Discovery of novel small molecule inhibitors of lysine methyltransferase G9a and their mechanism in leukemia cell lines

Article abstract of DOI:10.1016/j.ejmech.2016.06.028

Lysine methyltransferase G9a regulates the transcription of multiple genes by primarily catalyzing mono- and di-methylation of histone H3 lysine 9, as well as several non-histone lysine sites. An attractive therapeutic target in treating leukemia, knockou

Full text of DOI:10.1016/j.ejmech.2016.06.028

European Journal of Medicinal Chemistry 122 (2016) 382e393
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Discovery of novel small molecule inhibitors of lysine
methyltransferase G9a and their mechanism in leukemia cell lines
,
,
Shukkoor M. Kondengaden a 1, Liu-fei Luo b 1, Kenneth Huang a, Mengyuan Zhu a,
Lanlan Zang a c, Eudoxie Bataba a, Runling Wang c, Cheng Luo d, Binghe Wang a,
,
Keqin Kathy Li a *, Peng George Wang a
,
b,
, **
a Chemistry Department and Center for Diagnostics and Therapeutics, Georgia State University, Atlanta, GA 30303, USA
b State Key Laboratory of Medical Genomics, Shanghai Jiao Tong University, Shanghai, China
c Tianjin Key Laboratory on Technologies Enabling Development of Clinical Therapeutics and Diagnostics (Theranostics), School of Pharmacy, Tianjin
Medical University, Tianjin 300070, China
d Drug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai,
China
a r t i c l e i n f o
a b s t r a c t
Article history:
Lysine methyltransferase G9a regulates the transcription of multiple genes by primarily catalyzing mono-
and di-methylation of histone H3 lysine 9, as well as several non-histone lysine sites. An attractive
therapeutic target in treating leukemia, knockout studies of G9a in mice have found dramatically slowed
proliferation and self-renewal of acute myeloid leukemia (AML) cells due to the attenuation of HoxA9-
dependent transcription. In this study, a series of compounds were identified as potential inhibitors
through structure-based virtual screening. Among these compounds, a new G9a inhibitor, DCG066, was
confirmed by in vitro biochemical, and cell based enzyme assays. DCG066 has a novel molecular scaffold
unlike other G9a inhibitors presently available. Similar to G9a’s histone substrate, DCG066 can bind
directly to G9a and inhibit methyltransferase activity in vitro. In addition to suppressing G9a methyl-
transferase activity and reducing histone H3 methylation levels, DCG066 displays low cytotoxicity in
leukemia cell lines with high levels of G9a expression, including K562. This work presents DCG066 as an
inhibitor of G9a with a novel structure, providing both a lead in G9a inhibitor design and a means for
probing the functionality of G9a.
Received 3 February 2016
Received in revised form
13 June 2016
Accepted 15 June 2016
Available online 21 June 2016
Keywords:
Lysine methyltransferase
G9a inhibitor
Leukemia
MALDI-TOF assay
InChIKey:
PMPKMTDYPOAEEH-UHFFFAOYSA-N
© 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction
is unusual [2]. Frequently, the methylation of H4 lysine 20 (H4K20),
H3 lysine 9 (H3K9), and H3 lysine 27 (H3K27) are associated with
Epigenetic regulation plays a vital role in gene transcription and
expression, occurring not through alterations in the genetic code,
but rather by modifications to histones or other vital proteins.
Among the regulatory processes utilized in posttranslational
modifications, histone modification is crucial [1]. Comprised of at
least eight variations, including acetylation, methylation, phos-
phorylation, sumoylation and so forth, the causal relation between
histone methylation in gene expression and histone lysine residues
transcription suppression, whereas methylation of H3 lysine 4
(H3K4), H3 lysine 36 (H3K36) and H3 lysine 79 (H3K79) tends lead
to transcription activation [2e8]. Of interest, H3K9-methylation
correlates with transcription suppression, gene silencing and
genomic stability. H3K9-methylation is regulated by histone lysine
N-methyltransferase EHMT2 (G9a) and analogous protein GLP (G9a
like protein). G9a is responsible for catalyzing the mono- and di-
methylation of H3K9 in euchromatic regions. G9a is also capable
of methylating H3K27, lysine 373 of p53 and other gene expression
regulators. However in several cancers, particularly chronic
myeloid leukemia, G9a and homologous protein GLP are found to
be overexpressed with the proliferation of cancer cells suppressed
through knocking out G9a [9,10]. Because G9a and GLP also play a
significant role in gametogenesis and embryonic development
[11e14], G9a/GLP inhibitors have been used with success in cell
* Corresponding author. Chemistry Department and Center for Diagnostics and
Therapeutics, Georgia State University, Atlanta, GA 30303, USA.
** Corresponding author.
E-mail addresses: kli10@gsu.edu (K.K. Li), pwang11@gsu.edu (P.G. Wang).
These authors contributed equally.
1
http://dx.doi.org/10.1016/j.ejmech.2016.06.028
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
383
reprograming studies and in producing inducible pluripotent stem
cells (iPSCs) by various groups [15e18]. As displayed in Table 1,
current G9a inhibitors can be classified into three overall cate-
gories: type I with BIX-01294 and its derivatives [19,20], type II
with BIX-0138 and its derivatives [21], and type III with fungal
metabolite chaetocin [22,23]. Reported in 2007, BIX-01294 was the
first G9a selective inhibitor that did not target the binding pocket of
cofactor S-adenosyl-methionine (SAM) [24]. Since the discovery of
BIX-01294, many G9a inhibitors including UNC0224, UNC0638,
UNC0642, E72 and A-366 have emerged based on the pharmaco-
phore of BIX-01294 [19,21,25e28]. All of the current G9a inhibitors
possess the same quinazoline core and display similar biological
activity, performing well in enzyme and cellular assays.
Of particular interest is the inhibitors belonging to the type I and
III categories, as clinical trials have indicated their high efficacy in
inhibiting the proliferation of leukemia cells. By attenuating
HoxA9-dependent transcription, G9a inhibitors are able to reduce
the self-renewal and proliferation of acute myeloid leukemia (AML)
cells [20]. Previous studies have indicated that mutations involved
in DNA methylation could be responsible for mediating abnormal
self-renewal and differentiation in leukemia stem cells (LSCs) and,
discovery of targeted drugs for epigenetic enzymes could provide a
new route for leukemia treatment [29,30]. G9a has emerged as an
attractive drug target for leukemia. However, current inhibitors of
G9a are derivatives of parent compound BIX-01294 generated
through manual or structure-activity relationship (SAR) analysis,
offering limited design space for new inhibitors. Thus there is an
interest in searching for new scaffolds capable of inhibiting its ac-
tivity. This study aims to search for new scaffolds for G9a inhibition
through a combination of virtual and enzymatic screening.
High-throughput and structure-based virtual screenings (SBVS)
are widely utilized methods in the discovery of lead structures.
Through SBVS, DCG066 was identified as an inhibitor with binding
affinity comparable to that of the native substrates. In vitro
biochemical and cell based assays verified that DCG066 had a level
of activity comparable to that of BIX-01294, in addition to induction
of cell apoptosis. In studying inhibitors for G9a, the expression
levels of G9a in varying cancer cell lines were also examined to
determine which cell line would be most appropriate for in vitro
assays. From these observations, a trend emerged; the expression
levels of G9a in most hematologic cancer cells were much higher
than in normal blood cells.
Since the overexpression of G9a is related to several cancers
[31], we analyzed the G9a mRNA and protein level using
Table 1
Structures of known G9a inhibitors.

384
S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
quantitative real-time PCR and western blotting, respectively, in
order to quantify expression levels across different cell lines. K562,
a leukemia cell line, was found to have G9a expression levels higher
than all solid tumor cells, and thus was used in further in vitro cell
assays. DCG066 decreases di-methylation levels of histone H3
lysine 9 (H3K9Me2), inhibits cell proliferation and induces cell
apoptosis. Displaying comparable activity to BIX-01294 and lower
cytotoxicity in leukemia cells, DCG066 offers possibilities as both a
probe for the functions of G9a and as a novel lead in G9a inhibitor
design.
flowed over the chip surface in the concentration range of
10e100 M. Any interactions between the compounds and G9a
m
would be detected as changes of response units (RU) in the sen-
sorgram. Among the 125 candidates, 14 were found to interact with
G9a (Fig. S1A). The sensorgram response to the candidate com-
pounds ranged from 10RU to approximately 200RU. Compounds
that failed to display any interactions with G9a were excluded from
subsequent testing.
Having recognized that 14 of the compounds screened showed
significant binding with G9a, we employed a tritium (H3)-labeled
radioactive methylation assay to test the inhibition of enzymatic
2. Results and discussion
activity [34]. 10 mM solutions of the compounds were used for
radioactive assays and four of these compounds displayed more
than 50% inhibition at this concentration. Of these four compounds,
DCG066 displayed inhibition comparable to that of BIX-01294
(Table 2). As indicated in Table 2, all four compounds possess mo-
lecular structures that do not fall into any of the three categories of
existing G9a inhibitors. In an enzyme inhibition assay, DCG066 also
displayed similar activity as BIX-01294 (Fig. 2A and B). Tentatively,
the newly discovered DCG066 could be classified as a type IV
inhibitor.
2.1. Virtual screening
A schematic representation of the overall procedure is pre-
sented in Fig. 1. The crystal structure of G9a in complex with
UNC0638 (PDB entry: 3RJW) was selected to be the binding pocket
in docking studies. Chain A of the protein and the corresponding
UNC0683 were both extracted from 3RJW to serve as the template
for the binding pocket model. In generating the library of com-
pounds to screen, we searched for a public database containing a
large number of small molecule compounds that would be available
for subsequent in vitro screening. The SPECS database (http://www.
specs.net/), containing 200,000 small molecule compounds, was
found to meet the criteria. The database was then filtered for
compounds with log P value smaller than 5 by Pipeline Pilot 7.5,
leaving about 90,000 compounds that would likely be soluble in
aqueous solution to a certain degree. In evaluating which of the two
available docking methods (Glide 5.5, Gold 5.0) were to be used,
two criteria were employed: the root mean-square deviation
(RMSD) value and the enrichment factor. The RMSD value of
UNC0683 between the best-ranked pose by the docking methods
and the initial pose would provide a measure for the accuracy of the
docking algorithm in comparison with experimentally determined
results. The enrichment factor would measure the fraction of active
compounds found in a certain percentage divided by the fraction of
the screened library. Between these two criteria, Glide 5.5 was
found to be the best and was used in docking studies. The Glide XP
mode with default settings was used, following standard proce-
dure. The hit compounds were ranked based on G-score and the top
1000 compounds were selected for scaffold diversity analysis by
Pipeline Pilot 7.5 and visual inspection. Afterwards, candidate
molecules were chosen based on the following criteria: (1) mo-
lecular weight of the ligand is lower than 600 Da, which would
benefit from further structural optimization, (2) both geometric
and chemical features correspond with the active site of G9a, and
(3) the binding poses and chemical structures were reasonable.
From this, 125 candidate molecules were identified for purchase
and further G9a inhibition activity assessment.
2.3. Chemistry
For confirmation and further detailed mechanistic studies, we
synthesized DCG066 (Scheme 1). Commercially available methyl 1-
benzylpiperidine-4-carboxylate (1) was used as the starting ma-
terial to synthesize
fluoromethyl)benzene
3
upon reaction with 1-bromo-4-(tri-
(2) after lithiation. Subsequent
hydrogenolysis afforded a free secondary amine (4), which was
converted to 6 upon nucleophilic substitution with para-nitro-
benzylchloride (5). Reduction of the nitro group to amine 7 fol-
lowed by reaction with dimethylcarbamic chloride yielded the
targeted compound DCG066.
2.4. MALDI-TOF mass spectrum-based biochemical assay
To confirm the result of the radioactive methylation assay,
MALDI-TOF mass spectrum was used to quantitatively determine
the methylation of the histone H3 peptides catalyzed by G9a
enzyme with and without DCG066. 100 nM of protein G9a, 2
substrate peptide and 1 M S-adenosyl methionine (SAM) were
added to 50 mM HEPES pH 8.0, 5 g/mL BSA and 0.1% -mercap-
toethanol with DCG066 at a final concentration of 10 M. After
30 min of reaction, TFA was added to the buffer to terminate the
reaction. 1 L of the reaction solution was used to detect the
mM of
m
m
b
m
m
modifications of histone H3 (1e24). The relative intensity of
H3K9me2 in the sample after the addition of DCG066 was found to
be lower than that without DCG066, indicating the ability for
DCG066 to inhibit H3K9-dimethylation (Fig. 2C and D).
2.2. In vitro screening for direct binding and enzymatic inhibition
2.5. DCG066 inhibits cancer cell proliferation, decreases histone H3
lysine 9 (H3K9) di-methyl levels, blocks cell cycle and induces
apoptosis
To verify whether the 125 selected compounds could inhibit G9a
activity, two different approaches were used to determine the
direct binding and the enzymatic inhibition. Surface plasmon
resonance (SPR) biosensor (Biacore 3000) was used to test for
binding affinity; this was followed by radioactive methylation assay
to test the inhibition of enzymatic activity.
As SPR is one of the best methods for studying label-free binding
affinity [32,33], it is frequently used in drug discovery. To verify
whether the 125 compounds could potentially inhibit G9a activity
through direct binding to the enzyme, we carried out the experi-
ments by using Biacore 3000. G9a was immobilized on the sensor
chip; then the ligand and the candidate compounds sequentially
In order to find a good system to test the G9a inhibitor’s inhi-
bition activity in cell, the distinction of G9a gene expression be-
tween cancer tissues and normal tissue were analyzed by using
GENT [35,36] (Gene Expression across Normal and Tumor tissues)
database (http://medical-genome.kribb.re.kr/GENT/). Meanwhile,
the G9a gene expression levels of all available cancer cell lines were
analyzed by using the databased of Broad-Novartis Cancer Cell Line
Encyclopedia
(http://www.broadinstitute.org/ccle/home).
Compared to normal cells, G9a expression levels of cancer cells are
higher, especially in malignant hematologic cancer cell lines

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
385
Fig. 1. Flowchart of small molecule virtual screening.
(Fig. 3A and B). To ensure that the cells lines used in our following
experiments have appropriate levels of G9a expressed, the G9a
mRNA expression levels were tested by using real time-PCR
(Fig. 3C) and G9a protein expression levels were checked by using
western blot (Fig. 3D). mRNA and protein levels of G9a in the tested
leukemia cells (HL60 and K562 cell lines) are higher than the levels
in solid cancer cells (A549 for lung cancer and HepG2 for liver
cancer). Although normal cells (HUVEC) has detectable G9a protein
expressed, it is obviously lower than that in leukemia cells.
Throughout the in vitro enzymatic activity assay, DCG066 is the
best compound in inhibiting G9a activity among the 125 tested
compounds. To verify the ability of these compounds to inhibit cell
proliferation, DCG066 was tested at varying concentrations in HL-
60, K562, A549 and HepG2 cell lines, and BIX-01294 was used as
a control (Fig. 4A). The half maximal inhibitory concentrations
(IC50) of DCG066 were roughly equivalent to BIX-01294 across the
cell lines, with K562 displaying the greatest sensitivity to DCG066
changes in H3K27 tri-methylation (supplement Fig. S3 B). Based on
these findings, it can be concluded that DCG066 is a G9a selective
inhibitor.
Upon confirming DCG066’s ability to inhibit K562 cell prolifer-
ation, the potential impact of DCG066 on the cell cycle and
apoptosis was investigated. After treating K562 cells with DCG066
and incubating for 24 h, cell cycle and apoptosis were detected
using flow cytometry. DCG066 was found to block the cell cycle at
stage G2/M in the K562 cells, potentially leading to cell apoptosis
(Fig. 4D, C).
2.6. DCG066 binding pocket similarity to BIX-01294 binding pocket
To examine the interactions between DCG066 and G9a, a kinetic
assay by SPR was used with concentrations of DCG066 ranging
from 25
in Wizard mode (Fig. 2D). To calculate the affinity constant, a steady
state affinity model was used (Kd ¼ 18.05 2.48 M).
mM to 300 mM. The test was performed using Biacore 3000
(IC50 ¼ 1.7 0.3
mM) (supplement Fig. S3 A).
DCG066 was also considered for its ability to decrease the level
of dimethylated histone substrate (H3K9me2). After treating
m
To better understand the binding mechanism, we conducted
molecular docking studies with the two positive controls
(UNC0638 and BIX-01294) and the selected 14 molecules from
virtual screening. Because UNC0638 is also the co-crystallized
ligand, we calculated RMSD between the re-docked ligand and
co-crystalized ligand. Our result showed that the RMSD value is
0.995 Å, suggesting our re-docking procedure is reliable. All 16 re-
docked molecules are ranked by “Docking Score” and “Enzymatic
Inhibit Activity” (supplementary S1B). Considering the overall
ranking, DCG066 has the highest performance, therefore DCG066
was chosen for the detailed study (supplementary S1B). Selected
K562 cells with DCG066 at varying concentrations (0.5 mMe2 mM)
and incubating for 72 h, histones were extracted from cell. The
quantified total histones were measured through immuno-blotting
by using anti-histone H3 antibody (Cell Signaling Technology) and
anti-histone H3 methylation antibodies (Abcam) to detect expres-
sion levels of the relevant histone methylation levels. DCG066
reduced histone H3 lysine 9 dimethylation (H3K9Me2) levels in a
similar manner as BIX-01294 (Fig. 4B). 1.0e2.0
mM DCG066
decreased H3K9-dimethylation in K562 cells, but displayed no

386
S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
Table 2
Chemical structures of identified small molecules and their inhibition potency.
(Tentative type IV)
Hypo-thesis
DCG001
SPECS ID
Compound structure
Inhibition ratioa
61
%
Cell inhibitiona
34
%
AH-034/32473019
DCG020
AE-848/11486036
74
65
DCG033
AK-968/40335567
74
91
DCG066
AK-105/40833871
98
88
97
98
BIX-01294
e
a
Compounds used in both G9a enzyme inhibition assay and cell proliferation inhibition assay had a concentration of 10 mM.
docking result images were rendered in supplementary S1C. We
also studied the interactions between DCG066 and G9a compared
to BIX-01294 and UNC0638 (Fig. 5). Top scored conformations were
chosen for analysis. Results indicated that UNC0683, BIX-01294 and
DCG066 all bind to the peptide substrate pocket (Fig. 5A and B).
From the model, BIX and DCG066 were shown to form hydrophobic
contacts with residues Y1085, Y1067, F1087, F1152, Y1154, R1157
and F1158 (Fig. 5CeF). In addition to these, DCG066 also forms extra
hydrogen bonds with Y1154 and R1157 (Fig. 5F), offering a potential
explanation of its binding.
HoxA9-dependent transcription [20].
Many of the current G9a selective inhibitors can be sorted into
three categories, all of which are derivatives of BIX-01294. In this
study, a number of small molecules from the SPECS Database were
screened using virtual screening to discover potential G9a in-
hibitors. Virtual screening followed by bioassays allowed us to
identify a new scaffold embedded in DCG066 for further inhibitor
development. Initially we investigated various cancer cell lines for
their G9a expression profile using immune assay, and found leu-
kemia cell lines to have the highest expression levels of target
protein, hence this cell line were used for the cell based assay. In-
hibition of cancer cell growth was tested by CCK8 and the IC50 of
3. Conclusion
DCG066 found to be comparable to BIX-01294 (1.7
0.1 and
3.9 0.3 respectively in K562 cell line). DCG066 also demonstrated
the ability to induce apoptosis through blocking the cell cycle at
stage G2/M in the K562 cells. Further experiments with DCG066
showed that compound decreased H3K9 dimethylation levels in
K562 cells.
To examine the interactions between DCG066 and the catalytic
domain of G9a more closely, a binding assay was performed by SPR
and the structure-function relationship was modeled using GLIDE.
H3K9 methylation is a crucial epigenetic regulator in gene
transcription and embryonic cell differentiation. Thus, G9a is
emerging as a drug target for leukemia treatments. Recent studies
have shown that G9a may selectively regulate the fast proliferating
myeloid progenitors, but displays no effects on hematopoietic stem
cells (HSCs). Primary acute myeloid leukemia (AML) cells from
patients are found to be sensitive to G9a inhibition, causing
reduction in proliferation and self-renewal through attenuating

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
387
Fig. 2. In vitro kinetic assay and enzyme activity inhibition of DCG066. (A) Binding assay of DCG066 and recombined G9a catalytic domain by Biacore 3000. In this assay, G9a protein
was immobilized on the CM5 sensor chip as ligand by amine coupling. The concentration of DCG066 was range from 25 M to 300 M. Since the sensorgram of each concentration
can reach equilibrium, we choose steady-state analysis to calculate the affinity constant (Kd) of DCG066. The Kd is approximately 18 M calculated by nonlinear least-square (NLLSQ)
fitting. (B) DCG066 can inhibit G9a methyltransferase activity with an IC50 of 6.5 M, which is similar to that of BIX-01294. (C, D) In MADI-TOF-MS, dimethyl level of synthetic
peptide of histone H3 (1e20) decreased with the addition of DCG066, while the monomethyl level increased.
m
m
m
m
Scheme 1. Synthesis of DCG066.

388
S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
Fig. 3. G9a expression in human tissues and cancer cell lines. (A) Pattern of G9a expression across diverse normal and tumor tissues was obtained from GENT (http://medical-
genome.kribb.re.kr/GENT/). The average G9a expression in cancer tissue is higher than normal tissues, particularly for hematology system. (B) Analysis of G9a expression level
in different cancer cell lines. The statistical data was obtained from CCLE (http://www.broadinstitute.org/ccle/home). G9a is expressed higher in leukemia cell lines than other
cancer cells. (C, D) The expression of G9a in different cancer cells (K562, HL-60, HepG2, A549 and HUVEC) was detected in both mRNA level (C) and protein level (D). Expression of
G9a in leukemia cells is higher than other cancer cells (A549 and HepG2) and normal cells (HUVEC).
DCG066 can bind to G9a with low dissociation constant
M) as measured by SPR. While DCG066 binds in the
peptide substrate pocket in G9a with a pose similar to BIX-01294,
DCG066 forms extra hydrogen bonds with residues Y1154 and
R1157, suggesting an explanation for the relative activity between
the two molecules.
600,000 resolution. All chromatographic purification was per-
formed on Biotage Isolera One using normal phase silica gel 60.
HPLC chromatogram for DCG066 compound was acquired using an
Agilent 6110 Series system with UV detector set to 254 nm. Samples
(Kd ¼ 18
m
were injected (5
mL) onto an Agilent Eclipse Plus 4.6 ꢀ 250 mm,
3.5 M, C18 column at room temperature. Sample was eluted using
m
In this study, a new lead compound for G9a inhibition with a
similar activity to BIX-01294 and a novel structure, DCG066, was
identified from a database of small molecules. This discovery pro-
vides both new avenues in G9a inhibitor design and a potential
probe for G9a. In future studies, we plan to examine the structure-
activity relationship (SAR) of DCG066 analogs to G9a in an effort to
improve potency and reduce toxicity.
a linear gradient of 5%e95% B (ACN þ 0.1% TFA) in 30 min with A
being H2O þ 0.1% TFA. DCG066 had >95% purity using the HPLC
method described above.
4.2. Chemistry
4.2.1. (1-Benzylpiperidin-4-yl)bis(4-(trifluoromethyl)phenyl)
methanol
4. Experimental section
Synthesis of the intermediate compound 6 was carried out ac-
cording to the previously reported procedure [37]. n-Butyllitium
solution (19 mL, 1.6 M in hexane) was added to 6.8 g of 4-
4.1. Materials and physical measurements
bromobenzotrifluride
2 (30 mmol) in diethyl ether. To this
All reactions were carried out under an atmosphere of dry ni-
trogen. All reagents and anhydrous solvents were obtained
commercially and used without further purification. 1H NMR
spectra were recorded at 400 MHz and 13C NMR spectra were
recorded at 100 MHz using Me4Si as an internal standard. High-
resolution mass spectra (MS) data were acquired using a Thermo-
Fisher Orbitrap Elite mass spectrometer with an electrospray
ionization (ESI) source. All the samples were run under FTcontrol at
mixture 2.4 (10 mmol) of methyl-1-benzylpiperidine-4-
g
carboxylate 1 were added dropwise at 5 ꢁC. The reaction mixture
was stirred at room temperature for 1 h and at 45 ꢁC for 2 h. The
crude product was obtained according to the reported procedures,
and purified by silica gel flash column chromatography using
0e50% ethyl acetate/hexane as eluent to yield a brown solid. (4.4 g,
77% yield). 1H NMR (CDCl3):
ArH), 7.33e7.30 (m, 5H, Benzyl ArH) 3.56 (s, 2H, Benzylic CH2),
d
7.64 (m, 8H, 2 ꢀ trifluromethylphenyl

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
389
Fig. 4. The anti-cancer effect of DCG066 in cancer cell. (A) DCG066 can inhibit the proliferation of several cancer cells, but effects in healthy cells are small. The cell viability rates
shown as bar, the half maximal inhibitory concentrations (IC50s) are shown as table under the bar Fig. 4A. The IC50 of DCG066 is similar to BIX-01294 and may act better on leukemia
cells compared to adherent cells, such as A549 and HepG2. (B) The test of histone methylation level in K562 with the treatment of DCG066. DCG066 can reduce histone H3 lysine 9
dimethylation (H3K9Me2) levels in a lower concentration than BIX-01294. (C, D) DCG066 can induce apoptosis of K562 cell, with the treatment of 2
mM DCG066 for 24 h.
Meanwhile, DCG066 can block cell cycle in G2/M stage in a low concentration. All these procedures were taken by flow cytometry assay.
3.00e2.97 (m, 2H, NeCH2eCH2), 2.54 (t,
NeCH2eCH2eCH), 2.10e2.07 (m, 2H, NeCH2eCH2), 1.60e1.45 (m,
150.9, 149.6, 139.4, 128.4 (2C),
J
¼
11.6 Hz, 1H,
1H, NeCH2eCH2eCH), 2.12e2.09 (m, 2H, NeCH2eCH2), 1.59e1.45
149.0 (2C), 147.1, 146.3,
129.4 (4C), 129.0, 126.0 (6C), 125.4 (2C), 125.3, 123.5 (2C), 79.3, 62.1,
53.7 (2C), 43.8, 23.2 (2C). HRMS (ESI) m/z: calcd for C27H24F6N2O3
[M þ H]þ 539.1769; found 539.1737.
(m, 4H, 2 ꢀ NeCH2). 13C NMR (CDCl3)
d
4H, 2 ꢀ NeCH2). 13C NMR (CDCl3)
d
128.1 (4C), 127.3 (2C), 126.1 (2C), 125.8 (2C), 125.3 (3C), 125.2 (2C),
81.4, 62.7, 53.8 (2C), 43.8, 26.2 (2C). HRMS (ESI) m/z: calcd for
C
27H25F6NO [M þ H]þ 494.1919; found 494.1883.
4.2.3. (1-(4-Aminobenzyl)piperidin-4-yl)bis(4-(trifluoromethyl)
phenyl)methanol
4.2.2. (1-(4-Nitrobenzyl)piperidin-4-yl)bis(4-(trifluoromethyl)
phenyl)methanol
SnCl2.2H2O (900 mg, 4.0 mmol) was added into the solution of
compound 6 (1.1 g, 2.0 mmol) in EtOH (20 mL) at r.t., a mild exo-
therm was observed. The resulting yellow solution was stirred for 2
days, and then the solvent was removed in vacuo. The residue was
made basic with a 2 N NaOH solution and the aqueous layer was
extracted with DCM (2 ꢀ 25 mL). The combined organic layers were
dried over Na2SO4 and then filtered. The solvent was removed in
vacuo to yield a pale yellow solid (0.65 g, 65% yield), which was
used in the next step without further purification [38]. 1H NMR
The mixture of compound 3 (2 g, 4.04 mmol) and 5 wt% Pd/C
(500 mg) in ethanol (200 mL) was stirred for 24 h at room tem-
perature under hydrogen balloon (1 atm). The reaction mixture was
filtered and concentrated to a sticky solid. HRMS (ESI) m/z: calcd for
C
20H19F6NO [M þ H]þ 404.1449; found 404.1426. Compound 4 was
used for next step without further purification. The mixture of
compound 4 (1.2 g, 3 mmol) and DIPEA (1.2 mL) in 10 mL of
methylene chloride were added into the solution of 4-
nitrobenzylchloride 5 (600 mg, 3.5 mmol) in 20 mL of methylene
chloride. The reaction mixture was stirred for 3 h at room tem-
perature; then saturated NaHCO3 was added. The mixture was
extracted with ethyl acetate, washed with water, dried and
concentrated to give the crude product, which was purified by
column chromatography using 0e40% ethyl acetate/hexane as an
eluent to yield a pale yellow solid. (1.1 g; 67% yield). 1H NMR
(CDCl3)
d
7.59 (m, 8H, 2 ꢀ trifluromethylphenyl ArH), 7.07 (d,
J ¼ 7.5 Hz, 2H, benzyl, m- ArH), 6.63 (d, J ¼ 7.5 Hz, 2H, benzyl, o-
ArH), 3.62 (s, 1H, OH), 3.45 (s, 2H, benzylic CH2), 3.27 (s, 2H, NH2),
2.97 (d, J ¼ 11.2 Hz, 2H, NeCH2eCH2), 2.49 (t, J ¼ 11.6 Hz, 1H,
NeCH2eCH2eCH), 2.03 (t, J ¼ 11.4 Hz, 2H, NeCH2eCH2), 1.69e1.50
(m, 2H, NeCH2), 1.43 (d, J ¼ 12.6 Hz, 2H, NeCH2). 13C NMR (CDCl3)
d
149.4 (2C), 145.6, 130.7 (4C), 129.1, 128.8, 126.7, 126.1 (2C), 125.4
(CDCl3):
d
8.17 (d, J ¼ 8.7 Hz, 2H, Benzyl, m- ArH), 7.61 (m, 8H,
(4C), 125.2 (2C), 114.9 (2C), 79.28, 62.5, 53.1 (2C), 43.8, 25.7 (2C).
HRMS (ESI) m/z: calcd for C27H26F6N2O [M þ H]þ 509.1983; found
509.2000.
2 ꢀ trifluromethylphenyl ArH), 7.50 (d, J ¼ 8.7 Hz, 2H, Benzyl, o-
ArH), 3.60 (s, 2H, Benzylic CH2), 2.92 (m, 2H, NeCH2eCH2), 2.52 (m,

390
S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
Fig. 5. Complex models of G9a with superimposed inhibitors (UNC0683, BIX-01294, DCG066). Left panel (A, C, E) shows the structural electrostatic surfaces (blue indicates positive
charge and red indicates negative charge) and small molecules shows as sticks. Right panel (B, D, F) shows detailed interactions. G9a shows as cartoon in green color and residues,
which could form interactions with UNC0683, BIX-01294 or DCG66 are shown as green sticks. (A, B) The binding pocket of crystal structure of G9a-UNC0683 (PDB: 3RJW) [19]. (C, D)
The computational model of G9a-BIX-01294. (E, F) The computer mimic model of G9a-DCG066. In these models, the interaction surfaces of all these compounds are very similar.
Both BIX-01294 and DCG066 can form hydrophobic interactions with Tyr1085, Tyr1067, Phe1087, Phe1152, Tyr1154, Arg1157 and Phe1158. DCG066 can also form hydrogen bond
with Tyr1154 and Arg1157 (F). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4.2.4. 3-(4-((4-(Hydroxybis(4-(trifluoromethyl)phenyl)methyl)
piperidin-1-yl) methyl)phenyl)-1,1-dimethylurea (DCG066)
Compound 7 (254 mg, 0.5 mmol) was dissolved in 20 mL of
methylene chloride. To this resulting solution was then added
DIPEA (0.18 mL, 1 mmol). After stirring for 30 min at 0 ꢁC, dimethyl
LigandStout for the further screening. The SPECS database was
filtered with log S > ꢂ4 to construct a database of theoretically
soluble compounds (log P < 5). The compounds were then docked
into G9a (PDB entry: 3RJW) at the UNC0638 binding pocket using
the GLIDE program in standard precision mode [39]. The com-
pounds with the highest binding characteristics were then pur-
chased from SPECS Corp (The Netherlands). The computer model of
compound DCG066 and BIX-01294 were docked into G9a (PDB
entry: 3RJW) using GLIDE in the XP mode. The final graphs were
drawn by Pymol software. Docking validation of 16 molecules was
done using Surflex-Dock 2.1 [40] by following our previous pro-
cedures [41]. RMSD value was calculated using Openeye Toolkit
[42].
carbamoyl chloride (54 ml, 0.6 mmol) was added and the reaction
was stirred for another 3 h. Saturated NaHCO3 solution was added
and the mixture was extracted with DCM (3 ꢀ 20 mL). The com-
bined organic phase was dried over Na2SO4 and concentrated un-
der reduced pressure. The residue was purified on silica gel column,
eluting with 5% MeOH in DCM (containing 0.5% Et3N) to give a
yellow solid (180 mg, 62% yield). 1H NMR (CDCl3):
d 7.57 (m, 8H,
2 ꢀ trifluromethylphenyl ArH), 7.30 (d, J ¼ 8.4 Hz, 2H, benzyl, m-
ArH), 7.16 (d, J ¼ 8.4 Hz, 2H, benzyl, o- ArH), 6.57 (s, 1H, OH), 3.47 (s,
2H, benzylic CH2), 2.99 (s, 6H, N(CH3)2), 2.91 (d, J ¼ 11.4 Hz, 2H,
NeCH2eCH2), 2.47 (t, J ¼ 11.6 Hz, 1H, NeCH2eCH2eCH), 2.03 (t,
J ¼ 11.0 Hz, 2H, NeCH2eCH2), 1.60e1.52 (m, 2H, NeCH2), 1.38 (d,
4.4. Biological evaluations
J ¼ 12.3 Hz, 2H, NeCH2). 13C NMR (CDCl3)
d 155.9, 149.5 (2C), 138.2,
4.4.1. Cloning, protein expression and purification
Mouse histone methyltransferase G9a (969e1263) cDNA was
amplified from the cDNA of BALB/c mouse thymus, and the frag-
ment was sub-cloned into a vector with a 6His-sumo tag. The
mouse G9a (mG9a) was expressed in Escherichia coli BL21 (DE3) by
129.8 (2C), 128.9, 128.7, 126.2 (5C), 125.4 (2C), 125.2 (2C), 122.7 (2C),
119.9 (2C), 79.2, 62.3, 53.2 (2C), 43.8, 36.4 (2C), 25.8 (2C). HRMS
(ESI) m/z: calcd for C30H31F6N3O2 [M þ H]þ 580.2354; found
580.2373.
the addition of 1 mM isopropyl-1-thio-
D-galactopyranoside (IPTG)
and incubated overnight at 16 ꢁC.
4.3. Pharmacophore-based screening and computational model
The 6His-sumo mG9a (969e1263) protein was purified using
the following procedure: harvested cell pellet was re-suspended in
Pharmacophore models were automatically constructed using
20 mM Tris (pH 8.0), 500 mM NaCl, 0.1% b-mercaptoethanol, and

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
391
1 mM PMSF. Cells were lysed by sonicating for 15 s with 6-s in-
tervals for a total of 15 min on an ice bath. The supernatant of cell
lysate was loaded onto a Niþ affinity column (Invitrogen) and
eluted with buffer (20 mM Tris-HCl pH 8.0, 500 mM NaCl, 20 mM
MB-231 were grown in Dulbecco’s modified Eagle’s medium
(DMEM) with 10% FBS.
4.4.6. Human G9a expression level analysis
imidazole, 0.1%
b
-mercaptoethanol, and 1 mM PMSF). The 6His-
In this study, G9a expression levels were analyzed both in
cancer cell lines and human tissues. CCLE (Cancer Cell Line Ency-
clopedia, www.broadinstitute.org/ccle/home) and GENT (Gene
Expression database of Normal and Tumor tissues, http://medical-
genome.kribb.re.kr/GENT/) were the databases that provided in-
formation regarding gene expression in cancer cell lines and hu-
man tissues, respectively.
sumo tag was cleaved from the column by adding ubiquitin-like-
specific protease 1 (ULP-1) at 4 ꢁC for 12 h. Wash buffer was then
run through the Niþ column again and the elution buffer collected.
Subsequently, advanced protein purification was done by HiTrap Q
HP sequential Superdex 200 10/300 GL. Elute of every step was
analyzed by SDS PAGE, stained by Coomassie brilliant blue (CBB).
4.4.2. Tritium-labeled radioactive methylation assay
4.4.7. Quantitative real-time PCR
The tritium labeled radioactive methylation assay was used to
test the inhibitory effects of the compounds on enzyme activity
For quantitative real-time PCR (qRT-PCR), total RNA was
extracted and reverse transcribed into cDNA. qRT-PCR was per-
formed in ABI PRISM 7500 using SYBR Green PCR Master mix
(Takara). The mRNA expression level of G9a was normalized to the
transcript level of GAPDH.
[34]. In the methylation assay, 2
mM biotin labeled peptide sub-
strate, 5
m
M [methyl-3H]-SAM (78 Ci/mmol, PerkinElmer), and
varied concentrations of inhibitor were pre-incubated in the reac-
tion buffer (50 mM HEPES pH 8.0, 10 mM NaCl, 1 mM DTT) for
30 min at room temperature. The reaction was initiated by adding
4.4.8. Cell viability analysis
recombination mG9a (969e1263) for a final concentration of 2
ml. 2 L of the reaction mixture was transferred to a 96 well plate
coated with avidin (Corning), incubated for 30 min in PBST with
m
g/
Cell growth inhibition was detected by CCK8 assay (Dojindo).
1 ꢀ 104 cell suspensions were dispensed on a 96-well plate, and
then pre-incubated for 4 h in a humidified incubator. Compounds
were dissolved in DMSO, diluted in DMEM culture medium for a
m
5 mM of unlabeled SAM. The plate was washed 3 times with 200 mL
PBST per well to remove the remaining 3H-SAM. Later, 50 mM HCl
was used to elute the avidin binding peptide. Finally the elution
concentration of 10 ꢀ stock solution, and added as 10
mL per well.
After incubation for 48 h, 10 L of CCK8 solution was added into the
m
buffer mixed with 200
Microbeta scintillation counter (PerkinElmer).
m
L scintillator fluid was analyzed by 1450
wells and the solution was allowed to incubate for 4 h in a CO2
incubator. Afterwards, the absorbance was measured at 450 nm.
4.4.3. MALDI-TOF-MS
4.4.9. Flow cytometry assays
The in vitro inhibition of G9a by the compounds was measured
by MALDI-TOF mass spectrum (4800 Plus MALDI TOF/TOF Analyzer,
Cell apoptosis was determined by dual staining with annexin V
conjugated to phycoerythrin (PE) and 7-amino-actinomycin
ABI). 100 nM purified mG9a, 2
and 1 M non-radioactive S-adenosyl methionine (Sigma) were
added into a reaction buffer (50 mM HEPES pH 8.0, 5 g/mL BSA
and 0.1% -mercaptoethanol) with or without an inhibitor for a
final concentration of 10 M. The reaction was incubated at room
temperature for an hour, and stopped by TFA addition.
The result of mass spectrum was analyzed using the Data Ex-
plorer (TM) software, providing peak area scores while a statistical
graph was drawn by Graphpad Prism 5.0.
m
M synthesized histone H3 (1e24)
(7AAD). K562 cells were treated with 0, 2, 4, 6, 8, 10 mM of DCG066.
m
After incubation for 24 h, cells were collected and stained with
annexin V-PE and 7AAD (BD Pharmingen) for 15 min in the dark
and analyzed by flow cytometry. Cells undergoing apoptosis were
identified as annexin Vþ and/or 7AAD þ cells.
m
b
m
Additionally, cells (1 ꢀ 106) were treated with DCG066 (0, 1, 2,
4 mM) or DMSO for cell cycle analysis. After 24 h of incubation, cells
were collected and washed with cold PBS twice, and then sus-
pended in 300 L PBS, fixed by 700 L ethanol. The fixed cells were
m
m
washed by PBS twice and re-suspended in PI/RNase Staining Buffer
(BD Pharmingen) and then incubated for 15 min before analysis.
Flow cytometry experiments were performed using a LSR II cy-
tometer (BD Pharmingen), and data were analyzed by using the
FlowJo 7.6.1 software.
4.4.4. Surface plasmon resonance
The ability of the small molecular compounds to bind to the
target enzyme was tested by using SPR (Biacore 3000, GE).
Approximately 5000RU of the target protein was immobilized on a
CM5 sensor chip. Test compounds were dissolved in running buffer
(PBS þ P20 þ 5%DMSO) at a concentration of 100
m
M.
4.4.10. Histone extraction and western blot
To screen the compounds, solutions were sequentially injected
for 1 min at the associated stage, dissociated in running buffer for
1 min, and the sensor chip was then regenerated by running buffer
for 1 min. In this test, an empty cycle per every 5 cycles was run.
From the Biacore evaluation software (BIA evaluation version 4.1),
the maximum binding values of every compound were obtained.
For the G9a catalytic domain kinetics assay, five concentrations
of the compounds were prepared and DMSO contents in these
samples were equivalent with 5%. The test was performed in the
Wizard mode; and the injection time and dissociation times were
recorded. The results were analyzed using the static analysis option
in the BIA evaluation software.
Histones from cell lysates were extracted using trichloroacetic
acid precipitation as described previously [43]. 1 ꢀ 107 cells were
harvested by centrifugation at 800 ꢀ g for 5 min, and suspended on
ice in hypotonic buffer (10 mM Tris-HCl pH 8.0, 1 mM KCl, 1.5 mM
MgCl2, 1 mM DTT, 1 mM PMSF) for 30 min. The cell lysate was
centrifuged at 13,000 ꢀ g for 10 min, and the supernatant was
discarded. The lysate pellets were re-suspended using 0.2 N H2SO4,
mixing at rotor at 4 ꢁC for 30 min, centrifuged at 13,000 ꢀ g for
10 min at 4 ꢁC with the precipitation discarded. 50% trichloroacetic
acid (TCA) was added to the supernatant dropwise, and then the
resulting solution was centrifuged for 20 min. The pellets were
washed twice in ice-cold acetone and then dissolved by adding
H2O. Histone content was quantified by a Bradford assay and purity
was tested by Coomassie blue staining.
4.4.5. Cell culture
All leukemia cells such as HL-60, K562, U937 and Kasumi-1 were
grown in RPMI 1640 with 10% fetal bovine serum (FBS), while
adherent cell, including A549, HepG2, HCT116, SW1990, and MDA-
Extracted histone was denatured by SDS loading buffer. Anti-
histone H3 antibody (rabbit), anti-histone H3 lysine 9 trimethyl
antibody (mouse), anti-histone H3 lysine 9 trimethyl antibody

392
S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
[19] M. Vedadi, D. Barsyte-Lovejoy, F. Liu, S. Rival-Gervier, A. Allali-Hassani,
V. Labrie, T.J. Wigle, P.A. Dimaggio, G.A. Wasney, A. Siarheyeva, A. Dong,
W. Tempel, S.C. Wang, X. Chen, I. Chau, T.J. Mangano, X.P. Huang,
C.D. Simpson, S.G. Pattenden, J.L. Norris, D.B. Kireev, A. Tripathy, A. Edwards,
B.L. Roth, W.P. Janzen, B.A. Garcia, A. Petronis, J. Ellis, P.J. Brown, S.V. Frye,
C.H. Arrowsmith, J. Jin, A chemical probe selectively inhibits G9a and GLP
methyltransferase activity in cells, Nat. Chem. Biol. 7 (2011) 566e574.
[20] B. Lehnertz, C. Pabst, L. Su, M. Miller, F. Liu, L. Yi, R. Zhang, J. Krosl, E. Yung,
J. Kirschner, P. Rosten, T.M. Underhill, J. Jin, J. Hebert, G. Sauvageau,
R.K. Humphries, F.M. Rossi, The methyltransferase G9a regulates HoxA9-
dependent transcription in AML, Genes & Dev. 28 (2014) 317e327.
[21] Y. Yuan, Q. Wang, J. Paulk, S. Kubicek, M.M. Kemp, D.J. Adams, A.F. Shamji,
B.K. Wagner, S.L. Schreiber, A small-molecule probe of the histone methyl-
transferase G9a induces cellular senescence in pancreatic adenocarcinoma,
ACS Chem. Biol. 7 (2012) 1152e1157.
[22] E. Iwasa, Y. Hamashima, S. Fujishiro, E. Higuchi, A. Ito, M. Yoshida, M. Sodeoka,
Total synthesis of (þ)-chaetocin and its analogues: their histone methyl-
transferase G9a inhibitory activity, J. Am. Chem. Soc. 132 (2010) 4078e4079.
[23] H. Chaib, A. Nebbioso, T. Prebet, R. Castellano, S. Garbit, A. Restouin, N. Vey,
L. Altucci, Y. Collette, Anti-leukemia activity of chaetocin via death receptor-
dependent apoptosis and dual modulation of the histone methyl-transferase
SUV39H1, Leukemia 26 (2012) 662e674.
(rabbit), and mono- and di-methyl arginine antibody (mouse) were
purchased from Abcam. Anti-histone H3 lysine 27 trimethyl anti-
body (rabbit) was purchased from Millipore.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China grants (81270622) to KLI, the China 973 Pro-
gram (2013CB733700 and 2011CB510102) to KLI and Georgia
Research Alliance funding to P.G. Wang.
Appendix A. Supplementary data
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.ejmech.2016.06.
028. These data include MOL files and InChiKeys of the most
important compounds described in this article.
[24] S. Kubicek, R.J. O’Sullivan, E.M. August, E.R. Hickey, Q. Zhang, M.L. Teodoro,
S. Rea, K. Mechtler, J.A. Kowalski, C.A. Homon, T.A. Kelly, T. Jenuwein, Reversal
of H3K9me2 by a small-molecule inhibitor for the G9a histone methyl-
transferase, Mol. Cell 25 (2007) 473e481.
References
[25] F. Liu, X. Chen, A. Allali-Hassani, A.M. Quinn, G.A. Wasney, A. Dong, D. Barsyte,
I. Kozieradzki, G. Senisterra, I. Chau, A. Siarheyeva, D.B. Kireev, A. Jadhav,
J.M. Herold, S.V. Frye, C.H. Arrowsmith, P.J. Brown, A. Simeonov, M. Vedadi,
J. Jin, Discovery of a 2,4-diamino-7-aminoalkoxyquinazoline as a potent and
selective inhibitor of histone lysine methyltransferase G9a, J. Med. Chem. 52
(2009) 7950e7953.
[26] F. Liu, D. Barsyte-Lovejoy, A. Allali-Hassani, Y. He, J.M. Herold, X. Chen,
C.M. Yates, S.V. Frye, P.J. Brown, J. Huang, M. Vedadi, C.H. Arrowsmith, J. Jin,
Optimization of cellular activity of G9a inhibitors 7-aminoalkoxy-quinazo-
lines, J. Med. Chem. 54 (2011) 6139e6150.
[27] F. Liu, D. Barsyte-Lovejoy, F. Li, Y. Xiong, V. Korboukh, X.P. Huang, A. Allali-
Hassani, W.P. Janzen, B.L. Roth, S.V. Frye, C.H. Arrowsmith, Discovery of an
in vivo chemical probe of the lysine methyltransferases G9a and GLP, J. Med.
Chem. 56 (21) (2013) 8931e8942.
[28] R.F. Sweis, M. Pliushchev, P.J. Brown, J. Guo, F. Li, D. Maag, A.M. Petros,
N.B. Soni, C. Tse, M. Vedadi, M.R. Michaelides, G.G. Chiang, W.N. Pappano,
Discovery and development of potent and selective inhibitors of histone
methyltransferase G9a, ACS Med. Chem. Lett. (2014), 140110075408001.
[29] X.-J. Yan, J. Xu, Z.-H. Gu, C.-M. Pan, G. Lu, Y. Shen, J.-Y. Shi, Y.-M. Zhu, L. Tang,
X.-W. Zhang, Exome sequencing identifies somatic mutations of DNA meth-
yltransferase gene DNMT3A in acute monocytic leukemia, Nat. Genet. 43
(2011) 309e315.
[1] T. Kouzarides, Chromatin modifications and their function, Cell 128 (2007)
693e705.
[2] A. Barski, S. Cuddapah, K. Cui, T.Y. Roh, D.E. Schones, Z. Wang, G. Wei,
I. Chepelev, K. Zhao, High-resolution profiling of histone methylations in the
human genome, Cell 129 (2007) 823e837.
[3] R. Cao, L. Wang, H. Wang, L. Xia, H. Erdjument-Bromage, P. Tempst, R.S. Jones,
Y. Zhang, Role of histone H3 lysine 27 methylation in Polycomb-group
silencing, Science 298 (2002) 1039e1043.
[4] A. Kirmizis, S.M. Bartley, A. Kuzmichev, R. Margueron, D. Reinberg, R. Green,
P.J. Farnham, Silencing of human polycomb target genes is associated with
methylation of histone H3 Lys 27, Genes & Dev. 18 (2004) 1592e1605.
[5] G. Schotta, M. Lachner, K. Sarma, A. Ebert, R. Sengupta, G. Reuter, D. Reinberg,
T. Jenuwein, A silencing pathway to induce H3-K9 and H4-K20 trimethylation
at constitutive heterochromatin, Genes & Dev. 18 (2004) 1251e1262.
[6] A. Shilatifard, Molecular implementation and physiological roles for histone
H3 lysine 4 (H3K4) methylation, Curr. Opin. Cell Biol. 20 (2008) 341e348.
[7] A.J. Ruthenburg, C.D. Allis, J. Wysocka, Methylation of lysine 4 on histone H3:
intricacy of writing and reading a single epigenetic mark, Mol. Cell 25 (2007)
15e30.
[8] A.G. Evertts, A.L. Manning, X. Wang, N.J. Dyson, B.A. Garcia, H.A. Coller, H4K20
methylation regulates quiescence and chromatin compaction, Mol. Biol. Cell
24 (2013) 3025e3037.
[30] K.K. Li, L.-F. Luo, Y. Shen, J. Xu, Z. Chen, S.-J. Chen, DNA methyltransferases in
hematologic malignancies, in: Seminars in Hematology, Elsevier, 2013, pp.
48e60.
[9] J. Huang, J. Dorsey, S. Chuikov, L. Perez-Burgos, X. Zhang, T. Jenuwein,
D. Reinberg, S.L. Berger, G9a and Glp methylate lysine 373 in the tumor
suppressor p53, J. Biol. Chem. 285 (2010) 9636e9641.
[31] M.B. Kunze, D.W. Wright, N.D. Werbeck, J. Kirkpatrick, P.V. Coveney,
D.F. Hansen, Loop interactions and dynamics tune the enzymatic activity of
the human histone deacetylase 8, J. Am. Chem. Soc. 135 (2013) 17862e17868.
[32] R. Karlsson, SPR for molecular interaction analysis: a review of emerging
application areas, J. Mol. Recognit. JMR 17 (2004) 151e161.
[10] M.W. Chen, K.T. Hua, H.J. Kao, C.C. Chi, L.H. Wei, G. Johansson, S.G. Shiah,
P.S. Chen, Y.M. Jeng, T.Y. Cheng, T.C. Lai, J.S. Chang, Y.H. Jan, M.H. Chien,
C.J. Yang, M.S. Huang, M. Hsiao, M.L. Kuo, H3K9 histone methyltransferase G9a
promotes lung cancer invasion and metastasis by silencing the cell adhesion
molecule Ep-CAM, Cancer Res. 70 (2010) 7830e7840.
[33] K.A. Jun Matsui, Noriaki Hara, Daisuke Miyoshi, Hidemi Nawafune,
Katsuyuki Tamaki, Naoki Sugimoto, SPR sensor chip for detection of small
molecules using molecularly imprinted polymer with embedded Gold nano-
particles, Anal. Chem. 77 (2005) 4282e4285.
[11] V. Russo, N. Bernabo, O. Di Giacinto, A. Martelli, A. Mauro, P. Berardinelli,
V. Curini, D. Nardinocchi, M. Mattioli, B. Barboni, H3K9 trimethylation pre-
cedes DNA methylation during sheep oogenesis: HDAC1, SUV39H1, G9a, HP1,
and Dnmts are involved in these epigenetic events, J. Histochem. Cytochem.
Off. J. Histochem. Soc. 61 (2013) 75e89.
[12] N. Feldman, A. Gerson, J. Fang, E. Li, Y. Zhang, Y. Shinkai, H. Cedar, Y. Bergman,
G9a-mediated irreversible epigenetic inactivation of Oct-3/4 during early
embryogenesis, Nat. Cell Biol. 8 (2006) 188e194.
[13] S. Epsztejn-Litman, N. Feldman, M. Abu-Remaileh, Y. Shufaro, A. Gerson,
J. Ueda, R. Deplus, F. Fuks, Y. Shinkai, H. Cedar, Y. Bergman, De novo DNA
methylation promoted by G9a prevents reprogramming of embryonically
silenced genes, Nat. Struct. Mol. Biol. 15 (2008) 1176e1183.
[14] Y. Ushijima, Y.H. Inoue, T. Konishi, D. Kitazawa, H. Yoshida, K. Shimaji,
H. Kimura, M. Yamaguchi, Roles of histone H3K9 methyltransferases during
Drosophila spermatogenesis, Chromosome Res. Int. J. Mol. Supramol. Evol.
Asp. chromosome Biol. 20 (2012) 319e331.
[34] J. Wang, L. Chen, S.H. Sinha, Z. Liang, H. Chai, S. Muniyan, Y.W. Chou, C. Yang,
L. Yan, Y. Feng, K.K. Li, M.F. Lin, H. Jiang, Y.G. Zheng, C. Luo, Pharmacophore-
based virtual screening and biological evaluation of small molecule inhibitors
for protein arginine methylation, J. Med. Chem. 55 (2012) 7978e7987.
[35] J. Barretina, G. Caponigro, N. Stransky, K. Venkatesan, A.A. Margolin, S. Kim,
C.J. Wilson, J. Lehar, G.V. Kryukov, D. Sonkin, A. Reddy, M. Liu, L. Murray,
M.F. Berger, J.E. Monahan, P. Morais, J. Meltzer, A. Korejwa, J. Jane-Valbuena,
F.A. Mapa, J. Thibault, E. Bric-Furlong, P. Raman, A. Shipway, I.H. Engels,
J. Cheng, G.K. Yu, J. Yu, P. Aspesi, M. de Silva, K. Jagtap, M.D. Jones, L. Wang,
C. Hatton, E. Palescandolo, S. Gupta, S. Mahan, C. Sougnez, R.C. Onofrio,
T. Liefeld, L. MacConaill, W. Winckler, M. Reich, N. Li, J.P. Mesirov, S.B. Gabriel,
G. Getz, K. Ardlie, V. Chan, V.E. Myer, B.L. Weber, J. Porter, M. Warmuth,
P. Finan, J.L. Harris, M. Meyerson, T.R. Golub, M.P. Morrissey, W.R. Sellers,
R. Schlegel, L.A. Garraway, The Cancer Cell Line Encyclopedia enables pre-
dictive modelling of anticancer drug sensitivity, Nature 483 (2012), 603e307.
[36] G. Shin, T.-W. Kang, S. Yang, S.-J. Baek, Y.-S. Jeong, S.-Y. Kim, GENT: gene
expression database of normal and tumor tissues, Cancer Inf. 10 (2011)
149e157.
[15] Y. Shi, C. Desponts, J.T. Do, H.S. Hahm, H.R. Scholer, S. Ding, Induction of
pluripotent stem cells from mouse embryonic fibroblasts by Oct4 and Klf4
with small-molecule compounds, Cell Stem Cell 3 (2008) 568e574.
[16] Y. Shi, J.T. Do, C. Desponts, H.S. Hahm, H.R. Scholer, S. Ding, A combined
chemical and genetic approach for the generation of induced pluripotent stem
cells, Cell Stem Cell 2 (2008) 525e528.
[37] S. Farooq, S. Trah, A. Jeanguenat, Derivatives of (1-benzyl-piperidine-4-yl)-
diphenyl-methanol and their use as pesticide, in: Google Patents, 2003.
[38] B. Lagu, M. Wachter, Quaternary Salt CCR2 Antagonists, U.S. Patent No.
7,799,824, 21 Sep. 2010.
[17] N.V. Mezentseva, J. Yang, K. Kaur, G. Iaffaldano, M.C. Remond, C.A. Eisenberg,
L.M. Eisenberg, The histone methyltransferase inhibitor BIX01294 enhances
the cardiac potential of bone marrow cells, Stem Cells Dev. 22 (2013)
654e667.
[39] R.A. Friesner, J.L. Banks, R.B. Murphy, T.A. Halgren, J.J. Klicic, D.T. Mainz,
M.P. Repasky, E.H. Knoll, M. Shelley, J.K. Perry, D.E. Shaw, P. Francis,
P.S. Shenkin, Glide: a new approach for rapid, accurate docking and scoring. 1.
[18] B. Feng, J.-H. Ng, J.-C.D. Heng, H.-H. Ng, Molecules that promote or enhance
reprogramming of somatic cells to induced pluripotent stem cells, Cell Stem
Cell 4 (2009) 301e312.

S.M. Kondengaden et al. / European Journal of Medicinal Chemistry 122 (2016) 382e393
Method and assessment of docking accuracy, J. Med. Chem. 47 (2004) 1) 468e476.
393
1739e1749.
[42] I. OpenEye Scientific Software, USA, www.eyesopen.com, OEChem, inSanta Fe,
NM, 2014.
[43] A.P. Butler, C.V. Byus, T.J. Slaga, Phosphorylation of histones is stimulated by
phorbol esters in quiescent Reuber H35 hepatoma cells, J. Biol. Chem. 261 (20)
(1986 Jul 15) 9421e9425.
[40] A.N. Jain, Surflex-Dock 2.1: robust performance from ligand energetic
modeling, ring flexibility, and knowledge-based search, J. Comput. Aided Mol.
Des. 21 (2007) 281e306.
[41] T.D. Leng, H.F. Si, J. Li, T. Yang, M. Zhu, B. Wang, R.P. Simon, Z.G. Xiong,
Amiloride analogs as ASIC1a inhibitors, CNS Neurosci. Ther. 22 (6) (2016 Jun