Rational design, synthesis and biological profiling of new KDM4C inhibitors

Article abstract of DOI:10.1016/j.bmc.2019.115128

The human histone demethylases of the KDM4 family have been related to diseases such as prostate and breast cancer. Majority of currently known inhibitors suffer from the low permeability and low selectivity between the enzyme isoforms. In this study, toxoflavin motif was used to design and synthesize new KDM4C inhibitors with improved biological activity and in vitro ADME properties. Inhibitors displayed good passive cellular permeability and metabolic stability. However, diminishing of redox liability and consequently non-specific influence on cell viability still remains a challenge.

Full text of DOI:10.1016/j.bmc.2019.115128

Journal Pre-proofs
Rational design, synthesis and biological profiling of new KDM4C inhibitors
Vatroslav Letfus, Dubravko Jelić, Ana Bokulić, Adriana Petrinić Grba, Sanja
Koštrun
PII:
S0968-0896(19)31071-5
https://doi.org/10.1016/j.bmc.2019.115128
BMC 115128
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
26 June 2019
12 September 2019
18 September 2019
Please cite this article as: V. Letfus, D. Jelić, A. Bokulić, A. Petrinić Grba, S. Koštrun, Rational design, synthesis
and biological profiling of new KDM4C inhibitors, Bioorganic & Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.bmc.2019.115128
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.
© 2019 Elsevier Ltd. All rights reserved.

Graphical Abstract
Rational design, synthesis and biological
profiling of new KDM4C inhibitors
Leave this area blank for abstract info.
Vatroslav Letfus*, Dubravko Jelić, Ana Bokulić, Adriana Petrinić Grba, Sanja Koštrun
Fidelta Ltd., Prilaz baruna Filipovića 29, 10000 Zagreb, Croatia
Advanced hit
Starting point
IC50=541 nM
Rational
design
IC50=8 nM
Optimized
ADME

Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com
Rational design, synthesis and biological profiling of new KDM4C inhibitors
Vatroslav Letfus*, Dubravko Jelić, Ana Bokulić, Adriana Petrinić Grba, Sanja Koštrun
Fidelta Ltd., Prilaz baruna Filipovića 29,10000 Zagreb, Croatia
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
The human histone demethylases of the KDM4 family have been related to diseases such as
prostate and breast cancer. Majority of currently known inhibitors suffer from the low
permeability and low selectivity between the enzyme isoforms. In this study, toxoflavin motif was
used to design and synthesize new KDM4C inhibitors with improved biological activity and in
vitro ADME properties. Inhibitors displayed good passive cellular permeability and metabolic
stability. However, diminishing of redox liability and consequently non-specific influence on cell
viability still remains a challenge.
Available online
Keywords:
Epigenetics
KDM4C enzyme
Inhibitor
2009 Elsevier Ltd. All rights reserved.
Drug design
Computational chemistry
protein structures includes a JMJN and JMJC domains, two PHD
and two Tudor domains.
1. Introduction
KDM4A, B and C are capable of catalyzing demethylation of
Chromatin is a complex aggregation of DNA and histone proteins
that allows large amounts of DNA to fit in the restricted space of
a eukaryotic cell's nucleus.1 Dynamic regulation of its structure is
needed to maintain the processes that require access to DNA in
response to physiological and environmental changes.2
lysine residues 9 (H3K9) and 36 (H3K36) on histone 3, while
KDM4D is selective for the demethylation of only H3K9.
Moreover, KDM4A and KDM4C, compared to KDM4B, seem to
be much more catalytically active. Thus, the five KDM4 family
members exhibit different biochemistry and are thought to perform
distinct physiological functions.11-12
Methylation of lysine amino acid residues on histones plays a
significant role in the epigenetic regulation of gene expression by
applying different transcription factors to the chromatin.3-5
Homeostasis of these methylation motifs by different histone
lysine methyl transferase and histone lysine demethylase enzymes
(KDMs) is of key importance to preserve proper cellular
development and to enable differentiation of cells.6-7
Dysregulation of histone lysine methylations has been linked to
deviant gene expression and malignant transformations.8-10
Histone lysine methylation is a reversible process. The removal of
methyl groups is catalyzed by two enzyme families. The first is
flavin-dependent lysine specific demethylase (LSD/KDM1)
family that demethylate mono- and dimethyl lysine substrates. The
second family is comprised of several enzymes, which are
members of the Jumonji C (JmjC) domain containing Fe2+ and α-
ketoglutarate dependent di-oxygenase enzymes which can
additionally convert trimethyl lysine substrates to their dimethyl
forms. These enzymes are further classified into seven sub-
families (KDM2-8) based on their sequence similarities and
domain architectures.
Over-expressions of KDMs are implicated in various diseases 12-16
and are considered to promote oncogenesis in several cancer types,
mainly prostate, breast and colon.16,17 Furthermore, results from
Garcia et al. suggest that the histone demethylase activity of
KDM4C enzyme is essential for breast cancer progression given
its role in the maintenance of chromosomal stability and cell
growth, thus emphasizing it as a promising therapeutic target.18
So far, only a few KDM inhibitors are reported, most of which
function via active site metal chelation.15 The majority of known
inhibitors are analogues of the 2-oxoglutarate co-factor (2-OG),
such as N-oxalyl glycine, 2,4-pyridine dicarboxylic acid, or
consist of strong metal chelators, such as 2,2-bipyridines or
hydroxyquinolines.19 Although these compounds may be potent
biochemical inhibitors, their physicochemical properties limit
ADME properties, especially cell permeability. Two papers from
GSK reported inhibitors with enhanced properties and cell
permeability, but these compounds still have significant activity
against related histone demethylases.20-21 The combination of
moderate biochemical activity and limited cell permeability leaves
space for further investigation and design of novel compounds as
chemical tools or potential drug candidates.
The KDM4 sub-family consists of five enzymes, KDM4A,
KDM4B, KDM4C, KDM4D and KDM4E. They share extensive
homology at their N-terminal part harboring the catalytic domain.
KDM4A-C share more than 50% sequence identity and their

Toxoflavin has been reported as KDM4A22 (IC50=2.5 µM) and
KDM4C23 inhibitor (IC50= 0.5 µM). Five-fold increase in activity
is observed for KDM4C, which makes the toxoflavin a good initial
hit for the design of KDM4C inhibitors. Anticancer activity was
observed in both liquid and solid tumor cells, with a dose- and
time-dependent effect.22 However, a strong redox liability
(poisonous peroxide-generating capacity) of toxoflavin limits its
use.24
In this paper, toxoflavin was used as a starting scaffold for the
design and synthesis of a focused library diversified at position 3
(Figure 1). Effect of the different aromatic substituents on the
activity towards KDM4C enzyme; proliferation of two different
cell lines and finally influence on the redox and ADME properties
was investigated.
O
O
O
O
R
R
H
H
N
N
N
N
N
NH2
NH2
O
N
H
N
O
N
H
N
i
( )
O
N
H
Cl
ii
( )
3a-3m
1
2
O
N
O
N
N
R
R
N
N
N
N
N
iii
(
)
O
O
N
H
4a-4m
3a-3m
Scheme 1. Synthetic route for the preparation of focused
toxoflavin library; (i): methylhydrazine, EtOH, μW, 100°C, 10
min; (ii): aldehyde, EtOH, μW, 100°C, 10 min, y=26-89%; (iii):
NaNO2, acetic acid/water, 0°C to RT, y=15-78%.
2. Materials and Methods
2.1. Chemistry
O
HO
OH
O
B
O
N
O
C
Br
O
O
O
HN
General procedure: Toxoflavin based compounds were
synthesized by a procedure described by Todorovic et al.25
6-chloro-3-methyl-1H-pyrimidine-2,4-dione (1) (1 eq) was
dissolved in EtOH (2 mL) followed by addition of
methylhydrazine (2.2 eq). The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C to
obtain product 2 without isolation. Aldehyde was added to the
reaction mixture and again submitted to microwave irradiation for
10 minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3.
(ii)
(i)
H2
N
H2N
HN
O
O
O
O
N
HN
(iii)
O
N
H
HN
O
5a
Scheme 2. Synthetic route for the preparation of novel scaffold 5a;
(i): Pd(PPh3)4, Na2CO3, μW, 100°C, 30 min, y=82%; (ii): DMF,
0°C to RT, y=78%; (iii): NaOEt, EtOH, 60°C, y=73%.
O
O
N
N
N
N
Product 3 (1 eq; 1 mmol) was dissolved in a mixture of acetic acid
(7 mL) and water (0.7 mL). The reaction mixture was cooled to
0°C and sodium nitrite (1.5 eq) was added. After 6-18 hours, the
reaction mixture was diluted with water and extracted with DCM.
Organic layers were combined, dried over sodium sulphate,
filtered and evaporated under reduced pressure to obtain crude
product. Crude product was purified by flash chromatography on
a silica column. The gradient was set to 0-10% MeOH/EtOAc in
25 column volumes. Appropriate fractions were combined and
evaporated under reduced pressure to obtain cyclic product 4.
N
(i)
N
O
N
N
O
N
N
H
5b
4a
Scheme 3. Synthetic route for the preparation of novel scaffold 5b;
(i): DMF, 50°C, y=92%.
O
N
N
N
N
N
N
N
O
N
N
(i)
O
N
5c
4a
O
4
6
Scheme 4. Synthetic route for the preparation of novel scaffold 5c;
(i): NaOH, 10% in H2O, 60°C, y=63%
N
3
N
N
2
O
N
N
2.1.1. (6-[[(E)-benzylideneamino]-methyl-amino]-3-methyl-1H-
pyrimidine-2,4-dione, (3a): 6-chloro-3-methyl-1H-pyrimidine-
2,4-dione (1 eq; 100 mg) was dissolved in EtOH (2 mL) and
therafter methylhydrazine (2.2 eq; 72.06 μL) was added. The
reaction vessel was sealed and submitted to microwave irradiation
for 10 minutes at 100°C. Then benzaldehyde (1.6 eq; 105.75 mg)
was added to the reaction mixture and again submitted to
microwave irradiation for 10 minutes at 100°C. The reaction
mixture was cooled to room temperature and filtered off. The
resulting precipitate was washed with methanol, then diethyl ether
and dried in vacuo to give title product 3a in 89% yield; purity
100% by UPLC (UV, 254 nm). 1H NMR (DMSO-d6, 500 MHz, δ
1
8
Figure 1: Numeration of toxoflavin scaffold.

ppm): 3.11 (s, 3H), 3.35 (s, 3H), 5.25 (s, 1H), 7.36-7.46 (m, 3H),
7.97 (s, 1H), 7.98 (s, 2H), 10.66 (s, 1H). ESI+: (M+H)+=259.52
m/z.
chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and thereafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C.
Methyl 1-(4-formylphenyl)cyclopropanecarboxylate (1.6 eq;
203.51 mg) was added to the reaction mixture and again submitted
to microwave irradiation for 10 minutes at 100°C. The reaction
mixture was cooled to room temperature and filtered off. The
resulting precipitate was washed with methanol, then diethyl ether
and dried in vacuo to give title product 3f in 62% yield; purity
100% by UPLC (UV, 254 nm). 1H NMR (DMSO-d6, 500 MHz, δ
ppm): 1.22 (s, 2H), 1.50 (s, 2H), 3.11 (s, 3H), 3.34 (s, 3H), 3.55 (s,
3H), 5.24 (s, 1H), 7.38 (d, 2H, J=8.22 Hz), 7.90 (d, 2H, J=7.76
Hz), 7.97 (s, 1H), 10.65 (s, 1H). ESI+: (M+H)+=357.61 m/z.
2.1.7.(3-methyl-6-[methyl-[(E)-2-
2.1.2.
(4-[(E)-[methyl-(3-methyl-2,4-dioxo-1H-pyrimidin-6-
yl)hydrazono]methyl]benzonitrile, (3b): 6-chloro-3-methyl-1H-
pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in EtOH (2
mL) and therafter methylhydrazine (2.2 eq; 72.06 μL) was added.
The reaction vessel was sealed and submitted to microwave
irradiation for 10 minutes at 100°C. Then 4-formylbenzonitrile
(1.6 eq; 130.67 mg) was added to the reaction mixture and again
submitted to microwave irradiation for 10 minutes at 100°C. The
reaction mixture was cooled to room temperature and filtered off.
The resulting precipitate was washed with methanol, then diethyl
ether and dried in vacuo to give title product 3b in 72% yield;
1
purity 100% by UPLC (UV, 254 nm). H NMR (DMSO-d6, 500
MHz, δ ppm): 3.11 (s, 3H), 3.37 (s, 3H), 5.30 (s, 1H), 7.88 (d, 2H,
J=8.0 Hz), 8.01 (s, 1H), 8.19 (d, 2H, J=8.11 Hz), 10.93 (s, 1H).
ESI+: (M+H)+=284.51 m/z.
phenylethylideneamino]amino]-1H-pyrimidine-2,4-dione, (3g):
6-chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and therafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C. 2-
phenylacetaldehyde (1.6 eq; 119.72 mg) was added to the reaction
mixture and again submitted to microwave irradiation for 10
minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3g in 63% yield; purity 99% by UPLC (UV, 254 nm). 1H
NMR (DMSO-d6, 500 MHz, δ ppm): 3.09 (s, 3H), 3.15 (s, 3H),
3.71 (d, 2H, J=5.4 Hz), 5.11 (s, 1H), 7.20-7.26 (m, 1H), 7.26-7.35
(m, 5H), 10.41 (s, 1H). ESI+: (M+H)+=273.55 m/z.
2.1.3.
(6-[[(E)-(4-hydroxyphenyl)methyleneamino]-methyl-
amino]-3-methyl-1H-pyrimidine-2,4-dione, (3c): 6-chloro-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in
EtOH (2 mL) and therefer methylhydrazine (2.2 eq; 72.06 μL) was
added. The reaction vessel was sealed and submitted to microwave
irradiation for 10 minutes at 100°C. 4-hydroxybenzaldehyde (1.6
eq; 121.91 mg) was added to the reaction mixture and again
submitted to microwave irradiation for 10 minutes at 100°C. The
reaction mixture was cooled to room temperature and filtered off.
The resulting precipitate was washed with methanol, then
diethylether and dried in vacuo to give title product 3c in 61%
yield; purity 100% by UPLC (UV, 254 nm). 1H NMR (DMSO-d6,
500 MHz, δ ppm): 3.11 (s, 3H), 3.31 (s, 3H), 5.18 (s, 1H), 6.80 (d,
2H, J=8.2 Hz), 7.80 (d, 2H, J=8.2 Hz), 7.89 (s, 1H), 9.83 (s, 1H),
10.50 (s, 1H). ESI+: (M+H)+=275.51 m/z.
2.1.8. 8(6-[[(E)-1H-indol-3-ylmethyleneamino]-methyl-amino]-
3-methyl-1H-pyrimidine-2,4-dione, (3h): 6-chloro-3-methyl-1H-
pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in EtOH (2
mL) and thereafter methylhydrazine (2.2 eq; 72.06 μL) was added.
The reaction vessel was sealed and submitted to microwave
irradiation for 10 minutes at 100°C. 1H-indole-3-carbaldehyde
(1.6 eq; 169.57 mg) was added to the reaction mixture and again
submitted to microwave irradiation for 10 minutes at 100°C. The
reaction mixture was cooled to room temperature and filtered off.
The resulting precipitate was washed with methanol, then diethyl
ether and dried in vacuo to give title product 3h in 28% yield;
2.1.4.
4(6-[[(E)-(4-methoxyphenyl)methyleneamino]-methyl-
amino]-3-methyl-1H-pyrimidine-2,4-dione, (3d): 6-chloro-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in
EtOH (2 mL) and therafter methylhydrazine (2.2 eq; 72.06 μL)
was added. The reaction vessel was sealed and submitted to
microwave irradiation for 10 minutes at 100°C. 4-
methoxybenzaldehyde (1.6 eq; 135.67 mg) was added to the
reaction mixture and again submitted to microwave irradiation for
10 minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3d in 87% yield; purity 99% by UPLC (UV, 254 nm). 1H
NMR (DMSO-d6, 500 MHz, δ ppm): 3.11 (s, 3H), 3.80 (s, 3H),
5.19 (s, 1H), 6.99 (d, 2H, J=8.44 Hz), 7.90-7.95 (m, 3H), 10.60 (s,
1H). ESI+: (M+H)+=289.58 m/z.
1
purity 100% by UPLC (UV, 254 nm). H NMR (DMSO-d6, 500
MHz, δ ppm): 3.12 (s, 3H), 3.36 (s, 3H), 5.24 (s, 1H), 7.14-7.24
(m, 2H), 7.47 (d, 1H, J=7.75 Hz), 7.97 (s, 1H), 8.03 (d, 1H, J=7.55
Hz), 8.25 (s, 1H), 10.05 (s, 1H), 11.62 (s, 1H). ESI+:
(M+H)+=298.56 m/z.
2.1.9.
(3-methyl-6-[methyl-[(E)-1H-pyrrolo[2,3-b]pyridin-3-
6-
ylmethyleneamino]amino]-1H-pyrimidine-2,4-dione, (3i):
chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and therafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C. 1H-
pyrrolo[2,3-b]pyridine-3-carbaldehyde(1.6 eq; 145.63 mg) was
added to the reaction mixture and again submitted to microwave
irradiation for 10 minutes at 100°C. The reaction mixture was
cooled to room temperature and filtered off. The resulting
precipitate was washed with methanol, then diethylether and dried
in vacuo to give title product 3i in 26% yield; purity 95% by UPLC
2.1.5.
(4-[(E)-[methyl-(3-methyl-2,4-dioxo-1H-pyrimidin-6-
yl)hydrazono]methyl]benzenesulfonamide, (3e): 6-chloro-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in
EtOH (2 mL) and theafter methylhydrazine (2.2 eq; 72.06 μL) was
added. The reaction vessel was sealed and submitted to
microwave irradiation for 10 minutes at 100°C. 4-
formylbenzenesulfonamide (1.6 eq; 184.56 mg) was added to the
reaction mixture and again submitted to microwave irradiation for
10 minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3e in 34% yield; purity 100% by UPLC (UV, 254 nm). 1H
NMR (DMSO-d6, 500 MHz, δ ppm): 3.12 (s, 3H), 3.37 (s, 3H),
5.29 (s, 1H), 7.41 (s, 2H), 7.85 (d, 2H, J=8.1 Hz), 8.02 (s, 1H),
8.17 (d, 2H, J=8.1 Hz), 10.85 (s, 1H). ESI+: (M+H)+=338.51 m/z.
1
(UV, 254 nm). H NMR (DMSO-d6, 500 MHz, δ ppm): 3.12 (s,
3H), 3.36 (s, 3H), 5.23 (s, 1H), 7.23 (m, 1H), 8.11 (s, 1H), 8.22 (s,
1H), 8.29-8.32 (m, 1H), 8.41 (d, 1H, J=8.07 Hz), 10.25 (s, 1H),
12.14 (s, 1H). ESI+: (M+H)+=299.56 m/z.
2.1.10.
(3-methyl-6-[methyl-[(E)-4-
pyridylmethyleneamino]amino]-1H-pyrimidine-2,4-dione, (3j):
6-chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and therafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
2.1.6.
(methyl-[4-[(E)-[methyl-(3-methyl-2,4-dioxo-1H-
pyrimidin-6-
yl)hydrazono]methyl]phenyl]cyclopropanecarboxylate, (3f): 6-

submitted to microwave irradiation for 10 minutes at 100°C.
Pyridine-4-carbaldehyde (1.6 eq; 106.74 mg) was added to the
reaction mixture and again submitted to microwave irradiation for
10 minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3j in 45% yield; purity 99% by UPLC (UV, 254 nm). 1H
NMR (DMSO-d6, 500 MHz, δ ppm): 3.12 (s, 3H), 3.37 (s, 3H),
5.32 (s, 1H), 7.92-7.97 (m, 3H), 8.62 (d, 2H, J=4.57 Hz), 10.91 (s,
1H). ESI+: (M+H)+=260.54 m/z.
chromatography on silica column. The gradient was set to 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4a in
78% yield, purity 94% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.29 (s, 3H), 4.06 (s, 3H), 7.59 (s, 3H), 8.20
(s, 2H). 13C NMR (DMSO-d6, 150 MHz, δ ppm): 28.3, 43.3,
127.0, 127.5, 129.6, 131.7, 134.8, 149.9, 150.5, 160.2, 161.2. ESI+:
(M+H)+=270,50 m/z. UV: λmax (MeCN)=282 nm; 298 nm.
Published by Kyani et al.26
2.1.15. (4-(1,6-dimethyl-5,7-dioxo-pyrimido[5,4-e][1,2,4]triazin-
3-yl)benzonitrile, (4b): 4-[(E)-[methyl-(3-methyl-2,4-dioxo-1H-
pyrimidin-6-yl)hydrazono]methyl]benzonitrile(1 eq; 1 mmol) was
dissolved in a mixture of acetic acid (7 mL) and water (0.7 mL).
The reaction mixture was cooled to 0°C and sodium nitrite (1.5 eq)
was added. After 6 hours, the reaction mixture was diluted with
water and extracted with DCM. Organic layers were combined,
dried over sodium sulphate, filtered and evaporated under reduced
pressure to obtain crude product. Crude product was purified by
flash chromatography on silica column. The gradient was set to 0-
10% MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4b in
22% yield, purity 96% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.27 (s, 3H), 4.06 (s, 3H), 8.07 (s, 2H), 8.34
(s, 2H). ESI+: (M+H)+=295.54 m/z. UV: λmax (MeCN) =285 nm;
298 nm. Published by Todorovic et al.25
2.1.11.
(3-methyl-6-[methyl-[(E)-(1-methyl-2-oxo-4-
pyridyl)methyleneamino]amino]-1H-pyrimidine-2,4-dione, (3k):
6-chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and thereafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C. 1-
methyl-2-oxo-pyridine-4-carbaldehyde (1.6 eq; 136.65 mg) was
added to the reaction mixture and again submitted to microwave
irradiation for 10 minutes at 100°C. The reaction mixture was
cooled to room temperature and filtered off. The resulting
precipitate was washed with methanol, then diethyl ether and dried
in vacuo to give title product 3k in 41% yield; purity 94% by
1
UPLC (UV, 254 nm). H NMR (DMSO-d6, 500 MHz, δ ppm):
3.11 (s, 3H), 3.31 (s, 3H), 3.49 (s, 3H), 5.23 (s, 1H), 6.32-6.37 (m,
1H), 7.82 (d, 1H, J=6.62 Hz), 7.94 (s, 1H), 8.54 (d, 1H, J=7.16
Hz), 10.71 (s, 1H). ESI+: (M+H)+=290.53 m/z.
2.1.16.
e][1,2,4]triazine-5,7-dione,
(3-(4-hydroxyphenyl)-1,6-dimethyl-pyrimido[5,4-
(4c): 6-[[(E)-(4-
2.1.12.
(6-[[(E)-[2-(dimethylamino)phenyl]methyleneamino]-
(3l): 6-
methyl-amino]-3-methyl-1H-pyrimidine-2,4-dione,
hydroxyphenyl)methyleneamino]-methyl-amino]-3-methyl-1H-
pyrimidine-2,4-dione(1 eq; 1 mmol) was dissolved in a mixture of
acetic acid (7 mL) and water (0.7 mL). The reaction mixture was
cooled to 0°C and sodium nitrite (1.5 eq) was added. After 6
hours, the reaction mixture was diluted with water and extracted
with DCM. Organic layers were combined, dried over sodium
sulphate, filtered and evaporated under reduced pressure to obtain
chloro-3-methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was
dissolved in EtOH (2 mL) and thereafter methylhydrazine (2.2 eq;
72.06 μL) was added. The reaction vessel was sealed and
submitted to microwave irradiation for 10 minutes at 100°C. 2-
(dimethylamino)benzaldehyde (1.6 eq; 148.66 mg) was added to
the reaction mixture and again submitted to microwave irradiation
for 10 minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethylether and dried in vacuo to give title
product 3l in 38% yield; purity 100% by UPLC (UV, 254 nm). 1H
NMR (DMSO-d6, 500 MHz, δ ppm): 2.70 (s, 6H), 3.11 (s, 3H),
3.36 (s, 3H), 5.26 (s, 1H), 7.06 (t, 1H, J=7.56 Hz), 7.13 (d, 1H,
J=8.14 Hz), 7.34 (t, 1H, J=7.85 Hz), 8.03 (s, 1H), 8.19 (d, 1H,
J=7.85 Hz), 10.56 (s, 1H). ESI+: (M+H)+=302.59 m/z.
crude product.
Crude product was purified by flash
chromatography on silica column. Gradient was set 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4c in
28% yield, purity 85% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.27 (s, 3H), 4.01 (s, 3H), 6.94 (d, 2H,
J=7.83 Hz), 8.03 (d, 2H, J=8.17 Hz), 10.17 (s, 1H). ESI+:
(M+H)+=286.51 m/z. UV: λmax (MeCN) =227 nm. Published by
Zeller et al.27
2.1.13.
(6-[[(E)-(4-fluorophenyl)methyleneamino]-methyl-
amino]-3-methyl-1H-pyrimidine-2,4-dione, (3m): 6-chloro-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 100 mg) was dissolved in
EtOH (2 mL) and thereafter methylhydrazine (2.2 eq; 72.06 μL)
was added. The reaction vessel was sealed and submitted to
2.1.17.
e][1,2,4]triazine-5,7-dione,
(3-(4-methoxyphenyl)-1,6-dimethyl-pyrimido[5,4-
(4d): 6-[[(E)-(4-
methoxyphenyl)methyleneamino]-methyl-amino]-3-methyl-1H-
pyrimidine-2,4-dione (1 eq; 1 mmol) was dissolved in a mixture
of acetic acid (7 mL) and water (0.7 mL). The reaction mixture
was cooled to 0°C and sodium nitrite (1.5 eq) was added. After 6
hours, the reaction mixture was diluted with water and extracted
with DCM. Organic layers were combined, dried over sodium
sulphate, filtered and evaporated under reduced pressure to obtain
crude product.
chromatography on silica column. Gradient was set 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4d in
15% yield, purity 92% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.27 (s, 3H), 3.85 (s, 3H), 4.03 (s, 3H), 7.14
(d, 2H, J=8.65 Hz), 8.13 (d, 2H, J=8.37 Hz). 13C NMR (DMSO-
d6, 150 MHz, δ ppm): 28.7, 43.2, 55.9, 115.2, 125.5, 128.8, 146.8,
149.7, 151.7, 154.5, 159.5, 162.3. ESI+: (M+H)+=300.53 m/z. UV:
λmax (MeCN) =305 nm. Published by Zeller et al.27
2.1.18. 4-(1,6-dimethyl-5,7-dioxo-pyrimido[5,4-e][1,2,4]triazin-
3-yl)benzenesulfonamide, (4e): 4-[(E)-[methyl-(3-methyl-2,4-
dioxo-1H-pyrimidin-6-
microwave irradiation for 10 minutes at 100°C.
4-
fluorobenzaldehyde (1.6 eq; 123.68 mg) was added to the reaction
mixture and again submitted to microwave irradiation for 10
minutes at 100°C. The reaction mixture was cooled to room
temperature and filtered off. The resulting precipitate was washed
with methanol, then diethyl ether and dried in vacuo to give title
product 3m in 60% yield; purity 100% by UPLC (UV, 254 nm).
1H NMR (DMSO-d6, 500 MHz, δ ppm): 3.11 (s, 3H), 3.34 (s, 3H),
5.23 (s, 1H), 7.23-7.29 (m, 2H), 7.98 (s, 1H), 8.06 (m, 2H), 10.76
(s, 1H). ESI+: (M+H)+=277.53 m/z.
Crude product was purified by flash
2.1.14.
(1,6-dimethyl-3-phenyl-pyrimido[5,4-e][1,2,4]triazine-
5,7-dione, (4a): 6-[[(E)-benzylideneamino]-methyl-amino]-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 1 mmol) was dissolved in
a mixture of acetic acid (7 mL) and water (0.7 mL). The reaction
mixture was cooled to 0°C and sodium nitrite (1.5 eq) was added.
After 6 hours, the reaction mixture was diluted with water and
extracted with DCM. Organic layers were combined, dried over
sodium sulphate, filtered and evaporated under reduced pressure
to obtain crude product. Crude product was purified by flash
yl)hydrazono]methyl]benzenesulfonamide(1 eq; 1mmol) was

dissolved in a mixture of acetic acid (7 mL) and water (0.7 mL).
The reaction mixture was cooled to 0°C and sodium nitrite (1.5 eq)
was added. After 6 hours, the reaction mixture was diluted with
water and extracted with DCM. Organic layers were combined,
dried over sodium sulphate, filtered and evaporated under reduced
pressure to obtain crude product. Crude product was purified by
flash chromatography on silica column. The gradient was set to 0-
10% MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4e in
17% yield, purity 89% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.29 (s, 3H), 4.07 (s, 3H), 7.54 (s, 2H), 8.03
(d, 2H, J=8.30 Hz), 8.36 (d, 2H, J=8.30 Hz). ESI+:
(M+H)+=349.54 m/z. UV: λmax (MeCN) =234 nm; 296 nm.
2.1.22. (5-hydroxy-2,8-dimethyl-4-(1-methyl-2-oxo-4-pyridyl)-
2,3,5lambda5,8,10-pentazabicyclo[4.4.0]deca-1(10),3,5-triene-
7,9-dione, (4k): 3-methyl-6-[methyl-[(E)-(1-methyl-2-oxo-4-
pyridyl)methyleneamino]amino]-1H-pyrimidine-2,4-dione (1 eq;
1 mmol) was dissolved in a mixture of acetic acid (7 mL) and water
(0.7 mL). The reaction mixture was cooled to 0°C and sodium
nitrite (1.5 eq) was added. After 6 hours, the reaction mixture was
diluted with water and extracted with DCM. Organic layers were
combined, dried over sodium sulphate, filtered and evaporated
under reduced pressure to obtain crude product. Crude product
was purified by flash chromatography on silica column. The
gradient was set to 0-10% MeOH/EtOAc. Appropriate fractions
were combined and evaporated under reduced pressure to obtain
title product 4k in 32% yield, purity 97% by UPLC (UV, 254 nm).
1H NMR (DMSO-d6, 500 MHz, δ ppm): 3.17 (s, 3H), 3.50 (s, 3H),
3.84 (s, 3H), 6.39 (m, 1H), 7.68 (d, 2H, J=7.07 Hz), 7.99 (d, 2H,
J=6.39 Hz). ESI+: (M+H)+=317.54 m/z. UV: λmax (MeCN) =265
nm; 321 nm.
2.1.19.
e][1,2,4]triazin-3-yl)phenyl]cyclopropanecarboxylate,
methyl 1-[4-[(E)-[methyl-(3-methyl-2,4-dioxo-1H-pyrimidin-6-
yl)hydrazono]methyl]phenyl]cyclopropanecarboxylate(1 eq;
(methyl-1-[4-(1,6-dimethyl-5,7-dioxo-pyrimido[5,4-
(4f):
1
mmol) was dissolved in a mixture of acetic acid (7 mL) and water
(0.7 mL). The reaction mixture was cooled to 0°C and sodium
nitrite (1.5 eq) was added. After 6 hours, the reaction mixture was
diluted with water and extracted with DCM. Organic layers were
combined, dried over sodium sulphate, filtered and evaporated
under reduced pressure to obtain crude product. Crude product
was purified by flash chromatography on silica column. The
gradient was set to 0-10% MeOH/EtOAc. Appropriate fractions
were combined and evaporated under reduced pressure to obtain
title product 4f in 21% yield, purity 99% by UPLC (UV, 254 nm).
1H NMR (DMSO-d6, 500 MHz, δ ppm): 1.27 (s, 2H), 1.54 (s, 2H),
3.28 (s, 3H), 3.57 (s, 3H), 4.05 (s, 3H), 7.55 (d, 2H, J=7.89 Hz),
8.12 (d, 2H, J=7.89 Hz). ESI+: (M+H)+=368.60 m/z. UV: λmax
(MeCN)=282 nm; 304 nm.
2.1.23
.(3-[2-(dimethylamino)phenyl]-1,6-dimethyl-
(4l): 6-[[(E)-[2-
pyrimido[5,4-e][1,2,4]triazine-5,7-dione,
(dimethylamino)phenyl]methyleneamino]-methyl-amino]-3-
methyl-1H-pyrimidine-2,4-dione (1 eq; 1 mmol) was dissolved in
a mixture of acetic acid (7 mL) and water (0.7mL). The reaction
mixture was cooled to 0°C and sodium nitrite (1.5 eq) was added.
After 6 hours, the reaction mixture was diluted with water and
extracted with DCM. Organic layers were combined, dried over
sodium sulphate, filtered and evaporated under reduced pressure
to obtain crude product. Crude product was purified by flash
chromatography on silica column. The gradient was set to 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4l in 43%
yield, purity 91% by UPLC (UV, 254 nm). 1H NMR (DMSO-d6,
500 MHz, δ ppm): 2.55 (s, 3H), 2.65 (s, 3H), 3.27 (s, 3H), 4.01 (s,
3H), 6.97-7.05 (m, 1H), 7.08-7.15 (m, 1H), 7.38-7.48 (m, 2H).
ESI+: (M+H)+=313.57 m/z. UV: λmax (MeCN) =239 nm.
2.1.20.
5,7-dione,
(3-benzyl-1,6-dimethyl-pyrimido[5,4-e][1,2,4]triazine-
(4g): 3-methyl-6-[methyl-[(E)-2-
phenylethylideneamino]amino]-1H-pyrimidine-2,4-dione (1 eq; 1
mmol) was dissolved in a mixture of acetic acid (7 mL) and water
(0.7 mL). The reaction mixture was cooled to 0°C and sodium
nitrite (1.5 eq) was added. After 6 hours, the reaction mixture was
diluted with water and extracted with DCM. Organic layers were
combined, dried over sodium sulphate, filtered and evaporated
under reduced pressure to obtain crude product. Crude product
was purified by flash chromatography on silica column. The
gradient was set to 0-10% MeOH/EtOAc. Appropriate fractions
were combined and evaporated under reduced pressure to obtain
title product 4g in 38% yield, purity 85% by UPLC (UV, 254 nm).
1H NMR (DMSO-d6, 500 MHz, δ ppm): 3.23 (s, 3H), 3.93 (s, 3H),
4.22 (s, 1H), 4.46 (s, 1H), 7.20-7.34 (m, 5H). 13C NMR (DMSO-
d6, 150 MHz, δ ppm): 28.2, 28.7, 42.6, 127.1, 129.0, 129.5, 138.2,
150.0, 150.4, 155.8, 161.1, 165.0. ESI+: (M+H)+=284.51 m/z. UV:
λmax (MeCN) =264 nm. Published by Kyani et al.26
2.1.24.
e][1,2,4]triazine-5,7-dione,
(3-(4-fluorophenyl)-1,6-dimethyl-pyrimido[5,4-
(4m): 6-[[(E)-(4-
fluorophenyl)methyleneamino]-methyl-amino]-3-methyl-1H-
pyrimidine-2,4-dione (1 eq; 1 mmol) was dissolved in a mixture
of acetic acid (7 mL) and water (0.7 mL). The reaction mixture
was cooled to 0°C and sodium nitrite (1.5 eq) was added. After 6
hours, the reaction mixture was diluted with water and extracted
with DCM. Organic layers were combined, dried over sodium
sulphate, filtered and evaporated under reduced pressure to obtain
crude product.
Crude product was purified by flash
chromatography on silica column. The gradient was set to 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporated under reduced pressure to obtain title product 4m in
62% yield, purity 100% by UPLC (UV, 254 nm). 1H NMR
(DMSO-d6, 500 MHz, δ ppm): 3.28 (s, 3H), 4.05 (s, 3H), 7.39-
7.48 (m, 2H), 8.19-8.28 (m, 2H). ESI+: (M+H)+=288.51 m/z. UV:
λmax (MeCN) =285 nm. Published by Todorovic et al.25
2.1.21.(1,6-dimethyl-3-(4-pyridyl)pyrimido[5,4-e][1,2,4]triazine-
5,7-dione,
(4j):
3-methyl-6-[methyl-[(E)-4-
pyridylmethyleneamino]amino]-1H-pyrimidine-2,4-dione (1 eq; 1
mmol) was dissolved in a mixture of acetic acid (7 mL) and water
(0.7 mL). The reaction mixture was cooled to 0°C and sodium
nitrite (1.5 eq) was added. After 6 hours, the reaction mixture was
diluted with water and extracted with DCM. Organic layers were
combined, dried over sodium sulphate, filtered and evaporated
under reduced pressure to obtain crude product. Crude product
was purified by flash chromatography on silica column. The
gradient was set to 0-10% MeOH/EtOAc. Appropriate fractions
were combined and evaporated under reduced pressure to obtain
title product 4j in 26% yield, purity 91% by UPLC (UV, 254 nm).
1H NMR (DMSO-d6, 500 MHz, δ ppm): 3.29 (s, 3H), 4.07 (s, 3H),
8.08 (s, 2H), 8.82 (s, 2H). ESI+: (M+H)+=271,53 m/z. UV: λmax
(MeCN) =284 nm.
2.1.25. 3-methyl-6-phenyl-1H-quinazoline-2,4-dione, (5a): Step
1: The reaction vessel was charged with methyl 2-amino-5-bromo-
benzoate (200 mg), phenylboronic acid (106 mg), Na2CO3 (276.40
mg) and Pd(PPh3)4 (100.46 mg). 1,4-dioxane (3 mL) and water (1
mL) were added. The reaction vessel was sealed, vacuumed,
backfilled with argon and submitted to microwave irradiation for
30 minutes at 100°C. The reaction mixture was diluted with water
and extracted with DCM. Organic layers were combined, dried
over sodium sulphate, filtered and evaporated under reduced
pressure to obtain crude product. Crude product was purified by
flash chromatography on silica column. The gradient was set to 0-
50% EtOAc/cyclohexane in 25 CVs. Appropriate fractions were
combined and evaporated under reduced pressure to obtain
intermediate 5i; methyl 2-amino-5-phenyl-benzoate in 82% yield,

purity 95% by UPLC (UV, 254 nm). Step 2: Methyl 2-amino-5-
phenyl-benzoate (5i) (200 mg) was dissolved in dry DMF (2 mL).
Inhibition activity against KDM4C enzyme was tested for newly
synthetized compounds using the BPS Bioscience (San Diego,
USA) biochemical assay kit. The KDM4C chemiluminescent
assay kit was designed to measure activity of the KDM4C using a
highly specific antibody that recognizes the demethylated
substrate. The experimental procedure comprises of three steps on
a microtiter plate to detect methyltransferase: 1) KDM4C enzyme
was incubated with a sample containing assay buffer for one hour;
2) primary antibody was added; 3) the plate was treated with an
HRP-labelled secondary antibody followed by addition of the HRP
substrate to produce chemiluminescence that was measured using
a chemiluminescence reader (EnVision, Perkin Elmer). Newly
synthetized compounds were tested at 10 µM in primary screening,
and in dose-response range in secondary screening, in order to
produce IC50 values (10 µM starting compound concentration, 1:5
dilution), and results were presented as a IC50 values of the
KDM4C enzyme. 2,4-pyridinedicarboxylic acid was used as a
control compound with literature known activity. Assays for
determination of percentage of inhibition and IC50 values were
performed following manufacturers protocol.30
The
reaction
mixture
was
cooled
to
0°C,
methylimino(oxo)methane (106 mg) was added and allowed to
slowly warm to room temperature. After 3 hours, the reaction was
diluted with water and extracted with EtOAc. Organic layers were
combined, dried over sodium sulphate, filtered and evaporated
under reduced pressure to obtain crude product. Crude product
was purified by flash chromatography on silica column. The
gradient was set to 0-50% EtOAc/cyclohexane in 25 CVs.
Appropriate fractions were combined and evaporated under
reduced pressure to obtain intermediate 5ii methyl 2-
(methylcarbamoylamino)-5-phenyl-benzoate in 77% yield, purity
75% by UPLC (UV, 254 nm). Step 3: Methyl 2-
(methylcarbamoylamino)-5-phenyl-benzoate (5ii) (20 mg) was
dissolved in NaOEt, 20% solution in ethanol. Reaction mixture
was heated to 60°C. After 3 hours reaction mixture was cooled to
room temperature, diluted with water and extracted with DCM.
Organic layers were combined, dried over sodium sulphate,
filtered and evaporated under reduced pressure to obtain crude
product. Crude product was purified by flash chromatography
using silica column. Gradient was set 0-50% EtOAc/cyclohexane
in 25 CVs. Appropriate fractions were combined and evaporated
under reduced pressure to obtain product 5a, 3-methyl-6-phenyl-
1H-quinazoline-2,4-dione in 73% yield, purity 95% by UPLC
2.3. Influence on cell proliferation of HepG2 and A549 cell
lines
1
(UV, 254 nm). H NMR (DMSO-d6, 500 MHz, δ ppm): 3.28
(s,3H), 7.27 (d, 1H, J=5.06Hz), 7.37 (t, 1H, J=7.45Hz), 7.48 (t,
2H, J=7.45 Hz), 7.68 (d, 2H, J=3.99 Hz), 7.98 (d, 1H, J=4.07 Hz),
8.14 (s, 1H), 11.51 (s, 1H). 13C NMR (DMSO-d6, 150 MHz, δ
ppm): 27.6, 114.6, 116.4, 125.0, 126.8, 128.0, 129.6, 133.7, 134.9,
139.1, 139.2, 150.7, 162.6. ESI+: (M+H)+=253.14 m/z. UV: λmax
(MeCN) =236 nm; 268 nm.
2.3.1. HepG2 cell line
Proliferation of human Caucasian hepatocyte carcinoma (HepG2)
cell line was investigated after 24 hour incubation with compounds
by a modified version of Mosmann.31. A mother plate (Storplate-
384- deep-well-V plate, Perkin Elmer) with 1 : 2 serial dilutions (9
concentrations; starting from 25 µM concentration) of the
compounds in pure DMSO was prepared in duplicate from
compound stock solutions (1 mg/mL). 500 nL of compounds
solution was transferred from the mother plate to the test plate by
using a nanovolume liquid handling instrument Mosquito (3019-
0002, TTP Labtech). Cells were distributed in 384-well plates, at
concentrations of 2 × 104 cells per well, in 50 μL of cell culture
medium. Control conditions (cells/media) were spiked with equal
final DMSO concentrations of 1% DMSO. Cells were incubated
for 24 h at 37°C, 5% CO2 and 95% humidity. Cell growth was
estimated by the MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-
carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium]
assay. After adding 5 μL of MTS, plates were incubated for 2 hours
at 37°C. Absorbance at 492 nm was measured with a plate reader.
The reference control compound for MTS assay was
staurosporine, which showed acceptable standard value for this
assay. Data was evaluated using predefined IC50 equation from
GraphPad Prism program, version 5.3 for Windows (GraphPad
Software, San Diego, CA).
2.1.26. 6-methyl-3-phenyl-1H-pyrimido[5,4-e][1,2,4]triazine-5,7-
dione, (5b): Method described by Nagamatsu et al.28 1,6-dimethyl-
3-phenyl-pyrimido[5,4-e][1,2,4]triazine-5,7-dione
(4a)
was
dissolved in DMF and heated to 50°C. After 3 hours, the reaction
mixture was cooled to room temperature, diluted with water and
extracted with DCM. Organic layers were combined, dried over
sodium sulphate, filtered and evaporated under reduced pressure
to obtain crude product. Crude product was purified by flash
chromatography on silica column. The gradient was set to 0-10%
MeOH/EtOAc. Appropriate fractions were combined and
evaporeted under reduced pressure to obtain title product 5b in
92% yield, purity 100% by UPLC(UV, 254 nm). 1H NMR
(DMSO-d6, 500 MHz, δ ppm): 3.29 (s, 3H), 7.61 (m, 3H), 8.40
(m, 2H), 12.82 (s, 1H). ESI+: (M+H)+=256.08 m/z. UV: λmax
(MeCN) =273 nm; 302 nm. Published by Nagamatsu et al.28
2.1.27.1,5-dimethyl-3-phenyl-imidazo[4,5-e][1,2,4]triazin-6-one
(5c): Method described by Yoneda et al.29 1,6-dimethyl-3-phenyl-
pyrimido[5,4-e][1,2,4]triazine-5,7-dione (4a) was dissolved in
NaOH, 10% solution in water (1 mL). The reaction mixture was
heated to 60°C, stirred and after 2 hours the reaction mixture was
cooled to room temperature, diluted with water and extracted with
DCM. Organic layers were combined, dried over sodium sulphate,
filtered and evaporated under reduced pressure to obtain crude
product. Crude product was purified by flash chromatography
2.3.2. A549 cell line
Proliferation of adenocarcinomic human alveolar basal epithelial
(A549) cell line was investigated after 72 hours of incubation with
compounds. Compounds were tested in eight consecutive three-
fold dilutions starting from 30 µM in duplicates placed on two
different plates. Side control conditions (cells/media) were spiked
with equal final DMSO concentrations of 0.3% DMSO. 100 µL of
A549 cells was added, in concentration of 1 × 104 cells per well.
Cells were incubated for 72 h at 37°C, 5% CO2 and 95% humidity.
Determination of proliferation inhibition was performed using
Alamar blue (Invitrogen, Cat. No. DAL1100), according to the
manufacturer instructions. After 72 hours, 10 µL of Alamar blue
per well is added. Cells were then incubated 4 hours at 37ºC.
Fluorescence was measured using the EnVision2104
using silica column.
The gradient was set to 0-10%
MeOH/EtOAc. The appropriate fractions were combined and
evaporeted under reduced pressure to obtain title product 5c in
63% yield, purity 97% by UPLC (UV, 254 nm). 1H NMR (DMSO-
d6, 500 MHz, δ ppm): 3.33 (s, 3H), 3.98 (s, 3H), 7.53 (m, 3H),
8.21 (m, 2H). ESI+: (M+H)+=242.11 m/z. UV: λmax (MeCN) =280
nm; 334 nm. Published by Yoneda et al. 29
2.2. Biological activity evaluation in vitro

(PerkinElmer, 2104-0020). Data calculations were performed
using Microsoft Office Excel software. IC50 values of compounds
were determined by plotting % of inhibition against the Log10
compound test concentration using a four-parametric sigmoidal
curve fit with a variable slope in GraphPad Prism version 5.3 for
Windows (GraphPad Software, San Diego, CA).
All samples were quantified using tandem mass spectrometry
coupled to liquid chromatography. Samples were analysed on a
Sciex API4000 Triple Quadrupole Mass Spectrometer (Sciex,
Division of MDS Inc., Toronto, Canada) coupled to a Shimadzu
Nexera X2 UHPLC frontend (Kyto, Japan). Samples were injected
onto a UHPLC column (HALO2 C18, 2.1x20 mm, 2 µm or Luna
Omega 1.6 µm Polar C18 100A, 30x2.1 mm) and eluted with a
gradient at 50°C. The mobile phase was composed of
acetonitrile/water mixture (9/1, with 0.1 % formic acid) and 0.1 %
formic acid in deionized water. The flow rate was 0.7 mL/min and
under gradient conditions, leading to a total run time of 1.5–2 min.
Positive ion mode with turbo spray, an ion source temperature of
550°C and a dwell time of 150 ms were utilized for mass
spectrometric detection. Quantitation was performed using
multiple reaction monitoring (MRM) at the specific transitions for
each compound.
2.4. MDCKII-MDR1 permeability assay
MDCKII-hMDR1 cells were obtained from Solvo Biotechnology,
Hungary. DMEM, Fetal bovine serum, Glutamax-100,
Antibiotic/Antimycotic, DMSO, Dulbecco’s phosphate buffer
saline, MEM Non-essential amino acids were purchased from
Sigma (St. Louis, MO, USA).
Bi-directional permeability and P-glycoprotein substrate
assessment were investigated in Madin-Darby canine epithelial
cells with over-expressed human MDR1 gene (MDCKII-MDR1),
coding for P-glycoprotein. Experimental procedures, as well as
cell culture conditions, were the same as previously described.32
Briefly, compounds (10 µM, 1% DMSO v/v) in duplicate were
incubated at 37°C for 60 min with cell monolayer on 24-well
Millicell inserts (Millipore, Burlington, MA, USA) without and
with the P-glycoprotein inhibitor Elacridar ( 2 µM, International
Laboratory, USA). Inhibition of P-glycoprotein was verified by
amprenavir (Moravek Biochemicals Inc, Brea, CA, USA) and
monolayer integrity by Lucifer yellow (Sigma, St. Louis, MO,
USA). Compound concentrations were measured by LC-MS/MS
and Lucifer yellow was measured on an Infinite F500 (Tecan,
Männedorf, CH) using excitation of 485 nm and emission of 530
nm.
2.7. Computational methods
Known KDM4C inhibitors were retrieved from the Chembl data
base.33 Field-based alignment of representative structures has been
performed by the Forge software34 from the Cresset set of tools.
Activity atlas34 was subsequently used to build up the qualitative
generalized pharmacophoric model.
Available crystal structures of protein KDM4C and co-crystalized
KDM4C protein with various inhibitors were downloaded from
the Protein data bank.35 Representative ligands from the 2XML,
5FJH and 5FJK X-ray structure were used. The ligand docking
studies were carried out using Glide docking protocol36-37 within
Schroedinger software suite38 with and without extra precision
(SP, XP). Docking was performed using metal-ligand constrains
and octahedral coordination. Binding poses were refined and
binding energies estimated using MM-GBSA39-40 protocol and
OPLS3 force-field with flexible residues distance being 6 Å.
Applicability of MMGBSA and QM/MM methods for this class of
enzymes was investigated on KDM4A as a case study. Seven
KDM4A inhibitors were selected for which activities as well as
co-crystal structures were publicly available (5F2S, 5F2W, 5F5I,
5F32, 5F3I, 5F3E, 5F39).41
2.5. Metabolic stability
Mouse liver microsomes were obtained from Corning Life
Sciences (Corning, USA). DMSO, nicotinamide adenine
dinucleotide phosphate (NADP), glucose-6-phosphate, glucose-6-
phosphate dehydrogenase, magnesium chloride, propranolol,
caffeine, diclofenac, phosphate buffer saline (PBS) were
purchased from Sigma (St. Louis, MO, USA). Acetonitrile (ACN)
and methanol (MeOH) were obtained from Merck (Darmstadt,
Germany). Testosterone was purchased from Steraloids (Newport,
RI, USA).
Quantum Mechanics
/
Molecular Mechanics (QM/MM)
calculations were done with QSite within the Schroedinger
software.38 KDM4A complexes were treated with the Zn (II) and
Fe (II) in the metal complexes, Zn being in the single spin state
and Fe(II) was investigated in the both singlet and quintet spin
states. KDM4C complexes were investigated with the Ni(II) in the
triplet state. QM part of the metal complex was treated with
B3LYP-D3 functional and 6-31+(2d,p) basis set42-46 for H,C,O and
N atoms and lav3p++(2d,p) basis set for metal ions.47 MM part
was treated with OPLS_2005 force field.48 The NBO analysis49 has
been done on the optimized geometries. The second order
perturbation theory analysis of the Fock matrix in the NBO basis
was used to estimate strength of metal ligand donor-acceptor
interactions.
Metabolic stability was assessed in mouse liver microsomes.
Compounds (final concentration of 1 µM, 0.03% DMSO v/v) were
incubated in duplicate in phosphate buffer (50 mM, pH 7.4) at
37°C together with mouse liver microsomes in the absence and
presence of the NADPH cofactor (0.5 mM nicotinamide adenine
dinucleotide phosphate, 5 mM glucose-6-phosphate, 1.5 U/mL
glucose-6-phosphate dehydrogenase and 0.5 mM magnesium
chloride). Incubation and sampling was performed on a Freedom
EVO 200 (Tecan, Männedorf, CH) at 0.3, 10, 20, 30, 45 and 60
min. The reaction was quenched using 3 volumes of a mixture of
ACN/MeOH (2:1) containing internal standard (diclofenac),
centrifuged and supernatants were analyzed using LC-MS/MS.
Metabolic activity of microsomes was verified by simultaneous
analysis of several controls including testosterone, propranolol and
caffeine. The in vitro half-life (t1/2) was calculated using GraphPad
Prism non-linear regression of % parent compound remaining
versus time. In vitro clearance, expressed as µl/min/mg, was
estimated from the in vitro half-life (t1/2), and normalized for the
protein amount in the incubation mixture and assuming 52.5 mg
of protein per gram of liver.
Electronic structure and orbital levels were calculated from the
optimized geometries at the M06-2x/6-31G(2d,p)+ level. M06-2x
functional was used as it was previously shown that it reproduces
well ionization energies and electron affinities50-52
3. Results and discussion
Compounds for biological evaluation were synthesized in 3
campaigns as described in section 2.1. In the first campaign linear
2.6. LC-MS/MS analysis

molecules 3a-3n were produced. The second campaign gave their
bi-cyclic pairs based on toxoflavin scaffold (4a-4k). In the third
campaign a scaffold hopping exercise was done in order to directly
modify bi-cyclic toxoflavin scaffold (Table 1b, compounds 5a-
5c).
QM/MM methods, to study geometries and binding energies for
KDM4 class of proteins. As the similarity of the catalytic domains
of KDM4A and KDM4C is 84%, we assumed that results for
KDM4A inhibitors will be transferable to KDM4C inhibitors as
well.
Modelling of the metalloenzymes and metal-ligand interactions is
highly challenging due to the inadequacies of the force-field to
properly account for the coordination bonds and the need to treat
the system, or at least the part of the system by quantum-chemical
methods.53-55 Furthermore, an optimal choice of the QM method
used in the QM/MM approach is also important in order to obtain
high-quality data.39-41 Current DFT functionals are in general
improvements over HF (Hartree-Fock), LDA (local density) and
even MP2 calculations, however these improvements are not
systematic and often good agreement with experimental data has
to be tested empirically for different systems of interest.56-57 It has
been repeatedly reported that B3LYP-D3 functional is the optimal
choice for the exploration of metalloenzyme-ligand complexes53-
54, 58-59 and it has been used throughout this study.
O
N
O
N
N
R
R
N
N
N
N
O
O
N
N
4a-4m
3a-3m
R
Compound
3a, 4a
3b, 4b
3c, 4c
Compound
3h
R
N
N
3i
N
N
3j, 4j
3k, 4k
3l, 4l
O
N
O
In this study, the QM part was composed from the metal ion first
coordination sphere including crystal water molecules, His278,
triade His190, Thr191, Glu192 and the ligand molecule.
3d, 4d
3e, 4e
O
N
N
S
X-ray structures were used as a starting point for the MMGBSA
and QM/MM optimizations and the interaction energies were
calculated as a difference between energies of a complex and
energies of a free ligand and the protein. Since QM/MM
optimization calculations are time-demanding, especially in the
case of slow convergent calculations for flat potential energy
surfaces and higher spin states, single point QM/MM calculations
have also been performed on both MMGBSA and x-ray structures
to investigate applicability of computationally cheaper workflow.
Predicting experimental binding energies or inhibitor potential of
ligands is a highly challenging task. It has been shown previously
that experimental activity data can be estimated reasonably well as
a linear combination of QM/MM interaction energies and
desolvation-characterizing changes in the solvent-accessible
surface areas.55 Authors have used linear interaction energy
approximation and calculated binding free energies as a linear
combination of the differences Δ in the van der Waals energies,
electrostatic energies and the solvent accessible surface areas
(SASA) obtained from molecular dynamics (MD) or Monte Carlo
simulations approximatively describing desolvation penalties and
hydrophobic effects as a result of amount of SASA buried upon
complexation:
S
N
S
O
3f, 4f
3m, 4m
O
F
3g, 4g
Table 1a. Synthesized compounds3a-3m and 4a-4m.
Compound
5a
5b
5c
〈
〉
〈
〉
∆푮풃 = 휶 × 횫 푬풗풅푾 + 휷 × 횫 푬풆풍 + 휸 × 횫〈푺푨푺푨〉 + 휿
O
N
O
N
N
N
N
N
N
N
N
N
O
N
N
O
O
A similar method for the treatment of hydration of more complex
molecules required the use H-bond donor and acceptor counts.60 In
order to explore if further simplifications could be used, while still
obtaining at least semi-quantitative agreement with the
experimental activities, we assumed that the protein desolvation
energies are similar for all examined complexes and will be
accounted for by regression coefficients. Conformational changes
of the ligands upon binding are also assumed to be small due to the
rigidity of the compounds and SASA was approximated with the
SASA for the free compound in the water. QM/MM energies were
combined linearly with the desolvation contribution approximated
by different descriptors describing ligand solvent-accessible
Table 1b. Synthesized compounds5a-5c.
3.1. Calibrating computational workflow – KDM4A case study
A recently published set of highly homologous KDM4A protein
co-crystal complexes has been used in order to explore
applicability of Glide docking protocol, as well as MMGBSA and

surface areas; SASA (solvent-accessible surface area), FOSA
(Hydrophobic component of the SASA), FISA (Hydrophilic
component of the SASA), PISA (π component of the SASA) HBA,
HBD, dipole, volume and globularity.61 Experimental activities,
expressed as pIC50 values, were correlated with QM/MM energies
pyrazolo[1,5-a]pyrimidine derivative from the 5FJK fills one
coordination position and therefore contains two water molecules
in the metal coordination sphere. Additional H-bond interactions
are formed with the side chains on the opposite side of the binding
pocket: Tyr134, Lys208 and Asn200 as shown on the Figure 5.
Lys243 is displaced and the salt bridge between Lys243 and
Asp137 is broken in the 5FJK structure, in comparison with 5FJH
and 2XML structures, keeping the entrance channel open. Also,
Tyr 134 is displaced to the bottom of the binding pocket enabling
ligand to position across the binding site, form pi-pi stacking with
the phenyl ring of Tyr134 and H-bonding interactions with
Lys208. Hydroxy groups of Ser198 and Ser290 are assumed to be
differently oriented due to the H-bond network with coordinated
water molecules. Available x-ray structures therefore demonstrate
different binding coordinations and significant protein
conformational changes due to induced fit of diverse inhibitor
chemotypes.
62
and QikProp descriptors described above using PLS (partial-
least square method).63 Good correlation has been obtained for the
set of 7 KDM4A-ligand complexes as shown on Figure 2. Values
of descriptors used in the model development are given in the
Table S1 (Supplementary data). Correlation of the QM/MM
interaction energies with pIC50 values were R2=0.73 and 0.71 for
the SP and fully optimised geometries.
H2O
His190
Inh/H2O
Ni2+
Inh
His278
Glu192
Figure 3. Coordination of Ni2+ ion forming octahedral
coordination.
3.3. Field-based analysis of known KDM4C inhibitors
Known KDM4C inhibitors were retrieved from the Chembl
database.33 Based on the activity range and structural diversity, 40
compounds were selected for field-based alignment with ligands
from the 5FJH and 5FJK X-ray structures. Subsequently, a field-
based pharmacophoric model was built using the Activity atlas
tool from Cresset suite of software. These results were used
together with the SBDD information in order to design a focused
library based on the toxoflavin scaffold. As shown in Figure 4., a
major electrostatic contribution for the KDM4C inhibitors are
negative electrostatic fields from the nitrogen lone pairs
contributing to the metal-ion coordination, as well as electron rich
groups (carboxylic acid in the case of 5FJH ligand) that form a
strong H-bond network with Tyr134, Lys208 and Asn200. This
was also seen from the available x-ray structures. In addition, the
average shape of actives was calculated. The toxoflavin scaffold
was aligned with the reference 5FJH ligand to explore possible
binding modes and further modifications. Field-based analysis
showed that toxoflavin fits well within the generalized
pharmacophore model and that position C3 is suitable for further
diversification. Compound 4a was used as representative molecule
to explore if the aromatic ring could be accommodated within the
average shape of actives. It was clear that the alignment is good,
however, both monodentate and bidentate binding will have to
compromise direct metal coordination and fit into the region
populated by investigated active compounds. Therefore, induced
fit and the side-chain displacement of some amino-acids in the
binding pocket could be expected in order to accommodate the
designed compounds, as already observed in the 5FJK structure as
well as some of the available KDM4A co-crystal structures.
Figure 2. Correlation between experimental and predicted pIC50
values for a set of KDM4A inhibitors.
The strength of the ligand coordination to the metal ion was
additionally quantified in terms of the second order perturbation
theory analysis of the Fock matrix in the NBO basis.
Delocalization of the N(4) lone-pair electron density into the
unfilled valence-shell nonbonding orbitals of the metal ion ranges
from 44-64 kcal/mol for 5F2S, 5F2W, 5F51 and 5F32 ligands as
shown in Table S1.
Results for the KDM4A case study encouraged us to proceed with
the QM/MM calculations of the interaction energies for the
designed set of potential KDM4C inhibitors based on the
toxoflavin scaffold.
3.2. Key features of KDM4C active site
KDM4C is a metalloprotein containing Fe(II) ion which is
substituted with Ni(II) in co-crystallization studies.64 It belongs to
alpha-ketoglutarate-dependent hydroxylase superfamily and
requires α-ketoglutarate as a cofactor to exert its biological
activity. Three x-ray structures were publicly available: 2XLM,
5FJK and 5FJH.
Ni(II) ion has an octahedral coordination filled with His 190, His
278, Glu192 from the protein and one or two water molecules
depending if the inhibitor is monodentate of bidentate ligand
(Figure 3). α-ketoglutarate acts as a bidentate ligand as well as
isonicotinic acid derivative from the 5FJH structure, while

5FJK ligand-blue; compound 4a in monodentate and bidentate
mode – magenta.
Figure 4. Field-based pharmacophoric features of KDM4C
inhibitors: a) average negative field of actives, b) average shape of
actives, c) alignment of 5FJH ligand with toxoflavin, d) alignment
of different conformations for compound 4a.
Figure 6. a) MMGBSA optimized geometry of compound 4a
aligned with 5FJH ligand and corresponding 2D diagram. b)
MMGBSA optimized geometry of compound 4a aligned with
5FJK ligand and corresponding 2D diagram
3.4. Investigation of KDM4C interactions with toxoflavin
derivatives
MMGBSA and QM/MM studies. Complexes obtained by
docking were further minimized using MMGBSA and used in the
QM/MM calculations where QM part was kept fixed and MM part
was minimized. Final geometries are shown on Figure 6; bidentate
binding mode of compound 4a is shown on Figure 6a and
monodentate on Figure 6b. In both cases, major interactions with
the protein are formed by metal coordination of 8-N and 7-CO
oxygen and H-bonds between 5-CO oxygen and Lys208 and
Tyr134. NBO analysis showed that delocalization of the nitrogen
and oxygen lone-pair electron density into the unfilled valence-
shell nonbonding orbitals of the metal ion were in the range of
KDM4A complexes; 22 and 45 kcal/mol; while electron
delocalization was much stronger for the 5FJH and 5FJK ligands;
162 and 138 kcal/mol, respectively.
Docking studies. Potential binding modes for the proposed
toxoflavin-based focused library were investigated by docking
into both, 5FJH/2XML and 5FJK X-ray structures. In the case of
5FJH structure, bidentate and monodentate binding modes were
explored. For the 5FJK structure, only the monodentate binding
mode was investigated with two coordinated water molecules. To
enable bidentate binding into 5FJH structure, Lys243 side-chain
was displaced as in the 2XML and 5FJK structure. Water
orientations and possible H-bonded networks were optimized
using a refinement procedure from the protein preparation wizard
and interactive H-bond optimizer. The SP docking protocol with
constrains defined as metal-ligand interaction and octahedral
coordination gave the best results.
Proposed docking pose for the toxoflavin itself into 2XML is
shown on Figure 5a confirming findings from the ligand-based
analysis that diversification vector could be placed at the C3 atom
on the toxoflavin scaffold. Docking of compound 4a resulted in
bidentate binding within the 5FJH and monodentate binding into
5FJK structure using SP accuracy level as shown in the Figure 5b.
Figure 5b also illustrates different binding modes and different
sub pockets filled with ligands from currently available X-ray
structures.
Correlation between experimentally determined inhibitory
activities with QM/MM energies calculated for the both
monodentate and bidentate binding is shown in section 3.5.1.
3.5 In vitro biological profiling
3.5.1. Inhibition activity against KDM4C enzyme. Most of the
linear compounds displayed inhibition of 40% or less at
concentration of 10 μM. Compounds 3c, 3d and 3l showed
inhibition in range 40-70%. Among the best linear compounds
were 3a and 3m with an inhibition over 70%. All cyclic molecules
showed inhibition of 100%, and therefore they were tested in a
secondary screening assay to calculate the IC50. Results are
presented in Table 3. All cyclic molecules displayed inhibition
activity in the nanomolar range. Compared to the published
activity of toxoflavin (IC50 = 541 nM, CHEMBL578512),
compounds 4a-4m showed improved activity, probably due to the
additional interactions of substituted phenyl ring with the protein.
However, binding efficiency of toxoflavin derivatives were not
improved over the toxoflavin scaffold (Table S2), pointing to non-
specific hydrophobic interactions of additional phenyl ring and
inability to form perfect octahedral coordination to the metal ion.
Figure 5. a) Proposed toxoflavin binding mode in the active site
of KDM4C enzyme (PDB:5FJH used). b) Alignment of 2XML
ligand-yellow; KDM4A inhibitors-green; 5FJH ligand-orange;

In order to modulate ADMET properties of molecules based on
toxoflavin and introduce additional chemical diversity, scaffold
hopping exercise was performed. Several structural modifications
have been made (Table 1b) with intention to reduce redox
potential and retain good activity. Although a drop in IC50 values
in the biochemical assay to the micromolar range was observed
(Table 3), both 5b and 5c could be considered as a novel starting
points for further medicinal chemistry efforts.
Table 3. Inhibition activity on KDM4C, passive permeability, P-
gp substrate indication, microsomal stability (Clint), and influence
on cell proliferation on HepG2 and A549 cell-lines for tested
compounds; toxoflavin (IC50 = 541 nM, CHEMBL578512).
Cpd
KDM4
C
IC50
[nM]
Passive
permeability(
A2B)
Papp [x10-6
cm/sec]
P-gp
substr
ate
Clint
[µl/min/
mg]
Hep
G2
IC50
[uM]
A549
IC50
(µM)
Figure 7. Correlation between experimental and predicted pIC50
values for a set of KDM4C inhibitors based on toxoflavin scaffold.
4a
4b
130
11
34.7
-
-
<12.5
<12.5
0.8
7
0.8
0
1.1
1.9
Results from in vitro ADME profiling including MDCKII-MDR1
permeability and microsomal metabolic stability are summarized
in the Table 3.
26.8
4c
4d
242
8
13.2
34.5
-
-
<12.5
<12.5
1.1
0.8
0
5.6 >30
1.2 1.3
5.1 >30
5.8
1.1
3.5.2. MDCKII-MDR1 permeability assay. The majority of
tested compounds showed a high passive permeability, with
Papp(A2B) values in the presence of P-gp inhibitor ranging from
10.8 to 34,5x10-6 cm/s. Two compounds were classified as low
permeable, 4e and 4k with permeability values <2 x10-6 cm/s.
According to the literature65-66 the permeability of the majority of
compounds tested was not influenced by P-glycoprotein, i.e.,
efflux ratio (Papp(BA)/Papp(AB)) was lower than 2 and/or in the
presence of inhibitor was not decreased by more than 50 % when
compared with efflux ratio without P-gp inhibitor. Only compound
4f was classified as possible P-gp substrate.
3.5.3. Metabolic stability. Incubation with mouse liver
microsomes showed that all tested compounds were stable
molecules with clearance values below 12.5 µl/min/mg protein,
except for 5a which showed high clearance profile. In addition,
compound 4k showed matrix instability and therefore the Clint
value could not be obtained.
4e
4f
4g
4j
4k
4l
4m
42
NA
29
19
NA
16
0.8
24.9
25.3
13.4
1.1
-
+
-
-
-
<12.5
<12.5
<12.5
11.6
71850 4.5
<12.5
<12.5
1.8
4.3
8.8
5.7
1.6
18.0
30.6
-
-
2.1
0.8
2
81
5a
>100
00
7626
1265
0
27.3
-
229.1 >25 >30
5b
5c
35.5
40.6
-
-
<12.5 >25 >30
13.1
>25 >30
Toxofla
vin
0.0
54133
16.0
-
<12.5
4.5
22
3.6. Redox potential of toxoflavin derivatives
Experimental activities expressed as pIC50 values were correlated,
with QM/MM energies calculated for the both monodentate and
bidentate binding and QikProp descriptors described, using
partial-least square method (as described in section 3.1. for
KDM4A case study). Results are shown in Figure 7 and Table S2.
A statistical model similar to that of KDM4A data set has been
developed with Eint, dipole and FOSA (hydrophobic part of
solvent accessible surface area) as descriptors. The developed
model can be further used to estimate activity of new potential
KDM4C inhibitors structurally similar to toxoflavin scaffold.
In order to better understand the electronic structure of prepared
toxoflavin derivatives and investigate possible link with measured
influence on cell viability, their redox potentials have been
calculated.
As already published for flavin scaffold, redox potential can be
modulated by substituents with different properties.67 Toxoflavin
as a natural product is known to possess redox properties that result
in the reduction of a cell viability. The mechanism of reaction is
similar to those of flavins and it was shown to proceed over the
following mechanism:
NADH + Toxoflavin→ NAD++ Toxoflavin-H2
Toxoflavin-H2 + O2 → Toxoflavin + H2O2
where the hydrogen peroxide as a toxic species is produced.68
Toxoflavin derivatives investigated in this paper are planar in the
oxidized form and bent to ±142° in the reduced form due to the
nitrogen pyramidalization as already demonstrated for flavin
derivatives.69-70

Although, the redox potential is the net effect of both processes,
reduction as a first step is expected to be slower than for the
toxoflavin scaffold, similarly to early work on flavines by Hall et.
al.
As expected, HOMO orbital of reduced form resembles the
LUMO orbital of the oxidized form. It is π* orbital mostly
localized on the diazaallene system and lone pair of the
neighbouring nitrogen atom. Both, HOMO orbital of the oxidised
form and LUMO of the reduced form also have contributions from
the electron density distribution on the phenyl substituent.
Figure 8. Planar and butterfly bent geometries of Ph substituted
toxoflavin derivative (3a) in oxidized and reduced form.
It is also interesting to note that the lone-pair of the N8 from the
diazaallene system populates HOMO of the oxidised form while
the other N4 lone pair is part of the LUMO orbital. This might
explain preferential coordination of N8 to the metal ion that we
found from the docking and QM/MM.
Influence on the redox potential by R1 substitutions of phenyl
scaffold was investigated in terms of the changes of the LUMO
energy levels for the oxidized form and HOMO energy levels of
the reduced form as shown on Figure 9. DFT theory at the M06-
2x/6-31(2d,p)+ level was used for the geometry optimisation and
calculation of the electronic properties as described in the method
section.
A decrease of the redox potential is expected to be a net effect of
the destabilization of the LUMO orbital of the oxidized form in
comparison to the toxoflavin (as it will require more energy to fill
in additional electrons) and stabilization of the HOMO orbital of
the reduced form (as it will require more energy to extract the
electrons) as shown on Figure 9.
Figure10. Frontier orbitals for the oxidized and reduced forms of
compound 3a.
Predicted reduction and oxidation potentials of toxoflavin
derivatives were compared with measured and calculated redox
potentials for polycyclic aromatic hydrocarbons as shown on
Figure 11.71-72 Predicted redox potentials were calculated from the
frontier orbital energies using regression equation for PAHs.
Reduction potentials of investigated toxoflavin derivatives are in
the range of the potentials measured and calculated for PAHs,
while oxidation potentials of reduced forms are higher than those
of PAHs.
Figure 9. Schematic presentation of the desired influence of the
substitution on the toxoflavin scaffold in order to decrease the
redox potential.
Indeed, LUMO orbitals of investigated derivatives in the oxidised
form are destabilised for prepared analogues in comparison to
toxoflavin. This indicates lower reduction affinity. However,
negative energy values indicate that reduction is still an energy-
favoured process for investigated toxoflavin derivatives.
Comparing significantly higher orbital energies of acyclic
compounds, where the diazaallene system is not existing, with
cyclic Ph-substituted toxoflavin derivatives, the question still
remains if this LUMO destabilisation is significant enough to
decrease redox potential and potential non-specific toxicity of Ph-
substituted compounds. However, the HOMO orbitals of reduced
forms are also destabilised by substitutions making them more
prone to the oxidation, except in case of compounds 3b and 3e.

influenced by the same parameters with addition of LUMO orbital
shifts of oxidised form for investigated compounds. This is in line
with the longer time scale in the experimental setup for A549 cell
lines where unspecific toxic effects could have more influence on
the measured IC50 values. It is important to note that for A549 cell
proliferation model toxoflavin was excluded since none of
envisaged independent variables could not explain significant
difference between cell proliferation on HepG2 and A549 cell-
lines. As stated previously, this could be indication that non-
specific toxicity of toxoflavin due to its redox potential is reduced
for the investigated toxoflavin derivatives.
Figure 11. Plot of computed HOMO/LUMO energies (eV) vs.
experimental E1/2 avg. Circled are values for toxoflavin as a
reference.
3.7. Influence on HepG2 and A549 cell line proliferation
Proliferation of HepG2 and A549 cell lines after 24 and 72 hours
of incubation time with tested compounds was investigated.
Results are shown in Table 3. Compounds 4a-4m are found to
influence the cell proliferation on both cell lines with IC50 in low
µM range. Exceptions are compounds 4e and 4g that showed no
influence on proliferation of A549 cells. Compounds 5a-5c
showed no influence on cell proliferation on both cell-lines. These
results are in line with measured activity on KDM4C enzyme.
However, influence on the cell proliferation is expected to be a
combination of target based activity on KDM4C (and probably
other close analogues from the KDM4 class of enzymes) and non-
specific toxicity due to the oxido-reductive potential of the
toxoflavin scaffold. Since results on the HepG2 cell-line are
comparable with those for A549 cell-line, despite significantly
longer incubation time in case of A549 cell-line, this indicates that
major factor influencing the cell-line proliferation could be
specific, target based activity on KDM4C enzyme. Such
conclusion is further supported by results we obtained for
toxoflavin itself, where IC50 on A549 cell-line is two orders of
magnitude lower that on HepG2 cell-line, and also one order of
magnitude lower than its binding activity on KDM4C.
Figure 12. Inhibitory KDM4C enzyme activity vs. influence on
the cell proliferation for a) HepG2 and b) A549 cell-lines.
Experimental vs. predicted pIC50 values for c) HepG2 and d) A549
cell lines.
4. Conclusions
Focused library of non-carboxylate inhibitors of the KDM4C
histone lysine demethylase based on toxoflavin scaffold has
been designed, synthesized and profiled in biochemical and
cell-based assays. In order to minimize redox liability of
toxoflavin, substituents with different electronic properties
were introduced.
Several compounds were identified as potent inhibitors of
KDM4C with activity in low nM range in the target-specific
biochemical assay and in the µM range in the cell-based
assay. Active compounds possess improved in vitro ADME
properties with good passive cellular permeability and
metabolic stability.
Structural interactions of prepared inhibitors with the active
site of KDM4C have been analysed with the optimized
docking protocol, energetics of the interactions was predicted
by advanced QM/MM hybrid approach and redox potential
was calculated by hybrid DFT methods. Statistical analysis of
collected data has been performed in order to get mechanistic
insight into the major factors influencing cellular activity of
investigated compounds.
Correlation between inhibitory KDM4C enzyme activity and
influence on the cell proliferation is graphically shown on Figure
12a and b. In order to further explore major factors influencing
cellular proliferation, partial least square (PLS) analysis has been
performed to build up mechanistic models. Beside KDM4C
inhibitory activity, several structural parameters were calculated;
clogP, total polar surface area (TPSA), number of hydrogen bond
donors and acceptors (HBD, HBA). In addition, experimental
permeability (Papp (AB)) and calculated HOMO/LUMO orbital
shifts were used as independent variables. PLS models for the
influence on cell-proliferation for HepG2 and A549 cell-lines are
given on Figure 12c and d. Experimental values, calculated
parameters and predicted IC50 values are listed in Table S3. It is
interesting to note that HepG2 cell proliferation can be predicted
from specific KDM4C inhibitory activity, permeability into cells
and calculated logP values. Cell proliferation on A549 cells is

and therapeutic effect of novel demethylase inhibitor in breast
cancer. Am J Cancer Res 2015, 5 (4), 1519-1530.
However, full de-convolution of measured cellular activity to
distinguish activity on the KDM4C enzyme, redox liability
and non-specific influence on cell viability as well as possible
interactions with other targets still needs to be performed.
Nevertheless, interesting hit molecules with strong KDM4C
potency and good ADME profile are identified as a good
starting point for further optimization and characterization.
15.
Kawamura, A.; Münzel, M.; Kojima, T.; Yapp, C.;
Bhushan, B.; Goto, Y.; Tumber, A.; Katoh, T.; King, O. N. F.;
Passioura, T.; Walport, L. J.; Hatch, S. B.; Madden, S.; Müller,
S.; Brennan, P. E.; Chowdhury, R.; Hopkinson, R. J.; Suga, H.;
Schofield, C. J., Highly selective inhibition of histone
demethylases by de novo macrocyclic peptides. Nature Commun
2017, 8, 14773.
16.
Lin, H.; Li, Q.; Li, Q.; Zhu, J.; Gu, K.; Jiang, X.; Hu,
Conflicts of interest
Q.; Feng, F.; Qu, W.; Chen, Y.; Sun, H., Small molecule KDM4s
inhibitors as anti-cancer agents. J Enzyme Inhibit Med Chem
2018, 33, 777-793.
There are no conflicts to declare.
17.
Zhang, C.; Wang, Z.; Ji, Q.; Li, Q., Histone
demethylase JMJD2C: epigenetic regulators in tumors.
Oncotarget 2017, 8 (53), 91723-91733.
18.
Garcia, J.; Lizcano, F., KDM4C Activity Modulates
Cell Proliferation and Chromosome Segregation in Triple-
Negative Breast Cancer. Breast Cancer (Auckl) 2016, 10, 169-
175.
Notes and references
19.
Thinnes, C. C.; England, K. S.; Kawamura, A.;
Chowdhury, R.; Schofield, C. J.; Hopkinson, R. J., Targeting
histone lysine demethylases - progress, challenges, and the
future. Biochim Biophys Acta 2014, 1839 (12), 1416-1432.
1.
Zinchenko, A.; Berezhnoy, N. V.; Wang, S.;
Rosencrans, W. M.; Korolev, N.; van der Maarel, J. R. C.;
Nordenskiold, L., Single-molecule compaction of megabase-long
chromatin molecules by multivalent cations. Nucleic Acids Res
2018, 46 (2), 635-649.
20.
Westaway, S. M.; Preston, A. G.; Barker, M. D.;
Brown, F.; Brown, J. A.; Campbell, M.; Chung, C. W.; Diallo,
H.; Douault, C.; Drewes, G.; Eagle, R.; Gordon, L.; Haslam, C.;
Hayhow, T. G.; Humphreys, P. G.; Joberty, G.; Katso, R.;
Kruidenier, L.; Leveridge, M.; Liddle, J.; Mosley, J.; Muelbaier,
M.; Randle, R.; Rioja, I.; Rueger, A.; Seal, G. A.; Sheppard, R.
J.; Singh, O.; Taylor, J.; Thomas, P.; Thomson, D.; Wilson, D.
M.; Lee, K.; Prinjha, R. K., Cell Penetrant Inhibitors of the
KDM4 and KDM5 Families of Histone Lysine Demethylases. 1.
3-Amino-4-pyridine Carboxylate Derivatives. J Med Chem 2016,
59 (4), 1357-69.
2.
Bannister, A. J.; Kouzarides, T., Regulation of
chromatin by histone modifications. Cell Res 2011, 21 (3), 381-
95.
3.
lysine methylation. Nat Rev Mol Cell Biol 2005, 6 (11), 838-49.
4. Hublitz, P.; Albert, M.; Peters, A. H., Mechanisms of
transcriptional repression by histone lysine methylation. Int J
Dev Biol 2009, 53 (2-3), 335-54.
Martin, C.; Zhang, Y., The diverse functions of histone
5.
Cloos, P. A.; Christensen, J.; Agger, K.; Helin, K.,
21.
Westaway, S. M.; Preston, A. G. S.; Barker, M. D.;
Erasing the methyl mark: histone demethylases at the center of
cellular differentiation and disease. Genes Dev 2008, 22 (9),
1115-40.
6.
epigenetic states. Science 2010, 330 (6004), 612-6.
7. Pasini, D.; Bracken, A.; Agger, K.; Christensen, J.;
Hansen, K.; Cloos, P.; Helin, K., Regulation of Stem Cell
Differentiation by Histone Methyltransferases and Demethylases.
Cold Spring Harb Symp Quant Biol. 2008; 73, 253-63.
Brown, F.; Brown, J. A.; Campbell, M.; Chung, C.-W.; Drewes,
G.; Eagle, R.; Garton, N.; Gordon, L.; Haslam, C.; Hayhow, T.
G.; Humphreys, P. G.; Joberty, G.; Katso, R.; Kruidenier, L.;
Leveridge, M.; Pemberton, M.; Rioja, I.; Seal, G. A.; Shipley, T.;
Singh, O.; Suckling, C. J.; Taylor, J.; Thomas, P.; Wilson, D. M.;
Lee, K.; Prinjha, R. K., Cell penetrant inhibitors of the KDM4
and KDM5 families of histone lysine demethylases. 2.
Pyrido[3,4-d]pyrimidin-4(3H)-one derivatives. J Med Chem
2016, 59 (4), 1370-1387.
Bonasio, R.; Tu, S.; Reinberg, D., Molecular signals of
8.
Berry, W. L.; Janknecht, R., KDM4/JMJD2 histone
22.
Franci, G.; Sarno, F.; Nebbioso, A.; Altucci, L.,
demethylases: epigenetic regulators in cancer cells. Cancer Res
2013, 73 (10), 2936-42.
9.
emerging functions of histone demethylases. Curr Opin Genet
Dev 2008, 18 (2), 159-68.
10.
epigenetic 'language' of covalent histone modifications to cancer.
Br J Cancer 2004, 90 (4), 761-9.
11.
Identification and characterization of PKF118-310 as a KDM4A
inhibitor. Epigenetics 2017, 12 (3), 198-205.
Agger, K.; Christensen, J.; Cloos, P. A.; Helin, K., The
23.
Davies, M.; Nowotka, M.; Papadatos, G.; Dedman, N.;
Gaulton, A.; Atkinson, F.; Bellis, L.; Overington, J. P., ChEMBL
web services: streamlining access to drug discovery data and
utilities. Nucleic acids Res 2015, 43 (W1), W612-W620.
Hake, S. B.; Xiao, A.; Allis, C. D., Linking the
24.
Raoof, A.; Depledge, P.; Hamilton, N. M.; Hamilton, N.
S.; Hitchin, J. R.; Hopkins, G. V.; Jordan, A. M.; Maguire, L. A.;
McGonagle, A. E.; Mould, D. P.; Rushbrooke, M.; Small, H. F.;
Smith, K. M.; Thomson, G. J.; Turlais, F.; Waddell, I. D.;
Waszkowycz, B.; Watson, A. J.; Ogilvie, D. J., Toxoflavins and
Deazaflavins as the First Reported Selective Small Molecule
Inhibitors of Tyrosyl-DNA Phosphodiesterase II. J Med Chem
2013, 56 (16), 6352-6370.
Mallette, Frédérick A.; Richard, S., JMJD2A Promotes
Cellular Transformation by Blocking Cellular Senescence
through Transcriptional Repression of the Tumor Suppressor
CHD5. Cell Reports 2012, 2 (5), 1233-1243.
12.
Labbe, R. M.; Holowatyj, A.; Yang, Z. Q., Histone
lysine demethylase (KDM) subfamily 4: structures, functions and
therapeutic potential. Am J Transl Res 2013, 6 (1), 1-15.
13.
of Jumonji D2 (JMJD2/KDM4) histone demethylase
overexpression. Biochem Cell Biol 2013, 91 (6), 369-77.
14.
25.
Todorovic, N.; Giacomelli, A.; Hassell, J. A.;
Young, L. C.; Hendzel, M. J., The oncogenic potential
Frampton, C. S.; Capretta, A., Microwave-assisted synthesis of 3-
aryl-pyrimido[5,4-e][1,2,4]triazine-5,7(1H,6H)-dione libraries:
derivatives of toxoflavin. Tetrahedron Lett 2010, 51 (46), 6037-
6040.
Ye, Q.; Holowatyj, A.; Wu, J.; Liu, H.; Zhang, L.;
Suzuki, T.; Yang, Z.-Q., Genetic alterations of KDM4 subfamily

26.
Kyani, A.; Tamura, S.; Yang, S.; Shergalis, A.;
Dempster, S.; Boxall, K. J.; Jeganathan, F.; Hatch, S. B.;
Savitsky, P.; Velupillai, S.; Krojer, T.; England, K. S.; Sejberg,
J.; Thai, C.; Donovan, A.; Pal, A.; Scozzafava, G.; Bennett, J.
M.; Kawamura, A.; Johansson, C.; Szykowska, A.; Gileadi, C.;
Burgess-Brown, N. A.; von Delft, F.; Oppermann, U.; Walters,
Z.; Shipley, J.; Raynaud, F. I.; Westaway, S. M.; Prinjha, R. K.;
Fedorov, O.; Burke, R.; Schofield, C. J.; Westwood, I. M.;
Bountra, C.; Muller, S.; van Montfort, R. L.; Brennan, P. E.;
Blagg, J., 8-Substituted Pyrido[3,4-d]pyrimidin-4(3H)-one
Derivatives As Potent, Cell Permeable, KDM4 (JMJD2) and
KDM5 (JARID1) Histone Lysine Demethylase Inhibitors. J Med
Chem 2016, 59 (4), 1388-409.
Samanta, S.; Kuang, Y.; Ljungman, M.; Neamati, N., Discovery
and Mechanistic Elucidation of a Class of Protein Disulfide
Isomerase Inhibitors for the Treatment of Glioblastoma. Chem
Med Chem 2018, 13 (2), 164-177.
27.
Zeller, J.; Turbiak, A. J.; Powelson, I. A.; Lee, S.; Sun,
D.; Showalter, H. D. H.; Fearon, E. R., Investigation of 3-aryl-
pyrimido[5,4-e][1,2,4]triazine-5,7-diones as small molecule
antagonists of β-catenin/TCF transcription. Bioorg Med Chem
Lett 2013, 23 (21), 5814-5820.
28.
Nagamatsu, T.; Yamasaki, H., General syntheses of 1-
alkyltoxoflavin and 8-alkylfervenulin derivatives of biological
significance by the regioselective alkylation of reumycin
derivatives and the rates of transalkylation from 1-
alkyltoxoflavins into nucleophiles. J Chem Soc, Perkin Trans 1
2001, (2), 130-137.
42.
Becke, A. D., Density‐functional thermochemistry. III.
The role of exact exchange. J Chem Phys 1993, 98 (7), 5648-
5652.
43.
Lee, C.; Yang, W.; Parr, R. G., Development of the
29.
Yoneda, F.; Higuchi, M.; Nitta, Y., Synthesis of some
Colle-Salvetti correlation-energy formula into a functional of the
5H-imidazo[4,5-e]-as-triazin-6-(7H)ones (6-azapurin-8-ones). J
Heterocycl Chem 1980, 17 (5), 869-871.
electron density. Phys Rev B 1988, 37 (2), 785-789.
44.
Vosko, S. H.; Wilk, L.; Nusair, M., Accurate spin-
30.
Whetstine, J. R.; Nottke, A.; Lan, F.; Huarte, M.;
dependent electron liquid correlation energies for local spin
density calculations: a critical analysis. Can J Phys 1980, 58 (8),
1200-1211.
Smolikov, S.; Chen, Z.; Spooner, E.; Li, E.; Zhang, G.;
Colaiacovo, M.; Shi, Y., Reversal of histone lysine trimethylation
by the JMJD2 family of histone demethylases. Cell 2006, 125
(3), 467-81.
45.
Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.;
Frisch, M. J., Ab Initio Calculation of Vibrational Absorption
and Circular Dichroism Spectra Using Density Functional Force
Fields. J Phys Chem 1994, 98 (45), 11623-11627.
31.
Mosmann, T., Rapid colorimetric assay for cellular
growth and survival: application to proliferation and cytotoxicity
assays. J Immunol Methods 1983, 65 (1-2), 55-63.
46.
Hehre, W. J.; Ditchfield, R.; Pople, J. A., Self—
32.
Nožinić, D.; Milić, A.; Mikac, L.; Ralić, J.; Padovan, J.;
Consistent Molecular Orbital Methods. XII. Further Extensions
of Gaussian—Type Basis Sets for Use in Molecular Orbital
Studies of Organic Molecules. J Chem Phys 1972, 56 (5), 2257-
2261.
Antolović, R., Assessment of macrolide transport using PAMPA,
Caco-2 and MDCKII-hMDR1 assays. Croat Chem Acta 2010,
83.
33.
Gaulton, A.; Hersey, A.; Nowotka, M.; Bento, A. P.;
47.
Hay, P. J.; Wadt, W. R., Ab initio effective core
Chambers, J.; Mendez, D.; Mutowo, P.; Atkinson, F.; Bellis, L.
J.; Cibrian-Uhalte, E.; Davies, M.; Dedman, N.; Karlsson, A.;
Magarinos, M. P.; Overington, J. P.; Papadatos, G.; Smit, I.;
Leach, A. R., The ChEMBL database in 2017. Nucleic Acids Res
2017, 45 (D1), D945-D954.
potentials for molecular calculations. Potentials for K to Au
including the outermost core orbitals. J Chem Phys 1985, 82 (1),
299-310.
48.
Banks, J. L.; Beard, H. S.; Cao, Y.; Cho, A. E.; Damm,
W.; Farid, R.; Felts, A. K.; Halgren, T. A.; Mainz, D. T.; Maple,
J. R.; Murphy, R.; Philipp, D. M.; Repasky, M. P.; Zhang, L. Y.;
Berne, B. J.; Friesner, R. A.; Gallicchio, E.; Levy, R. M.,
Integrated Modeling Program, Applied Chemical Theory
(IMPACT). J Comput Chem 2005, 26 (16), 1752-80.
34.
Cheeseright, T.; Mackey, M.; Rose, S.; Vinter, A.,
Molecular field extrema as descriptors of biological activity:
definition and validation. J Chem Inf Model 2006, 46 (2), 665-76.
35.
Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.;
Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E., The
49.
Glendening, E. D..;Landis C. R.; Weinhold F, Natural
Protein Data Bank. Nucleic Acids Res 2000, 28 (1), 235-42.
bond orbital methods. Wiley Int Rev Comput Mol Sci 2012, 2 (1),
36.
Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye,
1-42.
L. L.; Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.;
Mainz, D. T., Extra precision glide: docking and scoring
incorporating a model of hydrophobic enclosure for protein-
ligand complexes. J Med Chem 2006, 49 (21), 6177-96.
50.
Armendáriz-Vidales, G.; Hernández-Muñoz, L. S.;
González, F. J.; de Souza, A. A.; de Abreu, F. C.; Jardim, G. A.
M.; da Silva, E. N.; Goulart, M. O. F.; Frontana, C., Nature of
Electrogenerated Intermediates in Nitro-Substituted Nor-β-
lapachones: The Structure of Radical Species during Successive
Electron Transfer in Multiredox Centers. J Org Chem 2014, 79
(11), 5201-5208.
37.
Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren,
T. A.; Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.;
Shelley, M.; Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S.,
Glide: a new approach for rapid, accurate docking and scoring. 1.
Method and assessment of docking accuracy. J Med Chem 2004,
47 (7), 1739-49.
51.
Marenich, A. V.; Ho, J.; Coote, M. L.; Cramer, C. J.;
Truhlar, D. G., Computational electrochemistry: prediction of
liquid-phase reduction potentials. Phys Chem Chem Phys 2014,
16 (29), 15068-106.
38.
39.
Schroedinger MacroModel, Schroeodinger, LLC, 2017.
Li, J.; Abel, R.; Zhu, K.; Cao, Y.; Zhao, S.; Friesner, R.
52.
Isegawa, M.; Neese, F.; Pantazis, D. A., Ionization
A., The VSGB 2.0 model: a next generation energy model for
high resolution protein structure modeling. Proteins 2011, 79
(10), 2794-812.
Energies and Aqueous Redox Potentials of Organic Molecules:
Comparison of DFT, Correlated ab Initio Theory and Pair
Natural Orbital Approaches. J Chem Theory Comput 2016, 12
(5), 2272-2284.
40.
Shivakumar, D.; Williams, J.; Wu, Y.; Damm, W.;
Shelley, J.; Sherman, W., Prediction of Absolute Solvation Free
Energies using Molecular Dynamics Free Energy Perturbation
and the OPLS Force Field. J Chem Theory Comput 2010, 6 (5),
1509-1519.
53.
Friesner, R. A., Modeling Metalloenzymes with Density
Functional and Mixed Quantum Mechanical/Molecular
Mechanical (QM/MM) Calculations: Progress and Challenges.
Encyclopedia of Inorganic Chemistry, 2009.
41.
Bavetsias, V.; Lanigan, R. M.; Ruda, G. F.; Atrash, B.;
McLaughlin, M. G.; Tumber, A.; Mok, N. Y.; Le Bihan, Y. V.;

54.
Vidossich, P.; Magistrato, A., QM/MM Molecular
66.
Mahar Doan, K. M.; Humphreys, J. E.; Webster, L. O.;
Dynamics Studies of Metal Binding Proteins. Biomolecules 2014,
4 (3), 616.
Wring, S. A.; Shampine, L. J.; Serabjit-Singh, C. J.; Adkison, K.
K.; Polli, J. W., Passive permeability and P-glycoprotein-
mediated efflux differentiate central nervous system (CNS) and
non-CNS marketed drugs. J Pharmacol Exp Ther 2002, 303 (3),
1029-37.
55.
Khandelwal, A.; Lukacova, V.; Comez, D.; Kroll, D.
M.; Raha, S.; Balaz, S., A Combination of Docking, QM/MM
Methods, and MD Simulation for Binding Affinity Estimation of
Metalloprotein Ligands. J Med Chem 2005, 48 (17), 5437-5447.
67.
Mueller, R. M.; North, M. A.; Yang, C.; Hati, S.;
56.
Siegbahn, P. E. M.; Blomberg, M. R. A., Transition-
Bhattacharyya, S., Interplay of Flavin's Redox States and Protein
Dynamics: An Insight from QM/MM Simulations of
Dihydronicotinamide Riboside Quinone Oxidoreductase 2. J
Phys Chem B 2011, 115 (13), 3632-3641.
Metal Systems in Biochemistry Studied by High-Accuracy
Quantum Chemical Methods. Chem Rev 2000, 100 (2), 421-438.
57.
Lundberg, M.; Siegbahn P.E.M., Agreement between
experiment and hybrid DFT calculations for O-H bond
dissociation enthalpies in manganese complexes. J Comput Chem
2005, 26 (7), 661-667.
68.
toxicity of toxoflavin. Biochim Biophys Acta 1961, 52, 502-8.
69. Hall., L. H.; Orchard, B. J.; Tripathy, S. K., The
Latuasan, H. E.; Berends, W., On the origin of the
58.
Lovell, T.; Himo, F.; Han, W.-G.; Noodleman, L.,
structure and properties of flavins: Molecular orbital study based
on totally optimized geometries. I. Molecular geometry
investigations. Int J Quantum Chem 1987, 31 (2), 195-216.
Density functional methods applied to metalloenzymes. Coord
Chem Rev 2003, 238-239, 211-232.
59.
Murphy R. B.; Phillipp D. M.; Friesner R. A, A mixed
70.
North, M. A.; Bhattacharyya, S.; Truhlar, D. G.,
quantum mechanics/molecular mechanics (QM/MM) method for
large-scale modeling of chemistry in protein environments. J
Comput Chem 2000, 21 (16), 1442-1457.
Improved Density Functional Description of the Electrochemistry
and Structure−Property Descriptors of Substituted Flavins. J
Phys Chem B 2010, 114 (46), 14907-14915.
60.
Duffy, E. M.; Jorgensen, W. L., Prediction of Properties
71.
Méndez-Hernández, D. D.; Tarakeshwar, P.; Gust, D.;
from Simulations:ꢀ Free Energies of Solvation in Hexadecane,
Octanol, and Water. J Am Chem Soc 2000, 122 (12), 2878-2888.
Moore, T. A.; Moore, A. L.; Mujica, V., Simple and accurate
correlation of experimental redox potentials and DFT-calculated
HOMO/LUMO energies of polycyclic aromatic hydrocarbons. J
Mol Model 2013, 19 (7), 2845-2848.
61.
62.
2018-2, S. R. QikProp, Schrödinger,, 2018.
Schrödinger Release 2019-1: QikProp, S., LLC, New
York, NY, 2019.
72.
Davis, A. P.; Fry, A. J., Experimental and computed
63.
64.
Biotech, S. S. Simca-P11.5, 2013.
Cecatiello, V., Pasqualato, S., Crystal Structures of
absolute redox potentials of polycyclic aromatic hydrocarbons
are highly linearly correlated over a wide range of structures and
potentials. J Phys Chem A 2010, 114 (46), 12299-304.
Human Jmjd2C Catalytic Domain Bound to Inhibitors. 2015.,
DOI: 10.2210/PDB5FJK/PDB
65.
Arpinelli, F.; Bamfi, F., The FDA guidance for industry
Click here to remove instruction text...
on PROs: the point of view of a pharmaceutical company. Health
Qual Life Out 2006, 4 (1), 85.

Conflicts of interest
There are no conflicts to declare.

Graphical Abstract
Rational design, synthesis and biological
profiling of new KDM4C inhibitors
Leave this area blank for abstract info.
Vatroslav Letfus*, Dubravko Jelić, Ana Bokulić, Adriana Petrinić Grba, Sanja Koštrun
Fidelta Ltd., Prilaz baruna Filipovića 29, 10000 Zagreb, Croatia
Advanced hit
Starting point
IC50=541 nM
Rational
design
IC50=8 nM
Optimized
ADME