (−) or containing 0.1, 1, and 10 mM ATP, respectively; (C) isolated 5-LO, 1147.5 50.1, 1090.0 40.2, and 177.9 28.7,
incubations containing 1, 2, and 10 mM ATP, respectively. *p < 0.05; **p < 0.01; ***p < 0.001 vs corresponding concentration (B) without ATP
(−) or (C) with 1 mM ATP; ANOVA + Bonferroni.
Figure 4. Effects of compound 15 on the activity of (A) COX-1, (B)
Figure 2. (A) Radical scavenging properties. Compound 15 (10 nmol)
or ascorbic acid (asc. acid, 5 nmol) was incubated with 5 nmol of
DPPH in 100 μL of EtOH for 30 min at RT, and the absorbance at
520 nm was measured. (B) Inhibition of 5-LO activity by compounds
15 and 25 is reversible. Purified 5-LO (about 10 μg/mL) was
incubated with 1 μM or 10 μM compound 15 or with 0.3 μM or 3 μM
compound 25, respectively, for 10 min at 4 °C. Aliquots of the 10 μM
compound 15 sample and of the 3 μM compound 25 sample were
diluted with assay buffer 10-fold (1(10) and 0.3(3), respectively),
while the concentration of the inhibitors in the other aliquots was not
altered (by 10-fold dilution with assay buffer containing the
corresponding concentrations of the inhibitors). Then, samples were
prewarmed for 30 s at 37 °C, and 20 μM AA and 2 mM CaCl2 were
added. After 10 min, 5-LO product formation was analyzed as
described. 5-LO products in 100% controls correspond to 799.1
147.4 ng/mL. Data are expressed as percentage of control (100%),
means + SE, n = 3.
COX-2, (C) mPGES-1, and (D) sEH. Compound 15 (10 μM) or
reference inhibitors (10 μM indomethacin, indo; 10 μM MK886; 100
nM AUDA) were added to the respective enzymes (A, B, D) or
microsomal enzyme preparation (C) 10 min prior starting the enzyme
reaction. Data are expressed as percentage of the remaining activity of
the uninhibited vehicle (0.1% v/v DMSO) control (100%). 100%
controls correspond to (A) 108.9 6.5 ng/mL 12-HHT; (B) 171.3
43 ng/mL 12-HHT; (C) 1.2 0.2 nmol PGE2 in 100 μL reaction
buffer; (D) Ex330, Em465: 28,724 8,022. Values are means SE, n =
3. *p < 0.05; **p < 0.01; ***p < 0.001 vs control; ANOVA +
Bonferroni.
major component of the β-catenin degradation complex and
plays a key inhibitory role in the Wnt pathway. In the cell,
GSK-3 phosphorylates β-catenin in a constitutive manner, thus
targeting it for ubiquitination and degradation by the
proteasome,29 which has been described also in monocytes.30
Accordingly, inhibition of GSK-3 induces stabilization of β-
catenin that is considered as a direct outcome and a
downstream activation of the canonical Wnt signaling pathway.
Thus, we performed reporter gene assays in HEK293
containing the Wnt/β-catenin activated reporter pSuperTOP-
FLASH (STF293 cells) as a read-out for inhibition of cellular
GSK-3 activity and observed significant effects by 15 at
concentrations of 0.5−1 μM (Figure 5A). Accordingly, 15 (1
Figure 3. Effects of compound 15 on the formation of products of 5-
LO (LTB4, LTB4 isomers, 5-HETE; red) and of 15-LO (12-HETE,
15-HETE; black) in human monocytes stimulated for 10 min at 37 °C
with A23187 plus AA (2.5 and 10 μM, respectively). The
preincubation with vehicle (0.1% v/v DMSO) or with compound 15
was performed for 10 min at 37 °C. Samples were analyzed by
UPLC−MS/MS, and data (peak area) were normalized on the internal
standard PGB1 and are expressed as percentage of control, means +
SE; n = 3.
Figure 5. Compound 15 inhibits cellular GSK-3 activity as analyzed by
(A) a Wnt reporter gene assay in STF293 cells and (B) by β-catenin
stabilization in human monocytes. Data are means + SD, n = 3,
duplicates.
cellular GSK-3 activity and downstream events at concen-
trations comparable to cellular 5-LO inhibition. GSK-3 is a
3719
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
μM) induced stabilization of β-catenin in human monocytes
(Figure 5B). This event was accompanied by a significant
reduction of pro-inflammatory cytokines (Table 5), with effects
in the high nanomolar to low micromolar range. Together, 15
is a dual inhibitor of 5-LO and GSK-3 with comparable potency
in intact cells.
ATP-binding site and showed selectivity over other AA-
metabolizing enzymes and high efficiency under different
stimulatory conditions. We classify this dual 5-LO/GSK-3
inhibition as a novel unconventional concept for anti-
inflammatory intervention. Though inhibition of GSK-3 has
been shown to attenuate asthma in mice,31 a possible benefit of
a dual 5-LO/GSK-3 inhibition in respiratory diseases is
unknown. On the other hand, this novel combination of
inhibitory activities may represent a promising strategy for
those diseases where both GSK-3 and 5-LO are involved, such
as Alzheimer’s disease, which could also result in new
therapeutic indications as compared to single approaches and
should be further investigated in future studies. However, as
GSK-3 inhibition causes up-regulation of the Wnt-pathway,
caution should be taken for patients e.g. with hyperproliferative
diseases where Wnt-signaling may play a crucial role.
a
Table 5. IC50 Values (μM) for Inhibition of Cytokine
Release by Compound 15
TNFα
IL-6
IL-8
>3
IL-1β
0.5 0.03
1.6 0.7
0.2 0.1
a
The IC50 values were calculated after sigmoidal concentration−
response fitting (variable slope) by GraphPad Prism software. Values
are means SE, n = 3. Intact human monocytes stimulated with 10
ng/mL LPS after preincubation for 30 min at 37 °C with vehicle (0.1%
DMSO) or compound 15. Cytokine amount in 100% (LPS-stimulated
and vehicle-treated) controls was (ng/mL) 1.7 0.3, 38.1 0.3, 24.6
2.9, and 20.9 3.1 for TNFα, IL-6, IL-8, and IL-1β, respectively.
EXPERIMENTAL SECTION
■
General Methods. The indirubin derivatives 2−6, 11 and 12 were
synthesized as previously described (see Scheme 1).32 Their structures
have been confirmed by NMR and MS and conformed to the
literature. Derivatives 7−10 and 13 have been synthesized as
previously described.24 Their structures have been confirmed by
NMR and MS and conformed to the literature. Compounds 14, 15,
and 16 were synthesized according to procedures described in the
literature.22,25,33 For the preparation of ethers 17−24, compound 15
was dissolved in anhydrous DMF under argon. Dibromoethane was
then added, and the mixture was stirred for 48 h at 50 °C. Addition of
water and subsequent filtration afforded the corresponding (2′Z-3′E)-
6-bromoindirubin-3′-[O-(2-bromoethyl)-oxime]. The latter was then
dissolved in anhydrous DMF under argon, and the corresponding
amine was added. The mixture was then submitted to microwave
CONCLUSIONS
■
We identified the 3,2′-bisindol core of the natural product
indirubin as a privileged and innovative chemotype for the
development of 5-LO inhibitors. Starting from indirubin-3′-
oxime (1) and considering several types of substitution, the 6-
bromine substitution turned out to be advantageous and led to
15 (6BIO), a GSK-3 inhibitor, with 10-fold improvement in the
potency (IC50 = 1.5 μM) being about equipotent to the only
approved 5-LO inhibitor zileuton (25). Compound 15 is a
direct and cell-permeable 5-LO inhibitor interfering with the
Scheme 1. General Synthesis of Indirubins Derivatives
3720
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
irradiation, in a CEM Single Mode microwave, at 90 °C (150 W) for
40 min to afford the corresponding ethers 17−24.14 The structures
were confirmed by NMR and MS and were in accordance with
literature. Condensation of correctly substituted isatin and 3-
acetoxyindoles afforded the indirubins in good yields (80−90%).
Oxime on the 3′-carbonyl has been quantitatively introduced by
reaction with hydroxylamine in refluxing pyridine.14 Derivatives 18 and
19 have been obtained in good yield (70−80%) by reaction of the
corresponding indirubin-3′-oxime with bromoethanol and 3-bromo-
1,2-propanediol, respectively. Derivatives 20−25 have been synthe-
sized in good yields (90−100%) in an already established two-step
procedure.14 The purity (>95%) of the compounds was determined by
combustion analysis.
the indicated compounds and lysed, and Firefly luciferase activities
were measured in duplicate as reported previously.38
Analysis of Cytokine Production in Monocytes. Monocytes
were incubated with vehicle (0.25% v/v DMSO) or compound 15 for
30 min at 37 °C and then left untreated or stimulated with LPS (10
ng/mL; 18 h). Cytokines were measured in the cell supernatants by
ELISA, according to the manufacturer’s instructions (R&D systems,
Minneapolis, Minnesota). The expression of β-catenin was evaluated
by Western blotting as described in the Supporting Information.
COX-1, COX-2, mPGES-1, and sEH Assays. Activities of isolated
ovine COX-1 and human COX-2, of human mPGES-1 from A549
cells, and of purified human sEH were analyzed as described in the
Supporting Information.
Statistics. Results are means
standard error (SE) or standard
Materials. Compound 1 was purchased from Enzo Life Sciences
deviation (SD) of n independent experiments. The IC50 values were
determined by sigmoidal concentration−response equation (Graph-
Pad Prism, San Diego, CA). Statistics was performed by repeated
measures one-way analysis of variance (ANOVA) followed by
Bonferroni, using a two-sided alpha level of 0.05 (*p < 0.05).
(Lorrach, Germany). Compound 25 was purchased from Sequoia
̈
Research Products (Oxford, U.K.). All other chemicals were purchased
from Sigma-Aldrich (Deisenhofen, Germany), unless stated otherwise.
HPLC and UPLC solvents were from VWR (Darmstadt, Germany).
Expression and Purification of 5-LO. 5-LO was expressed in E.
coli Bl21 (DE3) cells transformed with pT3−5LO,34 as indicated in
the Supporting Information. For ATP competition experiments, E. coli
cells were lysed, and the 40,000 × g supernatant (S40 fraction) was
preincubated with 0, 0.1, 1, or 10 mM ATP for 10 min a 4 °C. Then,
vehicle (0.1% v/v DMSO) or test compounds were added for 10 min
at 4 °C prior to addition of Ca2+ and AA (end concentrations, 1 mM
free Ca2+ and 20 μM AA) for 10 min at 37 °C. The reaction was
stopped, samples were extracted, and 5-LO products were analyzed by
HPLC as indicated in the Supporting Information. For ATP-agarose
binding experiments, the S40 fraction was incubated with ATP-agarose
(Sigma; Deisenhofen, Germany) for 10 min at 4 °C. After
centrifugation (1000 × g, 1 min, 4 °C), pellets were washed 3 times
with PBS containing 1 mM EDTA (PBS-EDTA) and then incubated
for 30 min at 4 °C with vehicle (0.1% DMSO) or test compounds, as
indicated. Then, samples were centrifuged (2000 × g, 5 min, 4 °C),
and pellets and supernatants were separated. Supernatants (eluates)
were centrifuged again and then mixed with SDS-PAGE sample
loading buffer. Pellets (ATP-agarose) were washed with PBS-EDTA
and incubated with SDS loading buffer to release bound 5-LO.
Samples were then analyzed by Western blotting as described in the
Supporting Information. Alternatively, 5-LO was partially purified by
ATP-agarose column for activity assays, as previously reported.34,35
After washing of the column (bed volume, 1 mL) with PBS-EDTA, 5-
LO enzyme was eluted with 10 mL of PBS-EDTA buffer containing 22
mM ATP (resulting in an eluate containing 20 mM ATP). The eluate
was diluted 1:25 to 1:100 with PBS-EDTA buffer for activity test, and
ATP concentration was accordingly adjusted to 1 mM, as standard
conditions.34,35 For ATP competition experiments on isolated 5-LO,
ATP concentrations were adjusted to 1, 2, and 10 mM ATP, as
indicated. Samples (0.5−2 μg of partially purified 5-LO, resulting in
about 750 ng/mL 5-LO products in an activity test performed with 20
μM AA) were incubated 10 min at 4 °C with vehicle or test
compounds, prewarmed for 30 s at 37 °C, and 2 mM CaCl2 and the
indicated concentrations of AA were added. The reaction was stopped,
and 5-LO products were analyzed as indicated in the Supporting
Information.
ASSOCIATED CONTENT
* Supporting Information
Supporting methods and figures S1 and S2. This material is
■
S
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
*Phone: +49-(0)3641-949811. E-mail: pergolacarlo@
googlemail.com.
■
*Phone: +49-(0)3641-949801. E-mail: oliver.werz@uni-jena.
de.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Harald Zimmer, Petra Wiecha, Monika Listing, and
Anna Lazari for assistance. This work was supported by the
Commission of the European Community through the
INsPiRE project (EU-FP7-REGPOT-2011-1, proposal
284460).
ABBREVIATIONS USED
■
AA, arachidonic acid; AUDA, 12-(3-adamantane-1-yl-ureido)-
dodecanoic acid; BIO, bromoindirubin-3′-oxime; COX, cyclo-
oxygenase; cPLA2, cytosolic phospholipase A2; DYRK, dual-
specificity tyrosine-regulated kinases; FLAP, 5-lipoxygenase-
activating protein; fMLP, N-formyl-methionyl-leucyl-phenyl-
alanine; GSK-3, glycogen synthase kinase-3; 5-HpETE, 5(S)-
hydroperoxy-6-trans-8,11,14-cis-eicosatetraenoic acid; IO, indir-
ubin-3′-oxime; 5-LO, 5-lipoxygenase; LT, leukotriene; mPGES-
1, microsomal prostaglandin E2 synthase-1; sEH, soluble
epoxide hydrolase
Determination of 5-LO Product Synthesis. Isolation and
stimulation of neutrophils and monocytes and determination of 5-
LO product formation were performed as described36 and as indicated
in the Supporting Information. Incubations were performed with 5 ×
106 and 2 × 106 cells/mL for neutrophils and monocytes, respectively.
Analysis of LO products was performed by HPLC or by UPLC−MS/
MS. For the values given in Tables 1−4, 5-LO products in vehicle
REFERENCES
■
(1) Radmark, O.; Werz, O.; Steinhilber, D.; Samuelsson, B. 5-
̊
Lipoxygenase: regulation of expression and enzyme activity. Trends
Biochem. Sci. 2007, 32, 332−341.
(DMSO) controls were cell-free, 787.2
61.5 ng/mL; intact cells,
(2) Peters-Golden, M.; Henderson, W. R., Jr. Leukotrienes. N. Engl. J.
Med. 2007, 357, 1841−1854.
236.5 101.8 and 514.8 48.8 ng/mL for A23187 and A23187 + AA
stimulated neutrophils, respectively.
Reporter Gene Assay. HEK-293 cells carrying a stably integrated
luciferase reporter construct under the control of 7 LEF/TCF binding
sites (STF293 cells,37 provided by Dr. Jeremy Nathans, The Johns
Hopkins University, Baltimore) were treated for 24 h with vehicle or
(3) Di Gennaro, A.; Haeggstrom, J. Z. Targeting leukotriene B in
inflammation. Expert Opin. Ther. Targets 2013, 18, 79−93.
(4) Pergola, C.; Werz, O. 5-Lipoxygenase inhibitors: a review of
recent developments and patents. Expert Opin. Ther. Pat 2010, 20,
355−375.
3721
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
(5) Werz, O.; Greiner, C.; Koeberle, A.; Hoernig, C.; George, S.;
Popescu, L.; Syha, I.; Schubert-Zsilavecz, M.; Steinhilber, D. Novel and
potent inhibitors of 5-lipoxygenase product synthesis based on the
structure of pirinixic acid. J. Med. Chem. 2008, 51, 5449−5453.
(6) Hofmann, B.; Franke, L.; Proschak, E.; Tanrikulu, Y.; Schneider,
P.; Steinhilber, D.; Schneider, G. Scaffold-hopping cascade yields
potent inhibitors of 5-lipoxygenase. ChemMedChem. 2008, 3, 1535−
1538.
sites of human 5-lipoxygenase using photoaffinity labeling and enzyme
immobilization. Biochemistry 1995, 34, 13603−13611.
(22) Polychronopoulos, P.; Magiatis, P.; Skaltsounis, A.-L.;
Myrianthopoulos, V.; Mikros, E.; Tarricone, A.; Musacchio, A.; Roe,
S. M.; Pearl, L.; Leost, M.; Greengard, P.; Meijer, L. Structural basis for
the synthesis of indirubins as potent and selective inhibitors of
glycogen synthase kinase-3 and cyclin-dependent kinases. J. Med.
Chem. 2004, 47, 935−946.
(7) Reddy, N. P.; Chandramohan Reddy, T.; Aparoy, P.; Achari, C.;
Sridhar, P. R.; Reddanna, P. Structure based drug design, synthesis and
evaluation of 4-(benzyloxy)-1-phenylbut-2-yn-1-ol derivatives as 5-
lipoxygenase inhibitors. Eur. J. Med. Chem. 2012, 47, 351−359.
(8) Wu, Y.; He, C.; Gao, Y.; He, S.; Liu, Y.; Lai, L. Dynamic modeling
of human 5-lipoxygenase-inhibitor interactions helps to discover novel
inhibitors. J. Med. Chem. 2012, 55, 2597−2605.
(9) Karg, E. M.; Luderer, S.; Pergola, C.; Buhring, U.; Rossi, A.;
Northoff, H.; Sautebin, L.; Troschutz, R.; Werz, O. Structural
optimization and biological evaluation of 2-substituted 5-hydrox-
yindole-3-carboxylates as potent inhibitors of human 5-lipoxygenase. J.
Med. Chem. 2009, 52, 3474−3483.
(23) Myrianthopoulos, V.; Magiatis, P.; Ferandin, Y.; Skaltsounis, A.
L.; Meijer, L.; Mikros, E. An integrated computational approach to the
phenomenon of potent and selective inhibition of aurora kinases B and
C by a series of 7-substituted indirubins. J. Med. Chem. 2007, 50,
4027−4037.
(24) Myrianthopoulos, V.; Kritsanida, M.; Gaboriaud-Kolar, N.;
Magiatis, P.; Ferandin, Y.; Durieu, E.; Lozach, O.; Cappel, D.;
Soundararajan, M.; Filippakopoulos, P.; Sherman, W.; Knapp, S.;
Meijer, L.; Mikros, E.; Skaltsounis, A.-L. Novel inverse binding mode
of indirubin derivatives yields improved selectivity for DYRK kinases.
ACS Med. Chem. Lett. 2013, 4, 22−26.
(25) Ribas, J.; Bettayeb, K.; Ferandin, Y.; Knockaert, M.; Garrofe-
Ochoa, X.; Totzke, F.; Schachtele, C.; Mester, J.; Polychronopoulos,
P.; Magiatis, P.; Skaltsounis, A. L.; Boix, J.; Meijer, L. 7-
Bromoindirubin-3′-oxime induces caspase-independent cell death.
Oncogene 2006, 25, 6304−6318.
(10) Laufer, S. A.; Augustin, J.; Dannhardt, G.; Kiefer, W. 6,7-
Diaryldihydropyrrolizin-5-yl)acetic acids, a novel class of potent dual
inhibitors of both cyclooxygenase and 5-lipoxygenase. J. Med. Chem.
1994, 37, 1894−1897.
(11) Koeberle, A.; Zettl, H.; Greiner, C.; Wurglics, M.; Schubert-
Zsilavecz, M.; Werz, O. Pirinixic acid derivatives as novel dual
inhibitors of microsomal prostaglandin E2 synthase-1 and 5-
lipoxygenase. J. Med. Chem. 2008, 51, 8068−8076.
(12) Meirer, K.; Rodl, C. B.; Wisniewska, J. M.; George, S.; Hafner,
A. K.; Buscato, E. L.; Klingler, F. M.; Hahn, S.; Berressem, D.;
Wittmann, S. K.; Steinhilber, D.; Hofmann, B.; Proschak, E. Synthesis
and structure-activity relationship studies of novel dual inhibitors of
soluble epoxide hydrolase and 5-lipoxygenase. J. Med. Chem. 2013, 56,
1777−1781.
(13) Meijer, L.; Skaltsounis, A. L.; Magiatis, P.; Polychronopoulos, P.;
Knockaert, M.; Leost, M.; Ryan, X. P.; Vonica, C. A.; Brivanlou, A.;
Dajani, R.; Crovace, C.; Tarricone, C.; Musacchio, A.; Roe, S. M.;
Pearl, L.; Greengard, P. GSK-3-selective inhibitors derived from Tyrian
purple indirubins. Chem. Biol. 2003, 10, 1255−1266.
(26) Siemoneit, U.; Pergola, C.; Jazzar, B.; Northoff, H.; Skarke, C.;
Jauch, J.; Werz, O. On the interference of boswellic acids with 5-
lipoxygenase: mechanistic studies in vitro and pharmacological
relevance. Eur. J. Pharmacol. 2009, 606, 246−254.
(27) Fischer, L.; Hornig, M.; Pergola, C.; Meindl, N.; Franke, L.;
Tanrikulu, Y.; Dodt, G.; Schneider, G.; Steinhilber, D.; Werz, O. The
molecular mechanism of the inhibition by licofelone of the
biosynthesis of 5-lipoxygenase products. Br. J. Pharmacol. 2007, 152,
471−480.
(28) Van der Donk, E. M.; Dubois, G. R.; Verhagen, J.; Veldink, G.
A.; Vliegenthart, J. F. Improved purification of 12-lipoxygenase from
rat basophilic leukemia cells and conditions for optimal enzyme
activity. Biochim. Biophys. Acta 1991, 1074, 443−447.
(29) Doble, B. W.; Woodgett, J. R. GSK-3: tricks of the trade for a
multi-tasking kinase. J. Cell Sci. 2003, 116, 1175−1186.
(14) Vougogiannopoulou, K.; Ferandin, Y.; Bettayeb, K.;
Myrianthopoulos, V.; Lozach, O.; Fan, Y.; Johnson, C. H.; Magiatis,
P.; Skaltsounis, A. L.; Mikros, E.; Meijer, L. Soluble 3′,6-substituted
indirubins with enhanced selectivity toward glycogen synthase kinase-3
alter circadian period. J. Med. Chem. 2008, 51, 6421−6431.
(15) Werz, O. Inhibition of 5-lipoxygenase product synthesis by
natural compounds of plant origin. Planta Med. 2007, 73, 1331−1357.
(16) Feisst, C.; Pergola, C.; Rakonjac, M.; Rossi, A.; Koeberle, A.;
Dodt, G.; Hoffmann, M.; Hoernig, C.; Fischer, L.; Steinhilber, D.;
(30) Martin, M.; Rehani, K.; Jope, R. S.; Michalek, S. M. Toll-like
receptor-mediated cytokine production is differentially regulated by
glycogen synthase kinase 3. Nat. Immunol 2005, 6, 777−784.
(31) Bao, Z.; Lim, S.; Liao, W.; Lin, Y.; Thiemermann, C.; Leung, B.
P.; Wong, W. S. Glycogen synthase kinase-3beta inhibition attenuates
asthma in mice. Am. J. Respir. Crit. Care Med. 2007, 176, 431−438.
(32) Ferandin, Y.; Bettayeb, K.; Kritsanida, M.; Lozach, O.;
Polychronopoulos, P.; Magiatis, P.; Skaltsounis, A.-L.; Meijer, L. 3′-
Substituted 7-halogenoindirubins, a new class of cell death inducing
agents. J. Med. Chem. 2006, 49, 4638−4649.
Franke, L.; Schneider, G.; Radmark, O.; Sautebin, L.; Werz, O.
̊
Hyperforin is a novel type of 5-lipoxygenase inhibitor with high
efficacy in vivo. Cell. Mol. Life Sci. 2009, 66, 2759−2771.
(33) Hoessel, R.; Leclerc, S.; Endicott, J. A.; Nobel, M. E.; Lawrie, A.;
Tunnah, P.; Leost, M.; Damiens, E.; Marie, D.; Marko, D.;
Niederberger, E.; Tang, W.; Eisenbrand, G.; Meijer, L. Indirubin, the
active constituent of a Chinese antileukaemia medicine, inhibits cyclin-
dependent kinases. Nat. Cell Biol. 1999, 1, 60−67.
(17) Pergola, C.; Jazzar, B.; Rossi, A.; Northoff, H.; Hamburger, M.;
Sautebin, L.; Werz, O. On the inhibition of 5-lipoxygenase product
formation by tryptanthrin: mechanistic studies and efficacy in vivo. Br.
J. Pharmacol. 2012, 165, 765−776.
(34) Fischer, L.; Szellas, D.; Radmark, O.; Steinhilber, D.; Werz, O.
̊
(18) Blazevic, T.; Schaible, A. M.; Weinhaupl, K.; Schachner, D.;
Nikels, F.; Weinigel, C.; Barz, D.; Atanasov, A. G.; Pergola, C.; Werz,
O.; Dirsch, V. M.; Heiss, E. H. Indirubin-3′-monoxime exerts a dual
mode of inhibition towards leukotriene-mediated vascular smooth
muscle cell migration. Cardiovasc. Res. 2014, 101, 522−532.
(19) Vougogiannopoulou, K.; Skaltsounis, A. L. From tyrian purple
to kinase modulators: Naturally halogenated indirubins and synthetic
analogues. Planta Med. 2012, 78, 1515−1528.
Phosphorylation- and stimulus-dependent inhibition of cellular 5-
lipoxygenase activity by nonredox-type inhibitors. FASEB J. 2003, 17,
949−951.
(35) Brungs, M.; Radmark, O.; Samuelsson, B.; Steinhilber, D.
̊
Sequential induction of 5-lipoxygenase gene expression and activity in
Mono Mac 6 cells by transforming growth factor beta and 1,25-
dihydroxyvitamin D3. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 107−111.
(36) Schaible, A. M.; Koeberle, A.; Northoff, H.; Lawrenz, B.;
Weinigel, C.; Barz, D.; Werz, O.; Pergola, C. High capacity for
leukotriene biosynthesis in peripheral blood during pregnancy.
Prostaglandins, Leukotrienes Essent. Fatty Acids 2013, 89, 245−255.
(37) Xu, Q.; Wang, Y.; Dabdoub, A.; Smallwood, P. M.; Williams, J.;
Woods, C.; Kelley, M. W.; Jiang, L.; Tasman, W.; Zhang, K.; Nathans,
J. Vascular development in the retina and inner ear: control by Norrin
(20) Zhang, Y. Y.; Hammarberg, T.; Radmark, O.; Samuelsson, B.;
̊
Ng, C. F.; Funk, C. D.; Loscalzo, J. Analysis of a nucleotide-binding
site of 5-lipoxygenase by affinity labelling: binding characteristics and
amino acid sequences. Biochem. J. 2000, 351, 697−707.
(21) Falgueyret, J. P.; Denis, D.; Macdonald, D.; Hutchinson, J. H.;
Riendeau, D. Characterization of the arachidonate and ATP binding
3722
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723

Journal of Medicinal Chemistry
Article
and Frizzled-4, a high-affinity ligand-receptor pair. Cell 2004, 116,
883−895.
(38) Weiske, J.; Albring, K. F.; Huber, O. The tumor suppressor Fhit
acts as a repressor of beta-catenin transcriptional activity. Proc. Natl.
Acad. Sci. U.S.A. 2007, 104, 20344−20349.
3723
dx.doi.org/10.1021/jm401740w | J. Med. Chem. 2014, 57, 3715−3723