LASSBio-1829 Hydrochloride: Development of a New Orally Active N-Acylhydrazone IKK2 Inhibitor with Anti-inflammatory Properties

Article abstract of DOI:10.1002/cmdc.201500266

Inhibitor of nuclear factor κB kinase 2 (IKK2) is suggested to be a potential target for the development of novel anti-inflammatory and anticancer drugs. In this work, we applied structure-based drug design to improve the potency of the inhibitor (E)-N′-(4-nitrobenzylidene)-2-naphthohydrazide (LASSBio-1524, 1 a: IC₅₀=20 μm). The molecular model built for IKK2 together with the docking methodology employed were able to provide important and consistent information with respect to the structural and chemical inhibitor characteristics that may confer potency to IKK2 inhibitors, providing important guidelines for the development of a new N-acylhydrazone (NAH) derivative. (E)-N′-(4-(1H-pyrrolo[2,3-b]pyridin-4-yl)benzylidene)-2-naphthohydrazide hydrochloride (LASSBio-1829 hydrochloride, 10) is a 7-azaindole NAH able to inhibit IKK2 with an IC₅₀ value of 3.8 μm. LASSBio-1829 hydrochloride was found to be active in several pharmacological inflammation tests in vivo, showing its potential as an anti-inflammatory prototype. Quelling the kinase: IKK2 has been considered a good target for the design of novel drug prototypes to treat chronic inflammatory diseases. Herein we report the successful use of a structure-based drug design strategy for the development of LASSBio-1829 hydrochloride (10), an IKK2 inhibitor (IC₅₀=3.8 μm) and a promising anti-inflammatory prototype.

Full text of DOI:10.1002/cmdc.201500266

DOI: 10.1002/cmdc.201500266
Full Papers
LASSBio-1829 Hydrochloride: Development of a New
Orally Active N-Acylhydrazone IKK2 Inhibitor with Anti-
inflammatory Properties
Isabella A. Guedes,[b] Rosana H. C. N. Freitas,[a] Natµlia M. Cordeiro,[c] Thaís S.
do Nascimento,[c] Tayna S. Valerio,[c] Patrícia D. Fernandes,[c] Laurent E. Dardenne,[b] and
Carlos A. M. Fraga*[a]
Inhibitor of nuclear factor kB kinase 2 (IKK2) is suggested to be
a potential target for the development of novel anti-inflamma-
tory and anticancer drugs. In this work, we applied structure-
based drug design to improve the potency of the inhibitor (E)-
N’-(4-nitrobenzylidene)-2-naphthohydrazide (LASSBio-1524, 1a:
IC50 =20 mm). The molecular model built for IKK2 together with
the docking methodology employed were able to provide im-
portant and consistent information with respect to the struc-
tural and chemical inhibitor characteristics that may confer po-
tency to IKK2 inhibitors, providing important guidelines for the
development of a new N-acylhydrazone (NAH) derivative. (E)-
N’-(4-(1H-pyrrolo[2,3-b]pyridin-4-yl)benzylidene)-2-naphthohy-
drazide hydrochloride (LASSBio-1829 hydrochloride, 10) is a 7-
azaindole NAH able to inhibit IKK2 with an IC50 value of 3.8 mm.
LASSBio-1829 hydrochloride was found to be active in several
pharmacological inflammation tests in vivo, showing its poten-
tial as an anti-inflammatory prototype.
Introduction
The IKK complex is formed by two catalytic subunits (the
serine/threonine kinases, IKK1 and IKK2) and one regulatory
subunit, IKKg (or NEMO).[1–3] This complex is involved in the ac-
tivation of NF-kB in response to apoptotic and inflammatory
processes.[4–6] Inhibition of the IKK complex is considered to be
a promising approach for the treatment of chronic inflamma-
tion and cancer.[7–13]
ed with the dimers of NF-kB in the cytosol. IkB is polyubiquiti-
nated and degraded, releasing the NF-kB dimers. These dimers
migrate into the cell nucleus, where they can exert their func-
tion as transcription factors assisting in the biosynthesis of
pro-inflammatory cytokines.[14] Activation of the biochemical
cascades of NF-kB is associated with the regulation of the
innate and adaptive immune responses, among other physio-
logical responses;[15,16] however, continuous stimulation of NF-
kB is directly related with some types of inflammatory and au-
toimmune diseases, and also with the genesis and survival of
tumor cells.[17,18] In this context, IKK2 has been suggested to be
a potential target for the development of novel candidates for
anti-inflammatory and cancer drugs.[19] The development of se-
lective IKK2 inhibitors is necessary because the inhibition of
IKK1 may promote severe side effects, especially within the ep-
idermis.[14,19–21]
The IKK2 subunit participates in the activation of nuclear
factor kB (NF-kB), which is essential for the full activation of
this transcription factor. When a cell receives pro-inflammatory
stimuli such as lipopolysaccharide (LPS) and tumor necrosis
factor-a (TNF-a), IKK2 is activated by phosphorylation. Then,
IKK2 phosphorylates the inhibitor of kB (IkB), which is associat-
[a] R. H. C. N. Freitas,+ Prof. C. A. M. Fraga
Laboratório de Avaliażo e Substâncias Bioativas (LASSBio)
Instituto de CiÞncias BiomØdicas (ICB)
The IKK2 subunit consists of three domains: the kinase, the
ubiquitin-like and the dimerization domains.[3] Nevertheless,
only the kinase domain (residues 15–300) was modeled be-
cause the main focus of this work is the development of inhibi-
tors that interact with the ATP binding site, which is located in
this catalytic domain.
Universidade Federal do Rio de Janeiro (UFRJ)
21941-902, Rio de Janeiro, RJ (Brazil)
E-mail: cmfraga@ccsdecania.ufrj.br
[b] I. A. Guedes,+ Prof. L. E. Dardenne
Laboratório Nacional de Computażo Científica (LNCC/MCTI)
Petrópolis, RJ (Brazil)
[c] N. M. Cordeiro,+ T. S. d. . Nascimento, T. S. Valerio, Prof. P. D. Fernandes
Laboratório de Farmacologia da Dor e da Inflamażo
ICB, UFRJ, Rio de Janeiro, RJ (Brazil)
Currently, some IKK2 structures deposited in the Protein
Data Bank (PDB codes 3RZF,[22] 3QA8,[22] 4E3C[23] and 4KIK[24]
)
have been solved at resolutions ranging from 2.8 to 4.0 Š. De-
spite their crucial role in elucidating some important structural
and functional properties of IKK2, these structures do not have
sufficient quality in the ATP binding site to be applied in the
structure-based drug design of IKK2 inhibitors. The structure
with the best resolution (PDB code 4KIK,[24] solved at 2.83 Š)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201500266.
This article is part of a Special Issue on Drug Discovery in the Field of
Autoimmune and Inflammatory Disorders. The complete issue can be
browsed via Wiley Online Library.
ChemMedChem 2016, 11, 234 – 244
234
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
has some disordered regions in the ATP binding site (e.g., G-
loop), which might influence the docking results because this
region is directly in contact with ATP and some inhibitors. For
these reasons, the molecular model of the IKK2 isoform was
constructed through a comparative modeling methodology
based on multiple templates. The constructed model was vali-
dated through diverse energetic and geometric tools and also
with the docking of ATP and other inhibitors from the litera-
ture.
Results and Discussion
Structural modeling of IKK2
The geometric analyses performed with MolProbity indicated
that the constructed model of IKK2 (Figure 1) presented no
outlier residues, and the QMEAN and ProSA results showed
satisfactory energetic and geometric quality (Supporting Infor-
mation Figure S1).
As previously published and discussed,[22,24] the IKK2 model
shares a common folding pattern with other serine/threonine
kinases: 1) a region rich in b-sheets (N terminus), and 2) a lobe
rich in a-helices (C terminus), connected by 3) a hinge loop.
The ATP binding site is located between the two lobes and is
mainly formed by the hinge, G-loop, and A-loop regions
(Figure 1).
Previous works from our research group reported the dis-
covery of the N-acylhydrazone (NAH) derivative LASSBio-1524
(1a), a selective IKK2 inhibitor with IC50 =20 mm and active in
pharmacological models of acute inflammation in vivo.[25] In
this study, the analogous compound 1b is reported, which has
a 4-carboxyphenyl ring, differing only at this point compared
with LASSBio-1524, which has a 4-nitrophenyl ring. This modifi-
The hinge is formed by two hydrogen bond acceptors (the
oxygen atoms from the main chain of Glu97 and Cys99) and
a hydrogen bond donor (the nitrogen atom from main chain
of Cys99) (Figure 1A). The gatekeeper residue in IKK2 is a me-
thionine (Met96), which is located between the end of the N-
terminal lobe and the beginning of the hinge. Given the struc-
tural similarity between the members of the kinase family, this
residue may undergo a conformational change in a similar
fashion to the JNK3 enzyme, providing access to an allosteric
binding site;[26] however, further investigations are needed to
confirm this hypothesis.
cation was deleterious, resulting in the inactivity of compound
1b; however, the previous docking studies were not able to
propose an explanation regarding this difference in activity.[25]
In this work, we constructed a novel molecular model for
the IKK2 kinase domain and carried out docking of compound
1b, LASSBio-1524 (1a) and a set of diverse inhibitors from the
literature to investigate the predicted binding properties that
could justify their differential potency and selectivity. The aim
of this work is to introduce structural modifications in the pro-
totype of the NAH series, LASSBio-1524 (1a), guided by the
docking results and in vitro ex-
The G-loop, also known as the P-loop (phosphate binding
loop), is a glycine-rich sequence followed by a conserved
lysine and a serine/threonine and is located between the b1
and the b2 sheets. This loop acts as a lid of the ATP binding
site, supporting phosphate transfer.[27] In IKK2, the sequence of
the G-loop consists of Gly24-Gly25-Phe26-Gly27.
The DFG sequence (Asp-Phe-Gly triad) is located in the A-
loop and is highly conserved between the members of the
kinase family; however, the IKK2 model has a DLG motif
(formed by Asp-Leu-Gly residues) (Figure 1B). This finding is in
periments to obtain more
potent and selective IKK2 inhibi-
tors.
The structure-based drug
design applied in our work led
us to the development of LASS-
Bio-1829 hydrochloride (10,
IC50 =3.8 mm),
a more potent
compound than the prototype
of the NAH series LASSBio-1524
(1a, IC50 =20 mm). LASSBio-1829
hydrochloride (10) was active
against inflammation in several
pharmacological tests in vivo,
significantly reducing TNF-a and
reactive oxygen species (ROS)
biosynthesis, corroborating its
potential as an anti-inflammatory
prototype.
Figure 1. IKK2 molecular model. The regions of the hinge, A-loop, and G-loop are highlighted in red. A) Hinge
region with the key amino acids Glu97 and Cys99, which form hydrogen bonds with the adenine ring of ATP and
Met96. B) DLG triad residues (Asp166, Leu167, and Gly168), and the salt bridge between Lys44 and Glu61.
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
235
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
accordance with the amino acid
sequence and with the 3D struc-
tures of IKK2, where the DLG
motif is also present. The change
of a phenylalanine to a leucine
residue might consist of an im-
portant point of selectivity for
the IKKs among other kinases.
The conserved pair of Lys44
(located in the b-sheet from the
N-lobe) and Glu61 (located in
the a-helix in the bottom of the
binding site) forms a salt bridge
in the IKK2 model, which is a fea-
ture of the active conformation
of
protein
kinases
(Fig-
ure 1B).[28,29]
Validating the IKK2 model
through the docking of ATP
The IKK2-ATP docking generated
three top-ranked conformations
with similar docking scores rang-
Figure 2. Predicted binding modes of A) LASSBio-1524 and B) compound 1b found by docking experiments with
IKK2 in the absence of water and metal ions. The aromatic side chains that interact with the inhibitor are repre-
sented as dots by the atomic van der Waals radii. The 2D ligand interaction diagrams generated with Maestro are
shown at right; note: the residue position number labels on the 2D ligand interaction diagrams are shifted by
nÀ14 relative to residue numbering at left and in the text.
ing
from
À16.511
to
À16.040 kcalmolÀ1
.
These con-
formations are very close with
the experimental structure of
the CDK2–ATP complex (PDB
code 1HCK[30], see Supporting In-
formation Figure S2). The adenosine moiety interacts through
two hydrogen bonds with the main chains of Glu97 and Cys99
of the hinge. These interactions are well conserved in the
family of protein kinases.[31] The ATP triphosphate group inter-
acts with the Mg2+ and the main chain of the residues from
the G-loop, whereas the ribose ring generally interacts with
Asp103 through hydrogen bonds.
a docking score of À4.27 kcalmolÀ1 and exhibits the nitro
group oriented toward the hinge region, the naphthyl group is
located between the Mg2+ binding site and the G-loop, and
the NAH group forms a hydrogen bond with Asp166. The
naphthyl group interacts with Phe26 from the G-loop and
Trp58 through stacking interactions. The predicted binding
mode of compound 1b (Figure 2B) has a docking score of
À2.98 kcalmolÀ1 and exhibits only a hydrogen bond between
the carboxylate group and the hinge region (main chain NH of
Cys99). Despite the overall similarity between the predicted
binding modes of compound 1b and LASSBio-1524, the
former is not hydrogen bonded with Asp166 and loses the
stacking interactions with Phe26 and Trp58.
Molecular docking of inhibitors in IKK2
According to the docking results, a predominance of hydrogen
bonds is observed between the inhibitors of IKK2 extracted
from the literature and the hinge region, which is in accord-
ance with the reported results for the ATP-competitive inhibi-
tors of diverse protein kinases (Supporting Information Fig-
ure S3).[32–34] In some cases, the inhibitors are able to interact
through two or three hydrogen bonds with the hinge, most
frequently with Glu97 and Cys99. The most potent inhibitors
were observed to interact through two hydrogen bonds with
the main chain of Cys99, including for the most potent inhibi-
tor tested (compound 16, IC50 =11 nm, see page S9 in Support-
ing Information). Potent inhibitors also usually interacted with
other important and conserved residues in the ATP binding
site, e.g., hydrogen bonds with Lys44 (from the b-sheet) and
Asp166 (from the DLG motif).
The docking results guided us toward the development of
the new NAH derivative by the positions of overlap between
LASSBio-1524 (1a) and a potent and selective IKK2 inhibitor 2
(IC50 =66 nm, compound 3[35] in Table S2, Supporting Informa-
tion), where it was observed that the ligands occupy and inter-
act with different regions in the active site of the target
enzyme (Figure 3). Thereby, LASSBio-1829 (9, Figure 4) was de-
signed from a molecular hybridization between LASSBio-1524
(1a) and 2 as the optimization strategy. We kept the NAH sub-
unit on account of it being a privileged structure,[36] and thus it
may contribute to the inhibitory activity. In subunit B, the bio-
isosteric exchange of the 2-aminopyrimidine ring by 7-azain-
dole (subunit B, Figure 4) was performed, once the azabicyclic
system possessed the structural requirements needed to form
The top-ranked poses of LASSBio-1524 (1a) and compound
1b are shown in Figure 2. LASSBio-1524 (1a, Figure 2A) has
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
236
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
In addition to the observed structural quality of the con-
structed IKK2 model, the success of the docking experiment in-
volving the ATP molecule (i.e., finding a very similar structure
with the experimental one from the CDK2-ATP complex) is an
important result to validate the docking protocol (structural
quality of the model and the parameters of the docking pro-
gram) used to predict the binding modes of the compounds
investigated in this work.
The predicted binding modes of LASSBio-1524 (1a) and
compound 1b (Figure 2) share a similar orientation inside of
the IKK2 binding side exhibiting the nitro and carboxylate
groups oriented toward the hinge region. Unlike the most
potent IKK2 inhibitors, LASSBio-1524 does not interact directly
with the hinge region via a hydrogen bond (4.7 Š between the
nitrogen atom of the nitro group and the oxygen of the main
chain of Glu97, and 5.7 Š between the nitro group and the ni-
trogen atom of the main chain of Cys99), which may imply
a low potency of this compound. This important result shows
a way to modify this compound to improve its potency. On
the other hand, the carboxyl group from compound 1b is pre-
dicted to be closer and hydrogen bonded to the
hinge; however, this negatively charged group is lo-
cated in a hydrophobic region inside of the IKK2
binding site. Furthermore, the NAH group does not
interact through hydrogen bonds with Lys44 or
Asp166. These features might be reflected in the
poor predicted score of this compound (À2.98 kcal
molÀ1) and are probably associated with the fact that
this compound is inactive against IKK2.
Figure 3. Overlap of the top-ranked docking pose of LASSBio-1524 (carbon
atoms in green) and inhibitor 2[34] (carbon atoms in light blue) in the ATP
binding site of IKK2. The hydrogen bonds formed by inhibitor 2 are shown
as yellow dashes.
The most favorable docking score of the neutral
LASSBio-1829 (9) (À9.72 kcalmolÀ1), associated with
the important interactions between 1) the 7-azain-
dole group and the hinge region, 2) the NAH group
with the conserved Asp166 and Lys44, and 3) the
stacking interactions with Phe26 from the G-loop
Figure 4. Design of LASSBio-1829 (9) from the molecular hybridization between LASSBio-
1524 (1a) and compound 2.[35]
(Figure 5), suggest that this compound may be more
potent than the prototype of the series LASSBio-1524
(1a).
the same interactions that the 2-aminopyrimidine subunit pos-
Chemistry
sessed in the target enzyme, acting as hydrogen bond accept-
or and donor. Additionally, subunit C (Figure 4) was removed
for synthetic ease.
The synthesis of LASSBio-1829 (9) was carried out as outlined
in Scheme 1. The starting material, 2-naphthoyl chloride 3, was
treated with methanol at room temperature to give the corre-
sponding methyl ester 4. Next, the ester 4 was reacted with
hydrazine hydrate in ethanol under reflux to produce 2-naph-
thohydrazide 5.[37] The desired aldehyde 8 was obtained
through a palladium-catalyzed cross-coupling reaction (Suzuki–
Miyamura reaction).[38] The Suzuki reaction is an efficient and
economical method for synthesizing new Csp2ÀCsp2 bonds, par-
ticularly in the synthesis of new leads.[39,40] For this synthesis,
appropriate aryl bromide 7 and formylbenzene boronic acid 6
reacted in the presence of sodium carbonate and palladium
catalyst (PdCl2(PPh3)2) in acetonitrile and water under micro-
wave irradiation. After column chromatography, the aldehyde
8 was obtained in moderate yield.[41] Next, the synthesized al-
dehyde 8 was condensed with 2-naphthohydrazide 5, without
LASSBio-1829 (9, Figure 5) has one neutral and one proton-
ated form predicted by Epik, with a predicted pKa equal to
4.38Æ1.47 (Figure 5B) and predicted state penalties of 0.0007
and 4.3996 kcalmolÀ1 for the neutral and ionized forms, re-
spectively. The neutral form of LASSBio-1829 has a better dock-
ing score than its ionized form (À9.719 and À1.595 kcalmolÀ1).
According to the docking results in the IKK2 model (Fig-
ure 5A), the neutral form of LASSBio-1829 shows the 7-azain-
dole group interacting through two hydrogen bonds with the
Cys99 from the hinge, similar to the docking of compound 2.
Moreover, 9 maintains the hydrogen bonds with the key resi-
dues Lys44 and Asp166 as well as the stacking interactions
with Phe26 from the G-loop, similar to the docking of LASS-
Bio-1524 (1a, Figure 2A).
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
237
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
different hydrogens being dis-
played within the same region.
One is the imine hydrogen of
NAH, and the other is the a-car-
bonyl hydrogen on the naphthyl
core. This proposition was con-
firmed by analysis of the HSQC
spectrum of LASSBio-1829 (see
Supporting Information page
S18). The hydrogen at 8.6 ppm is
attached to the carbon in
147.68 ppm, corresponding to
the imine carbon of NAH. The
other hydrogen in 8.57 ppm is
attached to
a
carbon at
143.16 ppm, referring to the
carbon on the naphthyl nucleus.
Pharmacology
LASSBio-1829 (9) was evaluated
for inhibition of IKK2 through in
vitro tests by measuring the
phosphorylation of phospho-IkB-
Figure 5. A) Docking results of the neutral (green) and protonated (cyan with transparency) forms of the com-
pound LASSBio-1829 (left) and 2D ligand interaction diagram of its neutral form (right); note: residue position
number labels on the 2D ligand interaction diagram are shifted by nÀ14 relative to residue numbering at left and
in the text. B) Neutral and protonated structures of the NAH derivative LASSBio-1829 and their state penalties pre-
dicted by Epik.
a-Ulight
substrate
using
a human recombinant enzyme
expressed in Sf21 cells.[44] Un-
fortunately, this compound did
not exhibit inhibitory activity.
This might be due to the sensi-
tivity of in vitro experiments to
the aqueous solubility of the
compounds tested.[45] This hy-
pothesis prompted us to pro-
pose that a lack of activity of
this compound could be due to
low water solubility. To improve
the aqueous solubility of the
compound, we proposed the
synthesis of LASSBio-1829 hydro-
chloride (10). As expected, the
hydrochloride 10 was 100-fold
more soluble in water than the
free base 9 (Table 1). In the same
in vitro test, LASSBio-1829 hy-
drochloride (10) is active (IC50 =
Scheme 1. Reagents and conditions: a) MeOH, room temperature, 2 h, 86%; b) NH2NH2, H2O, 80% EtOH, reflux, 4 h,
82%; c) PdCl2(PPh3)2, Na2CO3, CH3CN, H2O, MW irradiation, 5 min, 65%; d) EtOH, 24 h, 75–80%; e) CH3Cl, HCl(g), 15–
30 min, 15–50%.
the usual acid catalyst, to generate LASSBio-1829 (9) in good
3.8 mm) and more potent than the initial prototype, LASSBio-
1524 (IC50 =20 mm) (Table 1).
yield.[37] The synthesis of the hydrochloride 10 was performed
by bubbling hydrochloric acid gas in a chloroform solution
containing LASSBio-1829 (9) (Scheme 1).
The double bond of NAH can be in the E or Z configura-
tion.[42] As already well established,[43] the absence of duplica-
tion of the signals relating to amide and imine hydrogens of
LASSBio-1829 (9) in the 1H NMR spectrum is a strong indication
that only the diastereomer presenting the relative configura-
Subcutaneous air pouch (SAP) model
LASSBio-1829 (9) and hydrochloride LASSBio-1829 (10) were
tested in in vivo models to confirm their possible anti-inflam-
matory action by inhibiting the NF-kB-IKK2 pathway. The first
model selected was the subcutaneous air pouch (SAP) in
mice[47] because this model has been described to activate NF-
1
tion E was formed. The imine hydrogen atom in H NMR spec-
tra may appear to be doubled, but this profile results from two
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
238
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
inhibitors, because two IKK2 inhibitors, LASSBio-1524 (1a)[25]
and SC-514[49] both at doses of 30 mmolkgÀ1, showed similar
potency to that of LASSBio-1829 hydrochloride (Figure 6), re-
ducing cell migration by 23 and 46%, respectively.
Table 1. Aqueous solubility of the compounds and their inhibitory activi-
ty against IKK2.
[a]
Compound
Aq. Sol. [mgmLÀ1
]
IC50 [mm][b]
LASSBio-1524 (1a)
LASSBio-1829 (9)
LASSBio-1829·HCl (10)
Staurosporine[c]
1.20”10À3
5.90”10À4
5.54”10À2
ND
20
ND
3.8
0.52
TNF-a measurement
LASSBio-1829 hydrochloride (10) also showed an interesting
anti-TNF-a activity in the air pouch exudate at a dose of
30 mmolkgÀ1, inhibiting TNF-a migration by 46% (Figure 7). A
decrease in TNF-a can be due to the IKK2 inhibitory action,
whereas TNF-a expression is controlled by the NF-kB pathway,
which involves activation of IKK2.[50]
[a] Solubility was determined by UV/Vis spectroscopy as described by
Schneider et al.[46] [b] Assay performed at Cerep (France). [c] Reference
compound for IKK assay. ND: not determined.
kB in murine cells due to the inflammatory process induced by
carrageenan.[48]
Fortunately, both of the compounds showed good anti-in-
flammatory profiles in this model because they were both able
to decrease, with differential potency, cell migration to the air
pouch. At all of the doses tested, LASSBio-1829 (9) provided
a decrease in cell migration, presenting a dose–response effect
(Figure 6); however, only at the highest dose of LASSBio-1829,
Figure 7. Effect of LASSBio-1829 hydrochloride (10) on the levels of TNF-
a from exudate from the SAP. Animals were pretreated orally 60 min before
carrageenan injection in the SAP, with compound 10 (10, 30, or
100 mmolkgÀ1), LASSBio-1524 (30 mmolkgÀ1), SC-514 (30 mmolkgÀ1), or vehi-
cle. Results are expressed as the mean ÆSD of TNF-a concentration. Statisti-
cal significance (p<0.05) was calculated by analysis of variance (ANOVA) fol-
lowed by Bonferroni post-test: #relative to the vehicle group that received
PBS in SAP; *relative to the vehicle group that received carrageenan in SAP.
Figure 6. Effect of LASSBio-1829 (9) and LASSBio-1829 hydrochloride (10) on
leukocyte migration induced by carrageenan in the subcutaneous air pouch
(SAP). Animals were pretreated orally 60 min before carrageenan injection in
the SAP with test compounds (at 10, 30, and 100 mmolkgÀ1), dexamethasone
(1.5 mmolkgÀ1 i.p.), LASSBio-1524 (30 mmolkgÀ1), SC-514 (30 mmolkgÀ1 p.o.),
or vehicle. Results are expressed as the mean ÆSD of the number of total
leukocytes. Statistical significance (p<0.05) was calculated by analysis of var-
iance (ANOVA) followed by Bonferroni post-test: #relative to the vehicle
group that received phosphate-buffered saline (PBS) in SAP; *relative to the
vehicle group that received carrageenan in SAP.
Nitrate production
To access the inflammatory mediators involved with the NF-kB
pathway that are produced during the inflammatory process
and a possible effect of LASSBio-1829, the amount of nitric
oxide (NO) that accumulated in the SAP exudate as nitrate was
quantified. Figure 8 shows that the animals pretreated orally
with vehicle and injected with phosphate-buffered saline (PBS)
in SAP had significantly lower nitrate production. In the ani-
mals that received carrageenan injection in the SAP, pretreat-
ment with the lowest dose of LASSBio-1829 hydrochloride was
able to significantly decrease the production of NO relative to
vehicle administration; however, higher concentrations of
LASSBio-1829 hydrochloride were not able to decrease the
production of NO.
100 mmolkgÀ1, was a significant decrease in cell migration to
the pouch observed, by ~48%. As expected, LASSBio-1829 hy-
drochloride (10) was approximately twofold more potent than
the respective free base 9. At 30 and 100 mmolkgÀ1 LASSBio-
1829 hydrochloride (10) showed a significant decrease in cell
migration, at 29 and 35%, respectively (Figure 6). The fact that
the compounds were orally delivered in this model supports
the hydrochloride synthesis strategy not only to enhance
aqueous solubility but also to increase the intestinal absorp-
tion and bioavailability of the compound, resulting in a better
cellular inhibition profile. The inhibitory behavior of LASSBio-
1829 hydrochloride is consistent with that expected for IKK2
Production of reactive oxygen species
The production of reactive oxygen species (ROS) is closely re-
lated to NF-kB activation.[51,52] Therefore, the production of
ROS was quantified in the animal air pouch treated with LASS-
Bio-1829 hydrochloride. The compound had an excellent
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
239
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 10. Effect of LASSBio-1829 hydrochloride (10) during the first and
second phases of the formalin-induced licking response model. Animals
were pretreated orally with compound 10 (30 mmolkgÀ1), LASSBio-1524
(30 mmolkgÀ1), acetylsalicylic acid (ASA, 1000 mmolkgÀ1), or vehicle. Statisti-
cal significance (p<0.05) was calculated by ANOVA followed by Bonferroni
post-test: *relative to the vehicle group.
Figure 8. Effect of LASSBio-1829 hydrochloride (10) in nitric oxide (NO) pro-
duction induced by carrageenan in the SAP. Animals were pretreated orally
60 min before carrageenan injection in the SAP with compound 10 (10, 30,
or 100 mmolkgÀ1), LASSBio-1524 (30 mmolkgÀ1), SC-514 (30 mmolkgÀ1), or ve-
hicle. Results are expressed as the mean ÆSD of NO concentration. Statisti-
cal significance (p<0.05) was calculated by ANOVA followed by Bonferroni
post-test: #relative to the vehicle group that received PBS in SAP; *relative
to the vehicle group that received carrageenan in SAP.
tion of the IKK2-NF-kB pathway is not related to analgesia;
however, during the second stage, where there is significant
inflammatory involvement, the hydrochloride was able to mini-
mize the licking number, once again confirming its good anti-
inflammatory profile (Figure 10). LASSBio-1524 was also admin-
istered at the same dose (30 mmolkgÀ1) as LASSBio-1829 and
demonstrated pharmacological activity only during the second
stage of the assay. This finding supports the hypothesis that
IKK2 inhibitors are active only during the inflammatory phase
in the formalin test. Acetylsalicylic acid (ASA), the positive con-
trol used in this test, reinforces this postulate. ASA has been
reported to inhibit IKK2[53] and, similar to LASSBio-1524 and
LASSBio-1829, showed activity only during the inflammatory
phase of the test.
Figure 9. Effect of LASSBio-1829 hydrochloride (10), LASSBio-1524
(30 mmolkgÀ1), and SC-514 (30 mmolkgÀ1) on reactive oxygen species (ROS)
production from leukocytes collected from the SAP 24 h after carrageenan
injection and stimulated with phorbol myristate acetate (PMA). Reading was
performed on a flow cytometer counting 10000 events, and values shown
are the geometric mean of fluorescence intensity expression of 2’,7’-dichlor-
ofluorescein (DCF) in the FL-1 channel. Statistical significance (p<0.05) was
calculated by ANOVA followed by Bonferroni post-test: #relative to the
PMA-unstimulated vehicle group; *relative to the PMA-activated vehicle
group.
Conclusions
In this work, we generated a three-dimensional model of IKK2
and exploited the profile of the ATP binding site through ATP
and inhibitor docking studies. The molecular model built for
IKK2 together with the docking methodology employed pro-
vided important and consistent information with respect to
the structural and chemical inhibitor characteristics that may
confer potency to IKK2 inhibitors. The docking results not only
provided an explanation of the activity of compound LASSBio-
1524 (1a) but also furnished important guidelines for the de-
velopment of a new NAH inhibitor of IKK2. The docking results
for LASSBio-1829 (9) predicted the presence of an important
interaction with the hinge, the conserved residues Lys44 and
Asp166, and stacking interactions with Phe26 from the G-loop.
Furthermore, the analyses of the protonation states and the
docking results indicate that LASSBio-1829 interacts with the
enzyme in its neutral form. An important finding is that the
aqueous solubility of LASSBio-1829 plays an important role in
the determination of its potency. In fact, the inhibition of IKK2
through in vitro tests showed that only the hydrochloride of
LASSBio-1829 (10) was active with an IC50 of 3.8 mm, showing
more potent inhibition than the NAH prototype, LASSBio-1524.
Additionally, during in vivo tests for inflammation with partici-
dose–response profile, significantly decreasing ROS production
at concentrations of 10 and 30 mm (Figure 9). LASSBio-1829
demonstrated a better pharmacological profile than LASSBio-
1524 and SC-514 at 30 mmolkgÀ1, consistent with its higher in
vitro potency in relation to the other two IKK2 inhibitors.
Formalin test
LASSBio-1829 hydrochloride was submitted to the formalin
test. The dose selected was 30 mmolkgÀ1 because this dose ap-
peared to be the most promising in the other tests. During the
first phase of the formalin test, no significant action was ob-
served. This lack of activity was expected because the inhibi-
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
240
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Molecular docking simulations: Protein–ligand molecular docking
is a widely used strategy that aims to predict the binding modes
and affinities of one or more small molecules (ligands) in the bind-
ing site of a particular receptor (protein).[68] The molecular docking
experiments with the ATP cofactor were performed in the IKK2
binding site to validate the model built. The top ranked positions
of the docking experiments were compared with the experimental
conformation of ATP complexed with CDK2 (PDB code 1HCK[30]).
The docking experiments successfully determined the experimental
binding mode of ATP only in the presence of a water molecule
and the metal ion. This finding might indicate that the presence of
at least one metal ion and a water molecule is important for ATP
interaction in the IKK2 ATP binding site.
pation of the NF-kB-IKK2 pathway, both compounds, LASSBio-
1829 and LASSBio-1829 hydrochloride, were active; however,
LASSBio-1829 hydrochloride is the most promising compound,
showing superior potency. This result is likely due to their su-
perior physicochemical profile, which can be seen in improved
absorption, bioavailability and therapeutic action.
Experimental Section
Computational methods
The docking studies of 18 inhibitors of IKK2 extracted from the
BindingDB database[69] were performed to achieve a better under-
standing of the properties that may confer their distinct potency
and selectivity. The previously synthesized LASSBio-1524 (1a) and
compound 1b[25] were docked into the IKK2 model to investigate
the binding mode differences that could justify their distinct activi-
ty. The structures of the 18 inhibitors selected from the literature
were obtained from the BindingDB[69] and PubChem[70] databases
(Table S2), whereas the ATP and LASSBio compounds were manual-
ly designed in the Maestro Suite 2011 (version 9.2, Schrçdinger
LLC, New York, NY (USA), 2011). Only the E diastereomers of the
LASSBio compounds were considered for the docking studies. All
of the structures were prepared with LigPrep (version 2.5, Schrç-
dinger LLC, 2011), and the protonation states were determined
with Epik at pH 7.4Æ4.0.[71,72] The docking studies were performed
with the Glide software (version 5.7, Schrçdinger LLC, 2011) using
the Extra Precision Docking mode.[73]
Structural modeling of IKK2: Comparative modeling is a method-
ology that aims to generate 3D models of a protein sequence of
interest (target) based on the primary sequences and using struc-
tures of similar proteins as templates.[54,55] The molecular model of
IKK2 was constructed with Modeller ver. 8[56] through the multiple
templates approach. The amino acid sequence of the protein
target (IKK2) from Homo sapiens was extracted from the bank of
sequences UniProt[57] and stored with the code O14920. Because
no high-quality structure is available that covers the entire se-
quence to be used as a template, only the catalytic domain of IKK2
was built (residues 15–300). Furthermore, the main focus of this
work is the development of inhibitors that interact with the ATP
binding site, which is located in the catalytic domain.
The search for sequences of homologous protein kinases to be
used as templates was carried out with BLASTP[58] using the
BLOSUM62 matrix. For each isoform, a set of structures to be used
in building the model was selected according to the following cri-
teria: 1) high percentage of sequence identity (>30%), 2) experi-
mental structure with the ATP binding site well determined,
3) structural resolution less than or equal to 2.5 Š, and 4) absence
of mutations. According to these criteria, three structures were se-
lected as templates (PDB codes: 3BQR,[59] 2A2A,[60] and 1YHW[61]) for
building the IKK2 model (see details in Table S1, Supporting Infor-
mation).
Chemistry
General information: See the Supporting Information for details
regarding reagents and equipment.
4-(1H-Pyrrolo[2,3-b]pyridin-4-yl)benzaldehyde (8): In a glass cylin-
der suitable for microwave reactions 4-bromo-1H-pyrrolo[2,3-b]pyr-
idine (7, 0.2 g, 1.16 mmol, 1 equiv), 4-formylbenzeneboronic acid
(6, 0.19 g, 1.27 mmol, 1,1 equiv), sodium carbonate (0.36 g,
3.74 mmol, 3 equiv) and catalyst (PdCl2(PPh3)2) (0.04 g, 5.78 mmol,
0,05 equiv) were added. The reagents were then dissolved in 6 mL
of a 1:1 solution of CH3CN and H2O, and the cylinder was inserted
into the a microwave reactor, where it was heated for 5 min at
1508C. After this time, TLC analysis revealed that the major prod-
uct was the desired aldehyde 8. The contents of the flask were
transferred to a round-bottom flask, and the solvent was evaporat-
ed under reduced pressure. The solid obtained was purified by
column chromatography with the eluent n-hexane/EtOAc (5–50%)
The three top-ranking models according to the DOPE energy pro-
vided by the Modeller program were selected for the step of loop
optimization. Finally, the structural quality of the best energy
model was analyzed using the MolProbity,[62] QMEAN,[63] and
ProSA[64,65] tools. MolProbity is largely employed to predict outlier
residues on protein structures. ProSA uses force-field-based poten-
tials to evaluate the accuracy of the protein structure according to
statistical analyses of known structures of proteins. QMEAN is a vali-
dation tool that evaluates the quality of the protein structure ac-
cording to geometrical factors.
1
The presence of conserved water and metal ions in the ATP bind-
ing site was analyzed through superimposition of diverse kinase–
ATP/ADP/ANP complexes. Despite the high difference in the posi-
tion of the metal ions between the complexes analyzed, kinetics
studies demonstrated that a first cation is essential for phosphory-
lation of the ATP in most protein kinases.[66] Nonetheless, this
cation access is believed to provide the cofactor binding site com-
plexed with ATP, justifying its absence in most inhibitor–kinase
complexes.
to provide the desired aldehyde 8 in 65% yield. H NMR (200 MHz,
[D6]DMSO): d=11.90 (1H, s), 10.10 (1H, s), 8.35 (1H, d, J=6 Hz),
8.11 (2H, d, J=8 Hz), 8.02 (2H, d, J=8 Hz), 7.61 (1H, m), 7.29 (1H,
d, J=6 Hz), 6.66 ppm (1H, d, J=4 Hz); 13C NMR (50 MHz,
[D6]DMSO): d=194.7, 149.4, 145.1, 143.7, 140.5, 136.6, 131.3, 130.0,
128.4, 118.5, 115.6, 100.2 ppm; IR (KBr): n˜ =3135 and 3079 (NÀH),
2772 and 2877 (CÀH), 1705 cmÀ1 (C=O).
(E)-N-(4-(1H-Pyrrolo[2,3-b]pyridin-4-yl)benzylidene)-2-naphtho-
hydrazide (9, LASSBio-1829): To a solution of 2-naphthohydrazide
(5, 0.17 g, 92.9 mmol, 1 equiv) in absolute EtOH (15 mL) aldehyde
8 was added (0.19 g, 86.9 mmol, 1 equiv). The mixture was stirred
at room temperature for 24 h. Afterward, the solvent was partially
concentrated at reduced pressure, and the resulting mixture was
poured into ice and cold water. The precipitate formed was filtered
The importance of water and metal ions in the ATP binding site
was analyzed through docking studies of ATP in the presence and
absence of water and ions. The water molecule (W397) and Mg2+
ion were extracted from the structure of the death-associated pro-
tein kinase 1 complexed with ANP (PDB code 3F5U[67]).
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
241
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
out and dried under vacuum, producing the desired N-acylhydra-
zone in 80% yield. H NMR (400 MHz, [D6]DMSO): d=12.13 (1H, s),
kine concentrations were calculated using a standard curve and
are expressed as pgmLÀ1
.
1
11.80 (1H, s), 8.58 (1H, s), 8.56 (1H, s), 8.29 (1H, d, J=4 Hz), 8.01–
8.05 (2H, m), 7.99 (2H, d, J=8 Hz), 7.94 (2H, d, J=8 Hz), 7.88 (2H,
d, J=8 Hz), 7.61–7.66 (2H, m), 7.56 (1H, s), 7.24 (1H, d, J=4 Hz),
6.67 ppm (1H, s); 13C NMR (50 MHz, [D6]DMSO+CDCl3): d=163.5,
149.0, 147.5, 142.8, 140.1, 139.7, 134.4, 134.3, 132.1, 130.6, 128.9,
128.6, 128.1, 127.9, 127.7, 127.7, 126.9, 126.7, 124.3, 117.3, 114.1,
99.1 ppm; IR (KBr): n˜ =1648 (C=O), 3440, 3199, 3138 cmÀ1 (NÀH);
HPLC: 96.56% purity; MS: m/z=391.1 [M+H]+.
Production of nitric oxide: NO produced in the SAP supernatant
was quantified according to the technique of the conversion of ni-
trate to nitrite.[46] The SAP samples were deproteinized and then
admixed to a sample of sodium phosphate (0.5m, pH 7.2), ammo-
nium formate (2.4m, pH 7.2), and E. coli. After incubation for 2 h at
378C, centrifugation was performed at 10000 rpm for 10 min.
Equal portions of the supernatant and Griess reagent were incu-
bated for 10 min,[75] and the absorbance was measured spectro-
photometrically at l 540 nm. The nitrate concentration values are
expressed in nanomolar and were calculated from a standard
curve of sodium nitrate.
(E)-N-(4-(1H-Pyrrolo[2,3-b]pyridin-4-yl)benzylidene)-2-naphtho-
hydrazide hydrochloride (10, LASSBio-1829 hydrochloride): In
a flask containing 0.70 g LASSBio-1829 (9), CHCl3 was added
(150 mL), and the solution was transferred to magnetic stirring. Hy-
drochloric acid gas was bubbled into the flask for 30 min. During
the reaction the hydrochloride precipitated as a yellow solid. After
the solvent was evaporated under reduced pressure, Et2O was
added to the flask and vacuum filtered, obtaining LASSBio-1829
Determination of reactive oxygen species (ROS) production: Leuko-
cytes collected in the SAP were placed in tubes (106 cells) in
a volume of 1 mL. Incubation was performed at 378C and 5% CO2
for 1 h. Test compounds were then added at concentrations of 1,
10, and 30 mm and incubated for 30 min at 378C and 5% CO2. The
cells were treated with 10 nm phorbol myristate acetate (PMA) and
incubated for 45 min at 378C and 5% CO2. The cells were then
added to 2’-7’diclorodihidrofluoescein diacetate (DCF-DA, 2 mm),
followed by another incubation for 30 min at 378C. DCF-DA quickly
permeates the cell membrane, and once in the intracellular space,
is hydrolyzed by plasma esterase, to be converted into 2’7’-dichlor-
ofluorescein (DCFH), which is a non-fluorescent compound that is
impermeable to the cell membrane. In the presence of ROS, DCFH
is oxidized within the cell and produces a fluorescent compound,
2’,7’-dichlorofluorescein (DCF), which remains in the extracellular
space.[76] The emitted fluorescence is captured in the FL-1 channel
of a flow cytometer and is expressed in arbitrary units (au).
1
hydrochloride (10) in 50% yield. H NMR 200 MHz, [D6]DMSO): d=
12.18 (1H, s), 8.61 (1H, s), 8.58 (1H, s), 8.41 (1H, d, J=6 Hz), 7.95–
8.23 (8H, m), 7.63–7.67 (3H, m), 7.35–7.42 (1H, m), 6.78 ppm (1H,
m); HPLC: 96.44% purity.
Pharmacology
Animals: The experimental groups were composed of 5–8 female
Swiss Webster mice of 20–25 g donated by the Animal Production
Centre of the Institute Vital Brazil. The experimental protocols fol-
lowed the rules advocated by Law 11.794, of October 8, 2008 by
the National Council of Animal Experimentation Control (CONCEA)
and were approved by the Ethics Committee of Animal Use
(CEUA), Science Center Health/UFRJ number DFBCICB015–04/16.
Formalin test: Animals received an injection of formalin (2.5% v/v)
into their left hind paws. The time that the animals spent licking
their paws after injection was recorded. The nociceptive response
develops during two phases: 0–5 min after the formalin injection
(first phase, neurogenic pain response) and 15–30 min after the
formalin injection (second phase, inflammatory pain response).
Preparation and administration of compounds: LASSBio-1829 and
LASSBio-1829 hydrochloride were prepared in a stock solution of
100 mmol in 1 mL DMSO and stored at À208C. The compounds
were administered orally at doses of 10, 30, or 100 mmolkgÀ1 in
a final volume of 100 mL vehicle (Polysorbate 80). LASSBio-1524
and the selective and reversible IKK2 inhibitor SC-514[74] were ad-
ministered orally at a single dose of 30 mmolkgÀ1. The standard
anti-inflammatory medications used were dexamethasone
(1.5 mmolkgÀ1) and acetylsalicylic acid (ASA, 1000 mmolkgÀ1, p.o.).
Statistical analysis: All of the experiments were composed of 6–8
randomly selected animals per group. The results are presented as
the mean Æstandard deviation (SD). Statistical significance was cal-
culated by analysis of variance (ANOVA) followed by Bonferroni
post-test, using GraphPad Prism ver. 5.0 software; significance was
considered at p<0.05.
Subcutaneous air pouch (SAP) model: The formation of an air pouch
in the dorsum of mice was carried out upon injection of 10 mL
sterile air a 1% carrageenan solution, a phlogistic agent, to induce
cell migration.[47] After 3 days, over 7 mL of sterile air were injected.
On day 6, the animals were treated with compounds, and 60 min
later, a sterile 1% carrageenan injection was made in the formed
cavity. A negative control group was treated with vehicle 60 min
before receiving the injection of the sterile carrageenan solution at
SAP, and the positive control group received dexamethasone and
SC-514. 24 h post-carrageenan injection, animals were euthanized,
and the cavity was washed with 1 mL phosphate-buffered saline
(PBS). The number of cells from the collected exudate was then de-
termined with an automatic cell counter (Poch-100iV Diff, Sysmex).
The exudate was centrifuged at 1000 rpm for 10 min at 48C. The
supernatant was stored at À208C until subsequent measurements.
Acknowledgements
The authors thank Instituto Nacional de CiÞncia e Tecnologia de
Fµrmacos e Medicamentos (INCT-INOFAR) (grants 573.564/2008-6
and 170.020/2008), Fundażo Carlos Chagas Filho de Amparo à
Pesquisa do Estado do Rio de Janeiro (FAPERJ), and Conselho Na-
cional de Desenvolvimento Científico e Tecnológico (CNPq) for fi-
nancial support. The authors also thank the Institute Vital Brazil
for animal donation, Coordenadoria de AperfeiÅoamento de Pes-
soal de Nível Superior (CAPES) for fellowships to N.S.C. and
R.H.C.N.F. , and Alan Minho (Institute of Biomedical Sciences,
UFRJ) for technical support.
TNF-a measurement: Quantification of TNF-a was carried out for
the SAP exudate. The ELISA kit BD ELISA OptEIATM was used, and
the TNF-a concentration was determined according to the manu-
facturer’s recommendations (B&D Biosciences). The absorbance
was measured at l 450 nm using a microplate reader, and the cyto-
Keywords: anti-inflammatory agents
docking · N-acylhydrazones · structure-based drug design
· IKK2 · molecular
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
242
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[1] D. M. Rothwarf, E. Zandi, G. Natoli, M. Karin, Nature 1998, 395, 297–300.
[2] S. Yamaoka, G. Courtois, C. Bessia, S. T. Whiteside, R. Weil, F. Agou, H. E.
Kirk, R. J. Kay, A. Isral, Cell 1998, 93, 1231–1240.
[3] E. Zandi, D. M. Rothwarf, M. Delhase, M. Hayakawa, M. Karin, Cell 1997,
91, 243–252.
Dunah, P. A. Snodgrass-belt, H. M. Arnold, A. G. Taveras, K. J. Rhodes,
R. H. Scannevin, ACS Med. Chem. Lett. 2011, 2, 786–791.
[40] H. Suzuki, I. Utsunomiya, K. Shudo, T. Fujimura, M. Tsuji, I. Kato, T. Aoki,
A. Ino, T. Iwaki, ACS Med. Chem. Lett. 2013, 4, 1074–1078.
[41] Y. Gong, W. He, Org. Lett. 2002, 4, 3803–3805.
[4] R. M. Townsend, J. Postelnek, V. Susulic, K. W. McIntyre, D. J. Shuster, Y.
Qiu, F. C. Zusi, J. R. Burke, Transplantation 2004, 77, 1090–1094.
[5] J. A. Schmid, A. Birbach, Cytokine Growth Factor Rev. 2008, 19, 157–165.
[6] A. Israel, Cold Spring Harbor Perspect. Biol. 2010, 2, a000158–a000158.
[7] J.-L. Luo, H. Kamata, M. Karin, J. Clin. Invest. 2005, 115, 2625–2632.
[8] C. H. Lee, Y.-T. Jeon, S.-H. Kim, Y.-S. Song, Biofactors 2007, 29, 19–35.
[9] M. C.-T. Huo, D.-F. Lee, W. Xia, L. S. Golfman, F. Ou-Yang, J.-Y. Yang, Y.
Zou, S. Bao, N. Hanada, H. Saso, R. Kobayashi, M. C. Hung, Cell 2004,
117, 225–237.
[10] H. J. Kim, N. Hawke, A. S. Baldwin, Cell Death Differ. 2006, 13, 738–747.
[11] J. Strnad, J. R. Burke, Trends Pharmacol. Sci. 2007, 28, 142–148.
[12] S. Uota, M. Zahidunnabi Dewan, Y. Saitoh, S. Muto, A. Itai, A. Utsuno-
miya, T. Watanabe, N. Yamamoto, S. Yamaoka, Cancer Sci. 2012, 103,
100–106.
[42] G. Palla, G. Predieri, P. Domiano, C. Vignali, W. Turner, Tetrahedron 1986,
42, 3649–3654.
[43] G. Palla, C. Pelizzi, G. Predieri, Gazz. Chim. Ital. 1982, 112, 339–341.
[44] Q. K. Huynh, H. Boddupalli, S. A. Rouw, C. M. Koboldt, T. Hall, C. Somm-
ers, S. D. Hauser, J. L. Pierce, R. G. Combs, B. A. Reitz, J. A. Diaz-Collier,
R. A. Weinberg, B. L. Hood, B. F. Kilpatrick, C. S. Tripp, J. Biol. Chem.
2000, 275, 25883–25891.
[45] J. B. Dressman, C. Reppas, Eur. J. Pharm. Sci. 2000, 11, S73–S80.
[46] P. Schneider, S. S. Hosseiny, M. Szczotka, V. Jordan, K. Schlitter, Phyto-
chem. Lett. 2009, 2, 85–87.
[47] L. J. R. P. Raymundo, C. C. Guilhon, D. S. Alviano, M. E. Matheus, A. R. An-
toniolli, S. C. H. Cavalcanti, P. B. Alves, C. S. Alviano, P. D. Fernandes, J.
Ethnopharmacol. 2011, 134, 725–732.
[48] L. Ellis, C. C. Y. Soo, C. J. Morris, B. L. Kidd, P. G. Winyard, Ann. Rheum.
Dis. 2000, 59, 303–307.
[49] A. Baxter, S. Brough, A. Cooper, E. Floettmann, S. Foster, C. Harding, J.
Kettle, T. McInally, C. Martin, M. Mobbs, M. Needham, P. Newham, S.
Paine, S. St-Gallay, S. Salter, J. Unitt, Y. Xue, Bioorg. Med. Chem. Lett.
2004, 14, 2817–2822.
[50] S. C. Gupta, C. Sundaram, S. Reuter, B. B. Aggarwal, BBA-Gene Regul.
Mech. 2010, 1799, 775–787.
[13] Y. Yamamoto, R. B. Gaynor, J. Clin. Invest. 2001, 107, 135–142.
[14] S. Ghosh, M. Karin, Cell 2002, 109, S81–96.
[15] Q. Li, I. M. Verma, Nat. Rev. Immunol. 2002, 2, 725–734.
[16] S. Vallabhapurapu, M. Karin, Annu. Rev. Immunol. 2009, 27, 693–733.
[17] G. Courtois, T. D. Gilmore, Oncogene 2006, 25, 6831–6843.
[18] A. Oeckinghaus, S. Ghosh, Cold Spring Harbor Perspect. Biol. 2009, 1,
a000034.
[19] M. Karin, M. Delhase, Semin. Immunol. 2000, 12, 85–98.
[20] G. W. Peet, J. Li, J. Biol. Chem. 1999, 274, 32655–32661.
[21] Y. Hu, V. Baud, T. Oga, K. I. Kim, K. Yoshida, M. Karin, Nature 2001, 410,
710–714.
[22] G. Xu, Y.-C. Lo, Q. Li, G. Napolitano, X. Wu, X. Jiang, M. Dreano, M. Karin,
H. Wu, Nature 2011, 472, 325–330.
[23] S. Polley, D.-B. Huang, A. V. Hauenstein, A. J. Fusco, X. Zhong, D. Vu, B.
Schrçfelbauer, Y. Kim, A. Hoffmann, I. M. Verma, G. Ghosh, T. Huxford,
PLoS Biol. 2013, 11, e1001581.
[24] S. Liu, Y. R. Misquitta, A. Olland, M. A. Johnson, K. S. Kelleher, R. Kriz, L. L.
Lin, M. Stahl, L. Mosyak, J. Biol. Chem. 2013, 288, 22758–22767.
[25] C. M. Avila, A. B. Lopes, A. S. GonÅalves, L. L. da Silva, N. C. Romeiro,
A. L. P. Miranda, C. M. R. Sant’Anna, E. J. Barreiro, C. A. M. Fraga, Eur. J.
Med. Chem. 2011, 46, 1245–1253.
[26] G. Scapin, S. B. Patel, J. Lisnock, J. W. Becker, P. V. LoGrasso, Chem. Biol.
2003, 10, 705–712.
[27] A. P. Kornev, N. M. Haste, S. S. Taylor, L. F. T. Eyck, Proc. Natl. Acad. Sci.
USA 2006, 103, 17783–17788.
[28] R. Roskoski, Expert Opin. Drug Discovery 2013, 8, 1165–1179.
[29] K. C. Duong-Ly, J. R. Peterson, Curr. Protoc. Pharmacol. 2013, 60, 2.9.1–
2.9.14.
[51] C. Bubici, S. Papa, K. Dean, G. Franzoso, Oncogene 2006, 25, 6731–
6748.
[52] M. J. Morgan, Z. Liu, Cell Res. 2011, 21, 103–115.
[53] M.-J. Yin, Y. Yamamoto, R. B. Gaynor, Nature 1998, 396, 77–80.
[54] M. A. Martí-Renom, A. C. Stuart, A. Fiser, R. Sµnchez, F. Melo, A. Sali,
Annu. Rev. Biophys. Biomol. Struct. 2000, 29, 291–325.
[55] M. S. Madhusudhan, M. A. Marti-Renom, N. Eswar, B. John, U. Pieper, R.
Karchin, M. Shen, A. Sali, Comparative Protein Structure Modeling (Ed.:
J. M. Walker), The Proteomics Protocols Handbook, Humana Press Inc.,
Totowa, NJ, 2005, 831–860.
ˇ
[56] A. Sali, T. L. Blundell, J. Mol. Biol. 1993, 234, 779–815.
[57] The UniProt Consortium, Nucleic Acids Res. 2015, 43, D204–D212.
[58] S. F. Altschul, W. Gish, W. Miller, E. W. Myers, D. J. Lipman, J. Mol. Biol.
1990, 215, 403–410.
[59] “Crystal Structure of Human Death Associated Protein Kinase 3 (DAPK3)
in Complex with an Imidazo-Pyridazine Ligand”, P. Filippakopoulos, P.
Rellos, O. Fedorov, F. Niesen, A. C. W. Pike, E. S. Pilka, F. von Delft, C. H.
Arrowsmith, A. M. Edwards, J. Weigelt, S. Knapp, PDB ID 3BQR: http://
pdb.org/pdb/explore/explore.do?structureId=3bqr,
DOI:
10.2210/
pdb3bqr/pdb.
[60] “High-Resolution Analysis of the Autoinhibited Conformation of Human
DRP-1 Kinase”, P. Kursula, M. Wilmanns, PDB ID 2A2A: http://pdb.org/
pdb/explore/explore.do?structureId=2a2a, DOI: 10.2210/pdb2a2a/pdb.
[61] M. Lei, M. A. Robinson, S. C. Harrison, Structure 2005, 13, 769–778.
[62] V. B. Chen, W. B. Arendall III, J. J. Headd, D. A. Keedy, R. M. Immormino,
G. J. Kapral, L. W. Murray, J. S. Richardson, D. C. Richardson, Acta Crystal-
logr. Sect. D 2010, 66, 12–21.
[30] U. Schulze-Gahmen, H. L. De Bondt, S.-H. Kim, J. Med. Chem. 1996, 39,
4540–4546.
[31] C. Zhang, S.-H. Kim in Computational and Structural Approaches to Drug
Discovery: Ligand–Protein Interactions (Eds.: R. Stroud, J. Finer-Moore),
Royal Society of Chemistry, Cambridge, 2007, pp. 349–365.
[32] L. Garuti, M. Roberti, G. Bottegoni, Curr. Med. Chem. 2010, 17, 2804–
2821.
[63] P. Benkert, T. Schwede, S. C. E. Tosatto, BMC Struct. Biol. 2009, 9, 35.
[64] M. Wiederstein, M. J. Sippl, Nucleic Acids Res. 2007, 35, W407–410.
[65] M. J. Sippl, Proteins 1993, 17, 355–362.
[66] J. A. Adams, Chem. Rev. 2001, 101, 2271–2290.
[67] L. K. McNamara, D. M. Watterson, J. S. Brunzelle, Acta Crystallogr. Sect. D
2009, 65, 241–248.
[33] L. K. Gavrin, E. Saiah, MedChemComm 2013, 4, 41–51.
[34] D. Fancelli, D. Berta, S. Bindi, A. Cameron, P. Cappella, P. Carpinelli, C.
Catana, B. Forte, P. Giordano, M. L. Giorgini, S. Mantegani, A. Marsiglio,
M. Merone, J. Moll, V. Pitallà, F. Roletto, D. Severino, C. Soncini, P. Storici,
R. Tonani, M. Varasi, A. Vulpetti, P. Vianello, J. Med. Chem. 2005, 48,
3080–3084.
[68] I. A. Guedes, C. S. de Magalh¼es, L. E. Dardenne, Biophys. Rev. 2014, 6,
75–87.
[35] A. H. Bingham, R. J. Davenport, L. Gowers, R. L. Knight, C. Lowe, D. A.
Owen, D. M. Parry, W. R. Pitty, Bioorg. Med. Chem. Lett. 2004, 14, 409–
412.
[36] C. D. Duarte, E. J. Barreiro, C. A. M. Fraga, Mini-Rev. Med. Chem. 2007, 7,
1108–1119.
[37] P. C. Lima, L. M. Lima, K. C. da Silva, P. H. LØda, A. L. de Miranda, C. A. M.
Fraga, E. J. Barreiro, Eur. J. Med. Chem. 2000, 35, 187–203.
[38] A. Suzuki, Pure Appl. Chem. 1994, 66, 213–222.
[69] T. Liu, Y. Lin, X. Wen, R. N. Jorissen, M. K. Gilson, Nucleic Acids Res. 2007,
35, D198–D201.
[70] E. E. Bolton, Y. Wang, P. A. Thiessen, S. H. Bryant, Annu. Rep. Comput.
Chem. 2008, 4, 217–241.
[71] J. R. Greenwood, D. Calkins, A. P. Sullivan, J. C. Shelley, J. Comput.-Aided
Mol. Des. 2010, 24, 591–604.
[72] J. C. Shelley, A. Cholleti, L. L. Frye, J. R. Greenwood, M. R. Timlin, M. Uchi-
maya, J. Comput.-Aided Mol. Des. 2007, 21, 681–691.
[39] H. Peng, T. Talreja, Z. Xin, J. H. Cuervo, G. Kumaravel, M. J. Humora, L.
Xu, E. Rohde, L. Gan, M.-Y. Jung, M. N. Shackett, S. Chollate, A. W.
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
243
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[73] R. A. Friesner, R. B. Murphy, M. P. Repasky, L. L. Frye, J. R. Greenwood,
T. A. Halgren, P. C. Sanschagrin, D. T. Mainz, J. Med. Chem. 2006, 49,
6177–6196.
[76] N. Srivastava, V. K. Gonugunta, M. R. Puli, A. S. Raghavendra, Planta
2009, 229, 757–765.
[74] N. Kishore, C. Sommers, S. Mathialagan, J. Guzova, M. Yao, S. Hauser, K.
Huynh, S. Bonar, C. Mielke, L. Albee, R. Weier, M. Graneto, C. Hanau, T.
Perry, C. S. Tripp, J. Biol. Chem. 2003, 278, 32861–32871.
[75] L. C. Green, D. A. Wagner, J. Glogowski, P. L. Skipper, J. S. Wishnok, S. R.
Tannenbaum, Anal. Biochem. 1982, 126, 131–138.
Received: June 18, 2015
Revised: August 5, 2015
Published online on August 25, 2015
ChemMedChem 2016, 11, 234 – 244
www.chemmedchem.org
244
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim