Design, synthesis and biological evaluation of glutathione peptidomimetics as components of anti-Parkinson prodrugs

Article abstract of DOI:10.1021/jm800239v

Plethoras of CNS-active drugs fail to effect their pharmacologic response due to their in vivo inability to cross the blood-brain barrier (BBB). The classical prodrug approach to overcome this frailty involves lipophilic derivatives of the polar drug, but we herein report a novel approach by which endogenous transporters at BBB are exploited for brain drug delivery. The crucial role played by glutathione in pathogenesis of Parkinson's and the presence of its influx transporters at the basolateral membrane of BBB served as the basis for our anti-Parkinson prodrug design strategy. A metabolically stable analogue of glutathione is used as a carrier for delivery of dopamine and adamantamine. An account of successful syntheses of these prodrugs along with their transport characteristics and stability determination is discussed.

Full text of DOI:10.1021/jm800239v

J. Med. Chem. 2008, 51, 4581–4588
4581
Design, Synthesis and Biological Evaluation of Glutathione Peptidomimetics as Components of
Anti-Parkinson Prodrugs
Swati S. More† and Robert Vince*
Center for Drug Design, Academic Health Center, and Department of Medicinal Chemistry, College of Pharmacy, UniVersity of Minnesota,
8-123A WeaVer Densford Hall, 308 HarVard Street SE, Minneapolis, Minnesota 55455
ReceiVed March 5, 2008
Plethoras of CNS-active drugs fail to effect their pharmacologic response due to their in vivo inability to
cross the blood-brain barrier (BBB). The classical prodrug approach to overcome this frailty involves
lipophilic derivatives of the polar drug, but we herein report a novel approach by which endogenous
transporters at BBB are exploited for brain drug delivery. The crucial role played by glutathione in
pathogenesis of Parkinson’s and the presence of its influx transporters at the basolateral membrane of BBB
served as the basis for our anti-Parkinson prodrug design strategy. A metabolically stable analogue of
glutathione is used as a carrier for delivery of dopamine and adamantamine. An account of successful syntheses
of these prodrugs along with their transport characteristics and stability determination is discussed.
Introduction
quinone radicals) has been implicated in the progressive
degeneration of dopaminergic neurons, which in turn is one of
the principal causes of Parkinson’s disease. Glutathione, with
the help of enzymes glutathione peroxidase and reductase, forms
detoxification machinery against these oxidative species. We
therefore sought to develop a prodrug strategy targeting glu-
tathione transporters at BBB to deliver anti-Parkinson drugs that
normally can not cross the BBB.
Unfavorable physicochemical properties hinder the transport
of several entities across the blood-brain barrier (BBB)a and
thereby limit their utility as CNS-active agents. The brain
microvessel endothelium is a barrier toward the passive transport
of hydrophilic substances into the brain. Such obstacles to the
CNS penetration of therapeutic agents have in the past been
circumvented by approaches that bank on the synergy of more
than one disciplines; e.g., chemistry-based (lipophilic prodrugs),
biology-based approaches (carrier/receptor mediated transport),
and drug delivery based (intrathecal/interstitial/olfactory deliv-
ery).1 A prodrug approach that encompasses covalent linkage
of the drug to a carrier that can get recognized by a specific
transporter protein and hence get transported has garnered
interest in recent years.
While the classical prodrug approach to improve membrane
permeability of polar drugs uses lipophilic derivatives, increasing
their passive membrane penetration, the targeted prodrug
approach2 exploits transporters that mediate transport of polar
nutrients such as amino acids and peptides. Therefore, targeting
a specific membrane transporter is particularly rewarding when
the drug in question is polar or charged. Several examples of
the application of this promising strategy have been reported
in the literature. The best known example of carrier-mediated
drug delivery is the transport of L-dopa, a precursor of the
neurotransmitter dopamine, in the treatment of Parkinson’s
disease by the large neutral amino acid transporter (LNNA).2,3
Among various membrane transporters, peptide transporters are
attractive targets in prodrug design due to their broad substrate
specificity4 and high capacity.5 The important role played by
the antioxidant glutathione in the pathogenesis of Parkinson’s
disease6 and presence of glutathione transporters at the BBB
caught our attention.7 Oxidative stress (reactive oxygen species
such as the superoxide radical, hydroxide radical, and semi-
Levodopa (L-dopa) remains as the gold standard for the
treatment of Parkinson’s disease. Although L-dopa enters CNS
via the large neutral amino acid transporters at BBB and is
enzymatically cleaved in the brain to release dopamine, a large
percentage of circulating L-dopa does not penetrate the BBB
and, as a consequence, undergoes peripheral decarboxylation
to generate metabolites that cause dopamine-related side effects.8
While this study was in progress, glycoconjugates of dopamine
and L-dopa were explored by Bonina and De Caprariis9 as a
means to increase their BBB permeability by carrier- mediated
transport (GLUT 1). Several peptidyl prodrugs of dopamine and
L-dopa have also been shown to increase the BBB permeability
of L-dopa along with improved bioavailability.5,10 In our
attempts at increasing BBB penetration of dopamine, we focused
on the glutathione uptake transporters that are located on the
luminal side of the BBB. The broad substrate specificity
displayed by these transporters provides ample opportunity for
rational prodrug design.
The design of glutathione transporter targeted prodrug (Figure
1) involved three components: the carrier, glutathione (GSH),
the active drug, and a suitable linker for conjugation of the
carrier with the drug molecule. A significant impediment for
the direct utilization of GSH as a carrier is the presence of
γ-glutamyl transpeptidase (γ-GT), which would cleave the
γ-glutamyl-cysteinyl peptide bond, rendering these prodrugs
inactive. Incorporation of the metabolically stable urea analogue
of glutathione 3, as a carrier molecule, would be of added benefit
due to its known stability against γ-GT cleavage11,12 and
retention of crucial structural features necessary for recognition
by glutathione transporters. In addition, release of such an
antioxidant molecule would most likely contribute in attenuating
the damage by the oxidative stress generated in Parkinson’s
disease.
* To whom correspondence should be addressed. Phone: 612-624-9911.
Fax: 612-624-0139. E-mail: vince001@umn.edu.
†
Current address: Department of Biopharmaceutical Sciences, University
of California, San Francisco, San Francisco, CA 94158. Phone: 415-514-
4363. E-mail: swati.more@ucsf.edu.
a Abbreviations: BBB, blood-brain barrier; LNAA, large neutral amino
acid transporter; GSH, glutathione; MDCK, Madin Darby canine kidney;
γ-GT, γ-glutamyl transpeptidase.
10.1021/jm800239v CCC: $40.75
2008 American Chemical Society
Published on Web 07/24/2008

4582 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
More and Vince
Figure 1. Prodrug design rationale.
Figure 2. Structures of anti-Parkinson’s prodrugs of adamantamine (1) and dopamine (2). Shown in green is the carrier, metabolically stable
glutathione analogue; in blue is the linker, mercaptopyruvic acid, and in purple is the active drug moiety.
Scheme 1. Retrosynthetic Strategy for Prodrugs 1 and 2
Versatility was regarded to be paramount in the structural
design of these conjugates because the linkage of a pharmaco-
logically benign or beneficial carrier to a multitude of CNS-
active drugs was the ultimate goal. The success of any prodrug
design is dependent on the identity of the linker between the
carrier and the drug. Such a linker should fulfill two require-
ments: first, it should be able to release the active drug at the
target site, preferably by a selective enzymatic reaction, and
second, the cleavage products should be nontoxic. Because of
these aspects, mercaptopyruvate, the metabolite of cysteine, was
especially attractive. The carrier glutathione analogue was
attached to the linker via a heterodisulfide linkage due to the
known stability of this linkage in plasma and the abundance of
enzyme disulfide reductase in the brain compared to other body
compartments.13 The amine functional group of dopamine/
adamantamine was attached to the linker by an amide bond,
which is expected to be cleaved by peptidases in the brain with
reasonable stability in the plasma.
Specifically, this account includes our synthetic efforts to
arrive at these anti-Parkinson’s prodrugs 1 and 2 (Figure 2).
The biological evaluation of these prodrugs in terms of their
preliminary BBB permeability prediction and their ability to
convert into an active drug form are also discussed.
Our initial attempts for the synthesis of the mercaptopyru-
vamides 4 by alkylation and/or peptide coupling of commercially
available 3-bromopyruvic acid were thwarted by the highly
Chemistry
Our initial disconnections were centered around the creation
of the heterodisulfide linkage adjacent to the R-ketoester, an
arrangement that presented a considerable synthetic challenge.
Sensitivity of this link to reductive and nucleophilic protocols
led us to plan its creation toward the end of the synthesis
(Scheme 1). This effectively divided our prodrugs into two
fragments: the urea tripeptide 3 and the mercaptolactamides 4.
We recently have described the synthesis of a derivative of urea
tripeptide 3 that bears protecting groups at its N, C, and S
termini.12
reactive R-bromoketone functionality that caused numerous side-
reactions. We were unable to sequentially introduce these groups
through the oxidation of a pyruvamide and subsequent R-bro-
mination because the latter transformation could not be effected
in our hands. A biomimetic route to these molecules though
oxidative deamination of cysteine also was unsuccessful.
Alkylation of 3-bromopyruvamides (obtained via the dimeth-
ylketal of 3-bromopyruvic acid) with thioacetate, although
successful, was unfruitful because the hydrolysis of the thioa-
mide function resulted in equilibration of aldehyde and thioal-

GSH Peptidomimetics as Anti-Parkinson Prodrugs
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4583
Scheme 2. Equilibration of Aldehyde and Thioaldehydea
a Reagents and conditions: (a) LiOH, THF/H2O; (b) NH4OH, CH3OH; (c) hydrazine hydrate; (d) 2N HCl.
Scheme 3. Synthesis of R-Hydroxy-ꢀ-mercaptopropanamide 9a
another thiol-like dithiothreitol provided the solution to this
problem affording the thiol 13 in 96% yield.15
Both the thiols 9 and 13 were then subjected to hetero-
disulfide formation reaction with the thiol of urea tripeptide
14. The retrosynthetic plan in Scheme 1 depicts activation
of tripeptide thiol by a suitable electrophile, followed by
attack of this active intermediate by thiols 9 or 13 to arrive
at the heterodisulfide linkage.
Electrophiles like iodine, mesyl chloride, or NBS led to
significant formation of the undesired homodisulfide of the
tripeptide thiol, which was a result of the greater electrophilicity
of the activated thiol than the electrophile itself. This was not
found to be the case with molecular chlorine because the S-Cl
intermediate is not more electrophilic than chlorine gas and also
because chlorine can be removed completely under reduced
pressure. A conceptually similar but more practical solution was
found in the use of the commercially available methoxycarbonyl
sulfenyl chloride,16 which gave a clean formation of the active
intermediate from tripeptide thiol 14 (Scheme 5).
a Reagents and conditions: (a) RuO2, NaIO4, CCl4, CH3CN, cat. H2O;
(b) 1-aminoadamantane, isobutyl chloroformate NMM, CH2Cl2, 52% over
2 steps; (c) HSBn, nBuLi, BF3 ·OEt2, THF, 56%; (d) HSTrt, nBuLi,
BF3 · OEt2, THF, 69%; (e) AgNO3, pyridine, CH3OH, 54%; (f) Et3SiH,
CF3COOH, CH2Cl2, 89%.
Following the methodology that we described in our previous
report,12 synthesis of tripeptide 14 was accomplished from Boc-
L-asparagine. The active intermediate formed by reaction of thiol
14 with methoxycarbonyl sulfenyl chloride16 was attacked by
thiols 9 and 13 to afford heterodisulfides 15a and 15b. Oxidation
of secondary hydroxyl in 15a and 15b with Dess-Martin
periodinane gave prodrug precursors 16a and 16b. Global
deprotection of adamantamine prodrug precursor 16a using TFA
gave the adamantamine prodrug 1 in 72% yield. Similarly for
dopamine prodrug, precursor 16b was subjected to TBS-
deprotection followed by deprotection of N-Boc and tert-butyl
esters using TFA to afford the dopamine prodrug 2.
dehyde functional groups (Scheme 2), thereby leading to rapid
degradation of the product to dithianes and other aldol-
condensation polymeric products.
Building upon the knowledge garnered from these studies,
we formulated a sequence of disconnections that segregated the
presence of the incompatible R-bromo or a thiol and a keto
functionality across intermediates.
The ꢀ-mercapto-R-hydroxypropionamide derivatives in Scheme
1 could be envisioned as being formed by the opening of the
corresponding epoxyamides with a thiol nucleophile. Synthesis
of such an epoxyamide started with oxidation of (rac)-glycidol
with RuO2/NaIO4 (Scheme 3) to the oxirane carboxylic acid
(5),14 which was coupled to adamantamine, resulting in amide
6. Opening of the epoxide 6 with a thiol nucleophile inadvert-
ently revealed problems with regioselectivity. Use of benzyl
mercaptan as a nucleophile resulted in two inseparable regioi-
somers (7a and 7b) by attack at the R- and ꢀ-carbon of the
carbonyl of the amide, respectively. The electron-withdrawing
nature of the amide seems to override the steric factor, resulting
in the two regioisomers. One way to side-step this problem was
to use a sterically bulky thiol nucleophile to annul the electron-
withdrawing power of the amide. As expected, trityl mercaptan
afforded a single regioisomer 8 in 69% yield. Deprotection of
trityl protection under Lewis or Brønsted acidic conditions
resulted in thiol 9.
Results and Discussion
Biological Evaluation of Anti-Parkinson’s Prodrugs. Our
prodrug design rationale is based on two primary abilities of
the prodrugs (Figure 1): (1) the ability to cross the BBB through
recognition by glutathione transporters and allow their penetra-
tion into the brain, (2) the ability to release the active drug,
once inside the brain. We chose to evaluate the prodrugs for
the aforementioned properties in two systems: one designed to
predict their BBB permeability and inhibit the transport of GSH
to ascertain the utilization of same transport mechanism by both
of them and, second, to examine their convertibility into active
drug form.
These optimized reaction conditions were applied for the
dopamine linker as well (Scheme 4). TBS-protected dopamine
10 was coupled to acid 5, forming amide 11. Regioselective
opening of the epoxide 11 was achieved with trityl mercaptan,
followed by Lewis acid deprotection of the trityl protection.
Neutralization of the silver salt formed after trityl deprotection
with silver nitrate proved tricky due to acid lability of the TBS-
protecting groups. Exchanging the silver salt with an excess of
Directional Transport Studies of Prodrugs 1 and 2
across MDCK Cell Monolayer. For testing the ability of our
prodrugs to cross BBB using a carrier-assisted transport
mechanism, we performed an in vitro directional transport
experiment using the BBB cell model, the Madin Darby canine
kidney (MDCK II) cell line.17 A simple and precise fluorimetric
quantitation method was developed for measurement of the
prodrugs in these experiments. The presence of a primary amino

4584 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Scheme 4. Synthesis of R-Hydroxy-ꢀ-mercaptopropanamide 13a
More and Vince
a Reagents and conditions: (a) TBSCl, imidazole, CH2Cl2, 86%; (b) 5, isobutyl chloroformate, NMM, CH2Cl2, 56%; (c) HSTrt, nBuLi, BF3 ·OEt2, THF,
65%; (d) AgNO3, pyridine, trimethyl ammonium chloride, CH3OH, 24%; (e) AgNO3, pyridine, dithiothreitol, EtOH/CH2Cl2, 96%.
Scheme 5. Synthesis of Prodrugs 1 and 2a
a Reagents and conditions: (a) ClSOOCH3, CH3OH; ( b) 9 or 13, Et3N, CH3OH; (c) Dess-Martin periodinane, CH2Cl2; (d) TFA/CH2Cl2 (1:1); (e) TBAF,
THF.
group in our molecules prompted us to use fluorescamine,18
which forms fluorescent covalent complexes with amines and
allows for quantitation of the concentration of prodrugs as a
function of the fluorescence emitted by the complex. Alhough
this method gave a good estimation of prodrug concentration,
the instability of reagent fluorescamine in its aqueous solution
(in which the directional transport experiment is performed) and
the low sensitivity of this method (high nanomolar range) led
us to search for another similar reagent. o-Phthalaldehyde is
another such reagent that forms fluorescent isoindoles with
primary and secondary amines in presence of an added thiol
reagent (like ꢀ-mercaptoethanol or N-acetylcysteine).19 The
stability of the o-phthalaldehyde in an aqueous solution, high
sensitivity (high picomolar range), and reproducibility of this
method made it an ideal detection method for our directional
flux experiment.
To our knowledge, the MDCK cell line has never been
studied for glutathione transport. It was therefore essential to
first ascertain the existence of functional glutathione transporters
in this cell line. In our first experiment, we studied the transport
of glutathione itself through an MDCK cell monolayer, expect-
ing to display the general trend of the transport from apical
(blood) f basal (brain) > basal (brain) f apical (blood). At a
concentration of 100 µM, we observed that the transport of
glutathione from the apical to the basal side was approximately
three times greater than the transport in the reverse direction
(Data not shown). We also noted a saturation of transporters at
a 1 mM concentration of glutathione.
A similar transport experiment was performed on adaman-
tamine prodrug 1 at 100 µM concentration. In this case also,
we observed that the influx of prodrug 1 into the basal
compartment was much greater than its efflux (Figure 3).
Furthermore, to ensure that the prodrug was not being metabo-
lized as it crossed the MDCK cell monolayer, the integrity of
prodrug 1 was studied by HPLC. The presence of single peak
on the HPLC chromatogram corresponding to the intact prodrug
1 (cross-checked by doping this sample with an authentic sample
of 1 to obtain a single peak as well) verified the identity of the
substance transported (Figure S1, part A, Supporting Informa-
tion).
Directional transport experiment with dopamine prodrug 2
displayed apical f basal transport > basal f apical transport
(Figure 4). The integrity of the substance transported was
verified in a manner similar to 1. (Figure S1, part B, Supporting
Information). These preliminary experiments successfully dem-
onstrated the carrier-mediated transport of our prodrugs (1 and
2) in an in vitro BBB model.

GSH Peptidomimetics as Anti-Parkinson Prodrugs
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4585
Figure 3. Results of adamantamine prodrug 1 transport experiment across MDCK cell monolayer.
Figure 4. Results of dopamine prodrug 2 transport experiment across MDCK cell monolayer.
Inhibition of Glutathione Uptake by Glutathione-Based
Prodrugs. We next examined the ability of the prodrugs to
inhibit glutathione transport across MDCK cell monolayer as a
means of corroborating on the same transport mechanism used
by glutathione as well as prodrugs. For this, uptake of 3H-labeled
glutathione at a 100 nM concentration by MDCK cell monolayer
was determined in the absence and presence of prodrugs (1, 2;
at 1000, 100, and 10 nM). 14C-Mannitol was used as an internal
standard to govern the tightness of the MDCK monolayer
(Figure 5).
intervals. Lack of a chromophore makes detection of adaman-
tamine impossible by HPLC with simple UV detection. For this
reason, only the dopamine prodrug 2 was studied in this
experiment. The progress of the incubations was monitored by
HPLC (UV detector, λ ) 280 nm) analyzing the disappearance
of the prodrug 2 and the formation of dopamine. The half-life
of this prodrug was about 6 h in brain homogenate, while in
blood plasma, it had t1/2 of 3 h, indicating its probable
breakdown by plasma peptidases. Table 1 reports the half-life
(t1/2) of prodrug 2 in the different incubation media.
The decrease in glutathione transport observed in the presence
of added prodrug (1 or 2) when compared to its transport in the
absence of prodrugs was suggestive of the utilization of the same
transport mechanism by both, glutathione and the prodrugs.
Second, the negligible transport of mannitol, the internal
standard, supported the fact that the observed transport of
glutathione was due to utilization of the transporters and not
by mere passive diffusion through leaky junctions in the MDCK
cell monolayer.
In Vitro Stability Studies: Convertibility of Prodrugs
into Active Drugs. To determine the stability of prodrugs in
physiological media and their ability to deliver the active drug
after brain uptake, prodrugs were simultaneously incubated in
buffer (pH 7.0), rat plasma, and rat brain homogenate.9 In this
experiment, the prodrug 2 was incubated with these biological
samples at 37 °C and aliquots were taken at various time
The appearance of a peak corresponding to the active drug
dopamine, balanced by a proportionate decrease in the intensity
of the peak corresponding to the prodrug 2, can be seen in the
HPLC chromatogram (Figure S2, Supporting Information).
Conclusion
For the targeted delivery of anti-Parkinson drugs into the CNS
using the glutathione transport system, the glutathionyl prodrugs
1 and 2 must show affinity for the glutathione transporter at
BBB, be able to release the active drug at the target site, and
possess a good stability balance between the periphery and the
brain. The results of the directional transport of prodrugs and
the inhibition of glutathione uptake by the prodrugs are
suggestive of the utilization of the same transport mechanism
to get inside the brain. Although these prodrugs fulfilled the

4586 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
More and Vince
Figure 5. Inhibition of glutathione uptake across MDCK cell monolayer by prodrugs 1 (left) and 2 (right) [control ) buffer; 3H-GSH ) 3H-
glutathione].
Table 1. Half-life of Prodrug 2
dried over Na2SO4, and evaporated to dryness. The residue obtained
was dissolved in ethyl acetate and passed through a plug of silica
gel with the aid of EtOAc. Combined eluent was evaporated to
dryness and the residue (ESI-LRMS for 15a 796.4 (M + Na)+
and for 15b 1026.5 (M + Na)+) was carried over to the oxidation
step without any further purification or characterization. Crude
yields of ca. 65% were obtained in both of the cases.
t1/2 in h
buffer (pH 7.2)
plasma
brain homogenate
>12
3
51/2-6
To a anhydrous CH2Cl2 (5 mL) solution of alcohol 15a or 15b
(0.30 mmol) at room temperature under argon, Dess-Martin
periodinane reagent (0.45 mmol) was added all at once. The reaction
mixture was allowed to stir for an hour (indicating completion of
reaction by TLC), after which the reaction mixture was quenched
with 2-propanol and evaporated to dryness to get a white solid.
The solid residue was partitioned between EtOAc (20 mL) and 10
mL of saturated NaHCO3 solution containing 5% sodium thiosul-
fate. The EtOAc layer was separated and washed with brine, dried
over Na2SO4, and evaporated to obtain an oil that was purified by
silica gel column chromatography.
targeted-prodrug criteria with a sustained release of the active
drug in brain extract, exhibition of labilily in plasma is an issue
that needs to be addressed in the further development of this
chemical strategy.
To conclude, these studies resulted in an efficient practical
synthetic route toward new glutathione-based prodrugs. The
chemistry involved in the synthesis of each component of these
prodrugs (carrier, active drug, and linker) circumvents many of
the potential instability problems associated with the intermedi-
ates leading to these biologically important building blocks. The
biological data indeed support our hypotheses and pioneers the
utilization of metabolically stable glutathione as a carrier for
drug targeting into the brain.
Boc-γ-Gla{-Cys[3-(N-adamantylpyruvamide)thio]-Gly-OtBu}-
OtBu (16a). Yield 70%, clear oil, hexanes/EtOAc 1:1 then 1:4, Rf
1
0.60 (EtOAc/hexanes, 3:2); [R]D -6.4 (c 0.94, CHCl3). H NMR
(300 MHz, CDCl3) δ 7.11, 6.77, 5.98, 5.50, 5.35 (5s, 5H, NH),
4.60 (m, 1H, R-CH:Gla), 4.17 (m, 1H, R-CH:Cys), 3.90-3.78 (m,
4H,SCH2C(dO),CH2:Gly),3.62-3.35(m,3H,ꢀ-CH2:Gla,ꢀ-CHAHB:
Cys), 2.97 (m, 1H, ꢀ-CHAHB:Cys), 2.08-1.62 (m, 15H, CH2:Adm,
CH:Adm), 1.43, 1.41, 1.36 (3s, 27H, C(CH3)3). 13C NMR (75 MHz,
CDCl3) δ 193.5, 171.1, 170.1, 168.7, 158.9, 158.1, 156.0 (C(dO)),
82.7, 82.4, 82.3 (C(CH3)3), 60.6 (R-C:Gla), 53.1 (R-C:Cys),
43.1-28.2 (ꢀ-C:Gla, CH2:Gly, SCH2C(dO), ꢀ-C:Cys, C(CH3)3,
C:Adm). ESI-HRMS m/z 794.3479 (M + Na)+; C35H57N5O10S2 +
Na+ requires 794.3439.
Experimental Section
General Procedures. All commercial chemicals were used as
supplied unless otherwise indicated. Dry solvents (THF, Et2O,
CH2Cl2, and DMF) were dispensed under argon from an anhydrous
solvent system with two columns packed with neutral alumina or
molecular sieves. Flash chromatography was performed with
Silica-P flash silica gel (silicycle, 230-400 mesh) with indicated
mobile phase. All reactions were performed under inert atmosphere
of ultrapure argon with oven-dried glassware. H and 13C NMR
1
Boc-γ-Gla{-Cys[3-(N-(O,O′-di(tert-butyl-dimethylsilyl)dopami-
nylpyruvamide)-thio]-Gly-OtBu}-OtBu (16b). Yield 73%, clear oil,
hexanes/EtOAc 1:1 then 1:3, Rf 0.45 (EtOAc/hexanes, 3:2); [R]D
-17.3 (c 1.33, CHCl3). 1H NMR (300 MHz, CDCl3) δ 7.18, 7.03
(2s, 2H, NH), 6.57-6.44 (m, 3H, Ar), 5.86, 5.44, 5.29 (3s, 3H,
NH), 4.48 (q, J ) 3.0 Hz, 6.3 Hz, 1H, R-CH:Gla), 4.04 (m, 1H,
R-CH:Cys), 3.78-3.73 (m, 4H, SCH2C(dO), CH2:Gly), 3.46-3.13
(m, 5H, ꢀ-CH2:Gla, ꢀ-CHAHB:Cys, CH2CH2NH), 2.85-2.82 (m,
1H, ꢀ-CHAHB:Cys), 2.56 (t, J ) 3.6 Hz, 2H, CH2CH2NH), 1.28,
1.27, 1.24 (3s, 27H, C(CH3)3), 0.79 (s, 18H, SiC(CH3)3), 0.00 (s,
12H, SiCH3). 13C NMR (75 MHz, CDCl3) δ 196.1, 175.1, 174.1,
172.7, 163.9, 161.9, 159.9 (C(dO)), 150.9, 149.7, 135.1, 125.5,
125.2 (CAr), 86.6, 86.3, 84.1 (C(CH3)3), 64.4 (R-C:Gla), 57.1 (R-
C:Cys), 46.6, 46.2, 45.2, 44.9, 38.6, 36.7 (ꢀ-C:Gla, CH2:Gly, ꢀ-C:
Cys, SCH2C(dO), CH2CH2NH, CH2CH2NH), 33.7, 32.4, 32.1,
32.0, 30.0 (C(CH3)3), 22.5, 22.4 (SiC(CH3)3), 0.0 (SiCH3). ESI-
spectra were recorded on a Varian 300 MHz spectrometer. High-
resolution mass data were acquired on a Bruker Bio TOF-II
spectrometer capable of positive and negative ion ESI source, using
PPG or PEG as internal standards.
General Procedure for Heterodisulfide Bond Formation. A
methanolic solution of tripeptide 14 (0.35 mmol) was added
dropwise to a solution of Cl-SCOOCH3 (0.42 mmol) in MeOH (3
mL) at 0 °C over 20 min. The reaction mixture was allowed to stir
at 0 °C for another 1 h and allowed to warm to room temperature.
A CH3OH solution of thiol 9 or 13 (1.05 mmol) was added to the
above solution at once, followed by 0.70 mmol of Et3N. The
reaction mixture was stirred under argon at room temperature
overnight, and the solvent was evaporated to dryness to obtain a
sticky residue that was taken in EtOAc (30 mL). Organic layer
was washed with 10% citric acid solution (20 mL), saturated
NaHCO3 solution (20 mL), and brine. EtOAc layer was separated,

GSH Peptidomimetics as Anti-Parkinson Prodrugs
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4587
HRMS m/z 1024.4648 (M + Na)+; C45H79N5O12S2Si2 + Na+
requires 1024.4597.
transport experiment, the upper compartment of the Transwell was
designated as the apical (A) and the lower compartment designated
was the basal (B) side.
TFA.H2N-γ-Gla{-Cys[3-(N-adamantylpyruvamide)thio]-Gly-
OH}-OH (1). The protected tripeptide 16a (80 mg, 0.10 mmol) was
treated with a mixture TFA and CH2Cl2 (1:1 v/v, 6 mL) at room
temperature for 4 h. Concentration of the reaction mixture followed
by trituration with freshly distilled diethyl ether afforded a white
solid. Recrystallization of the solid from a aqueous ethanol gave
analytically pure 1 (53 mg, 77%). Rf 0.45 (butanol/acetic acid/H2O,
12:5:3); [R]D -5.7 (c 0.65, 1N HCl). 1H NMR (300 MHz, CDCl3/
CF3COOD) δ 4.53 (m, 1H, R-CH:Cys), 4.32 (m, 1H, R-CH:Gla),
4.16-3.82 (m, 4H, SCH2C(dO), CH2:Gly), 3.52-3.41 (m, 2H,
ꢀ-CH2:Gla), 3.21-2.98 (m, 2H, ꢀ-CH2:Cys), 2.21-1.70 (m, 15H,
CH2:Adm, CH:Adm). 13C NMR (75 MHz, CDCl3/CF3COOD) δ
195.1, 174.3, 172.6, 169.8, 159.5, 158.4 (C()O)), 53.4 (R-C:Cys),
51.4 (R-C:Gla), 44.6 (ꢀ-C:Gla), 42.8 (CH2:Gly), 41.2-29.3
(SCH2C(dO), ꢀ-C:Cys, C:Adm). ESI-HRMS m/z 560.1871 (M +
H)+; C22H33N5O8S2 + H+ requires 560.1843. Reverse phase HPLC
was run on Varian Microsorb column (C18, 5 µm, 4.6 mm × 250
mm) using two solvent systems with 0.5 mL/min flow rate and
detected at 220/254 nm. Solvent system 1: 0.04 M TEAB
(triethylammonium bicarbonate) in water/70% acetonitrile in water
) 1/1, tR ) 5.39 min, purity ) 98.57%. Solvent system 2: 0.04 M
TEAB in water/70% acetonitrile in water ) 20 - 100% B linear,
tR ) 6.21 min, purity ) 96.34%.
TFA·H2N-γ-Gla{-Cys[3-(N-(O,O′-di(tert-butyl-dimethylsilyl)-
dopaminylpyruvamide)thio]-Gly-OH}-OH (2). To a solution of 16b
(95 mg, 0.09 mmol) in THF (5 mL) was added commercial TBAF
(1 M solution in THF, 0.18 mL, 0.18 mmol) at 0 °C. After the
mixture was stirred at room temperature for 30 min, the reaction
was quenched by adding H2O. The resulting mixture was diluted
with EtOAc. The layers were separated; the organic layer was
washed with brine, dried (Na2SO4), and evaporated in vacuo. The
residue was purified by silica gel flash chromatography to obtain
the TBS-deprotected precursor tripeptide (64 mg, 87% yield) that
was used directly in the subsequent reaction. The TBS-deprotected
tripeptide obtained above (60 mg, 0.07 mmol) was dissolved in a
mixture of TFA and CH2Cl2 (1:1 v/v, 5 mL) at room temperature
and stirred for 4 h. The reaction mixture was then concentrated
and the residue was triturated with ether and EtOAc to get product
as fluffy solid. Recrystallization of this solid from a water/ethanol
mixture gave analytically pure 2 (53 mg, 77% yield). Rf 0.3
(butanol/acetic acid/H2O, 12:7:3); [R]D -10.3 (c 0.36, 1N HCl).
1H NMR (300 MHz, CDCl3/CF3COOD) δ 6.76-6.54 (m, 3H, Ar),
4.57 (m, 1H, R-CH:Cys), 4.24 (m, 1H, R-CH:Gla), 3.98-3.81 (m,
4H, SCH2C(dO), CH2:Gly), 3.52-3.34 (m, 4H, ꢀ-CH2:Gla,
CH2CH2NH), 3.12-2.89 (m, 2H, ꢀ-CH2:Cys), 2.69 (m, 2H,
CH2CH2NH). 13C NMR (75 MHz, CDCl3/CF3COOD) δ 195.4,
174.6, 172.4, 170.7, 160.5, 158.2 (C(dO)), 147.2, 146.8, 133.5,
123.5, 121.8 (CAr), 54.4 (R-C:Cys), 49.9 (R-C:Gla), 45.9 (ꢀ-C:Gla),
42.4 (CH2:Gly), 41.6, 39.3, 37.7, 36.9 (ꢀ-C:Cys, SCH2C(dO),
CH2CH2NH, CH2CH2NH).ESI-HRMS m/z 562.1298 (M + H)+;
C20H27N5O10S2 + H+ requires 562.1272. Reverse phase HPLC was
run on Varian Microsorb column (C18, 5 µm, 4.6 mm × 250 mm)
using two diverse solvent systems with 0.5 mL/min flow rate and
detected at 220/254 nm. Solvent system 1: 0.04 M TEAB
(triethylammonium bicarbonate) in water/70% acetonitrile in water
) 2/3, tR ) 7.06 min, purity ) 95.43%. Solvent system 2: 0.04 M
TEAB in water/70% acetonitrile in water ) 20-100% B linear, tR
) 7.53 min, purity ) 98.43%.
For the experiment, the medium was aspirated and the cells were
washed and preincubated with the assay buffer for 30 min. The
assay buffer was then replaced with assay buffer containing
glutathione/prodrug 1 or 2 in the donor side, i.e., the apical side
(1.5 mL), to determine flux from the apical to the basal side or the
basal side (2.6 mL) to determine flux from the basal to the apical
side. Drug-free fresh assay buffer was placed on the receiver side.
The experimental wells were incubated on an orbital shaker at 37
°C for the duration of the experiment (180 min) except while
drawing samples and replacing the assay buffer. Then 100 µL
samples were drawn from the receiver side at 0, 30, 60, 90, 120,
150, and 180 min and replaced with drug free fresh assay buffer.
Similarly, 100 µL samples were drawn from the donor side at 0
and 180 min and replaced with the assay buffer containing
glutathione.
All the aliquots were analyzed by fluorimetric assay based on
using o-phthalaldehyde (OPA). OPA reagent solution (complete)
was made by addition of 20 µL of ꢀ-mercaptoethanol to 5 mL of
OPA reagent solution (incomplete) immediately prior to its use.
Then 15 µL of each aliquot was mixed with 150 µL of OPA reagent
solution (complete) in a Costar 96-well black opaque plate. Samples
were allowed to incubate for 2 min with moderate shaking at room
temperature and the fluorescence intensity was measured using a
SynergyHT fluorescence plate reader with a 360 nm, 40 nm
bandwidth, excitation filter, and a 460 nm, 40 nm bandwidth,
emission filter. The sensitivity setting was at 45, and the data was
collected from the top. The percentage transport across the
monolayer was determined by dividing the fluorescence reading at
various time intervals with the fluorescence measured at the start
of the experiment. The integrity of glutathione/prodrug 1 or 2
transported across MDCK cell monolayer was verified by HPLC.
Assay buffer: NaCl (7.13 g), NaHCO3 (2.10 g), glucose (1.8 g),
HEPES (2.38 g), KCl (0.224 g), MgSO4 (0.295 g), CaCl2 (0.206
g), K2HPO4 (0.070 g) in 1000 mL H2O adjusted to pH 7.4.
HPLC system: Varian ProStar 210; UV detection: 220/254 nm;
column: Varian Microsorb C18, 5 µm, 4.6 mm × 150 mm; solvent
system: linear gradient 20-100% solvent B [solvent A ) 0.04 M
TEAB (triethylammonium bicarbonate) in water; solvent B ) 70%
acetonitrile in water]; flow rate: 0.5 mL/min; retention time (tR):
prodrug 1 6.20 min; prodrug 2 7.52 min.
In Vitro Blood and Brain Stability Studies. Plasma. Blood was
drawn from a rat though a heart puncture after anesthetization. Ten
µL of 0.5 M EDTA was added per mL of blood. Samples were
centrifuged at 14000 RPM for 15 min to separate plasma, placed
on ice, and used immediately.
Brain Homogenate. The brain was removed and homogenized
in cold 0.1 M phosphate buffer of pH 7.2 with a Dounce tissue
homogenizer (1:5 w/v). Samples were then placed on ice and used
immediately.
Procedure for the Stability Studies. 100 µL of 20 mM compound
in phosphate buffer was added to 2 mL of buffer (phosphate buffer,
pH 7.2), plasma, or brain homogenate and gently vortexed. Samples
were incubated at 37 °C and 200 µL aliquots were removed every
half hour. Then 200 µL of acetonitrile was added to each aliquot
and vortexed. Samples were centrifuged for 15 min to remove
proteins and the supernatants were analyzed by HPLC.
HPLC System: Waters Delta 600; UV detection: 220/254 nm;
column: Atlantis DC18, 5 µm, 4.6 mm × 150 mm; solvent system:
linear gradient 5-50% solvent B [solvent A ) 0.1% heptafluo-
robutyric acid (HFBA) in water; solvent B ) acetonitrile]; flow
rate: 1.0 mL/min; retention time (tR): dopamine 11.2 min; prodrug
2 12.2 min.
In Vitro Directional Transport Experiment Using MDCK
Cell Monolayer. The influence of the blood brain barrier (BBB)
on permeabilities of prodrugs was determined by MDCK cell
monolayers growing on a permeable support. MDCK cells
(Madin-Darby canine kidney cells) were seeded into 6-well
Transwell inserts in the presence of a suitable growth medium
(DMEM supplemented with 10% fetal bovine serum, penicillin (100
U/mL), and streptomycin (100 µg/mL)) at 3.0 × 105 cells per well
and placed in the CO2 TC incubator to grow for 3 days. The cells
were grown until they reached confluency, which was evaluated
by visual observation under a microscope. In our directional
Inhibition of Glutathione Uptake by Prodrugs 1 and 2. Proce-
dure same as in the directional transport experiment. Here, the
transport of [Glycine-2-3H]-glutathione from the apical to the basal
direction at 100 nM concentration was challenged with 0, 10, 100,
and 1000 nM concentrations of prodrugs 1 or 2. In each well, D-
[1-14C]-mannitol (4 nM) was added as an internal standard. Then

4588 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
More and Vince
Glutathione Across the Blood-Brain Barrier in the Rat. J. Clin. InVest.
1990, 85, 2009–2013. (b) Kannan, R.; Yi, J. R.; Tang, D.; Li, Y.;
Zlokovic, B. V.; Kaplowitz, N. Evidence for the Existence of a
Sodium-Dependent Glutathione (GSH) Transporter. J. Biol. Chem.
1996, 271, 9754–9758.
100 µL aliquots were taken at 15 min intervals for the first hour
and every 30 min thereafter up to 3 h from the basal (bottom)
chamber and replaced with assay buffer to keep the solution volume
constant. The aliquots taken during the experiment were combined
with Ecolite scintillation cocktail and the radioactivity was measured
with Beckman Coulter LS 6500 scintillation counter. The percentage
transport of glutathione across the monolayer was determined by
dividing the radioactivity reading at various time intervals with the
radioactivity measured at the start of the experiment.
(8) Wang, H.; Lee, J.; Tsai, M.; Lu, H.; Hsu, W. Synthesis and
Pharmacological Activities of a Novel Tripeptide Mimetic Dopamine
Prodrug. Bioorg. Med. Chem. Lett. 1995, 5, 2195–2198.
(9) Bonina, F.; Puglia, C.; Rimoli, M. G.; Melisi, D.; Boatto, G.; Nieddu,
M.; Calignano, A.; La Rana, G.; De Caprariis, P. Glycosyl Derivatives
of Dopamine and L-dopa as Anti-Parkinson Prodrugs: Synthesis,
Pharmacological Activity and In Vitro Stability Studies. J. Drug.
Targeting 2003, 11, 25–36.
(10) Bai, J. P. F. pGlu-L-Dopa-Pro: A Tripeptide Prodrug Targeting the
Intestinal Peptide Transporter for Absorption and Tissue Enzymes for
Conversion. Pharm. Res. 1995, 12, 1101–1104.
(11) Vince, R.; Brownell, J.; Akella, L. B. Synthesis and Activity of γ-(L-
γ-azaglutamyl)-S-(p-bromobenzyl)-L-cysteinylglycine: A Metabolically
Stable Inhibitor of Glyoxalase I. Bioorg. Med. Chem. Lett. 1999, 9,
853–856.
Acknowledgment. This work was supported by a grant from
the Center for Drug Design, Academic Health Center, University
of Minnesota. We acknowledge Prof. William Elmquist (De-
partment of Pharmaceutics, University of Minnesota) and his
students Mr. Naveed Shaik and Nagdeep Giri for their valuable
discussions and help in performing the transport experiments.
We thank Ms. Christine Dreis in our laboratory for her help
with the biological assays and Dr. Christine Solomon for help
in procuring rat blood and brain samples.
(12) More, S. S.; Vince, R. A Metabolically Stable Tight-binding Transition-
State Inhibitor of Glyoxalase-I. Bioorg. Med. Chem. Lett. 2006, 16,
6039–6042.
(13) Bickel, U.; Kang, Y. S.; Pardridge, W. M. In Vivo Cleavability of a
Disulfide-Based Chimeric Opioid Peptide in Rat Brain. Bioconjugate
Chem. 1995, 6, 211–218.
(14) Hughes, P.; Clardy, J. Total synthesis of 3(S)-carboxy-4(S)-hydroxy-
2,3,4,5-tetrahydropyridazine, an unusual amino acid constituent of
luzopeptin A. J. Org. Chem. 1989, 54, 3260–3264.
Supporting Information Available: Experimental details for
the synthesis of 5-14 and 18-26, in vitro BBB permeability assay
and in vitro stability assay of prodrugs. This material is available
free of charge via the Internet at http://pubs.acs.org.
(15) Lipta´k, A.; Sajtos, F.; Ja´nossy, L.; Gehle, D.; Szila´gyi, L. A General
Method for the Synthesis of Sugar 2-C-Sulfonic Acids by 1 f 2
Arylthio Group Migration in Acid-Sensitive Thioglycosides. Direct
Transformation of Thiotrityl Ethers into C-Sulfonic Acids. Org. Lett.
2003, 5, 3671–3674.
(16) Rietman, B. H.; Peters, R. F. R.; Tesser, G. I. A Facile Method for
the Preparation of S-(Alkylsulfenyl)cysteines. Synth. Commun. 1994,
24, 1323–1332.
(17) Gumbleton, M.; Audus, K. L. Progress and Limitations in the Use of
In Vitro Cultures to Serve as a Permeability Screen for the Blood-
Brain Barrier. J. Pharm. Sci. 2001, 90, 1681–1698.
(18) Udenfriend, S.; Stein, S.; Bo¨hlen, P.; Dairman, W.; Leimgruber, W.;
Weigele, M. Fluorescamine: A Reagent for Assay of Amino Acids,
Peptides, Proteins, and Primary Amines in the Picomole Range.
Science 1972, 178, 871–872.
References
(1) Pardridge, W. M. Blood-brain Barrier Drug Targeting: The Future
of Brain Drug Development. Mol. InterVentions 2003, 3, 90–105.
(2) Han, H. K.; Amidon, G. L. Targeted Prodrug Design to Optimize Drug
Delivery. AAPS PharmSci. 2000, 2, 1–11.
(3) Bonina, F.; Lanza, M.; Montenegro, L.; Salerno, L.; Smeriglio, P.;
Trombetta, D.; Saija, A. Transport of Alpha-Tocopherol and Its
Derivatives Through Erythrocyte Membranes. Pharm. Res. 1996, 13,
1343–1347.
(4) Fujita, T.; Kishida, T.; Wada, M.; Okada, N.; Yamamoto, A.; Leibach,
F. H.; Ganapathy, V. Functional characterization of brain peptide
transporter in rat cerebral cortex: identification of the high-affinity type
H+/peptide transporter PEPT2. Brain Res. 2004, 997, 52–61.
(5) Daniel, H.; Rubio-Aliaga, I. An Update on Renal Peptide Transporters.
Am. J. Physiol. Renal Physiol. 2003, 284, F885–F892.
(6) Lotharius, J.; Brundlin, P. Pathogenesis of Parkinson’s Disease:
Dopamine, Vesicles and R-Synuclein. Nat. ReV. Neurosci. 2002, 3,
932–942.
(19) Benson, J. R.; Hare, P. E. o-Phthalaldehyde: Fluorogenic Detection
of Primary Amines in the Picomole Range. Comparison with Fluo-
rescamine and Ninhydrin. Proc. Natl. Acad. Sci. U.S.A. 1975, 72,
619–622.
(7) (a) Kannan, R.; Kuhlenkamp, J. F.; Jeandidier, E.; Trinh, H.; Oakhtens,
M.; Kaplowitz, N. Evidence for Carrier-mediated Transport of
JM800239V