2-Aminothiazole Derivatives as Selective Allosteric Modulators of the Protein Kinase CK2. 2. Structure-Based Optimization and Investigation of Effects Specific to the Allosteric Mode of Action

Article abstract of DOI:10.1021/acs.jmedchem.8b01765

Protein CK2 has gained much interest as an anticancer drug target in the past decade. We had previously described the identification of a new allosteric site on the catalytic α-subunit, along with first small molecule ligands based on the 4-(4-phenylthiazol-2-ylamino)benzoic acid scaffold. In the present work, structure optimizations guided by a binding model led to the identification of the lead compound 2-hydroxy-4-((4-(naphthalen-2-yl)thiazol-2-yl)amino)benzoic acid (27), showing a submicromolar potency against purified CK2α (IC₅₀ = 0.6 μM). Furthermore, 27 induced apoptosis and cell death in 786-O renal cell carcinoma cells (EC₅₀ = 5 μM) and inhibited STAT3 activation even more potently than the ATP-competitive drug candidate CX-4945 (EC₅₀ of 1.6 μM vs 5.3 μM). Notably, the potencies of our allosteric ligands to inhibit CK2 varied depending on the individual substrate. Altogether, the novel allosteric pocket was proved a druggable site, offering an excellent perspective to develop efficient and selective allosteric CK2 inhibitors.

Full text of DOI:10.1021/acs.jmedchem.8b01765

Subscriber access provided by EKU Libraries
Article
2-Aminothiazole derivatives as selective allosteric modulators of
the protein kinase CK2. Part 2: Structure–based optimization and
investigation of effects specific to the allosteric mode of action
Benoît Bestgen, Irina Kufareva, Weiguang Seetoh, Chris Abell, Rolf W. Hartmann, Ruben
Abagyan, Marc Le Borgne, Odile Filhol, Claude Cochet, Thierry LOMBERGET, and Matthias Engel
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01765 • Publication Date (Web): 28 Jan 2019
Downloaded from http://pubs.acs.org on January 29, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 74
Journal of Medicinal Chemistry
1
1
2
3
4
5
6
7
8
9
TITLE PAGE
2
-Aminothiazole derivatives as selective allosteric
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
modulators of the protein kinase CK2. Part 2:
Structure–based optimization and investigation of
effects specific to the allosteric mode of action
a-e
f
g
g
h
Benoît Bestgen , Irina Kufareva , Weiguang Seetoh , Chris Abell , Rolf W. Hartmann ,
f
a
c-e
c-e
Ruben Abagyan , Marc Le Borgne , Odile Filhol , Claude Cochet , Thierry Lomberget
a, Matthias Engel b,*
aUniversité de Lyon, Université Lyon 1, Faculté de Pharmacie - ISPB, EA 4446 Bioactive
Molecules and Medicinal Chemistry, SFR Santé Lyon-Est CNRS UMS3453 - INSERM US7,
9373 Lyon Cedex 8, France
6
bSaarland University, Pharmaceutical and Medicinal Chemistry, Campus C2.3, 66123
Saarbrücken, Germany
cInstitut National de la Santé et de la Recherche Médicale, U1036, 38000 Grenoble, France
dCommissariat à l’Energie Atomique, Institute of Life Sciences Research and Technologies,
Biology of Cancer and Infection, 38000 Grenoble, France
eUniversity of Grenoble Alpes, Unité Mixte de Recherche-S1036, 38000 Grenoble, France
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 74
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
fUniversity of California, San Diego, Skaggs School of Pharmacy and Pharmaceutical
Sciences, La Jolla, CA 92093, USA
gDepartment of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW,
UK
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hHelmholtz Institute for Pharmaceutical Research Saarland (HIPS), Department of Drug
Design and Optimization, Campus C2.3, 66123 Saarbrücken, Germany
KEYWORDS
CK2; kinase inhibitors; allosteric inhibition; kinase selectivity: anti-cancer
ABSTRACT
Protein CK2 has gained much interest as an anti-cancer drug target in the last decade. We had
previously described the identification of a new allosteric site on the catalytic α-subunit, along
with first small molecule ligands based on the 4-(4-phenylthiazol-2-ylamino) benzoic acid
scaffold. In the present work, structure optimizations guided by a binding model led to the
identification of the lead compound 2-hydroxy-4-((4-(naphthalen-2-yl)thiazol-2-
yl)amino)benzoic acid (27), showing a submicromolar potency against purified CK2α (IC =
5
0
0.6 µM). Furthermore, 27 induced apoptosis and cell death in 786-O renal cell carcinoma
cells (EC50 = 5 µM) and inhibited STAT3 activation even more potently than the ATP-
competitive drug candidate CX-4945 (EC s: 1.6 µM vs. 5.3 µM). Notably, the potencies of
5
0
our allosteric ligands to inhibit CK2 varied depending on the individual substrate. Altogether,
the novel allosteric pocket was proved a druggable site, offering an excellent perspective to
develop efficient and selective allosteric CK2 inhibitors.
ACS Paragon Plus Environment

Page 3 of 74
Journal of Medicinal Chemistry
3
1
2
3
4
5
6
7
8
9
INTRODUCTION
The protein kinase CK2 has an exceptional position within the human kinome due to
the significant proportion of the cellular phosphoproteome that can be attributed to this
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
kinase. This ubiquitously expressed protein kinase is involved in the activation of several
pro-oncogenic pathways that are critical for cell proliferation, differentiation and survival.2
CK2 overexpression has been widely correlated with a poor prognosis for patients with
prostate cancer, breast carcinoma, glioblastoma, leukemias, gastric carcinoma or renal cell
3
carcinoma (see ref. for a recent review). Indeed, CK2 promotes prosurvival and antiapoptotic
4
signals by interfering with NF-κB, Wnt and Akt pathways. The crucial role of CK2 as a
cancer driver makes it an interesting target in cancer therapy, and CK2 inhibition proved to be
an effective method to induce tumor regression.5
Optimal activity toward most protein substrates is reached when CK2 is assembled as
a tetramer consisting of a dimer of regulatory subunits (CK2β) and two catalytic subunits,
6
CK2α and/or CK2α’. The catalytic subunit of CK2 is believed to be constitutively active,
while the regulatory subunit modulates the selectivity toward a subset of substrate proteins.7, 8
However, there is clear evidence that the catalytic CK2α subunit also exists as a population
independent from CK2β in cells.8, 9
Most of the known CK2 inhibitors bind to the ATP pocket and show an ATP-
competitive behavior in enzymatic assays.10 CX-4945 (5-(3-chlorophenylamino)-
benzo[c][2,6]naphthyridine-8-carboxylic acid, silmitasertib) is one of the most potent CK2
1
1
inhibitors (IC = 14 nM) and is currently in phase II clinical trials. X-ray crystallographic
5
0
analysis revealed that CX-4945 binds to the ATP-binding pocket of CK2α in the active
1
2
conformation. Looking for splicing regulation through phosphorylation of serine/arginine
rich (SR) proteins, Kim et al. discovered a CK2-independent implication of CX-4945 in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 74
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cellular regulation of SR (serine/arginine–rich) protein phosphorylation.13 Indeed, this
compound was found to be a strong inhibitor of Cdc2-like kinase-2 (Clk-2) (IC = 4 nM),
5
0
1
3
almost fourfold more potent than toward CK2α (IC = 14 nM). In addition, CX-4945 was
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
identified as a potent inhibitor of the dual-specificity tyrosine-phosphorylation regulated
1
4
kinases (Dyrk) 1A and 1B with IC s of 6.8 and 6.4 nM, respectively. These off-target
5
0
activities of silmitasertib, potentially causing adverse effects, underline the need for potent
and more selective CK2 inhibitors.
Therefore, continuous efforts are spent to identify alternative druggable sites on CK2.
Given the large conformational flexibility of CK2α in the glycine-rich loop, the hinge/helix
αD region, and the β4β5 loop at the CK2β interface,15-17 it is likely that several catalytically
inactive conformations exist in the dynamic equilibrium, some of which can be stabilized by
the binding of small molecule ligands to unique allosteric pockets. Indeed, one such small
pocket, termed αD pocket, was identified on the large C-terminal lobe of the catalytic
18
subunit. The most elaborated αD pocket ligand, CAM4712 (MW 453 g/mol), exhibited a
moderate potency in the cell-free assay (IC = 7 µM) and good efficacy in cells, while the
5
0
selectivity over other kinases required further optimization.19
Recently, we identified 2-aminothiazole derivatives as novel allosteric inhibitors of
20
CK2α; exemplarily shown are compounds 1-4 (Figure 1). Using complementary methods, it
was demonstrated that compound 3 binds in an allosteric pocket adjacent to the ATP binding
site, between the glycine-rich loop and the αC-helix. A preliminary hit optimization led to
compound 4, exhibiting an IC of 3.4 µM. However, it had yet to be shown that the new CK2
5
0
modulator class can be developed further into more potent drugs, which is often a limitation
with allosteric target sites that were not evolutionary designed for high-affinity interactions
with small molecules.21, 22 Therefore, we carried out a compound optimization guided by a
binding model, as will be described below. The cellular effects of the allosteric CK2
ACS Paragon Plus Environment

Page 5 of 74
Journal of Medicinal Chemistry
5
1
2
3
4
5
6
7
8
9
modulators were investigated and compared to that of the ATP-competitive reference
inhibitor CX-4945. Our results revealed a distinct modulation of several signal transduction
pathways by our new compounds, attributable to their novel mode of action.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Chemistry. The aminothiazole core was prepared according to Hantzsch-type
syntheses from 2-bromoacetophenones and thiourea intermediates (Scheme 1). Most of the 2-
bromoacetophenones 6a-n were commercially available or easily prepared 6a-b from
commercially available acetophenones 5a-b through a bromination step, as described in the
literature.23, 24 Thioureas 8a-g were obtained after reaction of isothiocyanate intermediates
with ammonia solution. The latter were synthesized after reaction of the primary amine of
aniline derivatives 7a-g with carbon disulfide, in presence of triethylamine, followed by
2
5
iodine oxidation. Hantzsch syntheses were carried out in ethanol and the resulting products
were directly obtained as solids, after their precipitation with water. Semi-preparative HPLC,
recrystallization, or washing procedures led to the final products with a LC-UV (254 nm)
purity higher than 95%. In one case, the ethyl ester derivative was obtained instead of the free
carboxylic acid. A saponification step with NaOH was then carried out to obtain the desired
product 10. In this way, aminothiazoles (9-16 and 18-27) were obtained in varying yield, with
the exception of compound 17 which was commercially available.
Compound 30 could not be obtained through Hantzsch-type synthesis, maybe due to
the spatial proximity between the primary amine and the carboxylic acid of 2-aminobenzoic
acid, which prevents the formation of the isothiocyanate intermediate. Consequently, a
nucleophilic substitution between 2-bromo-4-(3-nitrophenyl)thiazole (28) and 2-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 74
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aminobenzoic acid (29) in hydrochloric acid was performed to obtain the desired 2-((4-(3-
nitrophenyl)thiazol-2-yl)amino)benzoic acid (30) albeit in a low 2 % yield (Scheme 2).
Several modifications were considered on the carboxylic acid function (Scheme 3).
The methyl ester derivative 31 was easily obtained through a classical acidic esterification
step. Sulfonamide derivatives 32-33 were obtained according to peptide coupling procedures
whereas acetoxymethyl prodrugs 34-36 were accessed from carboxylic acid derivatives (2 or
18-19) and bromomethyl acetate through a previously described method.26
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Optimization of potency. The preliminary Structure-Activity Relationships (SAR)
obtained from a small series of analogs of the original hit compound 1 (Figure 1) had
suggested that the three aryl ring system as well as the carboxylic acid function are needed for
CK2α inhibition.20
In order to gain further insight into essential key features and to optimize the
inhibitory potency, we synthesized a targeted set of derivatives to systematically explore the
SAR around the new scaffold. First, we varied the position and number of the essential
carboxyl function (compounds 9 and 30, Table 1). Although this did not further improve the
activity, placing the carboxylic acid group in the ortho-position gave compound 30, exhibiting
a potency similar to that of the para-regioisomer 2. This suggested that different protein side
chains might be available for interaction with the carboxylate, which, however, cannot easily
reach the carboxylate when it is in the meta-position, since compound 1 (Figure 1) was four
times less potent than 2 or 30. Changing the benzoic acid to a phenyl acetic acid moiety
(
derivative 10) was detrimental to the activity, indicative of a specific interaction of the
carboxylate that is strictly dependent on the position. Altogether, the p-position appeared most
suitable for the placement of the essential acidic function.
Next, we modified the second phenyl ring in the thiazole 4-position while keeping the
p-benzoic acid moiety constant (Table 2). Various substituents were introduced to modulate
ACS Paragon Plus Environment

Page 7 of 74
Journal of Medicinal Chemistry
7
1
2
3
4
5
6
7
8
9
the electron density of the phenyl ring: 3-methoxy (compound 11), 4-methoxy (compound
2), 2,4-dimethoxy (compound 13), 2,4-dihydroxy (compound 14) and 5-bromo-2-hydroxy
1
(compound 15). The rather flat SAR observed with these and previous derivatives indicated
that the electron density of the phenyl ring was not relevant to the binding affinity. For
example, compounds 12 and 14 (Table 2), carrying electron–donating substituents, exhibited
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0
the same potency as the 3-nitrophenyl derivative 1 (cf. Figure 1). We also synthesized and
tested the pyridinyl derivative 17 in order to explore more polar heterocycles with a ring
dipole moment similar to the nitrophenyl compound. However, the measured IC value of 19
5
0
µM was unfavorable, suggesting that the polar H-bond acceptor nitrogen in the ring is not
complementary to the electrostatic potential of the binding pocket. In the previously
synthesized compound 4 (Figure 1), bromine as a large hydrophobic substituent had a positive
impact on the potency, suggesting that an extension of the aromatic ring (cf. compounds 19,
20, 21, 22) or exchange by a bulky substituent such as tert-butyl (compound 16) could be
tolerated by the binding pocket and further increase the potency. The bulky tert-butyl moiety
in 16 could not successfully replace the aromatic ring, as indicated by the strong decrease of
the CK2α inhibition. On the other hand, some of the bicyclic aromatic systems were good
substitutes for the potentially unstable or toxic 3-bromophenyl moiety, with the naphthalenyl
derivative 19 being the most potent (Table 2). More polar bicycles, such as chromene (20)
and 2-benzofuran (21), were less active than 19, underlining the preference of a hydrophobic
aromatic moiety by this part of the binding pocket. Interestingly, a plain thiophene ring (18,
IC = 5 µM) led to an almost twofold increase of potency compared with the original phenyl
5
0
congener (3, Figure 1). This promising result prompted us to introduce a 5-benzothiophene
substituent, however, no significant improvement of the inhibitory activity was noted (22,
IC = 4 µM).
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 74
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
So far, the naphthalenyl derivative had shown the highest potency among the aryl
aminothiazole compounds. In order to obtain insight in the potential binding mode with
CK2α, we started docking simulations using the available crystal structures of human CK2α
and the CK2α /β holoenzyme. A prior inspection of the different PDB entries revealed a very
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
high mobility of the P-loop, with the tip residue Tyr50 spanning a distance of more than 20 Å
2
7
6
(
Tyr-OH group) between its extreme locations in 4UB7 and 1JWH , with many intermediate
positions showing up in the other CK2 crystal structures. In X-ray structures displaying active
CK2α, Tyr50 occludes the lipophilic allosteric pocket (e. g., in the active symmetric CK2α /β
2
2
tetramer in 4MD728 and in the CK2α/AMPPNP complex in 3NSZ ). Contrastingly, the
12
allosteric pocket was accessible and gave reasonable poses in our preliminary docking runs
data not shown) in the inactive conformations of CK2α, 3FWQ29 and 4UB7 , and the
27
(
6
30
holoenzyme structures 1JWH and 4DGL , displaying distorted asymmetric CK2α2/β2
2
8
tetramers organized in inactive trimers. Because it was more logic to correlate the alanine
scanning mutational analysis (cf. Figure 4) with docking simulations both carried out with the
isolated CK2α subunit, we selected PDB entry 3FWQ for further docking studies. The
docking poses with the ten highest scores were evaluated for plausibility taking into account
2
0
all previous results : i) the binding pocket is located close to the ATP binding site without or
with only marginal overlap; ii) the binding pocket is delimited by residues Val73, Lys74 and
Lys77; iii) the observed pose fits to the STD-NMR–based binding epitope mapping as
previously determined with 4. The most consistent binding model of 19 in the 3FWQ
structure (Chain B) displayed a tight fitting of the naphthalene ring in the hydrophobic pocket,
with an additionally fixed orientation due to a double CH–π interaction of Gly177 with the
naphthalene cycle (Figure 5A).
The predicted tighter fitting of the naphthalene moiety imposed a higher constraint on
the orientation of the residual molecule than it was the case with compounds 1-3, potentially
ACS Paragon Plus Environment

Page 9 of 74
Journal of Medicinal Chemistry
9
1
2
3
4
5
6
7
8
9
shifting the relative position of the carboxylate. Given that an optimal ionic interaction was
crucial for the overall binding affinity, it seemed worthwhile to revisit the carboxylic acid
substitution on the 4-naphthalen-2-yl-thiazol-2-yl-arylamines. Several derivatives 23-27 were
synthesized (Table 3) and evaluated for their CK2α inhibitory properties. Compounds 23
without acid function and 24 with a 4-cyano substitution were lacking CK2α inhibition, thus
confirming the necessity of an acidic function also in the naphthalenyl series. However, in
contrast to the situation with the 3-nitrophenyl series (Table 1), where the p-carboxyl
derivative was substantially more potent than the meta analog (cf. compounds 1 and 2), the
carboxylic acid function was now even slightly preferred in the meta position (Table 3,
comparison between 25 and 19). Because our docking simulation had predicted a CH–π
interaction of the benzoic acid moiety with Val73, we aimed at increasing the π-electron
density using a methoxy subsituent. Compared with the parent compound 25, the 6-methoxy-
3-benzoic acid analog 26 exhibited a reduced potency on CK2α, which might be explained
either by a steric clash of the methoxy group with the binding pocket, unfavorable
conformational effects or the decreased acidity of the essential carboxyl function. To avoid
these potential problems, while yet increasing both electron density and acidity of the benzoic
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
acid moiety, we synthesized and tested the salicylic acid analog. Salicylic acid is known for
its more than 16 fold stronger dissociation in water to the anion compared with benzoic acid,31
which should lead to an enhancement of the ionic interactions. In agreement with these
considerations, the salicylic acid derivative 27 exhibited a 5-fold higher potency than the
corresponding benzoic acid analog 19, reaching a submicromolar IC value (IC = 0.6 µM,
5
0
50
Table 3). As expected, docking simulations predicted a binding mode for 27 similar to that of
19, with a slightly shifted orientation of the salicylic acid head (Figure 5B). Also considering
the higher ionization rate, the binding pose obtained for 27 was in line with the observed
increase in potency.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 74
1
0
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Carboxyl group replacement and prodrug synthesis. Since our goal was to develop
inhibitors suitable for cellular studies, we further considered bioisosteric replacements of the
carboxylic acid as well as a prodrug approach, in order to circumvent problems associated
with poor cell permeabilities of carboxylic acid derivatives.32, 33 A set of appropriate
derivatives of 2 and 19 were synthesized and tested for their cellular efficacy using the clear
cell renal cell carcinoma cell line 786-O (Table 4). This cell line was chosen because it
displays a strong decrease of cell viability following to inhibition of the intracellular CK2.5, 34
The acylsulfonamides 32 and 33, which were supposed to preserve the acidic properties (pKa
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
≈
4-5), did not reach the cell-free potency of the parent compound 19; however, the tosyl
derivative 33 was more efficient than 19 in decreasing the cell viability at 25 µM (87% vs.
3% reduction) and as efficient as the best compound of this series, 27. Altogether, this might
indicate a better cellular permeability of the methylbenzenesulfonamide 33 compared with the
4
3
5
carboxylic acid derivatives, as could be expected based on previous experience. On the other
hand, it could not be ruled out that off-target interactions in the cells in addition to CK2α
inhibition could be responsible for the effects on cell viability. Therefore, in light of the
markedly higher potency of 27 toward purified CK2α, we concluded that any effects exerted
by this compound in cells will more likely be mediated through CK2 inhibition.
26
The acetoxymethyl ester prodrugs 34-36 were expectedly less potent than the parent
compounds under cell-free conditions, however, none of them led to an improvement in
cellular efficacy compared with 19 (Table 4). Further experiments evaluating the stability of
these prodrugs have to be carried out before envisaging new derivatives.
Investigation of the mode of action. Before continuing cellular studies, it was
important to verify that our optimized lead compound 27 still exhibits non-ATP competitive
behavior in the CK2α inhibition assay, as demonstrated earlier for the starting compounds 2
2
0
and 4. The Lineweaver-Burk analysis for 27 is depicted in Figure 2A, confirming a non-
ACS Paragon Plus Environment

Page 11 of 74
Journal of Medicinal Chemistry
1
1
1
2
3
4
5
6
7
8
9
ATP competitive mechanism of CK2 inhibition, with a K value of 0.18 µM.
i
It was also of interest to investigate a potential preference of the inhibitor for either the
catalytic CK2α subunit or the holoenzyme CK2α β . In addition, two substrate peptides,
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
representative of differentially recognized CK2 substrate proteins, were also tested: a CK2β–
independent peptide substrate (peptide 29) and a CK2β–dependent peptide substrate (peptide
M, derived from eIF2β). As shown in Figure 2B, the resulting three experimental settings
gave similar results, indicating that i) 27 is able to inhibit CK2α and the holoenzyme complex
with the same potency, thereby ruling out any binding to the CK2α–β subunit interface
3
6
pocket, and ii) the inhibition mode is not sensitive against extended substrate interactions as
exerted by the eIF2β–derived peptide M.37
Binding characterization using Surface Plasmon Resonance (SPR) and
isothermal titration calorimetry (ITC) analysis. To determine the affinity constants of our
allosteric inhibitors, we performed SPR analysis after coupling CK2α to the biochip surface.
Intriguingly, the analysis revealed that the K values fully correlated with the IC s of the
d
50
respective compounds, and moreover, that both parameters even showed mostly similar
2
absolute values, as indicated by the slope of 1.1 in the IC vs. K plot (r = 0.91, Figure S1,
50
d
Supporting Information). Moreover, these findings corroborated that compound binding in the
enzymatic inhibition assay was not subject to competition neither with ATP nor substrate
peptide, which would have raised the IC over the K . In addition, the close correlation
5
0
d
between these parameters indicated that the binding of the allosteric ligands consistently leads
to full inhibition, in line with the stabilization of an inactive conformation of CK2α.
To complement the SPR analysis data, the thermodynamic characteristics of binding
were determined for 27 using ITC (Figure 3). Under these conditions, we obtained a K value
d
of 0.11 µM (±15 % S.D.) for 27, which was somewhat lower than that found by SPR analysis
(0.3 µM), possibly reflecting that the allosteric binding site was better accessible when the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 74
1
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
protein was in solution than when it was immobilized on the SPR sensor chip. The
thermodynamic parameters calculated for ΔH (-6.9 ± 0.17 kcal/mol) and TΔS (2.53 ± 0.29
kcal/mol) indicated a mainly enthalpy–driven binding, in agreement with the essential
contribution of the ionic interaction to the binding affinity. As described above, the
replacement of the carboxylate in 2 by the methyl ester function in 31 had abolished the
inhibitory potency (Table 1), consistent with a surface localization of the ester function,
where water molecules strongly compete with the potential H-bond interactions of the ligand.
A comparison of the thermodynamic binding signature of 27 with that of the
previously analyzed compound 4-(4-(pyridin-2-yl)thiazol-2-ylamino)benzoic acid (compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0
9 in ref. ) (ΔH = -13.3 kcal/mol; TΔS = - 6.6 kcal/mol; K (ITC) = 14.4 µM; Figure S2)
d
revealed that the increase in binding affinity achieved with 27 was solely attributable to the
enhancement of hydrophobic interaction. The exchange of the too polar pyridin-2-yl moiety
by naphthalenyl in 27 turned the negative entropic contribution to a positive one, however, it
can also be seen that this was accompanied by a decrease in binding enthalpy. This
phenomenon, known as enthalpy–entropy compensation, is not uncommon in lead
optimization programmes; based on previous experience, the thermodynamic profile of 27,
with its still predominant enthalpy term, is in an ideal range for a lead compound, predicting
that the binding affinity can be further enhanced by increasing ΔS, which is usually more
easily achieved than the subsequent optimization of enthalpy.38
Mapping of the allosteric binding site targeted by 27. To this end, compound 27
was screened against a panel of 13 GST-CK2α mutants, possessing point mutations mainly
located in the area of the binding site as described previously for compound 4 (Figure 4,
2
0
Table S1). Three CK2α mutants were significantly less sensitive toward inhibition by 27
than the other mutants or the wild type enzyme: Val73Ala, Lys77Ala and His160Ala (Figure
4
) and therefore selected for dose–response experiments. Compared with the wild type GST-
ACS Paragon Plus Environment

Page 13 of 74
Journal of Medicinal Chemistry
1
3
1
2
3
4
5
6
7
8
9
CK2α (IC50 = 0.6 µM), the IC s of compound 27 were significantly elevated with the
50
following three mutants: a 2.8-fold increase was noted for CK2α-Val73Ala (IC = 1.7 µM), a
5
0
4.3-fold increase for CK2α-Lys77Ala (IC = 2.6 µM), and a 1.7-fold increase for CK2α-
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
His160Ala (IC = 1 µM).
5
0
The loss of potency toward the Val73Ala and Lys77Ala mutants can be explained on
the basis of the predicted direct interaction of the respective side chains with 27, thereby
corroborating these docking results (Figure 5B). The loss of the Lys74 side chain in the silent
Lys74Ala mutant might be compensated by Lys77 and Arg191, which could rotate to the
former positions of Lys74 and Lys77, respectively. Surprisingly, the exchange of Lys71 by
Ala also did not influence the affinity of 27 to the binding pocket. It could be speculated that
this exchange leads to a conformational rearrangement of side chains close to the binding
pocket, e.g. involving Ser51 and Leu70, which could then replace the interactions formerly
predicted for Lys71 by OH–π and/or CH– contacts. It is noteworthy that a reduced activity
against the Lys77Ala and His160Ala mutants was observed with both 27 and the previously
described analogue 4-[4-(3-Bromo-phenyl)-thiazol-2-ylamino]-benzoic acid (compound 7 in
2
0
ref. ), corroborating that the compounds were targeting the same binding site. In addition,
the repeated identification of His160 in our alanine scan strengthens the previous hypothesis
that this residue, being too far away to contact the inhibitor itself, may play a role in the
formation and/or stabilization of the allosteric pocket in the inactive conformation. 20 In the
original coordinates of the PDB entry 3FWQ, His160 is relatively close to Tyr50 (~5.6 Å),
and a further convergence of the two residues was observed during the docking simulations
(Figure 5). Thus, at least in some conformations, His160 may have the function to fix the
position of Tyr50, which forms one of the pocket walls, thus stabilizing the hydrophobic part
of the allosteric binding site. Altogether, the CK2α mutant screening results supported the
predicted binding model shown in Figure 5.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 74
1
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Compound 27 is a selective inhibitor of CK2. Selectivity is an important feature for
kinase inhibitors to avoid off-target related side effects.39, 40 Allosteric kinase inhibitors of the
so-called type III target binding sites that are less conserved among the protein kinase families
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
than the ATP binding pocket, usually providing high selectivity for the kinase of interest.39
Therefore, a pronounced initial selectivity should be observed with an inhibitor targeting an
allosteric site, even before optimization of this parameter. For our study, compound 27 was
evaluated against a selected panel of 54 kinases (Supplementary Table S2), including closely
related kinases from the CMGC family as well as kinases from other branches of the kinome.
Only four other kinases besides CK2α were inhibited by more than 50% in the presence of 5
µM 27 (8.3 times IC against CK2α): ErbB2 (70%), Aurora-B (62%), Pim-1 (60%) and PI3K
5
0
(
53%). It was noteworthy that the main off-targets of CX-4945, Clk2 and Dyrk1A, were not
or only weakly inhibited by 27, respectively. The overall selectivity of 27, as expressed by a
4
1
Gini coefficient of 0.78, was higher than that of CX-4945 (Gini coefficient = 0.67) and of
other CK2 inhibitors described in the literature.34, 42, 43 The high selectivity of 27 is consistent
with binding to an allosteric site that is not strongly conserved in the human kinome.
The selectivity of the most potent compound from our first study, 4-[4-(3-Bromo-phenyl)-
thiazol-2-ylamino]-benzoic acid (compound 7 in ref. 20), was only evaluated in a smaller
screening panel (32 kinases); however, comparing the data from the 20 kinases common to
both screenings, it can be concluded that our optimization efforts led to an improved
selectivity: Under comparable screening conditions, the most strongly affected off-targets
from the first study, GSK3β, CK1γ1, ACVR1 (ALK) and Clk4, which were all inhibited from
2
0
6
0 to 70 %, displayed only 41, 32, 0 and 23 % inhibition, respectively, in the presence of 27.
Further kinases, including Clk2, Abl, CHK2, CDK1/cyclinB and Clk1, which had shown
2
0
some weak inhibition (from 20 to 30 %) with the previous analogue , were not affected by
ACS Paragon Plus Environment

Page 15 of 74
Journal of Medicinal Chemistry
1
5
1
2
3
4
5
6
7
8
9
2
7. Only Pim-1 and EGFR were inhibited slightly stronger by 27 (60 and 21 %, respectively)
2
0
than by the previous compound (43 and 0 % inhibition, respectively).
Inhibition of CK2 cellular activity and induction of apoptosis by 27. CK2
inhibitors are known to decrease the viability of cancer cells.5, 34 Consequently, to assess the
cellular potency of 27, the compound was screened against several cell lines of diverse tissue
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
origins (Table 5), and the half maximal growth-inhibitory concentration (GI ) was
50
determined. Among the tested cancer cell lines, 27 exhibited GI values ranging from 5 µM
5
0
in the clear cell renal cell carcinoma cell line 786-O to 20 µM in prostate cancer cells
DU145). Importantly, compound 27 was clearly less toxic against the non-cancerous
(benign) kidney cells RPTEC, displaying a 5-fold higher GI (27 µM) than against the kidney
(
5
0
cancer cell line 786-O. This result is consistent with the dependency of cancer cells on a high
4
level of CK2α expression for promoting cell growth and survival, as reported earlier. Of
note, the ATP-competitive inhibitor CX-4945, which is currently evaluated in clinical trials,
was similarly effective against 786-O cancer cells (GI = 6 µM, Supplementary Figure S3),
5
0
but showed an overall less pronounced tumor cell selectivity than 27 in our assays; for
instance, CX-4945 was acting weaker against the tumor cell lines MCF7, DU145 and U138
than against RPTEC cells (Supplementary Figure S4). Furthermore, while 27 showed no
difference in potency between the non-tumor fibrocystic mammary epithelial cell line
MCF10A and the mammary tumor cell line MCF7, CX-4945 inhibited the cell growth of
MCF10A cells even 5 times stronger than that of the MCF7 tumor cells (Figure S4). Overall,
the in vitro tumor cell selectivity of 27 was higher than that of CX-4945.
7
86-O cells are known to be resistant to classical chemotherapy and are frequently
used as a model for clear cell renal cell carcinoma, as they share many characteristics with the
clinical tumor cells.44, 45 Moreover, it was already demonstrated that the inhibition of CK2α
4
6
efficiently blocks the growth of these renal cancer cells. Therefore, the 786-O cell line was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 74
1
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
selected as a model for the in-depth characterization of the effects of 27 in cancer cells.
First, the potency of 27 to inhibit the CK2 activity in intact cells was evaluated. Two
protein substrates, Akt1 and α-catenin, known to be phosphorylated by CK2 on residues
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Ser12947 and Ser641 , respectively, were used as a reporter of CK2 cellular activity. A
48
concentration–dependent decrease in phosphorylation of Akt1 (Ser129) and α-catenin
(
Ser641) was observed after a 24 h treatment with compound 27 (Figure 6A). In addition,
numerous dead cells were observed microscopically during the cell viability experiments
data not shown). Since the CK2-catalyzed phosphorylation of Akt1 on Ser129 was associated
(
4
7
with a hyperactivation of the latter kinase, we analyzed whether Akt downstream signaling
was also affected by our CK2 inhibitor. Indeed, the phosphorylation of the Akt1
phosphorylation site (Thr145) on p21/Waf1 was reduced in the presence of 27 in a
concentration-dependent manner (EC = 4.0 µM, Figure 6A). Altogether, we can conclude
5
0
that compound 27 inhibited CK2 cellular activity, blocked Akt1 downstream signaling and
decreased cell viability.
In light of the inhibitory effects of 27 on the Akt1 activity, we surmised that the
observed cell death is due to the induction of apoptosis. Two apoptosis markers, the
appearence of cleaved PARP and the decrease in survivin protein levels, were analyzed by
Western Blotting. Both hallmarks of apoptosis induction were observed simultaneously after a
2
4 h treatment by compound 27 (Figure 6B). The concentration of 27 needed for apoptosis
induction (about 4.6-4.9 µM) was consistent with the GI (about 5 µM) and the cellular
5
0
inhibition of CK2 activity (about 3 µM as monitored with P-Akt1(Ser129)), suggesting that
CK2 inhibition was responsible for the induction of apoptosis and the subsequent cell death.
Notably, the EC of 27 for the inhibition of cellular CK2, at least according to the Phospho-
5
0
Akt1-Ser129 reporter signal (EC = 3.2 µM), was only about 5 times higher than the IC
50
5
0
toward purified CK2α (0.6 µM) (Figure 6A). In contrast, under the same experimental
ACS Paragon Plus Environment

Page 17 of 74
Journal of Medicinal Chemistry
1
7
1
2
3
4
5
6
7
8
9
conditions, we observed a 70-fold drop of potency with CX-4945 in living cells (IC cellular
5
0
CK2 ≈ 1 µM / IC cell-free CK2 = 14.7 nM) (Figure 7C), which is in the expected range for
5
0
4
9
ATP-competitive inhibitors of kinases with K (ATP) values in the low µM range (K (ATP)
m
m
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
for CK2 ≈ 10 µM ). Assuming that the cellular uptake of both compounds was comparable,
the distinct behavior of 27 added to the evidence that it was acting in a non-ATP competitive
manner in living cells. A similar behavior had been observed with CI-1040, a highly selective
and non-ATP competitive inhibitor of MEK1/2, that showed a ratio of about 5 to 6-fold
51, 52
between enzymatic potency (IC = 17 nM) and cellular efficacy (EC = 100 nM).
5
0
50
The inhibitory potency of the allosteric ligands is substrate-dependent. Given that
more than 300 proteins were identified as potential CK2 substrates, it is fair to ask how the
1
catalytic activity towards these numerous substrates is regulated. Although CK2β was found
to modulate substrate specificity of the catalytic subunit upon complex formation,7, 8 it is
5
3
unlikely that this is the only physiological mechanism of controlling the CK2 activity. On
the basis of our data and previous speculations about the potential role of Tyr50 in the
transition between CK2 conformational states29, 50 it might be hypothesized that adoption of
an inactive conformation is not only artificially induced by our ligands but also has a
physiological, regulatory function. Postulating that either active or inactive states of CK2
could be stabilized in vivo by interaction with proteins or other effectors, the binding affinities
of different CK2 conformations toward our allosteric inhibitors are expected to vary. In
support of this, we obtained first hints from the finding that the phosphorylation of α-catenin-
Ser641 in cells was less efficiently suppressed by 27 than the phosphorylation of Akt1-Ser129
(
Figure 6A), whereas both CK2 target sites were invariantly affected by the bona fide ATP-
competitive inhibitor CX-4945 (Figure 7C and D). To rule out that this result was due to a
particularity of 27, we repeated the same experiment with the previous, structurally somewhat
different compound 4,20 that did not contain the naphthalene ring and carried a 3-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 74
1
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
carboxyphenyl instead of the salicylic acid (Figure 1). With 4, the differential response of the
substrate phosphorylations toward the inhibitor became even more evident (Figure 7A and B):
While the Phospho-α-catenin(Ser641) signal showed a steep initial decrease (EC for the
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
partial inhibition = 6 µM), about half of the phosphorylations could not be inhibited further by
, or this required much higher concentrations (Figure 7B), basically confirming the tendency
4
already seen with 27 (cf. Figure 6A). In contrast, the CK2–catalyzed Ser129-phosphorylation
on Akt1 could be suppressed to a greater extent by 4, exhibiting an overall EC of 16 µM.
5
0
The latter correlated well with the EC = 25 µM for the growth inhibition of the 786-O cells
5
0
by this compound (data not shown).
Next, we aimed at corroborating the above observations by analyzing the
phosphorylation of further known CK2 protein substrates (nucleolin, Six1 and the CK2β
subunit), using defined cell-free conditions (Figure 8A). Compound 27 inhibited the Six1 and
the CK2β phosphorylation with about the same efficacy (IC50 ≈ 1 µM) as the canonical
peptide substrate used in the previous dose–response assay (IC = 0.6 µM), whereas a clear
5
0
drop of potency was observed in the case of nucleolin (IC50 ≈ 3.8 µM) (Figure 8A).
Comparable potency differences against Six1 and CK2β phosphorylations on the one hand
(IC s ≈ 7-8 µM) and nucleolin phosphorylation on the other hand (IC50 ≈ 42 µM) were also
5
0
noted for compound 4, just shifted to higher concentration ranges (Figure S5A). Under the
same conditions, CX-4945 inhibited the phosphorylation of nucleolin, Six1 and the CK2β
subunit with IC s between 2 and 6 nM, in the range of that found with the peptide substrate
5
0
(IC = 5 nM).
50
The above data were in agreement with the expectation that CX-4945 can enter and
block the ATP-binding site in the fully active state of CK2 as well as in less active
conformational states. On the other hand, our allosteric inhibitor might bind with reduced
affinity when the fully active conformation of the kinase is stabilized through interaction with
ACS Paragon Plus Environment

Page 19 of 74
Journal of Medicinal Chemistry
1
9
1
2
3
4
5
6
7
8
9
a particular substrate or other effectors, as it could be the case with nucleolin. Similarly, it can
be speculated that a certain part of the cellular α-catenin pool might be associated with CK2 in
a tight complex, thereby stabilizing the active conformation of the CK2α subunit and
rendering it less sensitive towards allosteric inhibition. It should be noted that an interaction
with substrates was proposed earlier by Lolli and Battistutta as a possible mechanism to
activate the autoinhibited CK2α2/β2 oligomers found in cells.54
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Another major survival pathway which is constitutively upregulated in cancer with the
participation of CK2 is the STAT3 pathway. Because of its importance as a target for
potential anticancer therapies, we wondered whether the CK2 recruited to this pathway would
be sensitive toward our allosteric inhibitors or show only a low response like in the case of
nucleolin. STAT3-Tyr705 phosphorylation is the critical step in triggering the transcription
factor activity of STAT3, while Ser727 phosphorylation is involved in the regulation of the
5
5
phospho-Tyr705 levels. Inhibition or overexpression of CK2 was shown to influence the
Ser727 phosphorylation levels, but, dependent on the cell type, either a positive56, 57 or
5
5
negative correlation with the Tyr705 phosphorylation status was observed. Because STAT3-
Tyr705 phosphorylation had been consistently diminished in all cell types after treatment by
CK2 inhibitors, we used an anti-phospho-STAT3-Tyr705 antibody to analyze the effects of
2
7 and CX-4945 in 786-O cells. Interestingly, 27 had a somewhat stronger impact on the
STAT3-Tyr705 phosphorylation levels than CX-4945 (Figure 6C), suggesting that the cellular
CK2 pool engaged in the STAT3 activation pathway is – at least in the 786-O renal cancer
cells – more amenable to allosteric inhibition than to ATP binding site-directed inhibition.
In summary, our results indicated that the sensitivity of cellular CK2 toward allosteric
inibition varied to a considerable extent, depending on the protein substrate and/or the
signalling complex environment.
Different subcellular localization of CK2 after treatment of cells by 27 vs. CX-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 74
2
0
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
945. Next we analyzed whether the differing sensitivity of the CK2–mediated nucleolin
phosphorylation toward allosteric vs. ATP-competitive inhibition also translates into a
different localization of the nucleolin/CK2 complex in cells after treatment with the two
inhibitor types. Nucleolin is an abundant nucleolar protein known to bind to CK2α with very
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
8
high affinity. Due to this complex formation, a significant portion of CK2α in growing cells
5
9
is usually concentrated in the nucleoli. CK2 phosphorylates the bipartite nuclear localization
signal near the N-terminus of nucleolin, thus promoting the translocation of the nucleolin–
CK2 complex to the nucleolus.60 In consequence, the phosphorylation status of the CK2
phospho-acceptor site in nucleolin is expected to determine the subcellular localization of the
high-affinity complex, which can be monitored by a fluorescent label. Using an EGFP-CK2α
construct, we analyzed whether compound 27 and CX-4945 differently influence the
subcellular localization of CK2α in HeLa cells. Indeed, we found that CX-4945 at 8 µM
induced a significant translocation of EGFP-CK2α from the nucleoli to the nuclear matrix and
the cytoplasm compartments (Figure 8B), whereas in the presence of compound 27, the
nucleolar localization pattern of EGFP-CK2α was mostly retained at a concentration of 15
µM. A comparable difference was also found with 4, at a higher concentration (40 µM, Figure
S5B). These results were consistent with the observed potency differences between compound
CX-4945 and 4 or 27, respectively, with respect to inhibition of the nucleolin phosphorylation
in the cell free assay. Thus, the alternative mode of action of our compounds, that rendered its
potency substrate–dependent, further translated into a different impact on the subcellular
localization pattern of a high affinity CK2α/substrate complex when compared with the ATP-
competitive inhibitor CX-4945.
ACS Paragon Plus Environment

Page 21 of 74
Journal of Medicinal Chemistry
2
1
1
2
3
4
5
6
7
8
9
CONCLUSIONS
Supported by a binding model (Figure 5A), the previous hit compounds20 were
systematically optimized to achieve a submicromolar potency with compound 27. Hence, we
were able to demonstrate that the allosteric site targeted by our CK2 modulators is druggable
and holds the potential for the development of high affinity ligands without having to
critically increase the molecular weight. This view is particularly supported by the high
cellular efficacy of 27: Although the IC of 0.6 µM was still about 40 times higher than that
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
of the CK2 inhibitor CX-4945, which is currently in clinical phase II trials, the cellular
potency of 27 was even superior in one of the analyzed CK2–dependent pathways, the
STAT3 activation (EC = 1.6 µM). The comparably high cellular activity of the allosteric
50
inhibitor 27 can probably be attributed to the absence of binding competition with the high
intracellular ATP levels (~2–5 mM), which severly limited the cellular efficacy of CX-4945
on the other hand. As expected for an inhibitor targeting a non-ATP binding site, the kinase
selectivity profile was also superior to that of CX-4945. Although the allosteric inhibitors
were modulators of CK2 (also see below) with a varying impact on specific pathways rather
than being total blockers of activity, the most important anticancer effects of CK2 inhibitors,
i. e. induction of apoptosis and cell growth inhibition, were fully retained. Hence, there is an
excellent perspective to develop highly efficient and selective allosteric CK2 inhibitors by
optimizing further the binding affinity of the current lead compound 27.
Most notably, we could demonstrate that the potencies of 27 and 4 to inhibit CK2
depend on the individual substrate; particular pools of cellular CK2 that are engaged by
different signaling pathways varied in their sensitivity towards allosteric but not ATP binding
site–directed inhibition. In view of this particular outcome, the 2-aminothiazolyl benzoic acid
derivatives will have a pharmacological profile distinct from any ATP-competitive CK2
inhibitor, which is worth being investigated further in in vivo models. It will also be
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 74
2
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
interesting to examine whether fewer side effects can be expected for the novel compounds
due to their less complete CK2 inhibition.
For most kinases, the equilibrium between the active and inactive state of the catalytic
domain is regulated by phosphorylation of the activation loop. Nevertheless, it is increasingly
recognized that even kinases that were a priori activated through phosphorylation exist in a
dynamic equilibrium between multiple conformations in solution.61-64 Therefore, it can be
assumed that CK2, which is not regulated at all by phosphorylation of a typical regulatory
element on the catalytic subunit, also exists in a dynamic equilibrium between different
conformations. Other kinases, for which the activation loop phosphorylation is not sufficient
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
5
66
or not the only way to fully activate the catalytic domain, include PDK1 and Aurora-A .
Notably, the active conformation in these cases is stabilized by transient interactions with
6
5
substrate proteins (PDK1 ) or with a complex binding partner, which is the spindle-
6
7
associated protein Tpx2 in the case of Aurora-A. Analogously, our data provided first hints
that the interactions of CK2 with its substrates might influence the conformational state of the
catalytic domain; e.g., nucleolin, one of the substrates analyzed in the present study, might a
priori stabilize the active conformation of CK2α, as it was found to counteract the allosteric
inhibition by our compounds (Figures 8 and S5). On the other hand, a substrate–induced
conformational stabilization is not necessarily sufficient to ensure steady phosphorylation of
the substrate by CK2: in the case of nucleolin, it was reported that an additional interaction
between the growth factor FGF-2 and CK2β in the CK2–nucleolin complex was required to
68
trigger the phosphorylation. Nevertheless, our novel ligands might be useful as tools to
probe the conformational state of different CK2 populations bound in complexes with
individual substrates, in order to elucidate the relevance of CK2 conformational stabilization
for the regulation of its activity.
ACS Paragon Plus Environment

Page 23 of 74
Journal of Medicinal Chemistry
2
3
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
Chemistry.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-((4-(3-nitrophenyl)thiazol-2-yl)amino)benzoic acid (1), 4-((4-(3-Nitrophenyl)thiazol-2-
yl)amino)benzoic acid (2) and 3-((4-(pyridin-2-yl)thiazol-2-yl)amino)benzoic acid (17) were
purchased from Sigma-Aldrich. All commercially available chemicals, 2-bromo-1-(3-
nitrophenyl)ethanone (6c), 2-bromo-1-(2,4-dimethoxyphenyl)ethanone (6d), 2-bromo-1-(2,4-
dihydroxyphenyl)ethanone (6e), 2-bromo-1-(4-bromo-2-hydroxyphenyl)ethanone (6f), 1-
bromo-3,3-dimethylbutan-2-one (6g), 2-bromo-1-(thiophen-3-yl)ethanone (6h), 2-bromo-1-
(naphthalen-2-yl)ethanone (6i), 3-(2-bromoacetyl)-2H-chromen-2-one (6j), 1-(benzofuran-2-
yl)-2-bromoethanone (6k), 1-(benzo[b]thiophen-5-yl)-2-bromoethanone (6l), other starting
materials and solvents were purchased from Sigma-Aldrich, Acros Organics, Fischer
scientific or Alfa Aesar and were used without further purification. Reactions were monitored
by Thin Layer Chromatography (TLC Silica gel 60 F254) purchased from Merck and
observed under UV light (254 nm). Purification by semi-preparative HPLC was carried out on
an Agilent 1200 series HPLC system (Agilent Technologies) using an Agilent C18 column
(30 x 100 mm/10 µm) as stationary phase and a gradient of water and acetonitrile as eluent.
1
13
H and C NMR spectra were obtained in dimethylsulphoxide-d , acetone-d or CDCl on a
6
6
3
1
13
Bruker DRX-500 instrument, operating at 500 MHz for the H and at 125 MHz for the C at
300K, using residual signal of deuterated NMR solvent as internal reference.69 Chemical
shifts are reported in parts per million (ppm), multiplicity of the signals are indicated by
lower-case letters (singlet s, doublet d, triplet t, quadruplet q, multiplet m, broad singlet br s,
or combination of letters). DEPT 135 was used to determine carbon multiplicity. Analytical
HPLC was performed using a SpectraSYSTEMTM (ThermoFisher) with a Macherey-Nagel
C18 column (3 ˣ 125 mm/5 µm). The purity of the final compounds was determined using
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 74
2
4
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HPLC–coupled ElectroSpray Ionisation (ESI) mass spectrometry (Finnigan Surveyor MSQ
Plus mass spectrometer, ThermoFisher). The UV traces at 254 nm thus obtained were used to
calculate the purity of the final compounds, which was at least 95 % in all cases. Melting
points (mp) of solids were determined by SMP3 melting point apparatus from Bibby Sterling.
High-resolution mass spectra (HRMS) were recorded on a Bruker MicroTOF Q mass
spectrometer.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General procedure 1 for the preparation of 2’-bromo-acetophenone analogs from
23
acetophenones. Bromine (1 eq) was added dropwise over 5 min to a stirred and warmed
(
40°C) solution of acetophenone derivative (1 eq) in CHCl . At the end of the addition, the
3
mixture was cooled down at RT; Et O was then added, followed by a saturated aqueous
2
solution of NaHCO and the resulting mixture was then stirred for 30 min. The organic layer
3
was separated, washed with a saturated aqueous solution of NaHCO , dried with Na SO and
3
2
4
filtered. The volatiles were removed under reduced pressure to afford the product as a solid,
which was used in the next step without further purification.
2
’-Bromo-3-methoxy-acetophenone (6a) .The reaction was carried out according to
general procedure 1, scale: 3-methoxyacetophenone (3.33 mmol, 459 µL), bromine (3.33
mmol, 172 µL), CHCl (10 mL), Et O (100 mL), to afford the title product as a yellow solid
3
2
(3.33 mmol, 762 mg). Yield = quantitative. LC-UV purity = 75%. The compound was not
+
1
ionized by ESI . H-NMR (500 MHz, CDCl ) δ = 3.70 (s, 3H, CH -O), 4.3 (s, 2H, CH -Br),
3
3
2
6
.99 (ddd, J = 1.0; 2.5; 8.2 Hz, 1H), 7.24 (t, J = 8.2 Hz, 1H), 7.34 (dd, J = 1.6, 2.5 Hz, 1H),
7.39 (ddd, J = 1.0, 1.6, 7.6 Hz, 1H); spectral data are in accordance with literature data.70
2’-Bromo-4-methoxy-acetophenone (6b).The reaction was carried out according to
general procedure 1, scale: 4-methoxyacetophenone (3.33 mmol, 500 mg), bromine (3.33
mmol, 172 µL), CHCl (10 mL), Et O (100 mL), to afford the title product as a grey solid
3
2
ACS Paragon Plus Environment

Page 25 of 74
Journal of Medicinal Chemistry
2
5
1
2
3
4
5
6
7
8
9
+
(
3.33 mmol, 765 mg). Yield = quantitative. LC-UV purity = 78%. LC-MS (ESI ) m/z =
+
1
2
29.31 for [M+H] . H-NMR (500 MHz, CDCl ) δ = 3.89 (s, 3H, CH -O), 4.40 (s, 2H, CH -
3 3 2
Br), 6.95 – 6.97 (m, 2H), 7.96 – 7.98 (m, 2H); spectral data are in accordance with literature
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
data.71, 72
General procedure 2 for the preparation of thiourea analogs from amines. Carbon
disulfide (3 eq) was slowly added to a mixture of aminobenzoic acid derivative (1 eq) in a 1/1
-1
THF/H O (0.7 mol.L ) solvent mixture and Et N (2.5 eq). The resulting mixture was stirred
2
3
at RT for 24h. Then iodine (1.05 eq) in THF (5 mL) was added dropwise at 0 °C and the
mixture was stirred for 3 h. A 1M HCl aqueous solution (7.5 mL) and Na SO (0.2 eq) were
2
3
added and the stirring was continued for 15 min. The aqueous layer was extracted with EtOAc
(3 x 50 mL). The combined organic layers were washed twice with brine (25 mL) and dried
on Na SO . After filtration, the volatiles were removed under reduced pressure to give the
2
4
isothiocyanate intermediate as a slightly yellow solid, which was converted to the thiourea
derivative by stirring at RT in NH OH (100 mL) for 6 h. NH OH refers to a 35% w/w
4
4
ammonia solution in water. After removal of the solids by filtration on a sintered-glass funnel,
the filtrate was concentrated to afford the thiourea product, which was used in the next step
without further purification.
Phenylthiourea (8a). Phenylisothiocyanate (8.37 mmol, 1 mL) was added to a
solution of NH OH (40 mL) and stirred at RT for 12 h. After removal of volatile in vacuo,
4
(aq)
the white residue was triturated in water and filtered to obtain the title product as a white solid
+
(8.15 mmol, 1.24 g). Yield = 97%. LC-UV purity > 99%. LC-MS (ESI ) m/z = 153.1 for
+
1
[
M+H] . H-NMR (500 MHz, DMSO-d ): δ = 7.12 (t, J = 7.3 Hz, 1H), 7.32 (t, J = 7.3 Hz,
6
2
H), 7.40 (d, J = 7.6 Hz, 2H), 9.70 (s, 1H); spectral data are in accordance with literature
data.73
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 74
2
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
-(4-Thioureidophenyl)acetic acid (8b). The reaction was carried out according to
the general procedure 2, scale: 2-(4-aminophenyl)acetic acid (3.31 mmol, 500 mg), to afford
the title product as an off-white solid (3.1 mmol, 651 mg). Yield = 93%. LC-UV purity =
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
+ 1
9
2
7%. LC-MS (ESI ) m/z = 211.05 for [M+H] . H-NMR (500 MHz, DMSO-d ): δ = 3.34 (s,
6
H), 5.00 (br s, 3H), 7.15 (d, J = 8.5 Hz, 2H), 7.30 (d, J = 8.2 Hz, 2H), 7.58 (br s, 1H), 10.2
1
3
(
br s, 1H). C-NMR (125.7 MHz, DMSO-d ): δ = 43.0, 122.7, 129.2, 133.4, 137.2, 173.9,
6
181.0. mp = 139-141 °C.
4
-Thioureidobenzoic acid (8c). The reaction was carried out according to the general
procedure 2, scale: 4-aminobenzoic acid (7.3 mmol, 1 g), to afford the title product as an off-
white solid (7.3 mmol, 1.43 g). Yield = quantitative. LC-UV purity = 80%. The compound
+
1
was not ionized by ESI . H-NMR (500 MHz, DMSO-d ): δ = 7.56 – 7.58 (m, 2H), 7.80 –
6
.82 (m, 2H); spectral data are in accordance with literature data.74
7
1
-(4-Cyanophenyl)thiourea (8d). A solution of 4-cyanophenyl isothiocyanate (6.24
mmol, 1 g) in NH OH (80 mL) was stirred at RT for 12 h. After cooling at 0 °C, the
4
(aq)
mixture was filtrated to afford the title product as an off-white solid (4.85 mmol, 860 mg).
+
1
Yield = 78%. LC-UV purity > 99%. The compound was not ionized by ESI . H-NMR (500
MHz, DMSO-d ): δ = 7.75 (m, 4H), 10.07 (s, 1H); spectral data are in accordance with
6
literature data.75
3
-Thioureidobenzoic acid (8e). The reaction was carried out according to the general
procedure 2, scale: 3-aminobenzoic acid (7.3 mmol, 1 g), to afford the title product as an off-
white solid (7.3 mmol, 1.43 g). Yield = quantitative. LC-UV purity = 96%. The compound
+
1
was not ionized by ESI . H-NMR (500 MHz, DMSO-d ): δ = 7.28 (t, J = 7.8 Hz, 1H), 7.57 –
6
7
.59 (m, 2H), 8.04 (d, J = 6.5 Hz, 1H), 8.92 (br s, 1H), 10.96 (br s, 1H). mp = 183-185 °C.
ACS Paragon Plus Environment

Page 27 of 74
Journal of Medicinal Chemistry
2
7
1
2
3
4
5
6
7
8
9
4
-Methoxy-3-thioureidobenzoic acid (8f). The reaction was carried out according to
the general procedure 2, scale: 3-amino-4-methoxybenzoic acid (7.3 mmol, 1.22 g) to afford
the title product as an off-white solid (7.3 mmol, 1.65 g). Yield = quant. LC-UV purity =
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+ 1
9
3
1
7%. LC-MS (ESI+) m/z = 226.79 for [M+H] . H-NMR (500 MHz, DMSO-d ): δ = 3.80 (s,
6
H), 5.33 (br s, labile H), 6.97 (d, J = 8.8 Hz, 1H), 7.52 (br s, 1H), 7.70 (dd, J = 1.9, 8.5 Hz,
1
3
H), 8.10 (s, 1H). C-NMR (125.7 MHz, DMSO-d ): δ = 55.7, 110.4, 126.5, 127.5, 127.7,
6
129.6, 154.2, 169.3, 181.6. mp = 241-243 °C.
2
-Hydroxy-4-thioureidobenzoic acid (8g). The reaction was carried out according to
the general procedure 2, scale: 4-aminosalicylic acid (7.3 mmol, 1.12 g) to afford the title
product as an off-white solid (7.3 mmol, 1.55 g). Yield = quantitative. LC-UV purity = 70%.
+
1
The compound was not ionized by ESI . H-NMR (500 MHz, DMSO-d ): δ = 6.73 (dd, J =
6
2
7
.2, 8.2 Hz, 1H), 6.96 (s, 1H), 7.30 – 7.50 (br s, labile H), 7.60 (d, J = 8.2 Hz, 1H), 7.56 –
.70 (br s, labile H), 10.00 (s, 1H). mp = 185-187 °C, melting point is in accordance with lit.
data (179-180 °C).76
General procedure 3 for the preparation of 2-aminothiazole analogs. An ethanolic
solution of 2’-bromo-acetophenone derivative and aryl thiourea derivative at eqimolar ratio,
corrected with respect to LC-UV purity, was heated at 80 °C for 3-12 h (reaction monitoring
by TLC). The mixture was then cooled down to RT and water (twice more than EtOH v/v)
was added. The mixture was stirred for 30 min and then filtrated. The obtained solid was
washed twice with water to afford the product as a solid. When necessary, purification was
performed (i) by recrystallization or washing with boiling MeCN (ii) by recrystallization or
washing in H O/MeOH (iii) by HPLC preparative to obtain the product with LC-UV purity
2
higher than 95%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 74
2
8
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
-(3-Nitrophenyl)-N-phenylthiazol-2-amine (9). The reaction was carried out
according to the general procedure 3, scale: 2-bromo-3’-nitro-acetophenone (0.66 mmol, 161
mg), phenylthiourea (0.66 mmol, 100 mg), absolute EtOH (5 mL). After removal of the
solvents under reduced pressure, water (10 mL) and MeOH (2 mL) were added. The pH of the
solution was adjusted to 9 with solid Na CO and then stirred for 1h. After filtration and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
washing of the residual solid with water (3 x 5 mL), the product was obtained as a yellow
+
solid (0.66 mmol, 196 mg). Yield = quantitative. LC-UV purity = 99%. LC-MS (ESI ) m/z =
+ 1
2
6
98.06 for [M+H] . H-NMR (500 MHz, DMSO-d ): δ = 6.99 (t, J = 6.9 Hz, 1H), 7.36 (t, J =
6
.6 Hz, 2H), 7.65 – 7.71 (m, 4H), 8.14 (d, J = 7.3 Hz, 1H), 8.35 (d, J = 6.9 Hz, 1H), 8.69 (s,
1
3
1H), 10.39 (s, 1H). C-NMR (125.7 MHz, DMSO-d ): δ = 105.6 (CH), 116.9 (2 x CH), 119.8
6
(
CH), 121.4 (CH), 121.9 (CH), 129.0 (2 x CH), 130.1 (CH), 131.7 (CH), 135.9 (C), 140.9 (C),
+
1
47.6 (C), 148.2 (C), 163.5 (C). mp = 136-138 °C. HRMS (ESI ): m/z calcd for C H N O S
1
5
12
3
2
+
[M+H] : 298.0645; found: 298.0633.
2-(4-(4-(3-Nitrophenyl)thiazol-2-ylamino)phenyl)acetic acid (10). The reaction was
carried out according to the general procedure 3, scale: 2-bromo-3’-nitro-acetophenone (0.48
mmol, 117 mg), 2-(4-thioureidophenyl)acetic acid (0.48 mmol, 100 mg), EtOH (5 mL) to
afford ethyl 2-(4-((4-(3-nitrophenyl)thiazol-2-yl)amino)phenyl) (0.41 mmol, 158 mg). Then, a
mixture of ethyl 2-(4-((4-(3-nitrophenyl)thiazol-2-yl)amino)phenyl)acetate in THF (5 mL)
and NaOH 0.5 M (1.3 mmol, 2.6 mL) was heated at 80 °C for 4 h. After removal of THF
under reduced pressure, 1M HCl(aq) was added until pH = 1. The product was obtained after
filtration as a black solid (0.22 mmol, 78 mg). Yield = 73%. LC-UV purity = 96%. LC-MS
+
+ 1
(
ESI ) m/z = 355.74 for [M+H] . H-NMR (500 MHz, DMSO-d ): δ = 3.53 (s, 2H), 7.24 (d, J
6
=
8.2 Hz, 2H), 7.64 (d, J = 8.2 Hz, 2H), 7.69 (s, 1H), 7.73 (t, J = 7.9 Hz, 1H), 8.16 (d, J = 7.9
1
3
Hz, 1H), 8.37 (d, J = 7.9 Hz, 1H), 8.70 (s, 1H), 10.43 (s, 1H). C-NMR (125.7 MHz, DMSO-
d ): δ = 39.7 (CH ), 105.5 (CH), 116.9 (CH), 119.8 (CH), 122.0 (CH), 128.0 (C), 129.0 (CH),
6
2
ACS Paragon Plus Environment

Page 29 of 74
Journal of Medicinal Chemistry
2
9
1
2
3
4
5
6
7
8
9
1
30.2 (CH), 131.7 (CH), 135.9 (C), 139.5 (C), 147.6 (C), 148.2 (C), 163.6 (C), 172.8 (C). mp
+
+
=
205-207 °C. HRMS (ESI ): m/z calcd for C H N O S [M+H] : 356.0700; found:
17 14 3 4
356.0691.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4-((4-(3-Methoxyphenyl)thiazol-2-yl)amino)benzoic acid (11). The reaction was
carried out according to the general procedure 3, scale: 2-bromo-3’-methoxy-acetophenone
(
0.66 mmol, 152 mg), 4-thioureidobenzoic acid (0.51 mmol, 100 mg), EtOH (10 mL).
Purification was performed by washing the residue with boiling MeCN (5 mL) and hot
filtration to give a crude solid which was then purified by flash chromatography
(
cyclohexane/Acetone 50/50) to afford the title product as a yellow solid (0.09 mmol, 29 mg).
+
+ 1
Yield = 17%. LC-UV purity = 96%. LC-MS (ESI ) m/z = 327.10 for [M+H] . H-NMR (500
MHz, DMSO-d ): δ = 3.83 (s, 3H), 6.91 (dd, J = 8.2, 2.2 Hz, 1H), 7.36 (t, J = 8.2 Hz, 1H),
6
7
.48 (s, 2H), 7.53 (d, J = 7.9 Hz, 1H), 7.81 (d, J = 8.8 Hz, 2H), 7.94 (d, J = 8.8 Hz, 2H), 10.69
1
3
(
s, 1H), 12.56 (br s, 1H). C-NMR (125.7 MHz, DMSO-d ): δ = 55.0 (CH ), 104.5 (CH),
6 3
111.1 (CH), 113.2 (CH), 115.8 (CH), 118.1 (CH), 122.8 (C), 129.7 (CH), 130.8 (CH), 135.6
C), 144.8 (C), 150.0 (C), 159.5 (C), 162.1 (C), 167.0 (C). mp = 246-248 °C. HRMS (ESI+):
(
+
m/z calcd for C H N O S [M+H] : 327.0798; found: 327.0786.
1
7
15
2
3
4
-((4-(4-Methoxyphenyl)thiazol-2-yl)amino)benzoic acid (12). The reaction was
carried out according to the general procedure 3, scale: 2-bromo-4’-methoxy-acetophenone
0.71 mmol, 163 mg), 4-thioureidobenzoic acid (0.51 mmol, 100 mg), absolute EtOH (10
(
mL). Purification was performed by washing with boiling MeCN (5 mL) and hot filtration to
afford the title product as a grey solid (0.15 mmol, 49 mg). Yield = 29%. LC-UV purity =
+
+ 1
9
3
7
6%. LC-MS (ESI ) m/z = 327.09 for [M+H] . H-NMR (500 MHz, DMSO-d ): δ = 3.80 (s,
6
H), 7.00 (d, J = 7.3 Hz, 2H), 7.29 (s, 1H), 7.81 (d, J = 7.9 Hz, 2H), 7.87 (d, J = 7.9 Hz, 2H),
1
3
.93 (d, J = 7.9 Hz, 2H), 10.65 (s, 1H), 12.54 (br s, 1H, O-H). C-NMR (125.7 MHz, DMSO-
d ): δ = 55.1 (CH ), 101.9 (CH), 114.0 (2 x CH), 115.9 (2 x CH), 122.7 (C), 127.1 (2 x CH),
6
3
ACS Paragon Plus Environment