Synthesis and biological evaluation of the disulfide form of the glutathione analogue γ-(L-glutamyl)-L-cysteinyl-L-aspartyl-L-cysteine

Article abstract of DOI:10.1016/S0045-2068(03)00022-1

By using the chain to chain mode of cyclization the title glutathione analogue γ-Glu-Cys-Asp-Cys (4), containing the 11-membered disulfide ring -Cys-Asp-Cys- replacing the native -Cys-Gly fragment, has been synthesized and characterized together with its reduced dithiol form γ-Glu-Cys-Asp-Cys (5). The activity of (4) with γ-glutamyl-transferase and glutathione reductase has been evaluated and compared with those of the two conformationally restricted glutathione analogues (2) and (3) previously reported.

Full text of DOI:10.1016/S0045-2068(03)00022-1

BIOORGANIC
CHEMISTRY
Bioorganic Chemistry 31 (2003) 109–121
www.elsevier.com/locate/bioorg
Synthesis and biological evaluation of
the disulfide form of the glutathione analogue
c-(
L
-glutamyl)-
L
-cysteinyl-
L
-aspartyl-
L
-cysteine
a
Ivana Cacciatore, Antonio Di Stefano, Silvestro Dupre,
a
b
ꢀ
c
a,
b
Enrico Morera, Francesco Pinnen, * and Alessandra Spirito
a
Dipartimento di Scienze del Farmaco, Universitaꢀ ‘‘G. DÕAnnunzio,’’ Via dei Vestini 31, Chieti 66100, Italy
Dipartimento di Scienze Biochimiche and Centro di Biologia Molecolare del C.N.R.,
b
Universitaꢀ ‘‘La Sapienza,’’ P.le A. Moro 5, Roma 00185, Italy
Dipartimento di Studi Farmaceutici and Centro di Studio per la Chimica del Farmaco del C.N.R.,
c
Universitaꢀ ‘‘La Sapienza,’’ P.le A. Moro 5, Roma 00185, Italy
Received 28 March 2002
Abstract
By using the chain to chain mode of cyclization the title glutathione analogue
(4), containing the 11-membered disulfide ring
replacing the native -Cys-Gly frag-
ment, has been synthesized and characterized together with its reduced dithiol form c-Glu-
Cys-Asp-Cys (5). The activity of (4) with c-glutamyl-transferase and glutathione reductase
has been evaluated and compared with those of the two conformationally restricted glutathi-
one analogues (2) and (3) previously reported.
Ó 2003 Elsevier Science (USA). All rights reserved.
Keywords: Amino acids; Conformational constraint; Cyclic disulfides; Cyclotripeptides; Dithiols;
Glutathione analogues
1. Introduction
Disulfide bridges represent important evolutionarily conserved structural motifs
in many natural peptides and proteins, including hormones, enzymes, growth fac-
tors, toxins, and immunoglobulins [1]. In recent years intensive efforts have been
* Corresponding author. Fax: +39-0871-3555267.
E-mail address: fpinnen@unich.it (F. Pinnen).
0045-2068/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0045-2068(03)00022-1

110
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
made to develop mimetics of bioactive natural peptides containing cyclic disulfide
frameworks in order to obtain analogs with better therapeutical potential than the
endogenous prototypes. This strategy provides a useful way to stabilize specific back-
bone conformations and limit the rotational space of the side chains [2,3].
In the context of our studies on glutathione (GSH, 1)1 analogues [4–6] we recently
adopted this approach by synthesizing the two disulfide analogs 2 and 3 characterized
by the presence of the 8-membered disulfide ring
and the achiral and Ca;atetra-
substituted 4-amino-1,2-dithiolane-4-carboxylic acid, respectively [7,8]. These modifi-
cations introduce new reactivities such as the participation in redox equilibria
involving intra- or intermolecular disulfide forms [9–11], and have biological and con-
formational consequences resulting from the stabilization of specific backbone confor-
mations and the restriction of the freely rotating side-chain of cysteine.
As part of our study on conformationally restricted GSH analogues we report
here the synthesis of a new tetrapeptide model in which the backbone constraint is
obtained by the incorporation of a cyclic fragment, namely
which replaces
the native C-terminal -Cys-Gly- dipeptide. In the resulting GSH analogue
4, the presence of the C-terminal carboxyl group, which is a prerequisite
for the GSH activity, is maintained due to the presence of the aspartic acid residue.
In contrast with previously reported cyclic GSH analogues, in which the residue
of the glutamic acid is part of the ring system [12], in 4 the c-glutamyl moiety is lo-
cated outside the ring and is not directly involved in the conformational constraint.
Furthermore, due to the presence of the two adjacent cysteine residues, the tetrapep-
tide molecule could participate in intramolecular or intermolecular dithiol–disulfide
interchange reactions.
Although several 11-membered disulfide cyclic tripeptide
systems have
been synthesized and studied [13–16], the tetrapeptide 4 is, to the best of our knowl-
edge, the first synthetic c-glutamyl derivative of an 11-membered disulfide cyclotri-
peptide to be reported. It should be also noted that a c-glutamyl derivative
analogous to 4, namely
ent in the bovine brain [17].
has been detected among the neuropeptides pres-
Data concerning the activity of 4 with two glutathione dependent enzymes, c-glut-
amyltranspeptidase (c-GT, EC 2.3.2.2) and glutathione reductase (GR, EC 1.6.4.2),
are reported and compared with that of analogues 2 and 3. Interest in c-GT is due to
the metabolic stability of GSH and its analogues [18] as well as in its expression and
function in carcinogenesis [19]. The GR, on the other hand, is a flavoprotein that plays
a key antioxidant role through its intracellular function of maintaining a
high GSH–GSSG ratio [20]. GSH analogues able to interfere with its functions may as-
sist in obtaining information about its role in the development of oxidative stress [21].
1
Abbreviations used: Abbreviations follow the recommendations of the IUPAC-IUB Commision on
Biochemical Nomenclature as given in Eur. J. Biochem. 138 (1984) 9–37. Additional abbreviations: DBU,
1,8-diazabicyclo 5.4.0 undec-7-ene; DCCI, N,N-dicyclohexylcarbodiimide; GPNA, c-glutamyl-p-nitroani-
lide; HETCOR, heteronuclear chemical shift correlation experiment; NADPH, b-nicotinamide adenine
dinucleotide phosphate reduced; NMM, N-methylmorpholine; THF, tetrahydrofuran; Trt, triphenylm-
ethyl; ðn-BuÞ3P, tri-n-butylphosphine; GSH, glutathione reduced form; GSSG, glutathione oxidized;
c-GT, c-glutamyltranspeptidase; GR, glutathione reductase.

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
111
2. Chemistry
The first synthetic strategy adopted to prepare the GSH analogue 4 was the incor-
poration of the c-Glu residue onto the preformed 11-membered cyclic disulfide. Ac-
cordingly,
was already found in the case of the 8-membered cyclic disulfide
was synthesized as suitable intermediate. However, as
[7], the
N-deprotection represented a critical step and notwithstanding the use of different de-
blocking agents, the desired Cys-Asp-(OtBu)-Cys-OtBu was only obtained in low yield
and was contaminated by impurities and degradation products. As an alternative route
to 4, coupling of the c-Glu residue with the linear Cys-Asp-Cys fragment was consid-
ered (Scheme 1). In order to achieve satisfactory yields, the use of the tert-butyl esters,
removable under acidic conditions, was adopted [6]. Thus (see Scheme 1) the key inter-
mediate tetrapeptide c(Boc-Glu-OtBu)-Cys(Trt)-Asp(OtBu)-Cys(Trt)-OtBu 11 was
synthesized by a stepwise elongation using the N,N-dicyclohexylcarbodiimide (DCCI)
method. For the N-deprotection of the Fmoc-intermediates 7 and 9, 1,8-diazabicyclo
5.4.0 undec-7-ene (DBU) in CH2Cl2 was chosen [6]. Selective removal under mild con-
ditions (I2 in MeOH, room temperature) of the Trt protecting groups under high dilu-
tion conditions [7] allowed direct conversion of 11 to the N-Boc cyclic [1,3] disulfide 12.
It should be noted that the yield (35%) of this deprotection-cyclization step is signifi-
cantly lower than that previously reported by us for the 8-membered cyclic disulfide
2. In the present case a more pronounced tendency to form oligomeric products was
observed and this led to the adoption of higher dilution conditions (10À3 M vs
10À2 M ca.). Subsequent removal in a single step of all protecting groups, followed
by treatment of the resulting hydrochloride with aqueous ammonia, afforded the
GSH analogue 4 in good yields. Reductive cleavage of the disulfide link of 4 with tri-
n-butyl phosphine in water/n-propanol solution [5] gave the dithiol form 5 in 70% yield.
The purity of 4 was determined by HPLC and TLC. Under the conditions adopted for
HPLC analysis, compound 4 eluted as a single peak (retention time: 17.7 min).

112
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
Scheme 1. Reagents and conditions. (a) Cys(Trt)OtBu, DCCI, THF 0 °C, then 5 °C, 16 h; (b) DBU,
CH2Cl2, rt; 15 min; (c) Fmoc-Cys(Trt)-OH, DCCI, THF, 0 °C, 4 h, then 5 °C, 16 h; (d) DBU, CH2Cl2,
rt, 15 min; (e) Boc-Glu-OtBu, DCCI, THF, 0 °C, 4 h, then 5 °C, 16 h; (f) I2, MeOH, rt, 4 h; (g) HCl, ethyl
acetate, rt, 1 h; (h) 1 N aq. NH3, rt, 30 min; (i) ðn-BuÞ P, n-PrOH-H2O (2:1), rt, 1 h.
3
The structure assigned to the GSH analogues 4 and 5 is in accordance with the
spectroscopic properties (see Table 1). In particular, the NMR spectra of the linear
dithiol form 5 reveal a single set of resonances. On the contrary, the spectral data of
the cyclic disulfide 4 show the existence of more than one set of signals, thus indicat-
ing the presence (under the conditions of the NMR experiment) of slowly intercon-
verting multiple conformational states. In particular, analogue 4 showed three
complete sets of signals (relative intensity 1:0.7:0.3) which, on the basis of accurate
HPLC analysis, can be attributed to different conformers. A characteristic feature

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
113
Table 1
1H- and 13C-NMR parametersa of compounds 4b and 5
Residue
(4)
c-Glu-Cys-Asp-Cys (5)
dH
dC
54.26
dH
dC
Glu
Ca
3.59 t (5.12)
1.90–2.12 m
2.24–2.50 m
3.59 t (5.15)
1.90–2.05 m
2.30–2.41 m
54.25
30.17
37.90
Cb
26.35
31.53
Cc
COO
CONH
–
–
176.70c
177.67c
172.47d
176.04d
Cys
Ca
4.45–4.54 m
2.75–2.90 m
–
53.42
38.86
174.79c
4.16–4.22 m
2.70–2.90 m
56.68
26.39
175.03d
Cb
CO
Asp
Ca
4.30–4.42 m
2.50–2.70 m
–
51.21
37.62
171.85c
4.51–4.59 m
2.45–2.70 m
51.60
31.45
172.32d
Cb
CO
Cys
Ca
4.45–4.55 m
3.00–3.20 m
–
52.44
38.06
174.16c
4.31–4.40 m
2.70–2.90 m
55.90
26.28
174.19d
Cb
CO
a In D2O at 25 °C. J/Hz in parentheses.
b Data refer to the predominant conformer.
c Assignments may be interchanged.
d Assignments may be interchanged.
of the 13C-NMR spectrum of 4 is represented by the resonances of the two Cys
b-carbon atoms (38.06 and 38.86 ppm) which are downfield shifted compared with
the corresponding atoms in the dithiol derivative 5 (26.28 and 26.39 ppm). This spec-
tral behaviour represents, as recently noted [22], a reliable diagnostic feature for dis-
tinguishing between cysteine and cystine containing forms. Mass spectra of 4 and 5
show the most abundant peaks correspond to M þ Hþ species (467 and 469 m=z,
respectively). Furthermore, high resolution analysis of the respective molecular ion
region demonstates the monomeric nature of 4 and 5.
3. Materials and methods
TLC was performed on Merck 60 F254 silica gel plates. HPLC was performed with
a Waters system equipped with a M-600 multisolvent system and a 2487 dual waves
UV detector linked to a Millenium 32 data station using a Merck C-18 reversed-
phase column (125 Â 4 mm) ‘‘Purospher’’ with 3 lm packing material. The solvent
used was 0.1% aqueous trifluoroacetic acid/methanol 80/10 at the flow rate of
1.0 ml/min. The eluant was detected at 210 nm. Column chromatography was carried
out using Merck 60 silica gel (230–400 mesh). Optical rotations were obtained at

114
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
25 °C with Perkin–Elmer 241 polarimeter. IR spectra were recorded employing a
1
Perkin–Elmer 983 FTIR 1600 spectrophotometer. H- and 13C-NMR spectra were
determined on a Varian VXR 300-MHz spectrometer (d expressed in ppm). The as-
sigment of 1H- and 13C-NMR resonances of analogues 4 and 5 (Table 1) was
achieved with standard 2D NMR spectroscopic techniques (HETCOR). The mass
spectra of 4 and 5 have been obtained in electrospray (ES) conditions by a LCQ
(Finnigan, San. Jose, CA, USA) instrument. The 60 lM solution of the sample in
CH3OH:H2O (50:50) containing 1% acetic acid was directly injected into the instru-
ment at the flow rate of 5 ll/min.
Cys(Trt)-OtBu, Fmoc-Cys(Trt)-OH, Fmoc-Asp(OtBu)-OH and Boc-Glu-OtBu
€
were obtained from Novabiochem (Laufelfingen, Switzerland, EU).
3.1. Fmoc-Asp(OtBu)-Cys(Trt)-OtBu (7)
DCCI (3.6 g, 17.5 mmol) was added portionwise at 0 °C to a stirred solution of
Fmoc-Asp(OtBu)-OH 6 (7.2 g, 17.5 mmol) and Cys(Trt)-OtBu (7.4 g, 17.5 mmol) in
THF (30 ml). After 3 h at 0 °C and 16 h at 5 °C, the reaction mixture was filtered
and the resulting solution was evaporated under vacuum. The residue was taken
up in ethyl acetate and the organic layer washed with 1 M KHSO4, saturated aque-
ous Na2CO3 and H2O. The residue obtained after drying and evaporation was chro-
matographed on silica gel using cyclohexane/diethyl ether 1:1 as eluant to yield
dipeptide di-tert-butyl ester 7 as a white foam (7.4 g, 52%). ½aŠ ¼ þ20:0° (c ¼ 1,
D
CHCl3). Rf ¼ 0:42 (cyclohexane/diethyl ether, 1:1). IR (nujol): 3310, 1725, 1670,
1
1490cmÀ1. H NMR (CDCl3) d 1.45 (18H s, 6Â CH3), 2.58 (2H, m, Cys b-CH2),
2.77 (2H, d, Asp b-CH2), 4.15–4.45 (3H, m, Fmoc CH–CH2), 4.40 (1H, m, Cys a-
CH), 4.60 (1H, m, Asp a-CH), 6.01 (1H, d, J ¼ 8:25 Hz, Asp NH), 7.13 (1H, d,
J ¼ 7:70 Hz, Cys NH), 7.10–7.82 (23H, m, aromatic protons).
3.2. Asp(OtBu)-Cys(Trt)-OtBu (8)
DBU (1.4 g, 8.6 mmol) was added at room temperature to a solution of the di-tert-
butyl ester 7 (7.0 g, 8.6 mmol) in CH2Cl2 (65 ml). After 20 min the solution was evap-
orated to dryness and the residue chromatographed on silica gel using CHCl3=
MeOH 99:1 as eluant to give pure N-deprotected dipeptide ester 8 as a yellow oil
(4.8 g, 95%). ½aŠ ¼ þ11:0° (c ¼ 1, CHCl3). Rf ¼ 0:42 (CHCl3=MeOH, 99:1). IR (nu-
D
jol): 3370 (brd), 1725, 1675, 1505cmÀ1. H NMR (CDCl3) d 1.45 (18H, s, 6ÂCH3),
1
1.80 (2H, s, NH2), 2.51 (2H, m, Cys b-CH2), 2.70 (2H, m, Asp b-CH2), 3.65 (1H, m,
Asp a-CH), 4.42 (1H, m, Cys a-CH), 7.10–7.51 (15H, m, aromatic protons), 7.84
(1H, d, J ¼ 8:10 Hz, Asp NH).
3.3. Fmoc-Cys(Trt)-Asp(OtBu)-Cys(Trt)-OtBu (9)
DCCI (1.6 g, 7.6 mmol) was added portionwise at 0 °C to a stirred solution of Fmoc-
Cys(Trt)-OH (4.45 g, 7.6 mmol) and the di-tert-butyl ester of the dipeptide 8 (4.5 g,
7.6 mmol) in THF (18 ml). After 3 h at 0 °C and 16 h at 5 °C, the reaction mixture

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
115
was filtered and the resulting solution was evaporated under vacuum to give a residue
which was taken up in ethyl acetate and worked up as described for compound 7. Pu-
rification by silica gel column chromatography using cyclohexane/diethyl ether 1:1 as
eluant afforded pure tripeptide di-tert-butyl ester 9 as a white foam (8.3 g, 94%).
½aŠ ¼ þ18:0° (c ¼ 1, CHCl3). Rf ¼ 0:29 (cyclohexane/diethyl ether, 1:1). IR (nujol):
D
3320 (brd), 1725, 1675, 1505cmÀ1. 1H-NMR (CDCl3) d 1.42 (18H, s, 6Â CH3), 2.41
and 2.71 (4H, m, 2Â Cys b-CH2), 2.78 (2H, m, Asp b-CH2), 3.65 (1H, m, Cys a-
CH), 4.15–4.45 (3H, m, Fmoc CH–CH2), 4.18 (1H, m, Cys a-CH), 4.63 (1H, m, Asp
a-CH), 4.90 (1H, d, J ¼ 7:05 Hz, Cys NH), 7.03 (1H, d, J ¼ 7:91 Hz, Asp NH), 7.20
(1H, obs, Cys NH), 7.15–7.81 (38H, m, aromatic protons).
3.4. Cys(Trt)-Asp(OtBu)-Cys(Trt)-OtBu (10)
DBU (1.1 g, 6.9 mmol) was added at room temperature to a solution of the trip-
eptide di-tert-butyl ester 9 (8.0 g, 6.9 mmol) in CH2Cl2 (53 ml). After 20 min the so-
lution was evaporated to dryness and the residue chromatographed on silica gel
using CHCl3=MeOH 99:1 as eluant to give pure N-deprotected tripeptide ester 10
as a white foam (5.1 g, 79%). ½aŠ ¼ þ10:5° (c ¼ 1, CHCl3). Rf ¼ 0:36 (CHCl3=
D
MeOH 99:1). IR (nujol): 3345 (brd), 1725, 1675, 1500cmÀ1. H NMR (CDCl3) d
1
1.49 (18H, s, 6Â CH3), 1.51 (2H, br, NH2), 2.45–2.75 (6H, m, 2Â Cys b-CH2 and
Asp b-CH2), 2.96 (1H, m, Cys a-CH), 4.30 (1H, m, Cys a-CH), 4.69 (1H, m, Asp
a-CH), 7.09 (1H, d, J ¼ 6:01 Hz, Cys NH), 7.11–7.50 (30H, m, aromatic protons),
8.11 (1H, d, J ¼ 7:72 Hz, Asp NH).
3.5. c(Boc-Glu-OtBu)-Cys(Trt)-Asp(OtBu)-Cys(Trt)-OtBu (11)
DCCI (1.1 g, 5.3 mmol) was added portionwise at 0 °C to a stirred ice-cold solu-
tion of Boc-Glu-OtBu (1.6 g, 5.3 mmol) and the tripeptide di-tert-butyl ester 10 (5 g,
5.3 mmol) in THF (30 ml). After 3 h at 0 °C and 16 h at 5 °C, the reaction mixture was
filtered and the resulting solution was evaporated under vacuum to give a residue
which was taken up in ethyl acetate and worked up as described for compound 7.
Purification by silica gel column chromatography using CHCl3/diethyl ether 1:1 as
eluant afforded pure tetrapeptide tri-tert-butyl ester 11 as a white foam (6.2 g,
96%). ½aŠ ¼ þ11:5° (c ¼ 1, CHCl3). Rf ¼ 0:68 (CHCl3/diethyl ether, 1:1). IR (nujol):
D
3405 (brd), 1725, 1665, 1495cmÀ1. H NMR (CDCl3) d 1.42 (9H, s, 3Â CH3), 1.45
1
(27H s, 9Â CH3), 1.81 (2H, m, Glu b-CH2), 2.21 (2H, m, Glu c-CH2), 2.30–2.70
(4H, m, 2Â Cys b-CH2), 2.70 (2H, m, Asp b-CH2), 3.85 (1H, m, Cys a-CH), 4.17
(2H, m, Cys a-CH and Glu a-CH), 4.70 (1H, m, Asp a-CH), 5.22 (1H, d,
J ¼ 8:24 Hz, Glu NH), 7.05 (1H, d, J ¼ 7:90 Asp NH), 7.10 (1H, J ¼ 6:00, Cys
NH), 7.21–7.59 (30H, m, aromatic protons), 7.40 (1H, obs, Cys NH).
3.6.
(12)
The tetrapeptide tri-tert-butyl ester 11 (6.0 g, 4.9 mmol) in MeOH (1500 ml) was
added portionwise at room temperature over 1 h to a stirred solution of I2 (2.5 g,

116
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
9.8 mmol) in MeOH (3500 ml). After 4 h under stirring the reaction mixture was
cooled at 0 °C and decolorized with 1 M Na2S2O3. The residue obtained after re-
moval of the solvent was partitioned between H2O and ethyl acetate and the organic
layer was washed with 1 M Na2S2O3 and H2O. Drying and evaporation followed by
purification by silica gel column chromatography (CHCl3=i Pr-OH 97:3 as eluant) of
the resulting crude product afforded pure disulfide ester 12 (1.2 g, 35%).
½aŠ ¼ À22:0° (c ¼ 1, CHCl3). Rf ¼ 0:35 (CHCl3=i Pr-OH, 97:3). IR (CHCl3): 3415
D
(brd), 1725, 1690, 1505cmÀ1. H NMR (CDCl3) d 1.45 (36H s, 12Â CH3), 1.83
1
(2H, m, Glu b-CH2), 2.30 (2H, m, Glu c-CH2), 2.76 (2H, m, Cys b-CH2), 2.91–
3.29 (4H, m, Cys b-CH2 and Asp b-CH2), 4.13 (1H, m, Glu a-CH), 4.54 (1H, m,
Cys a-CH), 4.61–4.89 (2H, m, Cys a-CH and Asp a-CH), 5.25 (1H, d, J ¼ 7:92 Hz,
Glu NH), 6.86 (1H, d, J ¼ 7:03 Hz, Cys NH), 7.40 (1H, d, J ¼ 6:32 Hz, Cys NH),
7.94 (1H, d, J ¼ 8:8 Hz, Asp NH).
3.7.
(4)
An ice-cooled solution of the protected disulfide 12 (1.2 g, 1.6 mmol) in EtOAc
(11 ml) was treated with dry HCl gas and was then stirred for 1 h at room temperature.
The igroscopic solid which separated was collected and repeatedly washed with diethyl
ether to give 0.7 g of HClÅ
which was used without further purification.
The hydrochloride (0.7 g, 1.4 mmol) was dissolved in 1 M aqueous NH3 (20 ml) at
room temperature. After 30 min the aqueous solution was concentrated and sub-
jected to column chromatography on Sephadex LH-20 using H2O as eluent to afford
disulfide 4 as a white amorphous solid (0.5 g, 80%). ½aŠ ¼ À47:0° (c ¼ 1, H2O).
D
Rf ¼ 0:23 (n-BuOH/acetic acid/H2O, 4:5:1). IR (KBr): 3415 (brd), 1725, 1665,
1530, 1410, 1225cmÀ1. MS M þ Hþ ¼ 467 m=z.
3.8. c-Glu-Cys-Asp-Cys (5)
A solution of the above reported disulfide 4 (0.5 g, 1.0 mmol) in a mixture of n-
PrOH/H2O (2:1) (24 ml) was brought to pH 8.5 with 25% aqueous NH3 and
flushed with nitrogen. Tri-n-butyl phosphine (0.3 g, 1.5 mmol) was added and the
stoppered flask was stirred at room temperature. After 1 h the reaction mixture
was repeatedly washed with CHCl3 and the pH of the aqueous solution adjusted
to 6.0 using 1 M HCl. The solution was concentrated and subjected to column
chromatography on Sephadex using H2O=MeOH 2:1 as eluant to afford compound
5 as a vitrous foam (0.33 g, 70%). ½aŠ ¼ À56:0° (c ¼ 1, H2O). Rf ¼ 0:38 (n-BuOH/
D
acetic acid/H2O, 4:5:1). IR (KBr): 3430 (brd), 3265, 1650, 1520, 1385, 1120cmÀ1. MS
M þ Hþ ¼ 469 m=z.
4. Biochemistry
c-GT from hog kidney (about 3 U/mg) and from bovine kidney (250 U/mg) were
supplied by Fluka and Sigma, respectively. GR from bakers yeast (type III, 260 U/

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
117
mg) and from wheat germ (type II, 0.08 U/mg) were obtained from Sigma. GPNA,
Gly–Gly, GSSG, GSH, NADPH were purchased from Sigma. In vitro assays of
c-GT and GR activity were performed according to Meister et al. [23] and Carlberg
et al. [24], respectively.
5. Results and discussion
In order to test the putative biological activity of glutathione analogues 2–4, in-
hibition studies have been performed on two enzymes: c-GT and GR. The first en-
zyme catalyzes the transfer of the c-glutamyl moiety from many donor molecules
including GSH [25] to a large number of acceptor substrates [26,27]. The second en-
zyme is a flavoprotein catalyzing the NADPH-dependent reduction of glutathione
disulfide (GSSG). The reaction is essential for maintenance of glutathione levels [24].
5.1. GSH analogues 2–4 as c-GT inhibitors
The inhibition of the c-glutamyl binding site of this enzyme has been reported pre-
viously [5,6]. According to Meister et al. [23] the reaction on the donor site (c-glut-
amyl binding) was performed using pseudo-first-order conditions, with a saturating
concentration of the acceptor molecule ([Gly–Gly] ¼ 150 mM). Table 2 shows the re-
sidual enzymatic activity in the presence of specific inhibitors at three different con-
centrations. The two different enzyme batches did not show any differences in
activity. Similar results were obtained with inhibitor 2 using a non-saturating con-
centration of the acceptor molecule ([Gly–Gly] ¼ 15 mM), which is the concentration
usually reported for enzyme activity determinations [23]. Compound 2 is more active
than any of the other compounds tested.
Preincubation of the enzyme with inhibitors at 37 °C shows an effect only for
compound 2 (Fig. 1). The observed improved inhibition with this compound,
which depends from the preincubation time, excludes the possibility that it is
acted on as substrate. A double reciprocal plot of inhibition data for compound
2 shows a non-competitive inhibition pattern at the c-glutamyl site (data not
Table 2
Residual activities of c-GT in the presence of different glutathione analogues 2–4a
Compound
Inhibitor concentration
1 mM (%)
1.4 mM (%)
2.8 mM (%)
2
3
4
79
90
98
68
–
62
85
95b
98
a Enzyme concentration was about 50 mU/ml. GPNA as substrate and the inhibitor were added si-
multaneously to prime the reaction. Data are averages of at least three determinations. The experiments
were carried out in 0.1 M Tris–HCl buffer, pH 8.0, 23 °C.
b At 6 mM concentration of 4, a residual activity of 85% is obtained.

118
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
Fig. 1. Plot of residual enzymatic activity of c-GT in the presence of two different inhibitors. Enzyme
(about 50 mU/ml) was preincubated in 0.1 M Tris–HCl buffer, pH 8.0 at 37 °C for the given time, in the
presence of *: no inhibitor present, r: [2] ¼ 1.4 mM, j: [3] ¼ 1 mM. The activity was then measured by
adding substrates [GPNA] ¼ 1 mM and [Gly–Gly] ¼ 150 mM.
shown) with an approximate Ki value of 0.7 mM. An analogous non-competitive
inhibition was reported also for the c-( -c-oxaglutamyl) analogue of GSH [6],
L
whereas GSH and the c-aza analogue exhibited competitive patterns [5]. These
data suggest different interactions with the c-glutamyl site of the enzyme by these
glutathione analogues.
Compound 3 was also tested as substrate of c-GT. After incubation at 37 °C for
1 h in the presence of the enzyme, the appearance of glutamic acid was monitored by
amino acid analysis. The activity of the enzyme using compound 3 was about 2% of
that using GPNA or using the natural substrate GSH [28]. Similar results were ob-
tained using compound 4.
5.2. GSH analogues 2–4 as GR inhibitors
The above glutathione analogues 2–4 were tested also using GR. Both enzyme
batches were tested for inhibition. The experimental data (Table 3) show the residual
activity of enzyme for three different concentrations of inhibitors. Very little inhibi-
tion was observed. Preincubation of the tested compounds with the enzyme for up to
60 min has no additional effect on activity. Compound 4 was also tested as putative
substrate with NADPH. The reaction followed at 340 nm against suitable blank
showed that the disulfide bond is not reduced by GR.

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
119
Table 3
Residual activities of GR in the presence of different glutathione analogues 2–4a
Compound
Inhibitor concentration
0.5 mM (%)
1.0 mM (%)
2.0 mM (%)
2
3
4
93
83
NTb
85
73
85
90
NT
NT
a Enzyme concentration was about 50 mU/ml. The compound to be tested, [GSSG] ¼ 1 mM and
[NADPH] ¼ 0.1 mM are added in the order and the reaction is followed at 340 nm. Data are averages of at
least three determinations. The experiments were carried out at three concentrations, in 0.2 M Kþ-phos-
phate buffer, pH 7.0, 23 °C.
b NT, not tested.
6. Conclusion
We report here the synthesis of novel glutathione analogues in which the typ-
ical c-glutamyl residue is bound to a Cys-Asp-Cys tripeptide fragment. Both
forms of the resulting tetrapeptide, i.e., the cyclic 11-membered disulfide 4 and
the linear reduced dithiol 5, have been isolated and fully characterized. Interest
in these new type of glutathione analogues is due to the presence of the two cys-
teine residues within a single molecule which in the natural glutathione are
responsible of the dithiol–disulfide oxidation and reduction as well as its antioxi-
dant and scavenger properties. As a consequence of this structural feature and
in contrast to what happens in the GSH–GSSG system, the disulfide bond reduc-
tion and oxidation may be involved in a unimolecular instead of a bimolecular
reaction.
The biological activity of the cyclic disulfide analogues 2–4 has been tested with c-
GT and GR. A wide variety of c-glutamyl derivatives has been in fact tested with
these enzymes and an effective interaction has been shown with compounds in which
the moieties attached to the c-glutamyl group have considerably different structures
[27]. By taking into account the structural analogy of 3 with the native GSH, we can
reasonably argue that the low activity of this analogue is due to an unfavourable al-
teration of the backbone conformation involving the replacement of the native cys-
teine with the Ca;atetrasubstituted 4-amino-1,2-dithiolane-4-carboxylic acid residue.
a-Amino acid residues possessing a highly substituted a-carbon atom, here a spiro
ring system, have restricted conformational freedom. This in turn leads to a more
specific secondary structure capable of strongly influencing both the recognition
and binding of macromolecular targets [29,30]. In the case of the two cyclic disulfide
compounds 2 and 4, containing the 8- and the 11-membered ring, respectively, the
results clearly show that large ring structures with the c-glutamyl residue do not fa-
vour interactions with the tested enzymes; only the 8-membered compound 2 shows
in fact a weak inhibitory activity on c-GT. Thus, in the case of compounds 2 and 4,
in contrast with 3, the impairment in the accessibility to the active site due to steric
reasons seems to be more likely.

120
I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
As shown in Table 3, GR is not able to reduce the disulfide bond of the analogues
nor is it strongly inhibited by these compounds. Because it is well known that GR
prefers open chain dimeric or trimeric substrates [31,32], we can reasonably argue
that the lack of activity of disulfide analogs 2–4 is due to their monomeric nature
which is not compatible with binding at the GR active site.
A final consideration concerns the structure of the GSH analogues 4 and 5. It is
interesting to note that a strategy of structural modification of the GSH molecule
analogous to that which led to 4 and 5, is used by several pathogenic protozoa of
the Trypanosoma and Leishmania genera.These protozoa do not depend on the
GSH–GSSG system for their thiol redox balance and antioxidant defense. Instead
they utilize a specific GSH derivative, N1,N8-diglutathionyl-spermidine, or trypano-
thione, in which the two Cys residues remain covalently bound through a spacer
fragment, both in the reduced and oxidized form, as is the case for 4 and 5 [33]. Thus,
the synthesis of the tetrapeptides 4 and 5 represents a preliminary step towards a new
type of GSH–GSSG analogue which possess significant structural analogy with the
redox system that is operative in several pathogenic microorganisms and may assist
in the development of therapeutic strategies.
Acknowledgments
ꢀ
Financial support from Ministero dellÕUniversita e della Ricerca Scientifica e Tec-
nologica (Cofin 2000) (Italy). We express our appreciation to Prof. Gino Lucente
ꢀ
(Universita ‘‘La Sapienza’’ Rome) for his helpful comments. The authors thank
Dr. C. Foppoli (Centro di Biologia Molecolare del CNR, 00185 Rome, Italy) for
ꢀ
the amino acid analysis and Dr. M. Ricciutelli (Universita degli Studi di Camerino,
Dipartimento di Scienze Chimiche, 62032, Camerino, Italy) for his expert technical
assistance.
References
[1] I. Annis, B. Hargittai, G. Barany, in: G.B. Fields (Ed.), Methods in Enzymology, vol. 289, Academic
Press, New York, 1997, pp. 198–221.
[2] H. Shao, C-W. Lee, Q. Zhu, P. Gantzel, M. Goodman, Angew. Chem. Int. Ed. Engl. 35 (1996) 90–92.
[3] K. Shreder, L. Zhang, M. Goodman, Tetrahedron Lett. 39 (1998) 221–224.
[4] G. Luisi, A. Calcagni, F. Pinnen, Tetrahedron Lett. 34 (1993) 2391–2392.
ꢀ
[5] A. Calcagni, S. Dupre, G. Lucente, G. Luisi, F. Pinnen, D. Rossi, Int. J. Pept. Protein Res. 46 (1995)
434–439.
[6] A. Calcagni, S. Dupre, G. Lucente, G. Luisi, F. Pinnen, D. Rossi, A. Spirito, Arch. Pharm. Pharm.
ꢀ
Med. Chem. 329 (1996) 498–502.
[7] A. Calcagni, G. Lucente, G. Luisi, F. Pinnen, D. Rossi, Amino Acids 17 (1999) 257–265.
[8] E. Morera, M. Nalli, F. Pinnen, D. Rossi, G. Lucente, Bioorg. Med. Chem. Lett. 10 (2000)
1585–1588.
[9] N. Izumiya, J.P. Greenstein, Arch. Biochem. Biophys. 52 (1954) 203–216.
[10] S. Capasso, L. Mazzarella, T. Tancredi, A. Zagari, Biopolymers 23 (1984) 1085–1097.
[11] F. Siedler, S. Rudolph-Bohner, M. Doi, H.-J. Musiol, L. Moroder, Biochemistry 32 (1993) 7488–7495.

I. Cacciatore et al. / Bioorganic Chemistry 31 (2003) 109–121
121
[12] L. Sheh, B.L. Chen, C.F. Chen, Int. J. Pept. Protein Res. 35 (1990) 574–581.
€
[13] B. Schmidt, S. Lindman, W. Tong, G. Lindeberg, A. Gogoll, Z. Lai, M. Thornwall, B. Synnergren,
A. Nilsson, C.J. Welch, M. Sohtell, C. Westerlund, F. Nyberg, A. Karlen, A. Hallberg, J. Med.
ꢀ
Chem. 40 (1997) 903–919.
[14] R. Kishore, P. Balaram, Biopolymers 24 (1985) 2041–2043.
[15] G. Callahan, J.F. Newlander, J.L. Burgess, D.S. Eggleston, A. Nichols, A. Wong, W.F. Huffman,
Tetrahedron 49 (1993) 3479–3488.
[16] M. Sato, J.Y.H. Lee, H. Nakanishi, M.E. Johnson, R.A. Chrusciel, M. Kahn, Biochem. Biophys.
Res. Commun. 187 (1992) 999–1006.
[17] T. Sasagawa, T. Okuyama, Bunseki Kagaku 27 (1978) 621–662.
[18] A.E.P. Adang, J. Brussee, A. van der Gen, G.J. Mulder, J. Biol. Chem. 266 (1991) 830–836.
[19] M.H. Hanigan, Chem-Biol. Interact. 111–112 (1998) 333–342.
[20] R.H. Schirmer, R.L. Krauth-Siegel, G.E. Schulz, in: D. Dolphin, R. Poulson, O. Avramovic (Eds.),
Glutathione: Chemical, Biochemical, and Medical Aspects-Part A, John Wiley, New York, 1989, pp.
554–596.
[21] D.L. Vander Jagt, L.A. Hunsaker, T.J. Vander Jagt, M.S. Gomez, D.M. Gonzales, L.M. Deck, R.E.
Royer, Biochem. Pharmacol. 53 (1997) 1133–1140.
[22] D. Sharma, K. Rajarathnam, J. Biomol. NMR 18 (2000) 165–171.
[23] A. Meister, S.S. Tate, O.W. Griffith, in: W.B. Jacoby (Ed.), Methods in Enzymology, vol. 77,
Academic Press, New York, 1981, pp. 237–242.
[24] I. Carlberg, B. Mannervik, in: W.B. Jacoby (Ed.), Methods in Enzymology, vol. 113, Academic Press,
New York, 1985, pp. 484–490.
[25] A. Meister, A. Tate, Annu. Rev. Biochem. 45 (1976) 559–604.
[26] M. Orlowski, S.S. Meister, Biochim. Biophys. Acta 73 (1963) 679–681.
[27] S.S. Tate, A. Meister, J. Biol. Chem. 249 (1974) 7593–7602.
[28] L.M. Shaw, J.W. London, L.E. Petersen, Clin. Chem. 24 (1978) 905–915.
[29] C. Toniolo, M. Crisma, N. Fabiano, P. Melchiorri, L. Negri, J.A. Krause, D.S. Eggleston, Int. J.
Pept. Protein Res. 44 (1994) 85–95.
[30] I. Torrini, M. Paglialunga Paradisi, G. Pagani Zecchini, G. Lucente, E. Gavuzzo, F. Mazza, G.
Pochetti, S. Traniello, S. Spisani, Biopolymers 42 (1997) 415–426.
[31] W. Janes, G. Schulz, Int. J. Pept. Protein Res. 36 (1990) 134–138.
[32] M. Styblo, D.J. Thomas, Biochem. Pharmacol. 49 (1995) 971–977.
[33] A.F. El-Waer, T. Benson, K.T. Douglas, Int. J. Pept. Protein Res. 41 (1993) 141–146.