Glutathione-like tripeptides as inhibitors of glutathionylspermidine synthetase. Part 1: Substitution of the glycine carboxylic acid group

Article abstract of DOI:10.1016/S0960-894X(02)00489-4

Glutathionylspermidine synthetase/amidase (GspS) is an essential enzyme in the biosynthesis and turnover of trypanothione and represents an attractive target for the design of selective anti-parasitic drugs. We synthesised a series of analogues of glutathione (L-γ-Glu-L-Leu-Gly-X) where the glycine carboxylic acid group (X) has been substituted for other acidic groups such as tetrazole, hydroxamic acid, acylsulphonamide and boronic acid. The boronic acid appears the most promising lead compound (IC₅₀ of 17.2 μM).

Full text of DOI:10.1016/S0960-894X(02)00489-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2553–2556
Glutathione-like Tripeptides as Inhibitors of
Glutathionylspermidine Synthetase. Part 1: Substitution of the
Glycine Carboxylic Acid Group
Katie Amssoms,a Sandra L. Oza,b Esteban Ravaschino,c Abdellah Yamani,a
Anne-Marie Lambeir,d Padinchare Rajan,a Gunther Bal,a Juan Bautista Rodriguez,c
Alan H. Fairlamb,b Koen Augustynsa and Achiel Haemersa,*
aDepartment of Medicinal Chemistry, University of Antwerp (UIA), Belgium
bDepartment of Biochemistry, University of Dundee, Scotland, UK
cDepartamanto de Quimica Organica, FCEyN, Universidad de Buenos Aires, Argentina
dDepartment of Medical Biochemistry, University of Antwerp (UIA), Belgium
Received 3 April 2002; revised 15 May 2002; accepted 24 June 2002
Abstract—Glutathionylspermidine synthetase/amidase (GspS) is an essential enzyme in the biosynthesis and turnover of trypa-
nothione and represents an attractive target for the design of selective anti-parasitic drugs. We synthesised a series of analogues of
glutathione (l-g-Glu-l-Leu-Gly-X) where the glycine carboxylic acid group (X) has been substituted for other acidic groups such as
tetrazole, hydroxamic acid, acylsulphonamide and boronic acid. The boronic acid appears the most promising lead compound (IC50
of 17.2 mM). # 2002 Elsevier Science Ltd. All rights reserved.
The development of chemotherapy for the treatment of
parasitic diseases such as sleeping sickness, Chagas dis-
ease and leishmaniasis has been hindered by the close
similarities between parasite and host metabolism.
However, one major difference occurs in the biochem-
istry of defence mechanisms against oxidative damage.1
inhibitors of trypanothione reductase have been devel-
oped and the enzyme has been validated as a target for
the design of trypanocidal compounds. The properties
of trypanothione metabolism as a drug target have been
reviewed.3
The biosynthesis of trypanothione from glutathione and
spermidine is catalysed by glutathionylspermidine syn-
thetase (GspS) and trypanothione synthetase (TryS).4,5
As carbon–nitrogen ligases, both enzymes use ATP and
are proposed to form an acyl phosphate at the glycine
carboxylate of glutathione, activating it for nucleophilic
attack by the primary amines of spermidine.
Parasitic protozoa of the order Kinetoplastida are pro-
tected against damage by oxidants and toxic heavy
metals by a unique thiol-redox cycling.2 This system
uses trypanothione [N1,N8-bis(glutathionyl)spermidine]
and is analogous to the glutathione system that operates
in humans and almost all other aerobic organisms.
However, it offers a number of distinct features that
may be exploited for selective attack. Trypanothione is
the pivotal compound in a concerted cascade of redu-
cing equivalents from NADPH as initial donor to a
peroxide acceptor, requiring three distinct enzymes,
specific to the Kinetoplastida: trypanothione reductase,
tryparedoxin and tryparedoxin peroxidase. Several
GspS also contains an amidase domain, hydrolysing
glutathionylspermidine into glutathione and spermi-
dine. Similar to other ATP-dependent ligases, phospho-
nates,6,7 phosphonamidates,7 and phosphinates8 derived
from glutathione-like peptides are inhibitors of GspS
from Crithidia fasciculata and the analogous enzyme
from Escherichia coli. These compounds are thought to
behave as substrate analogues or mimic the proposed
tetrahedral intermediate formed during the attack of
spermidine on the acyl phosphate.
*Corresponding author. Tel.: +32-3-820-2717; fax: +32-3-820-2739;
e-mail: achiel.haemers@ua.ac.be
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00489-4

2554
K. Amssoms et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2553–2556
The results obtained with glutathione-derived tripepti-
dylphosphonic acids such as 1 (Ki of 60 mM-linear non-
competitive inhibition) on the isolated enzyme from C.
fasciculata6 prompted us to start the synthesis and eva-
luation of related analogues in which the C-terminal
carboxylic acid group is substituted with other acidic
groups (Fig. 1).
tetrazole (3), boronic acid (4), acylsulphonamide (5) and
hydroxamic acid (6) and non-acidic groups such as
amide (7) and a carboxylic ester (8).
All pseudopeptides (Scheme 1) were prepared in a simi-
lar way. N-Z-l-g-Glu(OBn)-l-Leu( 9) was prepared by
coupling N-Z-l-Glu-OBn and l-Leu-Otert-butyl with
TBTU and deprotection with TFA.
We confined the current series to l-g-Glu-l-Leutripep-
tides, since we knew from previous substrate-selectivity
studies that the g-glutamyl moiety is essential for
recognition by GspS and cannot be omitted. l-Leuwas
the best substitute for l-Cys in the phosphonopeptide
inhibitors and was used also in this series.6 Hence, tri-
peptide l-g-Glu-l-Leu-Gly (2) was used as lead peptide.
Carboxylic substitutes were acidic groups such as in
Tripeptide 2 was synthesized as described.9 Ester 8 was
prepared from 2 by DCC/HONSucatalyzed esterification.
The tetrazole peptide 3 was prepared from 5-amino-
methyl-1-benzyl-tetrazole (10), synthesized from N-Boc-
protected aminoacetonitrile.10 The carbonitrile function
reacted with sodium azide to a tetrazole ring which was
N1-protected with a benzyl group. Coupling with protected
l-g-Glu-l-Leuand deprotection by hydrogenolysis
afforded l-g-Glu-l-Leu-5-aminomethyl-tetrazole (3).
The boronic acid peptide 4 was prepared from 2-bis
(trimethylsilyl)aminomethyl-4,4,5,5-tetramethyl-1,3,2-
dioxaborolane (11). The latter was prepared by amina-
tion of a chloroalkylboronate (prepared from 2-iso-
propyloxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane) with
lithium hexamethyldisilazane11 and coupled to pro-
tected l-g-Glu-l-Leu( 9) with TBTU. Deprotection by
hydrogenolysis and acid work up afforded l-g-Gly-
l-Leu-aminomethylboronic acid HCl salt (4).
The peptide sulphonamide 5 was synthesized by the
reaction of BOP-activated Z-glycine and the corre-
sponding sulphonamide. Deprotection with HBr-HOAc
afforded acylsulphonamide 12, which was TBTU
coupled to protected l-g-Glu-l-Leu( 9). Final depro-
tection with hydrogenolysis afforded l-g-Glu-l-Leu-
GlyNHSO2 p-tolyl (5).
Figure 1. Analogues of glutathione (l-g-Glu-l-Leu-Gly-X).
Scheme 1. Synthesis of l-g-Glu-l-Leu-Gly-X peptides 3–6: (a) NaN3, NH4Cl/DMF; (b) BnBr, Et3N/CH2Cl2; (c) TFA/CH2Cl2; (d) BrCH2Cl, BuLi/
DMF; (e) lihexamethyldisilazane/THF; (f) BOP, NH2SO2R, Et3N/DMF; (g) HBr/HOAc; (h) NH2OBn, DCC, HOBt, Et3N/DMF, CH2Cl2; (i) TFA;
(j) TBTU, Et3N/DMF or (6) DPPA, Et3N/DMF; (k) H2, Pd/C or (3) H2, Pd/C, 5 h, 20 psi or (5) H2, Pd(OH)2/C (during the preparation of 4 the
free boronic acid is formed during acid workup).

K. Amssoms et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2553–2556
2555
Table 1. Substrate and inhibitory activity of l-g-Glu-l-Leu-NHCH2-X against glutathionylspermidine synthetase (GspS) and the amidase free
mutant
X
Wild-type enzyme
C79A mutant
% substratea % inhibitiona IC50 (mM) Ki (mM) Ki0 (mM) % substratea % inhibition IC50 (mM) Ki (mM) Ki0 (mM)
b
1
2
3
4
5
6
7
8
P¼O(OH)2
0
104c
0
1
2
44
93
102
86
0
58
95
2
76
0
0
60ꢀ9
NA
60ꢀ9
COOH
NA
NA
125
1
0
NA
NA
5-tetrazole
B(OH)2 HCl
CONHSO2p-Tol
CONHOH
CONH2 TFA
COOC4H9
138ꢀ12
17.2ꢀ0.8 81ꢀ35
18ꢀ2
2
98
3.2ꢀ0.3 2.5ꢀ0.1
105ꢀ9 NA
NA
0
22d
89d
26
3
a% activity compared to 0.5 mM GSH.
bValue taken from ref 6.
cApparent Km of 34ꢀ10 mM at concentrations <0.4 mM.
dThis compound behaves as a competitive substrate with a Km of 245ꢀ16 mM and a Vmax that is 15% of that obtained with glutathione.
NA, not applicable.
The hydroxamic acid 6 was prepared from N-(N-Boc-
glycyl)-O-benzylhydroxylamine (prepared with DCC/
HOBt).12 Boc-deprotection with TFA to N-glycyl-O-
benzylhydroxylamine (13), DPPA coupling with pro-
tected l-g-Glu-l-Leu( 9) and hydrogenolysis afforded l-
g-Glu-l-Leu-GlyNHOH (7). The TBTU reaction
between glycinamide and N-Z-l-g-Glu(OBn)-l-Leu( 9)
and deprotection furnished amide 7.
We conclude that substituting the carboxylic acid group
in l-g-Glu-l-Leu-Gly (2), a glutathione related sub-
strate of GspS, affords inhibitory compounds in ana-
logy to the already reported l-g-Glu-l-Leu-Gly-
phosphonic acid (1). The hydroxamate analogue 6 was
the most potent competitive inhibitor of the synthetase
(Ki 2.5 mM), but this was rapidly inactivated by hydro-
lysis by the amidase domain of the enzyme to form a
substrate analogue l-g-Glu-l-Leu-Gly. Since this pep-
tide is itself a substrate for the synthetase, this results in
a mixed substrate-inhibitor pattern. l-g-Glu-l-Leu-Gly-
boronic acid (4) appears the most promising inhibitor.
It is not inactivated by the amidase domain and behaves
as a mixed inhibitor of the synthetase (Ki0 and Ki of 18
and 81 mM, respectively). Other substitutions behave as
substrates, because they are hydrolysed by the amidase
domain of GspS, or are not active.
All compounds13 were tested as substrates and inhibi-
tors of recombinant C. fasciculata GspS purified from
E. coli14 (Table 1). Tripeptide 2 behaved as a substrate
under our assay conditions, whereas this had been
found to be an inhibitor when tested at 5 mM under the
conditions used by De Craecker et al.9 A more
detailed kinetic analysis revealed that compound 2 is
indeed a substrate at low concentrations but showed
pronounced substrate inhibition at higher concentra-
tions, reconciling this apparent discrepancy. Tetrazole
3 was not active as a substrate but was a poor inhi-
bitor with an IC50 of 138 mM. Boronic acid 4
behaved as a pure inhibitor, showing an IC50 of
about 17 mM. p-Tolylsulphonamide 5 did not show
any activity.
Boronic acid (4) shows stronger inhibitory properties
against GspS than the lead phosphonic acid compound (1).
Unfortunately, these compounds do not show any
trypanocidal nor leishmanicidal activity in vitro.
Our results, however, provide interesting data for the
further investigation of GspS as a target in anti-para-
sitic drug design. Efforts to improve the inhibitory
activity of these compounds and to develop pro-drugs
of these inhibitors are currently in progress.
Hydroxamic acid 6 behaved as a substrate as well as an
inhibitor of the wild-type enzyme. However, when
assayed against a mutant form of GspS (C79A) that is
devoid of amidase activity,15 the compound no longer
displayed activity as a substrate, but retained its potent
inhibitory properties. HPLC analysis confirmed that the
wild-type enzyme hydrolysed the hydroxamate moiety
of 6 to produce the free carboxylate analogue 2, which is
a substrate rather than an inhibitor. The t1/2 value of the
hydrolysis by the amidase domain of the wild type
enzyme was about 5.5 min with essentially complete
hydrolysis after 1 h.
Acknowledgements
This work received support from the European Com-
mission—INCO-DC, the Fund of Scientific Research
(Flanders, Belgium) and the Institute for Promotion of
Innovation by Science and Technology in Flanders
(I.W.T.). A.H.F. is supported by the Wellcome Trust.
Amide 7 and butyl ester 8 were also completely hydro-
lysed by the amidase domain within 1 h to form 2 and
therefore also behaved as substrates for the synthetase
domain of GspS. A longer ester was however more
slowly hydrolysed and showed weak activity as a
substrate as well as being a weak inhibitor.
References and Notes
1. Fairlamb, A. H.; Blackburn, P.; Ulrich, P.; Chait, B. T.;
Cerami, A. Science 1985, 227, 1485.

2556
K. Amssoms et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2553–2556
2. Fairlamb, A. H.; Cerami, A. Ann. Rev. Microbiol. 1992, 46,
695.
3. Augystyns, K.; Amssoms, K.; Yamani, A.; Rajan, P. K.;
Haemers, A. Curr. Pharm. Des. 2001, 7, 1117.
4. Smith, K.; Nadeau, K.; Bradley, M.; Walsh, C.; Fairlamb,
A. H. Protein Sci. 1992, 1, 874.
5. Tetaud, E.; Manai, F.; Barrett, M. P.; Nadeau, K.; Walsh,
C. T.; Fairlamb, A. H. J. Biol. Chem. 1998, 273, 19383.
6. Verbruggen, C.; De Craecker, S.; Rajan, P. K.; Jiao, X.-Y.;
Borloo, M.; Smith, K.; Fairlamb, A. H.; Haemers, A. Bioorg.
Med. Chem. Lett. 1996, 6, 253.
7. Chen, S.; Lin, C. H.; Kwon, D. S.; Walsh, C. T.; Coward,
J. K. J. Med. Chem. 1997, 40, 3842.
8. Chen, S.; Lin, C. H.; Walsh, C. T.; Coward, J. K. Bioorg.
Med. Chem. Lett. 1997, 7, 505.
9. De Craecker, S.; Verbruggen, C.; Rajan, P. K.; Smith, K.;
Haemers, A.; Fairlamb, A. H. Mol. Biochem. Parasitol. 1997,
84, 25.
10. Tilley, J. W.; Danko, W.; Loveu, K.; Wagner, R.; Swistok,
J.; Makofske, R.; Michalewski, J.; Triscari, J.; Neslon, D.;
Weatherford, S. J. Med. Chem. 1991, 34, 1125.
11. Martichonok, V.; Jones, B. J. Biorg. Med. Chem. 1997, 5,
679.
CH-Leu), 26.4 (b-CH2-Glu), 31.4 (g-CH2-Glu), 39.8 (b-CH2-
Leu), 51.4 (a-CH-Leu), 54.4 (a-CH-Glu), 63.0 (CH2), 174.1,
175.3, 176.2 (CO). MS (ES+) m/z 317, 318 [M+H]+. HPLC
(B) Rt: 23 min; purity: >99%.
(6)1H NMR (400 MHz) (D2O) d: 1.03 (d, 3H, J=5.9, d-
CH3-Leu), 1.08 (d, 3H, J=5.6, d-CH3-Leu), 1.70–1.90 (m, 3H,
b-CH2- and g-CH-Leu), 2.20–2.40 (m, 2H, b-CH2-Glu), 2.60–
2.75 (m, 2H, g-CH2-Glu), 3.92 (m, 1H, a-CH-Glu), 4.04 (2H,
m, CH2), 4.48 (m, 1H, a-CH-Leu). MS (ES+) m/z 333
[M+H]+. HPLC (B) Rt: 17.1 min; purity: >99%.
14. All compounds were initially tested as substrates and/or
inhibitors at 0.5 mM using recombinant wild-type or C79A
mutant C. fasciculata glutathionylspermidine synthetase-ami-
dase. The relative activity as a substrate is the activity expres-
sed as the % of activity measured with 0.5 mM glutathione.
The % inhibition was determined according to the following
equation:
ꢀ
ꢁ
vi
% inhibition ¼ 1 À
:100
vo
Initial velocity measurements were made at 8–10 inhibitor
concentrations, ranging from 0 to 2.5 mM, and 0.5 mM glu-
tathione. Spermidine (2 mM), Mg2+ (10 mM) and ATP (2
mM) were present in saturating conditions. The assays were
conducted in 100 mM HEPES buffer pH 7.3 at 25 ꢁC with 40
nM GspS. Assay mixtures were incubated for 5 min prior to
initiation with spermidine. IC50 values were determined
according to the following two parameter equation, where the
lower data limit is 0, that is the data are background corrected,
and the upper data limit is 100, that is the data are range
corrected.
12. Odake, S.; Okayama, T.; Obata, M.; Morikawa, T.; Hat-
tori, S.; Hori, H.; Nagai, Y. Chem. Pharm. Bull. 1990, 38,
1007.
13. All compounds gave satisfactory spectroscopic data: We
report the NMR, MS and HPLC data obtained with the most
active compounds. HPLC on RP18 with A: H2O–TFA (0.1%)
and B: CH3CN–TFA (0.1%) in a 35 min linear gradient (sys-
tem I: from 10 to 90% B or system II: from 0 to 20% B).
1
(3) H NMR (400 MHz) (CD3OD) d: 0.90 (d, 3H, J=5.63,
d-CH3-Leu), 0.94 (d, 3H, J=5.63, d-CH3-Leu), 1.54–1.74 (m,
3H, b-CH2- and g-CH-Leu), 2.05–2.16 (m, 1H, b-CH2-Glu),
2.17–2.28 (m, 1H, b-CH2-Glu), 2.45–2.61 (m, 2H, g-CH2-Glu),
3.68–3.87 (m, 1H, a-CH-Glu), 4.31–4.45 (m, 1H, a-CH-Leu),
4.60–4.80 (2H, m, CH2). 13C NMR (CD3OD) d: 21.7 (d-CH3-
Leu), 23.5 (d-CH3-Leu), 25.9 (g-CH-Leu), 27.5 (b-CH2-Glu),
32.5 (g-CH2-Glu), 41.5 (b-CH2-Leu), 53.5 (a-CH-Leu), 54.9
(a-CH-Glu), 58.3 (CH2), 130.0 (C), 175.0, 175.4 (CO). MS
(FAB+) m/z 342 [M+H]+. HPLC (A) Rt: 6.3 min; purity:
95%.
100
ꢀ
ꢁ
y ¼
s
x
1 þ
IC50
In this equation, s is a slope factor. The equation assumes that
y falls with increasing x.
Initial velocity measurements were determined at three
inhibitor concentrations and a serial dilution of glutathione,
ranging from 0.03 to 1 mM. From these data, Michaelis–
Menten and Lineweaver–Burk plots were constructed, and
kinetic inhibitory values were determined by non-linear least
squares regression analysis using the Grafit1 programme.
Inhibitors gave the best fit with mixed or competitive type
inhibition. The indicated errors are the standard errors on the
fit.
1
(4) H NMR (400 MHz) (D2O + two drops (CD3)2CO) d:
0.80 (d, 3H, J=5.36, d-CH3-Leu), 0.84 (d, 3H, J=5.98, d-
CH3-Leu), 1.42–1.69 (m, 3H, b-CH2- and g-CH-Leu), 1.92–
2.18 (m, 2H, b-CH2-Glu), 2.32–2.45 (m, 2H, g-CH2-Glu),
3.61–3.72 (m, 1H, a-CH-Glu), 4.25–4.35 (m, 1H, a-CH-Leu),
4.55–4.78 (m, 2H, CH2). 13C NMR (D2O + two drops
(CD3)2CO) d: 20.9 (d-CH3-Leu), 22.4 (d-CH3-Leu), 24.5 (g-
15. Oza, S. L.; Ariyanayagam, M. R.; Fairlamb, A. H. Bio-
chem. J. 2002, 364, 679.