Pyrido[2, 3-d]pyrimidin-7(8H)-ones as new selective orally bioavailable Threonine Tyrosine Kinase (TTK) inhibitors

Article abstract of DOI:10.1016/j.ejmech.2020.113023

A series of pyrido [2, 3-d]pyrimidin-7(8H)-ones were designed and synthesized as new selective orally bioavailable Threonine Tyrosine Kinase (TTK) inhibitors. One of the representative compounds, 5o, exhibited strong binding affinity with a K_d value of 0.15 nM, but was significantly less potent against a panel of 402 wild-type kinases at 100 nM. The compound also potently inhibited the kinase activity of TTK with an IC₅₀ value of 23 nM, induced chromosome missegregation and aneuploidy, and suppressed proliferation of a panel of human cancer cell lines with low μM IC₅₀ values. Compound 5o demonstrated good oral pharmacokinetic properties with a bioavailability value of 45.3% when administered at a dose of 25 mg/kg in rats. Moreover, a combination therapy of 5o with paclitaxel displayed promising in vivo efficacy against the HCT-116 human colon cancer xenograft model in nude mice with a Tumor Growth Inhibition (TGI) value of 78%. Inhibitor 5o may provide a new research tool for further validating therapeutic potential of TTK inhibition.

Full text of DOI:10.1016/j.ejmech.2020.113023

Journal Pre-proof
Pyrido[2, 3-d]pyrimidin-7(8H)-ones as New Selective Orally Bioavailable Threonine
Tyrosine Kinase (TTK) Inhibitors
Minhao Huang, Yongjun Huang, Jing Guo, Lei Yu, Yu Chang, Xiaolu Wang, Jinfeng
Luo, Yanhui Huang, Zhengchao Tu, Xiaoyun Lu, Yong Xu, Zhimin Zhang, Zhang
Zhang, Ke Ding
PII:
S0223-5234(20)30995-8
DOI:
https://doi.org/10.1016/j.ejmech.2020.113023
EJMECH 113023
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 20 March 2020
Revised Date: 8 May 2020
Accepted Date: 10 November 2020
Please cite this article as: M. Huang, Y. Huang, J. Guo, L. Yu, Y. Chang, X. Wang, J. Luo, Y. Huang,
Z. Tu, X. Lu, Y. Xu, Z. Zhang, Z. Zhang, K. Ding, Pyrido[2, 3-d]pyrimidin-7(8H)-ones as New Selective
Orally Bioavailable Threonine Tyrosine Kinase (TTK) Inhibitors, European Journal of Medicinal
Chemistry, https://doi.org/10.1016/j.ejmech.2020.113023.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Masson SAS.

Graphical Abstract

Pyrido[2, 3-d]pyrimidin-7(8H)-ones as New
Selective Orally Bioavailable Threonine
Tyrosine Kinase (TTK) Inhibitors
†
,‡,ξ
,#
Ψ
,#
Ψ
,#
†
Ψ
Ψ
Minhao Huang,
Yongjun Huang, Jing Guo, Lei Yu, Yu Chang, Xiaolu Wang,
†
,ξ
†,ξ
†,
ξ
Ψ
†,ξ
Jinfeng Luo, Yanhui Huang, Zhengchao Tu, Xiaoyun Lu, Yong Xu, Zhimin
Ψ
,*
Ψ
,*
Ψ,*
Zhang, Zhang Zhang, Ke Ding
†
Guangdong Provincial Key Laboratory of Biocomputing, Joint School of Life Sciences,
Guangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences;
Guangzhou Medical University, Guangzhou 510530, China
‡
University of Chinese Academy of Sciences, No. 19 Yuquan Road, Beijing 100049,
China
ξ
Guangzhou Regenerative Medicine and Health Guangdong Laboratory (GRMH-GDL),
Guangzhou 510530, China
Ψ
International Cooperative Laboratory of Traditional Chinese Medicine Modernization
and Innovative Drug Discovery of Chinese Ministry of Education (MOE), Guangzhou
City Key Laboratory of Precision Chemical Drug Development, School of Pharmacy,
Jinan University, 601 Huangpu Avenue West, Guangzhou 510632, China
#
These authors contributed equally.
*
Corresponding authors, E-mail: zhangzm@jnu.edu.cn (Zhimin Zhang), 601 Huangpu
Avenue West, Guangzhou 510632, China; zhang_zhang@jnu.edu.cn (Zhang Zhang), 601
Huangpu Avenue West, Guangzhou 510632, China; dingke@jnu.edu.cn (Ke Ding), 601
Huangpu Avenue West, Guangzhou 510632, China.

KEYWORD: Pyrido[2,3-d]pyrimidin-7(8H)-ones, threonine tyrosine kinase (TTK),
orally bioavailable, inhibitor
ABSTRACT: A series of pyrido[2, 3-d]pyrimidin-7(8H)-ones were designed and
synthesized as new selective orally bioavailable Threonine Tyrosine Kinase (TTK)
inhibitors. One of the representative compounds, 5o, exhibited strong binding affinity
with a K value of 0.15 nM, but was significantly less potent against a panel of 402
d
wild-type kinases at 100 nM. The compound also potently inhibited the kinase activity of
TTK with an IC value of 23 nM, induced chromosome missegregation and aneuploidy,
5
0
and suppressed proliferation of a panel of human cancer cell lines with low µM IC50
values. Compound 5o demonstrated good oral pharmacokinetic properties with a
bioavailability value of 45.3% when administered at a dose of 25 mg/kg in rats. Moreover,
a combination therapy of 5o with paclitaxel displayed promising in vivo efficacy against
the HCT-116 human colon cancer xenograft model in nude mice with a Tumor Growth
Inhibition (TGI) value of 78%. Inhibitor 5o may provide a new research tool for further
validating therapeutic potential of TTK inhibition.
INTRODUCTION
Threonine tyrosine kinase (TTK, also known as Monopolar Spindle 1, Mps1) is a
dual-specificity protein kinase which was first discovered in budding yeast in the early
1
990s [1]. TTK has been identified as a core component of the spindle assembly
checkpoint (SAC) and plays a crucial role in accurate separation of sister chromatids
during mitosis to avoid aneuploidy [2-4]. TTK is essential for the activation and
maintenance of SAC functions. Upon recruitment by Aurora B to unattached kinetochore,

TTK phosphorylates kinetochore null protein 1 (Knl1) and consequently forms a
tetrameric mitotic checkpoint complex (MCC). MCC inhibits anaphase promoting
complex (APC/C) to restrain chromosome segregation and mitotic exit and prevents
premature initiation of anaphase [2, 5-8]. In addition to its role in mitosis, TTK is also
heavily involved in centrosome duplication [8-11], DNA damage checkpoint response
[12-16], ciliogenesis [17], meiosis [18, 19], cell transformation and chromosomal
instability (CIN) [20].
TTK was identified as one of the top 25 overexpressed genes in a variety of human
cancers [21]. For instance, elevated levels of TTK have been frequently detected in colon
cancer [22, 23], breast cancer including TNBC (triple-negative breast cancer) [24-27],
ovarian cancer [28], hepatocellular carcinoma [29-31], bladder cancer [32], pancreatic
cancer [33] and glioblastoma [34-36], and is closely associated with poor prognosis of
these diseases. TTK mediated aneuploidy is a hallmark of 70% – 95% of human cancer
cells in the top 10 most commonly affected organs [37]. Selective knockdown of TTK by
siRNA led to apoptosis of breast cancer cells with high aneuploidy levels while not
obviously affecting nonmalignant cells [27]. The cytotoxic differentiation between tumor
and normal cells implies that TTK inhibition could be considered as a new attractive
wide-spectrum strategy for cancer therapy.
A number of TTK inhibitors with different kinome selectivity profiles have been
reported from both pharmaceutical companies and academic institutes [38-44]. At least 5
molecules have been advanced to Phase I or II clinical investigation, including S-81694
(structure undisclosed), BAY-1161909 (1) [45, 46], BAY-1217389 (2) [45, 46],

CFI-402257 (3) [47] and BOS-172722 (4) [48] (Figure 1). Both inhibitors 1 and 2 were
developed from Bayer AG. The compounds exhibited single digit nM IC values against
5
0
TTK when assayed at high ATP concentrations, and potently suppressed the proliferation
of HeLa cervix carcinoma cells with IC50 values of 260 and 7.6 nM, respectively.
However, they only displayed moderate monotherapy efficacy with tumor growth
inhibition rates (TGI) less than 50% in various mouse tumor models possibly due to their
limited tolerability [46]. However, significant in vivo antitumor efficacy was achieved by
combining the inhibitors with paclitaxel, [45, 46] and both compounds have been
advanced to Phase I clinical trial in combination with intravenous paclitaxel
(NCT02138812, NCT02366949). Compound 3 is another highly potent and selective
TTK inhibitor with low nM IC50 potency inhibiting TTK kinase. This compound strongly
suppresses proliferation of MDA-MB-468, HCT-116 and OVCAR-3 cancer cells with
IC values of 2.0, 2.0 and 4.0 nM, respectively, and demonstrated a TGI value of 68%
5
0
following daily oral dosing of 6.5 mg/kg/day in a xenograft model of HCT-116 cancer
cells as a single-agent treatment [47]. Clinical efficacy of 3 as a monotherapy is being
actively investigated in patients with advanced solid tumors (NCT02792465). Another
promising TTK inhibitor 4 is currently in a Phase 1/1b clinical trial in combination with
paclitaxel. The compound exhibited an IC50 value of 11 nM and displayed extraordinary
kinome selectivity against a panel of more than 400 kinases [48]. Despite encouraging
progress on TTK inhibitor development, no clinical data has been disclosed to date. It is
still highly desirable to identify new selective TTK inhibitors with differentiated chemical
structures and good selectivity profiles to further validate TTK as a therapeutic target.

Figure 1. Chemical structures of the clinically investigated TTK inhibitors.
Herein, we report our efforts to discover pyrido[2, 3-d]pyrimidin-7(8H)-ones as new
orally bioavailable selective TTK inhibitors.
MOLECULAR DESIGN
L858R/T790M
Compound 5a is a covalent EGFR
inhibitor with an IC value of 91.8 nM
50
discovered in our laboratory [49]. Kinome selectivity investigations revealed that this
compound exhibited moderate TTK binding affinity with a K value of 190 nM. Since
d
compounds with a pyrido[2, 3-d]pyrimidin-7(8H)-one scaffold tend to exhibit a good
kinase selectivity [50], we utilized 5a as an initial lead compound for new selective TTK
inhibitor discovery.
A preliminary computational study suggested that compound 5a could bind to the ATP
binding site of TTK with a similar “U-shaped” configuration to that of EGFR in a type-I
binding mode (Figure 2). The aminopyrimidine moiety of compound 5a interacts with

Gly605 or Met793 at the hinge region of TTK or EGFR, respectively, through a classical
bidentate hydrogen bonding network. A hydrophobic sandwich of the pyrido[2,
3
-d]pyrimidin-7(8H)-one scaffold with a ceiling Ala551 residue and a floor consisted of
Leu654 and Ile663 is formed to stabilize binding of the molecule with TTK protein.
These interactions are similar to those formed with the Ala743 and Leu844 residues of
EGFR. The hydrophilic phenylpiperazine moiety is exposed to a solvent accessible
surface, which might only slightly influence the binding affinity. While the acrylamide of
5
a forms a covalent bond with Cys797 of EGFR, it instead remains close to Ala651 and
Asp608 of TTK. The acrylamide moiety may also form hydrophobic interactions with the
surrounding residues (Val678, Ile663, Pro673 and Leu654) in the flexible loop of the
TTK protein. Based on these observations, it was hypothesized that the acrylamide group
in 5a could be replaced by a saturated carboxyl amide to achieve potent and selective
binding with TTK over EGFR.
L858R/T790M
Figure 2. Predicted binding modes of compound 5a in EGFR
and TTK. (A)
L858R/T790M
Predicted binding mode of 5a with EGFR
kinase (PDB code: 5GMP); (B)
Predicted binding pose of 5a with TTK kinase (PDB code: 5EH0). The ligand is shown in

a gray stick representation.
CHEMISTRY
Synthesis of the designed pyrido [2, 3-d] pyrimidin-7(8H) -ones is outlined in Scheme
1
and Scheme 2. Briefly, commercially purchased ethyl 4-chloro-2-(methylthio)
pyrimidine-5-carboxylate (6) was reacted with different Boc-protected aliphatic amines to
generate intermediates 7, which were then reduced by lithium aluminum hydride (LiAlH4)
2
to yield alcohols 8. Compounds 8 were oxidized by manganese dioxide (MnO ) to
produce aldehydes 9. Compounds 9 were reacted with different Grignard reagents to
yield intermediates 10 which were then oxidized to give derivatives 11. The key
intermediates 12 were prepared by a tandem condensation-cyclization of 11 with
(2-methoxy-2-oxoethylidene) triphenylphosphorane. With intermediates 12 in hand, the
designed molecules 5e-5g, 5v and 5w were readily prepared by following previously
reported methods [51].
a
Scheme 1. Synthesis of compounds 5e-5g, 5v and 5w

a
Reagents and conditions: (a) K CO (2.0 eq.), MeCN, 0 °C to rt, 75-90%; (b) LiAlH
2
3
4
(
2.0 eq.), tetrahydrofuran (THF), -40 °C to 0 °C, 45-55%; (c) MnO
2
(10.0 eq.),
dichloromethane (DCM), rt, 80-95%; (d) Grignard reagent (2.0 eq.), THF, -78 °C, argon,
5-70%; (e) MnO (10.0 eq.), DCM, rt, 75-90%; (f) (2-Methoxy-2-oxoethylidene)
triphenylphosphorane (1.5 eq.), toluene, 110 °C, argon, 2 d, 18-35%; (g)
4
2
3
2
-Chloroperoxybenzoic acid (m-CPBA) (3.0 eq.), DCM, 0 °C, 3.0 hrs, 80-96%; (h)
-methoxy-4-(4-methylpiperazin-1-yl) aniline (1.1 eq.), trifluoroacetic acid (TFA) (1.2
eq. ), 2-Butanol (2-BuOH), 95 °C, argon, 18 hrs; (i) TFA (10.0 eq. ), DCM, rt, 2 hrs,
0-35% for two steps; (j) propionyl chloride (1.1 eq.), N,N-Diisopropylethylamine (DIEA)
2
(1.2 eq.), DCM, 0 °C, 30 minutes, 50-83%.
Alternatively, commercially available 5-bromo-2, 4-dichloropyrimidine (15) was

reacted with different Boc-protected aliphatic amines to generate intermediates 16 which
were reacted with various acrylic acids 17 under palladium-catalyzed conditions to
produce derivatives 18. Using 18 as the key intermediates, compounds 5b-5d, 5h-5u and
5
x-5z were prepared by following similar procedures to the previous protocols [49].
a
Scheme 2. Synthesis of compounds 5b-5d, 5h-5u and 5x-5z
a
Reagents and conditions: (a) K CO (2.0 eq.), MeCN, 0 °C to rt, 75-96%; (b) two steps,
2
3
1
) α,β-unsaturated carboxylic acid (5.0 eq.), DIEA (10.0 eq.), bis(benzonitrile) palladium
(
II) dichloride (0.05 eq.), tri-o-tolylphosphine (0.05 eq.), anhydrous THF, 70 °C, argon,
1
6 hrs; 2) Ac
-methoxy-4-(4-methylpiperazin-1-yl) aniline (1.1 eq.), TFA (1.2 eq.), 2-BuOH, 95 °C,
OH (1.3
eq.), 2-(7-aza-1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
HATU) (1.2 eq.), DIEA (1.5 eq.), N,N-Dimethylformamide (DMF), rt, 2 hrs, 60-80%; (f)
R Cl (1.1 eq.), DIEA (1.2 eq.), DCM, 0 °C, 30 minutes, 56-90%.
2
O (0.3 eq.), 90 °C, argon, 24 hrs, 32-60% for two steps; (c)
2
argon, 18 hrs; (d) TFA (10.0 eq. ), DCM, rt, 2 hrs, 33-40% for two steps; (e) R
1
(
1

RESULT AND DISCUSSION
Based on preliminary computational investigation, the acrylamide moiety of compound
5
a was first saturated to yield the corresponding propionamide 5b to prevent Michael
reaction with EGFR or other proteins containing similarly positioned cysteine residues.
Given the technical difficulties in determining TTK kinase inhibitory activity [52], the
binding affinity of the novel compounds was primarily evaluated by utilizing an active
site-directed competition binding assay (conducted by Eurofins DiscoveryX Corporation,
San Diego, CA, USA) to guide the structural optimization [53]. Inhibitor 3 was utilized as
a positive reference compound and exhibited a binding constant K
d
value of 0.071 nM
(Table 1). Highly consistent with our predictions, compound 5b exhibited a similar TTK
binding affinity to that of compound 5a with a K value of 145 nM, while its potency
d
against EGFR was significantly decreased (K > 1000 nM). Utilizing compound 5b as a
d
lead molecule, new compounds were designed and synthesized by replacing the
pyrrolidinyl linker moiety with a variety of heterocyclic groups (5c-5i). This confirmed
that the linker group had significant impact on TTK binding activity (Table 1). Although
replacement of the S-pyrrolidinyl linker with its R- isomer (5d) only caused a 2-fold loss
of potency, an azetidinyl substitution (5c) demonstrated almost totally abolished TTK
binding potency. The 4-piperidyl linked molecule (5g) exhibited strong binding with TTK
with a K value of 9.3 nM, but the R- (5e) or S- (5f) 3-piperidyl substituted molecules
d
were approximately 27 and 21 fold less potent, respectively. Encouragingly, when a cis-
(5h) or trans- (5i) 4-amino-cyclohexyl linker was introduced, the resulting compounds
tightly bound TTK with K values of 0.71 nM and 2.8 nM, respectively.
d

A computation study was also conducted to rationalize the diverse impacts of the
different linkers on the TTK binding affinity (Figure 3 and Supporting Information Figure
S1). It was shown that the three more potent molecules, i.e. 5g, 5h and 5i, demonstrated
overall similar binding poses to that of compound 5b (Figure 3). However, the increased
volume of the linkers present in 5g, 5h and 5i may favor improved hydrophobic
interactions with Ala651, Met671 and Pro673, when compared to those of 5b. It was also
shown that compounds with S- pyrrolidinyl (5b) and 3-piperidyl (5f) linkers would
achieve a better fit into the hydrophobic sandwich consisting of Ala651 and Met671 in
TTK (Supporting Information Figure S1). Alternately, steric conflict was observed
between 5c and Pro673 of TTK, which may explain its significant loss of binding affinity
(Supporting Information Figure S1B). Additionally, hydrogen-bond interaction networks
were also observed between the carboxyl amide of the inhibitors and TTK. For instance,
the propionamide moiety of 5h was sandwiched between Met671 and Ala651 to form
strong hydrogen-bond interactions that benefit binding with TTK, while only one
hydrogen bond could form in this region for compounds 5g or 5i. The computational
study further suggested that the ethyl moiety in 5h is located in a hydrophobic cavity
formed by Asn652, Ile663, Gln670, Met671, Pro673 and Val678 in the loop region, and
that a hydrophobic group with increased size could be accommodated in this pocket to
potentially improve binding affinity further. Thus, several new molecules with different
hydrophobic tails (5j-5r) were designed and synthesized. The results showed that while
the acetyl analog 5j demonstrated similar potency to that of 5h, compounds with larger
hydrophobic groups indeed exhibited improved binding affinities with TTK. For instance,

the cyclopropyl (5k), iso-propyl (5l), and phenyl (5q) substituted compounds displayed
improved binding affinities with K values of 0.41, 0.50 and 0.30 nM, respectively.
d
Significantly, compounds with tert-butyl (5m), cyclopentyl (5n) and cyclohexanyl (5o)
substituents exhibited the strongest affinity with K values of 0.20, 0.12 and 0.15 nM,
d
respectively. It was also noteworthy that this position was able to accommodate large
hydrophobic groups such as adamantyl (5p) or 1-naphthyl (5r) substituents (K values of
d
0
.56 and 1.1 nM, respectively). We have predicted that the carbonyl group in the alkyl
amide moiety could form a hydrogen bond with Met671 (Figure 3) improving binding
with TTK, consistent with this removal of the carbonyl group in 5k (to give 5s) caused an
approximate 15-fold loss of potency (5s exhibited a K value of 6.1 nM). Per-residue
d
binding free energy decomposition clearly suggested that the predicted hot-spot residues
(i.e., Asn654, Pro673 and Val678 in loop region) contributed greatly to the binding of 5o
with TTK (Supporting Information, Figure S2).
The Structure-Activity Relationship (SAR) investigation also revealed that the
2
’-position in the west phenyl ring favors hydrophobic group substitution. For instance,
the 2’-methoxyl (5h) and 2’-methyl (5ha) derivatives exhibited TTK binding affinity
with K values of 0.71 and 0.63 nM, respectively. Whereas, the corresponding
d
3
d
’-substituted molecules 5hb and 5hc were 6-fold less potent (K values of 4.2 and 3.9
nM, respectively). Our computational investigation suggested that additional interactions
of the 2’-methoxyl or 2’-methyl group with Cys604 and Gln541 residues of TTK may
contribute to the improved potencies of 5h and 5ha (Supporting Information, Figure S3).
Our previous studies have suggested that a hydrophobic substituent at the 5-position of

the pyrido[2, 3-d]pyrimidin-7(8H)-one ring system could contribute greatly to the target
specificity of the compounds, although it might be detrimental to a compound’s potency
[50]. Similar to previous observations [54], removal of the 5-methyl group in 5h resulted
in a 2-fold potency improvement (5t). However, when the 5-methyl group was replaced
by an ethyl (5u) or cyclopropyl (5v) group, the resulting molecules exhibited 2- and 11-
fold less potency, respectively. Most notable, the 5-phenyl compound (5w) displayed a Kd
value of 825 nM, which was approximately 1162-fold less potent than compound 5h.
Interestingly, the 5-methyl group of 5h could be moved to the 6- position to provide a
3
-fold potency improvement (5x). The corresponding cyclopentanecarbonyl and
cyclohexanecarbonyl substituted derivatives 5y and 5z also exhibited similar binding
potency to that of 5n and 5o, with K values of 0.17 and 0.16 nM, respectively.
d
Kinase inhibitory activity of selected compounds was further determined by utilizing a
direct filter-binding radiometric kinase activity assay to validate their TTK inhibitory
potency (conducted by Eurofins Cerep SA, France) (Table 2). Compounds 5n, 5o, 5y and
5
z exhibited strong TTK kinase inhibition with IC50 values of 19, 23, 11 and 20 nM,
respectively, which were comparable to that of compound 3 (IC50, 17 nM). Highly
consistent with the measured weaker binding affinities of compounds 5v and 5b they also
displayed poorer kinase inhibitory activities with IC50 values of 50 and >1000 nM,
respectively. Collectively, compounds 5n, 5o, 5y and 5z represented the most potent TTK
inhibitors in this series displaying comparable potency to that of the clinical candidate
compound 3.
a
Table 1. Binding affinity of compounds 5a-5z and 5ha-5hc against TTK

K
d
value of
TTK (nM)
.071
0.072,0.070)
d
K value of
Compd.
L
-
R1
R2
-
R3
-
Compd.
L
R1
R2
R3
H
TTK (nM)
0
0.50
3
-
5l
�
Me
(
(0.46, 0.54)
1
90
0.20
5
a
�
Me
Me
H
H
5m
5n
�
�
Me
Me
H
H
(
(
180, 200)
(0.21, 0.18)
1
45
0.12
5
b
�
160, 130)
(0.11, 0.12)
>
1000
(>1000, >10
0)
10
280, 330)
0
.15
5
c
�
�
Me
Me
H
H
5o
5p
�
�
Me
Me
H
H
(
0.13, 0.17)
0
3
0.56
5
d
(
(0.61, 0.51)
2
50
0.30
5
e
�
�
�
�
�
�
�
Me
Me
Me
Me
Me
Me
Me
H
H
H
H
H
H
H
5q
5r
5s
�
�
�
�
�
�
�
Me
Me
Me
H
H
H
H
H
H
H
H
(
(
240, 260)
(0.26, 0.34)
2
00
1.1
5
f
160, 240)
(0.75, 1.5)
9
.3
6.1
5
5
g
(
8.8, 9.8)
(5.4, 6.8)
0
.71
0.41
h
5t
(
0.61, 0.80)
(0.47, 0.34)
2
.8
1.5
5
i
5u
5v
5w
Et
(
2.8, 2.8)
（1.4, 1.5）
0
.63
cycli
c-Pr
7.9
5
ha
hb
(
0.57, 0.69)
(6.5, 9.3)
4
.2
825
5
Ph
(
3.6, 4.8)
(870, 780)

3
.9
0.29
5
hc
�
�
�
Me
Me
Me
H
H
H
5x
5y
5z
�
�
�
H
H
H
Me
Me
Me
(
3.7, 4.0)
(0.32, 025)
0
.72
0.17
5
j
(
(
0.66, 0.78)
(0.19, 0.15)
0
.41
0.17
5
k
0.45, 0.36)
(0.2, 0.13)
a
The binding affinity (K ) was determined by utilizing an active site-directed competition
d
binding assay (conducted by Eurofins DiscoveryX Corporation, San Diego, CA, USA).
The data is the mean from two independent experiments. The independent values are
presented in the brackets.
a
Table 2. In vitro TTK Kinase Inhibition of the selected compounds
Compd.
3
5b
5n
5o
5v
5y
11
5z
20
1
7
>1000
19
23
50
IC50 (nM)
(17, 17)
(>1000, >1000)
(17, 20) (21, 24) (52, 48)
(12, 9)
(21, 18)
a
The isolated enzyme inhibitory potency (IC50) was determined by utilizing a direct
filter-binding radiometric kinase activity assay (conducted by Eurofins Cerep SA, France).
The data is the mean from two independent experiments. The independent values are
presented in the brackets.

Figure 3. Representatives of the predicted binding modes of compounds 5g, 5h and 5i in
the active site of TTK (PDB code 5EH0). (A) The predicted binding mode of 5g in TTK;
(B) The predicted binding mode of 5h in TTK; (C) The predicted binding mode of 5i in
TTK; (D) Superposition of compounds 5b, 5g, 5h and 5i.
The kinome selectivity of compounds 5o and 5z were also evaluated against a panel of
4
68 kinases (including 403 wild-type kinases) at a single concentration of 100 nM, which
is over 600 times their respective K values against TTK. Only a small number of other
d
kinases were inhibited by 5o. 22 out of the 403 kinases exhibited more than 65%

inhibition at the tested concentration with 5 (i.e. DAPK3, IRAK1, MARK3, PIK3C2G
and PIK3CD) displaying more than 90% inhibition (Table 3). The results also showed
that 5z exhibited relatively poorer kinome selectivity than 5o. 47 potential off-target
kinases (>65% inhibition) were identified with 19 showing more than 90% inhibition
(Supporting Information Table S4). Further binding affinity determinations confirmed
that 5o bound with DAPK3 and MARK3 with K values of 13 and 11 nM, respectively,
d
being 87- and 73- fold greater than that that observed for TTK, respectively (Table 3).
However, compound 5o did not show strong binding affinity with IRAK1, PIK3C2G and
d
PIK3CD (K values > 1000 nM) as was initially suggested from the 100 nM
single-concentration screen. These data collectively supported 5o as a highly selective
TTK inhibitor.
a
Table 3 Kinome selectivity profiles of 5o
Target
PIK3CD
DAPK3
MARK3
JNK3
%Inhibition
K (nM)
Target
TTK
%Inhibition
K (nM)
d
d
100
94.5
92
>1000
98.1
94.4
91.8
87
0.15(0.13, 0.17)
13 (12, 13)
PIK3C2G
IRAK1
PHKG2
JNK2
>1000
11(11, 11)
>1000
87
-
-
-
-
-
-
-
-
-
-
-
-
STK16
87
86
DAPK1
SNARK
DCAMKL1
MELK
85
ULK2
85
84
DMPK
PIPK5
JNK1
81
80
77
77
75

STK39
CSNK2A1
DAPK2
75
69
65
-
-
-
GAK
70
68
-
-
ULK1
a
Kinome selectivity profile of compound 5o against a panel of 468 kinases at a
concentration of 100 nM. Kinases with > 65% inhibition are listed. K values of major
d
potential “off-target” kinases for 5o are also reported (conducted by Eurofins DiscoveryX
Corporation, San Diego, CA, USA).
TTK is a kinetochore-associated kinase essential for vertebrate mitotic checkpoint
death. TTK inhibition leads to impairment of SAC, and induces chromosome
missegregation and aneuploidy [43, 55-58]. To examine the effect of compound 5o on
chromosome segregation, HCT-116 human colorectal cancer cells were treated with 5o
and the number of chromosome segregation errors present in mitotic cells was
determined by immunostaining for centromeres (centromere protein B, CENP-B, green),
spindle (β-tubulin, red) and nucleus (4',6-diamidino-2-phenylindole, DAPI, blue). Two
less active inhibitors (i.e. 5c and 5b) and compound 3 were also included as the negative
and positive controls, respectively. It was shown that treatment with 5o (333.3 nM)
caused an obvious increase in chromosome missegregations when compared to untreated
cells (Figure 4A, 4B). SAC inactivation and gross chromosome segregation errors caused
by 5o had an impact on the cell cycle and survival of human cancer cells. Compound 5o
induced dose-dependent dysregulation of the cell cycle, resulting in an increase in the
frequency of cells exhibiting aneuploid DNA content (Figure 4C). Compound 5o-induced
aneuploidy was accompanied by a progressive accumulation of apoptotic cells which

were detected by flow cytometry analysis (Figure 5E) after 48 hours of treatment.
Western blot analysis also suggested that 5o induced activation of apoptosis-related
proteins, such as caspase 3, caspase 9 and poly ADP-ribose polymerase (PARP) (Figure
5
D) in a dose-dependent manner. Phosphorylated histone H3, a biomarker for SAC
over-ride and TTK inhibition [55, 58, 59], was also dose-dependently suppressed by 5o
Figure 5D). Not surprisingly, the positive control 3 exhibited similar aneuploidy
(
induction in HCT-116 human colorectal cancer cells as determined by the increase of
aneuploid DNA content, while the target-inactive compounds 5b and 5c did not show any
obvious effect (Supporting Information, Figure S4).
Figure 4. Compound 5o inhibits the function of TTK in HCT-116 human colorectal
cancer cells. (A) Compound 5o causes aberrant chromosomal segregation in HCT-116

human colorectal cancer cells. Mitotic figures of HCT-116 human colorectal cancer cells
treated with compound 5o or DMSO for 24 hrs are displayed. Centromeres are stained in
green, β-tubulin in red, and DAPI in blue; (B) Percentage of anaphase and telophase cells
with lagging/ no lagging chromosomes (mean from two independent experiments, fifty
cells were examined per treatment condition). Percentage of lagging chromosomes:
DMSO: 3.8 ± 1.1%; 111.1 nM compound 5o: 16.1 ± 4.6%; 333.3 nM compound 5o: 60.6
±
1.9%; (C) Compound 5o induces aneuploidy in HCT-116 human colorectal cancer cells.
HCT-116 human colorectal cancer cells were treated with DMSO (control) or 5o for 48
hrs. DNA content was assessed by flow cytometry.

Figure 5. Effects of 5o on proliferation and colony formation in cancer cells. (A) 5o
inhibits proliferation of cancer cells in vitro; (B) 5o suppresses colony formation in
HCT-116 human colorectal cancer cells. Cells were stained by crystal violet after
treatment with DMSO (control) or 5o for 7 days; (C) Colony numbers of assay were
described in Figure 5B; (D) 5o inhibits phosphorylation of histone H3 and activates

apoptosis-related proteins. HCT-116 human colorectal cancer cells were treated with
DMSO (control) or 5o for 48 hrs. Representative results were shown and similar results
were obtained in other independent trials; (E) 5o induced apoptosis in HCT-116 human
colorectal cancer cells. HCT-116 human colorectal cancer cells were treated with DMSO
(control) or 5o for 48 hrs before DNA labeling by Annexin V and 7-AAD and cell
apoptosis analysis by flow cytometry.
The anti-proliferative activity of 5o against HCT-116 and COLO-205 human colorectal
cancer cells, and A549 and NCI-H460 human lung cancer cells was also determined by
using a Cell Counting Kit-8 (CCK-8) assay. It was shown that compound 5o potently
inhibits the growth of human cancer cells with IC values of 0.41, 0.28, 4.40 and 1.09
50
µ
M, respectively (Figure 5A, Supporting information, Table S1). Additionally, 5o also
exhibited moderate inhibition of HCT-116 human colorectal cancer cell colony formation
(
Figure 5B).
Pharmacokinetic parameters of the representative molecules 5o and 5z were also
determined in SD rats to evaluate their drug-like properties (Table 4). It was shown that
both compounds exhibited good pharmacokinetic properties with high plasma exposure
(
AUC), reasonable half-life (T ) and acceptable oral bioavailability (F) values. For
1/2
instance, compounds 5o and 5z demonstrated AUC(0-∞) values of 13,021 and 33,346
g/L)·h, T1/2 values of 6.0 and 7.6 hrs, and oral bioavailability of 45.3% and 39.6%,
(µ
respectively, at an oral dose of 25 mg/kg.
a
Table 4. Pharmacokinetic parameters of 5o and 5z in SD rats
Com
pd.
AUC(0-∞)
(( g/L)·h)
route
T
1/2 (h)
C
max
(
µ
g/L)
CLz (mL/h/kg)
F (%)
µ

IV
3.58
2525.52
±500.46
5307.22
±449.55
946.43
±76.46
(
5 mg/kg)
±0.03
5
o
PO
6.01
1228.70
±63.51
13021.02
±1419.37
45.25
±4.76
(
25mg/kg)
±0.30
IV
6.20
4739.97
±733.38
15453.16
±3580.87
336.79
±86.06
(
5 mg/kg)
PO
±0.95
5
z
7.62
2304.63
±304.53
33346.28
±2227.15
39.63
±2.71
(
25 mg/kg)
±0.75
a
SD rat (male, three animals per group) weighing 190-230 g were used for the study.
In vivo efficacy of 5o was further investigated as a single-agent and in combination
with paclitaxel in a xenograft mouse model of HCT-116 cancer cells. CB17-SCID mice
bearing HCT-116 tumor xenografts were treated with paclitaxel (20 mg/kg,
intraperitoneal (ip), quaque week (QW)), 5o (50 mg/kg, po, QD, 2 days on/4 days off)
alone, or with a combination of paclitaxel (20 mg/kg, ip, QW) and 5o (50 mg/kg, po, QD,
2
days on/4 days off), respectively. Compound 5o or paclitaxel monotherapy exhibited
moderate tumor growth inhibition with TGI values of 29% and 52%, respectively.
Encouragingly, the combination strongly improved the in vivo efficacy with a TGI of
7
8%. This might be due to the fact that a TTK inhibitor could reduce paclitaxel-induced
mitotic arrest and accelerate progression of cells through mitosis by weakening SAC
activity, which will eventually lead to increased chromosomal segregation errors and cell
death [46, 60, 61]. It is noteworthy that all treatment protocols were well tolerated with
body weight loss values less than 15% and no significant damage to liver and kidney
function observed (Figure 6C, 6D, Supporting information Table S2, Table S3).

Figure 6. In vivo anticancer study of 5o in combination with paclitaxel against HCT-116
tumor xenografts. CB17-SCID mice (n = 5) bearing HCT116 tumor xenografts were
treated with 5o, paclitaxel, a combination of 5o and paclitaxel, and vehicle, respectively.
*
P < 0.05, **P<0.01 (Student’s t-test), compared with vehicle. (A) Tumor volume change
in HCT-116 xenograft study; (B) Tumor weight of each treatment group at the end point
of treatment; (C) Body weight change in HCT-116 xenograft study; (D) Hematoxylin and
eosin (H&E) staining images of the dissected kidney, liver and tumor at the end point of
treatment.
CONCLUSIONS
In summary, a series of new pyrido[2, 3-d]pyrimidin-7(8H)-ones were designed and
synthesized as new orally bioavailable selective TTK inhibitors. One of the representative

compounds, 5o, tightly binds TTK with a K
d
value of 0.15 nM and potently suppresses
TTK kinase function with an IC value of 23 nM. Compound 5o also displays reasonable
5
0
kinome selectivity against a panel of 403 wild-type kinases at 100 nM. Additionally,
compound 5o exhibits good oral pharmacokinetic properties with a Cmax value of 1228.70
µ
g/L, a T1/2 value of 6.0 hrs, and a bioavailability value of 45.3%, respectively, at an
administered dose of 25 mg/kg in SD rats. Further biological investigation demonstrates
that compound 5o potently induces chromosome missegregation and aneuploidy,
exhibiting promising anti-proliferative activity in HCT-116 human colorectal cancer cells
with an IC50 value of 0.41
µM. Combination of 5o with paclitaxel demonstrated
promising in vivo efficacy in an HCT-116 xenograft model with a tumor growth
inhibition (TGI) value of 78%. Inhibitor 5o may be utilized as a new research tool for
further biological investigation to validate the therapeutic potential of TTK inhibition.
EXPERIMENTAL SECTION
General Methods for Chemistry. All reagents and solvents used were purchased from
commercial sources without further purification. Flash chromatography was performed
using 300 mesh silica gel. All reactions were monitored by thin-layer chromatography
1
(
TLC) using silica gel plates with fluorescence F254 and UV light visualization. H NMR
spectra were recorded on a Bruker AV-400 spectrometer at 400 MHz or a Bruker AV-500
1
3
spectrometer at 500 MHz. C NMR spectra were recorded on a Bruker AV-500
spectrometer at 125 MHz. Coupling constants (J) are expressed in hertz (Hz). Chemical
shifts (δ) of NMR are reported in parts per million (ppm) units relative to an internal
standard (TMS). High resolution ESI-MS on an Applied Biosystems Q-STAR Elite

ESI-LC-MS/MS mass spectrometer. Purity of all of the final compounds were determined
by reverse phase high performance liquid chromatography (HPLC) analysis to be >95%.
HPLC instrument: Dionex Summit HPLC (column, Diamonsil C18, 5.0
µm, 4.6 mm ×
2
50 mm (Dikma Technologies); detector,PDA-100 photodiode array; injector, ASI-100
autoinjector; pump, p-680A). A flow rate of 1.0 mL/min was used with mobile phase of
MeOH in H O with 0.1% modifier (ammonia, v/v).
2
General Procedure for the Preparation of Compounds 5e-5g, 5v and 5w.
Ethyl4-(((1S,4S)-4-((tert-butoxycarbonyl)amino)cyclohexyl)amino)-2-(methylthio)
pyrimidine- 5-carboxylate (7v). To a suspension of ethyl 4-chloro-2- (methylthio)
pyrimidine-5-carboxylate (31.4 g, 135 mmol) and K CO (37.3 g, 270 mmol) in MeCN
2
3
(200 ml), tert-butyl((1S, 4S)-4-aminocyclohexyl)carbamate (32.2 g, 150 mmol) was
added in batches within 30 minutes. The mixture was stirred at room temperature
overnight. After the reaction was completed as monitored by TLC, the solution was
concentrated by rotary evaporation, and the residue was diluted with ethyl acetate (EA)
and water. The aqueous layer was re-extracted with EA for 3 times. The combined
2 4
organic layers were washed with water and brine, dried over Na SO , and concentrated in
vacuo. The residue was purified by flashing column chromatography (SiO , 10-20% EA
2
1
in petroleum ether) to give 7v (white solid, 42.2 g, yield 76%). H NMR (400 MHz,
CDCl3)
δ 8.62 (s, 1 H), 8.46 (d, J = 6.1 Hz, 1 H), 4.55 (s, 1 H), 4.32 (m, 3 H), 3.62 (s,
1
H), 2.51 (s, 3 H), 1.87-1.77 (m, 6 H), 1.53 (m ,2 H), 1.45 (s, 9 H), 1.37 (t, J = 7.0 Hz, 3
+
H). MS (ESI) m/z 411.2 [M+H] .

tert-Butyl((1
S,4S)-4-((5-(hydroxymethyl)-2-(methylthio)pyrimidin-4-yl)amino)cycl
ohexyl)carbamate (8v). A suspension of LiAlH (14.0 g, 368.3 mmol) in anhydrous THF
4
(100 mL) was added dropwise to a solution of 7v (75.5 g, 184.1 mmol) in anhydrous THF
(300 mL) under stirring for 1 hr at -40 °C. The mixture was then slowly warmed to 0 °C,
2
and stirred for another 2 hrs. 14 mL H O was added dropwise to quenched the reaction,
and further 14 mL 15% aqueous NaOH and 14 mL H O was added consequently. The
2
mixture was stirred at room temperature for 10 minutes and was diluted with EA and
DCM, filtered through celite. The combined organic layers were dried over Na SO , and
2
4
concentrated in vacuo. The residue was purified by flashing column chromatography
1
(
SiO , 2.5- 5% MeOH in DCM) to give 8v (white solid, 37.2 g, yield 55%). H NMR
2
(400 MHz, CDCl3)
δ 7.54 (s, 1 H), 6.17 (d, J = 6.8 Hz, 1 H), 4.66 (s, 1 H), 4.46 (s, 2 H),
4
.18 (s, 1 H), 3.58 (s, 1 H), 2.45 (s, 3 H), 1.75 (m, 6 H), 1.49 (m, 2 H), 1.42 (s, 9 H). MS
+
(
ESI) m/z 369.1 [M+H] .
tert-Butyl((1 ,4 )-4-((5-formyl-2-(methylthio)pyrimidin-4-yl)amino)cyclohexyl)ca
S
S
rbamate (9v). MnO (40.0 g, 46.2 mmol) was added to a solution of 8v (17.0 g, 46.2
2
mmol) in DCM (200 ml) in three times within 1.5 hrs. The mixture was stirred at room
temperature overnight, and filtered through celite. The organic layer was concentrated in
2
vacuo, and the residue was purified by flashing column chromatography (SiO , 2.5 - 5%
1
MeOH in DCM) to give 9v (white solid, 15.7 g, yield 93%). H NMR (400 MHz, CDCl3)
δ
9.69 (s, 1 H), 8.79 (d, J = 6.4 Hz, 1 H), 4.59 (s, 1 H), 4.34 (s, 1 H), 3.63 (s, 1 H), 2.52 (s,
+
3
H), 1.83 (m, 6 H), 1.56 (m, 2 H), 1.44 (s, 9 H). MS (ESI) m/z 367.2 [M+H] .

tert-Butyl((1
yl)amino)cyclohexyl)carbamate (10v). Cyclopropylmagnesium bromide in THF (1 N,
5.8 mL, 25.8 mmol) was added dropwise to a solution of 9v (4.7 g, 12.9 mmol) in
anhydrous THF (25 mL) at -78 °C. The mixture was then slowly warmed to -40 °C, and
stirred for another 2 hrs. The reaction was quenched with aqueous NH Cl and extracted
with DCM. The combine organic layers were dried over Na SO , and concentrated in
S,4S)-4-((5-(cyclopropyl(hydroxy)methyl)-2-(methylthio)pyrimidin-4-
2
4
2
4
2
vacuo. The residue was purified by flashing column chromatography (SiO , 2.5- 5%
1
MeOH in DCM) to give 10v (white solid, 3.3 g, yield 63%). H NMR (400 MHz, CDCl3)
δ
1
1
4
7.80 (s, 1 H), 6.48 (d, J = 6.6 Hz, 1 H), 4.54 (s, 1 H), 4.22 (s, 1 H), 3.77 (d, J = 8.6 Hz,
H), 3.60 (s, 1H), 2.60 (br, 1 H), 2.48 (s, 3 H), 1.78 (m, 6 H), 1.50 (m, 2 H), 1.45(s, 9 H),
.34 (m, 1 H), 0.71 (m, 1 H), 0.63 (m, 1 H), 0.46 (m, 1 H), 0.27 (m, 1 H). MS (ESI) m/z
+
09.7 [M+H] .
tert-Butyl((1
S,4S)-4-((5-(cyclopropanecarbonyl)-2-(methylthio)pyrimidin-4-yl)ami
no)cyclohexyl)carbamate (11v). MnO (3.80 g, 43.7 mmol) was added to a solution of
2
1
0v (1.78g, 4.4 mmol) in DCM (20 ml) in three times within 1.5 hrs. The mixture was
stirred at room temperature overnight, and filter through celite. The organic layer was
concentrated in vacuo, and the residue was purified by flash column chromatography
1
(
SiO
2
, 2.5 - 5% MeOH in DCM) to give 11v (white solid, 1.42 g, yield 80%). H NMR
(400 MHz, CDCl3)
δ
9.52 (d, J = 7.0 Hz, 1H), 8.79 (s, 1 H), 4.56 (s, 1 H), 4.33 (m, 1 H),
3
1
.60 (s, 1 H), 2.52 (br, 4 H), 1.84 (m, 2 H), 1.77 (m, 4 H), 1.51 (m, 2 H), 1.44 (s, 9 H),
+
.18 (m, 2 H), 1.00 (m, 2 H). MS (ESI) m/z 407.5 [M+H] .