Synthesis of S-alkyl L-homocysteine analogues of glutathione and their kinetic studies with γ-glutamyl transpeptidase

Article abstract of DOI:10.1016/j.bmcl.2004.04.072

A series of S-alkyl L-homocysteine analogues of glutathione was synthesized with varied oxidation state of the sulfur and tested for inhibition of rat kidney γ-glutamyl transpeptidase (GGT). The strong selectivity of the enzyme with respect to the sulfur oxidation state reveals important information for the development of powerful competitive inhibitors.

Full text of DOI:10.1016/j.bmcl.2004.04.072

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3451–3455
Synthesis of S-alkyl L-homocysteine analogues of glutathione
and their kinetic studies with c-glutamyl transpeptidase
Christian Lherbet, Christian Gravel and Jeffrey W. Keillor*
Department of Chemistry, Universitꢀe de Montrꢀeal, C.P. 6128, Succ. Centre-ville, Montrꢀeal, Que., Canada H3C 3J7
Received 20 February 2004; accepted 21 April 2004
Available online
Abstract—A series of S-alkyl L-homocysteine analogues of glutathione was synthesized with varied oxidation state of the sulfur and
tested for inhibition of rat kidney c-glutamyl transpeptidase (GGT). The strong selectivity of the enzyme with respect to the sulfur
oxidation state reveals important information for the development of powerful competitive inhibitors.
Ó 2004 Elsevier Ltd. All rights reserved.
c-Glutamyl transpeptidase (GGT, EC 2.3.2.2) is a
highly glycosylated heterodimeric enzyme bound to the
Cl
O
H3N
extracellular membrane of cells. It can be found in
mammals,1 bacteria2 and plants3 and is present in high
concentrations in kidney, liver and brain1. It has been
implicated in cellular detoxification processes, leukotri-
ene biosynthesis, anti-cancer prodrug activation and
physiological disorders including Parkinson’s disease4
and perturbation of apoptosis.5;6 This enzyme also plays
a key role in the metabolism of glutathione (GSH), its in
vivo substrate, by catalyzing its cleavage between the c-
glutamyl and Cys-Gly moieties. In the first step of its
catalytic cycle, GGT is acylated by transfer of a c-
glutamyl moiety from a donor substrate.7 This acyl
group is then transferred during the deacylation step to
acyl acceptor substrates such as a-amino acids and
peptides (transamidation), or to a water molecule
(hydrolysis). The important residues responsible for
catalysis by mammalian GGT are not well known. The
catalytic nucleophile of bacterial GGT has been identi-
fied by mass spectrometry as Thr-391, the N terminal
residue of the small subunit. This residue is conserved in
all GGTs, suggesting that GGT could be a member of
the Ntn-hydrolase family.8
OH
O
H
N
S
n
O
OH
O
n = 0,1,2
L-homocysteine analogues of GSH, studied herein.
Figure 1. S-Alkyl
GGT inhibitors, proposing that they resemble GSH, the
in vivo substrate. Testing of these analogues revealed a
promising competitive inhibitor (
Herein we report the synthesis of a series of S-alkyl L-
homocysteine GSH analogues, wherein the sulfur is
present in three distinct oxidation states: as a thioether,
a sulfoxide and the corresponding sulfone (Fig. 1).
Inhibition constants are reported for these sulfur
derivatives for the GGT-mediated reaction of the
chromogenic donor substrate -glutamic acid c-p-ni-
troanilide (L-GPNA), with the efficient acceptor sub-
strate Gly-Gly. The affinity of these S-alkyl inhibitors
can be compared with the corresponding S methyl (i.e.
-Met) derivatives at the same oxidation state.
L
-SPG, Fig. 1, n ¼ 1).9
L
L
In the course of our previous investigation of the
mechanism of acylation of GGT, we reported the dis-
covery of S-alkyl sulfoxides as a new class of reversible
Scheme 1 depicts the synthesis of a thioether analogue of
GSH (L-TEPG, 4). The Boc- -homocysteine methyl
ester was prepared in three steps from -homocystine as
reported in the literature.9 The first step from the
L
L
L
-
homocysteine derivative is the condensation with
bromide 1, previously synthesized in one step according
to a literature procedure,9 to obtain 2. The next step is
* Corresponding author. Tel.: +1-514-343-6219; fax: +1-514-343-7586;
e-mail: keillorj@chimie.umontreal.ca
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.04.072

3452
C. Lherbet et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3451–3455
BocHN
COOR
S
BocHN
COOMe
SH
i
O
H
N
OtBu
O
R = Me (2)
ii
R = H
(3)
Cl
H3N
COOH
S
O
H
N
Br
iii
OtBu
O
H
O
(1)
N
OH
O
(4)
Scheme 1. Synthesis of
(b) HCl (0.1 N), 100% (two steps).
L
-TEPG (4). Reagents and conditions: (i) K2CO3, (1), DMF, 71%; (ii) NaOH, THF/H2O; (iii) (a) TFA, CH2Cl2 at 0 °C;
the hydrolysis of the methyl ester to provide 3 and re-
moval of the Boc protecting group and tert-butyl ester
with TFA to give the desired compound 4 with a high
overall yield.
Scheme 3 describes the synthesis of an aryl analogue (
STG, 9) of the previously synthesized9 sulfoxide com-
pound ( -SPG). An aromatic moiety was introduced in
the place of the propyl moiety present in -SPG, since
L-
L
L
hippurate (benzoylglycine) is known to be one of the
best activators of the hydrolysis reaction catalyzed by
GGT. This activator has been proposed to be bound at
the same place as the Cys-Gly moiety of GSH, where we
The sulfone GSH analogue (L-SnPG, 6) shown in
Scheme 2 was synthesized from compound 3. Sulfone 5
was obtained directly through the oxidation of the S
alkyl
L
-homocysteine derivative 3 by hydrogen peroxide
propose that the propylglycine moiety of
bound. Analogue 9 was thus designed combining rec-
ognition elements of donor substrate -GPNA, the
sulfoxide inhibitor LSPG and the activator hippurate.
L-SPG may be
in presence of acetic acid with a modest final yield of
51%. In an attempt to improve this yield, a second
method was tried, involving the initial oxidation of 2
with sodium periodate to obtain the sulfoxide derivative.
All attempts to push the reaction further to obtain the
sulfone gave no further disappearance of the starting
material, even with heating. Thus the sulfoxide inter-
mediate was oxidized further using hydrogen peroxide in
acetic acid with heating. After two days, the desired
sulfone was obtained in 53% yield over these two steps.
Subsequent hydrolysis of the methyl ester in alkaline
media provided 5, followed by removal of the Boc
protecting group and the tert-butyl ester by the standard
method (TFA) to obtain sulfone 6.
L
The first step in the synthesis of L-STG is the conden-
sation of the -homocysteine derivative with com-
L
mercially available b-bromotoluic acid. The aryl
carboxylate derivative was then condensed with glycine
tert-butyl esterÆHCl in the presence of the coupling agent
TBTU to give the thioether derivative 7. Subsequent
oxidation of 7 with NaIO4 provided sulfoxide 8 in good
yield. The remaining steps represent the straightforward
deprotection of the molecule with sodium hydroxide to
remove the a-methyl ester, followed by removal of the
BocHN
COOH
BocHN
COOH
S
i
O
O
H
N
H
N
S
O
O
OtBu
OtBu
O
O
(5)
(3)
v
ii
iii, iv
Cl
BocHN
COOMe
S
H3N
COOH
S
O
O
H
N
H
O
N
OtBu
OH
O
O
O
(2)
(6)
Scheme 2. Synthesis of L-SnPG (6). Reagents and conditions: (i) H2O2, AcOH, THF, 51%; (ii) NaIO4, THF/H2O, 91%; (iii) H2O2, AcOH, MeOH,
58%; (iv) NaOH, THF/H2O, 81%; (v) (a) TFA, CH2Cl2 at 0 °C; (b) HCl (0.1 N), 100%.

C. Lherbet et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3451–3455
3453
BocHN
COOMe
SH
BocHN
COOMe
S
O
i, ii
OtBu
N
H
O
(7)
Cl
H3N
COOH
S
BocHN
COOMe
O
O
iii
iv, v
OtBu
OH
N
H
N
H
S
O
O
O
O
(9)
(8)
Scheme 3. Synthesis of
L
-STG (9). Reagents and conditions: (i) b-bromotoluic acid, K2CO3, DMF; (ii) glycine t-butyl esterÆHCl, DIEA, TBTU,
CH3CN, 44% (two steps); (iii) NaIO4, THF/H2O, 92%; (iv) NaOH, THF/H2O, 94%; (v) (a) TFA, CH2Cl2 at 0 °C; (b) HCl (0.1 N), 93%.
Boc protecting group and tert-butyl ester with TFA to
give the desired product 9.
with the donor substrate binding site of the enzyme.
Other analogues of GSH have been synthesized17 where
the isopeptide bond has been replaced with a dialkyl
amino linkage, and these have also demonstrated good
inhibition.
All synthetic reactions were followed by TLC and the
1
synthetic intermediates were characterized by H and
13C NMR and MS.
L-GPNA was synthesized as previ-
ously described.10 Compounds (4),11 (6)12 and (9)13 were
selected for kinetic evaluation14 as inhibitors of GGT,
purified from rat kidney as previously described.15
Parameters from Michaelis–Menten plots were re-plot-
ted against the concentrations of each inhibitor. The Ki
of sulfone 6 was calculated from its IC50 value using the
In order to probe the importance of the propylglycine
moiety, it is also instructive to compare sulfone 6 (
SnPG) and -methionine sulfone. The methionine
derivative is unable to inhibit GGT but -SnPG, bearing
L-
L
L
the propylglycine moiety, shows inhibition for which a
Ki of (2.9 0.6) mM was calculated. This calculation was
made assuming the inhibition was competitive, since the
structure of inhibitor 6 is very similar to those of thio-
KM of -GPNA as 590 lM and presuming competitive
L
inhibition. The results from these tests are reported in
Table 1.
ether 4 and sulfoxide (
known competitive inhibitors against
sulfonamide analogue of -SnPG, which could inhibit
L
-SPG) derivatives, which are
L
-GPNA. The
Compounds 4, 6 and 9 may be categorized as analogues
L
of
stituent at the d-position, and according to the oxidation
state of the sulfur. -TEPG (4) is thus a GSH analogue
L
-methionine or GSH, depending on the alkyl sub-
GGT irreversibly, has been synthesized, but no kinetic
data has been given as proof of the efficacy of this
molecule.18 Compared to the sulfoxide, the second S@O
double bond in the sulfone may lead to a loss of affinity
for the donor binding site, perhaps due to steric hin-
drance. Finally, the favourable effect of the propylgly-
cine moiety is further confirmed by comparing
L-methionine sulfoxide with L-SPG; the GSH analogue
is 100 times more efficient as a competitive inhibitor
than the corresponding methionine derivative.
L
wherein the c-amide bond has been replaced with a
thioether functionality. Inhibition studies of this com-
pound demonstrate competitive inhibition with a Ki of
(520 40) lM, around 10 times higher than that of the
corresponding sulfoxide (
L
-SPG, Ki ¼ ð53 ꢀ 3ÞlM).9 By
comparison with
L
-methionine, the affinity of LTEPG is
around 50-fold greater due to the pendant propylglycine
moiety, which resembles the Cys-Gly moiety of GSH,
the in vivo substrate of GGT. This moiety is evidently
critically important to favourable binding interactions
As discussed above, hippurate is one of the best acti-
vators of GGT-mediated hydrolysis, purportedly bind-
ing in the same place as the Cys-Gly moiety of GSH,19
so a benzoylglycine moiety was employed in sulfoxide
derivative 9 as a mimic of the Cys-Gly fragment. It was
Table 1. Type of inhibition and Ki for each compound tested in
presence of rat GGT at pH 8.0 and 37 °C
found that L-STG (9) acts as a competitive inhibitor;
Vmax did not vary as a function of inhibitor concentra-
Compounds
Type of inhibition Ki (mM)
L
L
L
L
L
L
L
-Methionine
-Methionine sulfoxide
-Methionine sulfone
-TEPG (4)
Competitive
Competitive
––
26.9a
5.9 0.2b
N/ob;c
tion (data not shown), but KM varied by the factor of
½Iꢁ
ð1 þ Þ, giving a Ki of (225 30) lM (Fig. 2). This aro-
Ki
matic derivative is thus a less efficient inhibitor than the
sulfoxide derivative bearing a propylglycine moiety (L-
Competitive
Competitive
––
0.52 0.04
0.053 0.003b
3.5 0.6d
0.23 0.03
-SPG
-SnPG (6)
-STG (9)
SPG). However, its affinity may still be better than
GSH, the natural substrate of GGT, whose KM value is
around 500 lM.20 This demonstrates that the sulfoxide
group is nevertheless a good pharmacophore for the
inhibition of GGT as a substitute for the labile c-glut-
amyl amide bond of GSH.
Competitive
a See Ref. 16.
b See Ref. 5.
c N/o: No inhibition observed up to 20 mM.
d Ki calculated from IC50, presuming competitive inhibition.

In summary, we have synthesized a series of S-alkyl L-
3454
C. Lherbet et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3451–3455
0.024
0.022
0.020
0.018
0.016
0.014
0.012
0.010
0.008
0.006
0.004
Figure 4. X-ray structure of the picrate salt of the inhibitory SCSS
methionine sulfoxide.
0
100
200
300
400
500
600
700
800
[I] (µM)
homocysteine derivatives with different oxidation states
of the sulfur. The sulfoxide derivative is 10 times more
efficient than the thioether compound, which is in turn
10 times more efficient than the corresponding sulfone.
Further comparison suggests that the propylglycine
moiety contributes significantly to its affinity for rat
Figure 2. Replot for the competitive inhibition of
(9).
L-GPNA by L-STG
The inhibition efficiency displayed by sulfoxides com-
pared to thioethers or sulfones may be due to their
resemblance of the tetrahedral intermediate formed
during the acylation of GGT. Based on mutagenesis
experiments, two adjacent serines, conserved in all
GGTs, have been implicated in important active site
binding roles, including the stabilization of the oxyanion
of the tetrahedral intermediate of the acylation step.1 It
kidney GGT, providing in the best case (L-SPG) a low
micromolar Ki value. It is also important to note that
these sulfoxides were tested as 50/50 mixtures of S dia-
stereoisomers, whereas we have confirmed with the
L
-Met derivative that only the SS diastereoisomer is
recognized. This demonstrates the importance of
stereospecific synthesis of the SS diastereoisomer for
future work.
is conceivable that sulfoxide L-SPG could be bound and
stabilized in the same fashion (Fig. 3). Further interac-
tions could also exist between the partially positively
charged sulfur of the inhibitor and the threonine residue
that is the probable active site nucleophile.
Acknowledgements
While no three-dimensional structure of GGT presently
exists to verify this hypothetical binding mode, other
lines of evidence suggest that these interactions are
highly spatially orientated. We have previously demon-
strated that the interaction between GGT and sulfoxide
inhibitors is remarkably stereospecific with respect to
the configuration of the d stereocentre, in that one dia-
stereoisomer inhibits with a Ki of 3.4 mM, whereas the
other diastereoisomer is not recognized.9 Following a
literature protocol,21 we have isolated one of the two
diastereomers by crystallization as its picrate salt and
determined its absolute configuration by X-ray diffrac-
tion (Fig. 4).22 Subsequent testing of the optically pure
compound confirmed that the isolated SCSS diastereo-
ꢀ
ꢀ
X-ray crystallography was performed by Dr. F. Belan-
ꢀ
ger-Gariepy of the X-ray lab of the Universite de
Montreal. This work was supported by the Natural
Sciences and Engineering Research Council (NSERC)
of Canada. The authors also acknowledge postgraduate
ꢀ
scholarships from the Universite de Montreal (CL) and
ꢀ
ꢀ
NSERC (CG).
References and notes
1. Taniguchi, N.; Ikeda, Y. Adv. Enzymol. Relat. Areas Mol.
Biol. 1998, 239.
mer of L-Met sulfoxide is specifically bound by GGT.
2. Suzuki, H.; Kumagai, H.; Echigo, T.; Tochikura, T. J.
Bacteriol. 1989, 171, 5169.
3. Kasai, T.; Larsen, P. O. Proc. Chem. Org. Nat. Prod. 1980,
39, 173.
H3N
COO
S
4. Sian, J.; Dexter, D. T.; Lees, A. J.; Daniel, S.; Jenner, P.;
Marsden, C. D. Ann. Neurol. 1994, 36, 356.
5. Graber, R.; Losa, G. A. Int. J. Cancer 1995, 62, 443.
6. Del Bello, B.; Paolicchi, A.; Comporti, M.; Pompella, A.;
Maellaro, E. FASEB J. 1999, 13, 99.
δ
O
H
N
Thr-380
O
OH
H
O
O
ꢀ
7. Keillor, J. W.; Menard, A.; Castonguay, R.; Lherbet, C.;
H
O
H
Rivard, C. J. Phys. Org. Chem. 2004, 17, in press.
8. Inoue, M.; Hiratake, J.; Suzuki, H.; Kumagai, H.; Sakata,
K. Biochemistry 2000, 39, 7764.
O
Ser-450
Ser-451
9. Lherbet, C.; Keillor, J. W. Org. Biomol. Chem. 2004, 2,
238–245.
10. Lindsay, H.; Whitaker, J. F. OPPI Briefs 1975, 89.
Figure 3. Possible interactions between the inhibitor (L-SPG) and the
active site of rat kidney GGT.

C. Lherbet et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3451–3455
3455
11. (4) Gummy solid: ½aꢁ ꢂ 3:0 (c 0.32, MeOH); 1H NMR
pH 8.0. An IC50 value was found with different concen-
trations (0–10 mM) for sulfone (6) in the presence of a
fixed concentration of LGPNA (400 lM) and a saturating
concentration (20 mM) of Gly-Gly. Blank experiments
were done without enzyme. Ki values were determined by
re-plots of Vmax=KMobs versus [I] analyzed by Origin 6.0
curve-fitting software.
D
(D2O, 400 MHz) d 2.07–2.20 (m, 2H), 2.54 (t, J ¼ 7:0 Hz,
2H), 2.65 (t, 2H, J ¼ 7:1 Hz, 2H), 2.75 (t, J ¼ 6:8 Hz, 2H),
3.90 (s, 2H), 4.13 (m, 1H); 13C NMR (D2O, 100 MHz) d
26.7, 26.8, 29.7, 35.5, 41.3, 51.9, 141.9, 173.6, 175.3; m=z
265.1 (MHþ, C9H17N2O5S requires 265.0858).
12. (6) Gummy solid: ½aꢁ ꢂ 3:7 (c 0.28, MeOH); 1H NMR
D
ꢀ
(D2O, 400 MHz) d 2.36–2.47 (m, 2H), 2.85 (t, J ¼ 7:1 Hz,
2H), 3.39–3.52 (m, 2H), 3.56 (t, J ¼ 7:1 Hz, 2H), 3.97 (s,
2H), 4.21 (t, J ¼ 6:6 Hz, 1H); 13C NMR (D2O, 100 MHz) d
22.5, 27.7, 41.5, 48.3, 48.7, 51.5, 170.9, 172.7, 173.6; m=z
297.0 (MHþ, C9H17N2O7S requires 297.0756).
15. Menard, A.; Castonguay, R.; Lherbet, C.; Rivard, C.;
Roupioz, Y.; Keillor, J. W. Biochemistry 2001, 40, 12678.
16. Allison, R. D. Methods Enzymol. 1985, 113, 419.
17. Burg, D.; Filippov, D. V.; Hermanns, R.; van der Marel,
G. A.; van Boom, J. H.; Mulder, G. J. Bioorg. Med. Chem.
2002, 10, 195.
18. Luisi, G.; Calcagni, A.; Pinen, F. Tetrahedron Lett. 1993,
34, 2391.
19. Thompson, G. A.; Meister, A. J. Biol. Chem. 1980, 255,
2109.
13. (9) Gummy solid: ½aꢁ ꢂ 4:4 (c 0.32, MeOH); 1H NMR
D
(D2O, 400 MHz) d 2.33 (m, 2H), 2.93–3.08 (m, 2H), 4.11
(s, 2H), 4.15–4.22 (m, 1H + 1H), 4.27 (dd, J ¼ 13:2 Hz,
J ¼ 3:0 Hz), 7.40 (d, J ¼ 8:1 Hz, 2H), 7.75 (d, J ¼ 8:1 Hz,
2H); 13C NMR (CD3OD, 75 MHz) d 24.1 (d, mixture 50/
50 of diastereoisomers), 42.3, 46.2 (d), 52.2 (d), 56.2 (d),
128.5, 131.6, 133.6, 134.2, 171.0, 171.4, 174.1; m=z
343.0964 (MHþ, C14H19N2O6S requires 343.0948).
ꢀ
20. Rivard, C.; Keillor, J. W. M.Sc. Thesis Universite de
ꢀ
Montreal, 2003.
21. Lavine, T. F. J. Biol. Chem. 1947, 169, 477.
22. Absolute configuration was established by structure
determination using a stereocentre reference of known
absolute configuration and confirmed by anomalous
dispersion effects (sulfur atom) in diffraction measure-
ments on the crystal. The Flack parameter refined to
0.07(6) and the crystallographic data were deposited at the
Cambridge Crystallographic Data Centre as CCDC
236529.
14. Reactions were initiated by the addition of 3.38 mU of
GGT. Liberated p-nitroaniline was detected spectropho-
tometrically at 410 nm (e ¼ 8800 Mꢂ1 cmꢂ1) on a Cary 100
Bio spectrophotometer. For inhibition studies, different
concentrations (67–1400 lM) of LGPNA and a saturating
concentration (20 mM) of Gly-Gly were used in the
presence of varying concentrations (0–1 mM) of com-
pounds
L-TEPG (4) and L-STG (9) in 0.1 M Tris–HCl