Discovery and optimization of 1,3,5-trisubstituted pyrazolines as potent and highly selective allosteric inhibitors of protein kinase C-χ

Article abstract of DOI:10.1021/jm500521n

There is increasing evidence that the atypical protein kinase C, PKCχ, might be a therapeutic target in pulmonary and hepatic inflammatory diseases. However, targeting the highly conserved ATP-binding pocket in the catalytic domain held little promise to

Full text of DOI:10.1021/jm500521n

Article
pubs.acs.org/jmc
Discovery and Optimization of 1,3,5-Trisubstituted Pyrazolines as
Potent and Highly Selective Allosteric Inhibitors of Protein Kinase C‑ζ
Mohammad Abdel-Halim,†,§ Britta Diesel,‡ Alexandra K. Kiemer,‡ Ashraf H. Abadi,§ Rolf W. Hartmann,†
and Matthias Engel*,†
†Pharmaceutical and Medicinal Chemistry, Saarland University, Campus C2.3, D-66123 Saarbrucken, Germany
̈
‡Pharmaceutical Biology, Saarland University, Campus C2.2, D-66123 Saarbrucken, Germany
̈
§Department of Pharmaceutical Chemistry, Faculty of Pharmacy and Biotechnology, German University in Cairo, Cairo 11835, Egypt
S
* Supporting Information
ABSTRACT: There is increasing evidence that the atypical protein kinase C,
PKCζ, might be a therapeutic target in pulmonary and hepatic inflammatory
diseases. However, targeting the highly conserved ATP-binding pocket in the
catalytic domain held little promise to achieve selective inhibition. In the
present study, we introduce 1,3,5-trisubstituted pyrazolines as potent and
selective allosteric PKCζ inhibitors. The rigid scaffold offered many sites for
modification, all acting as hot spots for improving activity, and gave rise to
sharp structure−activity relationships. Targeting of PKCζ in cells was
confirmed by reporter gene assay, transfection assays, and Western blotting.
The strongly reduced cell-free and cellular activities toward a PIF-pocket
mutant of PKCζ suggested that the inhibitors most likely bound to the PIF-
pocket on the kinase catalytic domain. Thus, using a rigidification strategy and
by establishing and optimizing multiple molecular interactions with the
binding site, we were able to significantly improve the potency of the
previously reported PKCζ inhibitors.
the PKCι gene was lethal in early embryonic development.5
With respect to the PKCζ biological function, knockout studies
INTRODUCTION
■
The protein kinase C (PKC) family comprises the classical
revealed that PKCζ was essential for the development and
(cPKCα, -βI, -βII, and -γ), novel (nPKCδ, -ε, -η, and -θ), and
differentiation of B cells and Th2 cells.6 PKCζ plays a key role
atypical (aPKCζ and -λ/ι) isoenzymes, based on their
biochemical requirements for activation. Only cPKCs respond
to calcium, whereas both cPKCs and nPKCs are sensitive to
phorbol esters and the lipid second messenger diacylglycerol
(DAG). Finally, PKCζ and PKCι (termed PKCλ in the mouse)
were named atypical because they do not respond to either of
the named agents but might be activated by some lipids, such as
in autoimmune and chronic inflammatory diseases through the
coactivation of the transcriptional activity of NF-κB,4 which can
lead to exacerbated lymphocyte activities.7 PKCζ phosphor-
ylates Ser311 on the RelA subunit of NF-κB,8 which has been
shown to be required for full NF-κB transcriptional activity in
vivo and in cell culture experiments.4 Numerous transfection
experiments using kinase-defective dominant-negative mutants,
phosphatidyl inositols and ceramide (reviewed in ref 1). In
overexpression studies, or antisense experiments further
addition, aPKC activity is regulated by protein interaction
supported a role for PKCζ in the control of NF-κB
partners that also provide the required temporal and spatial
activation.9−14 Another important consequence of the Ser311
specificity, including the inhibitory protein Par-4 and the
adapter protein p62.2,3
phosphorylation by PKCζ was demonstrated by Levy et al.: it
triggered displacement of the histone methyltransferase GLP
Both atypical PKCs share a high degree of identity (84%) in
from RelA, which abrogated the GLP-mediated methylation of
the catalytic domain and are less conserved in the N-terminal
regulatory PB1 and C1 domains (53% identical). Although
both enzymes are similarly regulated and activated, not all of
their biological functions are redundant (see ref 3 for a recent
review). PKCζ knockout mice appeared grossly normal; only
the formation and maturation of secondary lymphoid organs
were altered during the first weeks after birth; however, most of
these defects were overcome in adult animals.4 These findings
suggested that targeting PKCζ with specific inhibitors might
not cause harmful toxic effects. On the other hand, knockout of
histone H3, thus finally enabling NF-κB to become transcrip-
tionally active.15 In addition, a rather cell type specific function
as an IκB kinase was demonstrated for PKCζ in the lung, thus
exerting additional control on the NF-κB activation in this
organ, but not in fibroblasts or T cells.4,6,8,16 In naive T cells
̈
and in mouse liver, PKCζ was not involved in NF-κB activation
Received: April 2, 2014
Published: July 24, 2014
© 2014 American Chemical Society
6513
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
Figure 1. Structures of the previously published allosteric inhibitors of PKCζ (the original codes in ref 37 were A, 1q; B, 1a; C, 1n) and the newly
designed pyrazoline scaffold as a rigidified analogue.
but rather found to be essential for the efficient activation of
Jak1 and the subsequent phosphorylation and nuclear trans-
location of Stat6.6,8
benzimidazole derivative was reported as a potent PKCζ
inhibitor which was less active toward PKCι; however, several
kinases from less related families were significantly inhibited.27
Recently, CRT0066854, a thieno[2,3-d]pyrimidine-based
chemical inhibitor of aPKCs, was reported. However, the
compound was about 4-fold more potent against PKCι than
PKCζ. In addition, it showed significant cross-reactivity with
PRK2 and CDK5.28 Titchenell et al. reported a series of non-
ATP-competitive aPKC inhibitors with unidentified binding
site, exhibiting IC50s in the low micromolar range.29 While
these 2-amino-3-carboxy-4-phenylthiophenes showed remark-
able selectivity against other PKC families, they inhibited both
aPKCι and aPKCζ with equal potency. Besides, it remains to be
shown whether the more than 3 orders of magnitude higher
activity of the compound series in cell-based vs cell-free assays
is due to inhibition of intracellular aPKCs or other
mechanisms.29
Inhibitors with a true allosteric mode of action are generally
believed to be more selective, because they are targeting less
conserved, often subtype-specific regulatory sites. However,
they frequently display only moderate affinities.30 Recently, the
first allosteric inhibitors of a PKC isoenzyme which bind to the
so-called PIF-pocket [also named the hydrophobic motif (HM)
pocket] of PKCζ were reported.31 These 4-benzimidazolyl-3-
phenylbutanoic acids exhibited a remarkable selectivity even
against the most closely related isoform, PKCι; however, the
potency (IC50 = 18 μM for compound A, Figure 1) required
optimization. Of note, the first efforts to develop PIF-pocket
ligands were directed to the phosphoinositide-dependent kinase
1 (PDK1) PIF-pocket, yielding the discovery and development
of compounds that are PDK1 activators and protein−protein
interaction inhibitors.32−36 The PIF-pockets of the AGC
kinases are highly conserved among species; e.g., the PIF-
pocket amino acid residues in rat and human PKCζ are
identical, but there are variable residues between different AGC
kinase isoforms that can be exploited to achieve isoform
selectivity; for instance, Leu328 in PKCζ is a phenylalanine in
PKCι.
Besides its direct implication in the control of transcription
factors such as NF-κB and Stat6, PKCζ was recently reported
to be involved in the regulation of gene expression at the
epigenetic level as well; it was found to physically interact with
and phosphorylate DNA methyltransferase 1, resulting in a
reduced activity of this enzyme.17 In agreement with this
observation, down-regulation of PKCζ signaling was identified
as a potential mechanism by which very low density
lipoproteins might silence pro-inflammatory genes in macro-
phage-like THP-1 cells, an effect which is known to be
mediated by de novo DNA methylation.18
The other atypical PKC, PKCι, was demonstrated to act as
an oncogene in lung cancer, colon carcinoma, and further types
of solid tumors (reviewed in ref 19). In a recent study, specific
inhibition of PKCι in the liver led to normalization of glucose
and triglyceride levels in vivo,20 thus potentially expanding the
scope for selective PKCι inhibitors to the metabolic syndrome
and to insulin-resistant states of obesity.
As a potential adverse side effect of PKCζ inhibitors,
impairment of the insulin−dependent glucose uptake might be
considered, because PKCζ was also demonstrated to play an
important role in the phosphorylation and activation of the
glucose transporter GLUT-4.21,22 However, all of these findings
were derived from experiments in rats or mice, whereas in
humans, PKCι and not PKCζ appears to be the major aPKC
expressed in skeletal muscle.23 Thus, PKCζ-selective inhibitors
might not suppress glucose uptake in human muscle tissue.
Alternatively, tissue-specific drug delivery might be a viable
approach to mitigate potential side effects.
In summary, an increasing body of evidence suggests that
PKCζ might be suited as a pharmacological target in allergic
and inflammatory diseases. Up to now, drug development
efforts toward this target were rather limited. Given the high
similarity of the ten PKC isoforms in their ATP-binding sites,
traditional approaches using ATP-competitive inhibitor strat-
egies did not appear very promising, because of the difficulty of
obtaining selective inhibitors.24 Hence, only a few small
molecule compounds were reported as PKCζ inhibitors, but
all with limited selectivity in particular against the most closely
related isoform, PKCι, including hydroxyphenyl-1-benzopyran-
4-ones25 and PKCzI257.3 [N-(4-((dimethylamino)methyl)-
benzyl)-1H-pyrrole-2-carboxamide].26 A 2-(6-phenylindazolyl)-
In the present study, we describe the design, synthesis, and
optimization of tri- and tetra-substituted pyrazoline derivatives
as novel allosteric inhibitors of PKCζ. Several compounds
potently inhibited the purified, recombinant enzyme and
blocked PKCζ-dependent signaling in U937 cells. Furthermore,
we demonstrate that the pyrazoline core is a versatile scaffold
6514
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
that allows independent fine-tuning of all substituents to
optimize potency, physicochemical properties, and stability.
RESULTS AND DISCUSSION
■
Design Concept. Some of the published allosteric PKCζ
inhibitors (e.g., those depicted in Figure 1B,C) were found to
be allosteric activators of PDK1 as well,37 suggesting that the
shapes of the PIF-pockets in active PDK1 and inactive PKCζ
might share some similarities. Moreover, since the binding
conformations of B and C in the PIF-pocket of PDK1 were
elucidated after cocrystallography,37 we reasoned that the
bound conformations of the small molecules could serve as
templates for the design of rigidified analogues. Such a V-
shaped conformation seemed compatible as well with binding
to the PIF-pocket of the active PKCι catalytic domain (PDB
code 3A8X) (see Figure 2A for a comparison with PDK1). For
PKCζ, no crystal structure has been published so far; however,
only two amino acids are different in the PIF-pockets of the
atypical PKCs: Phe321 and Asn282 in PKCι correspond to
Leu328 and His289 in PKCζ, respectively (Figure 2A).
A complementary V-shaped conformation could potentially
be achieved by utilizing pyrazoline as a central core; indeed, the
two aromatic rings in the 1- and 5-positions superimposed
quite well with the PIF-pocket-binding conformation of B
(Figure 2B). The presence of a sp3-hybridized atom at one of
the connection positions seemed important to conserve the
dihedral angle enclosed by the single bonds linked to the
aromatic rings in the PIF-pocket-binding conformation of B.
The presence of a carboxyl function was another potential
shortcoming of the previous inhibitor series with respect to cell
permeability. In this regard, the use of the pyrazoline scaffold
offered the advantage that the imine nitrogen could partially
replace the carboxyl function and its potential salt bridge by
establishing an H-bond interaction with the flexible Lys301 side
chain. Figure S2 in the Supporting Information provides a
flowchart illustrating the design concept and the subsequent
optimization steps.
Synthesis. The pyrazoline system was synthetically
accessible by a three-component reaction, allowing introduction
of three different substituents at positions 1, 3, and 5. This
flexible synthetic approach provided the possibility to optimize
all three positions independently using a diverse range of
substituted phenyls or other moieties. The synthesis of the
target compounds was carried out in two steps. In the first step,
a Claisen−Schmidt condensation was carried out between
aromatic aldehydes and pinacolone or acetophenone analogues
to afford the required enones. In the second step, these enones
were reacted to give the desired pyrazolines using the
previously reported regioselective synthesis of 1,3,5-triarylpyr-
azolines (Schemes 1−3).38 It is worth mentioning that in this
step, the use of the aryl hydazine as hydrochloride salt rather
than the free base was of great importance to increase the yield
of the cyclization reaction and to reduce the reaction side
products.
Figure 2. (A) Comparison of the PIF-pockets of PKCι and PDK1
(generated from the PDB entries 3A8W and 4A06, respectively). In
the atypical PKCι, two backbone carbonyl groups derived from Lys278
and Val281, respectively (red circles), are sequestered into the PIF-
subpocket1 (to the left) because the helix αB is distorted.
Contrastingly, in PDK1 the corresponding carbonyls (Lys115 and
Ile118) are involved in intrahelical interactions and thus are hardly
available for H-bonding with ligands. In addition, the electrostatic
potential in this region is much more negative in PKCι than in PDK1
(lower panel), also suggesting that H-bond donor moieties are
favorable in the respective subpocket of PKCι but not of PDK1. PKCι
is the most closely related isoform of PKCζ; the PIF-pocket residues
are identical except for Phe321 (Leu328 in PKCζ) and Asn282
(His289 in PKCζ); however, the latter residue seems too far away for
interactions with small ligands. Lys278, Val281, Val290, and Lys294
correspond to Lys285, Val288, Val297, and Lys301 in PKCζ,
respectively. The graphics were generated using PyMOL (V. 1.2,
www.pymol.org). In the PKCι structure, the intramolecularly bound
HM peptide was omitted for clarity. (B) Pyrazoline prototype
compound 3m (cyan) superimposed with the 3D structure of B
(orange) as extracted from the cocrystal with PDK1 (PDB entry
4A06). Note that the carboxyl side chain of compound B was not
resolved in the cocrystal structure.
In our early trials, p-hydroxybenzaldehyde was used in an
attempt to obtain the essential p-hydroxy group on the phenyl
enone derivatives. Unfortunately, the phenoxide anion, which
was generated due to the strong alkaline reaction conditions,
both hindered the reaction and reduced the yield dramatically.
Hence, we decided to use methoxy- or ethoxybenzaldehyde
derivatives, which afforded the respective enones in a very high
yield; the final compounds were then obtained after
deprotection using BBr3 (Scheme 2, step iii).
In an attempt to test the effect of other substituents at the 5-
phenyl, tert-butoxybenzaldehyde was used to get the enone E1
(Figure 3), which was then reacted with 4-chlorophenylhy-
drazine hydrochloride to yield compound D. Surprisingly,
compound 1a was the major product obtained while compound
D was obtained in a very low yield. This clearly suggested that
the tert-butyl ether group was spontaneously cleaved to give the
corresponding p-hydroxyl group under the reaction conditions
1
used. This fact was confirmed by comparing H NMR and13C
6515
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (i) 10% KOH, MeOH, ice cooling and then room temperature overnight; (ii) 1.5 equiv of Ar-NH-NH2·HCl, DMF, 85
°C, 5 h.
a
Scheme 2
a
Reagents and conditions: (i) 10% KOH, MeOH, ice cooling and then room temperature overnight; (ii) 1.5 equiv of Ar-NH-NH2·HCl, DMF, 85
°C, 5 h; (iii) 3−9 equiv of BBr3, CH2Cl2, −78 °C and then room temperature, 20 h.
NMR spectra of both E1 and 1a. The concomitant cleavage was
observed throughout the 33 reactions shown in Scheme 1, thus
providing a short and facile pathway for obtaining 5-(4-
hydroxyphenyl)pyrazolines in a good yield, eliminating the
need for deprotection and ether cleavage by BBr3.
The aniline derivatives were obtained in a very good yield by
synthesizing the respective nitro derivatives followed by
reduction using SnCl2·2H2O in refluxing methanol (Scheme
3). To introduce a methyl group at position 4 of the pyrazoline
derivative 3h, deprotonation of the acidic methylene group with
LDA was done to give the respective anion, which was instantly
trapped with methyl iodide to afford the methyl-substituted
pyrazoline derivative 8.38 This compound was then deprotected
using BBr3 to give compound 9 (Scheme 4). Oxidation of the
pyrazoline ring in 3e to the corresponding pyrazole10 was
achieved using DDQ in refluxing benzene followed by
deprotection using BBr3 to afford compound 11 (Scheme 5).
6516
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
a
Scheme 3
a
Reagents and conditions: (i) 10% KOH, MeOH, ice cooling and then room temperature overnight; (ii) 1.5 equiv of Ar-NH-NH2·HCl, DMF, 85
°C, 5 h; (iii) 5 equiv of SnCl2·2H2O, MeOH, reflux, 2 h; (iv) 1.5 equiv of CH3COCl, acetone, ice cooling, 2 h.
With respect to the determined IC50s, it should be noted that
the inhibition at compound concentrations lower than 0.2 μM
could not be accurately measured, because this was equal to the
concentration of the PKCζ protein in the assay.
The initially synthesized pyrazoline prototype compounds
(3m,n), were designed to have p-halogen substitution (at both
the 1- and 5-phenyl), which was found to enhance the potency
of the previously published PIF-pocket-directed PKCζ inhibitor
class (e.g., compounds A, B, and C in Figure 1) (see also the
flowchart in Figure S2, Supporting Information). However,
they showed only little activity toward the purified enzyme
Figure 3. Structure of the enone intermediate E1, which yielded 1a
after reacting with p-chlorophenylhydrazine, via a spontaneous ether
cleavage of D.
(Table 1). Structural information derived from the homologous
Biological Activity. Structure−Activity Relationships
PIF-pocket of the crystallized PKCι suggested that it might be
possible to establish H-bond interactions with at least one of
the exposed backbone carbonyl groups of Val281 and Lys278
(V288 and K285 in PKCζ), respectively (Figure 2A). Hence,
some pyrazoline compounds were synthesized in which the p-
halogen-substituted phenyl ring was kept on one hand and, on
(SAR) in the Cell-Free Assay. All the initially synthesized
compounds were screened against recombinant, full length
PKCζ at a concentration of 62.5 μM; for compounds showing
more than 75% inhibition, the IC50 was determined (Tables 1
and S1, Supporting Information). At a later stage, favorable
features identified in the first compound series were combined.
a
Scheme 4
a
Reagents and conditions: (i) 1.5 equiv of LDA, 1.5 equiv of MeI, THF, −78 °C and then room temperature for 20 h; (ii) 6 equiv of BBr3, CH2Cl2,
−78 °C and then room temperature, 20 h.
6517
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
a
Scheme 5
a
Reagents and conditions: (i) 1.5 equiv of DDQ, benzene, reflux, 5 h; (ii) 3 equiv of BBr3, CH2Cl2, −78 °C and then room temperature, 20 h.
the other hand, a phenol was introduced in an attempt to
address the potentially conserved carbonyl groups. In
comparison, both carbonyls in the equivalent PDK1 PIF-
subpocket are considerably less accessible − if at all − and the
molecular electrostatic potential of the PIF-pocket in PDK1
does not favor the interaction with H-bond donors (Figure
2A).
The introduction of the phenolic OH was a successful
approach to revive the potency where the hit compound 1a was
able inhibit PKCζ by an IC50 of 10.7 μM, thus showing
somewhat higher activity than the most potent compound from
the previous series of allosteric PKCζ inhibitors, compound A
(named 1q in ref 31), for which we determined an IC50 of 15.9
μM. This value was close to the previously reported IC50 for
this compound (18 μM).31
considerably lower than that of 1a, suggesting that the acetyl
moiety was not well tolerated by the PIF-subpocket.
Compound 4a, a positional isomer of 1a, in which the
hydroxyl group was shifted to the meta-position, was
synthesized to investigate the optimal position of the hydroxyl
group. This modification led to a modest increase of activity by
roughly 2-fold, suggesting that a more efficient hydrogen
bonding might occur from the meta-position. In compounds
synthesized later, after optimization at other positions, it was
therefore logical to include a compound with a 3,4-dihydroxy
function (catechol derivative) to test if this might further
improve the potency by establishing a bidentate H-bond
interaction with the putative carbonyl oxygen or by targeting
even two different carbonyl functions. Indeed, compound 4g
showed a more than 15-fold increase in activity when compared
to the 4-hydroxy congener 2c.
To further increase the potency through fine-tuning at the
phenol ring, we decided to synthesize compounds 4d and 4e
bearing an additional m-fluoro or m-chloro substituent adjacent
to the essential 4-hydroxyl group. Both halogens triggered
superior activity; while the additional m-fluoro increased the
potency about 3-fold (4d compared to 1a), the m-chloro led to
a dramatic increase in activity by more than 15-fold (4e
compared to 1a). This beneficial effect of the neighboring
halogen substitution can be attributed to two major effects:
first, the electron-withdrawing effect, in particular of the
fluorine, increased the partial positive charge on the hydroxyl
hydrogen, thus enhancing the strength of the hydrogen bond.
Second, the presence of the halogen at the aromatic ring
restored its lipophilicity, thus increasing the entropy of binding
to the rather hydrophobic pocket. Obviously, the second effect
prevailed, as indicated by the larger impact of the chlorine. As
another potential benefit, this meta-halogen substitution was
also expected to improve the stability of the hydroxyl group
toward metabolic phase II conjugation reactions, which might
be important for in vivo activity.39,40
The 1-Phenyl. To test the importance of the halogen
substitution, we synthesized compound 1u, which lacked the 4-
chloro of 1a. The higher IC50 showed that the 4-chloro is
important for activity, which was consistent with the previous
observation with PIF-pocket-directed PKCζ inhibitors, in
which ring substitutions with halogens or alkyl had a large
impact on potency and possibly also on the selectivity.31 Several
compounds bearing various lipophilic electron-withdrawing
substituents, comprising 4-fluoro, 4-bromo, and 4-trifluoro-
methyl, were prepared (compounds 1b,c,d, respectively). All of
them showed inhibitory activity comparable to that of the 4-
chloro analogue (1a), and this finding was consistent with the
result obtained for compounds 2g,h,i, also substituted with 4-
chloro, 4-bromo, and 4-trifluoromethyl at the 1-phenyl ring,
Using the hit compound 1a as a template, numerous sites of
modification around the pyrazoline ring system were explored
in a trial to improve the inhibitory potency, as described in the
following (also illustrated in Figure S2, Supporting Informa-
tion).
The 5-Phenyl. As observed already in the initial screening,
the phenolic OH at the para-position of 1a was found to be
essential for activity; like chlorine and bromine in the prototype
compounds (3m,n), fluorine could not successfully replace this
OH either (compound 3l). Next, we tested several H-bond
acceptor (HBA) functions, including the lipophilic and
electron-withdrawing nitro function as well as the more
hydrophilic cyano group (compounds 5a and 3o, respectively),
which, however, completely abolished the inhibitory activity.
Replacing the phenolic OH in compound 2a by a methoxy
function (compound 3p) resulted in an almost inactive
compound as well. Altogether, these findings indicated that
the crucial role of the hydroxyl function came from its H-bond
donating ability. In the light of these results, the amino group
was conceivable as a possible bioisoster. Therefore, we
synthesized a series of aniline derivatives including compounds
6a−n. Surprisingly, none of these aromatic amine compounds
showed an activity comparable to that of the hit 1a (Table S1,
Supporting Information), giving clear indication that the
primary amine function cannot replace the hydroxyl in its
interaction with the target amino acid(s). Shifting the amino
function to the meta-position (compound 6c) did not recover
the potency either. The largely reduced activity might be
attributed to the lower δ+ character of the amine hydrogen
atoms or to the slightly different spatial orientation of the
aromatic amine. To increase the H-bond donor (HBD)
capacity of the amino function, the amino group in compound
6a was acetylated to yield the acetamido analogue of 1a
(compound 7a). This modification recovered some of the
activity compared with 6a, but the potency was still
6518
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
Table 1. Inhibition of Recombinant PKCζ and the NF-κB Pathway in Cells
NF-κB reporter gene assay
(U937 cells)
cell-free assay
% inhibn at
IC50 SD
(μM)
% inhibn at
IC50 SD
(μM)
a
a
compd
R1
R2
R3
R4
62.5 μM
5 μM
1a
1b
1c
1d
1e
1f
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-OH
4-F
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
phenyl
phenyl
4-Cl
H
91.5
95.5
93.2
94.2
88.8
54.9
47.9
96.8
98.1
96.1
91.7
57.1
57.6
71.0
68.0
59.9
65.5
98.8
97.8
91.7
84.3
95.6
90.5
92.0
68.4
42.8
90.8
97.0
95.2
99.4
82.2
10.7 0.54
9.4 0.19
11.4 1.25
8.8 0.53
12.6 1.13
ND
75.1
63.8
58.4
3.2 0.22
a
4-F
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
ND
4-Br
ND
b
4-CF3
4-CH3
ND
62.9
ND
ND
b
4-isopropyl
4-COOH
3-Cl
ND
40.7
64.6
73.5
89.7
73.7
23.5
48.7
50.2
51.8
41.5
12.3
92.0
90.1
84.2
38.5
50.8
42.2
68.2
26.5
48.2
38.7
52.8
64.2
76.3
70.2
54.1
42.3
3.7
ND
1g
1h
1i
ND
ND
5.2 0.67
2.2 0.09
2.7 0.08
3.5 0.35
ND
ND
3-F
ND
1j
3-CF3
3-CH3
2-Cl
2.7 0.32
ND
1k
1l
ND
1m
1n
1o
1p
1q
1r
1s
1t
2-F
ND
ND
2,4-dichloro
2,4-difluoro
2,4-dimethyl
2,6-dichloro
3-chloro-4-fluoro
3,4-difluoro
2,3,4-trifluoro
4-H
ND
ND
ND
ND
ND
ND
ND
ND
0.9 0.06
0.5 0.06
0.7 0.1
28.8 1.72
3.5 0.21
2.1 0.17
1.5 0.14
ND
1.9 0.18
0.9 0.13
3.1 0.12
ND
1u
2a
2b
2c
2d
2e
2f
4-Cl
ND
3-Cl
ND
2-hydroxyphenyl
2-methoxyphenyl
2-ethoxyphenyl
2-chlorophenyl
thiophen-2-yl
thiophen-2-yl
thiophen-2-yl
pyridin-2-yl
pyrrol-2-yl
4-aminophenyl
t-Bu
4-Cl
ND
4-Cl
ND
4-Cl
ND
ND
4-Cl
12.7 1.5
0.6 0.04
0.5 0.04
0.4 0.05
6.2 0.12
0.5 0.06
5.4 0.38
ND
ND
2g
2h
2i
4-Cl
ND
4-Br
ND
4-CF3
3-Cl
2.9 0.32
ND
2j
c
2k
2l
3-Cl
63.0
ND
4-Cl
98.0
0
ND
3l
4-Cl
ND
3m
3n
3o
3p
3q
4a
4b
4c
4d
4e
4f
4-Cl
t-Bu
4-Cl
11.8
39.0
20.7
21.1
58.8
95.6
40.9
72.2
98.5
94.5
ND
5.5
ND
4-Br
t-Bu
4-Cl
ND
9.3
ND
4-CN
4-OMe
4-OH
3-OH
4-OH
4-Cl
t-Bu
4-Cl
ND
20.1
2.9
ND
phenyl
4-Cl
ND
ND
Me
4-Cl
ND
50.8
89.9
35.0
37.3
95
ND
t-Bu
4-Cl
5.3 0.42
ND
2.0 0.14
ND
t-Bu
4-OH
4-OH
4-Cl
t-Bu
ND
ND
3-F,4-OH
t-Bu
3.8 0.49
0.7 0.11
0.3 0.04
<0.1
3.7 0.51
2.9 0.26
2.5 0.2
ND
3-Cl,4-OH
t-Bu
4-Cl
85
c
4-F,3-OH
t-Bu
3-Cl
75.0
92.8
62.0
71.7
75.1
81.7
87.4
c
4g
4h
4i
3,4-dihydroxy
3-F,4-OH
2-hydroxyphenyl
2-hydroxyphenyl
2-hydroxyphenyl
2-hydroxyphenyl
2,4-dihydroxyphenyl
4-Cl
89.7
c
3-Cl
64.7
2.9 0.26
11.1 0.44
6.6 0.13
<0.1
ND
d
3,5-difluoro-4-OH
2,3-difluoro-4-OH
3-F,4-OH
3-Cl
62.0
2.5 0.3
1.4 0.1
1.1 0.05
d
4j
3-Cl
69.4
c
4k
3-Cl
89.0
6519
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
Table 1. continued
NF-κB reporter gene assay
(U937 cells)
cell-free assay
IC50 SD
% inhibn at
% inhibn at
IC50 SD
(μM)
a
a
compd
R1
R2
R3
R4
62.5 μM
(μM)
5 μM
5a
7a
9
4-NO2
t-Bu
t-Bu
4-Cl
4-Cl
3-Cl
−
H
0
ND
5.9
ND
4-NHCOCH3
H
45.1
ND
25.5
73.0
93.0
6.4
ND
c
3-F,4-OH
2-hydroxyphenyl
CH3
81.8
0.72 0.05
17.4 1.9
15.9 2.3
ND
d
11
A
−
−
−
−
−
−
54.0
3.0 0.34
ND
d
−
54.6
a
b
Values are the mean of at least two experiments; standard deviation <15%. Compounds were toxic in the MTT assay (Table S4, Supporting
Information). Percent inhibition is shown at 10 μM. Percent inhibition is shown at 20 μM; ND: not determined.
c
d
respectively (further discussed below). Interestingly, replacing
the 4-chloro in 1a by 4-methyl (compound 1e) kept the activity
at a similar level, suggesting that mainly the lipophilic character
of the 4-substituent rather than its electronegativity is the
important factor for the improvement of the activity. This was
confirmed by synthesizing compounds 1g and 4b with polar
hydrophilic substituents (4-carboxy and 4-hydroxy, respec-
tively). The presence of such polar groups was obviously
deleterious to the activity, in agreement with their putative
binding to the entirely hydrophobic subpocket 2 lacking any
accessible HBD/HBA function (Figures 2 and 4, binding
model). However, sensitivity to steric hindrance at the 4-
position of this aromatic ring with regard to potency was found,
as replacing the 4-methyl in compound 1e by a bulkier
branched alkyl like isopropyl (compound 1f) significantly
decreased the activity.
Moving the lipophilic substituent from the para- to the meta-
position brought this group in an even better position for
interaction with the pocket; this was evident from compounds
1h,i,j,k, substituted by m-chloro, m-fluoro, m-trifluoromethyl,
and m-methyl, respectively, which showed about 2−4 times
higher potencies when compared to the para-substituted
analogues (1a,b and 1d,e). The finding that both the para-
and meta-substitution had a boosting effect prompted us to
synthesize some hybrid compounds with 3,4-disubstitutions.
Interestingly, compounds 1r, with 3-chloro-4-fluoro, and 1s,
with 3,4-difluoro-substituents, were among the most active
compounds, with submicromolar IC50s, indicating at least
additive effects of the lipophilic substituents at these two
positions. On the other hand, lipophilic substitution of the
ortho-position as in the case of compounds 1l (o-chloro) and
1m (o-fluoro) dramatically reduced the activity when compared
to the corresponding meta-substituted analogues 1h and 1i.
This unfavorable effect of the ortho-substitution was also
confirmed by the disubstituted analogues 1n (2,4-dichloro), 1o
(2,4-difluoro), 1p (2,4-dimethyl), and 1q (2,6-dichloro). This
was unexpected, since it was assumed that the ortho effect
would actually stabilize the biologically active conformation and
hence increase the binding affinity. It could be speculated that a
steric clash of the ortho-substituent with the border of the
hydrophobic subpocket might hinder binding of the com-
pounds or might push away the helix αC, thus impairing the
putative H-bond interaction of the pyrazoline imine with the
helix residue Lys301 (cf. Figure 4).
However, in compound 1t with 2,3,4-trifluoro substitution,
the effect of the o-fluoro on activity was minimal, maintaining a
submicromolar IC50 similar to that of the potent 3,4-difluoro
analogue (compound 1s). In 1t, favorable synergistic effects of
the p- and m-fluorine substitution seemed to outbalance the
adverse effect observed with the isolated ortho-substitution.
Such favorable interaction could potentially be explained by
dipole−dipole interactions with the peptide backbone from
Leu328 of the β-sheet bordering subpocket 2.
Substituent at Position 3. To explore the relative
importance of the bulky tert-butyl substituent in hit compound
1a, we started our modifications by contracting it to a methyl
group (compound 3q). Rather unexpectedly, this modification
caused a considerable drop of activity. This result was not
anticipated, because the third substituent of the pyrazoline core
was not expected to bind to the PIF-pocket, rather the binding
model predicted that it would point outside the pocket.
Nevertheless, the favorable effect of the tert-butyl group on the
Figure 4. Model for the hypothetical interaction of the triphenylpyr-
azoline analogues as supported by the SAR. The exemplary compound
2b (yellow) was docked into the PIF-pocket of PKCζ that was
modeled based on the PKCι coordinates (PDB entry 3A8X) using
MOE. In the PKCζ model, the PIF-pocket is formed by two mostly
hydrophobic subpockets (SP1 and SP2, red arrows) which
accommodate two phenyl rings (1-phenyl and 5-phenyl) of 2b. The
essential hydroxyl function interacts with the backbone carbonyl of
Val288, thus determining the binding orientation. The 1-(3-
chlorophenyl) ring binds to the merely hydrophobic subpocket 2
(SP2). Potential interactions with Lys301 comprise H-bonding with
the imine nitrogen and a cation−π interaction with the 3-phenyl ring.
White dotted lines indicate the electrostatic interactions with the
numbers denoting the distances (in Å) between the hydrogen and the
electron−rich partner.
6520
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
affinity could be explained on the basis of the binding model.
First, the strong positive inductive effect of the t-Bu might be
propagated via the conjugated double bond, thus increasing the
electron density and hence the H-bond acceptor strength of the
pyrazoline nitrogen. However, an even more important, parallel
effect of the bulky t-Bu group might be the shielding of this
potential H-bond with the pyrazoline nitrogen from water
molecules. This might be particularly relevant in the present
case of a highly water accessible surface pocket, where H-bonds
are strongly competed for and weakened by water molecules.
Schmidtke et al. showed that shielding of such an H-bond by an
ethyl “lid” provided by the ligand created a more lipophilic
binding groove that significantly stabilized and enhanced the H-
bond interaction.41 In the future, this concept might receive
broad attention, especially in the growing field of protein−
protein interaction inhibitors, though it might not seem logical
at first sight to introduce a (bulky) lipophilic residue at a water-
exposed position.
Replacing the t-Bu by an unsubstituted phenyl increased the
activity about 3 times (compounds 2a,b compared to 1a,h),
which might be explained by additional interactions exerted by
the aromatic π-electrons, such as CH−π or cation−π
interactions. Interestingly, it was previously reported that
such a cation−π interaction effectively enhanced the binding
affinity of a ligand by up to 8 kJ/mol at a fully solvent exposed
binding site, proving that such an interaction is not attenuated
by hydration effects.42 It might thus successfully substitute H-
bond interactions in cases where the binding site cannot be
protected from water molecules.
A further 2-fold increase in activity was achieved by
substituting this phenyl with a hydroxyl group at the o-position
(compound 2c compared to 2a), which might be attributed at
least partially to the mesomeric enhancement of the π-electron
system. In addition, calculations of the lowest energy conformer
of 2c (Figure S1, Supporting Information) suggested that the
hydroxyl forms an intramolecular H-bond with the pyrazoline
nitrogen, thus fixing the phenyl ring in a favorable position, also
resulting in a smaller entropic penalty upon ligand binding.
Furthermore, the hydroxyl oxygen is potentially available as an
alternative HBA for Lys301 (cf Figure 4, binding model). In
contrast, the loss of this conformational constraint in the
methoxy- and ethoxy-substituted analogues (2d and 2e,
respectively) was accompanied by a gradual decrease of the
potency. As could be confirmed by ab initio calculations, the
ether oxygen in these compounds was preferably pointing to
the opposite side due to electrostatic repulsion with the
pyrazoline nitrogen (Figure S1, Supporting Information).
Substitution of the 4-position with an amino group (2l) did
not increase the activity compared to the unsubstituted phenyl
(2a), apparently contradicting the presumed cation−π
interaction. However, a closer inspection of the calculated
electrostatic potential surface (Figure S1, Supporting Informa-
tion) revealed that the strong electron-donating effect of the
amino function slightly increased the electron density, even at
the p-hydroxy on the 5-phenyl, thus negatively affecting its H-
bond donor strength. The same effect was also observed in the
case of the 2-methoxy and 2-ethoxy substituents (2d,e), but not
with the 2-hydroxy analogue 2c or the unsubstituted analogue
2a, providing another explanation for the substantially
decreased activity of the two alkoxy derivatives.
reduction in potency (compound 2f), also corroborating that a
higher electron density was favorable for the activity.
Next, some heteroaryl rings were introduced instead of the
phenyl in compounds 2a,b, starting with a mildly basic
heterocycle like 2-pyridyl (compound 2j), which would
improve the physicochemical properties and could potentially
offer its nitrogen as HBA. However, the activity was reduced by
about 3-fold compared with that of 2b, indicating that an
electron-poor heterocycle was inferior to phenyl. Moreover, the
pyridine nitrogen apparently could not replace the pyrazoline
imine as HBA, which might be due to electrostatic repulsion
between the lone pairs of the nitrogen atoms, forcing the
pyridine nitrogen to rotate to a position opposite to the imine
(Figure S1, Supporting Information). On the other hand, the
thiophene ring in compounds 2g−i caused remarkable
increases in activity by more than 6-fold when compared to
the unsubstituted phenyl analogue (compare 2g with 2a). A
similar improvement was observed with the pyrrole analogue
2k (compared to 2b). The higher potency as conferred by the
five-membered heterocycles compared with the phenyl might
be attributed to the higher electron polarizability of these ring
systems, enabling superior van der Waals and π-electron-
dependent interactions. In summary, our SAR were fully
consistent with a cation−π interaction, as the π-electron density
mostly correlated with the potency, with the only exception
being 2l, as discussed above.
As can be concluded from our SAR, the electronic properties
and bulkiness of the substituents in position 3 allow the
adjustment of the electron density and water accessibility of the
pyrazoline nitrogen, respectively, but can also slightly influence
the H-bond donor capacity of the essential hydroxyl function.
Aromatic systems could serve as π-electron donors to Lys301 in
addition. Furthermore, since the 3-position of the pyrazoline
ring tended to be very tolerant with respect to a variety of
substituents like bulky alkyls, hydroxylated phenyl, and
heteroaryl rings, it was identified as a key position for the
optimization of the physicochemical properties such as water
solubility and log P.
Finally, the proposed binding model (Figure 4) was further
confirmed by synthesizing the inverted analogue of 1a in which
the substituents of the two aromatic rings were reversed
(compound 4c). This inversion resulted in a less active
compound, suggesting that the imine nitrogen interaction
might be lost if the phenol ring still binds to the same position
as with 1a, thus accounting for the drop in potency. In contrast,
the alternative binding mode with the phenol ring docking to
the other subpocket was unlikely, bearing in mind the huge loss
of activity observed with the bisphenol compound 4b.
Some Combined Favorable Features and Further
Modifications. In order to further enhance the binding
affinity, it was straightforward to combine some of the favorable
modifications identified in the compounds described above. For
instance, while the t-Bu was kept at position 3 of the pyrazoline,
the essential phenolic OH was shifted to the meta-position,
combined with an adjacent p-fluoro substituent, and a 3-chloro-
substituent was placed at the 1-phenyl. The resulting
compound 4f proved to be one of the most potent compounds,
exhibiting an IC50 of 0.3 μM. In addition, four compounds were
prepared (4h−k) in which we used the 2-hydroxyphenyl at
position 3 of the pyrazoline, as it had augmented the activity
before (compound 2c). As another beneficial effect, we
expected a reduction of the log P, while the o-hydroxyl should
be rather protected from phase II conjugation.43 We first
Replacing this 2-hydroxyl in 2c by a lipophilic, electron-
withdrawing group like chlorine led to a more than 8-fold
6521
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
a
combined this moiety with the 3-fluoro-4-hydroxyphenyl at the
5-position of pyrazoline (4h). However, in this case the fluorine
did not increase the activity in contrast to the 3-t-Bu analogue
4d. In compounds 4i,j, we added a second fluorine compared
with compound 4h, to yield a difluoro substitution adjacent to
the essential p-hydroxyl group at the 4-phenyl residue.
Unexpectedly, in compound 4i this second fluorine in 5-
position of the phenyl caused a more than 4-fold decrease in
activity compared with 4h, while in 4j the second fluorine in the
ortho-position still reduced the activity by half. Although this
difluoro substitution in the vicinity of the essential phenolic
OH was expected to increase its HBD ability, it decreases at the
same time the electron density on the aromatic ring, which
might have attenuated putative van der Waals and CH−π
interactions with the aliphatic PIF-pocket residues.
Table 2. Inhibition of PKCι at 20 μM
PKCι %
inhibn at
20 μM
PKCι %
inhibn at
20 μM
PKCι %
inhibn at
20 μM
compd
compd
compd
1a
1h
1i
0
2a
2b
2c
2h
2i
31.6
45.5
13.4
21.4
45.9
24.1
20.3
38.7
4a
4d
4e
4g
4h
4i
17.4
29.7
0
43.5
19.3
17.0
13.9
31.6
23.7
7.8
1j
44.1
25.2
45.2
39.2
23.1
1k
1r
1s
1t
2j
2k
2l
4j
9
a
Values are the mean of at least two experiments; standard deviation
<15%.
In compound 4k, an additional 4-hydroxyl group was
introduced to the 3-phenyl. Compared with 4h, this additional
OH caused a dramatic increase in activity by more than 29-fold,
yielding one of the most potent compounds of the present
series. This order of magnitude corresponded to an affinity gain
of more than 8.5 kJ/mol as estimated on the basis of the IC50s,
suggesting that the p-hydroxyl could be involved in a potential
H-bond with the target site, too.
Introduction of a Fourth Alkyl Substituent at Position
4 of the Pyrazoline. Afterward we decided to test the effect of
introducing a fourth small alkyl substituent such as methyl at
position 4 of the pyrazoline core. Hence, we prepared
compound 9 as the 4-methyl analogue of compound 4h.
Interestingly, this 4-methyl led to more than a 4-fold increase in
activity compared with 4h, which could be explained either by
additional hydrophobic interactions or by the effect of this 4-
substitution on the relative orientation of the neighboring
groups arising from the pyrazoline core or by a combination of
both effects. It is worth mentioning that alkylation of 1,3,5-
trisubstituted pyrazolines at position 4 was reported to increase
the chemical stability of the ring toward oxidation to the
corresponding pyrazole.38
Core Ring Planarity. To test the effect of core planarity
(pyrazoline vs pyrazole), the pyrazole derivative compound 11
was prepared by oxidation of compound 4e. Switching to the
planar pyrazole system caused more than a 24-fold drop in
activity (11 compared to 4e), confirming the importance of the
tetrahedral sp3-hybridized atom (C5) to conserve the dihedral
angle enclosed by 1- and 5-phenyl for the biologically active
conformation. Moreover, this finding was in agreement with a
non-ATP-competitive binding mechanism of the pyrazolines,
because if the ATP-binding pocket was the target site,
conversion to the flat pyrazole ring system should have rather
increased the binding affinity. Indeed, trisubstituted pyrazole
compounds were reported to be potent ATP-competitive
inhibitors of several serine/threonine kinases.44−47
other hand, the PIF-pocket in PKCι can be regarded as a kind
of PKCζ PIF-pocket mutant which differs in only two amino
acid residues (Leu328 is Phe and His289 is Asn in PKCι,
respectively).37 This difference in the PIF-pocket might account
for the lack of activity against PKCι, suggesting that our
compounds were targeting the PIF-pocket in PKCζ. It is worth
mentioning that even a single mutation of Leu328 into Phe was
sufficient to abolish the inhibitory activity of the previously
published allosteric inhibitors (PS186 and PS171 in ref 37).
Furthermore, compound 4f, the most potent compound among
the t-Bu analogues, was screened against all PKC isoforms and
some of the related AGC kinases. Interestingly, 4f did not show
a significant inhibition of any of the tested kinases at 10 μM
(Table S5, Supporting Information). Again, this selectivity for
PKCζ vs other PKC isoforms can be explained by the bulkier
Phe residue present in the PIF pocket of all PKCs, which is
replaced by the smaller Leu only in PKCζ.
Consistent with this assumption, a more direct evidence for
the allosteric mechanism and targeting of the PIF-pocket was
obtained by testing the inhibitory potency of 1a and three of
the more potent congeners (2c, 2h, and 4e) against PKCζ
mutated at a residue central to the PIF-pocket (PKCζ-
[Val297Leu]) (Table 3). The inhibitory potencies of
Table 3. Inhibition of PKCζ[Val297Leu] vs PKCζ[wt]
a
(Wild-Type)
IC50 (μM)
compd PKCζ[Val297Leu] PKCζ[wt] relative potency (wt vs mutant)
1a
2c
2h
4e
26.7
3.1
10.7
1.5
2.5
2.1
1.0
0.5
2.0
14.8
0.7
20.8
a
Values are the mean of at least two experiments; standard deviation
<15%.
Selectivity and Proof of the Allosteric Mechanism. We
screened 1a and most of the active compounds which showed
IC50s below 10 μM for their ability to inhibit the most closely
related isoform, PKCι (Table 2). Remarkably, 1a did not show
any inhibition, even at 40 μM, and all the compounds which
had shown submicromolar IC50s toward PKCζ did not exhibit
appreciable inhibition of PKCι at 20 μM, indicating that these
compounds were highly selective for PKCζ. In addition, this
finding strongly argued for a non-ATP-competitive mechanism,
since the ATP binding sites of both atypical PKCs are lined by
almost identical residues, rendering such a degree of selectivity
for ATP-pocket-directed compounds rather unlikely. On the
compounds 1a, 2c, and 2h were reduced by a factor of more
than 2, while for compound 4e, we noticed a more than 20-fold
increase of the IC50 against the mutant. The latter finding might
be explained by a steric clash of the m-chloro with the larger
Leu297 residue in the mutated PIF-pocket and was in
agreement with the proposed binding model as well (cf. Figure
4).
To further investigate the non-ATP-competitive mechanism,
we performed reporter dye assays using SYPRO Orange, a dye
that is known to change fluorescence dramatically when it binds
to hydrophobic patches of proteins, thus being applicable as a
6522
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
Figure 5. Allosteric inhibitors of PKCζ increase the exposure of hydrophobic patches. The fluorescence of the indicator dye SYPRO Orange was
measured in the presence of recombinant PKCζ (0.5 μM) and different concentrations of some pyrazoline compounds (1a, 2i, 4f), the previously
published allosteric inhibitor A, the ATP-competitive inhibitor Ro31-8220 (IX), or DMSO as a control. From each plotted value, the respective
fluorescence intensity obtained with the identical mixture but in the absence of PKCζ protein was subtracted. All reaction mixtures were incubated
for 4 min at 37 °C before measurement. The data shown are representative for a set of three independent experiments which gave essentially the
same results. Indicated are the relative fluorescence units with standard deviations.
conformational sensor (Figure 5).48 Our results suggested that
compounds 1a, 2i, and 4f as well as the previously published
allosteric inhibitor A could induce the exposure of hydrophobic
regions in the protein, to which the dye bound, thereby
increasing its fluorescence. In contrast, the ATP-competitive
PKC inhibitor Ro31-822049 did not trigger this effect. It is
likely that the allosteric inhibitors stabilize an inactive protein
conformation, which renders hydrophobic residues accessible
for the reporter dye.
In addition, HEK293 cell transfection experiments using
PKCζ wild-type and [Val297Leu] mutant expression plasmids
were performed, to investigate whether the compounds would
inhibit the activation of the target kinase in intact cells. After
transfection, the cells were treated with the allosteric inhibitor
4f and the ATP-competitive compound Ro31-8220, respec-
tively, in serum-free medium. After stimulation of the PI3K/
PDK1 signaling pathway by IGF-I, the recombinant proteins
were isolated from the cells and the enzymatic activity was
measured after extensive washing. We found that treatment of
the cells with 4f reduced the specific activity of the isolated
wild-type PKCζ in a concentration-dependent manner (Figure
6). This suggested that the conformational change for which we
provided evidence in the SYPRO Orange binding experiment
might either have prevented the activation by the upstream
kinase PDK1 or might have triggered deactivation by cellular
regulatory mechanisms, e.g., due to dephosphorylation. Such a
mechanism was observed earlier with allosteric inhibitors and
type II inhibitors, which shift the intracellular equilibrium
toward inactive conformations of the target kinases, thereby
promoting dephosphorylation, as exemplified for a type II
inhibitor of PDK1.50 In contrast, the enzymatic activity of the
PKCζ protein isolated from the Ro31-8220-treated cells was
not significantly altered compared to the DMSO-treated
control. Ro31-8220 was chosen for this transfection experiment
because it is more selective for PKCs than staurosporine and
does not inhibit PI3-kinase/PDK1-dependent signaling,51
which could have artificially blocked the potential intracellular
rephosphorylation and activation of the recombinant enzyme
by PDK1 during IGF-I stimulation. In addition, Ro31-8220 was
Figure 6. Compound 4f induces deactivation of PKCζ in cells.
HEK293 cells were transfected with GST-PKCζ wild-type and
[Val297Leu] mutant expression plasmids, and incubated with 4f, the
bisindolylmaleimide Ro31-8220, or DMSO, respectively, for 6 h in
serum-free medium. After stimulation with IGF-I for 30 min, the cells
were lysed and the recombinant proteins pulled down using
glutathione agarose. The purified PKCζ was eluted from the beads
using glutathione, and the enzymatic activity was measured in a
phosphorylation assay. All treatments were done in triplicate; shown
are the representative values of two independently performed
experiments.
reported to effectively inhibit PKCζ and PKCι in transfected
cells at concentrations of 2−3 μM.52,53
In order to prove that the sustained reduction of the PKCζ
activity was specifically dependent on the PIF-pocket, the same
experiment was performed in parallel using the PKCζ-
[Val297Leu] mutant expression plasmid. Indeed, treatment of
the transfected cells with 4f did not affect the specific activity of
the mutant PKCζ relative to the DMSO-treated control
samples, demonstrating that deactivation of the kinase required
an intact PIF-pocket. Altogether, our experimental data strongly
suggested that the allosteric PKCζ inhibitors acted via binding
6523
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
Figure 7. Influence of 4f on the activation loop phosphorylations of PKCζ vs PKCβ. U937 cells were serum-starved overnight, incubated with
compounds at the indicated concentrations or DMSO, and stimulated either by addition of 50 ng/μL TNFα for 30 min (A) or of 100 nM TPA for
10 min (B). The total cellular proteins were isolated and analyzed by immunoblotting using antibodies detecting the activation loop
phosphorylations. As a control for the amount of protein loaded, the blots were reprobed with anti-β-actin antibody. In parts A and B, one
representative result out of two separate experiments is shown.
to the PIF-pocket, both in the cell-free assay and in a cellular
setting.
the reporter enzyme with an IC50 of 3.2 μM (Table 1). Initially,
we checked whether the inhibition of another kinase known to
participate in the activation of NF-κB could have been
responsible for the observed effect. Hence, 1a was tested
toward all kinases associated with the TNF receptor-1 signaling
complex, the composition of which in U937 cells had been
elucidated.55 It was found that 1a did not inhibit any of the
purified kinases at a concentration of 10 μM (Table S2,
Supporting Information).
We then analyzed whether there was a correlation between
the compounds’ potencies under cellular and cell-free
conditions. To this end, the compounds were grouped into
high-, moderate-, and low-activity categories according to their
potencies to inhibit purified PKCζ (see Table S3, Supporting
Information). Then the Wilcoxon’s signed rank test was applied
Effects on the NF-κB Signaling Pathway in Intact
Cells. To analyze effects on downstream signaling, the U937
cell line was chosen, where PKCζ is strongly expressed and
essential for the NF-κB activation.10,54 Transfections were
performed using a reporter gene plasmid expressing luciferase
under control of NF-κB response elements to monitor PKCζ-
induced effects on NF-κB activation. All the compounds were
tested in the reporter gene assay at a 5 μM concentration, and
the IC50 was determined for compounds showing 75%
inhibition or higher (Table 1).
From the very beginning, we verified that the hit compound
1a was effective in this cell-based assay. We found that 1a was
able to disrupt the NF-κB activation cascade as monitored by
6524
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
by pairwise comparison of the corresponding cellular potencies.
It was found that compounds with low potency toward purified
PKCζ also displayed a significantly lower inhibition of the NF-
κB activation than compounds showing a moderate (P = 0.002)
or a strong inhibition (P = 0.000 17) of purified PKCζ. The
degree of significance was somewhat lower when comparing the
groups with moderate and high inhibition of PKCζ in the cell
free assay (P = 0.0273). However, the latter significance was
obtained only after excluding the compounds with plain 3-aryl
and heteroaryl moieties from the analysis, all of which had
uniformly impaired the cellular activity of the compounds when
compared to the 3-t-Bu analogues (see below). Altogether, the
statistical significance corroborated that PKCζ was also the
intracellular target of the pyrazoline compounds, consistent
with our data from the HEK293 cell transfection experiment.
As already anticipated by the statistical analysis, many
structural modifications caused the same trend both for the
cell-free and the cellular potencies (Table 1). In particular, the
phenolic OH at the 1-phenyl was essential for PKCζ cellular
inhibitory activity as well, since all compounds lacking the
phenolic OH were inactive (cf. compounds 3l−p and 5a).
Similarly, the aniline derivatives 6a−d and 6h were all less
active than their respective phenolic analogues (Table S1,
Supporting Information).
Influence on the Activation Loop Phosphorylations of
PKCζ vs PKCβ in Intact Cells. Next we wanted to analyze
whether the allosteric inhibitor 4f selectively blocked the
activation of PKCζ under the conditions used in the NF-κB
reporter gene assay. PKCβ was chosen as a control isoenzyme
that should not be affected by the inhibitor. It was previously
shown that TNFα signaling can enhance the interaction of the
scaffold protein p62 with PKCζ, which triggers an increased
phosphorylation of PKCζ at Thr410 in the activation loop (also
called the “T-loop”).56,57 Hence, in the first experiment, U937
cells were starved overnight, incubated with the test
compounds or controls, and stimulated by the addition of
TNFα. The total cellular proteins were then analyzed by
immunoblotting using antibodies that recognize the phosphory-
lated activation loops of PKCζ and PKCβI/II, respectively. It
was observed that 4f but not the ATP-competitive inhibitor
Ro31-8220 prevented the T-loop phosphorylation following
TNFα-treatment (Figure 7A). This result suggested that the
conformational change induced by 4f, as discussed above,
entailed a reduced phosphorylation of PKCζ at the activation
loop, thus diminishing the catalytic activity.
However, to our surprise, the equivalent phosphorylation at
the activation segment of PKCβI/II was also inhibited. Two
explanations were conceivable for this finding: (i) 4f was not
selective for PKCζ in a cellular environmentin contrast to
the cell-free conditionsor (ii) the phosphorylation and
activation of PKCβ were regulated dependent of the PKCζ
activity. To corroborate either of these hypotheses, the
experimental conditions were changed in order to analyze the
effect of the inhibitors in a distinct signaling pathway. The
phosphorylation of conventional PKCs, e.g., PKCα and PKCβ,
was reported to be enhanced in different cell lines after
treatment with 12-O-tetradecanoylphorbol-13-acetate
(TPA),58,59 whereas the atypical PKCs did not respond to
DAG and phorbol esters.59−61 Hence, starved U937 cells were
treated with the compounds as in the first experiment, except
that the stimulation was carried out with TPA. Under these
conditions, PKCζ was apparently not activated and the basal
level of Thr410 phosphorylation was not influenced by 4f
(Figure 7B). As expected, the conventional PKCβ isoform
responded to the TPA treatment by an increase of the
activation loop phosphorylation. Importantly, the latter was not
affected by the PKCζ inhibitor 4f, ruling out that the
compound nonselectively interfered with the activation of
PKCβ in cells, thus rendering the potential explanation ii
unlikely (see above). In contrast, the ATP-competitive inhibitor
Ro31-8220 blocked the activation of PKCβ at 10 μM, maybe
due to a nonspecific effect on the upstream kinases. That TNFα
can induce the phosphorylation and activation of PKCβ or
PKCα, as shown in Figure 7A, is rather unknown, although it
was reported in a few cases.62,63 It would be interesting to
investigate whether this activation is directly meditated by
PKCζ as the upstream kinase. Altogether, our results suggested
that 4f did not nonselectively inhibit the activation of the
nontarget kinase PKCβ, on the one hand, but that, on the other
hand, the complex inter-relationship of the regulation of PKC
activation in the cellular environment requires further
investigation.
Compound 1s, with 3,4-difluoro-substitution at the 1-phenyl,
was the most active compound in cells, with an IC50 of 0.9 μM.
It could be noticed that introduction of multiple fluorine atoms
at the 1-phenyl was an effective means to enhance the cellular
potency (compounds 1s,j and 2i,t).
Opposed to the cell-free assay, it was clearly observed that
the plain aryl and heteroaryls at position 3 of the pyrazoline
were not a superior replacement for the t-Bu group of 1a, since
compounds 2a,b,g,h,j,k showed equal or worse inhibition of
the reporter gene activity when compared to their t-Bu
analogues (1a,g). This could be possibly explained by an
increased nonspecific binding of the multiaryl ring system to
other cellular proteins. However, increasing the polarity of the
3-phenyl, e.g., by ortho-hydroxylation, resulted in higher
cellular potency when compared to the plain phenyl
(compounds 2c, 4h−j), with the best activity achieved by a
2,4-dihydroxysubstitution in 4k. This compound was among
the two most potent compounds in the cell free assay and
displayed the second highest cellular activity with an IC50 of 1.1
μM.
Unexpectedly, the pyrazole analogue 11 was equally active to
its respective pyrazoline 4e, which might be attributed to off-
target interactions, as the cell-free assay data showed a huge
reduction of PKCζ inhibition with 11. In addition, methyl-
substitution at position 4 of the pyrazoline (compound 9) did
not add to the activity in cells. The higher lipophilicity of
compound 9 might have decreased the cellular availability due
to reduced solubility in the assay medium and/or by increased
nonspecific adsorption (the additional methyl group in
compound 9 raised the calculated log P between 0.35 and
0.57 units, depending on the calculation method used).
Probably for the same reason, we noted that for the more
potent compounds, the cellular assay conditions seemed to
level the potencies. However, the two most active compounds
in cells were found among the most potent submicromolar
inhibitors in the cell free assay (compounds 1s and 4k).
Importantly, we were able to improve the cellular potency of
the previously reported allosteric PKCζ inhibitors by more than
1 order of magnitude.
Inhibition of NO Formation in RAW 264.7 Macro-
phage Cells. To provide evidence that our compounds were
indeed blocking the PKCζ-dependent NF-κB activation in cells,
we next analyzed a bona fide NF-κB-dependent downstream
process. PKCζ was reported to be essential for the NF-κB
6525
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
pathway activation in RAW 264.7 cells.64 In the same cell line,
NF-κB activation was required to induce transcription of
inducible nitric oxide synthase (iNOS) after LPS stimulation.65
Hence, iNOS is one of the enzymes that is transcriptionally
coregulated by the PKCζ signaling pathway; accordingly, our
allosteric PKCζ inhibitors were expected to suppress the
induction of iNOS and to reduce the release of its specific
enzymatic product, nitric oxide (NO), in the cell culture
medium. Indeed, we found that several of our most potent
compounds (4f, 2h, and 2i, see Table 4) but not two analogues
substitution at the 4-postion of the pyrazoline had a profound
effect on potency, in addition to its potential role in stabilizing
the pyrazoline ring against oxidation.
EXPERIMENTAL SECTION
■
Chemistry. Solvents and reagents were obtained from commercial
suppliers and used as received. A Bruker DRX 500 spectrometer was
used to obtain 1H NMR and 13C NMR spectra, and in a very few cases
a Varian Mercury VX 300 spectrometer was used for recoding the
spectra of some enones. The chemical shifts are referenced to the
residual protonated solvent signals and occasionally TMS was used as
a reference. At least 95% purity in all the tested compounds (Table 1
and Table S1, Supporting Information) was could be verified by means
of HPLC coupled with mass spectrometry. Mass spectra (HPLC−ESI-
MS) were obtained using a TSQ quantum (Thermo Electron Corp.)
instrument prepared with a triple quadrupole mass detector (Thermo
Finnigan) and an ESI source. All samples were inserted using an
autosampler (Surveyor, Thermo Finnigan) by an injection volume of
10 μL. The MS detection was determined using a source CID of 10 V
and carried out at a spray voltage of 4.2 kV, a nitrogen sheath gas
pressure of 4.0 × 105 Pa, a capillary temperature of 400 °C, a capillary
voltage of 35 V, and an auxiliary gas pressure of 1.0 × 105 Pa. The
stationary phase used was an RP C18 NUCLEODUR 100-3 (125 × 3
mm) column (Macherey&Nagel). The solvent system consisted of
water containing 0.1% TFA (A) and 0.1% TFA in acetonitrile (B). The
HPLC method used a flow rate of 400 μL/min. The percentage of B
started at 5%, was increased up to 100% during 16 min, was kept at
100% for 2 min, and was flushed back to 5% in 2 min. Melting points
were determined using a Mettler FP1 melting point apparatus and are
uncorrected.
Table 4. Inhibition of iNOS Induction in RAW 264.7
Macrophage Cells
a
a
compd
1l
% inhibn at 7.5 μM
IC50 (μM)
26
32
50
74
65
69
97
ND
ND
7.5
4.6
3.4
2.4
0.8
1u
1a
4f
2h
2i
curcumin
a
Values are mean of at least four experiments; standard deviation
≤10%. ND: not determined.
that were inactive or less active against purified human PKCζ
(1l and 1u) were effectively suppressing the LPS-triggered
production of NO at low micromolar concentrations (Table 4).
Furthermore, the hit compound 1a also triggered iNOS
suppression, however at a somewhat lower extent than 4f, 2h,
and 2i, consistent with the less potent inhibition of PKCζ.
Curcumin, a known inhibitor of iNOS induction,66 was
included as a positive control (Table 4). It should be
mentioned that an MTT assay was performed always in
parallel with the experiments, indicating that the tested
compounds were not toxic to the RAW macrophage cells at
7.5 μM (Table S4, Supporting Information).
General Synthetic Methods and Experimental Details for
Some Key Compounds. General Procedure for the Enone
Synthesis. To a an ice-cooled solution of the appropriate ketone
(10 mmol) in MeOH (50 mL) was added 10% KOH aq solution (30
mL), followed by gradual addition of the corresponding aryl aldehyde
(10 mmol). The mixture was left to attain room temperature and
stirred overnight. The solid product was filtered and washed three
times with MeOH/H2O mixture (5:3) and left to dry. In the case of
oily products the reaction mixture was extracted by CH2Cl2 (4 × 20
mL), and the combined organic layers were washed with water, filtered
over anhydrous MgSO4, evaporated under reduced pressure, and used
without further purification.
(E)-1-(4-(tert-Butoxy)phenyl)-4,4-dimethylpent-1-en-3-one (E1).
The compound was synthesized according to the procedure for the
enone synthesis using pinacolone and 4-(tert-butoxy)benzaldehyde to
give a yellow solid: yield 2 g (77%); mp 110−111 °C; 1H NMR (500
MHz, CDCl3) δ 7.63 (d, J = 15.6 Hz, 1H), 7.49−7.45 (m, 2H), 7.00−
6.95 (m, 3H), 1.36 (s, 9H), 1.20 (s, 9H); 13C NMR (125 MHz,
CDCl3) δ 204.22, 157.72, 142.55, 129.75, 129.21, 123.75, 119.28,
79.27, 43.16, 28.85, 26.38.
CONCLUSIONS
■
In summary, we have discovered 1,3,5-trisubstituted and
1,3,4,5-tetrasubstituted pyrazolines as a potent and selective
class of allosteric PKCζ inhibitors. As suggested by the strongly
reduced in vitro and cellular activities toward a PIF-pocket
mutant of PKCζ, the compounds were most likely directed to
the PIF-pocket on the kinase catalytic domain. Several cell-
based assays confirmed that PKCζ was also targeted in the
cellular environment. Furthermore, we could provide evidence
that the putative PIF-pocket-directed compounds affected the
phosphorylation of the activation loop in the catalytic domain.
Our novel PKCζ inhibitor scaffold proved to be rich in many
modifiable positions, all acting as hot spots for improving
binding characteristics and displaying sharp SAR (see Figure
S2, Supporting Information, for a graphical summary). The
phenolic group on the 5-phenyl was essential for the inhibitory
activity, with a catechol providing the best activity. The
presence of a lipophilic (halogen or alkyl) substituent on the 1-
phenyl proved to be essential for the generation of high
potency, with the meta-position being the most preferable for
monosubstitution and the 3,4-positions for disubstitution.
There was a substantial flexibility with respect to the
functionality tolerated at the 3-position of the pyrazoline,
with the 2,4-dihydroxyphenyl identified as one of the most
favorable motifs for activity. Furthermore, an additional methyl-
General Procedure for the Pyrazoline Synthesis. A mixture of the
enone derivative (2 mmol) and the corresponding phenylhydrazine
hydrochloride (3 mmol) in 15 mL of anhydrous DMF was heated to
85 °C for 5 h under argon atmosphere. The reaction solution was
cooled to room temperature and partitioned between 50 mL of diethyl
ether and 20 mL of water. The organic layer was separated and washed
with three 20-mL portions of water. The aqueous layers were
combined and extracted with three 20-mL portions of diethyl ether.
The organic layers were combined, dried over anhydrous Na2SO4, and
concentrated in vacuo. The residue was purified using column
chromatography or used in the next step without further purification.
4-(3-(tert-Butyl)-1-(4-chlorophenyl)-4,5-dihydro-1H-pyrazol-5-yl)-
phenol (1a). The title compound was prepared by reaction of (E)-1-
(4-(tert-butoxy)phenyl)-4,4-dimethylpent-1-en-3-one (E1) and 4-
chlorophenylhydrazine hydrochloride according to the general
procedure for the pyrazoline synthesis. The product was purified by
CC (CH2Cl2/hexane 4:1) to give a beige solid: yield 0.28 g (44%); mp
1
111−112 °C; H NMR (500 MHz, DMSO-d6) δ 9.35 (s, 1H), 7.24−
7.19 (m, 2H), 7.03−6.98 (m, 2H), 6.79−6.75 (m, 2H), 6.71−6.67 (m,
6526
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
2H), 5.06 (dd, J = 11.6, 6.3 Hz, 1H), 3.48 (dd, J = 17.5, 11.6 Hz, 1H),
2.66 (dd, J = 17.5, 6.4 Hz, 1H), 1.17 (s, 9H); 13C NMR (125 MHz,
DMSO-d6) δ 159.87, 156.53, 144.57, 132.56, 131.18, 126.87, 115.58,
114.47, 108.65, 62.71, 42.70, 33.43, 27.88; MS (ESI) m/z = 329.18 (M
+ H)+.
methoxyphenyl)-4,5-dihydro-1H-pyrazole (3h) in 10 mL of THF via
syringe at −78 °C. The mixture was stirred for 1 h. Afterward,
iodomethane (0.093 mL, 1.5 mmol) was added to the solution, and
the resulting mixture was stirred at −78 °C for 30 min and then left to
attain room temperature and stirred for 20 h. The reaction was
quenched with 10 mL of brine, the aqueous layer was separated and
extracted with two 10-mL portions of CH2Cl2, and the combined
organic layer was dried over MgSO4 and concentrated in vacuo. The
residue was purified by CC (CH2Cl2/hexane, 1:3) to afford 8 as an off-
4-(3-(tert-Butyl)-1-(3,4-difluorophenyl)-4,5-dihydro-1H-pyrazol-5-
yl)phenol (1s). The title compound was prepared by reaction of (E)-1-
(4-(tert-butoxy)phenyl)-4,4-dimethylpent-1-en-3-one (E1) and 3,4-
difluorophenylphenylhydrazine hydrochloride according to the general
procedure for pyrazoline synthesis. The product was purified by CC
(CH2Cl2/hexane 4:1) to give a light brown solid: yield 0.37 g (57%);
mp 95.8 °C; 1H NMR (500 MHz, DMSO-d6) δ 9.38 (s, 1H), 7.13 (dt,
J = 10.6, 9.2 Hz, 1H), 7.07−7.00 (m, 2H), 6.80−6.73 (m, 1H), 6.73−
6.68 (m, 2H), 6.55−6.50 (m, 1H), 5.04 (dd, J = 11.5, 6.7 Hz, 1H),
3.48 (dd, J = 17.5, 11.5 Hz, 1H), 2.67 (dd, J = 17.5, 6.7 Hz, 1H), 1.17
(s, 9H); 13C NMR (125 MHz, DMSO-d6) δ 160.32, 156.61, 149.47
1
white solid: yield 0.26 g (61%); mp 170−171 °C; H NMR (500
MHz, CDCl3) δ 7.95−7.91 (m, 1H), 7.45−7.39 (m, 1H), 7.25−7.22
(m, 1H), 7.17−7.14 (m, 1H), 7.14−7.07 (m, 3H), 7.02−6.97 (m, 2H),
6.88−6.84 (m, 1H), 6.83−6.78 (m, 1H), 4.68 (d, J = 7.2 Hz, 1H), 3.96
(s, 3H), 3.87 (s, 3H), 3.83−3.76 (m, 1H), 1.34 (d, J = 7.2 Hz, 3H);
1
13C NMR (125 MHz, CDCl3) δ 157.15, 152.98, 152.77 (d, JC−F
=
247.0 Hz), 147.09 (d, 2JC−F = 10.8 Hz), 146.06, 134.88 (d, 3JC−F = 5.3
1
2
3
4
Hz), 134.73, 131.74, 130.36, 130.05, 129.79, 126.24, 121.39 (d, 4JC−F
=
(dd, JC−F = 241.8, JC−F = 13.2 Hz), 143.01 (dd, JC−F = 8.7, JC−F
=1.2 Hz), 142.15 (dd, 1JC−F = 236.1, 2JC−F = 11.8 Hz), 132.40, 126.98,
117.28 (d, JC−F = 17.2 Hz), 115.63, 108.07 (dd, JC−F = 5.4, JC−F
2
3.5 Hz), 120.97, 118.80, 113.89, 113.61 (d, JC−F = 18.9 Hz), 113.58,
111.28 (d, 3JC−F = 9.8 Hz), 71.81, 56.26, 55.44, 53.58, 18.19; MS (ESI)
m/z = 424.81 (M + H)+.
2
3
4
=
2
2.7 Hz), 101.16 (d, JC−F = 21.8 Hz), 63.20, 42.86, 33.42, 27.84; MS
(ESI): m/z = 330.91 (M + H)+.
Biological Assays. Fluorescence Probing of Compound-Induced
Conformational Changes. Potential effects of the compounds on the
protein conformation were probed using the SYPRO Orange dye,
which strongly increases its fluorescence intensity upon binding to
solvent-exposed hydrophobic areas of proteins.48
Assays were performed in a final reaction volume of 25 μL,
containing 50 mM Tris-HCl, pH 7.4, 140 mM NaCl, 1 mM DTT, 0.8
times concentrated SYPRO Orange (Molecular Probes), 0.5 μM
PKCζ, and compounds in the indicated concentrations. Control
experiments were set up in the same manner, except that PKCζ was
omitted. Measurements were done in black 384 well plates (485 nm
excitation/585 nm emission) after heating of the samples for 4 min at
37 °C in a PolarStar plate reader (BMG Labtech, Germany). To
correct for fluorescence emission triggered by direct interaction
between the compounds and the dye, the control values obtained in
the absence of PKCζ protein were subtracted from the fluorescence
units generated in the presence of PKCζ.
General Procedure for the Ether Dealkylation. A 1 M BBr3
solution in CH2Cl2 (3−9 equiv) was added dropwise via a syringe
under nitrogen to a stirred solution of the methyl/ethyl ether
derivative (1 mmol, 1 equiv) in CH2Cl2 at −78 °C. Then the reaction
was maintained at −78 °C for 1 h. After that it was allowed to reach
room temperature and stirred for an additional 20 h. The mixture was
cooled to 0 °C and H2O was carefully added (15−25 mL). The
product was then repeatedly extracted with EtOAc, and the organic
layers were dried over anhydrous Na2SO4. Upon solvent removal the
residue was purified by column chromatography.
5-(3-(tert-Butyl)-1-(3-chlorophenyl)-4,5-dihydro-1H-pyrazol-5-yl)-
2-fluorophenol (4f). The title compound was prepared by
demethylation of 3-(tert-butyl)-1-(3-chlorophenyl)-5-(4-fluoro-3-me-
thoxyphenyl)-4,5-dihydro-1H-pyrazole (3f) using BBr3 (3 equiv)
according to the general procedure for ether dealkylation. The
product was purified by CC (CH2Cl2/hexane 4:1) to give a buff solid:
yield 0.23 g (66%); mp 101−102.5 °C; 1H NMR (500 MHz, DMSO-
d6) δ 9.85 (s, 1H), 7.12−7.05 (m, 2H), 6.88 (t, J = 2.1 Hz, 1H), 6.78
(dd, J = 8.5, 2.1 Hz, 1H), 6.71−6.63 (m, 3H), 5.15 (dd, J = 11.7, 6.0
Hz, 1H), 3.51 (dd, J = 17.6, 11.7 Hz, 1H), 2.69 (dd, J = 17.6, 6.0 Hz,
1H), 1.17 (s, 9H); 13C NMR (125 MHz, DMSO-d6) δ 160.47, 150.18
Protein Kinases and Kinase Assays. PKCζ was prepared and
purified as described before.31
The cell-free assay was done basically as previously described except
that reaction samples were incubated for 4 min at 37 °C before the
phosphorylation reactions were started by the addition of γ32P-ATP/
Mg2+.31,32 Similar conditions were used for the PKCι and mutant
PKCζ assays.
1
2
(d, JC−F = 240.5 Hz), 146.33, 145.22 (d, JC−F = 12.4 Hz), 138.88,
133.38, 130.33, 117.09, 116.59 (d, 3JC−F = 6.6 Hz), 116.38 (d, 2JC−F
=
18.4 Hz), 114.52, 111.81, 110.82, 62.23, 42.53, 33.45, 27.84; MS (ESI)
To investigate the effect of 4f on the activity of cellular PKCζ,
HEK293 cells were seeded in six-well plates and, at 80% confluency,
transfected using GST-PKCζ[wt] and GST-PKCζ[V297L] mutant
expression plasmids essentially as described.31 The next day, the
medium was exchanged to DMEM containing 0.1% FCS, and the test
compounds were added in different concentrations as indicated. After
6 h in the incubator, cells were stimulated by the addition of IGF-I (50
ng/mL final concentration) for 30 min at 37 °C. The cells were then
lysed, and the recombinant PKCζ was isolated and purified using
glutathione sepharose beads as previously described.31 Equal amounts
of eluted GST-PKCζ (50 ng per reaction) were assayed for enzymatic
activity as described using myelin basic protein as a substrate.32
Reporter Gene Assay. The human histocytic lymphoma cell line
U937 was transfected with a NF-κB reporter gene plasmid and the
assay performed exactly as previously described.31
Nitrite Assay (Griess Assay). Murine RAW 264.7 cells were seeded
in 96-well plates (8 × 104 cells/200 μL DMEM supplemented with
10% FBS and 1% antibiotics), cultured for 2 days under 5% CO2 at
37°, and then incubated with or without LPS in the absence or
presence of the test compounds or DMSO control for 20 h. As a
measure of NO synthesis, nitrite as the stable oxidation product was
quantified photometrically after its reaction to an azo-dye in the
supernatant essentially as described by Kiemer et al.67 Briefly, 100 μL
of cell culture supernatant was removed and combined with each of 90
μL of 1% sulfanilamide in 5% phosphoric acid and 90 μL of 0.1% N-
(1-naphthyl)ethylenediamine dihydrochloride in water in a 96-well
m/z = 346.89 (M + H)+.
4-(1-(3-Chlorophenyl)-5-(3-fluoro-4-hydroxyphenyl)-4,5-dihydro-
1H-pyrazol-3-yl)benzene-1,3-diol (4k). The title compound was
prepared by demethylation of 1-(3-chlorophenyl)-3-(2,4-dimethox-
yphenyl)-5-(3-fluoro-4-methoxyphenyl)-4,5-dihydro-1H-pyrazole (3k)
using BBr3 (9 equiv) according to the general procedure for the ether
dealkylation. The product was purified by CC (CH2Cl2/MeOH 99:1)
1
to give a beige solid: yield 0.22 g (55%); mp 183−184 °C; H NMR
(500 MHz, DMSO-d6) δ 10.41 (s, 1H), 9.89 (d, J = 17.4 Hz, 2H),
7.28−7.23 (m, 1H), 7.21−7.15 (m, 1H), 7.07 (dd, J = 9.9, 3.1 Hz,
1H), 6.93−6.86 (m, 3H), 6.79 (ddd, J = 8.4, 2.2, 0.8 Hz, 1H), 6.75
(ddd, J = 7.9, 2.0, 0.8 Hz, 1H), 6.37 (dt, J = 8.3, 2.2 Hz, 2H), 5.32 (dd,
J = 11.8, 6.1 Hz, 1H), 3.94 (dd, J = 17.7, 11.8 Hz, 1H), 3.21 (dd, J =
17.7, 6.1 Hz, 1H); 13C NMR (125 MHz, DMSO-d6) δ 160.14, 158.05,
1
2
151.73, 150.99 (d, JC−F = 241.8 Hz), 145.12, 144.22 (d, JC−F = 12.1
3
Hz), 133.59, 132.99 (d, JC−F = 5.0 Hz), 130.62, 129.58, 122.02 (d,
4JC−F = 2.9 Hz), 118.28, 118.02, 113.80 (d, JC−F = 18.9 Hz), 111.99,
2
111.22, 108.27, 107.71, 102.55, 60.78, 44.07; MS (ESI): m/z = 398.62
(M + H)+.
1-(3-Chlorophenyl)-5-(3-fluoro-4-methoxyphenyl)-3-(2-methoxy-
phenyl)-4-methyl-4,5-dihydro-1H-pyrazole (8). To a three-necked
flask containing 20 mL of dry THF was added 0.6 mL of a lithium
diisopropylamide (LDA) solution (2 M in THF/n-heptane/ethyl-
benzene) under argon atmosphere, followed by a solution of 411 mg
(1 mmol) of 1-(3-chlorophenyl)-5-(3-fluoro-4-methoxyphenyl)-3-(2-
6527
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
plate. Samples were then measured at 550 nm using a PolarStar plate
reader (BMG Labtech, Germany). In addition, a background
measurement was done at 620 nm, and the values obtained were
subtracted. Nitrite concentrations in the supernatants were determined
by comparison with a sodium nitrite standard curve. Experiments were
performed at last four times in triplicates.
To correct for any potential effects on the cell vitality, an MTT
assay was performed in parallel each time using the same cell line (cf.
Table S4, Supporting Information).
ABBREVIATIONS USED
■
Bcl-XL, B-cell lymphoma-extra large; BSA, bovine serum
albumin; CC, column chromatography; GLP, G9A-like protein;
GLUT-4, glucose transporter type 4; iNOS, inducible nitric
oxide synthase; IGF-I, insulin-like growth factor I; IκB, inhibitor
of κB; Jak1, Janus kinase-1; NF-κB, nuclear factor-κB; NO,
nitric oxide; OVA, ovalbumin; p62, sequestosome-1; Par-4,
prostate androgen responsive-4; PB1, phox and Bem 1p; PDK1,
phosphoinositide-dependent kinase-1; PH, pleckstrin homol-
ogy; PI3, phosphatidylinositol-3; PIF, PRK2-interacting frag-
ment; Stat6, signal transducer and activator of transcription-6;
Th2, T helper 2; TPA, 12-O-tetradecanoylphorbol-13-acetate.
Cell Treatments and Immunoblotting. U937 cells were cultured in
RPMI medium containing 10% fetal bovine serum (FBS, Gibco)
supplemented with 1% (v/v) antibiotic/antimycotic solution (Gibco)
in 12-well plates. The cells were serum-starved during 24 h before
treatment with compound 4f for 1 h. Then, depending on the
signaling pathway to be stimulated, TNFα (50 ng/mL) was added for
30 min or TPA (100 nM) for 10 min. The cells were harvested by
centrifugation, washed once with cold PBS, and lysed by the addition
of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS−
PAGE) sample buffer and boiling. For Western blot analysis, high
molecular weight DNA was sheared by passing through a 23-gauge
needle and syringe several times, and the samples were separated by
SDS−PAGE on 10% gel, and then transferred to Immobilon FL-
PVDF membranes (Millipore). The membranes were blocked in
blocking buffer for near-infrared Western blotting for 1 h and
incubated with one of the following primary antibodies in blocking
buffer for 3 h: antiphospho-PKCζ/λ(Thr410/403) (cat. no. 9378, Cell
Signaling, diluted 1:750), anti-phospho-PKCβI/II (Thr500) (cat. No.
ab5817, Abcam, diluted 1:750), or anti-β-actin antibody (cat. no. sc-
47778, Santa Cruz Biotechnology, diluted 1:1000). After washing, the
membranes were incubated with IRDye 680- or IRDye 800-conjugated
secondary antibodies for 1.5 h, washed again, and scanned with an
Odyssey infrared imaging system as reported previously.68
REFERENCES
■
(1) Newton, A. C. Protein kinase C: Poised to signal. Am. J. Physiol.
Endocrinol. Metab. 2010, 298, E395−402.
(2) Moscat, J.; Diaz-Meco, M. T.; Wooten, M. W. Of the atypical
PKCs, Par-4 and p62: Recent understandings of the biology and
pathology of a PB1-dominated complex. Cell Death Differ. 2009, 16,
1426−1437.
(3) Diaz-Meco, M. T.; Moscat, J. The atypical PKCs in inflammation:
NF-kappaB and beyond. Immunol. Rev. 2012, 246, 154−167.
(4) Leitges, M.; Sanz, L.; Martin, P.; Duran, A.; Braun, U.; Garcia, J.
F.; Camacho, F.; Diaz-Meco, M. T.; Rennert, P. D.; Moscat, J.
Targeted disruption of the zetaPKC gene results in the impairment of
the NF-kappaB pathway. Mol. Cell 2001, 8, 771−780.
(5) Bandyopadhyay, G.; Standaert, M. L.; Sajan, M. P.; Kanoh, Y.;
Miura, A.; Braun, U.; Kruse, F.; Leitges, M.; Farese, R. V. Protein
kinase C-lambda knockout in embryonic stem cells and adipocytes
impairs insulin-stimulated glucose transport. Mol. Endocrinol. 2004, 18,
373−383.
(6) Martin, P.; Villares, R.; Rodriguez-Mascarenhas, S.; Zaballos, A.;
Leitges, M.; Kovac, J.; Sizing, I.; Rennert, P.; Marquez, G.; Martinez, A.
C.; Diaz-Meco, M. T.; Moscat, J. Control of T helper 2 cell function
and allergic airway inflammation by PKCzeta. Proc. Natl. Acad. Sci. U.
S. A. 2005, 102, 9866−9871.
(7) Karin, M.; Greten, F. R. NF-kappaB: Linking inflammation and
immunity to cancer development and progression. Nat. Rev. Immunol.
2005, 5, 749−759.
(8) Duran, A.; Diaz-Meco, M. T.; Moscat, J. Essential role of RelA
Ser311 phosphorylation by zetaPKC in NF-kappaB transcriptional
activation. EMBO J. 2003, 22, 3910−3918.
(9) Diaz-Meco, M. T.; Berra, E.; Municio, M. M.; Sanz, L.; Lozano, J.;
Dominguez, I.; Diaz-Golpe, V.; Lain de Lera, M. T.; Alcami, J.; Paya,
C. V.; et al. A dominant negative protein kinase C zeta subspecies
blocks NF-kappa B activation. Mol. Cell. Biol. 1993, 13, 4770−4775.
(10) Folgueira, L.; McElhinny, J. A.; Bren, G. D.; MacMorran, W. S.;
Diaz-Meco, M. T.; Moscat, J.; Paya, C. V. Protein kinase C-zeta
mediates NF-kappa B activation in human immunodeficiency virus-
infected monocytes. J. Virol. 1996, 70, 223−231.
(11) Sontag, E.; Sontag, J. M.; Garcia, A. Protein phosphatase 2A is a
critical regulator of protein kinase C zeta signaling targeted by SV40
small t to promote cell growth and NF-kappaB activation. EMBO J.
1997, 16, 5662−5671.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and analytical data of compounds
1b−r,t,u, 2a−l, 3a−q, 4a−e,g−j, 5a−n, 6a−n, 7a, 9−11, and
E2−37; Molecular modeling and docking; Table S1, inhibition
of recombinant PKCζ and the NF-κB pathway in cells by
compounds (6a−n); Table S2, hit compound 1a does not
inhibit other potential target kinases in the NF-κB pathway nor
PKCβII; Table S3, Wilcoxon’s signed rank test; Table S4, MTT
toxicity assay using RAW 264.7 cells; Table S5, selectivity
profile of 4f vs the PKC family and further AGC kinases; Figure
S1, molecular electrostatic potentials; Figure S2, work-flow
chart. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +49 681 302 70312. Fax: +49 681 302 70308. E-mail:
ma.engel@mx.uni-saarland.de. Website: http://www.
pharmmedchem.de.
(12) Anrather, J.; Csizmadia, V.; Soares, M. P.; Winkler, H.
Regulation of NF-kappaB RelA phosphorylation and transcriptional
activity by p21(ras) and protein kinase Czeta in primary endothelial
cells. J. Biol. Chem. 1999, 274, 13594−13603.
(13) Martin, A. G.; San-Antonio, B.; Fresno, M. Regulation of nuclear
factor kappa B transactivation. Implication of phosphatidylinositol 3-
kinase and protein kinase C zeta in c-Rel activation by tumor necrosis
factor alpha. J. Biol. Chem. 2001, 276, 15840−15849.
(14) LaVallie, E. R.; Chockalingam, P. S.; Collins-Racie, L. A.;
Freeman, B. A.; Keohan, C. C.; Leitges, M.; Dorner, A. J.; Morris, E.
A.; Majumdar, M. K.; Arai, M. Protein kinase Czeta is up-regulated in
osteoarthritic cartilage and is required for activation of NF-kappaB by
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to the DAAD for the support of M.A.-H. in the
framework of the PPP funding. We also thank the Deutsche
Forschungsgemeinschaft (DFG, EN 381/2-3) for financial
support. The excellent technical assistance of Nadja Weber,
Tamara Paul, and Menna A. Hammam is also deeply
acknowledged.
6528
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
tumor necrosis factor and interleukin-1 in articular chondrocytes. J.
Biol. Chem. 2006, 281, 24124−24137.
(29) Titchenell, P. M.; Hollis Showalter, H. D.; Pons, J.-F. o.; Barber,
A. J.; Jin, Y.; Antonetti, D. A. Synthesis and structure−activity
relationships of 2-amino-3-carboxy-4-phenylthiophenes as novel
atypical protein kinase C inhibitors. Bioorg. Med. Chem. Lett. 2013,
23, 3034−3038.
(15) Levy, D.; Kuo, A. J.; Chang, Y.; Schaefer, U.; Kitson, C.;
Cheung, P.; Espejo, A.; Zee, B. M.; Liu, C. L.; Tangsombatvisit, S.;
Tennen, R. I.; Kuo, A. Y.; Tanjing, S.; Cheung, R.; Chua, K. F.; Utz, P.
J.; Shi, X.; Prinjha, R. K.; Lee, K.; Garcia, B. A.; Bedford, M. T.;
Tarakhovsky, A.; Cheng, X.; Gozani, O. Lysine methylation of the NF-
kappaB subunit RelA by SETD6 couples activity of the histone
methyltransferase GLP at chromatin to tonic repression of NF-kappaB
signaling. Nat. Immunol. 2011, 12, 29−36.
(30) Lamba, V.; Ghosh, I. New directions in targeting protein
kinases: Focusing upon true allosteric and bivalent inhibitors. Curr.
Pharm. Des. 2012, 18, 2936−2945.
(31) Frohner, W.; Lopez-Garcia, L. A.; Neimanis, S.; Weber, N.;
̈
Navratil, J.; Maurer, F.; Stroba, A.; Zhang, H.; Biondi, R. M.; Engel, M.
4-Benzimidazolyl-3-phenylbutanoic acids as novel PIF-pocket-target-
ing allosteric inhibitors of protein kinase PKCzeta. J. Med. Chem. 2011,
54, 6714−6723.
(16) Duran, A.; Rodriguez, A.; Martin, P.; Serrano, M.; Flores, J. M.;
Leitges, M.; Diaz-Meco, M. T.; Moscat, J. Crosstalk between PKCzeta
and the IL4/Stat6 pathway during T-cell-mediated hepatitis. EMBO J.
2004, 23, 4595−4605.
(17) Lavoie, G.; Esteve, P. O.; Laulan, N. B.; Pradhan, S.; St-Pierre, Y.
PKC isoforms interact with and phosphorylate DNMT1. BMC Biol.
2011, 9, 31.
(18) Rangel-Salazar, R.; Wickstrom-Lindholm, M.; Aguilar-Salinas, C.
A.; Alvarado-Caudillo, Y.; Dossing, K. B.; Esteller, M.; Labourier, E.;
Lund, G.; Nielsen, F. C.; Rodriguez-Rios, D.; Solis-Martinez, M. O.;
Wrobel, K.; Wrobel, K.; Zaina, S. Human native lipoprotein-induced
de novo DNA methylation is associated with repression of
inflammatory genes in THP-1 macrophages. BMC Genomics 2011,
12, 582.
(32) Engel, M.; Hindie, V.; Lopez-Garcia, L. A.; Stroba, A.; Schaeffer,
F.; Adrian, I.; Imig, J.; Idrissova, L.; Nastainczyk, W.; Zeuzem, S.;
Alzari, P. M.; Hartmann, R. W.; Piiper, A.; Biondi, R. M. Allosteric
activation of the protein kinase PDK1 with low molecular weight
compounds. EMBO J. 2006, 25, 5469−5480.
(33) Stroba, A.; Schaeffer, F.; Hindie, V.; Lopez-Garcia, L. A.; Adrian,
I.; Frohner, W.; Hartmann, R. W.; Biondi, R. M.; Engel, M. 3,5-
̈
Diphenylpent-2-enoic acids as allosteric activators of the protein kinase
PDK1: Structure−activity relationships and thermodynamic character-
isation of binding as paradigms for PIF-binding pocket-targeting
compounds. J. Med. Chem. 2009, 52, 4683−4693.
(19) Murray, N. R.; Kalari, K. R.; Fields, A. P. Protein kinase Ciota
expression and oncogenic signaling mechanisms in cancer. J. Cell.
Physiol. 2011, 226, 879−887.
(34) Wilhelm, A.; Lopez-Garcia, L. A.; Busschots, K.; Froehner, W.;
Maurer, F.; Boettcher, S.; Zhang, H.; Schulze, J. r. O.; Biondi, R. M.;
Engel, M. 2-(3-Oxo-1,3-diphenylpropyl)malonic acids as potent
allosteric ligands of the PIF pocket of phosphoinositide-dependent
kinase-1: Development and prodrug concept. J. Med. Chem. 2012, 55,
9817−9830.
(35) Hindie, V.; Stroba, A.; Zhang, H.; Lopez-Garcia, L. A.; Idrissova,
L.; Zeuzem, S.; Hirschberg, D.; Schaeffer, F.; Jorgensen, T. J.; Engel,
M.; Alzari, P. M.; Biondi, R. M. Structure and allosteric effects of low-
molecular-weight activators on the protein kinase PDK1. Nat. Chem.
Biol. 2009, 5, 758−764.
(20) Sajan, M. P.; Nimal, S.; Mastorides, S.; Acevedo-Duncan, M.;
Kahn, C. R.; Fields, A. P.; Braun, U.; Leitges, M.; Farese, R. V.
Correction of metabolic abnormalities in a rodent model of obesity,
metabolic syndrome, and type 2 diabetes mellitus by inhibitors of
hepatic protein kinase C-iota. Metabolism 2012, 61, 459−469.
(21) Standaert, M. L.; Galloway, L.; Karnam, P.; Bandyopadhyay, G.;
Moscat, J.; Farese, R. V. Protein kinase C-zeta as a downstream
effector of phosphatidylinositol 3-kinase during insulin stimulation in
rat adipocytes. Potential role in glucose transport. J. Biol. Chem. 1997,
272, 30075−30082.
(36) Stockman, B. J.; Kothe, M.; Kohls, D.; Weibley, L.; Connolly, B.
J.; Sheils, A. L.; Cao, Q.; Cheng, A. C.; Yang, L.; Kamath, A. V.; Ding,
Y. H.; Charlton, M. E. Identification of allosteric PIF-pocket ligands
for PDK1 using NMR-based fragment screening and 1H−15N TROSY
experiments. Chem. Biol. Drug Des. 2009, 73, 179−188.
(22) Farese, R. V. Function and dysfunction of aPKC isoforms for
glucose transport in insulin-sensitive and insulin-resistant states. Am. J.
Physiol. Endocrinol. Metab. 2002, 283, E1−11.
(23) Farese, R. V.; Sajan, M. P. Metabolic functions of atypical
protein kinase C: “Good” and “bad” as defined by nutritional status.
Am. J. Physiol. Endocrinol. Metab. 2010, 298, 385−394.
(37) Lopez-Garcia, L. A.; Schulze, J. O.; Frohner, W.; Zhang, H.; Suß,
̈
̈
E.; Weber, N.; Navratil, J.; Amon, S.; Hindie, V.; Zeuzem, S.;
Jørgensen, T. J. D.; Alzari, P. M.; Neimanis, S.; Engel, M.; Biondi, R.
M. Allosteric regulation of protein kinase PKCzeta by the N-terminal
C1 domain and small compounds to the PIF-pocket. Chem. Biol. 2011,
18, 1463−1473.
(24) Mochly-Rosen, D.; Das, K.; Grimes, K. V. Protein kinase C, an
elusive therapeutic target? Nat. Rev. Drug. Discovery 2012, 11, 937−
957.
(25) Yuan, L.; Seo, J. S.; Kang, N. S.; Keinan, S.; Steele, S. E.;
Michelotti, G. A.; Wetsel, W. C.; Beratan, D. N.; Gong, Y. D.; Lee, T.
H.; Hong, J. Identification of 3-hydroxy-2-(3-hydroxyphenyl)-4H-1-
benzopyran-4-ones as isoform-selective PKC-zeta inhibitors and
potential therapeutics for psychostimulant abuse. Mol. Biosyst. 2009,
5, 927−930.
(26) Wu, J.; Zhang, B.; Wu, M.; Li, H.; Niu, R.; Ying, G.; Zhang, N.
Screening of a PKC zeta-specific kinase inhibitor PKCzI257.3 which
inhibits EGF-induced breast cancer cell chemotaxis. Invest. New Drugs
2009, 28, 268−275.
(27) Trujillo, J. I.; Kiefer, J. R.; Huang, W.; Thorarensen, A.; Xing, L.;
Caspers, N. L.; Day, J. E.; Mathis, K. J.; Kretzmer, K. K.; Reitz, B. A.;
Weinberg, R. A.; Stegeman, R. A.; Wrightstone, A.; Christine, L.;
Compton, R.; Li, X. 2-(6-Phenyl-1H-indazol-3-yl)-1H-benzo[d]-
imidazoles: Design and synthesis of a potent and isoform selective
PKC-zeta inhibitor. Bioorg. Med. Chem. Lett. 2009, 19, 908−911.
(28) Kjaer, S.; Linch, M.; Purkiss, A.; Kostelecky, B.; Knowles, P. P.;
Rosse, C.; Riou, P.; Soudy, C.; Kaye, S.; Patel, B.; Soriano, E.;
MurrayRust, J.; Barton, C.; Dillon, C.; Roffey, J.; Parker, P. J.;
McDonald, N. Q. Adenosine-binding motif mimicry and cellular
effects of a thieno[2,3-d]pyrimidine-based chemical inhibitor of
atypical protein kinase C isoenzymes. Biochem. J. 2013, 451, 329−342.
(38) Huang, Y. R.; Katzenellenbogen, J. A. Regioselective synthesis of
ligands for the estrogen receptor. Org. Lett. 2000, 43, 61−64.
(39) Madsen, P.; Ling, A.; Plewe, M.; Sams, C. K.; Knudsen, L. B.;
Sidelmann, U. G.; Ynddal, L.; Brand, C. L.; Andersen, B.; Murphy, D.;
Teng, M.; Truesdale, L.; Kiel, D.; May, J.; Kuki, A.; Shi, S.; Johnson,
M. D.; Teston, K. A.; Feng, J.; Lakis, J.; Anderes, K.; Gregor, V.; Lau, J.
Optimization of alkylidene hydrazide based human glucagon receptor
antagonists. Discovery of the highly potent and orally available 3-
cyano-4-hydroxybenzoic acid [1-(2,3,5,6-tetramethylbenzyl)-1H-indol-
4-ylmethylene]hydrazide. J. Med. Chem. 2002, 45, 5755−5775.
(40) Kirk, K. L. Fluorination in medicinal chemistry: Methods,
strategies, and recent developments. Org. Process Res. Dev. 2008, 12,
305−321.
(41) Schmidtke, P.; Luque, F. J.; Murray, J. B.; Barril, X. Shielded
hydrogen bonds as structural determinants of binding kinetics:
Application in drug design. J. Am. Chem. Soc. 2011, 133, 18903−
18910.
(42) Sorme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.; Rini, J. M.;
Nilsson, U. J. Structural and thermodynamic studies on cation-Pi
interactions in lectin-ligand complexes: High-affinity galectin-3
inhibitors through fine-tuning of an arginine−arene interaction. J.
Am. Chem. Soc. 2005, 127, 1737−1743.
6529
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530

Journal of Medicinal Chemistry
Article
(43) Ebner, T.; Burchell, B. Substrate specificities of two stably
expressed human liver UDP-glucuronosyltransferases of the UGT1
gene family. Drug. Metab. Dispos. 1993, 21, 50−55.
(44) Davis, M. I.; Hunt, J. P.; Herrgard, S.; Ciceri, P.; Wodicka, L. M.;
Pallares, G.; Hocker, M.; Treiber, D. K.; Zarrinkar, P. P.
Comprehensive analysis of kinase inhibitor selectivity. Nat. Biotechnol.
2011, 29, 1046−1051.
(45) Wagner, G.; Laufer, S. Small molecular anti-cytokine agents.
Med. Res. Rev. 2006, 26, 1−62.
(46) Zhu, J.; Huang, J.-W.; Tseng, P.-H.; Yang, Y.-T.; Fowble, J.;
Shiau, C.-W.; Shaw, Y.-J.; Kulp, S. K.; Chen, C.-S. From the
cyclooxygenase-2 inhibitor celecoxib to a novel class of 3-
phosphoinositide-dependent protein kinase-1 inhibitors. Cancer Res.
2004, 64, 4309−4318.
(47) Hansen, J. D.; Grina, J.; Newhouse, B.; Welch, M.; Topalov, G.;
Littman, N.; Callejo, M.; Gloor, S.; Martinson, M.; Laird, E.;
Brandhuber, B. J.; Vigers, G.; Morales, T.; Woessner, R.; Randolph,
N.; Lyssikatos, J.; Olivero, A. Potent and selective pyrazole-based
inhibitors of B-Raf kinase. Bioorg. Med. Chem. Lett. 2008, 18, 4692−
4695.
(48) Niesen, F. H.; Berglund, H.; Vedadi, M. The use of differential
scanning fluorimetry to detect ligand interactions that promote protein
stability. Nat. Protoc. 2007, 2, 2212−2221.
(49) Davis, P. D.; Elliott, L. H.; Harris, W.; Hill, C. H.; Hurst, S. A.;
Keech, E.; Kumar, M. K. H.; Lawton, G.; Nixon, J. S.; Wilkinson, S. E.
Inhibitors of protein kinase C. 2. Substituted bisindolylmaleimides
with improved potency and selectivity. J. Med. Chem. 1992, 35, 994−
1001.
(50) Nagashima, K.; Shumway, S. D.; Sathyanarayanan, S.; Chen, A.
H.; Dolinski, B.; Xu, Y.; Keilhack, H.; Nguyen, T.; Wiznerowicz, M.;
Li, L.; Lutterbach, B. A.; Chi, A.; Paweletz, C.; Allison, T.; Yan, Y.;
Munshi, S. K.; Klippel, A.; Kraus, M.; Bobkova, E. V.; Deshmukh, S.;
Xu, Z.; Mueller, U.; Szewczak, A. A.; Pan, B. S.; Richon, V.; Pollock,
R.; Blume-Jensen, P.; Northrup, A.; Andersen, J. N. Genetic and
pharmacological inhibition of PDK1 in cancer cells: Characterization
of a selective allosteric kinase inhibitor. J. Biol. Chem. 2011, 286,
6433−6448.
(58) Maines, M. D.; Miralem, T.; Lerner-Marmarosh, N.; Shen, J.;
Gibbs, P. E. M. Human biliverdin reductase, a previously unknown
activator of protein kinase C betaII. J. Biol. Chem. 2007, 282, 8110−
8122.
(59) Kim, S.; Gao, Z.-G.; Soltysiak, K.; Chang, T.-S.; Brodie, C.;
Jacobson, K. P2Y6 nucleotide receptor activates PKC to protect
1321N1 astrocytoma cells against tumor necrosis factor-induced
apoptosis. Cell. Mol. Neurobiol. 2003, 23, 401−418.
(60) Ono, Y.; Fujii, T.; Ogita, K.; Kikkawa, U.; Igarashi, K.;
Nishizuka, Y. Protein kinase C zeta subspecies from rat brain: Its
structure, expression, and properties. Proc. Natl. Acad. Sci. U. S. A.
1989, 86, 3099−3103.
(61) Ways, D. K.; Cook, P. P.; Webster, C.; Parker, P. J. Effect of
phorbol esters on protein kinase C-zeta. J. Biol. Chem. 1992, 267,
4799−4805.
(62) Rigor, R. R.; Hawkins, B. T.; Miller, D. S. Activation of PKC
isoform beta(I) at the blood−brain barrier rapidly decreases P-
glycoprotein activity and enhances drug delivery to the brain. J. Cereb.
Blood Flow. Metab. 2010, 30, 1373−1383.
(63) Peng, J.; He, F.; Zhang, C.; Deng, X.; Yin, F. Protein kinase C-α
signals P115RhoGEF phosphorylation and RhoA activation in TNF-α-
induced mouse brain microvascular endothelial cell barrier dysfunc-
tion. J. Neuroinflammation 2011, 8, 28.
(64) Huang, X.; Chen, L.-Y.; Doerner, A. M.; Pan, W. W.; Smith, L.;
Huang, S.; Papadimos, T. J.; Pan, Z. K. An atypical protein kinase C
(PKCzeta) plays a critical role in lipopolysaccharide-activated NF-
kappaB in human peripheral blood monocytes and macrophages. J.
Immunol. 2009, 182, 5810−5815.
(65) Xie, Q. W.; Kashiwabara, Y.; Nathan, C. Role of transcription
factor NF-kappa B/Rel in induction of nitric oxide synthase. J. Biol.
Chem. 1994, 269, 4705−4708.
(66) Brouet, I.; Ohshima, H. Curcumin, an anti-tumor promoter and
anti-inflammatory agent, inhibits induction of nitric oxide synthase in
activated macrophages. Biochem. Biophys. Res. Commun. 1995, 206,
533−540.
(67) Kiemer, A. K.; Baron, A.; Gerbes, A. L.; Bilzer, M.; Vollmar, A.
M. The atrial natriuretic peptide as a regular of Kupffer cell functions.
Shock 2002, 17, 365−371.
(68) Hoppstaedter, J.; Diesel, B.; Eifler, L. K.; Schmid, T.; Bruene, B.;
Kiemer, P. A. K. Glucocorticoid-induced leucine zipper is down-
regulated in human alveolar macrophages upon Toll-like receptor
activation. Eur. J. Immunol. 2012, 42, 1282−1293.
(51) Standaert, M. L.; Galloway, L.; Karnam, P.; Bandyopadhyay, G.;
Moscat, J.; Farese, R. V. Protein kinase C-zeta as a downstream
effector of phosphatidylinositol 3-kinase during insulin stimulation in
rat adipocytes. Potential role in glucose transport. J. Biol. Chem. 1997,
272, 30075−30082.
(52) Limatola, C.; Schaap, D.; Moolenaar, W. H.; van, B. W. J.
Phosphatidic acid activation of protein kinase C-zeta overexpressed in
COS cells: Comparison with other protein kinase C isotypes and other
acidic lipids. Biochem. J. 1994, 304, 1001−1008.
(53) Numazawa, S.; Ishikawa, M.; Yoshida, A.; Tanaka, S.; Yoshida,
T. Atypical protein kinase C mediates activation of NF-E2-related
factor 2 in response to oxidative stress. Am. J. Physiol. 2003, 285,
C334−C342.
(54) Muller, G.; Ayoub, M.; Storz, P.; Rennecke, J.; Fabbro, D.;
Pfizenmaier, K. PKC zeta is a molecular switch in signal transduction
of TNF-alpha, bifunctionally regulated by ceramide and arachidonic
acid. EMBO J. 1995, 14, 1961−1969.
(55) Haas, T. L.; Emmerich, C. H.; Gerlach, B.; Schmukle, A. C.;
Cordier, S. M.; Rieser, E.; Feltham, R.; Vince, J.; Warnken, U.;
Wenger, T.; Koschny, R.; Komander, D.; Silke, J.; Walczak, H.
Recruitment of the linear ubiquitin chain assembly complex stabilizes
the TNF-R1 signaling complex and is required for TNF-mediated gene
induction. Mol. Cell 2009, 36, 831−844.
(56) Nigro, P.; Abe, J.-i.; Woo, C.-H.; Satoh, K.; McClain, C.; O’Dell,
M. R.; Lee, H.; Lim, J.-H.; Li, J.-d.; Heo, K.-S.; Fujiwara, K.; Berk, B. C.
PKC zeta decreases eNOS protein stability via inhibitory phosphor-
ylation of ERK5. Blood 2010, 116, 1971−1979.
(57) Kim, G.-Y.; Nigro, P.; Fujiwara, K.; Abe, J.-i.; Berk, B. C. p62
binding to protein kinase C zeta regulates tumor necrosis factor α-
induced apoptotic pathway in endothelial cells. Arterioscler. Thromb.
Vasc. Biol. 2012, 32, 2974−2980.
6530
dx.doi.org/10.1021/jm500521n | J. Med. Chem. 2014, 57, 6513−6530