Structure-based design, synthesis, and study of pyrazolo[1,5-a][1,3,5]triazine derivatives as potent inhibitors of protein kinase CK2

Article abstract of DOI:10.1016/j.bmcl.2007.05.041

The structure-based design, synthesis, and anticancer activity of novel inhibitors of protein kinase CK2 are described. Using pyrazolo[1,5-a][1,3,5]triazine as the core scaffold, a structure-guided series of modifications provided pM inhibitors with μM-le

Full text of DOI:10.1016/j.bmcl.2007.05.041

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4191–4195
Structure-based design, synthesis, and study
of pyrazolo[1,5-a][1,3,5]triazine derivatives as potent
inhibitors of protein kinase CK2
Zhe Nie,a,* Carin Perretta,a Philip Erickson,a Stephen Margosiak,b Robert Almassy,c
Jia Lu,b April Averill,b Kraig M. Yagera and Shaosong Chua,*
aDepartment of Medicinal Chemistry, Polaris Pharmaceuticals Inc., San Diego, CA 92121, USA
bDepartment of Protein Chemistry, Polaris Pharmaceuticals Inc., San Diego, CA 92121, USA
cDepartment of Structural Biology, Polaris Pharmaceuticals Inc., San Diego, CA 92121, USA
Received 27 February 2007; revised 11 May 2007; accepted 14 May 2007
Available online 18 May 2007
Abstract—The structure-based design, synthesis, and anticancer activity of novel inhibitors of protein kinase CK2 are described.
Using pyrazolo[1,5-a][1,3,5]triazine as the core scaffold, a structure-guided series of modifications provided pM inhibitors with
lM-level cytotoxic activity in cell-based assays with prostate and colon cancer cell lines.
Ó 2007 Elsevier Ltd. All rights reserved.
CK2 is a highly conserved and ubiquitous serine/threo-
nine protein kinase that only until recently has been con-
sidered as a possible target in cancer chemotherapy.1 It
has been shown to be elevated in a wide variety of can-
cers that have been examined2 and has been shown to be
correlated with the tumors’ aggressive growth. Targeted
over-expression of CK2 in transgenic mice results in
mammary tumorigenesis and CK2 has been demon-
strated to increase a cell’s oncogenic potential by sensi-
tizing a cell to transformation by other oncogenic
proteins.3 In addition, the recent interest stems in part
from new data concerning its involvement in cell prolif-
eration, protection from apoptosis, and its involvement
in angiogenesis, in part, through its regulatory function
on HIF-1.4 The modulation of CK2 activity has been
demonstrated to cause a decrease in cellular prolifera-
tion as well as an increase in the level of apoptosis ob-
served in cancer cells. This has been accomplished
through the use of small molecules, dominant negative
over-expression of kinase inactive mutants,5 anti-sense
methods,1 and small interfering RNA molecules (siR-
NA).6 These methods have been applied to tumor bear-
ing mice and in the best case, demonstrated complete
eradication of the PC3 human prostate cancer tumor.7
Clearly, modulation of CK2 protein levels or kinase
activity results in an interesting and promising pre-clin-
ical result.
The currently available CK2 inhibitors8 lack the po-
tency, physiochemical, and pharmacological properties
required to be successful in a clinical setting. The need
for new classes of CK2 inhibitors satisfying these chal-
lenges is clear. Using a structure-based design approach,
we have developed a series of pyrazolo[1,5-a][1,3,5]tri-
azine derivatives as potent inhibitors of CK2 that also
demonstrate cell-based anti-proliferative activity in mul-
tiple cancer cell lines.
The medicinal chemistry effort to find small molecule
inhibitors of CK2 started with lead identification utiliz-
ing both a protein structure specific library screen and
de novo rational design techniques. Initial screening of
our pre-selected library was accomplished using the hu-
man recombinant CK2a catalytic subunit. N2,N4-Diphe-
nyl-pyrazolo[1,5-a][1,3,5]triazine-2,4-diamine 1 (Fig. 1)
was found to be a potent inhibitor of CK2.
Keywords: CK2; Anticancer.
The active site of human and that of corn CK2 are vir-
tually identical with only minor conservative amino acid
substitutions. Corn CK2 (cCK2) protein readily co-crys-
*
Corresponding authors. Tel.: +1 858 731 3565; fax: +1 858 550
0526; e-mail addresses: zhe.nie@takedasd.com; shaosong.chu@
takedasd.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.041

4192
Z. Nie et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4191–4195
R
8
deeply buried, hydrophobic pocket lined by Val95 and
7
Phe113; (2) a large hydrophobic back pocket composed
of residues Val53 and Ile174; and (3) a pair of solvent
shielded polar residues Lys68 and Asp175 (Fig. 2.4).
Therefore, our initial efforts were focused on these re-
gions. Indeed, as expected, addition of a nitrile group
at the C8 position of 1 provided compound 2 with an
improvement of potency by ꢀ10-fold (Fig. 1). The
nitrile group interacts with hydrophobic residues Val95
and Phe113, which is confirmed by later solved co-crys-
tal structures. A focused library of C2 substitutions was
prepared next to explore the large back pocket.
6
N
1
N
N
4
5
1; R = H Ki = 0.26 μΜ (cCK2)
2
N
HN
N
3
2; R = CN Ki = 0.024 μΜ (cCK2) 0.015 μM (hCK2)
H
Figure 1. Pyrazolo[1,5-a][1,3,5]triazine inhibitors of CK2.
tallized within our laboratory and the resulting struc-
tural coordinates were therefore used in structure-based
˚
drug design of CK2 inhibitors. A 1.9 A resolution co-
According to Scheme 1, 5-amino-1H-pyrazole-4-carbo-
nitrile was first reacted with ethoxycarbonyl-isothiocya-
nate to give the corresponding pyrazolo-thiourea
derivative 3 at room temperature. Direct cyclization
failed to yield the desired pyrazolo-triazine. However,
once the thiol was methylated, it was cyclized smoothly
onto the pyrazole in the presence of DBU in DMF. C4
substitution with amines was carried out at 90 °C to give
6, followed by chlorination in refluxing phosphorus oxy-
chloride to provide 7. The chlorine could be displaced
easily with a variety of amines in NMP at 50 °C to pro-
vide 8.
crystal structure of 1 with cCK2 was solved in which
the pyrazolo-triazine core occupied the adenine binding
site and was sandwiched between Ile66 and Met163.
Two crucial hydrogen bonds between the inhibitor and
the backbone of Val116 in the proteins, hinge region an-
chored the inhibitor in the ATP binding site (Fig. 2.1).
The two phenyl groups of the inhibitor adopted a folded
conformation due to intra-molecular stacking interac-
tions, in which the C4-NHPh made hydrophobic inter-
actions with the side chains of Arg43, and the surface
amino acids Tyr115 and Asn118 (Fig. 2.2). The C2-
NHPh extended toward the substrate binding crevice
between the phosphate binding loop (Gly46 and
Arg47) and the catalytic loop (Fig. 2.3).
Using compound 7 as the building block, a focused li-
brary of 184 compounds with C2 substitutions was syn-
thesized in 96-deep well plates with 2 lmol/compound/
well. Without further purification, these library com-
pounds were screened against CK2 at 50 and 250 nM
concentrations. Compounds in the library that showed
>80% inhibition at 50 nM were re-synthesized and char-
acterized. The data are summarized in Table 1. In gen-
eral, these compounds show potent enzymatic
inhibition with the best Ki values in the low nM range.
It was recognized that three regions of the cCK2 binding
site were left largely untouched by the inhibitor: (1) a
CN
NH2
CN
NH2
NH
CN
NH2
a
b
N
N
N
N
O
N
O
N
H
S
N
S
O
O
4
3
CN
N
CN
N
N
N
e
d
c
N
S
N
OH
N
HN
N
OH
5
6
CN
CN
N
N
N
N
f
N
N
R1
Cl
N
HN
HN
N
N
R2
Figure 2. cCK2a active site complex with inhibitor 1 (PDB code:
2PVH). Four panels highlight active site residues that can potentially
interact with the pyrazole-triazine core or substitutions at the C2, C4
or C8 positions. All cCK2a residue numbers are derived from those of
the highly homologous human CK2a.
7
8
Scheme 1. Reagents and conditions: (a) (CO2Et)NCS, EtOAc/ben-
zene, rt; (b) MeI, NEt3, THF; (c) DBU, DMF; (d) Aniline, 90 °C; (e)
POCl3, reflux; (f) HNR1R2, i-Pr2NEt, NMP, 50 °C.

Z. Nie et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4191–4195
Table 1. Inhibition data of C2 substituted pyrazolo-triazines
4193
Co-crystal structures of compounds 8g, 8h and 8o in
complex with cCK2 were solved (Fig. 3). Interestingly,
these three inhibitors all adopted the folded conforma-
tion and occupied the same regions as observed in our
initial co-crystal structure (Fig. 2). Apparently, intra-
molecular hydrophobic stacking interaction between
C2 substitutions and C4 aniline dominates the confor-
mation of the C2 substitutions. Although these com-
pounds failed to reach the back pocket as we hoped,
the interaction of appropriately positioned C2-substitu-
tions (8g, 8h, and 8o) with adjacent protein residues such
as His160, Asp120, Phe121, Val45, and Met163 resulted
in significant enhancement of their inhibitory potency as
compared to the parent compound 2.
8
–NR1R2
hCK2a Ki (lM)
a
0.242
N
N
b
0.168
N
H
c
0.005
0.009
N
H
O
O
d
N
H
Clearly, in order to reach the back pocket, the C2-sub-
stitution needs to adopt an extended conformation,
which could be achieved by interrupting the intra-molec-
ular stacking interactions between C2 and C4 substitu-
tions, or designing an appropriately substituted
C2-group which would be induced-fit into the back
pocket. When C4-aniline was replaced with a flexible
group, a loss of potency was observed for compounds
9a, 9b, and 9c (Table 2). The co-crystal structure of 9a
in complex with cCK2 revealed that without a prefer-
ence for intra-molecular stacking interactions, C2 ani-
line indeed adopted the extended conformation, which
fit into the back pocket as expected (Fig. 3). Next,
compound 9d was synthesized with an acetamide on
the C2-phenyl ring designed to interact with the salt
bridge between Lys68 and Asp175. As predicted by
modeling, the increase in potency was profound with
9d being ꢀ20-fold more active than 9b. Compound 9e,
which has aromatic anilines on both C2 and C4
positions, was synthesized, and another ꢀ20-fold
enhancement in potency was achieved (Table 2).
N
e
0.023
N
H
N
f
0.045
0.008
N
H
N
H
N
H
g
N
H
h
i
0.005
0.055
Cl
N
j
0.013
0.012
N
H
k
N
H
N
N
l
0.014
0.126
m
N
N
n
o
0.011
0.024
N
N
N
N
p
0.007
N
a All Ki values reported in this paper were determined using human
CK2a (hCK2), except for 1 and 8a which were obtained from corn
CK2a.
Figure 3. cCK2a binary complex with 8h and inhibitors 8g, 8o, and 9a
superposed using their respective protein Ca atoms. (PDB codes:
2PVJ, 2PVK, 2PVL and 2PVM.)

4194
Z. Nie et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4191–4195
Table 2. Effect of C4 substitutions on the binding site of C2
substitutions
CN
N
N
N
R2
R1
N
9
9
a
R1
R2
hCK2 Ki (lM)
NH
0.36
N
N
H
N
H
b
c
0.137
3.4
N
H
N
H
N
N
H
N
H
O
NH
Figure 4. cCK2a with inhibitor 9e (PDB code: 2PVN) bound to the
active site in an extended conformation. Compound 9e accesses a
buried pocket with both hydrophobic and hydrophilic interactions.
d
0.008
N
H
N
H
O
azolo-triazine derivatives 11, followed by benzylation and
chlorination to give 12 and 13. Displacement of C4-
chloro on 13 with primary amines followed by mCPBA
oxidation yielded various C4-NHR derivatives 14. The fi-
nal products 15 were obtained by substitution of C2-sulf-
oxide with the amine at high temperature.
NH
e
0.00035
N
H
N
H
˚
The 2.0 A co-crystal structure for 9e in complex with
cCK2 revealed that, in addition to the normal interac-
tions between inhibitor and protein observed previously,
the C2 phenyl group indeed adopted an extended con-
formation and occupied the back pocket, making hydro-
phobic interactions with Val53 and Ile174. More
importantly, the acetamide group on the phenyl ring
adopted a cis conformation, forming three strong
hydrogen bonds with Asp175, Lys68, and a buried water
molecule (Fig. 4). Thus, all these highly favorable
hydrophobic and hydrogen bond interactions between
ligand and protein made 9e one of the most potent
CK2 inhibitors ever reported with Ki = 0.35 nM.
Table 3. Inhibitory activity of C4 and C8 substituted pyrazolo-
triazines
Compound
–X
–HNR
hCK2
Ki (lM)
HCT116
IC50 (lM)
9e
CN
CN
0.00035
0.00026
0.99
2.3
N
H
15a
N
H
15b
15c
CN
Me
0.0019
0.035
5.5
43
N
H
Further modifications at C4 and C8 positions of the pyr-
azolo-triazine core were carried out as illustrated in Table
3. Preparation of these pyrazolo-triazines starts with the
reaction between pyrazoles and ethoxycarbonyl-isothio-
cyanate as shown in Scheme 2. As discussed earlier in
Scheme 1 and by others,9 at room temperature, isothiocy-
anate reacted readily with the ring nitrogen of 3-amino-
pyrazole when X is an electron withdrawing group,
such as CN or Br. In contrast, when X is an electron
donating group, isothiocyanate reacted with both the ring
nitrogen and the 3-amino group of the pyrazole. How-
ever, by heating in EtOAc, 3-thiourea derivatives 10 were
the only products obtained due to its thermal stability.
Cyclization of 10 in the presence of base afforded the pyr-
N
H
15d
15e
15f
Me
Et
0.011
2
7.6
ND
18
N
H
N
H
Br
0.017
N
H

Z. Nie et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4191–4195
Table 4. Cell activity of selected pyrazolo-triazines
4195
X
X
X
O
a
N
b
NH
N
CN
N
H
N
N
N
NH
NH2
S
N
H
HN
O
N
N
SH
HO
N
N
O
X = CN, Me, Et, Br
HN
N
N
H
10
11
X
X
R
16
N
N
c
d
N
N
N
N
16
R
hCK2
HCT116
PC3
S
N
S
Cl
N
HO
Ki (nM) IC50 (lM) IC50 (lM)
a
b
c
d
e
4-N-Methylpiperazine 0.21
1.06
2.5
2.2
NI
39
1.4
12
13
O
3-NMe2
4-NEt2
3-COOH
0.55
0.5
ND
ND
ND
ND
X
X
NH
N
N
0.095
e,f
g
N
N
N
N
3-CONH(CH2)2NMe2 0.19
S
N
H
N
N
HN
HN
R
O
R
N
3
f
0.70
1.4
0.78
14
15
O
g
h
i
3-CONEt2
3-OEt
3-OCHF2
0.37
0.19
0.19
0.76
1.48
2.1
0.61
1.2
Scheme 2. Reagents and conditions: (a) (CO2Et)NCS, EtOAc, reflux;
(b) NH4OH, MeOH; (c) BnBr, i-Pr2NEt, NMP; (d) N,N-dimethylan-
iline, POCl3, reflux; (e) RNH2, NMP, rt; (f) mCPBA, CH2Cl2; (g) N-(3-
amino-phenyl)-acetamide, NMP, 160 °C.
1.7
NI, no inhibition; ND, not determined.
Structure activity relationships of compounds (15)
clearly indicated that a small cyclopropyl group at the
C4 position (15a) was at least as active as the phenyl
group (9a). However, small substitutions at the C8 posi-
tion, such as methyl (15d), ethyl (15e) or bromo (15f), re-
sulted in significant decreases in inhibitory potency,
probably because of their size and/or electronic proper-
ties (Table 3).
References and notes
1. Ahmad, K. A.; Wang, G.; Slaton, J.; Unger, G.; Ahmed, K.
Anti-Cancer Drugs 2005, 16, 1037.
2. Saha, S.; Bardelli, A.; Buckhaults, P.; Velculescu, V. E.; Rago,
C.; St Croix, B.; Romans, K. E.; Choti, M. A.; Lengauer, C.;
Kinzler, K. W.; Vogelstein, B. Science 2001, 294, 1343.
3. (a) Seldin, D. C.; Leder, P. Science 1995, 267, 894; (b)
Kelliher, M. A.; Seldin, D. C.; Leder, P. EMBO J. 1996, 15,
5160; (c) Landesman-Bollag, E.; Channavajhala, P. L.;
Cardiff, R. D.; Seldin, D. C. Oncogene 1998, 16, 2965; (d)
Channavajhala, P.; Seldin, D. C. Oncogene 2002, 21, 5280.
4. Mottet, D.; Ruys, S. P.; Demazy, C.; Raes, M.; Michiels, C.
Int. J. Cancer 2005, 117, 764.
5. Lebrin, F.; Chambaz, E. M.; Bianchini, L. Oncogene 2001,
20, 2010.
6. (a) Yamane, K.; Kinsella, T. J. Cancer Res. 2005, 65, 4362;
(b) Di Maira, G.; Salvi, M.; Arrigoni, G.; Marin, O.; Sarno,
S.; Brustolon, F.; Pinna, L. A.; Ruzzene, M. Cell Death
Differ. 2005, 12, 668.
As shown in Table 3, in an MTT cell-based assay, com-
pound 9e showed sub-lM inhibition against the
HCT116 cancer cell line. Therefore, a series of com-
pounds 16 with a variety of substitutions on the C4-aro-
matic ring were synthesized (Table 4). All of the
compounds are very potent inhibitors of human CK2
with Ki < 1 nM (many of them are more potent than
9e). Compounds 16a, 16f, and 16g strongly inhibit cell
growth when tested against HCT116 and PC3 cancer
cell lines. It should be noted that, there is still a big dis-
crepancy between their enzymatic potency and cell
growth inhibition, probably due to their low aqueous
solubility and poor cell membrane permeability. Further
optimization of this series of compounds is ongoing to
address these problems.
7. Slaton, J. W.; Unger, G. M.; Sloper, D. T.; Davis, A. T.;
Ahmed, K. Mol. Cancer Res. 2004, 2, 712.
8. (a) Sarno, S.; Ruzzene, M.; Frascella, P.; Pagano, M. A.;
Meggio, F.; Zambon, A.; Mazzorana, M.; Di Maira, G.;
Lucchini, V.; Pinna, L. A. Mol. Cell. Biochem. 2005, 276,
69; (b) Cozza, G.; Bonvini, P.; Zorzi, E.; Poletto, G.;
Pagano, M. A.; Sarno, S.; Donella-Deana, A.; Zagotto, G.;
Rosolen, A.; Pinna, L. A.; Meggio, F.; Moro, S. J. Med.
Chem. 2006, 49, 2363; (c) Battistutta, R.; Mazzorana, M.;
Sarno, S.; Kazimierczuk, Z.; Zanotti, G.; Pinna, L. A.
Chem. Biol. 2005, 12, 1211; (d) Sarno, S.; Salvi, M.;
Battistutta, R.; Zanotti, G.; Pinna, L. A. Biochim. Biophys.
Acta 2005, 1754, 263; (e) Zien, P.; Duncan, J. S.; Skierski,
J.; Bretner, M.; Litchfield, D. W.; Shugar, D. Biochim.
Biophys. Acta 2005, 1754, 271.
In summary, we have described the design and synthesis
of pyrazolo[1,5-a][1,3,5]triazines as novel CK2 inhibi-
tors. Using X-ray crystal structures of cCK2, protein
structure-based design enables us to identify the most
potent CK2 inhibitors with Ki < 1 nM. These com-
pounds also show strong inhibition against cancer cell
growth. Future work will be focused on the improve-
ment of their biophysical properties to yield drug-like
pre-clinical candidates for in-vivo animal studies.
9. Lubbers, T.; Angehrn, P.; Gmunder, H.; Herzig, S.;
Kulhanek, J. Bioorg. Med. Chem. Lett. 2000, 10, 821.