Novel bivalent inhibitors with sub-nanomolar affinities towards human glyoxalase i

Article abstract of DOI:10.1016/j.bmcl.2015.08.055

The zinc metalloenzyme glyoxalase I (GlxI) catalyzes the glutathione-dependent inactivation of cytotoxic methylglyoxal. Two competitive bivalent GlxI inhibitors, polyBHG2-62 (K_i = 1.0 nM) and polyBHG2-54 (K_i = 0.3 nM), were synthesiz

Full text of DOI:10.1016/j.bmcl.2015.08.055

Accepted Manuscript
Novel bivalent inhibitors with sub-nanomolar affinities towards human glyox-
alase I
Yankui Sang, Qing Shi, Mingguang Mo, Caixia Ni, Zonghe Li, Bichong Liu,
Qishan Deng, Donald J. Creighton, Zhe-Bin Zheng
PII:
S0960-894X(15)30007-X
http://dx.doi.org/10.1016/j.bmcl.2015.08.055
BMCL 23049
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
30 June 2015
14 August 2015
20 August 2015
Please cite this article as: Sang, Y., Shi, Q., Mo, M., Ni, C., Li, Z., Liu, B., Deng, Q., Creighton, D.J., Zheng, Z-B.,
Novel bivalent inhibitors with sub-nanomolar affinities towards human glyoxalase I, Bioorganic & Medicinal
Chemistry Letters (2015), doi: http://dx.doi.org/10.1016/j.bmcl.2015.08.055
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.els evier.c om
Novel bivalent inhibitors with sub-nanomolar affinities towards human
glyoxalase I
a
a
a
a
a
a
a
b*
Yankui Sang , Qing Shi , Mingguang Mo , Caixia Ni , Zonghe Li , Bichong Liu , Qishan Deng , Donald J. Creighton ,
a*
Zhe-Bin Zheng
a
Department of Antibiotic Research & Re-evaluation, Key Laboratory of Sichuan Province, Sichuan Industrial Institute of Antibiotics, Chengdu University,
Huaguan Road 168, Chengdu City, Sichuan Province 610052, P.R. China
b
Department of Chemistry & Biochemistry, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, MD, 21250, USA
ARTICLE INFO
ABSTRACT
Article history:
Received
The zinc metalloenzyme glyoxalase I (GlxI) catalyzes the glutathione-dependent inactivation of
Revised
cytotoxic methylglyoxal. Two competitive bivalent GlxI inhibitors, polyBHG2-62 (K
i
= 1.0 nM)
Accepted
and polyBHG2-54 (K = 0.3 nM), were synthesized based on the transition-state analog
i
Available online
S-(N-bromophenyl-N-hydroxycarbamoyl) glutathione (BHG). The most effective inhibitor,
polyBHG2-54, is the first subnanomolar inhibitor of GlxI, and is over 50-fold more potent than
BHG itself.
Keywords:
Glyoxalase I
Bivalent inhibitor
linker
Methylglyoxal
1
2+
Human glyoxalase I (GlxI) is a 42 kDa dimeric Zn
molecules of the transition state analog S-(N-chlorophenyl-
N-hydroxycarbamoyl)glutathione (CHG) through their
γ-glutamyl-NH groups with poly-β-alanyl tethers of different
length: [CHG(β-ala) ] suberate diamide (n = 1−7). The
strongest inhibitors of this antitumor target enzyme likely bind
simultaneously to the active site on each subunit, and give K
values as low as 0.96 nM (n = 6), a 50-fold tighter binding than
metalloenzyme that detoxifies methylglyoxal in vivo by
converting it into S-D-lactoylglutathione, which is then
converted to D-lactate by glyoxalase II. Since high activities of
GlxI are present in tumor tissues, inhibitors of GlxI increase the
accumulation of cytotoxic methylglyoxal, which results in
significant anti-tumor activity both in vitro and in vivo. A potent
and selective GlxI inhibitor could therefore potentially result in
an adjuvant to restore chemotherapy sensitivity in tumor cells.
2
2
,3
n
i
4
5
,12
the monomer inhibitor CHG (K = 46 nM). Cross-linking not
i
5
,6
only improves the binding affinity, but also improves the
selectivity by almost 100-fold for human GlxI (hGlxI) relative to
yeast GlxI (yGlxI). In the X-ray crystal structure of the hGlxI
complex with CHG, the γ-glutamyl-NH groups are exposed to
2
solvent, and are about 30 Å apart. However, a 70−80 Å tether
Linking two identical binding groups by a spacer unit has
been proposed to improve both the selectivity and the activity of
inhibitors compared with the corresponding univalent ligands,
and this approach has been verified experimentally. However,
optimization of the spacer unit between the binding groups
remains a challenge, and is crucial in fragment-based drug
design. In this paper, we extend our previous work on bivalent
transition-state analog inhibitors of human glyoxalase I (hGlxI)
7
8
-10
length was found to give the best inhibition.
Nevertheless, these inhibitors have two drawbacks that need
to be addressed: amidation of the γ-glutamyl-NH groups
2
11
1
0
decreases the binding affinity by 7 to 13-fold, and the bivalent
inhibitors have low solubility, making them unable to cross the
cell membrane. In this work, we modify the linker by replacing
the amide groups with ethylcarbonyl groups, which do not affect
the binding affinity of CHG. Two new inhibitors based on
to include two new competitive bivalent GlxI inhibitors in which
symmetric ligands are linked by linkers that differ in flexibility,
length, and water solubility.
We previously developed a new class of competitive inhibitors
of homodimeric human glyoxalase I by cross-linking two
5,12
S-(N-bromophenyl-N-hydroxycarbamoyl) glutathione (BHG)
exhibit inhibition constants of 1 nM or less, and one of these
—
———
Corresponding author Tel.: +86-28-8421-5041(Z-B. ZH);
E-mail: zhebinzheng@hotmail.com
∗

Figure 1.The chemical structures of newly synthesized bivalent inhibitors.
compounds has a PEG linker, which increases water solubility
bis-(N-hydroxysuccinimide ester) was used to adjust the linker
and should1 result in an enhanced ability to cross cell
spacing.
3-15
membranes.
5,12
y-carbamoyl) glutathione (BHG, K = 14 nM) is used instead
of CHG because of its better binding to GlxI.
i
The two competitive bivalent GlxI inhibitors polyBHG2-62
(
K
i
= 1 nM) and polyBHG2-54 (K
i
= 0.3 nM) (Figure 1) were
designed based on the
transition-state analog
S-(N-bromophenyl-N-hydroxycarbamoyl) glutathione (BHG) by
an examination of the X-ray crystal structure of human GlxI in
complex with one CHG at each active site (Figure 2, PDB code
10
1
QIN). CHG is a stable transition-state analog that chelates the
active site zinc and mimics the transition-state. We chose a
central scaffold of poly β-ala and suberic acid
bis(N-hydroxysuccinimide ester) for the linkage of the two
binding sites.
In our previous work, amide groups were used to link the
CHG units to the spacer groups. However, amidation of CHG
decreases the affinity of the inhibitor, as illustrated in Figure 3.
Both compounds A (K
i
= 330 nM) and B (K
i
= 130 nM) have
Figure 2. The structure of human GlxI in complex with CHG
inhibition constants that are significantly higher than that of
CHG itself (46 nM). Addition of an ethylcarbonyl substituent
(ball and stick model)
to give compound C (K = 46 nM), however, does not decrease
i
the inbibition constant relative to CHG itself. These results
indicate that the carbonyl group on the substituent is crucial. On
the other hand, the unfavorable effect of a carbonyl group near
the nitrogen may be rationalized by the weakened hydrogen
bonds/electrostatic interactions observed in the X-ray structure
O
O
NH
2
NH
O
NH
NH
O
OH
OH
OH
O
O
O
O
O
O
O
O
O
1
0
O
NH
Ki = 330 nM
O
NH Ki = 130 nM
O
NH
Ki = 46 nM
OH
H
N
H
N
H
N
(
PDB code 1QIN).
HO
OH
N
HO
OH
HO
S
Cl
S
N
Cl
S
N
Cl
The bivalent transition state analogs incorporate a symmetric
O
O
O
scaffold based on 1 and 3 (Scheme 1), in which the terminal
carbonyls are protected by pyrrolidine-2,5-dione. The
A
B
C
intermediate
O,O-(dipropylamine)diethyleneglycol,
1
was used as
a
flexible linker with
and suberic acid
Figure 3. Variation of Ki with substituents on CHG
To synthesize two inhibitors with different distances between
the active groups, a building block strategy was used (Schemes 2
The bivalent inhibitors PolyBHG2-62 and polyBHG2-54
exhibit much tighter binding ability than BHG itself, with K
i
1
7
and 3). PolyBHG2-62 was synthesized in five steps, as outlined
values of 1.0 nM and 0.3 nM, respectively. The tightest binding
compound (polyBHG2-54) is more than 50-fold more potent
than BHG, and three-fold more potent than the best bivalent
18
in Scheme 2, and synthesis of polyBHG2-54 followed a similar
strategy (Scheme 3). In Scheme 3, O,O-(dipropylamine)
diethyleneglycol was replaced by poly β-ala to generate the
more rigid intermediate 3, which was used to prepare
polyBHG2-54, in which the length of the linker is different from
PolyBHG2-62. For these inhibitors S-(N-bromophenyl-N-hydrox
6
inhibitor we obtained previously. Modification of the linker by
using ethylcarbonyl groups instead of amide groups on the BHG
results in the strongest inhibitor yet reported for hGlxI.
Importantly, use of the PEG linker in polyBHG2-62 retains most
of the inhibition toward GlxI, while providing enhanced water

solubility. It should be noted that, compounds polyBHG2-62 and
polyBHG2-54 have similar solubilities (26 mg/mL and 32
mg/mL, respectively). This combination of enhanced water
solubility and strong binding is an important step toward the
development of inhibitors for GlxI that can cross cell
membranes.
BHG2-54 binds about 12,300-fold more tightly. A comparison of
the inhibition constants of CHG and poly BHG2-54 for hGlxI
versus bGlxII shows that cross-linking increases binding
selectivity 8.6-fold, from 37-fold to 320-fold.
Table 1
Competitive inhibition constant (Ki) of CHG with hGlxI, yeast
GlxI (yGlxI), and bovine liver GlxII (bGlxII)
Table 1, shows the selectivities of the newly synthesized
compounds with different glyoxalase I enzymes, including
hGlxI, yeast GlxI (yGlxI), and bovine liver GlxII (bGlxII). CHG
was used as a positive control. Surprisingly, the two target
compounds exhibited higher levels of inhibition than CHG in the
hGlxI control. Poly BHG2-62 (Ki 1.0 nM) and poly BHG2-54
a
Ki, hGlxI
Ki, yGlxI
(µM)
3.6 ± 0.3
6.9 ± 0.2
3.7 ± 0.1
Ki, bGlxI
(nM)
1700 ± 17
138 ± 2
81 ± 5
Compound
(nM)
46 ± 4
d
d
CHG
poly BHG2-62
poly BHG2-54
1.0 ± 0.02
0.3 ± 0.02
(
Ki 0.3 nM) are almost 50-fold more active and 156-fold more
a
Conditions: sodium phosphate buffer, 50 mM, PH 7, 25 °C.
Mean (± SD) for triplicate determinations.
Mean (± SD) for duplicate determinations.
active than CHG, respectively. In addition, CHG binds 78-fold
less tightly to yGlxI than to hGlx1. Thus, cross-linking increases
the inhibitor selectivity by approximately 158-fold, as CHG
binds 78-fold more tightly to hGlxI than to yGlx, while poly
b
c
d
Ref 12.
Scheme 1. Preparation of symmetric scaffold 1 and 3: Reagents and conditions: (a) suberic acid bis (N-hydroxysuccinimide ester), DMF, 0 °C, 57%; (b) adipoyl
chloride, diisopropylethylamine (DIPEA), CH Cl , 0 °C, 48%; (c) ethylenediamine, DMF, 0 °C, 6%.
2
2
Scheme 2. Preparation of poly BHG2-62. Reagents and conditions: (a) acryloyl chloride, DIPEA, DMF, 0 °C, 35%; (b) BHG, DIPEA, DMF, 21%; (c) TFA,
3%; (d) 1, DIPEA, DMF, 72%.
5

O
O
H
N
O
b
H
a
NHBoc
NHR2
N
HO
BocHN
O
N
O
O
O
O
R2 = Boc
R2 = H
8
9
7
c
O
H
N
NH2
2
HN
O
O
6
d
H
H
OH
O
e
N
N
NHBoc
O
O
O
O
O
1
O
1
1
0
O
NH
H
N
HO
OH
N
S
Br
O
H
N
NH2
3
O
HN
O
OH
O
g
f
polyBHG2-54
O
O
O
1
2
O
NH
H
N
HO
OH
N
S
Br
O
Scheme 3. Preparation of polyBHG2-54. Reagents and conditions: (a) ethanolamine, DMF; (b) acryloyl chloride, DIPEA, DMF, 0 °C, 53%; (c) TFA; (d)
N-t-Boc-β-ala-N-hydroxysuccinimide ester, DIPEA, DMF, 55%; (e) 1) BHG, DIPEA, DMF; 2) TFA, 5.2%; (f) 1) N-t-Boc-β-ala-N-hydroxysuccinimide ester ,
DIPEA, DMF; 2) TFA, 25%; (g) 3, DIPEA, DMF, 11%.
Acknowledgements
8.
9.
Cox, K. J.; Shomin, C. D.; Ghosh, I. Future Med. Chem. 2011, 3,
9.
Du, D. M.; Paul R. Curr. Pharm. Design. 2004, 10, 3141.
2
This work is partially supported by the Natural Science
Foundation of China (81273353), and the National Science &
Technology Key Projects of China (2009ZX09103-085).
10. Zheng, Z. B.; Creighton, D. J. Org. Lett. 2003, 5, 4855.
1
1
1. Hajduk, P. J.; Greer, J. Nat. Rev. Drug Discov. 2007, 6, 211.
2. Murthy, N. S.; Bakeris, T.; Kavarana, M. J; Hamilton, D. S.; Lan,
Y.; Creighton, D. J. J. Med. Chem. 1994, 37, 2161.
3. Lyttle, M. H. Current Protocols in Nucleic Acid Chemistry. 2000,
4.6.1.
1
Supplementary data
1
4. Hill, Z. B.; Perera, B. G.; Maly, D. J. Mol. BioSyst. 2011, 7, 447.
Supplementary data associated with this article can be found in
the online version.
15. Shan, M.; Carlson, K. E.; Bujotzek, A.; Wellner, A.; Gust, R.;
Weber, M.; Katzenellenbogen, J. A.; Haag, R. Chem. Biol. 2013,
8
, 707.
1
1
6. More, S. S.; Vince, R. J. Med. Chem. 2009, 52, 4650.
7. Spectral data of PolyBHG2-62: H NMR (500 MHz, D
1
2
O Tris
buffer, δ), 7.67 (d, J =8.3 Hz, 4H), 7.54 (d, J =8.3 Hz, 4H), 4.53
m, 2H), 3.76 (m, 41H), 3.64 (m, 9H), 3.45 (t, J =5.2 Hz, 4H),
.22 (t, J = 6.7 Hz, 6H), 3.18 (m, 5H), 2.60 (m, 6H), 2.32 (m,
References and notes
(
3
+
1
2
.
.
Thornalley, P. J. Biochem. Soc. Trans. 2003, 31, 1343.
Cameron, A. D.; Olin, B.; Ridderström, M; Mannervik, B.; Jones,
T. A. EMBO J. 1997, 16, 3386.
10H), 2.24 (m, 2H). ESI-HRMS m/z 2031.6582 (M+H) ,
C H Br N O S +H requires 2031.6588.
+
8
2
128
2
12 33 2
1
1
2
8. Spectral data of PolyBHG2-54: H NMR (500 MHz, D O Tris
3
.
Creighton, D. J.; Hamilton, D. S. Arch. Biochem. Biophy. 2001,
buffer, δ), 7.69 (d, J = 8.7 Hz, 4H), 7.55 (d, J = 8.7 Hz, 4H), 4.75
(dd, J = 7.9, 4.12 Hz, 2H), 4.27 (m, 2H), 3.86 (d, J = 17.3Hz, 2H),
3
87, 1.
4
5
.
.
Thornalley, P. J.; Rabbani, N. Semin Cell Dev Biol. 2011, 22, 318.
Creighton, D. J.; Zheng, Z. B.; Holewinski, R.; Hamilton, D. S.;
Eisemant, J. L. Biochem. Soc. Trans. 2003, 31, 1378.
Creighton, D. J.; Hamilton, D. S.; Kavarana, M. J.; Sharkey, E.
M.; Eiseman, J. L. Drugs Future. 2000, 25, 385.
3
5
.85 (d, J = 17.3Hz, 2H), 3.72 (m, 7H), 3.49 (m, 12H), 3.38 (m,
H), 3.22 (m, 8H), 2.49 (m, 19H), 2.30 (m, 9H), 2.19 (m, 4H),
1.38 (m, 4H), 1.23 (m, 2H), 0.96 (m, 2H). ESI-HRMS m/z
+
+
6
7
.
.
1
2 18 30 2
977.5521 (M+H) , C76H110Br N O S +H requires 1977.5516.
Maeda, D. Y.; Mahajan, S. S.; Atkins, W. M.; Zebala, J. A.
Bioorg. Med. Chem. Lett. 2006, 16, 3780.

Graphical Abstract
Abstract—The zinc metalloenzyme glyoxalase I (GlxI) catalyzes the glutathione-dependent inactivation of cytotoxic
methylglyoxal. Two competitive bivalent GlxI inhibitors, polyBHG2-62(K
i
= 1.0 nM) and polyBHG2-54 (K = 0.3 nM), were
i
synthesized based on the transition-state analog S-(N-bromophenyl-N-hydroxycarbamoyl) glutathione (BHG). The most effective
inhibitor, polyBHG2-54, is the first subnanomolar inhibitor of GlxI, and is over 50-fold more potent than BHG itself.
Novel bivalent inhibitors with sub-nanomolar affinities
towards human glyoxalase I
Yankui Sang, Qing Shi, Mingguang Mo, Caixia Ni, Zonghe Li, Bichong Liu, Qishan
Deng, Donald J. Creighton, Zhe-Bin Zheng