γ-Glutamyl transpeptidase acylation with peptidic substrates: Free energy relationships measured by an HPLC kinetic assay

Article abstract of DOI:10.1039/b606914b

γ-Glutamyl transpeptidase (GGT, EC 2.3.2.2) is a highly glycosylated heterodimeric enzyme linked to the external cellular membrane that catalyzes the hydrolysis of glutathione as well as the transfer of its γ-glutamyl group to amino acids and dipeptides i

Full text of DOI:10.1039/b606914b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
c-Glutamyl transpeptidase acylation with peptidic substrates: free energy
relationships measured by an HPLC kinetic assay†
Myle`ne Morin, Caroline Rivard and Jeffrey W. Keillor*
Received 17th May 2006, Accepted 17th August 2006
First published as an Advance Article on the web 8th September 2006
DOI: 10.1039/b606914b
c-Glutamyl transpeptidase (GGT, EC 2.3.2.2) is a highly glycosylated heterodimeric enzyme linked to
the external cellular membrane that catalyzes the hydrolysis of glutathione as well as the transfer of its
c-glutamyl group to amino acids and dipeptides in a transpeptidation reaction. The measurement of
both the hydrolytic and transpeptidation activity of this important enzyme is a challenge, since its
native substrates are not highly chromogenic. We have developed an HPLC-based method for the
quantitative photometric detection of numerous enzyme substrates and products, after their
pre-column derivation with dabsyl chloride. The broad applicability of this method was demonstrated
in the kinetic investigation of transpeptidation reactions of rat kidney GGT with glutathione, its native
substrate, as well as a series of pertinent glutathione analogues. The pH-rate profile constructed for
glutathione confirmed the dependence on the ionisation state of at least two residues. Analysis of the
free-energy relationships in the series of synthetic peptidic substrate analogues revealed the importance
of enzyme–substrate interactions unrelated to amine leaving group basicity during the acylation step.
These results are further interpreted in the context of the recently published structure for a similar GGT.
In the past, several mechanistic studies have been carried out
Introduction
with GGT,1,11,15–18 but much remains to be learned concerning the
c-Glutamyl transpeptidase (GGT, EC 2.3.2.2) is a highly glycosy-
catalytic mechanism and the specific amino acid residues involved.
lated heterodimeric enzyme linked to the external surface of cells
A better comprehension of this mechanism will be important for
through a membrane anchor found in its large ∼46 kDa subunit,
the development of more efficient inhibitors for the treatment of
certain disorders.2,5–8 Until recently, the mechanism of GGT was
studied primarily using chromogenic substrate analogues such
as c-glutamyl-p-nitroanilide, from which the intensely yellow-
that is bound non-covalently to its small ∼22 kDa subunit.1
GGT plays roles in several biological processes, including the
biosynthesis of leukotrienes D4,2 cellular detoxification through
the formation of mercapturic acids,3 homeostasis of glutathione
(GSH), its in vivo substrate, and possibly the transport of amino
acids across the cellular membrane via the c-glutamyl cycle.4 This
enzyme has also been implicated in many physiological disorders
such as Parkinson’s disease,5 inhibition of apoptosis,6 diabetes7
and asthma.2,8 It is particularly concentrated in the kidney, brain
and pancreas.1
coloured p-nitroaniline is liberated as a reaction product.1,13,15,19
We have also studied other para-substituted c-glutamyl anilide
substrate analogues through the colorimetric determination of
released aniline by a discontinuous diazotization assay.15 These
studies have led to a greater understanding of the mechanism of
GGT acylation, but the relevance of these studies with respect to
the mechanism of the native enzyme reaction may be questioned,
considering the structural differences between a c-glutamyl anilide
and glutathione, the native substrate. However, the study of more
structurally similar, peptidic substrate analogues has been limited
by the lack of a convenient kinetic method for following their
reaction with GGT. A high-performance liquid chromatography
(HPLC)-based method has been reported20 for the specific detec-
tion of glutamate deriving from the hydrolase activity of GGT,
but this method is not applicable to the study of physiologically
relevant GGT-mediated transpeptidation reactions. In this paper,
we report the development of a complementary HPLC-based
method for the quantitative photometric detection of numerous
primary amine products of the acylation step, after their pre-
column derivation with (dimethylamino)azobenzenesulfonyl chlo-
ride (dabsyl chloride, DABS-Cl).
GGT catalyzes three different types of reactions via a modified
ping-pong mechanism.9,10 The first reaction, transpeptidation,
comprises the transient transfer of a c-glutamyl moiety from a
donor substrate, GSH, to form an intermediate acyl-enzyme, fol-
lowed by acyl transfer to an acceptor substrate—either an amino
acid or a dipeptide—to yield a c-glutamyl peptide derivative.1,11–13
At higher concentrations, GSH may react as both donor and
acceptor substrates, leading to autotranspeptidation, the second
type of catalytic reaction.9 The third catalytic reaction is hydrolysis,
resulting in the formation of glutamic acid when a water molecule
acts as an acceptor substrate.9,10,14
De´partement de chimie, Universite´ de Montre´al, C.P. 6128, Succ. Centre-
ville, Montre´al, QC H3C 3J7, Canada. E-mail: jw.keillor@umontreal. ca
† Electronic supplementary information (ESI) available: Chromato-
graphic gradient profiles, reversed-phase HPLC chromatogram of
the dabsyl chloride derivatives, standard curve for the concentra-
tion of L-c-glutamylglycylglycine and pH–rate kinetic data. See DOI:
10.1039/b606914b
This method permitted the study of rat kidney GGT-mediated
transpeptidation reactions from peptidic GSH substrate analogues
whose products cannot be detected simply and directly by
spectrophotometry (Scheme 1). A series of such analogues was
3790 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Scheme 1 Simplified mechanism of transpeptidation reactions mediated by GGT.
synthesized and used to investigate the free-energy relationships
and pH-rate effects of the acylation step of the catalytic transpep-
tidation process. Furthermore, the recent publication21 of the first
high-resolution structure of GGT allowed, for the first time, the
interpretation of the mechanistic consequences of the results of
our studies in the context of a model for the enzyme structure.
were prepared by slightly different synthetic routes. During the
preparation of L-c-glutamyl-2-(ethylthio)ethyl amide (16), benzyl
ester hydrolysis preceded removal of the Boc group (Scheme 3).
The synthesis of L-c-GluAlaGly (20) was effected according to a
simple peptide coupling with L-AlaGly according to a protocol
previously developed in the group.22 This synthesis, shown in
Scheme 4, involves the in situ preparation of a p-nitrophenyl ester
intermediate prior to peptide coupling, prior to hydrogenolysis of
the benzyl ester and removal of the Boc protecting group.
Results and discussion
Substrate analogue synthesis
HPLC analysis
All of the substrate analogues tested in the course of this study
were prepared in excellent yields by standard peptide coupling
methods. Most of the substrates were prepared according to
Scheme 2, involving TBTU coupling of various amines with the
c-carboxylate group of the a-benzyl ester of N-Boc-L-glutamic
acid. The details of this coupling are provided as the general
method A. This coupling was followed by removal of the Boc
group with TFA (general method B) and removal of the benzyl
ester by hydrogenolysis (general method C). Two substrates
The amino acids and peptides found in the enzymatic reaction
mixtures of these substrate analogues do not possess a strong
chromophore, so a common derivatization method was used to
permit their facile detection. Dabsyl chloride derivation of amino
acids is well-known23 and was optimized herein, with respect
to the pH of the amino acid solution and the solvent of the
dabsyl chloride stock solution used. Briefly, the optimal pH for
the derivatization reaction was determined to be 9.0.24,25 At lower
Scheme 2 Synthesis of the majority of c-glutamyl amides studied herein.
The Royal Society of Chemistry 2006 Org. Biomol. Chem., 2006, 4, 3790–3801 | 3791
This journal is
©

View Article Online
Scheme 3 Synthesis of L-c-glutamyl-2-(ethylthio)ethyl amide.
Scheme 4 Synthesis of L-c-GluAlaGly.
3792 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
pH, primary amine groups are predominantly protonated and
slow to react, whereas at higher pH, background hydrolysis of
dabsyl chloride becomes problematic. Concerning the co-solvent,
although dabsyl derivatizations have been carried out previously in
acetonitrile, we found that the greater solubility of dabsyl chloride
in acetone allows for higher concentrations to be used, resulting
in more rapid derivatization.
Subsequent separation of derivatized substrate analogues and
enzyme reaction products was accomplished by reversed-phase
HPLC. The elution gradients reported in the ESI† generally
permitted baseline resolution of the amino acid and peptide
derivatives. Making reference to an internal standard, the quan-
titative analysis of these reaction products from these HPLC
chromatograms was found to be linear and accurate over the
concentration range of 10–1000 lM in the enzymatic reaction
mixture.26 Although many amino acids were found to react rapidly
and reproducibly with DABS-Cl, glutamine was chosen as an
internal standard since it elutes from the HPLC column with
a retention time similar to, but distinct from the other reaction
components assayed in this application, providing a means for
the reliable determination of its concentration, normalizing for
variance in injection volume.27
In a typical application of this discontinuous assay to the kinetic
study of GGT-mediated reactions, enzymatic reaction mixtures
were prepared on such a scale as to provide several aliquots that
were removed at different times throughout the initial rate time
course. These aliquots were immediately quenched with TCA,
stopping the enzymatic reaction through precipitation of the
enzyme. Once a series of samples were accumulated, they were
analysed by HPLC, typically overnight, using an autosampler.
Admittedly, discontinuous assay methods are often tedious and
less convenient for application to kinetic analysis; however, the
availability of an autosampler allowed the injections and subse-
quent chromatographic separations to be completely automated,
rendering this method almost as practical as a continuous reaction
assay. In this way, kinetic data representing linear initial rates were
obtained rapidly for a variety of substrate analogue L-c-glutamyl
amides (see ESI†).
Fig. 1 Michaelis–Menten plot of the rate of GGT-mediated transpep-
tidation reaction between L-c-glutamylglycinamide and 20 mM GlyGly
at pH 8.0 and 37 ◦C. The standard deviations reported for the kinetic
parameters derive from the non-linear regression for the hyperbolic line
through data obtained from single kinetic runs.
curve thus obtained for GSH (Fig. 2) gives macroscopic kinetic
pKa values of 7.88 and 8.37, similar to those obtained previously by
another method,15 for ionizations of the E·S complex. This curve
illustrates the dependence of the catalytic mechanism for acylation
from the native donor substrate on the ionization states of (at least)
two residues, one of which must be in its basic form and another
that is active in its acidic form. The lower kinetic pKa value has
previously been attributed9,28 to a histidine imidazole. Subsequent
mutagenesis studies showed that replacement of the two histidine
residues of the small subunit of human GGT dramatically reduced
the efficiency of the mutant enzyme to 2% residual activity.28
This is certainly suggestive of the involvement of these residues
in a general acid–base mechanism15,16 but does not exclude the
possibility that other ionizable groups may be involved.29 For
example, other workers have suggested that the pH dependence of
the acidic limb reflects the pKa of GlyGly, the acceptor substrate
pH–Rate profile for rate-limiting acylation of GGT by GSH
◦
Kinetic studies were thus carried out at pH 8.0 and 37 C in the
presence of varied concentrations of L-c-glutamyl donor substrates
and a high concentration (20 mM, >6-fold KM) of GlyGly, an ex-
cellent acceptor substrate for GGT-mediated transpeptidation.9,13
This high concentration of GlyGly also assures that the deacyla-
tion step of the acyl-transfer mechanism is rapid and the preceding
acylation step is rate-limiting and manifested as kcat. Under these
conditions, the slower hydrolysis reaction and the adventitious
autotranspeptidation reaction of the c-glutamyl substrate are
apparently minimized.1,10,15,18,19 Initial rates measured this way
were plotted against the substrate concentration at which they
were obtained, resulting in Michaelis–Menten plots of which a
representative example is shown in Fig. 1.
The generality of the HPLC method also permitted the kinetic
evaluation of GSH, the non-chromogenic native donor substrate
of GGT. A pH–rate plot was constructed using GSH as donor
substrate and GlyGly as acceptor substrate, according to the
equation described in the Experimental section. The bell-shaped
Fig. 2 pH–Rate profile for the GGT-mediated reaction of GSH with
GlyGly at 37 ◦C.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3790–3801 | 3793
©

View Article Online
common to these studies, suggesting that the deprotonated form
of GlyGly (pKa 8.13) is the true acceptor.15,30 In either case, the
present work confirms the kinetic pKa values measured for the
GGT-mediated reaction of GSH with GlyGly. The general HPLC
method used herein could be applied to the study of a series of
acceptor substrates to test this hypothesis concerning the lower
kinetic pKa value. For example, at a relatively high concentration
of GSH (4 mM), the value of kcat for L-AlaGly as acceptor substrate
was determined to be 250 30 s−1(KM = 1.2 0.3 mM). However,
the work described herein focuses primarily on GGT acylation
using donor substrate analogues.
Brønsted plots for substrate analogues
The flexibility of the method described herein permitted further
investigation of the GGT-mediated transpeptidation reactions
between GSH or a series of L-c-glutamyl amide donor substrates
and GlyGly as acceptor substrate (Table 1). Also shown in Table 1
are the pKa values31,32 for the conjugate acids of the amine leaving
groups of the acylation reaction. It is apparent from Table 1 that
the relative reactivities of the different donor substrates studied
herein fall into three categories, their kcat values for rate-limiting
acylation separated roughly by an order of magnitude. The native
tripeptide substrate, GSH, undergoes transpeptidation turnover
most rapidly, while dipeptide substrates and finally simple c-
glutamyl alkyl amides are slower. This tendency, represented
in Fig. 3, does not correlate simply with the nucleofugality of
the leaving group during the acylation step. Reaction efficiency
is apparently dominated by unidentified interactions related to
gross substrate structure and affinity, rather than limited to the
Fig. 3 Brønsted plot for the acylation of GGT by a series of L-c-glutamyl
substrates in the presence of 20 mM GlyGly at pH 8.0 and 37 ◦C. pKa NH3+
refers to the pKa of the conjugate acid ammonium of the amine leaving
group. Data obtained previously for a series of anilides15 are shown as
circles. Data obtained in this work are shown in squares for a series of
alkyl amides, triangles for dipeptides and as a diamond for GSH (kcat
,
Table 1).
basicity of the departing amine. Simply stated, among the substrate
analogues tested, the turnover of the acylation step appears to
increase with the similarity of the leaving group to CysGly, the
initial product of the reaction with the native substrate GSH.
Also shown in Fig. 3 are kcat values previously obtained for
acylation of GGT by a series of c-glutamyl anilides.15 These
data were obtained under similar conditions, using a relatively
high concentration of GlyGly as an acceptor substrate, allowing
Table 1 Kinetic parameters (kcat, KM) determined for the series of c-glutamyl substrate analogues studied herein
kcatb/s−1
KMc/mM
0.67
k
cat/KM/mM−1 s−1
1480
a
Donor substrate
X
pKa NH3+
8.01
L-c-Glutamyl-L-cysteinylglycine (glutathione, GSH)
990
L-c-Glutamylglycine methyl ester
L-c-Glutamylglycinamide
7.92
8.23
172
2.78
3.30
61.9
95.9
29.1
L-c-Glutamylpropylamide
10.53
9.57
6.02
20.2
11.6
33.9
0.71
0.49
2.35
28.5
23.7
14.4
L-c-Glutamyl-2-(ethylthio)ethylamide
L-c-Glutamyl-2,2,2-trifluoroethylamide
a From ref. 31 and 32. b Relative error for kcat values was ≤6% (SD) from non-linear regression. c Relative error for KM values was ≤15% (SD) from
non-linear regression.
3794 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
comparison with the data obtained herein. In general, two
peculiarities of the anilide substrates are evident from this
comparison—their greater reactivity than alkyl amides and dipep-
tides and the distinctive free-energy relationship with the basicity
of their departing aniline groups. The enhanced reactivity of
anilides compared to alkyl amides toward acyl transfer reactions
is well-known, and they have long been studied as models for
amide hydrolysis reactions for this reason.33 It is therefore not
surprising that the anilides studied previously are more efficient
substrates than the alkyl amides and peptides studied herein. It is
clear from Fig. 4 that the positive Brønsted correlation for the most
basic anilines does not continue with the more basic amine leaving
groups studied herein, for example. The non-linear free-energy
relationship for the kcat values measured previously for the anilides
was ascribed to their concerted protonation and departure from
the tetrahedral intermediate formed during the acylation step.15
However, no obvious trend is apparent for the kcat values measured
for the substrate analogues studied herein.
GGT from a series of activated anilides, where nucleophilic attack
is apparently faster than the rate-limiting breakdown of the tetra-
hedral intermediate. Studies of non-enzymatic amide hydrolysis
(or the microscopic reverse reaction) have shown that either
formation or breakdown of the tetrahedral intermediate can be
rate-limiting and that related Brønsted slopes can be very weakly
negative.34,35 Alternatively, similar studies36,37 with chymotrypsin,
a hydroxyl-nucleophile enzyme that catalyzes analogous acyl
transfer reactions, have shown that weak Brønsted slopes may
be observed when different means of substrate activation cancel
each other. For example, decreasing electrophilicity within a series
of substrates may be compensated by increasing hydrogen bonding
in the oxyanion hole of chymotrypsin—or in this case, with the
conserved glycine residues of GGT (Fig. 5).
Fig. 5 Crystal structure of c-glutamyl acyl-enzyme intermediate of E. coli
GGT from ref. 21 (PDB entry 2DBW). Active site nucleophile Thr391,
hydrogen bonding Gly382 and Gly383 and conserved binding site residues
Thr409 and Asn411 are shown in stick form.
Fig. 4 Brønsted plot for the acylation of GGT by a series of L-c-glutamyl
substrates in the presence of 20 mM GlyGly at pH 8.0 and 37 ◦C. pKa NH3+
refers to the pKa of the conjugate acid ammonium of the amine leaving
group. Data are shown in squares for a series of alkyl amides, triangles for
dipeptides and as a diamond for GSH (kcat/KM, Table 1).
Structural considerations
Rather, a more linear correlation is observed for the Brønsted
plot constructed from the ratio of kcat/KM (Fig. 4), suggesting that
the rate-limiting step is the reaction of the free enzyme with the
substrate rather than the decomposition of a covalent tetrahedral
intermediate, as postulated for the anilides. From this plot it is
again apparent that the relative efficiencies of the different donor
substrates studied herein fall into at least two categories, their
specificity constants (kcat/KM) separated by roughly two orders
of magnitude. The native tripeptide substrate, GSH, undergoes
transpeptidation most efficiently, while the dipeptide and c-
glutamyl alkyl amide substrates are far less efficient. The near-
zero slope obtained for the substrate analogues suggests there is
very little development of charge on the nitrogen of the leaving
group at the transition state of the rate-determining step. Taken
together, the results shown in Fig. 3 and 4 are consistent with
rate-limiting nucleophilic attack of GGT on substrate analogues,
followed by rapid protonation and expulsion of the amine leaving
group to form the acyl enzyme. Again, this mechanism differs
markedly from that postulated previously15 for the acylation of
The differences between the mechanisms proposed for acylation
from anilides or from alkyl amides may be accounted for by
the structural differences between the series of substrates studied
and their accompanying interactions with the enzyme. Recently,
Fukuyama and coworkers published the first high resolution GGT
crystal structure for the E. coli enzyme.21 In their publication,
they draw attention to the exposed cavity beside the c-glutamyl
carbonyl of the crystalline acyl-enzyme intermediate (Fig. 5).
When we used their crystal structure to construct the tetrahedral
intermediate formed from nucleophilic attack on the native sub-
strate GSH, the CysGly leaving group could be manually docked
into the adjacent putative binding site, providing a compelling
image of a potentially relevant mechanistic intermediate (Fig. 6).
From this model it is apparent that the dimensions of the cavity are
appropriate for binding CysGly and that the departing nitrogen
is correctly positioned to be protonated by the ammonium group
of the N-terminal nucleophilic Thr391 residue. The floor of the
cavity is formed by residues such as Asn411 and Thr409 in
the E. coli enzyme that are conserved in the rat and human
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3790–3801 | 3795
©

View Article Online
Fig. 6 Stereo view of model of anionic GSH tetrahedral intermediate constructed from crystalline acyl-enzyme intermediate, intended to demonstrate
the scale of the putative Cys–Gly subsite. Residues can be identified from Fig. 5, with Thr409 and Asn411 shown in line form and all labels removed for
the sake of clarity. Image generated using Accelrys Discovery Studio Visualizer.
enzymes. Furthermore, in the model, Thr409 can form a close
dipole interaction with the cysteine sulfhydryl while Asn411 can
Experimental
Materials
form a hydrogen bond with the glycine NH of the leaving group.
The glycine carboxylate extends to the edge of the binding site
where its potential interactions with solvent may also contribute
to orienting the CysGly leaving group in the cavity. While these
interactions will be verified through more rigorous modelling
studies and/or future crystallographic efforts, the existence and
nature of the cavity allows meaningful interpretation of the kinetic
data presented herein.
The c-glutamyl anilide substrates studied previously would ap-
pear to owe their reactivity to their predisposition to nucleophilic
attack,33 leading to acylation through rate-limiting breakdown of
their tetrahedral intermediate15 without assistance from any spe-
cific interaction between the enzyme and the aniline leaving group.
Alkyl amides are less activated and subsequently less efficient
substrates such that nucleophilic attack may be the rate-limiting
step in their acylation of GGT. As the resemblance between these
substrate analogues and the native substrate GSH increases, their
efficiency increases also. This may be due to the ability of peptidic
substrates to take advantage of specific interactions in the binding
site identified by Fukuyama. Ultimately, the binding of CysGly
in the site would realise optimal interactions and imparts to
GSH its optimal reactivity. These interactions may help to distort
General information. Dabsyl chloride and synthetic starting
materials were purchased from Sigma-Aldrich; dabsyl chloride
was recrystallised from acetone.39 All aqueous solutions were
prepared from water purified and deionized by a Millipore Milli-
Q Biocell system. All HPLC solvents including acetone were
purchased from VWR. HPLC solvents were filtered by a Millipore
filter (0.22 lm). Sample preparation was carried out using 0.22 lm
filters and 2 mL micro tubes with caps purchased from Sarstedt
and automated sample injection using 300 lL crimp vials with
clear snap-close lids purchased from VWR. Separation and
detection was effected using a Hitachi D-7000 HPLC, comprising
a model L-7100 pump, a Higgins Targa C18 5 lM HPLC column
(250 × 4.6 mm, Chromabec), an automatic injection system (model
L-7200) and a photodiode detector (model L-7450).
Enzyme preparation. GGT was purified from rat kidney
according to a modified protocol15 based on the method published
by Tate and Meister.1,19
Synthesis of glutathione analogue substrates.
=
the c-glutamyl amide group, disrupting N–C O resonance and
Amide coupling reaction: general method A. Diisopropylethy-
lamine (DIEA) (2.32 mL, 13.3 mmol, 4.5 eq) was added dropwise
under an inert atmosphere at room temperature to a solution of
N-Boc-L-glutamic acid a-benzyl ester 1 (1.0 g, 2.9 mmol, 1.0 eq)
in 100 mL anhydrous acetonitrile. The solution was stirred for
about ten minutes prior to the addition of alkyl amine in the
form of its hydrochloride salt (8.9 mmol, 3.0 eq). After complete
dissolution, N,N,Nꢀ,Nꢀ-tetramethyl-O-(benzotriazol-1-yl)uronium
rendering the carbonyl more susceptible to nucleophilic attack,
a mechanism suggested to be important for amide cleavage
by proteases as well.38 This form of activation would be more
important than remote inductive effects expected from electron
withdrawing substituents on the amine leaving group, which may
explain why the slope of Fig. 4 is near zero, rather than negative
as expected for a series of substrates of varying electrophilicity.
3796 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
tetrafluoroborate (TBTU) (1.9 g, 5.9 mmol, 2.0 eq) was added
and the reaction mixture was stirred prior to addition of 100 mL
of ethyl acetate. The organic phase was washed with water (2 ×
30 mL), 0.1 M HCl solution (20 mL), a solution of 5% NaHCO3
(2 × 20 mL) and finally with saturated NaCl (20 mL). The organic
phase was dried over MgSO4, filtered and concentrated under
reduced pressure. The alkyl amide product was obtained as an
off-white solid.
Removal of Boc protecting group: general method B
To a solution of N-Boc-L-c-glutamylalkylamide a-benzyl ester 2–
5 (2.9 mmol, 1.0 eq) in 50 mL of anhydrous dichloromethane
was added, dropwise at 0 ◦C, trifluoroacetic acid (TFA) (4.5 mL,
59 mmol, 20 eq). The solution was stirred at 0 ◦C for about
30 minutes and then at room temperature for around 2 hours. Then
4.5 mL of TFA (20 eq) at 0 ◦C was added again. The solution was
stirred at 0 ◦C for another 30 minutes, and then again for 2 hours at
room temperature. The solvent was then removed under reduced
pressure and the crude product was washed with CH2Cl2 (2 ×
10 mL) and with cyclohexane (4 × 10 mL). This TFA salt of the
product amine was then dissolved in 0.1 M HCl (50 mL) and
washed with diethyl ether (3 × 15 mL). The aqueous phase was
evaporated under reduced pressure to give the product amine as a
yellowish-white HCl salt.
N-Boc-L-c-glutamyl-2,2,2-trifluoroethylamide a-benzyl ester (2).
The general coupling procedure was followed using 2,2,2-
trifluoroethylamine HCl (1.2 g, 8.9 mmol, 3.0 eq) and stirring
for 6 hours. Amide 2 was obtained as a beige solid (1.18 g,◦95%).
Rf: 0.66 (70 : 30 EtOAc–hexane). Mp 96–97 ◦C. [a]D20: −2.3 (c =
1.00; CHCl3). 1H-NMR (400 MHz, CD3OD) d (ppm): 7.37–7.33
(m, 5 H), 5.17 (q, J = 10.2 Hz, 2 H), 4.17 (m, 1 H), 3.89–3.85 (m,
2 H), 2.35 (t, J = 7.3 Hz, 2 H), 2.15–2.12 (m, 1 H), 1.94–1.92 (m,
1 H), 1.43 (s, 9 H). 13C-NMR (100 MHz, CDCl3) d (ppm): 172.3,
171.9, 155.9, 135.0, 128.5, 128.4, 128.2, 124.0 (q, JC–F = 278.6 Hz),
80.2, 67.1, 52.8, 40.0 (q, JC–F = 34.4 Hz), 38.4, 32.0, 28.0 HRMS
C19H25F3N2O5 (M + H)+ calc: 419.1716, found: 419.1783.
L-c-Glutamyl-2,2,2-trifluoroethylamide a-benzyl ester (6). The
general procedure for Boc group removal was followed using 1.2 g
of N-Boc-L-c-glutamyl-2,2,2-trifluoroethylamide a-benzyl ester 2
to give amine 6 as a yellowish-white solid HCl salt (0.98 g, 94%).
Rf: 0.13 (70 : 30 EtOAc–hexane). Mp 125–127 ◦C. [a]D20: +5.1◦ (c =
1.00; H2O). 1H-NMR (400 MHz, CD3OD) d (ppm): 7.41–7.36 (m,
5 H), 5.28 (m, 2 H), 4.16–4.13 (m, 1 H), 3.89–3.84 (m, 2 H), 2.51–
2.47 (m, 2 H), 2.21–2.17 (m, 2 H). 13C-NMR (100 MHz, CD3OD) d
N-Boc-L-c-glutamylglycine methyl ester a-benzyl ester (3). The
general coupling procedure was followed using glycine methyl ester
HCl (1.1 g, 8.9 mmol, 3.0 eq) and stirring for 5 hours. Amide 3
was obtained as a yellowish white solid (0.98 g, 82%). Rf: 0.48 (70 :
30 EtOAc–hexane). Mp 88–90 ◦C. [a]D20: +2.3◦ (c = 1.00; CHCl3).
1H-NMR (400 MHz, CD3OD) d (ppm): 7.39–7.31 (m, 5 H), 5.17
(q, J = 10.3 Hz, 2 H), 4.17 (m, 1 H), 3.91 (s, 2 H), 3.71 (s, 3 H),
2.35 (t, J = 7.6 Hz, 2 H), 2.15–2.13 (m, 1 H), 1.92 (m, 1 H),
1.43 (s, 9 H). 13C-NMR (100 MHz, CDCl3) d (ppm): 172.6, 170.8,
156.2, 135.7, 129.0, 128.9, 128.7, 80.5, 67.6, 53.3, 52.7, 41.7, 32.6,
29.4, 28.7. HRMS C20H28N2O7 (M + H)+ calc: 409.1897, found:
409.1965.
(ppm): 174.3, 170.0, 136.3, 129.9, 129.6 (2 peaks), 127.2 (q, JC–F
=
278.2 Hz), 69.2, 53.5, 41.6 (q, JC–F = 34.8 Hz), 31.8, 26.8. HRMS
C14H17F3N2O3 (M + H)+ calc.: 319.1191, found: 319.1260.
L-c-Glutamylglycine methyl ester a-benzyl ester (7). The gen-
eral procedure for Boc group removal was followed, but without
washing with ether or formation of the HCl salt, using N-Boc-L-c-
glutamylglycine methyl ester a-benzyl ester 3 (0.98 g, 2.4 mmoles,
1.0 eq) to give the TFA salt of amine 7 as a brownish oil (0.97 g,
95%). Rf: 0.08 (70 : 30 EtOAc–hexane). [a]2D0: +3.6◦ (c = 1.00;
H2O). 1H-NMR (300 MHz, CD3OD) d (ppm): 7.41–7.34 (m, 5 H),
5.29 (m, 2 H), 4.16 (m, 1 H), 3.91 (s, 2 H), 3.70 (s, 3 H), 2.49 (t, J =
7.0 Hz, 2 H), 2.25–2.10 (m, 2 H). 13C-NMR (100 MHz, CD3OD) d
(ppm): 174.5, 171.9, 170.1, 161.0 (q, JC–F = 38.0 Hz), 136.3, 129.8,
129.7, 129.6, 117.2 (q, JC–F = 287.9 Hz), 69.2, 53.6, 52.7, 41.9, 31.8,
27.0. LRMS C15H20N2O5 (M + H)+ calc.: 309.1, found: 309.1.
N-Boc-L-c-glutamylglycinamide a-benzyl ester (4). The gen-
eral coupling procedure was followed using glycinamide HCl
(0.33 g, 2.9 mmol, 2.0 eq) and stirring for 18 hours. Amide 4
was obtained as a beige sol◦id (0.38 g, 65%). Rf: 0.83 (70 : 30
EtOAc–hexane). Mp 43–45 C. [a]2D0: +4.5◦ (c = 1.00; CHCl3).
1H-NMR (400 MHz, CD3OD) d (ppm): 7.37–7.31 (m, 5 H), 5.17
(q, J = 9.5 Hz, 2 H), 4.18–4.16 (m, 1 H), 3.81 (q, J = 9.3 Hz,
2 H), 2.37 (t, J = 7.4 Hz, 2 H), 2.16–2.13 (m, 1 H), 1.91 (m, 1 H),
1.43 (s, 9 H). 13C-NMR (100 MHz, CDCl3) d (ppm): 173.0, 172.8,
172.4, 156.2, 136.6, 129.1, 128.9, 128.7, 128.5, 80.6, 67.7, 53.4, 43.3,
32.3, 28.7. HRMS C19H27N3O6 (M + H)+ calc: 394.1900, found:
394.1991.
L-c-Glutamylglycinamide a-benzyl ester (8). The general pro-
cedure for Boc group removal was followed using N-Boc-L-c-
glutamylglycinamide a-benzyl ester 4 (0.33 g, 0.8 mmol, 1.0 eq)
to give amine 8 as a yellowish-white solid HCl salt (0.28 g,◦82%).
Rf: 0.05 (70 : 30 EtOAc–hexane). Mp 57–60 ◦C. [a]D20: +5.8 (c =
1.00; H2O). 1H-NMR (400 MHz, CD3OD) d (ppm): 7.44–7.36 (m,
5 H), 5.30 (m, 2 H), 4.17 (t, J = 6.3 Hz, 1 H), 3.85 (m, 2 H),
2.50 (m, 2 H), 2.25 (m, 1 H), 2.18–2.14 (m, 1 H). 13C-NMR
(100 MHz, CD3OD) d (ppm): 174.5, 174.4, 170.2, 136.4, 129.9
(2 peaks), 129.8, 69.3, 53.7, 43.0, 32.0, 26.9. LRMS C14H19N3O4
(M + H)+ calc.: 294.1, found: 294.2.
N-Boc-L-c-glutamylpropylamide a-benzyl ester (5). The gen-
eral coupling procedure was followed using propylamine HCl
(0.85 g, 8.9 mmol, 3.0 eq) and stirring for 4 hours. Amide 5 was
obtained as a white solid (1.08 g, 96%). Rf: 0.62 (70 : 30 EtOAc–
hexane). Mp 87–88 ◦C. [a]D20: −3.5◦ (c = 1.00; CHCl3). 1H-NMR
(400 MHz, CD3OD) d (ppm): 7.37–7.34 (m, 5 H), 5.17 (m, 2 H),
4.17 (m, 1 H), 3.11 (m, 2 H), 2.27 (m, 2 H), 2.17–2.13 (m, 1 H),
1.91 (m, 1 H), 1.45 (m, 2 H), 1.43 (s, 9 H), 0.92 (m, 3 H). 13C-NMR
(100 MHz, CDCl3) d (ppm): 172.0, 171.8, 155.7, 135.2, 128.4,
128.3, 128.2, 79.8, 67.0, 53.1, 41.2, 32.4, 28.6, 28.1, 22.5, 11.2.
HRMS C20H30N2O5 (M + H)+ calc: 379.2155, found: 379.2242.
L-c-Glutamylpropylamide a-benzyl ester (9). The general pro-
cedure for Boc group removal was followed using N-Boc-L-c-
glutamylpropylamide a-benzyl ester 6 (1.08 g, 2.9 mmol, 1.0 eq)
giving 9 as a yellowish oil (1.01 g, 95%). Rf: 0.2 (70 : 30 EtOAc–
hexane). [a]2D0: +4.9◦ (c = 1.00; H2 O). 1H-NMR (400 MHz,
CD3OD) d (ppm): 7.44–7.35 (m, 5 H), 5.30 (m, 2 H), 4.15
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3790–3801 | 3797
©

View Article Online
(t, J = 6.4 Hz, 1 H), 3.11 (t, J = 7.3 Hz, 2 H), 2.42 (t, J =
6.9 Hz, 2 H), 2.19–2.13 (m, 2 H), 1.53–1.47 (m, 2 H), 0.91 (t,
J = 7.5 Hz, 3 H). 13C-NMR (100 MHz, CD3OD) d (ppm): 173.8,
170.1, 136.4, 129.9 (3 peaks), 69.2, 53.7, 42.3, 32.3, 27.2, 23.5, 11.7.
HRMS C15H22N2O3 (M + H)+ calc.: 279.1630, found: 279.1713.
to give final product ◦13 as a white solid (0.33 g, 50%). Mp
◦
1
86–89 C. [a]2D0: +14.3 (c = 1.00; H2O). H-NMR (400 MHz,
D2O) d (ppm): 4.05 (t, J = 6.1 Hz, 1 H), 3.14 (t, J = 6.9 Hz, 2 H),
2.50–2.45 (m, 2 H), 2.23–2.18 (m, 2 H), 1.53–1.48 (m, 2 H), 0.88
(t, J = 7.3 Hz, 3 H). 13C-NMR (100 MHz, D2O) d (ppm): 174.8,
172.7, 53.3, 42.1, 32.1, 26.7, 22.5, 11.4. HRMS C8H16N2O3 (M +
H)+ calc.: 189.1161, found: 189.1243.
Removal of benzyl protecting group: general method C
A
solution of L-c-glutamylalkylamide a-benzyl ester 6–9
N-Boc-L-c-glutamyl-2-(ethylthio)ethylamide a-benzyl ester (14).
N-Boc-L-glutamic acid a-benzyl ester 1 (0.6 g, 1.78 mmol, 1.0 eq)
was coupled with 2-(ethylthio)ethyl amine according to Method
A, with the following exceptions: 1.5 eq of alkyl amine and 4.0 eq
of DIEA were used (instead of 3.0 eq and 4.5 eq, respectively)
and the reaction with TBTU was allowed to proceed overnight.
Amide 14 was obtained as a beige solid (0.67 g, 88%). Rf: 0.23 (70 :
(2.5 mmol, 1.0 eq) in 50 mL methanol was added, under inert atmo-
sphere at room temperature, to palladium over carbon (10 wt%).
The reaction mixture was allowed to stir for 18 hours under
150 psi hydrogen in a hydrogenation bomb. Once the reaction
R
was complete, the reaction mixture was filtered over Celiteꢀ and
washed with methanol. The methanol was then removed under
reduced pressure and the crude reaction product was obtained
as an oil. The product was then purified by precipitation from a
methanol–ether mixture to give the final product as a white solid.
◦
30 EtOAc–hexane). Mp 62–65 C. [a]D: −1.5 (c = 0.01; CHCl3).
1H-NMR (300 MHz, CDCl3) d (ppm): 7.33 (m, 5 H), 5.13 (q, J =
12.9 Hz, 2 H), 4.31 (m, 1 H), 3.38 (q, J = 6.3 Hz, 2 H), 2.62 (t, J =
6.6 Hz 2 H), 2.50 (q, J = 7.4 Hz, 2 H), 2.26–2.21 (m, 2 H), 2.17–
2.14 (m, 1 H), 1.95–2.02 (m, 1 H), 1.41 (s, 9 H), 1.23 (t, J = 7.4,
3 H). 13C-NMR (75 MHz, CDCl3) d (ppm): 172.6, 172.4, 156.2,
135.7, 128.9, 80.4, 78.0, 67.5, 60.8, 53.6, 39.0, 32.9, 31.5, 30.0, 28.7,
25.9, 15.1. HRMS C21H32N2O5S (M + H)+ calc.: 424.5560, found:
424.2032.
L-c-Glutamyl-2,2,2-trifluoroethylamide (10). The general
method C for removal of the benzyl group was followed using
L-c-glutamyl-2,2,2-trifluoroethylamide a-benzyl ester 7 (0.88 g,
2.5 mmol, 1.0 eq) and 0.09 g of palladium. The crude reaction
product was obtained as an oil (0.6 g, 92%) and the purified
product 10 was obtained as a white solid (0.39 g, 60%). Mp
167–170 C (dec.). [a]2D0: +14.1◦ (c = 1.00; H2 O). 1H-NMR
◦
N-Boc-L-c-glutamyl-2-(ethylthio)ethylamide (15). A solution
of N-Boc-L-c-glutamyl-2-(ethylthio)ethylamide a-benzyl ester 14
(225 mg, 0.141 mmol, 1.0 eq) dissolved in a 1 : 1 mixture of
H2O–THF was cooled to 0 ◦C, followed by the addition of K2CO3
(219 mg, 1.58 mmol, 3 eq) to the reaction mixture. The mixture was
stirred initially at 0 ◦C and allowed to warm to room temperature
overnight. Water was then added (5 mL) and the reaction was
stirred for another 5 hours. The THF was then removed under
reduced pressure and the aqueous phase was acidified with 1 M
HCl and extracted with 25 mL EtOAc. The organic phase was
then dried over MgSO4, filtered and removed under reduced
pressure. The residue was dissolved in EtOAc and purified by
flash chromatography (80 : 15 : 5 EtOAc–hexane–HOAc). Acid
15 was thus obtained as a colourless gum (107 mg, 61%). Rf: 0.55
(80 : 15 : 5 EtOAc–hexane–HOAc). [a]D: +8.3 (c = 0.0029; CHCl3).
1H-NMR (300 MHz, CDCl3) d (ppm): 4.30 (m, 1 H), 3.48 (q, J =
6.1 Hz, 2 H), 2.70 (t, J = 6.6 Hz, 2 H), 2.57 (q, J = 7.4 Hz, 2
H), 2.44 (m, 1 H), 2.22–2.19 (m, 1 H), 2.17–2.04 (m, 1 H), 1.46
(s, 9 H), 1.28 (t, J = 7.4, 3 H). 13C-NMR (75 MHz, CDCl3) d
(ppm): 174.6, 174.0, 156.5, 80.7, 53.3, 39.2, 33.0, 30.6, 29.4, 28.8,
25.9, 15.2. HRMS C14H26N2O5 S (M + H)+ calc.: 334.4354, found:
334.1562.
(400 MHz, D2O) d (ppm): 3.84–3.79 (m, 3 H), 2.45–2.40 (m, 2 H),
2.08 (m, 2 H). 13C-NMR (100 MHz, D2O) d (ppm): 175.7, 172.3,
124.9 (q, JC–F = 278.0 Hz), 53.0, 41.0 (q, JC–F = 34.6 Hz), 31.7,
26.1. HRMS C7H11F3N2O3 (M + H)+ calc.: 229.0722, found:
229.0810.
L-c-Glutamylglycine methyl ester (11). The general method
C for removal of the benzyl group was followed using L-c-
glutamylglycine methyl ester a-benzyl ester 7 (0.97 g, 2.3 mmol,
1.0 eq). The crude reaction product, obtained as a yellow oil
(0.71 g, 92%) was further purified by precipitation from methanol–
ether to give final product 11 as a white solid (0.42 g, 55%). Mp
◦
◦
195 C (dec.). [a]2D0: +6.7 (c = 1.00; H2O). H-NMR (400 MHz,
D2O) d (ppm): 4.04 (s, 2 H), 3.81 (t, J = 6.1 Hz, 1 H), 3.77 (s,
3 H), 2.52 (t, J = 7.5 Hz, 2 H), 2.20–2.16 (m, 2 H). 13C-NMR
(100 MHz, D2O) d (ppm): 176.1, 174.6, 173.1, 54.8, 53.6, 42.0,
31.9, 27.0. HRMS C8H14N2O5 (M + H)+ calc.: 219.1, found: 219.1.
1
L-c-Glutamylglycinamide (12). The general method C for
removal of the benzyl group was followed using L-c-gluta-
mylglycinamide a-benzyl ester 8 (0.23 g, 0.7 mmol, 1.0 eq). The
crude reaction product, obtained as a yellow oil (0.16 g, 95%) was
further purified by precipitation from methanol–ether to give final
product 12 as a white solid (0.13 g, 76%). Mp 190 ◦C (dec.). [a]D20:
+11.3◦ (c = 1.00; H2O). 1H-NMR (400 MHz, D2O) d (ppm): 3.88
(t, J = 6.2 Hz, 1 H), 3.81 (s, 2 H), 2.45 (m, 2 H), 2.11–2.08 (m,
2 H). 13C-NMR (100 MHz, D2O) d (ppm): 175.8, 175.1, 172.5,
53.3, 42.9, 31.6, 26.2. HRMS C7H13N3O4 (M + H)+ calc.: 204.0906,
found: 204.0979.
L-c-Glutamyl-2-(ethylthio)ethylamide (16). To a solution of N-
Boc-L-c-glutamyl-2-(ethylthio)ethylamide 15 (38 mg, 0.11 mmol,
1.0 eq) dissolved in dichloromethane (10 mL), TFA (0.026 mL,
0.342 mmol, 3 eq) was added dropwise at 0 ◦C. After stirring for
20 minutes, the reaction mixture was allowed to warm to room
temperature. After stirring for another 90 minutes, another 3 eq
aliquot of TFA was added dropwise at 0 ◦C over 20 minutes. The
mixture was stirred for another 3.5 hours at room temperature
and another 3 eq aliquot of TFA was added dropwise at 0 ◦C. The
solution was allowed to warm to room temperature with stirring
overnight. Solvent was then removed under reduced pressure.
Cyclohexane was added and subsequently removed under reduced
L-c-Glutamylpropylamide (13). The general method C for
removal of the benzyl group was followed using L-c-
glutamylpropylamide a-benzyl ester 9 (0.94 g, 3.0 mmol, 1.0 eq).
The crude reaction product, obtained as a colourless oil (0.6 g,
89%) was further purified by precipitation from methanol–ether
3798 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
pressure 4 times. The residue was dissolved in CH2Cl2 and purified
by elution from a 3 cm plug of silica gel, through rapid washing
with 10 mL aliquots of hexane, EtOAc, CH2Cl2 and a subsequent
gradient of MeOH in CH2Cl2. A white solid was thus obtained
(23 mg, 54%). Rf: 0.56 (80 : 15 : 5 EtOAc–hexane–HOAc). Mp
198–201 C. H-NMR (300 MHz, D2O) d (ppm): 3.74 (t, J =
6.1 Hz, 1 H), 3.39 (t, J = 6.7 Hz, 2 H), 2.69 (t, J = 6.6 Hz, 2 H),
2.56 (q, J = 7.4 Hz, 2 H), 2.42–2.37 (m, 2 H), 2.14–2.09 (m, 2 H),
1.20 (t, J = 7.4 Hz, 3 H). 13C-NMR (75 MHz, D2O) d (ppm): 175.0,
174.3, 54.5, 38.9, 31.8, 30.2, 26.8, 25.3, 14.2. HRMS C9H18N2O3S
(M + H)+ calc.: 234.3222, found: 234.1038.
Cyclohexane was then added and subsequently removed under
reduced pressure 4 times. The resulting residue was then dissolved
in 25 mL EtOAc and extracted using ∼0.1 M aqueous HCl (3 ×
25 mL). The aqueous phase was then evaporated under reduced
pressure, giving the HCl salt of tripeptide 20 as a pale yellow gum
(31 mg, 80%). Rf: 0.07 (80 : 15 : 5 EtOAc–hexane–HOAc) [a]D:
+18.4 (c = 0.01; CHCl3). 1H-NMR (300 MHz, D2O) d (ppm): 4.27
(q, J = 8.0 Hz, 1 H), 3.96 (t, J = 7.1 Hz, 1 H), 3.89 (s, 2 H), 2.31
(t, J = 7.0 Hz, 2 H), 1.99–1.93 (m, 2 H), 1.05 (d, J = 7.9 Hz, 3 H).
13C-NMR (75 MHz, D2O) d (ppm): 176.1, 174.5, 173.4, 52.7, 50.0,
41.3, 31.1, 25.7, 16.9, 16.3. HRMS C10H17N3O6 (M + H)+ calc.:
275.2643, found: 275.1117.
◦
1
N-Boc-L-c-glutamyl-AlaGly a-benzyl ester (18). N-Boc-L-
glutamic acid a-benzyl ester 1 (100 mg, 0.296 mmol, 1.0 eq) was
dissolved in 4 mL anhydrous acetonitrile under an inert atmo-
sphere at 0 ◦C. Triethylamine (0.041 mL, 0.296 mmol, 1.0 eq) was
then added dropwise, followed by p-nitrophenyl chloroformate
(63 mg, 0.311 mmol, 1.05 eq). After stirring for 5 minutes, 4-
dimethylaminopyridine (DMAP) (4 mg, 0.030 mmol, 0.1 eq) was
added. After 45 minutes, excess p-nitrophenyl chloroformate was
hydrolysed upon addition of 1 mL of a solution of triethylamine
(0.164 mL, 1.184 mmol, 4.0 eq) in 4 mL water. After ∼10 minutes,
L-AlaGly (173 mg, 1.184 mmol, 4.0 eq) was added to the remaining
3 mL of the aqueous triethylamine solution, and the resulting
solution was added dropwise to the reaction mixture. The reaction
was allowed to stir for 2.5 hours. Solvents were then removed under
reduced pressure and the resulting residue was purified by flash
chromatography (85 : 10 : 5 EtOAc–MeOH–HOAc). Tripeptide 18
was obtained as a colourless gum (66 mg, 54%). Rf: 0.48 (85 : 10 : 5
EtOAc–MeOH–HOAc). [a]D: −4.2 (c = 0.005; CHCl3). 1H-NMR
(300 MHz, CD3OD) d (ppm): 7.30–7.36 (m, 5 H), 5.15 (q, J =
12.5 Hz, 2 H), 4.38 (m, 1 H), 4.17 (m, 1 H), 3.84 (s, 2 H), 2.34 (m,
2 H), 1.42 (s, 9 H), 2.15–2.12 (m, 1 H), 1.94–1.89 (m, 1 H), 1.34 (d,
J = 7.1 Hz, 3 H). 13C-NMR (75 MHz, CD3OD) d (ppm): 174.3,
173.5, 172.9, 157.1, 136.2, 128.6, 128.2, 79.7, 66.9, 53.7, 49.3, 31.8,
27.7, 27.1, 17.0. HRMS C22H31N3O8 (M + H)+ calc.: 465.4949,
found 465.2111.
Methods
Assay procedure. Enzyme activity was normalized15 using
the standard GGT specific activity assay. The progress of the
enzymatic reactions of the L-c-glutamyl amide substrate analogues
studied was followed by the formation of the transpeptidation
product, L-c-GluGlyGly. A test tube was charged with donor
substrate giving final concentrations of 0.05–3.5 mM, after
addition of 0.50 mL of a solution of 0.1 M glycylglycine, pH 8.0
(final concentration 20 mM) and dilution to 2.5 mL with 0.1 M
HEPES buffer (pH 8.0). (HEPES was chosen as an appropriate
reaction buffer since it does not contain a primary amine group
for adventitious reaction with DABS-Cl during the subsequent
derivatization step.) The solution was pre-incubated for 10 min at
37 ◦C and the reaction was initiated by the addition of 0.05–
1.19 units of GGT, depending on the substrate analogue. At
different times between 0 and 75 minutes, an aliquot of 195 lL
was taken from the reaction vessel and transferred to a 2 mL
microcentrifuge tube containing 60 lL of 40% trichloroacetic acid
(TCA), thus quenching the reaction.17 The enzyme precipitate
was removed by centrifugation at 10 000 rpm (10 600 g) for
5 min at 4 ◦C. The supernatant fraction was then removed and
placed in a second 2 mL microcentrifuge tube. A 20 lL aliquot
of the standard 50 mM glutamine solution (in 0.2 M sodium
bicarbonate, pH 9.0) was then added to each quenched reaction
solution as an internal standard (0.5 mM final concentration). The
concentration of derivatized reaction product in each quenched
sample was determined by HPLC (see below) and plotted against
reaction time to give linear initial rates. These acylation rates
(measured under conditions where deacylation is rapid) were used
to determine the kinetic parameters Vmax and KM according to the
standard hyperbolic Michaelis–Menten equation by non-linear
regression. The value for Vmax was converted to kcat by dividing
by the concentration of enzyme used. The kinetic parameters of
L-AlaGly as acceptor substrate were determined according to the
same general procedure, using final concentrations of L-AlaGly
between 0.2 and 5 mM and 4 mM of GSH as donor substrate.
N-Boc-L-c-glutamyl-L-alanylglycine (19). N-Boc-L-c-gluta-
myl-L-alanylglycine a-benzyl ester 18 (63 mg, 0.151 mmol) was
dissolved in ∼5 mL of methanol. Palladium on charcoal (6 mg,
10 wt%) was then added to the reaction mixture, that was allowed
to react overnight under 150 psi hydrogen in a hydrogenation
R
bomb. The reaction mixture was then filtered over Celiteꢀ and
washed with methanol. The solvent was then evaporated under
reduced pressure to give 19 as a pale white gum (49 mg, 99%). Rf:
0.21 (85 : 10 : 5 EtOAc–MeOH–HOAc). [a]D: −6.4 (c = 0.002;
1
MeOH). H-NMR (300 MHz, CD3OD) d (ppm): 4.38 (q, J =
7.1 Hz, 1 H), 4.08 (m, 1 H), 3.87 (s, 2 H), 2.36 (m, 2 H), 2.12 (m,
1 H), 1.95–1.89 (m, 1 H), 1.44 (s, 9 H), 1.36 (d, J = 7.1 Hz, 3
H). 13C-NMR (75 MHz, CD3OD) d (ppm): 175.6, 174.4, 173.9,
172.9, 157.0, 79.5, 53.9, 49.4, 41.5, 32.0, 27.8, 27.8, 17.0. HRMS
C15H25N3O8 (M + H)+ calc.: 375.3775, found: 375.1642.
Dabsyl chloride derivatization. Dabsyl derivatives were pre-
pared according to a modified26 procedure based primarily on the
method published previously by Chang et al.39 Firstly, the pH
of each aliquot was adjusted to 9.0 0.1 with NaOH. Next, an
equal volume of a freshly prepared 20 mM solution of DABS-Cl
in acetone was added to the solution. The resulting 2-fold molar
excess of DABS-Cl is sufficient to ensure complete derivatization.40
L-c-Glutamyl-L-alanylglycine (20). To a solution of N-Boc-
L-c-glutamyl-L-alanylglycine 19 (41 mg, 0.125 mmol, 1.0 eq)
dissolved in 15 mL CH2Cl2, TFA (0.127 mL, 0.752 mmol,
6.0 eq) was added dropwise over 20 minutes, with stirring, at
0 ◦C. The reaction mixture was then allowed to stir for 5 hours at
room temperature. Solvent was removed under reduced pressure.
◦
The mixture was then heated at 70 C for 10–15 min, to ensure
complete solubility, in 2 mL tubes equipped with airtight caps
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3790–3801 | 3799
©

View Article Online
to prevent the evaporation of acetone. Under these conditions,
derivatization is rapid and complete after heating to the point of
solubilization. Finally, a 1 : 1 (v/v) solution of 40 mM sodium
phosphate (pH 6.5)–ethanol was added, to a final volume of
2 mL, to dilute the derivatized sample prior to HPLC analysis.
Derivatized samples were found to be stable at◦room temperature
for several months41 and can be stored at −20 C for one year.39
Conclusions
In summary, we report a complementary method for kinetic
analysis of the GGT-mediated transpeptidation reactions of non-
chromogenic substrates, based on dabsyl derivatization of all
primary amine products followed by separation and quantification
by reversed-phase HPLC. This method is economical, simple,
reproducible, sensitive and precise. Furthermore, its flexibility
offers the potential for the analysis of enzymatic reactions of series
of substrates that do not contain a chromophore, in particular
those of the transamidases (EC 2.3.-.-) and peptidases (EC 3.4.-.-),
whose peptidic substrates and products are not easily visualized
quantitatively.
Using this kinetic technique, we obtained structure–function
data for GGT-mediated transpeptidation that are not easily
accessible by other methods. Kinetic data obtained for the
acylation of GGT with the native substrate GSH as well as
a series of dipeptide substrate analogues (with GlyGly as an
efficient acceptor substrate) indicate that their reactivity is not
dependent solely on the nucleofugality of the amine leaving group.
The remarkable efficiency of GSH as a donor substrate, taken
together with the higher turnover numbers obtained for dipetide
substrate analogues, seems to suggest there are important enzyme–
substrate affinity interactions that contribute more predominantly
to acylation efficiency. Preliminary homology modelling based on
the recently published21 crystal structure of the similar E. coli
enzyme suggests there may be specific interactions between the
Cys–Gly leaving group of GSH and an adjacent binding sub-site
that may contribute to the activation of GSH as an acyl-donor
substrate. Further studies using peptidic substrate analogues that
closely resemble the native substrates of GGT will be required to
explore this hypothesis.
HPLC analysis. Mixtures of dabsyl derivatives were separated
and analysed by HPLC according to a modified26 method based
on a previously published procedure.42 Samples were prepared
as indicated above and analysed on a reversed-phase column,
using an injection volume of 40 lL. The mobile phase for
separation of dabsyl derivatives was comprised of two solvents:
25 mM monobasic potassium phosphate (pH 6.8) (solvent A)
and an 80 : 20 mixture of acetonitrile–propan-2-ol (solvent
B). The chromatographic profile used to separate products of
the enzymatic reactions is shown in the ESI†. Note that for
the enzymatic reactions of L-c-glutamyl-2,2,2-trifluoroethylamide
and L-c-glutamylglycinamide, additional washing and column
reconditioning steps were added to the profile, also reported in
the ESI. In each case, the flow rate was 1 mL min1−1 and the
elution was carried out at ambient temperature.
Standard calibration plots. Authentic transpeptidation prod-
ucts were dissolved in sodium bicarbonate 0.2 M buffer (pH 9.0)
according to the desired concentration. To a 0.5 mL aliquot of
any given compound were added 20 lL of a standard 0.5 mM
glutamine solution and 0.52 mL of a freshly prepared acetone
solution of DABS-Cl, resulting in a final ≥2-fold excess of
DABS-Cl, sufficient to ensure complete derivatization according
to the method described above.40 The derivatized sample was
then diluted to a final volume of 2 mL with a solution of 1 : 1
40 mM sodium phosphate (pH 6.5)–ethanol prior to injection
onto the HPLC. Subsequent quantification was accomplished
by comparing the areas of the elution peaks of the analyte and
the internal standard glutamine, measured by their absorbance
at 420 nm. These curves displayed excellent linearity (see ESI†).
The plot of the ratio of the area of the elution peak of L-c-
GluGlyGly to that of L-Gln, the internal standard, versus the
injected concentration of L-c-GluGlyGly, had a slope of 1.69 ×
10−3 lM−1(R2 = 0.994). This effective extinction coefficient was
used subsequently to determine the concentration of product L-c-
GluGlyGly in the quenched reaction aliquots. The quantitative
measurement of L-c-GluAlaGly as a transpeptidation product
was accomplished similarly using peak heights, giving a slope of
0.853 × 10−3 lM−1(R2 = 0.999).
Abbreviations used
CHES, 2-(cyclohexylamino)ethanesulfonic acid; DABS, dabsyl,
4-dimethylaminoazobenzene-4ꢀ-sulfonyl; DIEA, diisopropylethy-
lamine; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid; MES, 4-morpholinoethanesulfonic acid; MOPS, 4-
morpholinopropanesulfonic acid; TBTU, N,N,Nꢀ,Nꢀ-tetramethyl-
O-(benzotriazol-1-yl)uronium tetrafluoroborate; TCA, trichloro-
acetic acid; TFA, trifluoroacetic acid.
Acknowledgements
The authors are grateful to Dr Roselyne Castonguay, formerly
of this laboratory, for her careful reading of this manuscript and
helpful suggestions. They also acknowledge the financial support
of the Natural Sciences and Engineering Research Council of
Canada.
pH–Rate curve. The protocol described above for enzyme
kinetic analysis was followed using 0.15–3.0 mM GSH and 20 mM
GlyGly in 100 mM of the following buffers:16 MES (pH 6.0),
MOPS (pH 7.0), HEPES (pH 8.0) and CHES (pH 8.5, pH 9.0,
pH 9.5). Kinetic parameters were determined from non-linear
regression to the Michaelis–Menten equation as described above,
and normalized kcat values were subsequently determined from
non-linear regression according to the following equation for two
ionizations of the Michaelis ES complex:
References
1 A. Meister, S. S. Tate and O. W. Griffith, c-Glutamyl transpeptidase,
Methods Enzymol., 1981, 77, 237–253.
2 K. Bernstro¨m, L. Orning and S. Hammarstro¨m, c-Glutamyl-
transferase transpeptidase, a leukotriene metabolizing enzyme, Meth-
ods Enzymol., 1982, 86, 38–45.
¨
max
k
cat
kcat
=
K
+
3 A. K. Godwin, A. Meister, P. J. O’Dwyer, C. S. Huang, T. C. Hamilton
and M. E. Anderson, High resistance to cisplatin in human ovarian
[H
]
K
[H
2
]
ES
+
+ 1 +
1
ES
3800 | Org. Biomol. Chem., 2006, 4, 3790–3801
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
cancer cell lines is associated with marked increase of glutathione
synthesis, Proc. Natl. Acad. Sci. U. S. A., 1992, 89, 3070–3074.
4 A. Meister, On the enzymology of amino acid transport, Science, 1973,
180, 33–39.
5 J. Sian, D. T. Dexter, A. J. Lees, S. Daniel, P. Jenner and C. D.
Marsden, Gluthatione-related enzymes in brain in Parkinson’s disease,
Ann. Neurol., 1994, 36, 356–361.
23 R. Knecht and J.-Y. Chang, Liquid chromatographic determination
of amino acids after gas-phase hydrolysis and derivatization with
(dimethylamino) azobenzenesulfonyl chloride, Anal. Chem., 1986, 58,
2375–2379.
24 J. Vendrell and F. X. Avile´s, Complete amino acid analysis of proteins
by dabsyl derivatization and reversed-phase liquid chromatography,
J. Chromatogr., 1986, 358, 401–413.
6 B. Del Bello, A. Paolicchi, M. Comporti, A. Pompella and E. Maellaro,
Hydrogen peroxide produced during c-glutamyl transpeptidase activity
is involved in prevention of apoptosis and maintenance of proliferation
in U937 cells, FASEB J., 1999, 13, 69–79.
25 J.-Y. Chang, Analysis of phosphor-amino acids and amino acid amides
at the picomole level using 4-dimethylaminoazobenzene-4ꢀ-sulfonyl
chloride, J. Chromatogr., 1984, 295, 193–200.
26 C. Rivard, MSc Thesis, Universite´ de Montre´al, 2003.
27 S. Van Dyck, A. Van Schepdael and J. Hoogmartens, Kinetic study of
c-glutamyltransferase activity by electrophoretically mediated micro-
analysis combined with micellar electrokinetic capillary chromatogra-
phy, Electrophoresis, 2002, 23, 2854–2859.
7 D. H. Lee, M. H. Ha, J. H. Kim, D. C. Christiani, M. D. Gross,
M. Steffes, R. Blomkoff and D. R. Jacobs, c-Glutamyltransferase
and diabetes—a 4 year follow-up study, Diabetologia, 2003, 46, 359–
364.
¨
8 L. Orning, S. Hammarstro¨m and B. Samuelsson, Leukotriene D: A
28 Y. Ikeda, J. Funjii and N. Taniguchi, Effects of substitutions of
the conserved histidine residues in human c-glutamyl transpeptidase,
J. Biochem., 1996, 119, 1166–1170.
slow reacting substance from rat basophilic leukemia cells, Proc. Natl.
Acad. Sci. U. S. A., 1980, 77, 2014–2017.
9 R. D. Allison, c-Glutamyl transpeptidase: Kinetics and Mechanism,
Methods Enzymol., 1985, 113, 419–437.
10 N. Taniguchi and Y. Ikeda, c-Glutamyl transpeptidase: Catalytic
mechanism and gene expression, Adv. Enzymol. Relat. Areas Mol. Biol.,
1998, 72, 239–278.
11 S. S. Tate and A. Meister, Interaction of c-glutamyl transpeptidase with
amino acids, dipeptides, and derivatives and analogs of glutathione,
J. Biol. Chem., 1974, 249, 7593–7602.
12 G. A. Thompson and A. Meister, Utilization of L-cystine by the
g-glutamyl transpeptidase-c-glutamylcyclotransferase pathway, Proc.
Natl. Acad. Sci. U. S. A., 1975, 72, 1985–1988.
13 G. A. Thompson and A. Meister, Interrelationships between the
binding sites for amino acids, dipeptides, and c-glutamyl donors in
c-glutamyl transpeptidase, J. Biol. Chem., 1977, 252, 6792–6798.
14 G. A. Thompson and A. Meister, Hydrolysis and transfer reactions
catalyzed by c-glutamyl transpeptidase; evidence for separate substrate
sites and for high affinity of L-cystine, Biochem. Biophys. Res. Commun.,
1976, 71, 32–36.
29 J. W. Keillor, R. Castonguay and C. Lherbet, c-Glutamyl transpeptidase
substrate specificity and catalytic mechanism, Methods Enzymol., 2005,
401, 449–467.
30 N. D. Cook and T. J. Peters, The effect of pH on transpeptidation
and hydrolytic reactions of rat kidney c-glutamyltransferase, Biochim.
Biophys. Acta, 1985, 832, 142–147.
31 N. D. Cook, K. P. Upperton, B. C. Challis and T. J. Peters,
The donor specificity and kinetics of the hydrolysis reaction of c-
glutamyltransferase, Biochim. Biophys. Acta, 1987, 914, 240–245.
32 A. Leblanc, C. Gravel, J. Labelle and J. W. Keillor, Kinetic studies
of guinea pig liver transglutaminase reveal a general base catalyzed
deacylation mechanism, Biochemistry, 2001, 40, 8335–8342.
33 V. Somayaji and R. S. Brown, Distorted amides as models for activated
=
peptide N–C O units produced during enzyme-catalyzed acyl transfer
reactions, J. Org. Chem., 1986, 51, 2676–2686.
34 M. L. Bender and R. J. Thomas, The concurrent alkaline hydrolysis
and isotopic oxygen exchange of a series of p-substituted acetanilides,
J. Am. Chem. Soc., 1961, 83, 4183–4189.
15 A. Me´nard, R. Castonguay, C. Lherbet, C. Rivard, Y. Roupioz and J. W.
Keillor, Nonlinear free energy relationship in general-acid-catalyzed
acylation of rat kidney c-glutamyl transpeptidase by a series of c-
glutamyl anilide substrate analogues, Biochemistry, 2001, 40, 12678–
12685.
16 R. Castonguay, C. Lherbet and J. W. Keillor, Kinetic studies of rat
kidney c-glutamyltranspeptidase deacylation reveal a general base-
catalyzed mechanism, Biochemistry, 2003, 42, 11504–11513.
17 J. W. Keillor, A. Me´nard, R. Castonguay, C. Lherbet and C. Rivard,
Pre-steady state kinetic studies of rat kidney c-glutamyl transpeptidase
confirm its ping-pong mechanism, J. Phys. Org. Chem., 2004, 17, 529–
536.
35 G. M. Blackburn and W. P. Jencks, The mechanism of aminolysis of
methyl formate, J. Am. Chem. Soc., 1968, 90, 2638–2645.
36 J. Urdis and A. Williams, Proteolytic enzymes: lipophilic binding sites
for specific and non-specific substrates of a-chymotrypsin, J. Chem.
Soc., Perkin Trans. 2, 1976, 686–690.
37 A. Williams, Chymotrypsin-catalyzed phenyl ester hydrolysis. Evidence
for electrophilic assistance on carbonyl oxygen, Biochemistry, 1970, 9,
3383–3390.
38 J. W. Keillor and R. S. Brown, Attack of zwitterionic thiolates on a
distorted anilide as a model for the acylation of papain by amides. A
simple demonstration of a bell-shaped pH–rate profile, J. Am. Chem.
Soc., 1992, 114, 7983–7989.
18 S. S. Tate and A. Meister, c-Glutamyl transpeptidase: catalytic,
structural and functional aspects, Mol. Cell. Biochem., 1981, 39, 357–
368.
19 S. S. Tate and A. Meister, c-Glutamyl transpeptidase from kidney,
Methods Enzymol., 1985, 113, 400–419.
20 N. D. Cook and T. J. Peters, A sensitive high-performance liquid
chromatography assay for c-glutamylhydrolase, Biochem. Soc. Trans.,
1985, 13, 1226–1227.
21 T. Okada, H. Suzuki, K. Wada, H. Kumagai and K. Fukuyama,
Crystal structures of c-glutamyltranspeptidase from Escherichia coli,
a key enzyme in glutathione metabolism, and its reaction intermediate,
Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 6471–6476.
39 J.-Y. Chang, R. Knecht and D. G. Braun, Amino acid analysis in the
picomole range by precolumn derivatization and high-performance
liquid chromatography, Methods Enzymol., 1983, 91, 41–48.
40 J.-K. Lin and J.-Y. Chang, Chromophoric labelling of amino acids with
4-dimethylaminoazobenzene-4ꢀ-sulfonyl chloride, Anal. Chem., 1975,
47, 1634–1638.
41 J.-Y. Chang, R. Knecht and D. G. Braun, A complete separation
of dimethylaminoazobenzenesulfonyl-amino acids, Biochem. J., 1982,
203, 803–806.
42 V. Stocchi, L. Cucchiarini, G. Piccoli and M. Magnani,
Complete high-performance liquid chromatographic separation
of 4-N,N-dimethylaminoazobenzene-4ꢀ-thiohydantoin and 4-
dimethylaminoazobenzen-4ꢀ-sulfonyl chloride amino acids utilizing
the same reversed-phase column at room temperature, J. Chromatogr.,
1985, 349, 77–82.
22 P. Gagnon, X. Huang, E. Therrien and J. W. Keillor, Peptide coupling of
unprotected amino acids through in situ p-nitrophenyl ester formation,
Tetrahedron Lett., 2002, 43, 7717–7719.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3790–3801 | 3801
©