Discovery of novel 9H-purin derivatives as dual inhibitors of HDAC1 and CDK2

Article abstract of DOI:10.1016/j.bmcl.2019.06.059

HDAC and CDK inhibitors have been demonstrated to be synergistically in suppressing cancer cell proliferation and inducing apoptosis. In this work, we incorporated the pharmacophore groups of HDACs and CDKs inhibitors into one molecule to design and synthesize a series of purin derivatives as HDAC/CDK dual inhibitors. The lead compound 6d, showing good HDAC1 and CDK2 inhibitory activity with IC₅₀ values of 5.8 and 56 nM, respectively, exhibited attractive potency against several cancer cell lines in vitro. This work may lead to the discovery of a novel scaffold and potential dual HDAC/CDK inhibitors.

Full text of DOI:10.1016/j.bmcl.2019.06.059

Accepted Manuscript
Discovery of novel 9H-purin derivatives as dual inhibitors of HDAC1 and
CDK2
Yu Yu, Dongzhi Ran, Junhao Jiang, Tao Pan, Yanrong Dan, Qiang Tang, Wei
Li, Lin Zhang, LinLing Gan, Zongjie Gan
PII:
S0960-894X(19)30441-X
https://doi.org/10.1016/j.bmcl.2019.06.059
BMCL 26537
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
23 May 2019
22 June 2019
28 June 2019
Please cite this article as: Yu, Y., Ran, D., Jiang, J., Pan, T., Dan, Y., Tang, Q., Li, W., Zhang, L., Gan, L., Gan, Z.,
Discovery of novel 9H-purin derivatives as dual inhibitors of HDAC1 and CDK2, Bioorganic & Medicinal
Chemistry Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.06.059
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

Discovery of novel 9H-purin derivatives as dual
inhibitors of HDAC1 and CDK2
Yu Yua, b, Dongzhi Rana, b, Junhao Jianga, b, Tao Pana,, Yanrong Dana, Qiang Tanga, b,
Wei Lia, b, Lin Zhanga,, LinLing Gana,, Zongjie Gana, b
*
a Department of Medicinal Chemistry, College of Pharmacy, Chongqing Medical University,
Chongqing 400016, PR China
b Chongqing Research Center for Pharmaceutical Engineering, Chongqing Medical University,
Chongqing 400016, PR China
*Corresponding author: ZongJie Gan
Email: gzj@cqmu.edu.cn
Tel.:86-23-68485161

Abstract HDAC and CDK inhibitors have been demonstrated to be synergistically in suppressing
cancer cell proliferation and inducing apoptosis. In this work, we incorporated the pharmacophore
groups of HDACs and CDKs inhibitors into one molecule to design and synthesize a series of purin
derivatives as HDAC/CDK dual inhibitors. The lead compound 6d, showing good HDAC1 and CDK2
inhibitory activity with IC50 values of 5.8 and 56 nM, respectively, exhibited attractive potency against
several cancer cell lines in vitro. This work may lead to the discovery of a novel scaffold
and potential dual HDAC/CDK inhibitors.
Keywords HDACs; CDKs; Dual inhibitors; Purin; Antitumor
Histone deacetylases (HDACs) are a family of enzymes which could catalyze the removal of acetyl
groups from the ε-amino groups of lysine residues presenting within the N-terminal extension of the
core histones.1 Such modifications are crucial for the regulation of gene expression and cellular
processes, including apoptosis, cell growth and differentiation.2,3 However, aberrant histone
deacetylation, owing to the increased activity and expression of HDACs activity and expression, has
been demonstrated in several kinds of cancer cell lines and correlated with aggressiveness, migration
and invasion of cancer.4,5 Thus, inhibition of HDACs has proven to be a valuable epigenetic strategy
for cancer treatment. Over the past decade, there have been intense efforts to identify novel small-
molecule HDACs inhibitors to address unmet medical needs. Up to now, there are already approved
HDACs inhibitors such as SAHA and panobinostat for the treatment of lymphoma and multiple
myeloma (Fig. 1).6,7 Despite the efficacy in hematological malignancies, HDACs inhibitors as
monotherapy in solid tumors have been proved less effective than hematological malignancies in vivo
and thus combination of HDACs inhibitors with other anticancer agents may result in the improved
anti-proliferation of tumor cells, which represents a new therapeutic approach to the malignancies
treatment. 8-10
O
H
N
O
OH
OH
N
N
H
H
O
O
SAHA
HN
N
H
O
N
NH2
H
Panobinostat
H
N
N
O
Entinostat
Figure 1. Structures of the approved HDACs inhibitors.
Cyclin-dependent kinases (CDKs) are key regulators of cell division cycle, since CDKs are often
deregulated in cancer, thus, CDKs inhibitors which target CDKs are considered to be a promising agent
for cancer treatment.11 Recently, studies have indicated that the inhibition of CDKs sensitizes cancer

cells to HDACs inhibitors which could induce apoptosis, thus combination of HDACs and CDKs
inhibitors may have statistically synergy in suppressing cancer cell proliferation. For instance, Gahr
and coworkers12 have reported that the combination of HDACs inhibitor entinostat with CDKs
inhibitor roscovitine could inhibit the cell cycle progression and increase the suppression of hepatoma
cell DNA synthesis. Furthermore, the synergistic combination between the inhibitors of CDKs and
HDACs was proved to overcome CDKs inhibitor resistance by activating p21 in ER-positive breast
cancer cells, which might be beneficial in the resistant cases of CDKs inhibitors.13 Therefore, following
the success of synergy between HDACs and CDKs inhibitors, we conceived that the incorporation of
the pharmacophore groups of HDACs and CDKs inhibitors into one molecule for the concurrent
inhibition of HDACs and CDKs would be an effective way to optimize HDACs and CDKs inhibitors’
efficacy in solid tumors and avoid unwanted side effects associated with the combination therapy, such
as drug-drug interaction, different physicochemical properties, or poor patient compliance. Lately,
Xiang et al14 have reported a series of 7H-pyrrolo[2,3-d]pyrimidine derivatives bifunctional targeting
on CDK4/9 and HDAC1, which demonstrated excellent antitumor efficacy against 4T1 tumor
xenograft models in vivo, suggesting the development of CDK/HDAC dual inhibitors could be of great
significance for malignant cancer therapy.
According to the crystal structure of CDK2 complexed with its selective analog roscovitine, 15 it is
found that the purin group of roscovitine occupies the ATP binding pocket, functioning as the mainly
binding moiety with ATP. On the other hand, it is well known that the pharmacophore of HDACs
inhibitors comprises three parts, including a zinc-binding group (ZBG), a linker and a surface
recognition “CAP” group.16 Generally, the CAP region is predominantly responsible for selectivity and
tolerates a large change. Therefore, we have chosen purin ring instead of indole ring as the “CAP”
group, utilized ethenylphenyl as the linker to link the ZBG and CAP groups with the aim to discover
novel HDAC/CDK dual inhibitors (Figure 2). With this approach,
a
novel scaffold
and potential HDAC/CDK dual inhibitor has been designed and presented in this work.
CAP
ZBG
Linker
O
OH
N
H
H
N
N
H
O
CH3
HN
OH
HDAC inhibitor panobinostat
HN
N
N
N
N
N
het
R
HN
H3C
HO
N
N
N
Dual inhibitors
N
N
H
CH3
H3C
CDK inhibitor roscovitine
Figure 2. Design strategy of HDAC/CDK dual inhibitors.

The preparation of target compounds (6) was described in Scheme 1. The target compounds were
synthesized from commercially available starting material 2, 6-dichloro-9H-purin (1) via a novel and
convenient route. Reaction of (1) with corresponding iodoalkanes in DMF afforded intermediate (2),
which was then condensed with 3- or 4-bromobenzyl amine under mild conditions providing
intermediate (3). The intermediates (4), synthesized by the substitution of (3) with heterocyclic amines
such as morpholine or piperidine, underwent standard Heck coupling with methyl acrylate in the
presence of Pd(OAC)2 to give the key intermediate (5). Finally, the ester groups of (5) reacted with
freshly prepared hydroxylamine in methanol to produce target compounds (6). The aniline substituted
compounds 6s-6t were prepared in the same way from the distinct reagent 4-bromoaniline in step b.
Br
Br
HN
Cl
N
HN
Cl
N
N
R1
a
c
N
N
R1
b
N
N
R1
N
N
N
N
N
R2
N
Cl
N
N
H
Cl
Cl
N
4
3
2
1
O
O
HN
OCH3
e
OH
d
HN
HN
N
N
R1
N
N
N
R2
N
R2
N
N
R1
6
5
Scheme 1. Reagents and conditions: a) iodoalkanes, K2CO3, DMF, 0°C to rt, 10 h; b) 3- or 4-
bromobenzyl amine, K2CO3, DMF, 90°C, 4 h; c) heterocyclic amines, 90°C, 3 h; d) methyl acrylate,
Pd(OAc)2, tri(o-tolyl)phosphine, CH3CN, DMF, 90°C, 6 h; e) NH2OH, KOH, CH3OH, 0°C, 2 h.
To explore the biological profile, all synthesized compounds were evaluated against HDAC1 and
CDK2 enzymes, and panobinostat, SAHA and roscovitine were employed as the positive control. The
results are summarized in Table 1 as follows: despite the decreased activities comparing to the positive
control panobinostat, all target compounds still demonstrated good inhibitory activities against HDAC1
at the concentrations of 10 nM and 1000 nM, while some of them could also inhibit CDK2 over 50% at
100 nM, indicating that most of these compounds are dual inhibitors. Among these derivatives,
compound 6d exhibited the most potent activities on CDK2 (IC50 = 56 nM), which was 4-fold more
potent than roscovitine (IC50 = 192 nM), and also showed a remarkable function against HDAC1 (IC50
= 5.8 nM). Furthermore, in order to investigate the effects of substituents at R2 group on the HDAC1
and CDK2 inhibition, compounds 6a, 6b, 6c and 6n were synthesized. The results showed that the
replacement of morpholine by pyrrolidine, piperidine N-methylpiperazine and 4-methylpiperidine led
to a decreased CDK2 inhibition, indicating that the morpholine group at R2 position is favorable for
CDK2 inhibition, and similar trends were observed in compounds 6e, 6r, 6s. In addition, with the

increasing size of R1 (6e - 6g), the inhibitory activity dramatically decreased, indicating that the
substituted group at R1 position was critical for its activity and ethyl group at R1 position seemed to be
optimal for CDK enzyme inhibitory activity. Interestingly, conversion of benzyl substituent to phenyl
group barely influenced the CDK2 or HDAC1 inhibitory activity (6s, 6t). Deserved to be noted, unlike
panobinostat, the compound 6e, which possess an ortho-aniline group instead of a para-aniline (6i),
showed higher binding affinity for HDAC1 probably due to the steric hindrance.
Table 1. Inhibition ratio of target compounds on HDAC1 and CDK2.
R3
HN
N
N
R2
N
N
R1
Inh% (HDAC1)
Inh% (CDK2)
10 nM 100 nM
IC50
IC50
Compd.
R1
R2
R3
(nM)
(nM)
10 nM
1000 nM
N
O
OH
6a
Et
-
-
-
-2.7
22.5
-
N
H
O
O
O
O
N
OH
OH
OH
OH
6b
6c
6d
6e
Et
Et
46
-
99
-
-
6.9
5.6
9.5
4.1
37.2
28.3
67.9
47.5
-
-
N
H
H
N
-
N
H
N
Et
66
63
97
98
5.8
7.0
56
127
N
H
O
N
i-Pr
N
H
O
N
O
N
O
OH
6f
41
97
-
2.0
33.0
-
N
H
O
O
O
OH
OH
OH
6g
6h
6i
n-Bu
Me
46
-
98
-
-
-
-
3.8
13.8
-
11.0
20.8
-
-
-
-
N
H
O
N
N
H
O
N
i-Pr
26
97
N
H
O
N
O
OH
6j
i-Pr
-
-
-
0.8
11.6
-
N
H
N
N
O
O
OH
OH
6k
6l
Me
42
31
100
99
-
-
-7.4
-1.2
1.6
3.7
-
-
N
H
H
N
i-Pr
N
H

O
O
N
OH
OH
6m
6n
Me
Et
-
-
11.5
4.4
16.8
15.5
-
-
N
H
-
-
N
N
-
-
N
H
N
N
O
O
OH
OH
6o
6p
i-Pr
69
-
101
-
-12.1
2.3
21.7
4.6
-
-
N
H
-
-
N
N
n-Bu
N
H
O
O
N
OH
OH
6q
6r
i-Pr
i-Pr
-
-
6.6
12.1
5.0
-
-
N
H
-
-
H
N
58
99
-0.8
N
H
O
O
N
O
OH
N
6s
6t
i-Pr
Me
47
-
100
-
-15.4
3.4
15.2
5.3
-
-
-
H
N
OH
N
H
-
SAHA
-
-
-
46
-
100
-
0.6
-
-
-
-
-
-
-
-
-
-
-
-
-
Panobinostat
roscovitine
-
-
-
-
-
193
-
Not determined
Encouraged by the favorable enzymatic potency, compounds 6d and 6e were selected to evaluate
their growth inhibitory activities against several human tumor cell lines, including human lung cancer
cells (A549), human liver cancer cells (HepG2) and human breast cancer cells (CAL-148). As shown in
table 2, though less efficient than the panobinostat, compound 6d and 6e still inhibited the proliferation
of all the tested cell lines at good degrees of inhibitory activities. HepG2 cell line seemed to be more
sensitive than A549 and CAL-148 lines, and compound 6d (IC50 = 0.77 ± 0.07 μM) and 6e (IC50 = 0.22
± 0.08 μM) exhibited better anti-proliferative activities than roscovitine revealing their potential for
further modification.
Table 2. Cell growth inhibitory activities of compounds on human cancer cell lines.
IC50 (μM) a
Compd.
A549
HepG2
CAL-148
6d
1.49 ± 0.07
1.26 ± 0.02
0.02 ± 0.001
>10
0.77 ± 0.07
0.22 ± 0..08
< 0.1
1.14 ± 0.21
0.39 ± 0..01
< 0.1
6e
Panobinostat
Roscovitine
15.76 ± 3.09
>10
a
Data are mean ± SD values from three independent experiments.

To elucidate the interaction mode of compound 6d, the docking studies of 6d in HDAC1 (PDB code:
4BKX) and CDK2 (PDB code: 2A4L) were performed on software Discovery Studio 2018 (Fig. 3).
Figure 3A showed the binding mode of 6d with the active site of HDAC1. Unsurprisingly, the
hydroxamic acid group chelated the Zn2+ in a coordinated manner and formed hydrogen bonds.
Meanwhile, the aliphatic linker occupied the lipophilic tube, and the cap group extended to a pocket
and interacted with the amino acid residues Pro29 and His28. On the other hand, as shown in the
binding mode of 6d with CDK2 in Figure 3B, the purin moiety of 6d occupied approximately the same
ATP-binding region of CDK2 as the purin ring of roscovitine. Ethyl group on the morphine ring
formed hydrophobic interactions with the residue Ala144 and Phe80, which was similar to the
isopropyl group in roscovitine. It is worth noting that the hydroxamic acid group also formed strong
hydrogen bond interactions with Glu8 and Lys20 residues, which may be the key contributor to the
high potency of 6d against CDK2 compared to roscovitine.
B
A
Fig 3. (A) The binding mode of compound 6d (purple) and panobinostat (yellow) in the active site of HDAC2 (PDB code:
4BKX). (B) The binding mode of compound 6d (purple) and roscovitine (green) in the active site of CDK2 (PDB code: 2A4L).
In conclusion, a series of fused-purin based novel HDAC1/CDK2 dual inhibitors were designed,
synthesized, and evaluated for bioactivity. The growth inhibitory activities of synthesized compounds
were tested using three human tumor cell lines. The potent lead compound 6d with good HDAC1 and
CDK2 inhibitory activity of IC50 = 5.8 and 56 nM, respectively, can effectively induce apoptosis of
these cancer cell lines in vitro, especially on HepG2 cell line. Above studies supported 6d as a good
and novel candidate compound targeting on HDAC1 and CDK2, which deserved further research.
Acknowledgements
We appreciate the financial support from National Natural Science Foundation of China (No.
21701018) and the Natural Science Foundation of Chongqing City (No. cstc2017jcyjAX0228).
References
1. Bolden JE, Peart MJ, Johnstone RW. Nat Rev Drug Discov. 2006; 5: 769.
2. Khan O, La Thangue NB. Immunol Cell Biol. 2012; 90: 85.
3. Zhang Y, Feng J, Jia Y, et al. J Med Chem. 2011; 54: 2823.
4. Korfei M, Skwarna S, Henneke I, et al. Thorax. 2015; 70: 1022.
5. Park SY, Jun JA, Jeong KJ, et al. Oncol Rep. 2011; 25: 1677.
6. Muller S, Kramer OH. Curr Cancer Drug Targets. 2010; 10: 210.

7. Andrews KT, Haque A, Jones MK. Immunol Cell Biol. 2012; 90: 66.
8. Lachenmayer A, Toffanin S, Cabellos L, et al. J Hepatol, 2012; 56: 1343.
9. Shao M, He L, Zheng L, et al. Bioorg Med Chem Lett. 2017; 27: 4051.
10. Santos, F P. Verstovsek, S. Curr Opin Hematol. 2013; 20: 123.
11. Indovina P. Curr Pharm Design. 2013; 19: 5327.
12. Gahr S, Peter G, Wissniowski TT, et al. Oncol Rep, 2008; 20: 1249.
13. Lee J, Lim B, Pearson T, et al. Cancer Res. 2018; 78: P5.
14. Li Y, Luo X, Guo Q, et al. J Med Chem. 2018; 61: 3166.
15. Azevedo WFD, Leclerc S, Meijer L, et al. Eur J Biochem. 1997; 243: 518.

Graphical abstract:
O
OH
N
H
IC50 (HDAC1) = 5.8 nM
IC50 (CDK2) = 56 nM
IC50 (HepG2) = 0.77 ± 0.07
HN


N
N
IC50 (CAL-148) = 1.14 ± 0.21
IC50 (A549) = 1.49 ± 0.07

N
N
N
H3C
O
6d
16.