Synthesis and activity of N-oxalylglycine and its derivatives as Jumonji C-domain-containing histone lysine demethylase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.03.098

N-Oxalylglycine (NOG) derivatives were synthesized, and their inhibitory effect on histone lysine demethylase activity was evaluated. NOG and compound 1 inhibited histone lysine demethylases JMJD2A, 2C and 2D in enzyme assays, and their dimethyl ester prodrugs DMOG and 21 exerted histone lysine methylating activity in cellular assays.

Full text of DOI:10.1016/j.bmcl.2009.03.098

Bioorganic & Medicinal Chemistry Letters 19 (2009) 2852–2855
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and activity of N-oxalylglycine and its derivatives as Jumonji
C-domain-containing histone lysine demethylase inhibitors
a
a
Shohei Hamada a, Tae-Dong Kim b, Takayoshi Suzuki a, , Yukihiro Itoh , Hiroki Tsumoto ,
*
a,
Hidehiko Nakagawa a, Ralf Janknecht b, , Naoki Miyata
*
*
a Graduate School of Pharmaceutical Sciences, Nagoya City University, 3-1 Tanabe-dori, Mizuho-ku, Nagoya, Aichi 467-8603, Japan
b Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73104, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
N-Oxalylglycine (NOG) derivatives were synthesized, and their inhibitory effect on histone lysine
demethylase activity was evaluated. NOG and compound 1 inhibited histone lysine demethylases
JMJD2A, 2C and 2D in enzyme assays, and their dimethyl ester prodrugs DMOG and 21 exerted histone
lysine methylating activity in cellular assays.
Received 3 February 2009
Revised 17 March 2009
Accepted 21 March 2009
Available online 26 March 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Histone lysine demethylase
N-Oxalylglycine
Inhibitor
Methylation of lysine residues in the N-terminal region of the
core histones plays a pivotal role in the regulation of gene expres-
sion.1 It has been reported that methylation of histone lysine resi-
dues occurs on lysine 26 of histone 1 (H1K26), H3K4, H3K9, H3K27,
H3K36, H3K79 and H4K20.2 Methylated lysine can be in a mono-,
di-, or tri-methylated form, and these different forms differentially
affect chromatin structure and are responsible for transcriptional
activation as well as silencing.3
In contrast to other histone modifications such as acetylation
and phosphorylation, histone methylation had been regarded as
irreversible because of the high thermodynamic stability of the
N–C bond. However, two classes of histone lysine demethylases
have recently been identified. One is lysine-specific demethylase
1, a flavin-dependent amine oxidase domain-containing enzyme.4
The other comprises the Jumonji C domain-containing histone
demethylases (JHDMs) such as JMJD2A, 2B, 2C and 2D.5 JHDMs
tate cancer.5,7 Therefore, JHDM inhibitors can be not only tools to
study the function of these enzymes, but also serve as potential
anticancer agents.
To date, only a few JHDM inhibitors have been identified
(Fig. 1). N-Oxalylglycine (NOG), the amide analogue of a-ketoglu-
tarate, has been reported to inhibit JMJD2C in an in vitro assay,7
but has not been tested in a cellular assay. Succinic acid has been
suggested to inhibit JMJD2D by product inhibition.8 In addition,
during the course of our study presented in this report, 2,4-lutidi-
nic acid, which is known as an inhibitor of other Fe(II)/a-ketoglu-
tarate dependent oxygenases, was reported as an inhibitor of
JMJD2A and 2E,9 although it was also not examined in a cellular as-
say. To the best of our knowledge, there is no report on selective
JHDM inhibitors. Therefore, we initiated a search for novel JHDM
inhibitors with the goal of drug discovery as well as finding new
tools for biological research. In this letter, we describe the design,
synthesis, JHDM inhibition activity and cellular activity of NOG and
its derivatives.
are Fe(II) and a-ketoglutarate dependent enzymes that oxygenate
methylated histone lysine residues and thereby cause their
demethylation after releasing formaldehyde, carbon dioxide and
succinic acid. The identification of these enzymes established that
histone methylation is reversibly regulated by histone lysine meth-
yltransferases and histone lysine demethylases.6
We designed JHDM inhibitors based on the crystal structure of
JMJD2A complexed with NOG and histone trimethylated lysine
JHDMs are implicated in cancer cell growth.2 For example, it
was shown that JMJD2C is associated with the growth of oesopha-
gal squamous cancer, and JMJD2A, 2B and 2C are involved in pros-
N
CO2H
O
CO2H
HO2C
HO2C
N
H
CO2H
CO2H
NOG
succinic acid
2,4-lutidinic acid
Figure 1. Structures of previously reported JHDM inhibitors.
* Corresponding authors. Tel./fax: +81 52 836 3407 (N. Miyata).
E-mail addresses: suzuki@phar.nagoya-cu.ac.jp (T. Suzuki), ralf-janknecht@
ouhsc.edu (R. Janknecht), miyata-n@phar.nagoya-cu.ac.jp (N. Miyata).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.03.098

S. Hamada et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2852–2855
2853
peptide (PDB ID 2OQ6).10 The crystal structure showed that the
NH2
NHBoc
NH2
NHBoc
NHBoc
a
c
b
d
oxalyl group of NOG interacts with Fe(II), and the other carboxyl
group forms a hydrogen bond with Tyr 132 in the active centre
of JMJD2A (Fig. 2, left). In addition, the trimethylamino group of
histone trimethylated lysine peptide is surrounded by Gly 170,
Tyr 175, Glu 190 and Ser 288. On the basis of this structure, we de-
signed potential selective JHDM inhibitors 1–4 (Fig. 3) in which
NOG is connected with a dimethylamino group through a linker
(Fig. 2, right). We anticipated that the nitrogen atom of the dimeth-
ylamino group forms a hydrogen bond with the hydroxyl group of
Tyr 175 or Ser 288,11,12 and the aryl linker interacts with the aro-
matic rings of Tyr 177, Phe 185 and Trp 208, which might lead to
the selective inhibition of JHDMs.
O2N
O2N
8
9
H2N
Me2N
10
11
2
e-g
Me2N
12
Scheme 2. Reagents and conditions: (a) (Boc)2O, Et3N, THF, 0 °C to rt, 82%; (b) H2,
Pd/C, EtOH, rt, 90%; (c) MeI, K2CO3, DMF, rt, 34%; (d) HCl, AcOEt, CH2Cl2, 0 °C to rt,
94%; (e) tert-butyl bromoacetate, Et3N, CH2Cl2, rt; (f) tert-butyl chloroglyoxylate,
Et3N, CH2Cl2, rt; (g) HCl, AcOEt, CHCl3, 0 °C to rt, 26% (three steps).
The synthesis of compound 1 is outlined in Scheme 1. The pri-
mary amine of 4-dimethylaminobenzylamine 5 was alkylated with
tert-butyl bromoacetate to give secondary amine 6. Treatment of
compound 6 with tert-butyl chloroglyoxylate afforded tertiary
amine 7. Removal of the two tert-butyl groups of 7 under acidic
conditions gave the desired dicarboxylic acid 1.
Scheme 2 shows the preparation of compound 2. Reaction of 4-
nitrophenethylamine 8 with (Boc)2O gave Boc-protected com-
pound 9. Reduction of the nitro group of 9 gave aniline 10. Aniline
10 was reacted with iodomethane to give dimethylamine 11. Then,
deprotection of 11 using hydrochloric acid yielded amine 12. Com-
pound 2 was prepared from amine 12 using the procedure de-
scribed for the synthesis of 1.
Preparation of compound 3 is shown in Scheme 3. Horner–
Wadsworth–Emmons reaction was applied to the conversion of
benzaldehyde 13 into cinnamonitrile 14. The double bond of 14
was hydrogenated to give compound 15. The cyano group of 15
was then reduced, and subsequent treatment with (Boc)2O gave
compound 16. N-Boc compound 16 was converted to compound
3 using the procedure described for the synthesis of 1 and 2.
Scheme 4 illustrates the synthesis of compound 4. Coupling be-
tween carboxylic acid 17 and NH3 in the presence of EDCI and
substrate
NH
O
Gly 170
Glu 190
Trp 208
Gly 170
Glu 190
Ser 288
Tyr 175
Ser 288
Me
Me
Me
Me
N
N
O
H
Tyr 175
Me
drug design Trp 208
linker
Tyr 177
O
Tyr 177
Phe 185
Phe 185
O
O
O
O
H
N
N
O
O
O
O
NOG
O
Fe2+
Fe2+
H
O
H
O
inhibitor
Tyr 132
Tyr 132
Figure 2. Interaction of NOG and trimethylated lysine substrate with JMJD2A (left),
and models for the binding of designed inhibitors (right).
O
CN
a
b
HO
N
O
Me2N
HO
N
O
H
Me2N
Me2N
Me2N
14
OH
OH
13
Me2N
O
O
CN
O
O
c
NHBoc
1
3
2
4
Me2N
HO
N
O
HO
N
O
16
15
Me2N
OH
OH
d-g
Me2N
O
O
3
O
O
Scheme 3. Reagents and conditions: (a) cyanomethylphosphoic acid diethyl ester,
NaH, THF, 0 °C, 66%; (b) H2, Pd/C, MeOH, 83%; (c) (i) BH3?SMe2, THF, reflux; (ii)
(Boc)2O, Et3N, THF, rt, 58%; (d) HCl, AcOEt, CH2Cl2, 0 °C to rt, 80%; (e) tert-butyl
bromoacetate, Et3N, CH2Cl2, rt; (f) tert-butyl chloroglyoxylate, Et3N, CH2Cl2, 0 °C, rt;
(g) HCl, AcOEt, CH2Cl2, 0 °C to rt, 12% (three steps).
Figure 3. Structures of compounds 1–4.
t-BuO
O
NH2
a
N
H
O
O
Me2N
a
b
Me2N
Me2N
Me2N
Me2N
OH
NH2
5
6
t-BuO
O
17
18
c
b
N
O
c-f
1
NHBoc
4
Ot-Bu
Me2N
O
19
7
Scheme 4. Reagents and conditions: (a) NH3, EDCI, HOBtÁH2O, DMF, H2O, rt, 54%;
(b) (i) LiAlH4, THF, reflux; (ii) (Boc)2O, Et3N, THF, rt, 87%; (c) HCl, AcOEt, 0 °C to rt;
(d) tert-butyl bromoacetate, Et3N, CH2Cl2, rt; (e) tert-butyl chloroglyoxylate, Et3N,
CH2Cl2, 0 °C, rt; (f) HCl, AcOEt, CH2Cl2, 0 °C to rt, 23% (four steps).
Scheme 1. Reagents and conditions: (a) tert-butyl bromoacetate, Et3N, CH2Cl2, rt,
54%; (b) tert-butyl chloroglyoxylate, Et3N, CH2Cl2, 0 °C, 58%; (c) HCl, AcOEt, CH2Cl2,
0 °C to rt, 59%.

2854
S. Hamada et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2852–2855
HOBt afforded amide 18. The reduction of the amide group of 18,
followed by the treatment with (Boc)2O gave compound 19. Com-
pound 4 was prepared from compound 19 using the procedure de-
scribed for the synthesis of 1 and 2.
For the evaluation of the enzyme inhibitory activity of NOG and
its derivatives 1–4, we initially generated GST fusion proteins of
the catalytic N-terminus of JMJD2A and 2C as well as of full-length
JMJD2D.13–16 Because JMJD2A, 2C and 2D have been reported to
demethylate H3K9me3,14 we confirmed the histone demethylating
activity of these three proteins in an in vitro assay and revealed
demethylation with an antibody recognizing trimethylated H3K9
(H3K9me3).17–20 As shown in Figure 4, all three GST-JMJD2 fusion
proteins completely removed H3K9me3, whereas no demethylat-
ing activity was observed with mutated GST fusion proteins13 in
which homologous histidine residues in the catalytic centre were
mutated to alanine. Using this system, we next assessed the impact
of NOG and putative JMJD2 inhibitors 1–4 on the in vitro demeth-
ylation activity of GST-JMJD2 fusion proteins.17 We first utilized
inhibitors at 1 mM concentration and observed that only NOG sig-
nificantly curtailed the demethylation activity of JMJD2A and
JMJD2C (Fig. 5, lane 4). However, at a higher inhibitor concentra-
tion of 3 mM, significant inhibitory effects were also noted with
compounds 1, 2, 4 and weakest with 3. In contrast, JMJD2D enzy-
matic activity was not only strongly inhibited by 1 mM NOG, but
also by 1 mM compounds 1, 2 and 3. Furthermore, compound 4
also displayed an inhibitory effect at 3 mM concentration. Thus,
all compounds tested were able to inhibit the demethylation activ-
ity of JMJD2 proteins.
Figure 5. In vitro JMJD2 inhibition assay of NOG and compounds 1–4. H4K20me1
was assessed as a control and revealed to be unchanged.
MeO
O
NH2
a
N
H
Me2N
Me2N
O
5
20
MeO
Next, we assessed if the inhibitors would display activity in
cells.14,21–25 We focused on the strongest inhibitors, NOG and com-
pound 1, and prepared dimethylester prodrugs of these com-
pounds to allow for uptake through the cell membrane. Scheme
5 shows the synthesis of compound 21, the dimethylester prodrug
of compound 1. Compound 5 was allowed to react with methyl
bromoacetate to give compound 20. Then, compound 20 was re-
acted with methyl chloroglyoxylate to give compound 21.
In this cellular assay, the accumulation of H3K9me3 and
H3K36me3 was examined, because JMJD2C has been reported to
demethylate both H3K9 and H3K36.14 As shown in Figure 6, over-
expression of JMJD2C resulted in robust demethylation of
H3K9me3 and H3K36me3 (compare lanes 1 and 5). Further, neither
100 mM dimethyl succinate (DMS), a previously reported JMJD2D
b
N
O
OMe
Me2N
O
21
Scheme 5. Reagents and conditions: (a) bromoacetic acid methyl ester, Et3N,
CH2Cl2, rt, 58%; (b) chloroglyoxylic acid methyl ester, Et3N, CH2Cl2, 0 °C, 35%.
Figure 6. Demethylation in cells. 293T cells transfected with or without Flag-
JMJD2C were treated with 100 mM dimethyl succinate (DMS), 2.5 mM oxalylgly-
cine dimethylester (DMOG), 2.5 mM compound 21 or DMSO as
a control.
Demethylation activity of JMJD2C was assessed by anti-H3K9me3 and anti-
H3K36me3 blotting. As loading controls, total H3 and actin levels were assessed,
and expression of JMJD2C revealed by anti-Flag blotting.
inhibitor,8 nor 2.5 mM compound 21 affected H3K9me3 and
H3K36me3 in the presence of JMJD2C. However, 2.5 mM dimethy-
lester of NOG (DMOG) resulted in enhanced H3K9me3 and
H3K36me3 levels both in the presence of JMJD2C and in the ab-
sence of overexpressed JMJD2C, indicating that DMOG represses
the demethylation activity of JHDMs including JMJD2C in cells.
On the other hand, compound 21 slightly enhanced H3K9me3 lev-
Figure 4. Demethylation ability of GST-JMJD2 fusion proteins (top) and Coomassie
stained gel (bottom) indicating that comparable amounts of GST-JMJD2 fusion
proteins were utilized.

S. Hamada et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2852–2855
2855
freezing in liquid nitrogen, GST fusion proteins were dialyzed against 10 mM
Hepes pH 7.4, 50 mM NaCl, 10% glycerol, 0.2 mM phenylmethylsulfonyl
fluoride and 1 mM DTT (Ref. 16).
els in the absence of overexpressed JMJD2C as compared with
DMSO and DMS, suggesting that 2.5 mM compound 21 selectively
inhibited the demethylase activity of a JHDM(s) other than JMJD2C
in cells.
14. Shin, S.; Janknecht, R. Biochem. Biophys. Res. Commun. 2007, 353, 973.
15. Knebel, J.; De Haro, L.; Janknecht, R. J. Cell. Biochem. 2006, 99, 319.
16. Shin, S.; Rossow, K. L.; Grande, J. P.; Janknecht, R. Cancer Res. 2007, 67, 7572.
In summary, we have designed and synthesized NOG deriva-
tives 1–426 and evaluated the JMJD2-inhibitory activity of NOG
and compounds 1–4 both in enzyme assays and in cellular assays
that have been established for this study. Our data suggest that
NOG and compound 1 show inhibitory activity against JHDMs
and are more potent than succinic acid. As far as we could deter-
mine, this is the first report demonstrating that NOG and its deriv-
ative inhibit JHDMs both in enzyme assays and in cellular assays.
The findings presented here should be valuable for further explor-
ative studies uncovering more potent and selective JHDM inhibi-
tors. Further investigations pertaining to NOG derivatives are
progressing and will be reported in due course.
17. Reactions were set up in 18
KCl, 2 mM ascorbate, 0.2 mM (NH4)2Fe(SO4)2 and either 1 mM (Fig. 4) or 10
(Fig. 5) -ketoglutarate employing 3 g histones (USB) and 0.5 g of indicated
GST fusion proteins (Ref. 8). Where indicated, 1.8 or 5.4 of 10 mM
ll of 50 mM Hepes pH 7.4, 50 mM NaCl, 50 mM
l
V
a
l
l
l
l
ll
inhibitor in 10% (v/v) DMSO or the vehicle DMSO itself was included in the
reaction mixture. Reactions proceeded at 37 °C for 3 h (Fig. 4) or 1 h (Fig. 5,
GST-JMJD2A and GST-JMJD2C) or 2 h (Fig. 5, GST-JMJD2D). After adding 18 ll
of 2Â Laemmli sample buffer and boiling for 10 min, supernatants were
employed for SDS–PAGE (Ref. 18). Western blotting was then performed as
described Ref. 19 utilizing antibodies from Upstate against H3K9me3 (07-442;
1:4000 dilution) or H4K20me1 (07-440; 1:2000 dilution). After incubation with
secondary antibody coupled to horseradish peroxidase, enhanced
chemiluminescence was employed to reveal proteins (Ref. 20).
18. Kim, T. D.; Shin, S.; Janknecht, R. Biochem. Biophys. Res. Commun. 2008, 366, 563.
19. Papoutsopoulou, S.; Janknecht, R. Mol. Cell. Biol. 2000, 20, 7300.
20. Janknecht, R. Oncogene 2003, 22, 746.
21. 293T cells were seeded in poly-
confluency. Then, cells were transiently transfected by the calcium phosphate
coprecipitation method essentially as described (Refs. 22,23). 1.7 of
pBluescript KS+ and 0.5
g of empty vector pEV3S or Flag-JMJD2C expression
L-Lysine coated 12 wells and grown to 25%
Acknowledgement
lg
l
This work was supported in part by a Grant-in-Aid for Scientific
Research from Japan Society for the Promotion of Science.
vector (Ref. 24) were utilized for transfection. 12 h after addition of the
precipitate to the cells, it was removed by washing twice with phosphate-
buffered saline (Ref. 25). Cells were then incubated in 600 ll DMEM + 10% fetal
calf serum for 12 h in a humidified atmosphere containing 10% CO2. Then,
medium was replaced with fresh one containing 1% of DMSO or inhibitors
dissolved in DMSO and cells were incubated for another 24 h. Cells were then
References and notes
1. Suzuki, T.; Miyata, N. Curr. Med. Chem. 2006, 13, 935.
2. Bannister, A. J.; Kouzarides, T. Nature 2005, 436, 1103.
3. Kubicek, S.; Jenuwein, T. Cell 2004, 119, 903.
4. Shi, Y.; Lan, F.; Matson, C.; Mulligan, P.; Whetstine, J. R.; Cole, P. A.; Casero, R. A.;
Shi, Y. Cell 2004, 119, 941.
5. Shi, Y. Nat. Rev. Genet. 2007, 8, 829.
6. Cole, P. A. Nat. Chem. Biol. 2008, 4, 590.
7. Cloos, P. A.; Christensen, J.; Agger, K.; Maiolica, A.; Rappsilber, J.; Antal, T.;
Hansen, K. H.; Helin, K. Nature 2006, 442, 307.
8. Smith, E. H.; Janknecht, R.; Maher, L. J. III. Hum. Mol. Genet. 2007, 16, 3136.
9. Rose, N. R.; Ng, S. S.; Mecinovic, J.; Liénard, B. M. R.; Bello, S. H.; Sun, Z.;
McDonough, M. A.; Oppermann, U.; Schofield, C. J. J. Med. Chem. 2008, 51, 7053.
10. Ng, S. S.; Kavanagh, K. L.; McDonough, M. A.; Butler, D.; Pilka, E. S.; Lienard, B.
M.; Bray, J. E.; Savitsky, P.; Gileadi, O.; von Delft, F.; Rose, N. R.; Offer, J.;
Scheinost, J. C.; Borowski, T.; Sundstrom, M.; Schofield, C. J.; Oppermann, U.
Nature 2007, 448, 87.
11. Ilczyszyn, M. J. Chem. Soc., Faraday Trans. 1994, 90, 1411.
12. Maiti, N. C.; Carey, P. R.; Anderson, V. E. J. Phys. Chem. A 2003, 107, 9910.
13. Glutathione S-transferase (GST) fusion proteins of amino acids 2–350 of
JMJD2A or its H188A mutant, of amino acids 2–352 of JMJD2C or its H190A
mutant and of JMJD2D or its H192A mutant (Ref. 14) were produced in
Escherichia coli and purified on glutathione-agarose beads (Ref. 15). Before
lysed in 80
DTT, 1% triton X-100 (pH 7.1) supplemented with 20
pepstatin A, 4 g/ml aprotinin and 1 mM phenylmethylsulfonyl fluoride. After
l
l of 10 mM Tris, 30 mM Na4P2O7, 200 mM NaCl, 50 mM NaF, 2 mM
l
g/ml leupeptin, 2 g/ml
l
l
45 min on ice, an equal volume of 2Â Laemmli sample buffer was added, the
extracts boiled for 12 min, debris removed by centrifugation and supernatants
employed for SDS–PAGE and Western blotting as described above. The
following antibodies at the indicated dilutions were used: H3K9me3 (1:4000;
07-442, Upstate), H3K36me3 (1:4000; Ab9050, Abcam), H3 (1:500; sc-10809,
Santa Cruz), Actin (1:500; A2066, Sigma) and Flag (1:5000; M2, Sigma).
22. Fuchs, B.; Inwards, C. Y.; Janknecht, R. Clin. Cancer Res. 2004, 10, 1344.
23. Goel, A.; Janknecht, R. J. Biol. Chem. 2004, 279, 14909.
24. Shin, S.; Janknecht, R. Biochem. Biophys. Res. Commun. 2007, 359, 742.
25. Wu, J.; Janknecht, R. J. Biol. Chem. 2002, 277, 42669.
26. The purity of compounds 1–4 and 21 was certified by elemental analysis.
1ÁHCl: Anal. Calcd for C13H16N2O5ÁHClÁ1/4H2O: C, 48.60; H, 5.49; N, 8.72.
Found: C, 48.89; H, 5.80; N, 8.43. 2ÁHCl: Anal. Calcd for C14H18N2O5ÁHCl: C,
50.84; H, 5.79; N, 8.47. Found: C, 50.56; H, 5.67; N, 8.48. 3ÁHCl: Anal. Calcd for
C15H21N2O5ÁHClÁ1/10H2O: C, 51.91; H, 6.17; N, 8.07. Found: C, 52.25; H, 6.14; N,
8.12. 4ÁHCl: Anal. Calcd for C13H16N2O5ÁHCl: C, 49.30; H, 5.41; N, 8.84. Found: C,
49.13; H, 5.59; N, 8.64. 21: Anal. Calcd for C15H20N2O5: C, 58.43; H, 6.54; N,
9.09. Found: C, 58.35; H, 6.52; N, 9.05.