Novel CDKs inhibitors for the treatment of solid tumour by simultaneously regulating the cell cycle and transcription control

Article abstract of DOI:10.1080/14756366.2019.1705290

A novel series of cyclin-dependent kinases (CDKs) inhibitors, which play critical roles in the cell cycle control and regulation of cell transcription, were synthesised. A systematic study of enzymatic and cellular assays led to the identification of compound X22 with a nanomolar potency against CDK4 and CDK9 and potent antiproliferative activities against a panel of tumour cell lines. X22 could induce cell cycle arrest and cell apoptosis in cancer cell lines. X22 dose-dependently inhibits signalling pathways downstream of CDKs in cancer cells. In vivo antitumor activity assays, oral administration of X22 led to significant tumour regression in mouse model without obvious toxicity. Superior anti-cancer efficacy in vitro and in vivo of X22 demonstrated combined depletion of cell cycle and transcriptional CDK all contributed to antitumor activity. Taken together, concomitant inhibition of cell cycle and transcriptional CDK activities provided valuable guide for further structural optimisation.

Full text of DOI:10.1080/14756366.2019.1705290

Journal of Enzyme Inhibition and Medicinal Chemistry
ISSN: 1475-6366 (Print) 1475-6374 (Online) Journal homepage: https://www.tandfonline.com/loi/ienz20
Novel CDKs inhibitors for the treatment of solid
tumour by simultaneously regulating the cell cycle
and transcription control
Xin Wang, Kaiyuan Deng, Cheng Wang, Yao Li, Tianqi Wang, Zhi Huang,
Yakun Ma, Peiqing Sun, Yi Shi, Shengyong Yang, Yan Fan & Rong Xiang
To cite this article: Xin Wang, Kaiyuan Deng, Cheng Wang, Yao Li, Tianqi Wang, Zhi Huang,
Yakun Ma, Peiqing Sun, Yi Shi, Shengyong Yang, Yan Fan & Rong Xiang (2020) Novel CDKs
inhibitors for the treatment of solid tumour by simultaneously regulating the cell cycle and
transcription control, Journal of Enzyme Inhibition and Medicinal Chemistry, 35:1, 414-423, DOI:
10.1080/14756366.2019.1705290
To link to this article: https://doi.org/10.1080/14756366.2019.1705290
©
2020 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group.
View supplementary material
Published online: 03 Jan 2020.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=ienz20

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
020, VOL. 35, NO. 1, 414–423
2
https://doi.org/10.1080/14756366.2019.1705290
SHORT COMMUNICATION
Novel CDKs inhibitors for the treatment of solid tumour by simultaneously
regulating the cell cycle and transcription control
a
ꢀ
a
ꢀ
a
a
a
a
a
a,e
Xin Wang , Kaiyuan Deng , Cheng Wang , Yao Li , Tianqi Wang , Zhi Huang , Yakun Ma , Peiqing Sun ,
a,b
d
a,b
a,b,c
Yi Shi , Shengyong Yang , Yan Fan and Rong Xiang
a
b
Department of Medicinal Chemistry, School of Medicine, Nankai University, Tianjin, China; 2011 Project Collaborative Innovation Center for
c
d
Biotherapy of Ministry of Education, Tianjin, China; State Key Laboratory of Medicinal Chemical Biology, Tianjin, China; Department of Medical
Oncology, Cancer Center, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Chengdu, China; Department of Cancer
e
Biology, Wake Forest Comprehensive Cancer Center, Wake Forest School of Medicine, Winston-Salem, NC, USA
ABSTRACT
ARTICLE HISTORY
A novel series of cyclin-dependent kinases (CDKs) inhibitors, which play critical roles in the cell cycle con-
trol and regulation of cell transcription, were synthesised. A systematic study of enzymatic and cellular
assays led to the identification of compound X22 with a nanomolar potency against CDK4 and CDK9 and
potent antiproliferative activities against a panel of tumour cell lines. X22 could induce cell cycle arrest
and cell apoptosis in cancer cell lines. X22 dose-dependently inhibits signalling pathways downstream of
CDKs in cancer cells. In vivo antitumor activity assays, oral administration of X22 led to significant tumour
regression in mouse model without obvious toxicity. Superior anti-cancer efficacy in vitro and in vivo of
X22 demonstrated combined depletion of cell cycle and transcriptional CDK all contributed to antitumor
activity. Taken together, concomitant inhibition of cell cycle and transcriptional CDK activities provided
valuable guide for further structural optimisation.
Received 25 September 2019
Revised 5 December 2019
Accepted 9 December 2019
KEYWORDS
CDK4; CDK9; inhibitor;
cancer
1
. Introduction
major interest of many groups. CDK9 forming heterodimeric com-
plex with subunit cyclin T or cyclin K phosphorylates the COOH-
terminal domain (CTD) of RNA polymerase II (RNAPII) to control
Cancer is a multigenetic disease with the hallmark that multiple
signalling pathways aberration, which often require multiple thera-
peutic interventions . Kinases mediate various cellular activities
due to their critical roles in cellular signalling, such as prolifer-
ation, apoptosis, transcription, differentiation and so on , which
have been demonstrated as promising drug targets for the treat-
ment of many diseases such as cancers . Kinase inhibitors are
widely employed in clinical oncology . Simultaneous inhibition of
different mechanisms using multi-kinase inhibitors could yield
superior efficacy, such as synergy effects, avoiding drug resistance
and so on .
Out of numerous kinases, cyclin-dependent kinases (CDKs) are
protein kinases involved in important cellular processes due to
the complexity of their roles. They regulate the cell division, apop-
tosis, transcription and differentiation, which involved in a number
of pathological conditions such as human cancer
divided into two groups based on their roles in cell cycle progres-
1
6
1
,2
the transcription progression . As a component of a larger pro-
tein complex called positive transcription elongation factor b
(P-TEFb), CDK9 stimulate transcription elongation of most protein
3
,4
1
7
coding genes . Phosphorylation of RNAPII is regarded as a marker
involved in a variety of human pathological conditions, such as
cancer. Inhibition of CDK9 results in reduced levels of antiapop-
5–7
8
,9
1
8,19
totic proteins of cancer
.
Dysregulation of CDKs-cyclin pathway resulting in uncontrolled
2
0–22
1
0
proliferation have been observed in various cancers
. Owing to
the important role of CDKs in the control of cell division, numer-
ous drugs targeting CDKs were designed and reported in the
2
3
clinic . Several selective CDK4/6 inhibitors have been reported in
the recent years. Pfizer’s palbociclib represents the first CDK
inhibitor approved by the US FDA in February 2015 to treat HRþ/
11,12
. CDKs are
2
4,25
HER2-breast cancer
. Novartis’ ribociclib and Lilly’s abemaciclib
1
3,14
also got a fast approval by FDA in March 2017 and September
sion and transcription regulation
CDK subtypes 1, 2, 4 and 6 mainly is involved in cell cycle and the 2017, respectively, to treat HRþ/HER2-breast cancer
second group including CDK subtypes 7, 8 and 9 are associated reported, both palbociclib and ribociclib as selective CDK4/6 inhib-
with transcription control . CDK4/6 binding to cyclin D forming itors need to combine with letrozole for the treatment of breast
CDK4/6-cyclin D complex, is an important component of cell cycle cancer
. The first group including
2
6,27
. While as
11
2
8,29
. Abemaciclib with additional kinase activities, such as
3
0,31
activation and mediates the transition from G to S phase, where CDK1, CDK2 and CDK9 shows unique single-agent activities
.
1
cells grow and synthesise proteins in preparation for DNA These studies indicate that monotherapy targeting of individual
1
5
synthesis . Among the transcriptional CDKs, CDK9 has attracted cell cycle CDKs may be insufficient for cancer therapy.
CONTACT Yan Fan
yanfan@nankai.edu.cn
Department of Medicinal Chemistry, School of Medicine, Nankai University, 94 Weijin Road, Tianjin, China; Rong
Xiang
rxiang@nankai.edu.cn Department of Medicinal Chemistry, School of Medicine, Nankai University, 94 Weijin Road, Tianjin 300071, China
ꢀ
These authors contributed equally to this work.
Supplemental data for this article can be accessed here.
ß 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
415
Simultaneous regulating the cell cycle and transcription control 2.4. Kinase inhibition assays
32–34
could provide superior anticancer efficacy
. In an effort to dis-
Kinase inhibition profiles were determined using KinaseProfiler
services provided by Eurofins, and ATP concentrations used the
Km of corresponding kinases.
cover novel CDKs inhibitors with high inhibition potency against
both cell cycle and transcriptional CDKs, we recently started a
medicinal chemistry study based on ribociclib, a latest and select-
ive CDK4 inhibitor and performed
a rational drug design.
According to the researches, many sulphur atom-containing com- 2.5. Cytotoxicity assay
pounds have immense importance in medicinal chemistry and
Cytotoxicity assay were conducted as described previously using a
Cell Counting Kit-8 (CCK-8) assay (#CK04, Dojindo, Kumamoto,
Japan) .
35
exhibit anticancer activity . Moreover, isothiocyanate plays a
prominent role in the field of pharmaceuticals with anticancer
37
36
effects in various cancer types . Sulphur atom or isothiocyanate-
containing derivatives are biologically active agents for drugs
used for the treatment of cancer. We reasoned to introduce sul- 2.6. Cell cycle assay
phur atom and isothiocyanate moiety based on pyrrolo-[2,3-d]
5
Cells were plated on six-well culture plates at a density of 5 ꢁ 10
pyrimidines-2-amine skeleton to proceed structural optimisation
on R, L and linker regions (Supplementary Figure S1). In our study,
we synthesised a series of novel compounds simultaneously medi-
ating the cell cycle and transcription control of CDKs with high
potent inhibition against solid tumour. Compound X22 targeting
CDK4 and CDK9 was discovered and exhibited potent antitumor
efficacy in vitro and in vivo.
cells/mL and were treated with the indicated concentrations of
X22 or ribociclib for 12, 24, 36 or 48 h after they adherence. Cells
were harvested and washed with phosphate buffered saline (PBS)
for three times and then fixed with ice cold 75% ethanol over-
night. The fixed cells were then washed with PBS and stained
with propidium iodide (50 mg/mL) in the presence of RNase A
0.5 mg) for 30 min at 37 C. The stained cells were then subjected
ꢃ
(
to flow cytometry (Modfit, BD) for cell cycle analysis.
2
. Materials and methods
2.7. Annexin V-FITC/PI apoptosis assay
2
.1. General methods for chemistry
5
Cells at a density of 3 ꢁ 10 cells/mL were seeded in six-well
plates and treated with compounds at different concentrations for
8 h. The cells were then harvested and washed twice with cold
PBS. Then the cells were subjected to an Annexin V/PI Apoptosis
Detection kit (BD Biosciences) for staining according to manufac-
ture’s instructions, and finally analysed by flow cytometry
(Modfit, BD).
The commercially obtained chemicals were used directly without
further purification. Solvents were purified and distilled following
the standard procedures. All the reactions were monitored by
thin-layer chromatography (TLC). The NMR spectra were taken on
4
1
a Bruker AV-400 MHz spectrometer (400 MHz for H and 101 MHz
1
3
for C) and chemical shifts were expressed in ppm downfield
using tetramethylsilane as the internal standard. High-resolution
mass spectra (HRMS) were performed on a VG ZAB-HS mass spec-
trometer under electron spray ionisation (ESI). All the derivatives
for testing bioactivity were purified to >95% purity which was
determined by HPLC analysis on a Shimadzu Prominence-i LC-
2.8. Western blotting
Protein extraction and western blotting methods were performed
as described previously . The antibodies used in this study
including anti-b-actin (Santa Cruz; sc-47778), anti-p53 (CST; #9282),
(CST; #9284), anti-Bax (CST; #2772), anti-Bcl-
2 (BD; #51-6511GR), anti-Rb (CST; #9309), anti-phospho-Rb
CST; #8516), anti-phospho-Rb
merase II CTD (Abcam; ab817), anti-RNA polymerase II CTD
3
7
2
030C 3D system (column, InertSustain C18, 4.6 ꢁ 250 mm, 5 lM;
Ser15
anti-phospho-p53
mobile phase, gradient elution of methanol/H
2
O (90:10); low rate,
Ser807/811
ꢃ
1
.0 ml/min; UV wavelength, 190 ꢂ 800 nm; temperature, 40 C;
Ser780
(
(CST; #9307), anti-RNA poly-
injection volume, 10 lL). The detailed synthesis of compounds is
phospho S5
presented in Supplementary Material.
phospho S2
(
Abcam; ab5131), anti-RNA polymerase II CTD
(Abcam;
ab5095).
2.2. Molecular docking
2.9. In vivo assay
Docking studies were finished by Discovery Studio 3.1 to explore
the predicted binding modes of compound X22 in CDK4, CDK6
The experimental procedures of the animal study were proved by
the Animal Care and Use Committee at Nankai University. 4T1
(PDB code: 4EZ5) and CDK9 (PDB code: 4BCF) respectively.
4
breast cancer cells (5 ꢁ 10 ) were injected in the mammary fat
Hydrogen atoms were added by Gold (version 5.0). The images
were created by PyMOL.
pads of 6 ꢂ 8 weeks old female BALB/c mice. Once the tumours
3
grew to a volume of approximately 100–150 mm , they were
placed into five treatment groups (n ¼ 3, with a total tumour num-
ber of 15). The mice were treated daily for 18 days via oral gavage.
Body weights and tumour size were determined every other day.
Tumour measurements were found using a digital vernier calliper,
2.3. Cell lines and cell culture
All the human cancer cell lines were obtained from ATCC. Cells and the volumes were determined using the following calculation:
2
were cultured in RPMI-1640 (BioInd) medium according to the (short ) ꢁ long ꢁ 0.5. Experiments were performed under an
instructions from ATCC, with the medium containing 10% FBS approved IACUC protocol according to federal and institutional
ꢃ
(
BioInd), 1% antibiotics (penicillin and streptomycin) at 37 C in an guidelines and regulations. Inhibition rate of tumour growth was
atmosphere of 5% CO
2
.
calculated using the following formula: 100 ꢁ f1 ꢂ [(tumour

4
16
X. WANG ET AL.
volumefinal ꢂ tumour volumeinitial) for X22-treated group]/[(tumour 3. Results and discussion
volumefinal ꢂ tumour volumeinitial) for the vehicle-treated group]g.
In our previous study, we found structural optimisation based on
the pyrrolo-[2,3-d] pyrimidines-2-amine scaffold could improve the
inhibitory activity against the cancer cell cycle as targeting CDK4/
3
8
6
and transcription as targeting CDK9 . As the noteworthy antitu-
2.10. Statistical analysis
mor functions of sulphur atom and isothiocyanate as mentioned
above, we introduced these moieties to form molecules to investi-
gate the structure–activity relationship. We changed the substitu-
ent at R position on the phenyl ring, as well as the functional
group L and length of spacer between the isothiocyanato and
Statistical analysis results were analysed values by GraphPad
Prism version 6.0 software. For Student t test and ANOVA,
p < 0.05 was considered statistically significant. Values were
2
expressed as means ± SEM. Significance was determined by v
test, others were determined by Student’s t-test. A value of phenyl ring. The pathways adopted for the synthesis of com-
p < 0.05 was used as the criterion for statistical significance. pounds X1–X23 were carried out as shown in Schemes 1–3.
ꢀ
ꢀꢀ
ꢀꢀindicates
indicates significant difference with p < 0.001,
p < 0.01,
The synthesis of the key intermediates 5a–5e was depicted in
Scheme 1. Firstly, refluxing the commercial obtained potassium
ꢀindicates p < 0.05.
a
a
ꢃ
ꢃ
Scheme 1. Synthesis of intermediates 5a–5e . Reagents and conditions: (A) Acetone, 70 C, overnight-; (B) Potassium thioacetate (3.0 eq), THF, 75 C, 5 h; (C) concen-
trated HCl, MeOH, 60 C, 4 h.
ꢃ
a
a
ꢃ
Scheme 2. Synthesis of intermediates 17a–18e, 19a–20e . Reagents and conditions: (A) K
2
CO
3
(2.0 eq), DMF, rt, 4.5 h; (B) m-CPBA (4.5 eq), DCM, 0 C to rt, 2 h; (C)
ꢃ
ꢃ
Fe/AcOH, MeOH, 60 C, 5 h; (D) Zn, MeOH, rt, 1.5 h; (E) Pd(OAc)
2
(0.10 eq), BINAP (0.06 eq), Cs
2
CO
3
(2.0 eq), 1,4-dioxane, 105 C, 7 h.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
417
a
a
2
Scheme 3. Synthesis of derivatives X1–X23 . Reagents and conditions: (A) Hydrazine hydrate (6.0 eq), MeOH, rt, 3 h; (B) CS (20.0 eq), dicyclohexylcarbodiimide (DCC,
ꢃ
ꢃ
ꢃ
1
.1 eq), THF, rt, overnight; (C) m-CPBA (1.5 eq) , DCM, ꢂ10 C to 0 C, 2 h; (D) m-CPBA (4.5 eq), DCM, 0 C to rt, 2 h.
phthalimide 1 with various dibromoalkane in acetone yielded the decrease enzymatic inhibitory activity on different levels, espe-
mono-bromides 3a–3e bearing the phthalimide tail meeting the cially for CDK4/6. We presumed whether the joint position L
different length for SAR. Then, 3a–3e were refluxed with potas- between the isothiocyanato tail and phenyl ring affected the
sium thioacetate (KSAc) in THF to afford thioacetates 4a–4e, fol- activity. Thus, X4 as a candidate for its poor potency on activity
lowing acidic hydrolysis reaction in methanol to give the thiols was selected to be oxidised to X5 and X6 structurally character-
5a–5e. In Scheme 2, prepared 5a–5e were arylated by p-fluoroni- ised as sulfoxide and sulphone, all of which were tested together
trobenzene with various R substituents (6a–6d) to give the to confirm the importance of the joint position L as shown in
intermediates 7a–7h. On the other hand, 5a–5e reacted with Table 1. As expected, X5–X6 displayed more notable increasing
1
-(bromomethyl)-4-nitrobenzene (11) in DMF under the alkaline inhibitory activity than X4 against all of three kinases. Oxidation
condition of K CO to obtain the 12a–12e. Especially emphasised state of sulphur at the joint position L might be superior than sul-
2
3
in dotted line cage, thioether intermediates 7a–7e and 12a–12e, phur, that was also be confirmed in the example of X1 versus X7
were oxidised absolutely to corresponding sulphone 8a–8e and (Table 1). Further, the derivatives bearing oxidation state of sul-
1
1
3a–13e. Subsequently, nitro group in intermediates 7a–8e and phur were tested the IC50 value against CDK4 and CDK9 due to
2a–13e need to be reduced to primary amine using either iron the high inhibition at 1 micromolar concentration (Table 2).
powder activated by acetic acid or zinc powder (9a–9e, 10a–10e, Compared with ribociclib, X7 exhibited excellent potency against
4a–14e, 15a–15e). After that, all the primary amine intermedi- CDK4/9 (IC : 7 and 26 nM, respectively). At the same time, a
1
5
0
ates described above underwent Buchwald–Hartwig coupling reac- derivative (X8) with different space between isothiocyanato and
tion with commercially available 16 by catalytic amount of phenyl ring was synthesised to determine the changing the activ-
palladium acetate to create the key intermediates 17a–18e and ity produced by the length of the tail. The results indicated that
19a–20e. As shown in Scheme 3, the derivatives except X5–X7 the length of the tail could cause changing on the enzymatic
were produced via two successive steps without intermediate inhibitory activity, but still need to be explored the rules of the
purification. 17a–20e first were refluxed with 80% hydrazine changing in detail.
hydrate in methanol to deprotect, following reacting with carbon
Based on preliminary research, we then adopted the H at R
0
disulphide and N,N -dicylohexylcarbodiimide to afford the isothio- position to design several derivates (X9–X13, Table 1) to focus on
cyanates derivatives X1–X4 and X8–X23. Finally, sulphone deriva- the effect of the sulphur and its oxidation state at L position and
tive X6 was obtained by oxidising X4 completely with an excess length of the tail on the enzyme test. Obviously, X9 with sul-
of m-CPBA. Sulfoxide derivatives X5 and X7 were obtained phone enhanced the inhibition against the three kinases, com-
through incomplete oxidation from X4 and X1, respectively, which pared with X8. In the example of compound X9–X13, the
were accomplished through controlling strictly the amount of 1.1 inhibition against CDK9 was keeping at a relative high level, but
equiv m-CPBA.
The results of the enzymatic-inhibition assays are listed in space between the isothiocyanato and sulphone structurally.
Tables 1 and 2 with ribociclib as the positive control. We expected In view of the importance of the L position and the length of
the efficacy activity on CDK4/6 was declined with the increasing
to explore an optimisation strategy for improving the inhibitory the tail, we attempt to insert a methylene between the L position
activity against cell cycle as targeting on CDK4/6 and simultan- and the phenyl leading to derivatives X14–X23 (Table 1).
eously targeting CDK9 for cell transcription. The structure–activity Excitingly, X19–X23 were proved high inhibition on all three kin-
relationships were discussed based on the optimisation at the R, L ases as well as no obvious difference for the enzymatic activity
substituents and the carbon tail length (shown as Formula I and II but X23. We summarised the structure–activity relationships as
in Table 1).
Firstly, we explored the R position with various substituents,
such as H, CH , OCH , CF
that H and CH could only maintain the potent against CDK9 potency for cancer treatment, we screened compounds exhibiting
compared with ribociclib, while the OCH and CF groups were better kinase inhibitory to do the cytotoxicity assays. The selected
shown in Supplementary Figure S2.
Since our goal here is to discover CDKs inhibitors with cell
3
3
3
(X1–X4, Table 1). The results showed cycle arrest and transcription blockade activities having high
3
3
3
not tolerable at all because of the cliff loss potency on all the compounds, including X7, X19, X20 and X22 were carried out
three kinases. In total, all the optimisation on R position would using standard CCK8 assay and the results are shown in Table 3.

4
18
X. WANG ET AL.
a
Table 1. Structure and enzymatic inhibition activity evaluation of compounds X1–X23.
Structure
R
Inhibition (%) at 1 lM
Compds.
Formula
L
n
CDK4
CDK6
CDK9
Ribociclib
X1
X2
X3
X4
X5
X6
X7
X8
–
I
I
I
I
I
I
I
I
I
I
I
I
–
H
CH
OCH
–
S
S
S
S
–
3
3
3
3
3
3
3
1
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
101 ± 0
35 ± 0.5
29 ± 2.1
8 ± 2.1
85 ± 1.2
63 ± 1.2
17 ± 1.2
5 ± 0.8
84 ± 0
76 ± 0
80 ± 0
3
3
33 ± 2.5
41 ± 0.8
97 ± 0.4
75 ± 0.4
99 ± 0
97 ± 0.4
97 ± 0
98 ± 0.4
98 ± 0.4
97 ± 0.4
96 ± 0.4
97 ± 0.4
96 ± 0
94 ± 0.4
94 ± 1.6
95 ± 0.4
98 ± 0.4
98 ± 0.4
98 ± 0
CF
CF
CF
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
3
3
3
9 ± 1.6
8 ± 3.3
SO
91 ± 0.8
35 ± 2.4
99 ± 0
70 ± 2.4
94 ± 1.6
90 ± 0.8
89 ± 0.4
83 ± 0
48 ± 1.2
14 ± 1.2
87 ± 1.6
64 ± 0.8
92 ± 1.2
88 ± 0.8
82 ± 0.8
75 ± 0.4
68 ± 0.4
86 ± 1.2
76 ± 0.8
73 ± 0
52 ± 0.4
52 ± 0.8
94 ± 0.8
94 ± 0.4
93 ± 0
SO
SO
S
SO
SO
SO
SO
SO
S
S
S
S
S
2
X9
2
X10
X11
X12
X13
X14
X15
X16
X17
X18
X19
X20
X21
X22
X23
a
2
2
2
2
I
74 ± 0
95 ± 0
II
II
II
II
II
II
II
II
II
II
83 ± 0.8
71 ± 0.8
61 ± 2.1
82 ± 1.2
98 ± 0
99 ± 0.4
98 ± 0.4
97 ± 0.4
95 ± 0.5
SO
SO
SO
SO
SO
2
2
2
2
2
91 ± 0
77 ± 1.6
98 ± 0
97 ± 0
Inhibition activities were determined using the KinaseProfiler of Eurofins. The data represent the mean values of two independ-
ent experiments.
a
Table 2. Enzymatic inhibition activity evaluation of selected compounds.
IC50 (nM)
IC50 (nM)
CDK6
Compds.
CDK4
CDK6
CDK9
Compds.
CDK4
CDK9
Ribociclib
X5
X7
X8
X9
13 ± 1.3
331 ± 33.3
26 ± 1.1
–
43 ± 2.6
90 ± 10.5
–
71 ± 14.1
197 ± 14.8
34 ± 5.3
7 ± 0.4
16
4 ± 0.4
7 ± 0.8
8 ± 0.8
13 ± 1.8
24 ± 3.1
16 ± 2.8
X15
X16
X17
X18
X19
X20
X21
X22
X23
–
–
–
–
–
–
29 ± 1.4
34 ± 1.4
55 ± 2.2
42 ± 1.6
6 ± 0.4
7 ± 1.0
7 ± 0.6
–
–
–
–
–
101 ± 4.7
17 ± 0.7
21 ± 1.4
24 ± 0.9
30 ± 3.9
75 ± 10.0
55 ± 1.6
42 ± 5.4
78 ± 13.4
126 ± 20.6
–
X10
X11
X12
X13
X14
a
–
–
–
–
–
–
–
10 ± 1.3
14 ± 0.9
79 ± 1.3
IC50 values were determined using the KinaseProfiler of Eurofins. The data represent the mean values of two independent experiments.
a
Table 3. In vitro cell growth inhibition (IC50) of selected compounds against multiple cancer cell lines.
IC50 (lM)
Cell lines
4T1
T-47D
MCF7
A549
H460
H1299
Ribociclib
X8
X12
X7
>10
6.23 ± 4.04
0.41 ± 0.10
0.16 ± 0.02
0.10 ± 0.02
0.72 ± 0.03
0.12 ± 0.01
0.11 ± 0.01
>10
7.46 ± 3.06
3.07 ± 0.11
1.57 ± 0.07
0.67 ± 0.03
1.46 ± 0.28
0.38 ± 0.02
0.21 ± 0.07
>10
5.46 ± 2.65
3.79 ± 0.07
1.31 ± 0.06
1.15 ± 0.10
0.60 ± 0.04
0.60 ± 0.24
0.57 ± 0.07
1.46 ± 0.34
0.46 ± 0.07
0.41 ± 0.10
0.69 ± 0.14
0.26 ± 0.19
0.083 ± 0.34
1.33 ± 0.15
0.57 ± 0.24
0.23 ± 0.04
0.16 ± 0.06
0.12 ± 0.07
0.11 ± 0.07
3.22 ± 0.22
2.79 ± 0.32
0.92 ± 0.15
1.45 ± 0.10
1.50 ± 0.39
1.80 ± 0.34
X19
X20
X22
a
The IC50 values are shown in the forms. The cytotoxic effect of compounds was assayed using CCK-8 assay with 72 h incubation. Data
are from three independent experiments.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
419
X8 and X12 with selective CDK9 inhibition and ribociclib as CDK4
inhibitor were assess the cancer cell inhibition assays as control.
IC50 values were detected on compounds against breast cancer
cell lines (4T1, T47D and MCF7) and lung cancer cell lines (A549,
H1299 and H460). The best tumour cell potency of these ana-
logues was obtained for molecule X22, which was also rational
inhibitor of CDK9 with IC50 value of 10 nM and CDK4 with IC50
value of 30 nM. The CDK4 inhibitor ribociclib alone was not potent
in these solid tumour cells. X22 displayed higher inhibition
potency than selective CDK4 inhibitor ribociclib and selective
CDK9 inhibitor X8 and X12. This result demonstrated that dual-
target compound had more efficient inhibition at the cellular
level. As the results of enzymatic and antitumor activity in cell
lines showed that X22 is the best in this class, we performed fur-
ther in-depth in vitro and in vivo biological studies using this
compound.
Molecular docking studies were carried out to investigate the
binding modes of compound X22. X22 is a potent CDK4 and
CDK9 inhibitor with IC50 value of 30 and 10 nM, respectively, and
also has weak inhibition activity against CDK6 with IC50 value of
1
26 nM. As shown in Figure 1(A), X22 binds to the ATP-binding
sites of CDK6 with two key hydrogen bonds formed by aminopyri-
midine and the backbone VAL101 residue, which retained the
binding mode of the scaffold of ribociclib and CDK6 as reported.
Given no 3D CDK4-ligand was reported, the structure of CDK6-lig-
and complex (PDB code: 4EZ5) was used as a template for hom-
ology modelling to generate the 3D structure of CDK4-ligand
(Figure 1(B)). Importantly, two additional hydrogen bonds were
formed because of the hydrophobic side chain containing isothio-
cyanate. The sulphur atom formed two hydrogen bonds with the
backbone of ALA16 and TYR17 at the front of the ATP binding
site of CDK4, respectively, which might explain the better inhibi-
tory activity of the compound X22 against CDK4 (IC50 ¼ 30 nM)
than CDK6 (IC ¼126 nM). In Figure 1(C), the 2-aminopyrimidine
5
0
formed two hydrogen bonds with the CYS106 residue of X22 in
CDK9 (VAL101 in CDK6, VAL in CDK4). As well, the nitrogen atom
formed a key hydrogen-bond with the residue of the LYS151 at
the front of the ATP binding pocket. At the back of the ATP bind-
ing site, the pyrrolo-[2,3-d]-pyrimidine scaffold exploited the
hydrophobic region close to the gatekeeper residue to form a
favourable p–p interaction in all three kinases region.
Interestingly, a hydrogen-bond was formed between the oxy-
gen of the sulphonyl of X22 with the backbone of the ASP109 in
another docking configuration (Figure 2(A)). At the front of the
ATP binding pocket in CDK9, GLY112 possessing the smallest resi-
due makes this hydrophobic site accommodate flexible side chain
Figure 1. Representation of the predicted binding modes of compound X22 with
like compound X22, while THR107 at the same position in CDK6 CDKs kinase domain. X22 is shown in green. Hydrogen bonds are shown in red.
(
A) Proposed binding mode of compound X22 with CDK6 (PDB code: 4EZ5),
shows a little crowded for flexible group (Figure 2(B)). The much
bigger hydrophobic site permit the compound X22 exits in mul-
tiple configurations docking CDK9 forming hydrogen bond for
CDK6 backbone is shown in light pink. (B) Proposed binding mode of compound
X22 with CDK4, which employed CDK6 (PDB entry: 4EZ5) as the template for
homology modelling, CDK4 backbone is shown in yellow. (C) Proposed binding
binding tightly. The difference might offer an explanation the mode of compound X22 with CDK9 (PDB code: 4BCF), CDK9 backbone is shown
in light blue.
stronger inhibitory potency compound X22 against CDK9 (IC ¼
5
0
1
0 nM). Thus, X22 as an ATP-competitive inhibitor potently binds
in the ATP pockets of CDKs kinases.
ribociclib and DMSO for 12, 36 and 48 h (Supplementary Figure
Given the strong inhibitory ability of X22 in MCF7 and A549 S3). These results confirmed that X22 induced the cell cycle arrest,
3
9
cells, we examined the effect of X22 on the cell cycle. Cells were which are consistent with our observations above . Compared
treated with compound X22, ribociclib and DMSO for 24 h. with CDK4 depletion alone, combined depletion CDK4 and CDK9
Representative flow cytometry patterns are shown in Figure 3. The induced substantial G /M arrest. These data indicated that X22
2
2
2
results showed that X22 significantly blocked the cell cycle at G /M could induce G /M phase arrest of the cell cycle in both breast
phase in a dose-dependent manner accompanied by decreases in and lung cancer cells.
4
0,41
S phase in both cell lines, as compared to the cells incubated with
DMSO. Ribociclib mainly blocked the cell cycle at G1 phase form complexes. Activations of these complexes are responsible
Figure 3). We also detected cell kinetics treated with X22, for the phosphorylation of retinoblastoma protein (Rb), which
As reported
, CDK4/6 preferentially bind D-type cyclins to
(

4
20
X. WANG ET AL.
Figure 2. (A) Representation of the predicted binding mode of compound X22 with CDK9 kinase domain (PDB code: 4BCF). CDK9 backbone is shown in grey. X22 is
shown in green. Hydrogen bonds are shown in red. (B) Representation of the hydrophobic pocket difference at the front of the ATP-binding site by overlaying CDK6
and CDK9. CDK6 is shown in grey. CDK9 is shown in orange.
2
Figure 3. X22 induced G /M phase arrest in breast and lung cancer cells. Cells were harvested after treatment with different concentrations of X22 or ribociclib for
2
4h. Cells were fixed by 70% ethanol and stained with propidium iodide. Data are represented as histogram on the left and the percentage of cell cycle distribution
are shown on the right. The assays were performed in triplicate.
allows cell cycle to proceed from G to S phase and further results
We next analysed whether X22 treatment promoted cell apop-
1
in cell proliferation. In contrast, CDK9 is not a typical Cdc-2-like tosis by Annexin V-FITC/PI staining. After 48 h incubation, X22
kinase and does not participate in cell cycle regulation. It can dose-dependently increased the proportion of apoptotic cells (Q2
form the P-TEFb complex, which is capable of phosphorylating late apoptotic and Q3 early apoptotic) in A549 cells, compared to
the CTD of the largest subunit of RNAPII and regulate the RNA the cells treated with DMSO or ribociclib. Similar results were also
transcription elongation.
observed in MCF7 cells. X22 at 0.1 to 0.5 mM increased the percen-
To further verify the inhibitory effect of compound X22 on tages of apoptotic cells by approximately 2.2 to 3.5-fold change,
CDKs, western blotting analysis was performed. We found that compared to DMSO-treated cells (Figure 5(A)). To further confirm
X22 treatment significantly suppressed the phosphorylation of Rb the apoptosis induction of compound X22, we examined the
at CDK4/6 specific site Ser 780 and CDK9 specific sites Ser 807/ expression of relevant proteins by Western blot. Cells were treated
8
11 in MCF7 and A549 cells respectively, which confirmed that with or without compound X22 and control for 48 h and then
X22 targeted on CDKs (Figure 4). We also detected the level of lysed and analysed. Given the crucial role of p53 in regulating cell
phosphorylated Ser2 (Ser2-P), a well-established cellular target of apoptosis, we detected whether compound X22 induced apop-
CDK9 on the RNAPII CTD during transcription elongation. tosis depending on the expression of p53 and its downstream tar-
Compared to cells treated with DMSO, X22 effectively inhibited gets. As illustrated in Figure 5(B), X22 treatment remarkably
the phosphorylation of Ser2 and Ser5 at 0.5 mM in MCF7 cells and increased the expressions of p53 and promoted the phosphoryl-
1
mM in A549 cells (Figure 4). Taken together, the results above ation of p53 at Ser15. We also demonstrated that X22-induced
revealed that X22 could specifically target CDKs, thus further reg- apoptosis in solid tumour cells, which reduced expression of the
ulating their downstream signalling proteins in cancer cells. anti-apoptotic protein Bcl-2 and enhanced expression of Bax.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
421
Figure 4. X22 suppressed the downstream signalling proteins of CDK4/9 in breast and lung cancer cells. Cells were incubated with the indicated concentrations of
X22 or ribociclib for 24 h. Proteins were extracted and analysed by western blotting.
Figure 5. X22 induced cell apoptosis in breast and lung cancer cells. (A) Cells were seeded in six-well plates and treated with the indicated concentrations of X22 or
ribociclib for 48 h. Cells were then stained with the AnnexinV-FITC Apoptosis Detection Kit, followed by flow cytometry analysis. Quantitative data are expressed as
mean ± SD of the percentages of apoptotic cells from three independent experiments. (B) Expressions of p53, phospho-p53 (Ser15), Bax, Bcl-2 in MCF-7 and A549 cells
ꢀ
ꢀꢀ
were detected by western blotting after the treatment with the indicated concentrations of ribociclib or X22 for 48 h. p < 0.05 versus DMSO; p < 0.01 versus
ꢀꢀꢀ
DMSO;
p < 0.001 versus DMSO.

4
22
X. WANG ET AL.
Figure 6. X22 significantly inhibited growth of breast cancer in homograft mouse models. (A) Tumour growth curve of 4T1 tumour-bearing mice in response to differ-
ent treatments. (B) Photographs of 4T1 homograft tumours. Mice were sacrificed 18days after tumour implantation. (C) The average tumour weight of excised tumours
at day 18. (D) Body weight curves of BALB/c mice in each group after injection.
These findings indicated that compound X22-induced apoptosis is inhibited CDK4 and CDK9 with IC50 values of 30 and 10 nM,
associated with upregulating pro-apoptotic protein p53 and Bax respectively, which showed marked antitumor activity. In in vitro
and downregulating anti-apoptotic protein Bcl-2 expression. cellar assays, X22 was capable of blocking the cell cycle at G /M
2
Collectively, X22 may induce cell apoptosis through p53/Bax/Bcl-2 phase resulting in decreased S-phase populations. Compared with
pathway in breast and lung cancer cells.
CDK4 depletion alone, combined depletion induced apoptosis in
To better assess whether X22 could efficiently suppress tumour cancer cells in a dose-dependent manner. Western blot assays
growth in vivo, homograft tumour models were established using confirmed substantial apoptosis after X22-induced CDK4 and
4
T1 cells. Oral treatment with X22 significantly suppressed the CDK9 depletion. Meanwhile, X22 reduced RNA polymerase II
tumour growth (Figure 6(A)). Representative photographs of expression and CTD phosphorylation. In in vivo assay, oral admin-
excised tumours at day 18 were shown in Figure 6(B). X22 treat- istration of compound X22 once-daily at 15, 30 or 60 mg/kg for
ment with 15 or 30 or 60 mg/kg substantially inhibited the tumour 18 days led to tumour regression without obvious toxicity. Thus,
growth with tumour inhibition rates of 51.8 ± 17.5%, 65.4 ± 11.7% reduced CDK4 activity in concert with depleted CDK9 activity may
and 70.3 ± 4.4%, respectively. In contrast, tumour volume in riboci- enhance the antiproliferative effects. This study suggest combined
clib-treated group decreased slightly compared with vehicle depletion of cell cycle and transcriptional CDK activities may be
group. Tumour growth inhibitions of 17.2 ± 5.2% were observed at
doses of 60 mg/kg of ribociclib. The potencies of ribociclib are
relatively weaker than those of X22. The average tumour weight
worthy of clinical further development for solid tumour therapy.
of excised tumours in X22 groups was obviously lighter than Disclosure statement
those in other groups (Figure 6(C)). A preliminary toxicity evalu-
The authors report no declarations of interest.
ation for X22 was also carried out. Mouse weights were moni-
tored twice per week over 18 days. There was no significant
difference among the five groups of mice (Figure 6(D)). No obvi- Funding
ous signs of toxicity were observed in the X22-treated groups,
This work was supported by the National Key Research and
Development Programme of China (2018YFE0114300) and the
National Natural Science Foundation of China (81972882 and
indicating that compound X22 is well tolerated in vivo.
8
1903099).
4
. Conclusions
To date, selective CDK inhibitors have produced only modest
activity against solid tumours. In this article, we discovered and
studied the medicinal chemistry of multi-CDKs inhibitors to assess
their relative activities against the cell cycle and transcriptional
CDKs. Several compounds inhibited CDK4 and CDK9 at low nano-
molar levels and exhibited good antiproliferative activities in a
panel of tumour cells. The most potent analogue X22 potently
References
1. Ciceri P, Muller S, O’Mahony A, et al. Dual kinase-bromodo-
main inhibitors for rationally designed polypharmacology.
Nat Chem Biol 2014;10:305–12.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
423
2.
3.
4.
5.
Hanahan D, Weinberg RA. Hallmarks of cancer: the next 24. DiPippo AJ, Patel NK, Barnett CM. Cyclin-dependent kinase
generation. Cell 2011;144:646–74.
Johnson LN, Lewis RJ. Structural basis for control by phos-
phorylation. Chem Rev 2001;101:2209–42.
Adams JA. Kinetic and catalytic mechanisms of protein kin-
ases. Chem Rev 2001;101:2271–90.
Huang M, Shen AJ, Ding J, Geng MY. Molecularly targeted
inhibitors for the treatment of breast cancer: past, present,
and future. Pharmacotherapy 2016;36:652–67.
25. Beaver JA, Amiri-Kordestani L, Charlab R, et al. FDA approval:
palbociclib for the treatment of postmenopausal patients
with estrogen receptor-positive, HER2-negative metastatic
breast cancer. Clin Cancer Res 2015;21:4760–6.
cancer therapy: some lessons from the past decade. Trends 26. Shah A, Bloomquist E, Tang S, et al. FDA approval: ribociclib
Pharmacol Sci 2014;35:41–50.
Ma WW, Adjei AA. Novel agents on the horizon for cancer
therapy. Ca Cancer J Clin 2009;59:111–37.
Sun C, Bernards R. Feedback and redundancy in receptor
tyrosine kinase signaling: relevance to cancer therapies.
Trends Biochem Sci 2014;39:465–74.
Wu P, Nielsen TE, Clausen MH. FDA-approved small-mol-
ecule kinase inhibitors. Trends Pharmacol Sci 2015;36:
for the treatment of postmenopausal women with hormone
receptor-positive, HER2-negative advanced or metastatic
breast cancer. Clin Cancer Res 2018;24:2999–3004.
6
.
.
2
2
7. FDA OKs abemaciclib for ERþ, HER2-breast cancer. Cancer
7
Discov 2017;7:OF1.
8. Im SA, Mukai H, Park IH, et al. Palbociclib plus letrozole as
first-line therapy in postmenopausal Asian women with
metastatic breast cancer: results from the phase III, random-
ized PALOMA-2 study. J Glob Oncol 2019;5:1–19.
9. Meattini I, Desideri I, Scotti V, et al. Ribociclib plus letrozole
and concomitant palliative radiotherapy for metastatic
breast cancer. Breast 2018;42:1–2.
0. Single-agent abemaciclib active in breast cancer. Cancer
Discov 2016;6:809–10.
1. Hafner M, Mills CE, Subramanian K, et al. Multiomics profil-
ing establishes the polypharmacology of FDA-approved
CDK4/6 inhibitors and the potential for differential clinical
activity. Cell Chem Biol 2019;26:1067–80.
8
.
.
422–39.
2
9
Baker SJ, Reddy EP. Targeted inhibition of kinases in cancer
therapy. Mt Sinai J Med 2010;77:573–86.
1
1
1
1
1
1
0. Krug M, Hilgeroth A. Recent advances in the development
of multi-kinase inhibitors. Mini Rev Med Chem 2008;8:
3
3
1312–27.
1. Asghar U, Witkiewicz AK, Turner NC, et al. The history and
future of targeting cyclin-dependent kinases in cancer ther-
apy. Nat Rev Drug Discov 2015;14:130–46.
2. Peyressatre M, Prevel C, Pellerano M, Morris MC. Targeting
cyclin-dependent kinases in human cancers: from small mol-
ecules to peptide inhibitors. Cancers 2015;7:179–237.
3
3
3
2. Lukasik PM, Elabar S, Lam F, et al. Synthesis and biological
evaluation of imidazo[4,5-b]pyridine and 4-heteroaryl-pyrimi-
dine derivatives as anti-cancer agents. Eur J Med Chem
3. Deshpande A, Sicinski P, Hinds PW. Cyclins and cdks in
development and cancer: a perspective. Oncogene 2005;24:
2
012;57:311–22.
3. Singh U, Chashoo G, Khan SU, et al. Design of novel 3-pyri-
midinylazaindole CDK2/9 inhibitors with potent in vitro and
in vivo antitumor efficacy in a triple-negative breast cancer
model. J Med Chem 2017;60:9470–89.
4. Cai D, Latham VM, Jr., Zhang X, Shapiro GI. Combined
depletion of cell cycle and transcriptional cyclin-dependent
kinase activities induces apoptosis in cancer cells. Cancer
Res 2006;66:9270–80.
2909–15.
4. Kalra S, Joshi G, Munshi A, Kumar R. Structural insights of
cyclin dependent kinases: Implications in design of selective
inhibitors. Eur J Med Chem 2017;142:424–58.
5. Harbour JW, Luo RX, Dei Santi A, et al. Cdk phosphorylation
triggers sequential intramolecular interactions that progres-
sively block Rb functions as cells move through G1. Cell
1999;98:859–69.
3
3
3
3
5. Surur AS, Schulig L, Link A. Interconnection of sulfides and
sulfoxides in medicinal chemistry. Archiv Der Pharmazie
1
6. Romano G. Deregulations in the cyclin-dependent kinase-9-
related pathway in cancer: implications for drug discovery
and development. ISRN Oncol 2013;2013:1–14.
7. Morales F, Giordano A. Overview of CDK9 as a target in can-
cer research. Cell Cycle 2016;15:519–27.
8. Wang S, Fischer PM. Cyclin-dependent kinase 9: a key tran-
scriptional regulator and potential drug target in oncology,
virology and cardiology. Trends Pharmacol Sci 2008;29:
2
019;352:e1800248.
6. Gupta P, Kim B, Kim SH, Srivastava SK. Molecular targets of
isothiocyanates in cancer: recent advances. Mol Nutr Food
Res 2014;58:1685–707.
7. Guo Q, Yu C, Zhang C, et al. Highly selective, potent, and
oral mTOR inhibitor for treatment of cancer as autophagy
inducer. J Med Chem 2018;61:881–904.
1
1
302–13.
8. Li Y, Luo X, Guo Q, et al. Discovery of N1-(4-((7-cyclopentyl-
1
2
2
9. Bose P, Simmons GL, Grant S. Cyclin-dependent kinase
inhibitor therapy for hematologic malignancies. Expert Opin
Investig Drugs 2013;22:723–38.
0. Vijayaraghavan S, Moulder S, Keyomarsi K, et al. Inhibiting
CDK in cancer therapy: current evidence and future direc-
tions. Targeted Oncol 2018;13:21–38.
1. Sarosiek T. Inhibitors of cyclin-dependent kinases (CDK) – a
new group of medicines in therapy of advanced breast can-
cer. Polski Merkuriusz Lekarski: organ Polskiego Towarzystwa
Lekarskiego 2018;44:5–9.
2. Heptinstall AB, Adiyasa I, Cano C, Hardcastle IR. Recent
advances in CDK inhibitors for cancer therapy. Future Med
Chem 2018;10:1369–88.
6
-(dimethylcarbamoyl)-7 H-pyrrolo[2,3- d]pyrimidin-2-yl)ami-
no)phenyl)-N8-hydroxyoctanediamide as a novel inhibitor
targeting cyclin-dependent kinase 4/9 (CDK4/9) and histone
deacetlyase1 (HDAC1) against malignant cancer. J Med
Chem 2018;61:3166–92.
3
9. Chiu HC, Huang WR, Liao TL, et al. Mechanistic insights into
avian reovirus p17-modulated suppression of cell cycle CDK-
cyclin complexes and enhancement of p53 and cyclin H
interaction. J Biol Chem 2018;293:12542–62.
40. Thangavel C, Boopathi E, Liu Y, et al. Therapeutic challenge
with a CDK 4/6 inhibitor induces an RB-dependent SMAC-
mediated apoptotic response in non-small cell lung cancer.
Clin Cancer Res 2018;24:1402–14.
2
2
3. Shapiro GI. Cyclin-dependent kinase pathways as targets for 41. Aleem E, Arceci RJ. Targeting cell cycle regulators in hema-
cancer treatment. J Clin Oncol 2006;24:1770–83. tologic malignancies. Front Cell Dev Biol 2015;3:16.