A metabolically stable tight-binding transition-state inhibitor of glyoxalase-I

Article abstract of DOI:10.1016/j.bmcl.2006.08.121

The design, synthesis, and enzyme kinetics evaluation of a transition-state inhibitor of glyoxalase-I is described. The union of the hydroxamic acid zinc-chelator with a urea isostere for the glu-cys amide bond led to a glutathione analog which retained inhibitory potency toward glyoxalase-I while possessing resistance toward γ-glutamyltranspeptidase mediated breakdown. This compound is viewed as a potential lead for the development of second-generation glyoxalase-I inhibitors wherein, the problems pertaining to metabolism and selectivity are overcome.

Full text of DOI:10.1016/j.bmcl.2006.08.121

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6039–6042
A metabolically stable tight-binding transition-state
inhibitor of glyoxalase-I
Swati S. More and Robert Vince*
Center for Drug Design, Academic Health Center, and Department of Medicinal Chemistry, College of Pharmacy,
University of Minnesota, 8-123A WDH, 308 Harvard St SE, Minneapolis, MN 55455, USA
Received 28 July 2006; revised 29 August 2006; accepted 29 August 2006
Available online 25 September 2006
Abstract—The design, synthesis, and enzyme kinetics evaluation of a transition-state inhibitor of glyoxalase-I is described. The
union of the hydroxamic acid zinc-chelator with a urea isostere for the glu–cys amide bond led to a glutathione analog which
retained inhibitory potency toward glyoxalase-I while possessing resistance toward c-glutamyltranspeptidase mediated breakdown.
This compound is viewed as a potential lead for the development of second-generation glyoxalase-I inhibitors wherein, the problems
pertaining to metabolism and selectivity are overcome.
ꢀ 2006 Elsevier Ltd. All rights reserved.
O
The aldehyde of pyruvic acid, methylglyoxal, has been
the subject of a large number of investigations due to
its cytotoxicity.1 This cytotoxic metabolite is transiently
H
H3C
Methyl Glyoxal
O
formed in a myriad of metabolic pathways, for example,
lipid peroxidation, glycolysis, and DNA metabolism.2
The removal of this ketoaldehyde from the body occurs
through the glyoxalase pathway involving the cofactor
O
G-SH
Glutathione
Glyoxalase-II
glutathione. Two key enzymatic steps are involved in
this process (Fig. 1): the isomerization of the hemi-
thioacetal of methylglyoxal and glutathione (GSH) to
S-D-lactoylglutathione (by glyoxalase-I) and its subse-
quent glyoxalase-II catalyzed transformation to D-lac-
tate, that is then actively transported out of the cell
and oxidized back to pyruvate which enters Kreb’s cy-
cle.3 This cascade limits the use of a-ketoaldehydes as
potential antitumor agents.
H3C
Hemithioacetal
H
S
HO
G
Glyoxalase-I
H
OH
OH
H3C
H
-
OH
OH
O
H3C
O
H3C
GS
H3C
OH
-
S
O
D-Lactate
G
O
GS
S-D-Lactoyl glutathione
Enediol(ate) Intermediate
Figure 1. Glyoxalase pathway.
In 1969, our laboratory was the first to propose the
obstruction of the glyoxalase pathway as a means to
cause accumulation of methylglyoxal in cancer cells,
eventually leading to their death.4 We subsequently
reported the competitive nature of S-alkylglutathione
analogs toward glyoxalase-I (Glx-I)5 and the discovery
of S-p-bromobenzyl glutathione (PBBG, 1) (Fig. 2) as
a highly potent Glx-I inhibitor.6 Because of their
charged nature, these inhibitors lacked sufficient cell
penetration to exhibit antitumor activity. In addition,
loss of enzyme inhibition occurred through the rapid
hydrolysis of the glu–cys amide bond of the glutathione
scaffold by c-glutamyltranspeptidase7 (Fig. 2). Years lat-
er Lo and Thornalley used ester prodrugs of our PBBG
(1), which were able to penetrate cell membranes and
inhibit the growth of human leukemia 60 (HL-60) cells.8
Keywords: Methylglyoxal; Glyoxalase; c-Glutamyltranspeptidase;
Glutathione; Hydroxamic acid; Transition-state inhibitor.
One of the shortcomings of these S-alkylglutathiones,
otherwise-promising Glx-I inhibitors, was their inability
*
Corresponding author. Tel.: +1 612 6249911; fax: +1 612 6240139;
e-mail: vince001@umn.edu
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.121

6040
S. S. More, R. Vince / Bioorg. Med. Chem. Lett. 16 (2006) 6039–6042
Br
S
O
OOC
O
N
H
NH3
HN COO
Br
Br
S
S
O
O
OOC
O
ROOC
O
N
H
N
H
N
H
NH3
HN COO
NH3
HN COOR
γ-Glutamyl
Transpeptidase
O
Br
Cell
Br
OOC
S
O-
NH3
O
+H3N
S
O
N
HN COO
OOC
O
NH3 HHN COO
Figure 2. Problems with existing glutathione-based glyoxalase-I inhibitors.
to selectively inhibit tumor cells when compared to nor-
mal human cells. This suggested that simple competitive
inhibitors of glyoxalase are perhaps inadequate tumor-
selective agents. This shortcoming was addressed by
Creighton and Murthy9 through the design of hydroxa-
mic acid-based transition state inhibitors (3, for exam-
ple). The rationale behind this design was thus: the
hydroxamic acid moiety acts as a close mimic of the ene-
diol(ate) intermediate10 of the Glx-I catalytic cycle
(Fig. 1), thus conferring upon these molecules a high
Glx-I inhibitory potency. Incorporation of a thioester
linkage into these compounds caused them to resemble
the glyoxalase-II (Glx-II) substrate (S-D-lactoylglutathi-
one, Fig. 1). Since the Glx-II activity is abnormally low
in certain types of cancer cells11,12 compared to normal
cells, Creighton and Murthy9,10 suggested that the inac-
tivation of compounds like 3 by Glx-II in tumor cells
will be very low, thus rendering the desired tumor-selec-
tivity from their reduced ability to hydrolyze these inhib-
itors. Although these inhibitors were in fact able to
restrict the growth of solid tumors in mice when admin-
istered as their ester prodrugs, the problem of their
breakdown by c-glutamyltranspeptidase still remained.
Recently, we initiated a project to revisit the design and
synthesis of these inhibitors with an aim to address var-
ious pharmacokinetic, metabolic, and synthetic prob-
lems that plague the development of this class of
molecules into clinically useful antitumor agents.13 In
this report, we wish to describe the application of our
urea-isostere strategy to prevent breakdown of the
hydroxamic acid inhibitors by c-glutamyltranspepti-
dase. Incorporation of the urea linkage and the
hydroxamic acid transition-state mimic into 3 gave the
tripeptide 4 (Fig. 3). Here, the urea-isostere was expect-
ed to provide metabolic stability to the glyoxalase-I
Br
OH
O
N
S
S
O
Br
O
O
HOOC
HOOC
O
N
X
N
H
H
NH2
HN
COOH
NH2
HN
COOH
1:X = CH2
2:X = NH
3
Increased resistance
towards enzymatic
break down
Tumor selectivity
OH
N
O
S
Br
O
O
HOOC
N
N
H
H
NH2
HN
COOH
4
Figure 3. Rationale for the design of 4.

S. S. More, R. Vince / Bioorg. Med. Chem. Lett. 16 (2006) 6039–6042
6041
inhibitor 4 along with conservation of glyoxalase-I
inhibitory activity.13 A good body of experimental evi-
dence exists for the presence of a significantly large
hydrophobic pocket in the active site of Glx-I, with
which the S-p-bromobenzyl substituent of PBBG (1)
interacts.5,14 This led us to conserve that particular
pharmacophore. We then set forth to develop an effi-
cient synthetic route to the tripeptide backbone and
the thiocarbamate linkage of compound 4.
tion (‘diimide reduction’) of 4-bromo nitrobenzene.18
Insertion of a carbonyl between thiol 13 and p-bro-
mophenylhydroxy amine (18) was accomplished by
sequentially treating 13 with phosgene (to form the thio-
chloroformate in situ) and then with 18 in dichlorometh-
ane. Under these conditions a 51% yield of the
thiocarbamate 14 was obtained. Global deprotection
of 14 with trifluoroacetic acid afforded the target mole-
cule 4.19
The synthesis of 4 is shown in Scheme 1. The N-a-t-Boc-
L-diaminopropanoic acid-t-butyl ester (9) fragment was
synthesized in four steps from Boc-L-asparagine (5).
The key step in this synthesis was a Hofmann rearrange-
ment of the terminal carboxamide, which was performed
according to the elegant protocol of Zhang15 that em-
ploys iodosobenzene diacetate (PIDA) as the oxidant.
The primary amine 6 thus formed was subjected to Cbz
protection and the free carboxylic acid was esterified
with t-BuOH and DCC to give 8. Hydrogenolysis of
the Cbz group afforded amine 9. Formation of the urea
linkage between 9 and L-cysteine(S–S-t-Bu) allyl ester
hydrochloride (16) was accomplished through the use
of carbonyl diimidazole (CDI).16 The S–S-tert-butyl-
protecting group for cysteine was chosen from the view-
point of orthogonal deprotection. Pd0 catalyzed depro-
tection of the allyl ester in urea dipeptide 10 gave the
free acid 11 which was coupled with glycine tert-butyl es-
ter hydrochloride in the presence of EDC, HOBt, and N-
methylmorpholine (NMM), to give the tripeptide 12 in
78% yield. Reductive cleavage of the disulfide protection
in 12 by tributylphosphine17 provided the thiol 13.
Compound 4 was examined for its ability to inhibit yeast
glyoxalase-I. Initial rates were determined by following
the increase in UV absorption at 240 nm (0.05 M phos-
phate buffer, pH 6.6) which corresponds to the isomeri-
zation process and formation of S-D-lactoyl glutathione
(Fig. 1). Methylglyoxal, GSH, 4, and buffer were added
to the cell and allowed to equilibrate at 30 ꢁC for 6 min
(to allow formation of hemimercaptal) before addition
of the enzyme. The concentrations of hemimercaptal
were calculated using the dissociation constant
3.1 · 10À3 M as previously determined for this equilibri-
um reaction.6 It has been shown previously that 3 is a
tight-binding competitive inhibitor of glyoxalase-I
(Ki = 1.2 0.2 lM).9 Similarly the urea analog 4 was
found to be a tight-binding competitive inhibitor with
Ki = 2.19 0.57 lM. The Km for the hemimercaptal of
glutathione and methylglyoxal as determined by this as-
say was 0.35 mM, in agreement with the literature val-
ue.13 Results of these experiments confirm that the
urea isostere is well tolerated by glyoxalase-I.
We then tested the stability of the urea linkage in 4 toward
c-glutamyltranspeptidase. Compound 4 was incubated at
37 ꢁC with equine kidney c-glutamyltranspeptidase (c-
GT, Sigma) in 200 mM AMPD buffer at pH 8.4 in the
p-Bromophenylhydroxylamine (18, an unstable solid)
was synthesized by a Rh-catalyzed transfer hydrogena-
S-StBu
O
COOR1
COOH
a
g
e
BuOtOC
BocHN
N
N
H
BocHN
COOR
NHR2
H
NHBoc
H2N
O
6 R1 =H , R2 =H
10
5
R= Allyl
f
b
7 R1 =H , R2 =C bz
11 R= H
c
8 R1 = tBu, R2 =C bz
9 R1 = tBu, R2 =H
SR
OH
N
d
O
O
S
H
Br
O
BuOtOC
COOtBu
N
i
N
N
BuOtOC
O
H
H
N
N
NHBoc
O
H
H
HN
COOtBu
NHBoc
12 R= S-StBu
13 R= H
14
h
j
ButS-S
OH
N
ButS-S
AllOOC
k, l
O
HOOC
16
15
S
NH2
NHBoc
Br
O
HOOC
O
N
N
H
H
NH3+
CF3COO-
HN
COOH
m
4
Br
NHOH
Br
NO2
18
17
Scheme 1. Synthesis of 4. Reagents and conditions: (a) PIDA, EtOAc/CH3CN/H2O (2:2:1), 67%; (b) CbzCl, KOH, K2CO3, THF/H2O (3:1), 94%; (c)
t-BuOH, DCC, DMAP, CH2Cl2, 84%; (d) H2, 10% Pd/C, CH3OH, 94%; (e) 16, CDI, NMM, CH2Cl2, 67%; (f) (PPh3)4Pd, morpholine, CH2Cl2, 80%;
(g) HClÆGlyOtBu, EDC, HOBt, NMM, CH2Cl2, 78%; (h) Bu3P, THF/H2O, 94%; (i) 18, COCl2, DIPEA, CH2Cl2, 0 ꢁC to rt, 51%; (j) 50% TFA in
CH2Cl2, 89%; (k) allyl bromide, benzene, reflux, 86%; (l) 4N HCl/dioxane, quant.; (m) 5% Rh–C, NH2NH2ÆH2O, THF, 71%.

6042
S. S. More, R. Vince / Bioorg. Med. Chem. Lett. 16 (2006) 6039–6042
3. (a) Thornalley, P. J. Biochem. J. 1990, 269, 1; (b)
Hertz, L.; Gerald, A. D. J. Neurosci. Res. 2004, 79,
11–18.
4. Vince, R.; Wadd, W. B. Biochem. Biophys. Res. Commun.
1969, 35, 593.
5. Vince, R.; Daluge, S. J. Med. Chem. 1971, 14, 35.
6. Vince, R.; Daluge, S.; Wadd, W. B. J. Med. Chem. 1971,
14, 402.
7. Vince, R.; Wolff, M.; Sanford, C. J. Med. Chem. 1973, 16,
951.
8. Lo, T. W. C.; Thornalley, P. Biochem. Pharmacol. 1992,
44, 2357.
9. Murthy, N. S. R. K.; Bakeris, T.; Kavarana, M. J.;
Hamilton, D. S.; Lan, Y.; Creighton, D. J. J. Med. Chem.
1994, 37, 2161.
10. Hamilton, D. S.; Creighton, D. J. J. Biol. Chem. 1992, 267,
24933.
11. Jerzykowski, T.; Winter, R.; Matuszewski, W.; Piskorska,
D. A. Int. J. Biochem. 1978, 9, 853–860.
12. (a) Ayoub, F.; Zaman, M.; Thornalley, P.; Masters, J.
Anticancer Res. 1993, 13, 151; (b) Antognelli, C.; Bald-
racchini, F.; Talesa, V. N.; Elisabetta, C.; Zucchi, A.;
Mearini, E. Cancer J. 2006, 12, 222–228.
13. Vince, R.; Brownell, J.; Akella, L. B. Bioorg. Med. Chem.
Lett. 1999, 9, 853.
14. Kalsi, A.; Kavarana, M. J.; Lu, T.; Whalen, D. L.;
Hamilton, D. S.; Creighton, D. J. J. Med. Chem. 2000, 43,
3981.
15. Zhang, L.; Kauffman, G. S.; Pesti, J. A.; Yin, J. J. Org.
Chem. 1997, 62, 6918.
16. Zhang, X.; Rodrigues, J.; Evans, L.; Hinkle, B.; Ballan-
tyne, L.; Pena, M. J. Org. Chem. 1997, 62, 6420.
17. Zhu, J.; Hu, X.; Dizin, E.; Pei, D. J. Am. Chem. Soc. 2003,
125, 13379.
18. Oxley, P. W.; Adger, B. M.; Sasse, M. J.; Forth, M. A.
Org. Syn. 1993, 8, 16.
presence of 40 mM of gly–gly as an acceptor peptide for
the released glutamic acid. The course of this incubation
was monitored by thin-layer chromatography. Excellent
separation of the metabolic products was obtained on sil-
ica gel TLC plates that were developed with n-butanol/
acetic acid/water (12:5:3) and sprayed with a solution of
fluorescamine in acetone. Glx-I inhibitors PBBG (1) and
3 were used as references in this assay. The results of this
experiment indicate that PBBG (1) and 3 were significant-
ly degraded after 15 min of incubation with c-GT. TLC of
incubation aliquots containing PBBG (1) and 3 showed a
new higher Rf spot corresponding to complete degrada-
tion to the respective dipeptides (cysteinylglycine) and a
new lower Rf spot corresponding to c-glutamylglycylgly-
cine, the other expected product from the transpeptidase
reaction.13 The identities of these spots were confirmed
by comparison with authentic samples of these dipep-
tides. Compound 3 after 1 h showed a further degradation
of cysteinylglycine dipeptide by release of the hydroxyl-
amine portion (a higher Rf spot). In contrast to the above,
compound 4 did not show any degradation even after
incubation for 15 h, thus confirming its stability against
c-GT mediated cleavage.
In summary, we have developed an efficient synthetic
route to the required urea-isostere containing hydroxa-
mic acid-based inhibitor 4. The target molecule, 4, was
found to retain the inhibitory potency of the corre-
sponding carbo-analog 3 against glyoxalase-I while
possessing resistance to cleavage by c-glutamyl
transpeptidase. The design of metabolically stable
glyoxalase-I inhibitors based on 4 may be useful in
potentiating the antitumor activity of a-ketoaldehydes.
We are currently designing and synthesizing ester pro-
drugs of 4 that would potentially possess the ability to
penetrate cell membranes and thus render 4 suitable for
testing against tumor cells.
19. Spectral data of compound 4: 1H NMR (300 MHz,
CD3OD-CF3COOD) d ppm 7.59 (d, J = 9.0, 2H, Ar),
7.49 (d, J = 8.7 Hz, 2H, Ar), 4.60 (q, J = 4.8, 8.4 Hz, 1H,
a-CH:Cys), 4.48 (q, J = 4.8, 7.8 Hz, 1H, a-CH:Dap), 4.17
(m, 2H, CH2:Gly), 3.68–3.62, 3.41–3.34 (2m, 2H, b-
CH2:Dap), 3.24–2.98 (m, 2H, b-CH2:Cys); 13C NMR
(75 MHz, CD3OD-CF3COOD) d 174.5, 172.6, 170.5,
159.2, 157.1 (C@O), 140.6 (CArNOH), 131.5 (CAr: ortho
to Br), 121.1 (CArBr), 117.7 (CAr: ortho to NOH), 52.4 (a-
C:Cys), 49.1 (a-C:Dap), 44.3 (b-C:Dap), 42.6 (CH2:Gly),
31.5 (b-C:Cys); ESI-HRMS m/z 522.0276 (M+H)+;
C16H20BrN5O8S + H+ requires 522.0294; reverse-phase
HPLC was run on Varian Microsorb column (C18, 5 l,
4.6 · 250 mm) using two solvent systems with 0.5 mL/min
flow rate and detected at 254 nm. Solvent system 1: 0.04 M
TEAB (triethylammonium bicarbonate) in water/70%
acetonitrile in water = 1/1, tR = 5.74 min, purity = 99.4%.
Solvent system 2: 0.04 M TEAB in water/70% acetonitrile
Acknowledgments
The authors thank Dr. Diana S. Hamilton (Late Prof.
Donald Creighton’s laboratory) at the University of
Maryland for generously providing us a sample of
compound 3.
References and notes
1. French, F. A.; Freelander, B. L. Cancer Res. 1958, 18, 172.
2. Kalapos, M. P. Toxicol. Lett. 1999, 110, 145.
in water = 10–100%
purity = 98.07%.
B
linear,
tR = 17.52 min,