Proline-glutamate chimeras in isopeptides. Synthesis and biological evaluation of conformationally restricted glutathione analogues

Article abstract of DOI:10.1016/S0968-0896(03)00041-5

The two novel diastereoisomeric glutathione analogues 1 and 2 have been designed and synthesized by replacing the native γ-glutamylic moiety with the conformational rigid skeleton of cis- or trans-4-carboxy-L-proline residue. Both analogues have been obtained by following the solution phase peptide chemistry methodologies and final reduction of the corresponding disulfide forms 13 and 14. The two analogues 1 and 2 have been tested towards γ-glutamyltranspeptidase (γ-GT) and human glutathione S-transferase (hGST P1-1). Both analogues 1 and 2 are completely resistant to enzymatic degradation by γ-GT. The S-transferase utilizes the analogue 2 as a good substrate while is unable to bind the analogue 1.

Full text of DOI:10.1016/S0968-0896(03)00041-5

Bioorganic & Medicinal Chemistry 11 (2003) 1677–1683
Proline–Glutamate Chimeras in Isopeptides. Synthesis
and Biological Evaluation of Conformationally Restricted
Glutathione Analogues
Mario Paglialunga Paradisi,a Adriano Mollica,a Ivana Cacciatore,b Antonio Di Stefano,b
Francesco Pinnen,b Anna Maria Caccuri,c Giorgio Ricci,c Silvestro Dupre,d
Alessandra Spiritod and Gino Lucentea,*
aIstituto di Chimica Biomolecolare, CNR Sezione di Roma, c/o Dipartimento di Studi Farmaceutici,
`
Universita di Roma ‘La Sapienza’, P.le A. Moro, 00185 Rome, Italy
`
b
Dipartimento di Scienze del Farmaco, Universita ‘G. D’Annunzio’, Via dei Vestini 31, 66100 Chieti, Italy
c
`
Dipartimento di Biologia, Universita ‘Tor Vergata’, Via della Ricerca Scientifica, 00133 Rome, Italy
`
Dipartimento di Scienze Biologiche, Universita di Roma ‘La Sapienza’, P.le A. Moro, 00185 Rome, Italy
d
Received 22 October 2002; accepted 3 January 2003
Abstract—The two novel diastereoisomeric glutathione analogues 1 and 2 have been designed and synthesized by replacing the
native g-glutamylic moiety with the conformational rigid skeleton of cis- or trans-4-carboxy-l-proline residue. Both analogues have
been obtained by following the solution phase peptide chemistry methodologies and final reduction of the corresponding disulfide
forms 13 and 14. The two analogues 1 and 2 have been tested towards g-glutamyltranspeptidase (g-GT) and human glutathione
S-transferase (hGST P1-1). Both analogues 1 and 2 are completely resistant to enzymatic degradation by g-GT. The S-transferase
utilizes the analogue 2 as a good substrate while is unable to bind the analogue 1.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
level of the g-Glu-Cys junction. By following this strat-
egy the first examples of GSH pseudopeptide analogues,
containing urea, urethane or sulfonamido groups replacing
the native g-glutamyl isopeptide bond, have been
described.11À13 Some of these analogues have sub-
sequently been exploited to obtain enzyme inhibitors
which are highly active and completely resistant to
enzymatic degradation by g-glutamyl-transpeptidase
(g-GT).14 As an extension of these studies we report
here synthesis and activity of the GSH analogues 1 and
2 in which the entire g-glutamylic moiety is replaced by
the residue of cis and trans 4-carboxy-l-proline, respec-
tively (Fig. 1).
The highly relevant and multifunctional roles played by
glutathione (g-l-glutamyl-l-cysteinyl-glycine, GSH) in a
variety of biological processes—including xenobiotic
detoxification and protection against radiations—con-
tinue to stimulate both theoretical and experimental
studies.1À8 Furthermore, the structural simplicity and
the low molecular weight, render the GSH molecule an
appealing target for chemical modification; thus, a very
large number of structural analogues have been
designed and synthesized.9,10 As a result of these inves-
tigations a better understanding of the interactions with
enzyme catalytic sites has been obtained and more
potent, selective and stable inhibitors have been devel-
oped. In this research field we reported recently design
and synthesis of a new group of GSH analogues
characterized by chemical modifications centered at the
Proline is the only proteinogenic aminoacid to possess a
naturally restricted conformation, characterized by the
presence of a pyrrolidine ring in which the N–Ca bond is
inserted; thus, use of the chiral and quite rigid proline
ring as template to support side-chain functionalities
typical of other aminoacids (namely proline–aminoacid
chimeras) is an actively pursued approach to obtain
peptidomimetics endowed with specific local restrictions
*Corresponding author. Tel.: +39-06-445-4218; fax: +39-06-491-491;
e-mail: gino.lucente@uniroma1.it
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00041-5

1678
M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
cis- and trans-N-(benzyloxycarbonyl)-4-carboxy-l-pro-
line-1-methyl esters Z-Cpc-OMe 3 and Z-Cpt-OMe 4
were synthesized as key intermediates (Scheme 1).
According to the procedure previously reported by
Bridges et al.17 commercially available trans-4-hydroxy-
l-proline was used as starting material to synthesize
both Z-Cpc(OMe)-OMe and Z-Cpt(OMe)-OMe
(Scheme 1). Subsequent regioselective hydrolysis of the
C-4 ester group represents a critical step. After several
attempts we found that a high selectivity and good
yields of both the cis and trans isomer 3 and 4 can be
obtained when the corresponding dimethyl esters are
treated with equimolar amount of 1N NaOH in aqu-
eous THF (Scheme 1). Structure as well as stereo-
chemistry of monomethyl esters 3 and 4 are clearly
reflected by their 1H NMR spectra. Thus, when the
spectrum of the cis monomethylester 3 is examined and
compared with those of the related dimethyl esters, the
singlet corresponding to the -OMe group at C-4 is
absent and only the couple of singlets, which are typical
of the -OMe group at the C-1 and are due, as expected,
to restricted rotation around the proline CO–N bond,
can be observed. Furthermore, the different resonance
patterns exhibited by cis and trans monoesters allow
exclusion of cross-contamination due to epimerisation
at one of the two centers during the hydrolysis. Ana-
logous considerations, deduced by the analysis of the 1H
NMR spectra, have been made concerning the stereo-
chemistry of the trans monomethylester 4.
Figure 1. Structure of GSH and conformationally restricted GSH
analogues 1 and 2.
and improved metabolic stability.15,16 Concerning pro-
line-glutamate chimeras, a literature examination shows
that, whereas cis- and trans-4-carboxy-l-prolines have
been synthesized and examined for their activity in
neurotransmission processes as glutamate conformer
mimics,17 their incorporation into biologically active
peptides is very limited and no data are available con-
cerning the use of proline as g-glutamylic residue tem-
plate.18 Thus, compounds 1 and 2 represent the first
reported examples of isopeptide analogues in which this
modification strategy has been applied. Finally, it can
be observed that both the new analogues maintain, as
compared with the native GSH, all the functional
groups capable to establish efficient non bonding inter-
actions; the only difference, concerning these properties,
is in fact represented by the presence of a secondary in
place of a primary amino group at the N-terminal end
of the molecule. Furthermore, the relevant confor-
mational restriction imposed by the pyrrolidine ring on
the backbone as well as the opposite spatial orientation
of the amide g-carbonyl in the two analogues may pro-
foundly influence the reactivity towards the GSH
dependent enzymes and render 1 and 2 useful probes to
get information on the structural features of the
involved active sites.
Synthesis of the two conformationally restricted GSH
analogues 1 and 2 were performed according to Scheme
2. In view of the mild protection conditions assured,19
the acetamidomethyl (Acm) group was chosen to pro-
tect the cysteine thiol function. Accordingly, the Acm-
derivatives 7 and 8 were synthesized by coupling Z-Cpc-
OMe or Z-Cpt-OMe 3 and 4, respectively, with
Cys(Acm)-Gly-OMe trifluoroacetate 6 (Scheme 2) by
following the solution phase peptide chemistry metho-
1
dologies. H NMR spectra of the two tripeptides 7 and
8 did not show evidence of diastereoisomeric species,
thus confirming the stereochemical integrity of starting
monoesters 3 and 4. Selective removal (I2 in MeOH,
room temperature) of the Acm protecting group
allowed direct conversion of 7 and 8 into the N-pro-
tected symmetrical disulfides 9 and 10. Subsequent
Results and Discussion
As useful route to the GSH analogues 1 and 2, the
incorporation of cis- and trans-4-carboxy-l-proline resi-
dues (Cpc and Cpt, respectively) into the Cys-Gly frag-
ment was considered. Thus, the two diastereoisomeric
Scheme 1. Reagents and conditions (see also ref 17): (a) ClCO2CH2Ph, NaHCO3, H2O, PhCH3, 16 h, rt; (b) EtOH, p-TsOH, 16 h, reflux; (c) TsCl, py,
72h, rt; (d) NaCN, DMSO, 80 ꢀC, 4 h; (e) HCl, MeOH, 84 h, rt; (f) Jones reagent, acetone, 10 min, À5 ꢀC then 2.5 h, rt; (g) NaBH4, MeOH, 20 h,
À5 ꢀC; (h) THF, NaOH 1 N, 0.5 equiv, 30 min, 0 ꢀC then 1 h, rt.

M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
1679
Scheme 2. Reagents and conditions: (a) DCCI, NMM, THF, 0 ꢀC, 3 h then 5 ꢀC, 16 h; (b) TFA, 1 h, rt; (c) isobutyl chloroformate, TEA, THF,
À15 ꢀC, 15 min then 0 ꢀC, 2h and 5 ꢀC, 16 h; (d) I2, MeOH, 4 h, rt; (e) NaOH 1 N, THF, 0 ꢀC, 20 min then 16 h, rt; (f) HBr/CH3COOH, 5 h, rt;
(g) 2N aq ammonia, rt, 20 min; (h) ( n-But)3P, n-PrOH–H2O, 3 h, rt.
removal in two steps of all protecting groups, followed
by treatment of the resulting hydrobromides with aqu-
eous ammonia gave the fully deprotected GSSG ana-
logues 13 and 14 in good yields. The desired GSH thiol
forms 1 and 2 were obtained by reductive cleavage of
the disulfide link with tri-n-butylphosphine (1.2equiv)
in water/n-propanol solution, by operating at pH 8.5
(aqueous ammonia).11 Mass spectra of 1 and 2 show the
most abundant peaks corresponding to M++1 species
(320 m/z). The structures assigned to the GSH ana-
logues 1 and 2 are in accordance with their spectro-
scopic properties (see Experimental). A characteristic
feature of the 13C NMR spectrum of 1 and 2 is repre-
sented by the resonance of the Cys b-carbon atom
which is shifted at higher field (32.45 and 25.43 ppm) as
compared with the corresponding atoms in the disulfide
precursors 13 and 14 (38.90 and 38.87 ppm, respec-
tively). This spectral behaviour represents, as recently
noted,20 a reliable diagnostic feature in order to distin-
guish between cysteine and cystine containing forms.
g-glutamyl compounds to a variety of acceptors such as
aminoacids and dipeptides (transpeptidation) and water
(hydrolysis).21,22 Its proposed physiological role
includes aminoacid and peptide transport as well as
detoxification reactions through mercapturic acids for-
mation. The reaction mechanism involves a covalent
g-glutamyl-enzyme intermediate and specificity studies
have shown that the g-glutamyl site (donor site) of the
enzyme accepts many g-glutamyl derivatives, including
d-forms as well as a-methyl derivatives.23 Other
modifications of the g-glutamyl backbone are found in
glutamine-based affinity labels for the same enzyme site.
The reported GSH analogues 1 and 2 show unusual
structural and stereochemical alterations at the level of
the g-glutamyl moiety and could therefore interfere with
the donor site of g-GT or even act as substrates. Thus
both 1 and 2 have been tested as putative inhibitors or
substrates of the enzyme.
The enzymatic activity in the presence of 1 mM con-
centration of the substrate l-g-glutamyl-p-nitroanilide
(GPNA) is completely unaffected by the presence of
compound 1 or 2 at 1 mM or 2mM concentration even
after preincubation at 37 ꢀC of the enzyme with the
putative inhibitor for 30 or 60 min.
Activity on ꢀ-Glutamyltranspeptidase (ꢀ-GT)
g-GT is a membrane bound enzyme which catalyzes the
transfer of the g-glutamyl residue of GSH and related

1680
M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
Thus, at these concentrations the tested compounds are
not inhibitors of the enzyme. Furthermore they do not
show detectable affinity for the g-glutamyl site, as would
be if they were affinity labels or even substrates, since an
apparent inhibition would be observed.
Conclusions
In conclusion, we report here an efficient synthesis of
the two novel diastereoisomeric glutathione analogues 1
and 2 containing the conformational rigid skeleton of
cis- or trans-4-carboxy-l-proline residue, respectively, in
place of the native g-glutamyl moiety. At the best of our
knowledge these analogues represent the first models in
which the conformational freedom of native GSH has
been reduced by introducing a ring equivalent of the
glutamate residue.
Activity on human glutathione S-transferase P1-1 (hGST
P1-1)
The glutathione S-transferases (GSTs, EC 2.5.1.18) are
a superfamily of multifunctional enzymes involved in
the cellular detoxification of xenobiotic compounds via
a conjugation reaction with GSH which is highly spe-
cific for the GST active site.24
As reported above the biological activity of analogues 1
and 2 has been tested on g-GT and GST. Compounds 1
and 2 are not substrates or inhibitors on g-GT; these
results clearly show that the conformational restricted
backbone of these analogues is apparently not compa-
tible with binding to the donor site of g-GT thus ren-
dering these models resistant to the enzymatic
degradation by this enzyme.
Enzymatic experiments were performed to check the
ability of the hGST P1-1 to replace GSH with the GSH
analogues 1 and 2. In the hGST P1-1 catalysed nucleo-
philic reaction with the electrophilic substrate 1-chloro-
2,4-dinitrobenzene (CDNB), the natural GSH was
replaced either by GSH analogue 1 or 2. In the presence
of 1, a specific activity not higher than 1.0 U/mg was
observed up to 3 mM analogue concentration. This
activity corresponds to about 1% of that found with the
natural substrate GSH. The absence of any competitive
inhibition toward GSH up to 10 mM concentration of
analogue 1 (data not shown) suggests that the lack of
activity is merely due to a scarce affinity for the GSH
binding site and not to a non-productive binding of
analogue 1. On the contrary, compound 2 showed a
quite good interaction with the hGST P1-1 active site
and it is actively conjugated to CDNB. The dependence
of the GST activity on the concentration of the GSH-
analogue (Fig. 2), identifies a Km=0.7Æ0.2mM, which
is about 3.5-fold higher than the Km for GSH
(KGmSH=0.2mM). At saturating analogue concen-
tration, the specific activity was 38Æ3.4 U/mg, which is
only 2-fold lower than the specific activity found with
the nucleophilic substrate GSH (80 U/mg at pH 6.5 and
25 ꢀC).
It is well known that the g-glutamyl moiety is the main
binding determinant for the GSH-GST interaction, giv-
ing multiple polar interaction with specific protein resi-
dues, that is Asp98, Gln64 and Ser65.25 The here
reported results indicate that the transoid spatial
arrangement involving the Ca and Cg pyrrolidine car-
boxy groups, present in the analogue 2, represents the
favoured conformation for binding to the GST active
site. However, further investigations are requested in
order to define which of the structural and confor-
mational features distinguishing analogues 1 and 2 from
the native GSH are specifically responsible of the
observed binding selectivity at the enzymatic active site.
It should be remembered here that the GTSs super-
family may play an important part in cellular resistance
to foreign, electrophilic toxins as well as to many anti-
cancer drugs which are also electrophilic in nature.26 In
particular, the hGST P1-1 isoform represents an attrac-
tive target since it has been found to be over-expressed
in a number of solid tumors, thus diminishing the effec-
tiveness of cytotoxic drugs.27 As a consequence, the
versatily of the here described synthetic approach which
relies upon the utilization of a chiral 4-carboxy-proline
skeleton as glutamate conformer mimic could give rise
to new interesting peptidase stable GSH inhibitors pro-
viding tumor-directed potentiation of conventional
antineoplastic agents.
Experimental
Chemistry
t
.
H-Gly-OBu HCl
Boc-Cys(Acm)-OH
and
were
obtained from Novabiochem. TLC was performed on
Merck 60 F254 silica gel plates developed with the fol-
lowing solvent system: (a) CHCl3/MeOH (95:5); (b)
CHCl3/i-PrOH (95:5); (c) n-BuOH/AcOH/H2O (3:2:2).
Column chromathography was carried out using Merck
60 silica gel (230–400 mesh). Optical rotations were
taken at 25 ꢀC with Perkin–Elmer 241 polarimeter. IR
Figure 2. Conjugation reaction between CDNB and the GSH ana-
logue 2. The solid line is the best fit of the experimental data to a
hyperbolic binding equation which fulfils a Km value of 0.7Æ0.2mM
and a Vmax of 38Æ3.4 U/mg.

M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
1681
spectra were recorded employing a Perkin–Elmer 1600
FTIR spectrophotometer. 1H and 13C NMR spectra
were determined on a Varian VXR 300 MHz instrument
(d expressed in ppm). 13C NMR data for compounds 1,
2, 13 and 14 were confirmed by APT experiments. The
mass spectra of 1 and 2 have been obtained in electro-
spray (ES) conditions by a LCQ (Finnigan, S. Jose, CA,
USA) instrument. The 60 mM solution of the sample in
CH3OH/H2O 50:50 containing 1% AcOH has been
directly injected at the flow rate of 5 mL/min. Abbre-
viations: THF, tetrahydrofuran; DCC, dicyclohexyl-
carbodiimide; NMM, N-methylmorpholine; EtOAc,
ethyl acetate; TEA, triethylamine; TFA, trifluoroacetic
acid; DMSO, dimethylsulphoxide.
b-CH2), 3.75 (3H, s, OCH3), 3.90–4.15 (2H, m, Acm
CH2), 4.30 (1H, m, Cys a-CH), 4.45–4.55 (2H, m, Gly
CH2), 5.65 (1H, d, J=7.5 Hz, Cys NH), 7.20 (1H, br t,
Gly NH), 7.35 (1H, br t, Acm NH).
Z-Cpc[-Cys(Acm)-Gly-OMe]-OMe (7) and Z-Cpt[-Cys
(Acm)-Gly-OMe]-OMe (8). The above reported di-
peptide ester 5 (3.3 g, 9.1 mmol) was dissolved in TFA
(10 mL). After 1 h at room temperature the solution was
evaporated to dryness and the residue repeatedly evapo-
.
rated with ether to give TFA H-Cys(Acm)-GlyOMe 6 (3.4
g, 100%) which was used without further purification.
To an ice-cold solution of Z-Cpc-OMe 3 or Z-Cpt-OMe
4 (2.3 g, 7.5 mmol) in dry THF (28 mL) TEA (1.1 mL,
7.5 mmol) and isobutyl chloroformate (1 mL, 7.5 mmol)
were added under stirring. After 15 min at À15 ꢀC,
methyl ester 6 (7.5 mmol) in TEA (1.1 mL, 7.5 mmol)
and dry THF (20 mL) were added to the mixture at
À15 ꢀC with stirring. After 2h at 0 ꢀC and 16 h at 5 ꢀC,
the reaction mixture was filtered and the resulting solu-
tion was evaporated under vacuum. Work up, as descri-
bed for compound 5, gave an oily product which was
purified on silica gel using CHCl3/MeOH (9:1) as eluant
to yield the corresponding pure tripeptide metilesters.
Z-Cpc-OMe (3) and Z-Cpt-OMe (4). N-(Benzyloxy-
carbonyl)-4-carboxy-l-proline dimethyl ester (3.2g, 10
mmol) was dissolved in THF (21 mL) and 1 N NaOH
(10 mL) was added dropwise at 0 ꢀC with stirring. After
30 min at 0 ꢀC and 1 h at room temperature, the reac-
tion mixture was concentrated and treated with ether
and saturated acqueous NaHCO3. Acidification of aqu-
eous layer with concentrated HCl caused precipitation
of the oily product which was extracted into EtOAc.
The residue obtained after drying and evaporation of
the organic layer was chromatographed on silica gel
using CH2Cl2/MeOH (99:1) as eluant to give the corris-
ponding monomethyl esters.
(7). White foam (2.3 g, 55%). Rf (a)=0.29; [a]D=À33ꢀ
(c 1, CHCl3); IR (KBr): 3300 br, 3065, 1745, 1655, 1540
1
cmÀ1; H NMR (CDCl3) d=2.05 (3H, s, CH3CO), 2.42
(3). Oil (2.0 g, 65%). Rf (a)=0.39; [a]D=À42ꢀ (c 1,
(1H, m, Cpc b-CHA), 2.53 (1H, m, Cpc b-CHB), 2.75–
3.01 (2H, m, Cys b-CH2), 3.05 (1H, m, Cpc g-CH), 3.60
(1H, m, Cpc d-CHA), 3.73 (6H, s, 2ÂOCH3), 3.97 (1H,
m, Cpc d-CHB), 3.90–4.15 (2H, m, Acm CH2), 4.40–4.60
(2H, m, Gly CH2), 4.50 (1H, m, Cpc a-CH), 4.80 (1H,
m, Cys a-CH), 5.02–5.21 (2H, m, PhCH2), 7.15 (2H, br,
Cys NH and Acm NH), 7.26–7.37 (5H, aromatics), 7.42
(1H, br t, Gly NH).
CHCl3); IR (nujol): 3500–2800 br, 1740, 1705, 1415
1
cmÀ1; H NMR (CDCl3) d=2.42 (1H, m, Cpc b-CHA),
2.51 (1H, m, Cpc b-CHB), 3.14 (1H, m, Cpc g-CH), 3.55
and 3.72(3H, s, OCH 3), 3.79–3.97 (2H, m, Cpc d-CH2),
4.41 (1H, m, Cpc a-CH), 5.02–5.21 (2H, m, PhCH2),
6.50 (1H, br, COOH), 7.26–7.37 (5H, aromatics).
(4). Oil (2.0 g, 65%). Rf (a)=0.46; [a]D=À48.9ꢀ (c 1,
CHCl3); IR (nujol): 3500–2800 br, 1740, 1710, 1410
(8). White foam (2.8 g, 68%). Rf (a)=0.32;
[a]D=À23.6ꢀ (c 1, CHCl3); IR (KBr): 3310 br, 1745,
1
cmÀ1; H NMR (CDCl3) d=2.23 (1H, m, Cpt b-CHA),
2.52 (1H, m, Cpt b-CHB), 3.26 (1H, m, Cpt g-CH), 3.59
and 3.76 (3H, s, OCH3), 3.72(1H, m, Cpt d-CHA), 3.89
(1H, m, Cpt d-CHB), 4.51 (1H, m, Cpt a-CH), 5.01–5.21
(2H, m, PhCH2), 5.70 (1H, br, COOH), 7.26–7.36 (5H,
aromatics).
1655, 1540 cmÀ1 1H NMR (CDCl3) d=2.05 (3H, s,
;
CH3CO), 2.22 (1H, m, Cpt b-CHA), 2.52 (1H, m, Cpt
b-CHB), 2.70 (1H, m, Cys b-CHA), 2.95 (1H, m, Cys b-
CHB), 3.15 (1H, m, Cpt g-CH), 3.60 (3H, s, OCH3), 3.68
(1H, m, Cpt d-CHA), 3.73 (3H, s, OCH3), 3.91 (1H, m,
Cpt d-CHB), 3.95–4.15 (2H, m, Acm CH2), 4.30 (1H, m,
Gly CHA), 4.56 (1H, m, Cpt a-CH), 4.60 (1H, m, Gly
CHB), 4.80 (1H, m, Cys a-CH), 5.01–5.21 (2H, m,
PhCH2), 6.80 (1H, br t, J=6.15 Hz, Gly NH), 6.97 (1H,
br t, J=7.90 Hz, Cys NH), 7.25–7.35 (5H, aromatics),
7.36 (1H, obs, Acm NH).
Boc-Cys(Acm)-Gly-OMe (5). To a stirred solution of
Boc-Cys(Acm)-OH (6.0 g, 20.5 mmol) in dry THF (8
.
mL) H-Gly-OMe HCl (2.6 g, 20.5 mmol) and NMM
(2.2 mL, 20.5 mmol) in dry THF (8 mL) were added at
0 ꢀC followed by portionwise addition of DCC (4.2g,
20.5 mmol) in dry THF (8 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 EtOAc and the organic layer
washed with 1 N KHSO4, saturated aqueous NaHCO3
and H2O. The residue obtained after drying and eva-
poration was cromatographed on silica gel using
CHCl3/MeOH (95:5) as eluant to give dipeptide
methylester 5 as a white foam (7.0 g, 94%). Rf (a)=0.33;
[a]D=+26.5ꢀ (c 1, CHCl3); IR (KBr): 3450, 3300, 1745,
[Z-Cpc(-Cys-Gly-OMe)-OMe]2 (9) and [Z-Cpt(-Cys-
Gly-OMe)-OMe]2 (10). To a stirred solution of the
above reported S-protected tripeptide esters (2.2 g, 4
mmol) in MeOH (48 mL) I2 (2.03 g, 8 mmol) in MeOH
(24 mL) was added portionwise at room temperature
during 45 min. After 4 h under stirring the reaction
mixture was cooled at 0 ꢀC and decoulorized with 1 N
Na2S2O3. The residue obtained after removal of the
solvent was partioned between H2O and EtOAc and the
organic layer was washed with 1 N Na2S2O3 and H2O.
Drying and evaporation followed by purification by
1690, 1500 cmÀ1
;
1H NMR (CDCl3) d=1.47 (9H, s,
3ÂCH3), 2.00 (3H, s, CH3CO), 2.90 (2H, m, Cys

1682
M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
silica gel column chromatography (CHCl3/i-PrOH 95:5
as eluant) of the resulting crude products afforded the
corresponding pure disulfide esters.
over KOH. The obtained oil was dissolved in anhydrous
MeOH; precipitation with anhydrous ether yielded
[HBr Pro(-Cys-Gly-OH)-OH]2 in quantitative yield
which was used as such.
(9). Foam (1.3 g, 34%). Rf (b)=0.29; [a]D=À18.5ꢀ (c 1,
CHCl3); IR (KBr): 3480, 3305, 1745, 1690, 1540 cmÀ1
;
The hydrobromide (0.8 g, 1.0 mmol) was dissolved in
2N aqueous NH 3 (18 mL) at room temperature. After 2
h, the aqueous solution was concentrated and subjected
to column chromatography on Sephadex LH-20 using
H2O/MeOH (2:1) as eluant to afford the corresponding
disulfides.
1H NMR (CDCl3) d=2.30 (2H, m, Cpc b-CHA), 2.59
(2H, m, Cpc b-CHB), 2.82 (2H, m, Cys b-CHA), 3.01
(2H, m, Cys b-CHB), 3.07 (2H, m, Cpc g-CH), 3.55 (6H,
s, OCH3), 3.72(H2 , m, Cpc
d-CHA), 3.76 (6H, s,
OCH3), 3.87 (2H, m, Cpc d-CHB), 4.00–4.15 (4H, m,
Gly CH2), 4.40 (2H, m, Cpc a-CH), 5.01–5.22 (4H, m,
PhCH2), 5.48 (2H, m, Cys a-CH), 7.10 (2H, br t,
J=10.45 Hz, Cys NH), 7.26–7.35 (10H, aromatics), 8.41
(2H, t, J=4.95 Hz, Gly NH).
(13). Foam (0.6 g, 85%). Rf (c)=0.54; [a]D=À30.7ꢀ (c
1, H2O); IR (KBr): 3240 br, 1680–1530, 1395, 1250
1
cmÀ1; H NMR (D2O) d=2.10 (2H, m, Cpc b-CHA),
2.62 (2H, m, Cpc b-CHB), 2.75–2.98 (4H, m, Cys
b-CH2), 3.26 (2H, m, Cpc g-CH), 3.37–3.56 (4H, m, Cpc
d-CH2), 3.65 (4H, AB q, J=18.0 Hz, Gly CH2), 4.08
(2H, m, Cpc a-CH), 4.46 (2H, m, Cys a-CH); 13C NMR
(D2O) d=33.07 (Cpc Cb), 38.90 (Cys Cb), 42.84 (Cpc
Cg), 43.51 (Gly Ca), 47.56 (Cpc Cd), 53.02(Cys C a),
61.39 (Cpc Ca), 171.79, 173.32, 173.66 and 176.38
(4ÂCO).
(10). Foam (1.6 g, 36%). Rf (b)=0.29; [a]D=À26.1ꢀ (c
1, CHCl3); IR (KBr): 3465, 3305, 1750, 1690, 1540
1
cmÀ1; H NMR (CDCl3) d=2.16 (2H, m, Cpt b-CHA),
2.58 (2H, m, Cpt b-CHB), 2.78–2.97 (4H, m, Cys
b-CH2), 3.15 (2H, m, Cpt g-CH), 3.67 (6H, s, OCH3),
3.74 (6H, s, OCH3), 4.00 (4H, m, Gly CH2), 4.02(4H,
m, Cpt d-CH2), 4.45 (2H, m, Cpt a-CH), 5.01–5.18 (4H,
m, PhCH2), 5.52(H2 , m, Cys
a-CH), 6.94 (2H, t,
J=6.60 Hz, Cys NH), 7.25–7.35 (10H, aromatics), 8.37
(2H, m, Gly NH).
(14). Foam (0.7 g, 70%). Rf (c)=0.32; [a]D=À32.4ꢀ (c
1, H2O); IR (KBr): 3300 br, 1690–1540, 1390, 1250
1
cmÀ1; H NMR (D2O) d=2.35 (4H, m, Cpt b-CH2),
[Z-Cpc(-Cys-Gly-OH)-OH]2 (11) and [Z-Cpt(-Cys-Gly-
OH)-OH]2 (12). The protected disulfide (1.2g, 1.52
mmol) was dissolved in dry THF (3.1 mL) and 1 N
NaOH (5.6 mL) was added under stirring at 0 ꢀC. After
30 min at 0 ꢀC and 18 h at room temperature, the pH
was adjusted to 6.0 by 1 N KHSO4 and the solution
evaporated in vacuo to dryness and dried over KOH.
The resulting residue gave the corresponding TLC-pure
disulfide acids which were used as such.
2.83 (2H, m, Cys b-CHA), 3.11 (2H, m, Cys b-CHB),
3.24 (2H, m, Cpt g-CH), 3.39 (2H, m, Cpt d-CHA), 3.51
(2H, m, Cpt d-CHB), 3.62(4H, s, Gly CH ), 4.14 (2H,
2
m, Cpt a-CH), 4.61 (2H, m, Cys a-CH); 13C vNMR
(D2O) d=33.09 (Cpt Cb), 38.87 (Cys Cb), 42.47 (Cpt
Cg), 43.62(Gly C a), 47.93 (Cpt Cd), 52.90 (Cys Ca),
61.44 (Cpt Ca), 171.61, 173.89, 173.97 and 176.33
(4ÂCO).
H-Cpc(-Cys-Gly-OH)-OH (1) and H-Cpt(-Cys-Gly-
OH)-OH (2). A solution of the foregoing disulfide acid
(0.4 g, 0.6 mmol) in a mixture of n-PrOH/H2O (2:1) (12
mL) was brought to pH 8.5 with 25% aqueous NH3 and
flushed with nitrogen. Tri-n-butyl phosphine (0.15 g,
0.72mmol) added and the stoppered flask stirred at
room temperature. After 3 h the reaction mixture was
repeatedly washed with CHCl3 and the pH of the aqu-
eous solution adjusted to 6.0 by 1 N HCl. The solution
was concentrated and subjected to column chromato-
graphy on Sephadex using H2O/MeOH (2:1) as eluant
to afford the corresponding reduced compounds.
(11). Foam (1.1 g, 96%). Rf (c)=0.57; [a]D=À17.5ꢀ (c
1, H2O); IR (KBr): 3600–3100, 1700, 1675, 1540 cmÀ1
;
1H NMR (DMSO-d6) d=2.15 (2H, m, Cpc b-CHA),
2.42 (2H, m, Cpc b-CHB), 2.90 (2H, m, Cys b-CHA),
2.95–3.15 (4H, m, Cys b-CHB and Cpc g-CH), 3.35 (4H,
m, Cpc d-CH2), 3.75 (4H, m, Gly CH2), 4.20 (2H, m,
Cpc a-CH), 4.55 (2H, m, Cys a-CH), 5.01–5.23 (4H, m,
PhCH2), 7.26–7.35 (10H, aromatics), 8.35 (4H, br, Gly
NH and Cys NH).
(12). Foam (1.4 g, 98%). Rf (c)=0.69; [a]D=À39.8ꢀ (c
1, H2O); IR (KBr): 3600–3100, 1700, 1675, 1540 cmÀ1
;
1H NMR (D2O) d=1.99 (2H, m, Cpt b-CHA), 2.30 (2H,
m, Cpt b-CHB), 2.70 (2H, m, Cys b-CHA), 3.05–3.25
(4H, m, Cys b-CHB and Cpt g-CH), 3.45–3.75 (8H, m,
Cpt d-CH2 and Gly CH2), 4.15 (2H, m, Cpt a-CH), 4.60
(2H, m, Cys a-CH), 4.90–5.10 (4H, m, PhCH2), 7.25–
7.35 (10H, aromatics), 8.35 (4H, br, Gly NH and Cys
NH).
(1). Vitrous foam (0.3 g, 98%). Rf (c)=0.39; [a]D=À27.9ꢀ
(c 1, H2O); IR (KBr): 3250 br, 1690–1530, 1395, 1255
cmÀ1; 1H NMR (D2O) d=2.11 (1H, m, Cpc b-CHA), 2.59
(1H, m, Cpc b-CHB), 2.74 (1H, m, Cys b-CHA), 2.88 (1H,
m, Cys b-CHB), 3.24 (1H, m, Cpc g-CH), 3.36–3.54 (2H,
m, Cpc d-CH2), 3.83 (2H, Gly CH2), 4.15 (1H, m, Cpc
a-CH), 4.43 (1H, m, Cys a-CH); 13C NMR (D2O)
d=32.45 (Cys Cb), 32.92 (Cpc Cb), 41.52(Gly C a), 42.71
(Cpc Cg), 47.70 (Cpc Cd), 52.98 (Cys Ca), 60.89 (Cpc
Ca), 172.56, 172.95, 173.46 and 173.70 (4ÂCO); MS
[H-Cpc(-Cys-Gly-OH)-OH]2 (13) and [H-Cpt(-Cys-Gly-
OH)-OH]2 (14). A solution of the above reported
N-protected disulfide acid (1.0 g, 1.1 mmol) was treated
with 30% HBr in glacial AcOH (20 mL) at room tem-
perature. After 5 h the solution was evaporated under
reduced pressure below 40 ꢀC and the residue was dried
+
(m/z) 32 0 (M +1).
(2). Vitrous foam (0.3 g, 90%). Rf (c)=0.43;
[a]D=À28.9ꢀ (c 1, H2O); IR (KBr): 3300 br, 1695–1540,

M. P. Paradisi et al. / Bioorg. Med. Chem. 11 (2003) 1677–1683
1683
1
1340, 1250 cmÀ1; H NMR (D2O) d=2.26 (2H, m, Cpt
4. Duckert, J. F.; Balas, L.; Rossi, J. C. Tetrahedron Lett.
2001, 42, 3709.
5. Cavero, M.; Hobbs, A.; Madge, D.; Motherwell, W. B.;
Selwood, D.; Potier, P. Bioorg. Med. Chem. Lett. 2000, 10,
641.
6. 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.
7. Burg, D.; Hameetman, L.; Filippov, D. V.; van der Marel,
G. A.; Mulder, G. J. Bioorg. Med. Chem. Lett. 2002, 12, 1579.
8. Li, Y.; Carter, D. E.; Mash, E. A. Synthetic Commun. 2002,
32, 1579.
b-CH2), 2.73 (2H, m, Cys b-CH2), 3.16 (1H, m, Cpt
g-CH), 3.30–3.47 (2H, m, Cpt d-CH2), 3.57 (2H, Gly
CH2), 4.10 (1H, m, Cpt a-CH), 4.34 (1H, m, Cys a-CH);
13C NMR (D2O) d=25.43 (Cys Cb), 32.97 (Cpt Cb),
42.18 (Cpt Cg), 43.28 (Gly Ca), 47.68 (Cpt Cd), 55.65
(Cys Ca), 61.07 (Cpt Ca), 171.53, 173.85, 173.95 and
+
176.37 (4ÂCO); MS (m/z) 32 0 (M +1).
Biological assays
9. Douglas, K. T. In Offprints from Glutathione: Chemical,
Biochemical and Medical Aspects; Dolphin, D.; Poulson, R.;
Avramovic, O. Eds.; Wiley & Sons: New York, 1989; Part A.
10. Lucente, G.; Luisi, G.; Pinnen, F. Il Farmaco 1998, 53,
721.
11. Calcagni, A.; Dupre, S.; Lucente, G.; Luisi, G.; Pinnen,
F.; Rossi, D. Int. J. Pept. Protein Res. 1995, 46, 434.
12. Calcagni, A.; Dupre, S.; Lucente, G.; Luisi, G.; Pinnen,
F.; Rossi, D.; Spirito, A. Arch. Pharm. Pharm. Med. Chem.
1996, 329, 498.
13. Luisi, G.; Calcagni, A.; Pinnen, F. Tetrahedron Lett. 1993,
34, 2391.
14. Vince, R.; Brownell, J.; Akella, L. B. Bioorg. Med. Chem.
Lett. 1999, 9, 853.
15. Sharma, R.; Lubell, W. D. J. Org. Chem. 1996, 61, 202,
and references cited therein.
16. Kitas, E. A.; Loeffer, B. M.; Daetwyler, S.; Dehmlow, H.;
¨
Aebi, J. D. Bioorg. Med. Chem. Lett. 2002, 12, 1727.
17. Bridges, R. J.; Stanley, M. S.; Anderson, M. W.; Cotman,
C. W.; Chamberlin, A. R. J. Med. Chem. 1991, 34, 717.
18. Herman, J. P. Eur. Patent 0 334 244 A2, 1989. Chem.
Abstr. 1990, 112, 119449.
Bovine kidney g-GT (about 26 U/mg), GPNA, H-Gly-
Gly-OH and CDNB were purchased from Sigma. In
vitro assay of g-GT activity was performed according to
Meister et al.28 at saturating concentration of the
acceptor molecule (H-Gly-Gly-OH=150 mM).
Human placenta GST P1-1 (EC 2.5.1.18) was expressed
in Escherichia coli and purified as previously descri-
bed.29 Enzymatic experiments concerning the activity of
GSH analogues 1 and 2 on hGST P1-1 were performed
at 25 ꢀC in the presence of 1 mM CDNB as acceptor
substrate. GST P1-1 activity was measured in 1 mL of
0.1 M K-phosphate buffer pH 6.5 containing 1 mM
CDNB, 0.1 mM EDTA and variable amounts of either
GSH or GSH analogue (1 or 2) ranging from 0.1 to 3
mM. Activity was recorded at 340 nm, where the
DNB-conjugate absorbs (e=9.6 mMÀ1 cmÀ1), using a
double-beam Uvikon spectrometer thermostatted at
25 ꢀC. One unit of GST activity is defined as the amount
of enzyme that catalyzes the formation of 1 mmol of
DNB-conjugate per min at 25 ꢀC.
19. Kamber, B. Helv. Chim. Acta 1971, 54, 927.
20. Sharma, D.; Rajarathnam, K. J. Biomol. NMR 2000, 18,
165.
21. Meister, A.; Tate, S. S. Ann. Rev. Biochem. 1976, 45, 559.
22. Tate, S. S.; Meister, A. Mol. Cell. Biochem. 1981, 39, 357.
23. Douglas, K. T. Adv. Enzymol. 1987, 59, 103.
24. Armstrong, R. N. Chem. Res. Toxicol. 1997, 10, 2 .
25. Oakley, A. J.; Lo Bello, M.; Battistoni, A.; Ricci, G.;
Rossjohn, J.; Villar, H. O.; Parker, M. W. J. Mol. Chem. 1997,
274, 84.
Acknowledgements
This work was partially supported by Ministero
dell’Universita e della Ricerca Scientifica e Tecnologica
(MIUR, 40%, Italy).
26. Salinas, A. M.; Wong, M. G. Curr. Med. Chem. 1999, 6,
279, and references cited therein.
References and Notes
27. O’Brien, M. L.; Tew, K. D. Eur. J. Cancer 1996, 32A, 967.
28. Meister, A.; Tate, S. S.; Griffith, O. W. In Methods
Enzymology; Jacoby, W. B., Ed.; Academic: New York, 1981;
Vol. 77, p 237.
29. Ricci, G.; Del Boccio, G.; Pennelli, A.; Lo Bello, M.; Pet-
ruzzelli, R.; Caccuri, A. M.; Barra, D.; Federici, G. J. Biol.
Chem. 1991, 266, 21409.
1. Lyon, R. P.; Atkins, W. M. J. Am. Chem. Soc., 2001, 123,
4408.
2. Rauk, A.; Armstrong, D. A.; Berges, J. Can. J. Chem. 2001,
79, 405.
3. Diaz, S.; Gonzales, A.; Gonzales de Riancho, S.; Rodrı-
guez, A. J. Organomet. Chem. 2000, 610, 2 5.