Discovery of pyrazole derivatives as cellular active inhibitors of histone lysine specific demethylase 5B (KDM5B/JARID1B)

Article abstract of DOI:10.1016/j.ejmech.2020.112161

KDM5B (also known as PLU-1 and JARID1B) is 2-oxoglutarate and Fe²⁺ dependent oxygenase that acts as a histone H3K4 demethylase, which is a key participant in inhibiting the expression of tumor suppressors as a drug target. Here, we present the discovery of pyrazole derivatives compound 5 by structure-based virtual screening and biochemical screening with IC₅₀ of 9.320 μM against KDM5B, and its subsequent optimization to give 1-(4-methoxyphenyl)-N-(2-methyl-2-morpholinopropyl)-3-phenyl-1H-pyrazole-4-carboxamide (27 ab), a potent KDM5B inhibitor with IC₅₀ of 0.0244 μM. In MKN45 cells, compound 27 ab can bind and stabilize KDM5B and induce the accumulation of H3K4me2/3, bona fide substrates of KDM5B, while keep the amount of H3K4me1, H3K9me2/3 and H3K27me2 without change. Further biological study also indicated that compound 27 ab is a potent cellular active KDM5B inhibitor that can inhibit MKN45 cell proliferation, wound healing and migration. In sum, our finding gives a novel structure for the discovery of KDM5B inhibitor and targeting KDM5B may be a new therapeutic strategy for gastric cancer treatment.

Full text of DOI:10.1016/j.ejmech.2020.112161

Journal Pre-proof
Discovery of pyrazole derivatives as cellular active inhibitors of histone lysine specific
demethylase 5B (KDM5B/JARID1B)
Bing Zhao, Qianqian Liang, Hongmei Ren, Xinhui Zhang, Yang Wu, Kun Zhang, Li-
Ying Ma, Yi-Chao Zheng, Hong-Min Liu
PII:
S0223-5234(20)30128-8
DOI:
https://doi.org/10.1016/j.ejmech.2020.112161
EJMECH 112161
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 21 December 2019
Revised Date: 15 February 2020
Accepted Date: 18 February 2020
Please cite this article as: B. Zhao, Q. Liang, H. Ren, X. Zhang, Y. Wu, K. Zhang, L.-Y. Ma, Y.-C.
Zheng, H.-M. Liu, Discovery of pyrazole derivatives as cellular active inhibitors of histone lysine specific
demethylase 5B (KDM5B/JARID1B), European Journal of Medicinal Chemistry (2020), doi: https://
doi.org/10.1016/j.ejmech.2020.112161.
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.

Enamine Library
PDB:5FYZ
≈2,000,000 compounds
our in-house library
≈ 500 compounds

Discovery of pyrazole derivatives as cellular active inhibitors of
histone lysine specific demethylase 5B (KDM5B/JARID1B)
Bing Zhao1, #, Qianqian Liang1, #, Hongmei Ren 1, #, Xinhui Zhang1, Yang
Wu1, Kun Zhang1, Li-Ying Ma1, *, Yi-Chao Zheng 1, * and Hong-Min Liu 1,
*
1Collaborative Innovation Center of New Drug Research and Safety Evaluation,
Henan Province; Key Laboratory of Technology of Drug Preparation (Zhengzhou
University), Ministry of Education of China; Key Laboratory of Henan Province for
Drug Quality and Evaluation; Institute of Pharmaceutical research and School of
Pharmaceutical Sciences, Zhengzhou University, Zhengzhou 450001, PR China
ABSTRACT: KDM5B (also known as PLU-1 and JARID1B) is 2-oxoglutarate and
Fe2+ dependent oxygenase that acts as a histone H3K4 demethylase, which is a key
participant in inhibiting the expression of tumor suppressors as a drug target. Here, we
present the discovery of pyrazole derivatives compound 5 by structure-based virtual
screening and biochemical screening with IC50 of 9.320µM against KDM5B, and its
subsequent
optimization
to
give
1-(4-methoxyphenyl)-N-(2-methyl-2-
morpholinopropyl)-3-phenyl-1H-pyrazole-4-carboxamide (27ab), a potent KDM5B
inhibitor with IC50 of 0.0244µM. In MKN45 cells, compound 27ab can bind and
stabilize KDM5B and induce the accumulation of H3K4me2/3, bona fide substrates of
KDM5B, while keep the amount of H3K4me1, H3K9me2/3 and H3K27me2 without
change. Further biological study also indicated that compound 27ab is a potent
cellular active KDM5B inhibitor that can inhibit MKN45 cell proliferation, wound
healing and migration. In sum, our finding gives a novel structure for the discovery of

KDM5B inhibitor and targeting KDM5B may be a new therapeutic strategy for
gastric cancer treatment.
INTRODUCTION
To data, some studies indicate that the occurrence and development of cancer are
closely related to epigenetics, including DNA and RNA modification, histone
modification and non-coding RNAs [1-5]. Among the diverse epigenetic mechanisms
that modify chromatin structure to regulate gene expression, histone modification
plays a significant role in gene transcription, genomic integrity and chromosome
stability [6-9]. Until now, two classes of lysine demethylases can remove methyl
groups from lysine residues: one is Histone lysine demethylase 1 (KDM1) subfamily,
the other is Histone lysine demethylase 2-7 (KDM2-KDM7) subfamily which contain
JmjC domain. KDM1 subfamily is flavin adenine dinucleotide (FAD) dependent
oxygenase and KDM2-7 subfamily is α-ketoglutarate (2-OG) /Fe2+ - dependent
oxygenase [10-12]. Of the JmjC containing KDMs, lysine demethylase 5B (KDM5B)
is a lysine demethylase that can erase methyl group from H3K4me1/2/3 [13], and it is
overexpressed and plays a vital role in diverse cancers [14, 15], such as breast cancer
[16-18], gastric cancer [19-22], hepatocellular carcinoma [23], and lung cancer [24,
25]. Hence, targeting KDM5B is a new therapeutic strategy for cancer treatment [26,
27], and a range of small-molecule inhibitors of KDM5B has been reported, the
majority of which are 2-OG competitive and coordinate to Fe2+ in the catalytic site,
for example compounds 1-4 [28-31] (Figure 1). Although several KDM5B inhibitors
have been developed, there is still no KDM5B inhibitors in clinic trials due to their
cell permeability, specificity and potency and so on. Hence, it is still a challenging
subject to develop potent KDM5B inhibitors to abrogate the abnormal overexpression
and activation of KDM5B.

Figure 1. Reported small molecule inhibitors of KDM5B.
High-throughput virtual screening (HTVS) is a powerful and efficient tool for drug
discovery [32-34]. In order to discover novel KDM5B inhibitor with new scaffold,
HTVS was performed to identify hit compound using known receptor structure (PDB
code: 5FYZ, resolution: 1.75 Å). Over 2 million compounds from the database
Enamine and our in-house compound library (approximate 500 compounds) were
screened, and hit compound 5 (IC50 = 9.320 µM) was found. Moreover, through
subsequent structural based optimization, compound 27ab was developed as a potent
KDM5B inhibitor with IC50 = 0.0244 µM
, which improves the potency for more than
382-fold, comparing to the hit compound 5.
Herein, we described the identification of pyrazole derivatives as KDM5B inhibitors
with the aid of HTVS and the subsequent structure-based optimization. Among our
synthesized compounds, compound 27ab (IC50 = 0.0244 µM) is a potent and cellular
active KDM5B inhibitor (Figure 2) that can induce the accumulation of H3K4me2/3
as a chemical tool. Besides, compound 27ab can also inhibit the proliferation, colony
formation and would healing of gastric cancer cell line MKN45. Our finding gives a
novel structure for the discovery of KDM5B inhibitor and targeting KDM5B may be
a new therapeutic strategy for gastric cancer treatment.

Enamine Library
PDB:5FYZ
≈2,000,000 compounds
our in-house library
≈ 500 compounds
Figure 2. Identification of hit compound 5 from virtual screening and subsequent
optimizations gives compound 27ab
RESULTS AND DISCUSSION
2.1. The general routes
The general routes of the designed compounds were depicted in Scheme 1-4. In this
work, the pyrazole scaffold was also replaced with pyridine and triazole. The compounds
13a-b and 14a-b of general routes are presented in Scheme 1. Treatment of ketones 6 with
tert-butyl hydrazinecarboxylate 7 in EtOH gave compounds 8a-b, which then reacted with
POCl3 in DMF [35, 36], affording compounds 9a-b [37]. Oxidization of 9a-b using KMnO4
yielded 10a-b. Compounds 13a-b and 14a-b were prepared via reaction of compound
10a-b and commercial compounds 11a-b with appropriately 2-morpholinoethan-1-
amine 12.

Scheme 1. Synthesis route of compounds 13a-b and 14a-b. Reaction conditions: a) cat.
AcOH, EtOH, reflux, 4h/TFA, DCM, 0.5h; b) POCl3, DMF; c) KMnO4, KOH; d) EDCI, HOBT,
DCM, rt, 1h.
As shown in Scheme 2, the 16 were prepared via prolonged heating of kinds of
ketones 6 and phenylhydrazines with different substituents 15. Treatment of compounds
16 with POCl3 in DMF, affording compounds 17, which then reacted with KMnO4 in
dioxane and water to yield the key intermediate derviatives18. In addition, as shown in
Scheme 3, compounds 23 and 25 were prepared via reaction of morpholine 21 and
trimethylsilanecarbonitrile 22 wizth appropriately acetone 19, cyclohexanone 20,
respectively. Treatment of compounds 23 and 25 with LiAlH4 in THF gave compounds 24
and 26. Compounds 18a-r were modified by reacting with compound 24, 26 and kinds of
amines, respectively, affording compounds 27a-ac.

Scheme 2. Synthesis route of intermediattes18a-r. Reaction conditions: a) cat.AcOH, EtOH,
reflux; b) POCl3, DMF; c) KMnO4, KOH; d) 1, 4-dioxane/H2O, rt, 12h;
Scheme 3. Synthesis route of compounds 27a-c and segments of amino. Reaction conditions:
a) EtOH, reflux, 5h; b) THF, LiAlH4, 0 , 6h; c) EDCI, HOBT, DCM, rt, 1h.

Scheme 4. Synthesis route of compounds 33a-b. a) cat.AcOH, EtOH, reflux; b) cat.I2, TBHP,
MeCN, 90 ; c) DCM, Dess-martin periodinane (DMP), 0 -rt; d) THF, t-BuOK, rt.
As shown in Scheme 4, treatment of benzaldehyde 28 with substituted
phenylhydrazine 15 in EtOH gave compounds 29a-b, which were allowed to react
with ethanolamine 29a-b via the iodine-catalyzed in MeCN to yield compounds 31a-
b. Treatment of compounds 31a-b with DMP in DCM, affording compounds 32a-b
[38], which then reacted with 2-methyl-2-morpholinopropan-1-amine 24 in THF,
under the presence of t-BuOK, to yield compounds 33a-b.
2.2. Evaluation of biological activity
2.2.1. KDM5B inhibitory activity and SARs
All synthesized compounds were evaluated for their inhibitory activity against
KDM5B using Amplified Luminescent Proximity Homogeneous Assay (ALPHA) and
CPI-455 was used as the positive control [31, 39]. The inhibitory activity results of
candidate compounds were shown in Table 1-4. The hit compound 5 identified from

the database Enamine was characterized to inhibit KDM5B with IC50 of 9.320 µM,
which encouraged further structural optimization.
Firstly, simplified chiral amine fragments and replacement of the pyrazole ring in 5
with pyridine ring led to the formation of compounds 13a-b and 14a-b (Table 1).
Clearly, compound 13a and 14a-b bearing pyridine ring, were found to be inactive
toward KDM5B. Compound 13b exhibited potent inhibitory activity with IC50 values
of 5.390 µM toward KDM5B. This gave us more confidence of modelled binding of 5
to the KDM5B, albeit compound 14a-b did not perform acceptable activity, which
suggested the importance of pyrazole ring for KDM5B inhibitory activity.
Table 1. Inhibitory effect of compounds 13a-b and 14a-b on KDM5B.
Inhibition at
Compd.
R1
IC50/µM
10 µM
21.53%
75.71%
13a
13b
CH3
>10
5.390
14a
CH3
33.65%
42.50%
-
>10
14b
>10
CPI-455
-
10.52 ± 0.23nM
As shown in Table 2, further structural modifications displayed that the identification
of pyrzole and benzene ring containing compounds 27a, which inactivated KDM5B
potency (IC50 = 0.3102 µM), performed about 18-fold more potency than compound
13b. In-depth structural optimizations focusing on replacement of R1 and R2 groups
affiliated the pyrzole core were performed, yield compounds 27a-27o. Most of them

showed improved KDM5B inhibitory activity. Compounds 27o showed good activity
with an IC50 value of 108.3nM, about 2-fold increase compared with compound 27a.
Compounds 27a-27h contain the same phenyl group (R2), replacement of R1 group
with furyl, pyridyl and phenyl groups with different substituents. At 10µM, the
inhibitory effect of compounds 27g, 27h was significantly decreased with less than
10% inhibition. The remaining compounds involving a hydrophobic R1 with
substituents displayed comparable and potent efficacy towards KDM5B. Interestingly,
modifications of R2 in compounds 27i-27o were found to be inactive against KDM5B
except compound 27o. Conversely, when pyrazole scaffold with a larger substituted
phenyl group, namely 27h and 27n, the inhibitory activity decreased significantly.
Table 2. Inhibitor effect of compounds 27a-o on KDM5B.
Inhibition at
IC50 (µM)/IR
Compd.
27a
R1
R2
10µM
H
H
H
H
H
H
H
H
76.42%
75.38%
81.26%
71.95%
82.32%
45.32%
9.25%
0.3102
0.3353
0.2371
0.5716
0.2382
1.3506
>10
27b
27c
27d
27e
p-chlorophenyl
p-methoxylphenyl
p-ethylphenyl
27f
27g
27h
o-methylphenyl
p-Isoproplphenyl
-5.81%
>10

27i
p-Cl
p-CH3
8.69%
66.78%
-7.78%
20.92%
20.03%
14.81%
93.05%
>10
0.8507
>10
27j
27k
27l
o-CH3
m-CH3
>10
27m
p-CH2CH3
p-CH(CH3)2
p-OCH3
>10
27n
>10
27o
0.1083
10.52 ± 0.23nM
CPI-455
-
Based on above findings, the third-round structural modification were primarily
focused on variations of R3. Interestingly, compound 27v inactivated KDM5B with
IC50 of 73.0 nM, exhibited about 4.2-fold increase in potency, comparable to
compound 27a (IC50 = 310.2 nM, Table 2). 4-Methoxyl-benzyl-substituted compound
27ab exhibited significantly increased potency against KDM5B with IC50 of 24.4nM,
exhibited about 4.5-fold in potency, comparable to compound 27o (IC50 = 108.3 nM,
Table 2). The remaining compounds with different R3 were slightly more potent than
compound 27a, while they were not significantly increased. Compounds 27ac bearing
the 2-methoxy benzyl group, displayed acceptable potency at nanomolar levels (IC50
= 93.20 nM), but was less potent than compound 27ab.
Table 3. Inhibitory effect of compounds 27p-ac on KDM5B.

Inhibition at
IC50 (µM)/IR
Compd.
27p
R1
R2
H
R3
10µM
67.01%
84.63%
92.28%
2.2359
0.3479
0.2004
27q
H
27r
H
27s
H
-1.95%
86.57%
>10
0.2320
27t
H
27u
p-OCH3
p-CH3
p-CH3
p-CH3
p-CH3
p-CH3
p-Cl
17.25%
94.23%
76.45%
82.13%
79.46%
82.36%
62.01%
95.78%
92.32%
>10
0.0730
27v
27w
27x
0.8507
0.2515
27y
0.3234
27z
0.1912
27aa
27ab
27ac
CPI-455
0.5409
p-OCH3
o-OCH3
0.0244
0.0932
-
10.52 ± 0.23nM
Finally, the effects of the core scaffold toward KDM5B were explored by scaffold
hopping strategy, and compounds 33a-b were compared. According to the Table 4,

the pyrazole scaffold was also replaced with triazole, leading to the generation of
compounds 33a and 33b, which were found to be inactive toward KDM5B (IC50 > 10
µM), highlight the essential pyrazole scaffold for the activity.
Table 4. Activity of compounds 27ab and 33a-b on KDM5B
.
Inhibition at IC50 (µM)
10µM
Compd.
27ab
Structure
95.78%
28.03%
28.89%
0.0244
>10
33a
33b
>10
CPI-455
-
10.52 ± 0.23nM
As we have identified compound 27ab as the most potent KDM5B inhibitor among
our synthesized compounds (Figure 3A), selectivity of compound 27ab against other
KDMs was also performed and results showed that compound 27ab may inhibit
KDM4A for 5.12%, 29.36% and 56.25% at 0.01 µM, 0.25µM and 1 µM, respectively;
KDM4C for 3.52%, 39.36% and 50.12% at 0.01 µM, 0.25µM and 1 µM, respectively;
KDM5A for 20.12%, 62.36% and 79.32% at 0.01 µM, 0.25µM and 1 µM,
respectively; KDM5C for 16.35%, 62.11% and 80.24% at 0.01 µM, 0.25µM and 1
µM, respectively and KDM6B for 3.57%, 12.37% and 19.57% at 0.01 µM, 0.25 µM
and 1 µM, respectively (Figure 3B), while compound 27ab may inhibit KDM4A with
IC50 = 0.9542 ± 0.0231 µM, KDM4C with IC50 = 1.1125 ± 0.0301 µM, KDM5A with

IC50 = 0.0249 ± 0.0011 µM, KDM5C with IC50 = 0.0248 ± 0.0019 µM, and inhibit
KDM6A with IC50 more than 1 µM. All these results suggested that compound 27ab
is a potent KDM5B inhibitor with acceptable selectivity against KDM5A/B/C, while
it performed poor activity against KDM4A/C and KDM6B.
Figure 3. Inhibitory activity of compound 27ab against KDMs. (A) Inhibitory curve of compound
27ab against KDM5B; (B) Inhibition rate of compound 27ab against KDM4A, KDM4C, KDM5A,
KDM5B, KDM5C and KDM6B at 0.01 µM, 0.25 µM and 1 µM, respectively. The data represent
the mean ± SD for three independent experiments.
2.3. Molecular docking studies
A molecular docking study was carried out to predict the binding mode of the studied
compounds with KDM5B using Molecular Operating Environment (MOE) platform
(v201608). KDM5B protein structure was extracted from Protein Data Bank with
code 5FYZ (PDB code: 5FYZ, Resolution: 1.75 Å) and was prepared by adding
hydrogen atoms, removing water molecules and small molecules ligand, the docking
studies were in accordance with the results of KDM5B inhibitory activity, and gave
more SAR information of pyrazole derivatives. The 2D binding modes of compound
27ab with KDM5B were shown in Figure 4A. We characterized that compound 27ab

fits well in the tight flat binding pocket of KDM5B, and the oxygen atom of
morpholine has directly bind with the Mn ion, the distance was 2.1 Å (Figure 4A/4B).
The phenyl with p-methoxyl group of compound 27ab would be located in the
junction of hydrophobic pocket and hydrophilic pocket, the phenyl was surrounding
by Tyr 488, Tyr 425, and Val 489 (Figure 4A), and had π-π stacking with benzene
ring of Tyr 488. Also, the oxygen atom of methoxyl formed hydrogen bond with Lys
517, the distance was 2.1 Å (Figure 4B). Addition, the nitrogen of acidamide had
hydrogen interation with Tyr 488 and the distance was 2.9 Å.
Figure 4. The molecular docking result of compound 27ab in binding with KDM5B (PDB
code: 5FYZ). (A) 2D results of compound 27ab catalytic site of KDM5B; (B) 3D close-up
views of hydrogen bond interactions between compound 27ab and KDM5B; Binding pocket
is indicated with transparent surface with compound 27ab shown in brown and protein shown
in cyan; Mn ion used as the crystallographic surrogate for the catalytic Fe2+is shown as purple
red sphere.
2.4. Compound 27ab is a target engaged KDM5B inhibitor in cells
As KDM5B was reported to be overexpressed in gastric cancer and is necessary for
gastric cancer cell proliferation and metastasis [22, 26], and gastric cancer is the fifth

most common cancer in the world and the third most frequent cause of neoplasm-
related deaths [22, 40]. Cellular activity of compound 27ab, one of our most potent
synthesized KDM5B inhibitors, was applied in gastric cancer cells. Before the cellular
activity study of compound 27ab, seven gastric cancer cells were collected and
subjected to evaluate the KDM5B expression in these cells. As shown Figure 5A and
5B, expression of KDM5B in MKN45 cell line was the highest among all the tested
gastric cancer cell lines. So, MKN45 was chosen for the following KDM5B inhibitor
study.
As compound 33a shares similar structure with compound 27ab, but not active
against KDM5B, hence, it was chosen as a negative control for some of the following
studies. First, cellular thermal shift assay was employed to assess the cellular target
engagement of compound 27ab in MKN45 cells. As shown in Figure 5C & D,
cellular KDM5B protein was denatured at 46 in MKN45 cells in the presence of
different concentrations of compound 27ab and compound 33a. After 1h treatment of
indicated compound, thermal stability of cellular KDM5B protein was investigated
and results revealed that compound 27ab can stabilize KDM5B in a dose dependent
manner (Figure 5C), while the negative control compound 33a failed to stabilize
KDM5B in MKN45 cells (Figure 5D), all these results confirmed the cellular target
engagement of compound 27ab in MKN45 cells.

Figure 5. Cellular engagement of compound 27ab. (A) Expression of KDM5B in
gastric cancer cell lines AGS, NCI-N87, MKN45, SGC-7901, MGC-803, HGC-27
and BGC-823 were evaluated by Western Blotting, GAPDH was used as loading
control; (B) Density quantification of KDM5B in the indicated gastric cancer cell
lines; (C) Thermal stability of KDM5B in MKN45 cells treated with compound 27ab
for 1 h and then heated at 46
for 3 min; (D) Thermal stability of KDM5B in
MKN45 cells treated with compound 33a for 1 h and then heated at 46 for 3 min.
GAPDH was used as a loading control. All experiments were repeated for three times.
2.5. Compound 27ab inactivates KDM5B in MKN45 Cells
As stated above, compound 27ab is a target engagement KDM5B inhibitor in cells,
hence, diverse substrates of methylated histone 3 were screened to evaluate its cellular
activity and cellular selectivity. As shown in Figure 6A, 5 days treatment of
compound 27ab, from 0.5 µM to 20 µM, for MKN45 induced significant

accumulation of KDM5B substrate H3K4me2/3, while the negative compound 33a
failed to induce the accumulation of H3K4me2/3. Nevertheless, amount of H3K4me1
was kept constant in the presence of both compound 27ab and compound 33a. On the
other hand, several substrates of KDM4 family and KDM6 family, including
H3K9me2, H3K9me3 and H3K27me2, were also investigated to explore their
accumulation with the compound treatment. Results in Figure 6A suggested that
compound 27ab failed to induce the expression of KDM4 family substrates
H3K9me2 and H3K9me3, and so did the negative compound. Similarly, amount of
H3K27me2, substrate of KDM6 family, was kept unchanged with the treatment of
either compound 27ab or the negative compound, even at high concentration as 20µM.
To further confirm the KDM5B inhibitory effect of compound 27ab in situ,
immunofluorescence using H3K4me2/3 antibody with indicated treatment were
performed with the aid of high content screening for quantification. As shown in
Figure 6B-C, compound 27ab induced the accumulation of H3K4me2 and H3K4me3
in situ in a dose dependent manner when it was applied to MKN45 cells for 5 days,
while the negative compound 33a induced hardly accumulation of H3K4me2 and
H3K4me3. To sum up, compound 27ab can inactivate KDM5B in MKN45 cells
without significant impact on the activity of KDM4 and KDM6 families.

Figure 6. Compound 27ab induced the accumulation of H3K4me2/3 in MKN45 cells.
(A) Western Blotting analysis of the amount of H3K4me1/2/3, H3K9me1/2 and
H3K27me2 in MKN45 cells with indicated treatment for 5 days; (B & C) High
content analysis of H3K4me2 (B) and H3K4me2 (C) in MKN45 cells that were
incubated with indicated compounds for 5 days. Compound 33a was used as a
negative control. Histone 3 was used as loading control. The data represent the mean
± SD for three independent experiments. *P < 0.05, **P < 0.01.
2.6. Compound 27ab inhibits MKN45 cells proliferation, wound healing and
migration in a time and dose dependent manner in MKN45 cells
As KDM5B contributes to gastric cancer cell proliferation and metastasis [22, 26], we
tried to investigate the biological role of compound 27ab in gastric cancer cells. First,

antiproliferation activity of compound 27ab was evaluated. After cells were treated
for 4 days, both compound 27ab and CPI-455 performed moderate antiproliferation
activity in MKN45 cells with IC50 as 26.72 ± 2.28 µM and 30.33 ± 1.71 µM (Figure
7A), respectively. Meanwhile, 6 days treatment of compound 27ab and CPI-455
inactivated the proliferation of MKN45 cells with IC50 as 20.56 ± 2.91 µM and 23.79
± 1.09 µM (Figure 7B), respectively, but no antiproliferation effect against normal
human gastric epithelial GES-1 cell line (antiproliferation rate < 10% even at 100
µM). On the other hand, compound 33a inhibited 10.32% and 13.24% of the
proliferation rate of MKN45 cells for 4 days and 6 days exposure, respectively, which
means compound 33a is inactive against MKN45 proliferation as a negative
compound. As KDM5B was identified as a migration promoter in gastric cancer,
further wound healing assay and transwell experiment were conducted. Results in
Figure 7C suggests that compound 27ab can inhibit the wound healing in a dose and
time dependent manner, while the negative compound 33a failed to inhibit the wound
healing. In the 20 µM group, compound 27ab showed equivalent activity as 20 µM
CPI-455 in inhibiting the wound healing. Besides, activity of compound 27ab against
cell migration was investigated. As the transwell experiment showed in Figure 7D,
different concentrations of compound 27ab can suppress the migration of MKN45
cells in a concentration dependent manner for 24h treatment, and compound 27ab
inhibited cell migration more potently than CPI-455 at 20 µM, while the negative
compound 33a failed to inhibit cell migration even at 20 µM. In a nutshell, compound

27ab can inhibit MKN45 cell proliferation, would healing and migration by targeting
KDM5B with similar potency as CPI-455.
Figure 7. Compound 27ab inhibited MKN45 cells proliferation and migration in a
time and dose dependent manner. (A) Proliferation assay of compound 27ab and CPI-
455 in MKN45 cells for 4 days treatment; (B) Proliferation assay of compound 27ab
and CPI-455 in MKN45 cells for 6 days treatment; (C) Wound healing assay in
MKN45 cells with indicated treatments at 0h, 6h, 12, 24h and 48h; (D) Migration
assay was performed with transwell experiment with indicated treatments for 24h.
Compound 33a was used as a negative control, and CPI-455 was used as a positive
control. All experiments were repeated for three times. The data represent the mean ±
SD for three independent experiments. *P < 0.05, **P < 0.01.
3. Conclusions
In this study, the (R)-N-(2-(2-chlorophenyl)-2-morpholinoethyl)-3-isopropyl-1H-
pyrazole-4-carboxamide (compound 5) was firstly discovered as a KDM5B inhibitor
based on HTVS and biochemical screening. Then, a series of pyrazole derivatives

were designed and synthesized through structural based optimization and obtained
compounds were applied to evaluate the inhibitory activities against KDM5B. Among
them,
1-(4-methoxyphenyl)-N-(2-methyl-2-morpholinopropyl)-3-phenyl-1H-
pyrazole-4-carboxamide (compound 27ab) performed good inhibitory activity against
KDM5B (IC50 = 0.0244 µM) and there was more than 382 folds increase in inhibitory
activity compared to hit compound 5. Further cellular study also indicated that
compound 27ab can stabilize KDM5B and induce the accumulation of H3K4me2/3 in
gastric cancer cell line MKN45, which suggests that compound 27ab is a potent and
cellular active KDM5B inhibitor. Further biological study also indicated that
compound 27ab can inhibit MKN45 cell proliferation, colony formation and wound
healing. Our finding gives a novel structure for the discovery of KDM5B inhibitor
and compound 27ab can be further developed and optimized as a chemical tool by
targeting KDM5B, also, targeting KDM5B may be a new therapeutic strategy for
gastric cancer treatment.
4. Materials and Methods
4.1. Chemistry
4.1.1. Materials
All the reagents and solvents used in the chemical synthesis were obtained from
1
commercial sources and were used without further purification. H NMR and 13C
NMR spectra data were obtained on Bruker AVANCE 400 M spectrometer (Bruker
Instruments, Inc.), Chemical shifts (δ) were reported in parts per million (ppm)
relative to tetramethylsilane (TMS) and J values were reported in Hertz. High

resolution mass spectra (HRMS) were recorded on a Waters Micromass Q-T of
Micromass spectrometer by electrospray ionization (ESI) (Water, Milford, MA) and
HPLC-Mass spectra were recorded on a Waters QDA. The NMR spectra data of
compounds was provided in Supporting Information. All the chemistry reactions were
monitored by thin layer chromatography (TLC) on silica gel. Column
chromatography was carried out at medium pressure using silica gel (200-300 mesh)
purchased from Qingdao Haiyang Chemical Co. Ltd. China. 1-Hydroxybenzo -
triazole (HOBT), Phosphorus Oxychloride and 1-(3-Dimethyl -aminopropyl)-3-
ethylcarbodiimide Hydrochloride (EDCI) were purchased from Beijing Yinuokai
Science and Technology Ltd. China. Compound 5 was purchased from Enamine
(product ID: Z647268608).
4.1.2. General procedure for preparation of the final compounds.
Procedure for preparation of compounds 13a-b, 14a-b
Substituted ketones (6, 5 mmol) was dissolved in ethanol (25 mL), then tert-butyl
hydrazinecarboxylate (7, 5.5 mmol) and CH3COOH (0.5 mmol) were sequentially
added. The solution was heated under reflux for 4 hours, monitoring the progress of
the reaction by TLC. After the reaction was completed, solvents were removed under
vacuum. The residue was dissolved in DCM (5 mL) and CF3COOH (7.5 mmol) were
added, the solution was kept stirring under rt for 0.5h, after completion of the reaction,
solvents were removed under vacuum, the remaining solid 8a-b was used in the next
step without further purification. DMF (3 mL) was slowly added phosphorus
oxychloride (1 mL) under ice bath, stirring for 1 h to get Vilsmeier-Haack Reagent.

Then adding the 8a-b (5 mmol) to the solution, stirring at room temperature for 30
min, then transferring to 80 for 4 h. After the reaction was over, the resulting
mixture was slowly poured into ice-cold water, a saturated solution of sodium
bicarbonate was added to neutralize the mixture, extracted with ethyl acetate. The
organic phase was dried over Na2SO4, filtered and the solvent was evaporated under
reduced pressure to afford crude product 9a-b. To a stirring suspension of 9a-b (5
mmol) in 10 mL of Dioxane/water (4:1), were added of KOH (6 mmol) and potassium
permanganate (6 mmol). The mixture was stirred at room temperature for 3 h. Then
slowly adding Hydrogen peroxide, MnO2 was filtered. The dioxane in the filtrate were
removed under reduced pressure and added water (10 mL), the solution was acidified
to pH 1-2 with concentrated HCl, and the crude product was purified by column
chromatography on silica gel to afford compounds 10a-b. To a solution of 10a-b (2
mmol) in dichloromethane (5 ml), EDCI (2.4 mmol) and HOBT (2.4 mmol) were
added and the reaction mixture was stirred for 20 mins. To the reaction mixture
compound 12 (2 mmol) was added and stirred overnight at room temperature. The
reaction was diluted with water (30 mL) and extracted with DCM (2×30mL), the
organic layers were combined, dried over MgSO4, filtered, and evaporated under
reduced pressure. The product 13a-b was purified by column chromatography
(EtOAc/hexane 3/1). Compounds 14a-b were prepared through similar procedure as
used for the synthesis of compound 13a-b except 10a-b were replaced by 11a-b.
Procedure for preparation of compounds 27a-ac.

To the solution of ethanol (40 mL) was added substituted ketones (6a, 10 mmol) and
phenylhydrazines (15a, 10 mmol) and CH3COOH (11 mmol) were sequentially added.
The solution was heated under reflux for 6-8 hours, monitoring the progress of the
reaction by TLC. After the reaction was completed, solvents were removed under
vacuum and the residue was diluted with water (25 mL) and filtered. The filter cake
was washed repeatedly with a small amount of ethyl acetate, the remaining solid
compound 16 was used in the next step without further purification. The mixture of
DMF (3 mL) and phosphorus oxychloride (1 mL) was stirred for 1 h under ice bath, to
get Vilsmeier-Haack Reagent. Then adding the compound 16 (5 mmol) to the solution,
stirring at rt for 30 min, then transferring to 80 ° C for 4 h. After the reaction was over,
the resulting mixture was slowly poured into ice-cold water, a saturated solution of
sodium bicarbonate was added to neutralize the mixture, obtained solid 17 precipitate
was filtered. To a stirred suspension of 17 (5 mmol) in 10 mL of Dioxane/water (4:1),
were added of KOH (6 mmol) and potassium permanganate (6 mmol). The mixture
was stirred at room temperature for 3 h. Then slowly adding Hydrogen peroxide, filter
out the MnO2 in the system. The dioxane in the filtrate were removed under reduced
pressure and added water (10 mL), the solution was acidified to pH 1-2 with
concentrated HCl, and the resulting precipitate was filtered to get compounds 18a.
Compounds 18b-r were prepared through similar procedure as used for the synthesis
of compound 18a and some of them require column chromatography to purify.
The mixture of morpholine (21, 50 mmol), trimethylsilanecarbonitrile (22, 60 mmol)
and acetone (19, 100mmol) or cyclohexanone (20, 100mmol) in alcohol (20mL) was

reflux for 5h and monitored by TLC. After completion of the reaction, the mixture
was concentrated under reduced pressure. The residue was diluted with ethyl acetate
and extracted with water. The organic phase was dried over Na2SO4, filtering and the
solvent was evaporated under reduced pressure to afford crude product compound 23
and 25 respectively, which was used in the next step without further purification.
Then compound 23 or 25 was dissolved in THF (8mL) at 0 , following the addition
of LiAlH4 (2eq). The reaction mixture was stirring for 6h under ice bath. After
completion of the reaction, the mixture was added saturated ammonium chloride
solution and 25% sodium hydroxide in aqueous solution (1: 1), and filtering, the crude
product was purified by column chromatography on silica gel to afford compounds 24
and 26, respectively. To a solution of 18a-r (2 mmol) in dichloromethane (5 ml),
EDCI (2.4 mmol) and HOBT (2.4 mmol) were added and the reaction mixture was
stirred for 20 minutes. To the reaction mixture compound 24 or 26 or different
substituted amino groups (2 mmol, shown in the Scheme 3) was added and stirred at
room temperature for 12 h. The reaction was diluted with water (30 mL) and extracted
with DCM (2×30mL), the organic layers were combined, dried over MgSO4, filtered,
and evaporated under reduced pressure. The product 27a-ac was purified by column
chromatography by using EtOAc/hexane. The reaction yield varies between 30-89%
depending on the amine of the reactants.
Procedure for preparation of compounds 33a-b
To the solution of ethanol (40 mL) was added benzaldehyde (28, 10 mmol) and
substituted phenylhydrazines (15j or 15o, 10 mmol) and CH3COOH (11 mmol) were

sequentially added. The solution was heated under reflux for 6-8 hours, monitoring
the progress of the reaction by TLC. After the reaction is complete, solvents were
removed under vacuum and the residue was diluted with water (25 mL) and filtered.
The filter cake was washed repeatedly with a small amount of ethyl acetate, the
remaining solid compounds 29a-b were used in the next step without further
purification. To the solution of MeCN (5ml) was added compounds 29a-b (5 mmol),
thanolamine (15 mmol) and followed by the addition of iodine (I2, 20 mol %) and tert-
butyl hydroperoxide (TBHP, 70% aq. 5 mmol). The formed slurry was stirred at 90°
C for 4 h, and monitored by (TLC). After cooling to rt and 20 mL water was added to
the mixture, then extracted by EtOAc for 3 times (30mL×3). The extract was washed
with 10% Na2S2O3 solution(w/w), dried over anhydrous Na2SO4 and the solvent was
removed in vacuo to provide a crude product compounds 31a-b, which was purified
by column chromatography on silica gel (PE/EA = 2/1).
Treatment of compounds 31a-b (5mmol) with DMP (Dess-Martin periodinane,) in
DCM (10 mL) at 0 , then remove ice-bath, the solution was stirring under room
temperature for 3h, monitoring the progress of the reaction by TLC. After the
reaction is complete, and the residue was diluted with water (25 mL), then extracted
by EtOAc for 3 times (20mL×3). The extract was washed with 10% Na2S2O3
solution(w/w), dried over anhydrous Na2SO4 and the solvent was removed in vacuo to
provide a crude product compounds 32a-b, which were used in the next step without
further purification.

To the solution of THF (10 mL) was added compounds 32a-b (5 mmol) and
compound 24 (5 mmol), then potassium tert-Butoxide (0.2eq) was sequentially added,
then the solution was stirring under room temperature for 3h, monitoring the progress
of the reaction by TLC. After the reaction is complete, solvents were removed under
vacuum to provide a crude product compounds 33a-b, which was purified by column
chromatography on silica gel (PE/EA = 3/1).
4.1.3. 3-methyl-N-(2-morpholinoethyl)-1H-pyrazole-4-carboxamide (13a)
White solid, yield 70%, Mp: 107.3-107.9°C. 1H NMR (400 MHz, CDCl3) δ 13.09, (s,
1H), 7.88 (s, 1H), 6.34 (s, 1H), 3.57-3.55 (m, 4H), 3.37-3.32 (m, 2H), 2.45-2.39 (m,
6H), 2.25 (s, 3H). 13C NMR (100.13 MHz, CDCl3) δ 161.84, 146.85, 139.64, 103.80,
66.15, 57.29, 53.17, 35.32, 10.28. MS (ESI) found: m/z 239.16 [M+H]+. HR-MS
(ESI): Calcd. C11H18N4O2, [M+H]+ m/z: 239.1508, found: 239.1519
4.1.4. N-(2-morpholinoethyl)-3-phenyl-1H-pyrazole-4-carboxamide (13b)
White solid, yield 89%, Mp: 100.1-100.3 °C. 1H NMR (400 MHz, CDCl3) δ 13.15, (s,
1H), 7.84 (s, 1H), 7.73-7.71 (m, 2H), 7.41-7.39 (m, 3H), 3.56 (s, 4H), 3.33 (s, 4H),
2.50-2.45 (m, 4H). 13C NMR (100.13 MHz, CDCl3) δ 163.12, 128.42, 127.98, 114.69,
65.90, 57.16, 53.02, 35.65. MS (ESI) found: m/z 301.20 [M+H]+. HR-MS (ESI):
Calcd. C16H20N4O2, [M+H]+ m/z: 301.1665, found: 301.1658.
4.1.5. 5-methyl-N-(2-morpholinoethyl) nicotinamide (14a)
White solid, yield 85%, Mp: 139.3-142.5°C. 1H NMR (400 MHz, CDCl3) δ 8.98-8.95
(t, 1H), 8.88 (s, 1H), 8.65 (s, 1H), 8.12 (s, 1H), 4.02 (s, 2H), 3.69-3.64 (m, 4H), 3.57(s,
13
2H), 3.35-3.32(m, 2H), 3.15 (s, 2H), 2.39(s, 3H). C NMR (100.13 MHz, CDCl3) δ
165.18, 158.43, 158.08, 151.29, 144.78, 136.37, 63.23, 55.11, 51.22, 33.75, 17.72.
MS (ESI) found: m/z 250.34 [M+H]+. HR-MS (ESI): Calcd. C13H19N3O2, [M+H]+
m/z: 250.1477, found:250.1449.

4.1.6. N-(2-morpholinoethyl)-5-phenylnicotinamide (14b)
1
White solid, yield 69%, Mp: 151.3-156.5°C. H NMR (400 MHz, CDCl3) δ 9.61 (s,
1H), 9.12 (s, 1H), 9.07 (s, 1H), 8.56 (s, 1H), 7.83-7.81 (d, J = 8 Hz, 2H), 7.59-7.57 (m,
2H), 7.52-7.48 (m, 1H), 4.04-4.01(m, 2H), 3.73-3.67 (m, 4H), 3.59-3.56(m, 2H),
13
3.37(s, 2H), 3.19(s, 2H). C NMR (100.13 MHz, CDCl3) δ 165.04, 149.13, 146.57,
135.86, 133.71, 129.71, 129.28, 128.80, 127.08, 63.26, 55.12, 51.28, 33.82. MS (ESI)
found: m/z 312.29 [M+H]+. HR-MS (ESI): Calcd. C18H21N3O2, [M+H]+ m/z:
312.1712, found: 312.1708.
4.1.7. N-(2-morpholinoethyl)-1, 3-diphenyl-1H-pyrazole-4-carboxamide (27a)
1
White solid, yield 82%, Mp: 161.6-162.5°C. H NMR (400 MHz, CDCl3) δ 8.51 (s,
1H), 7.73 (dd, J = 20.2, 7.3 Hz, 4H), 7.48 (d, J = 6.9 Hz, 5H), 7.34 (t, J = 7.2 Hz, 1H),
13
6.27 (s, 1H), 3.51 – 3.37 (m, 6H), 2.37 (t, J = 5.6 Hz, 2H), 2.25 (s, 4H). C NMR
(100.13 MHz, CDCl3) δ 162.62, 150.90, 139.40, 132.45, 131.01, 129.47, 129.13,
128.80, 127.30, 119.41, 118.30, 66.72, 56.58, 53.04, 35.65. MS (ESI): m/z 377.2
[M+H]+. HR-MS (ESI) calcd. C22H24N4O2, [M+H]+ m/z: 377.1977, found: 377.1976.
4.1.8. N-(2-morpholinoethyl)-1-phenyl-3-(thiophen-2-yl)-1H-pyrazole-4
-carboxamide (27b)
1
White solid, yield 43%, Mp: 132.2-133.1°C. H NMR (400 MHz, CDCl3) δ 8.5 (s,
1H), 7.75-7.73 (d, J = 8.0 Hz, 2H), 7.59-7.58 (d, J = 3.2 Hz, 1H), 7.49-7.43 (m, 3H),
7.36-7.32 (t, J = 8 Hz, 1H), 7.15-7.12 (dd, 1H), 6.75(s, 1H), 3.64-3.62 (m, 4H), 3.54-
3.50 (m, 2H), 2.56-2.53 (m, 2H), 2.45(m, 4H). 13C NMR (100.13 MHz, CDCl3) δ
162.456, 144.798, 139.173, 133.764, 130.825, 129.577, 128.282, 127.496, 127.409,
127.283, 119.392, 118.098, 66.553, 56.860, 53.151, 35.547. MS (ESI): m/z 383.26
[M+H]+. HR-MS (ESI) calcd. C20H22N4O2S, [M+H]+ m/z: 383.1542, found: 383.1535.
4.1.9. N-(2-morpholinoethyl)-1-phenyl-3-(pyridin-4-yl)-1H-pyrazole-4
-carboxamide (27c)