Acylation is rate-limiting in glycosylasparaginase-catalyzed hydrolysis of N4-(4′-substituted phenyl)-L-asparagines

Article abstract of DOI:10.1039/b301513k

Glycosylasparaginase catalyzes the hydrolysis of the N-glycosylic bond between N-acetyl-D-glucosamine and L-asparagine in the catabolism of glycoproteins. The mechanism has been proposed to resemble that of serine proteases involving an acylation step where a nucleophilic attack by a catalytic Thr residue on the carbonyl carbon of the N-glycosylic bond gives rise to a covalent β-aspartyl-enzyme intermediate, and a deacylation step to give the final products. The question posed in this study was: Is the acylation step the rate-limiting step in the hydrolysis reaction as in serine proteases? To answer this question a series of mostly new substituted anilides was synthesized and characterized, and their hydrolysis reactions catalyzed by glycosylasparaginase from human amniotic fluid were studied. Five N⁴-(4′-substituted phenyl)-L-asparagine compounds were synthesized and characterized: 4′-hydrogen, 4′-ethyl, 4′-bromo, 4′-nitro, and 4′-methoxy. Each of these anilides was a substrate for the enzyme. Hammett plots of the kinetic parameters showed that acylation is the rate-limiting step in the reaction and that upon binding the electron distribution of the substrate is perturbed toward the transition state. This is the first direct evidence that acylation is the rate-limiting step in the enzyme-catalyzed reaction. A Bronsted plot indicates a small, negative charge (-0.25) on the nitrogen atom of the leaving group anilines containing electron-withdrawing groups, and a small, positive charge (0.43) on the nitrogen atom of the leaving group anilines containing electron-donating groups. The free energy (incremental) change of binding (ΔΔG_b) in the enzyme-substrate transition state complexes shows that substitution of a substituted phenyl group for the pyranosyl group in the natural substrate results in an overall loss of binding energy equivalent to a weak hydrogen bond, the magnitude of which is dependent on the substituent group. The data are consistent with a mechanism for glycosylasparaginase involving rapid formation of a tetrahedral structure upon substrate binding, and a rate-limiting breakdown of the tetrahedral structure to a covalent β-aspartyl-enzyme intermediate that is dependent on the electronic properties of the substituent group and on the degree of protonation of the leaving group in the transition state by a general acid.

Full text of DOI:10.1039/b301513k

View Online / Journal Homepage / Table of Contents for this issue
Acylation is rate-limiting in glycosylasparaginase-catalyzed
hydrolysis of N 4-(4Ј-substituted phenyl)-L-asparagines†‡
Wenjun Du and John M. Risley*
Department of Chemistry, The University of North Carolina at Charlotte,
9201 University City Blvd., Charlotte, North Carolina 28223-0001, USA.
E-mail: jmrisley@email.uncc.edu
Received 7th February 2003, Accepted 14th April 2003
First published as an Advance Article on the web 2nd May 2003
Glycosylasparaginase catalyzes the hydrolysis of the N-glycosylic bond between N-acetyl--glucosamine and
-asparagine in the catabolism of glycoproteins. The mechanism has been proposed to resemble that of serine
proteases involving an acylation step where a nucleophilic attack by a catalytic Thr residue on the carbonyl carbon
of the N-glycosylic bond gives rise to a covalent β-aspartyl–enzyme intermediate, and a deacylation step to give the
final products. The question posed in this study was: Is the acylation step the rate-limiting step in the hydrolysis
reaction as in serine proteases? To answer this question a series of mostly new substituted anilides was synthesized
and characterized, and their hydrolysis reactions catalyzed by glycosylasparaginase from human amniotic fluid were
studied. Five N 4-(4Ј-substituted phenyl)--asparagine compounds were synthesized and characterized: 4Ј-hydrogen,
4Ј-ethyl, 4Ј-bromo, 4Ј-nitro, and 4Ј-methoxy. Each of these anilides was a substrate for the enzyme. Hammett plots
of the kinetic parameters showed that acylation is the rate-limiting step in the reaction and that upon binding the
electron distribution of the substrate is perturbed toward the transition state. This is the first direct evidence that
acylation is the rate-limiting step in the enzyme-catalyzed reaction. A Brønsted plot indicates a small, negative
charge (Ϫ0.25) on the nitrogen atom of the leaving group anilines containing electron-withdrawing groups, and a
small, positive charge (0.43) on the nitrogen atom of the leaving group anilines containing electron-donating groups.
The free energy (incremental) change of binding (∆∆Gb) in the enzyme–substrate transition state complexes shows
that substitution of a substituted phenyl group for the pyranosyl group in the natural substrate results in an overall
loss of binding energy equivalent to a weak hydrogen bond, the magnitude of which is dependent on the substituent
group. The data are consistent with a mechanism for glycosylasparaginase involving rapid formation of a
tetrahedral structure upon substrate binding, and a rate-limiting breakdown of the tetrahedral structure to a
covalent β-aspartyl–enzyme intermediate that is dependent on the electronic properties of the substituent group
and on the degree of protonation of the leaving group in the transition state by a general acid.
Introduction
The predominant N-glycosylic1 bond in nature is between
N-acetyl--glucosamine and asparagine.2 In the catabolism of
N-linked carbohydrate moieties to their constituent molecules,
glycosylasparaginase (GA, aspartylglucosaminidase (AGA),
N 4-(β-N-acetyl--glucosaminyl)--asparaginase; EC 3.5.1.26)
catalyzes the hydrolysis of the N-glycosylic (amide) bond
in β-N-acetylglucosaminyl--asparagine [(GlcNAc-)Asn] to
give aspartic acid and 2-acetamido-2-deoxy-β--glucopyrano-
sylamine (Scheme 1), which hydrolyzes non-enzymatically to
N-acetyl--glucosamine (GlcNAc) and ammonium ion.3
A
deficiency, or absence, of GA activity gives rise to aspartyl-
glycosaminuria (AGU), an autosomal recessive inherited
lysosomal storage disorder that results in the accumulation of
glycoasparagines, particularly (GlcNAc-)Asn, in lysosomes;
AGU is the most common disorder of glycoprotein metabol-
ism.3 A few kinetic studies have been reported for the enzyme.4
The crystal structures for human GA5 and Flavobacterium
meningosepticum GA6,7 have been reported, and show that GA
belongs to the superfamily of N-terminal nucleophile (Ntn)
amidohydrolases.8 Based on some of the kinetic studies and the
crystal structure of human GA, a catalytic mechanism was
Scheme 1
proposed5 that is analogous to the mechanism for serine pro-
teases (Scheme 2): upon binding of (GlcNAc-)Asn, nucleophilic
attack by an N-terminal Thr leads to release of 2-acetamido-2-
deoxy-β--glucopyranosylamine with formation of a covalent
β-aspartyl-threonine ester intermediate, which undergoes
hydrolysis to give -Asp and to regenerate the enzyme. The
crystal structures for GA show that the active site is funnel-
shaped with binding sites for the α-carboxyl and α-amino
groups on Asn at the bottom of the funnel, and the catalytic
Thr is located near the top of the funnel. As a result, there is
quite a range of structures for the R-group at the N 4-nitrogen
of Asn which are substrates for GA (Scheme 2): R can be H
† Electronic supplementary information (ESI) available: Supplemen-
tary Figures 1–5. See http://www.rsc.org/suppdata/ob/b3/b301513k/
‡ In memoriam. Dedicated to the memory of Dr. Terry L. Kruger,
Professor of Chemistry, Ball State University (Muncie, Indiana),
whose amazing undergraduate organic chemistry lectures changed a
career direction, and who was an important mentor in learning to do
research (J. M. R.).
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
1900

View Online
been reported only as a racemic mixture.16 The general pro-
cedure used to synthesize the five compounds was based on a
published procedure9,13 and is shown in Fig. 1. The α-amino
group of -aspartic acid was protected as the N 2-phthalyl
derivative (Step 1), and the protected acid was converted to
N 2-phthalyl--aspartic anhydride (1) (Step 2); reaction of
the protected anhydride with the specific aniline (Step 3) and
deprotection with aqueous hydrazine (Step 4) gave the product.
The use of the phthalyl group as a protecting group has been
reported to result in formation only of the N 4-asparagine,9,13 as
well as the N 5-glutamines,19 presumably due to steric hindrance
that precludes formation of the N 1-amides. No attempt was
made to optimize yields for the reactions. The syntheses of 2, 3
and 4 required 12 h, 9 h and 24 h, respectively, for Step 3. The
properties of the compounds are consistent with the successful
synthesis of each compound. The assignments of the NMR
signals are based on the 1H and 13C NMR data for (GlcNAc-)-
1
Asn,20 the H NMR data for acetanilides,21 and the 13C NMR
data for acetanilides.22 Our data are in general agreement with
these data.
Scheme 2
(Asn is a substrate), various monosaccharides to oligo-
saccharides (high mannose, complex, hybrid structures),
methylcoumarin, and various amino acids.4 Therefore, there is
generally little substrate specificity for the amide functional
group by GA as long as the α-carboxyl and α-amino groups on
Asn are not blocked.4 In order to support the proposed catalytic
mechanism, fundamental studies of the mechanism are
required. Evidence thus far would indicate that, in the two-step
reaction of acylation and deacylation, the acylation step is
probably the rate-limiting step, but there is no direct evidence
for this. In this paper we present the first direct evidence that the
acylation step is the rate-limiting step in the hydrolysis reaction
catalyzed by glycosylasparaginase.
Fig. 1 The synthesis of N 4-(4Ј-substituted phenyl)--asparagines was
done in 4 steps. Reaction of -Asp with phthalic anhydride to protect
the α-amino group and reaction with acetic anhydride gave the phthalyl
-aspartic anhydride (1). Reaction of the latter with 4-substituted
anilines and deprotection of the α-amino group gave the desired
compounds, which were characterized.
Numerous attempts to synthesize 5 were not successful with
no product detected. The electron-withdrawing nitro group
significantly decreases the nucleophilicity of 4-nitroaniline.
With an idea to increase the electrophilicity of the β-carbonyl
group of the anhydride, and possibly increase the nucleo-
philicity of 4-nitroaniline, glacial acetic acid (11% v/v) was
added to Step 3 in the general procedure. This addition resulted
in the formation of 5 after reaction for 8 h. The properties of 5
are consistent with its successful synthesis. We did not pursue a
study of the mechanism, but formation of 4Ј-nitroacetanilide
was not detected, which would indicate that a mixed anhydride
was not formed in the reaction.
When 4-methoxyaniline was used in Step 3, the product after
deprotection with aqueous hydrazine was an approximately
equal mixture of 6 and N 1-(4Ј-methoxyphenyl)--asparagine
(7). This was evident from the 1H NMR and 13C NMR spectra
(Supplementary Figure 1) where two sets of signals were
observed. One set of signals was observed for each isomer in the
1H NMR spectrum, while two signals for C-1, C-3, and C-4
(in -asparagine) were observed for the two isomers in the 13C
NMR spectrum. This is the first observation that the phthalyl
group as a protecting group for the α-amino group does not
preclude formation of an N 1-amide. Presumably this arises
from the electron-donating ability of the methoxy group that
makes 4-methoxyaniline quite nucleophilic; the reaction
required only 4 h, the shortest time for any of the anilines
used in this study. However, the formation of the N 1-amide may
be limited to N 2-phthalyl--aspartic anhydride because the
Results and discussion
Synthesis of N 4-(4Ј-substituted phenyl)-L-asparagines
Surprisingly, very few N 4-(4Ј-substituted phenyl)--asparagines
have been reported in the literature, and those that have
been are incompletely characterized with reports only of
melting points, IR data for one compound, and two optical
rotation values for one compound. N 4-Phenyl--asparagine
has been reported,9–15 racemic mixtures were synthesized for
N 4-phenyl-, N 4-(2Ј-methylphenyl)-, N 4-(3Ј-methylphenyl)-, N 4-
(4Ј-methoxyphenyl)-, N 4-(2Ј-carboxyphenyl)-, and N 4-(4Ј-
carboxyphenyl)-,-asparagine,16 and N 4-(4Ј-nitrophenyl)--
asparagine is commercially available,17 although no literature
citation to its synthesis and properties was found. [N 2-Cbz-
N 4-phenyl-, N 2-Fmoc-N 4-phenyl-, and N 2-Boc-N 4-phenyl--
asparagine and N 2-Cbz-N 4-(4Ј-methylphenyl)--asparagine
have been reported.18] We synthesized N 4-phenyl--asparagine
(2), N 4-(4Ј-ethylphenyl)--asparagine (3), N 4-(4Ј-bromo-
phenyl)--asparagine (4), N 4-(4Ј-nitrophenyl)--asparagine (5),
and N 4-(4Ј-methoxyphenyl)--asparagine (6); 3 and 4 are new
compounds, 5 has only been reported commercially,17 and 6 has
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
1901

View Online
Table 1 Kinetic parameters and relevant data
kcat/sϪ1
Km/µM
(kcat/Km)/µMϪ1 sϪ1
σpϪ a
pKa
b
Substrate
(GlcNAc-)Asn
5.0 ( 0.9)
2.0 ( 0.2)
1.9 ( 0.1)
1.2 ( 0.15)
1.7 ( 0.1)
9.0 ( 0.9)
215 ( 27)
1714 ( 112)
804 ( 35)
463 ( 46)
295 ( 15)
603 ( 40)
0.023
na
na
4-(4Ј-MeOPh)--Asn
4-(4Ј-EtPh)--Asn
4-(4Ј-HPh)--Asn
4-(4Ј-BrPh)--Asn
4-(4Ј-NO2Ph)--Asn
0.0012
0.0023
0.0025
0.0057
0.015
Ϫ0.26
Ϫ0.19
0.00
0.25
1.27
5.31
5.11
4.58
3.88
1.02
a Values from reference 25. b Values from reference 26.
formation of an N 1-amide from reaction of N 2-phthalyl--
glutamic anhydride with 4-methoxyaniline was not reported.19
Further studies with other anilines containing strongly elec-
tron-donating groups are necessary to understand this present
result. Separation of the two isomers was achieved quite easily
by taking advantage of the limited solubilities of each isomer in
aqueous solution at their isoelectric points. The N 4-amide
(6) has a lower isoelectric point (estimated as 5.5) than the
N 1-amide (7) (estimated as 7.5). The difference between the
isoelectric points was sufficiently large that the two isomers
were separated by dissolving the mixture in aqueous solution at
pH 11, and, upon acidification to pH 9, 7 precipitated. Upon
further acidification to pH 3, 6 precipitated. The 1H NMR and
13C NMR spectra of 6 are shown in Supplementary Figure 2.
The properties of 6 are consistent with its successful synthesis
and separation from its isomer.
Ϫ
Fig. 2 A Hammett plot of log (kcat) vs. σp is biphasic. These data
show that acylation is the rate-limiting step in the reaction. (For
deacylation as the rate-limiting step, the plot would be linear with a
slope = 0.) For substrates containing electron-donating groups the slope
ρ is Ϫ0.94 (r = 0.99) and for substrates containing electron-withdrawing
groups the slope ρ is 0.705 (r = 1.00).
Reaction of N 4-(4Ј-substituted phenyl)-L-asparagines with
glycosylasparaginase
In the hydrolysis of the N 4-(4Ј-substituted phenyl)--aspara-
gines, the released substituted aniline was measured color-
imetrically by a diazotization reaction as previously described.19
Prior to the enzyme kinetic studies, the time and temperature
for color development were studied, the absorbance spectrum
was recorded to determine the wavelength for measurement,
and the molar absorptivity value was calculated from plots of
absorbance against concentration. The average of triplicate
measurements for 12–16 concentrations of each aniline gave
excellent correlation coefficients, r ≥ 0.9994. The values for each
aniline are given in the Experimental section. The values for
4-bromoaniline have not been reported. The other values are in
general agreement with those previously reported.19 The molar
absorptivity value for aniline is slightly smaller, while the values
for 4-ethylaniline and 4-methoxyaniline are larger; the value for
4-methoxyaniline was checked several times.23
reaction catalyzed by GA (and an N-terminal amidohydrolase),
and strongly implies that acylation is the rate-limiting step in
the reaction of GA with (GlcNAc-)Asn. Linear regression of
the data for substrates containing electron-donating groups
gives a line with a correlation coefficient r = 0.99 and a slope
ρ = Ϫ0.94. Linear regression of the data for substrates contain-
ing electron-withdrawing groups gives a line with a correlation
coefficient r = 1.00 and a slope ρ = 0.705. A Hammett plot of
Ϫ
log (Km) vs. σp (not shown) is also biphasic, similar to Fig. 2;
for substrates containing electron-donating groups the slope
ρ = Ϫ1.99, while for substrates containing electron-withdrawing
groups Km is within experimental error independent of the
structure of the aniline leaving group. A Hammett plot of log
(kcat/Km) vs. σpϪ (Supplementary Figure 3) is linear with a corre-
lation coefficient r = 0.95 and a slope ρ = 0.65. This second-
order rate constant for the reaction of substrate with free
enzyme is not affected by nonproductive binding effects. The
positive slope is indicative of a favorable effect on the binding
step with increasingly electron-withdrawing substituents. The
substituent effect on the binding step suggests that the electron
distribution of the substrate is perturbed toward the transition
state upon binding to the active site.27
Because the mechanism for GA has been proposed to be
analogous to serine proteases, we compared our results with the
most applicable results reported for rate-limiting acylation reac-
tions utilizing anilides for chymotrypsin28,29 and urokinase.30
A biphasic Hammett plot for log (kcat) vs. σ as in Fig. 2 was
suggested for chymotrypsin by Bundy and Moore.28 There is
an indication of a biphasic plot in the data of Inagami et al.29
similar to our observation in Fig. 2. Inagami et al.29 also
reported that a plot of log (Km) vs. σϪ was scattered, but with
the same tendency as the plot of log (kcat) vs. σϪ; the plot
shows a general trend similar to our observation where Km
is within experimental error independent for electron-
withdrawing groups. For urokinase,30 a plot of log (kcat) vs. σϪ
for substrates containing electron-withdrawing groups was
linear with a slope ρ = 0.72 that is nearly identical to our result
In an early study it was reported that N 4-phenyl--asparagine
was not a substrate for GA from hog serum and hog kidney, but
was a competitive inhibitor.12 Recently N 4-(4Ј-nitrophenyl)--
asparagine (5) was shown to be a substrate for GA from
recombinant Flavobacterium meningosepticum.24 We found that
each of the five anilides was a substrate for GA from human
amniotic fluid. The kinetic parameters are given in Table 1; the
kinetic parameters for N 4-(4Ј-nitrophenyl)--asparagine (5) are
similar for GA from human amniotic fluid (Km 603 µM, kcat
9.0 sϪ1) and GA from recombinant Flavobacterium meningo-
septicum (Km 430 µM, kcat 2.3 sϪ1).24 No correlations between
the kinetic parameters and field/inductive (σI) constants, reson-
ance (σR) constants, Taft steric parameters, and hydrophobic
parameters25 were found.
A Hammett plot of log (kcat) vs. σpϪ (Fig. 2) is biphasic. The
value of kcat varies with substitution on the aniline leaving
group. This indicates that acylation is the rate-limiting step
in the hydrolysis of these substrates.27 If deacylation was the
rate-limiting step, the value of kcat would be independent of
the structure of the aniline leaving group with formation of a
β-aspartyl–enzyme intermediate. This is the first direct evi-
dence for a rate-limiting acylation step in the hydrolysis
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
1902

View Online
(Fig. 2); no correlation was evident for urokinase in plots of log
(Km) vs. σϪ or log (kcat/Km) vs. σϪ. Therefore, while the kinetic
data for GA show some similarities to Hammett correlations
in serine proteases, the overall relationship is not particularly
strong except for the biphasic relationship of log (kcat).
therefore the rate-limiting step in this hydrolysis reaction.
These results are in agreement with data for chymotrypsin that
support a rate-limiting breakdown of the tetrahedral structure
for anilide substrates to the acyl enzyme.32,33 Additional studies
will clarify further aspects of the mechanism.
The charge on the nitrogen atom of the aniline leaving group
at the transition state may be calculated from the slope of a
Brønsted plot.27 A plot of log (kcatX/kcatH) vs. pKa for the conju-
gate acid of the leaving group, i.e. the anilinium ion, is shown in
Supplementary Figure 4; the plot is biphasic as it must be
according to Fig. 2. The charge (slope) on the nitrogen atom in
Experimental
Materials and methods
Chemicals purchased from the following suppliers were:
phthalyl anhydride, aniline, 4-ethylaniline, 4-bromoaniline, 4-
nitroaniline, 4-methoxyaniline, and N-(1-naphthyl)ethylene-
diamine dihydrochloride from Acros; -aspartic acid from
Spectrum Chemical; deuterium oxide (99.9 atom% 2H), di-
methyl sulfoxide-d6 (99.9 atom% 2H), and acetone-d6 (99.9
the anilines containing electron-withdrawing groups is βlg
=
Ϫ0.25 (correlation coefficient r = 1.00 for the linear regression)
and on the nitrogen atom in the anilines containing electron-
donating groups is βlg = 0.43 (correlation coefficient r = 0.98 for
the linear regression). These data are similar in magnitude to
βlg ≈ 0.8 (calculated)29 and βlg = 1.27 for electron-donating
groups31 in the reaction of chymotrypsin, and βlg ≈ Ϫ0.3 (calcu-
lated)30 for electron-withdrawing groups in the reaction of
urokinase.
2
atom% H) from Cambridge Isotopes; sodium lauryl sulfate,
Concanavalin A-Sephrose 4B, Sephadex G-100, and bovine
serum albumin (fraction V) from Sigma; N 4-(2-acetamido-2-
deoxy-β--glucopyranosyl)--asparagine from Bachem (Cali-
fornia); 4-dimethylaminobenzaldehyde from Mallinckrodt. All
other chemicals were at least analytical grade. Human amniotic
fluid was obtained from the Chemistry Laboratory of the
Carolinas Medical Center (Charlotte, North Carolina), and
was stored at Ϫ20 ЊC.
The free energy (incremental) change of binding of the sub-
stituted anilides relative to the natural substrate in enzyme–
substrate transition state complexes, ∆∆Gb,4 was calculated at
37 ЊC from ∆∆Gb = ϪRTln [(kcat/Km)aniline/(kcat/Km)(GlcNAc-)Asn
]
and are plotted against Hammett constants (Supplementary
Figure 5); the plot is linear as it must be according to Sup-
plementary Figure 3, with a correlation coefficient r = 0.95 and
a slope ρ = Ϫ3.845. Substitution of the pyranosyl group in the
natural substrate with a substituted phenyl group results in an
overall loss in transition state binding energy between the
enzyme and substrate equivalent to a weak hydrogen bond.
Interestingly, the magnitude of the loss in binding energy
depends on the substituent group on the phenyl ring. For
electron-donating groups the loss of binding energy is greater
than for electron-withdrawing groups as shown in Supplemen-
tary Figure 5. Therefore, in terms of transition state binding
energy, these data would indicate that the greater the electron-
withdrawing potential on the phenyl ring, the closer the struc-
ture of the substituted anilide in the transition state might
resemble the natural substrate. The perturbation of electron
distribution of the substrates toward the transition state upon
binding indicated in Supplementary Figure 3 is supported by
the data in Supplementary Figure 5.
These data are consistent with a mechanism for GA
similar to that proposed for the transpeptidation reaction of
γ-glutamyl transpeptidase based on the electronic properties
of the substituent groups and on the degree of protonation of
the leaving group in the transition state.19 For GA (Scheme 2),
acylation is the rate-limiting step in the hydrolysis reaction.
Nucleophilic attack on the bound substituted anilide by the
N-terminal Thr residue is very fast with initial formation of a
high energy tetrahedral structure. The binding of this tetra-
hedral structure perturbs the electron distribution of the
substrate toward the transition state. The breakdown of the
tetrahedral structure is general acid-catalyzed where proton
transfer to the nitrogen of the leaving group aniline occurs
simultaneously with C–N bond cleavage. The α-amino group
on the catalytic Thr residue (i.e., the N-terminal group
of the subunit containing the catalytic Thr residue) has been
proposed to act as a general acid.5 The degree of proton
transfer and the degree of C–N bond cleavage depend on the
electronic properties of the substituent group on the aniline.
Electron-withdrawing groups favor C–N bond cleavage which
results in significant bond stretching in the transition state, and
make the nitrogen atom less basic which results in less proton
transfer in the transition state. Electron-donating groups favor
C–N bond cleavage which results in less bond stretching in the
transition state, and make the nitrogen atom more basic which
results in greater proton transfer in the transition state. The
formation of the covalent β-aspartyl–enzyme intermediate is
A GE 300 spectrometer was used to record NMR spectra. 1H
NMR spectra were recorded at 300.2 MHz in a 5 mm probe at
ambient temperature with a 5000 Hz sweep width, 30Њ pulse
angle, and an 8k (real) data block; no line-broadening factor
was applied to the accumulated FID. Natural abundance 13C
NMR spectra were recorded at 75.5 MHz in a 5 mm probe at
ambient temperature with a 20000 Hz sweep width, 30Њ pulse
angle, and an 8k (real) data block; protons were broad-band
decoupled and a line-broadening factor of 2.0 Hz was applied
to the accumulated FID. The error in the measured chemical
1
shifts was 0.002 ppm for H NMR and 0.033 ppm for 13C
NMR; J values are given in Hz and the error in the measured
coupling constants was 0.61. The (CD3)2SO–NaOD solvent
was prepared by adding a 1 M NaOD–D2O solution to di-
methyl sulfoxide-d6 to give a ratio of (CD3)2SO to NaOD–D2O
of 95:5 (v/v). Elemental analyses were done at Atlanta
Microlabs, Inc. (Norcross, Georgia).
N 2-Phthalyl-L-aspartic anhydride 1. The synthesis of 1 was
carried out as described with slight modification.13 Phthalic
anhydride (37.0 g, 250 mmol) was added over 5 min to a stirred
suspension of -aspartic acid (33.3 g, 250 mmol) in pyridine
(400 ml) at room temperature. The mixture was heated under
reflux for 4 h, after which the solution was cooled to room
temperature. Any solid material was removed by filtration. The
filtrate was evaporated under reduced pressure to give a light
yellow oil. Acetic anhydride (200 ml) was added to the oil. After
stirring the mixture for 1 h at room temperature, the solution
was cooled at 4 ЊC for 2 h. The resulting precipitate was
collected by filtration, washed with anhydrous diethyl ether
(3 × 50 ml), and recrystallized from 1,4-dioxane (300 ml) to give
1 (25.5 g, 42%) as white crystals. Mp 224–226 ЊC (from 1,4-
dioxane) (lit.,13 224–226 ЊC); δH(300.2 MHz; CD3COCD3;
acetone at 2.05 ppm) 3.45 (1H, dd, J2,3a 6.10 and J3a,3b Ϫ18.92,
3a-H), 3.65 (1H, dd, J2,3b 9.77 and J3a,3b Ϫ18.92, 3b-H), 5.80
(1H, dd, J2,3a 6.10 and J2,3b 9.77, 2-H), 7.95 (4H, s, Ph) (lit.,13
δH((CD3)2SO) 3.53 (2H, d, 3-H), 5.77 (1H, t, 2-H), 8.00 (4H, s,
Ph)); δC(75.5 MHz; (CD3)2SO–NaOD; dimethyl sulfoxide at
39.50 ppm) 33.36 (C-3), 47.28 (C-2), 123.75 (2 × C-3Ј), 131.07
(2 × C-2Ј), 135.11 (2 × C-4Ј), 166.57 (2 × C-1Ј), 169.33 (C-1),
169.95 (C-4).
N4-(4Ј-Substituted phenyl)-L-asparagines. The synthesis of 2,
3, 4, 5, and 6 was a modified procedure of those reported.13,19
To a suspension of 1 (3.0 g, 12.2 mmol) in anhydrous diethyl
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
1903

View Online
ether (80 ml) was added the (substituted)aniline (12.2 mmol).
The solution was stirred at room temperature (for the time
given below), after which the solvent was evaporated under
reduced pressure. The solid was dissolved in 0.1 M NaOH
(10 ml), the solution was acidified with 5 M HCl to pH 2.0,
and the resulting precipitate was collected and dissolved in
methanol (100 ml). Any insoluble material was removed by
filtration. Hydrazine (10 ml) was added and the mixture
was stirred at room temperature for 24 h. After the solvent
and excess hydrazine were evaporated under reduced pressure,
0.5 M HCl (150 ml) was added and the pH was adjusted to
1.0 with 5 M HCl. The resulting precipitate was removed
by filtration. To the filtrate was added solid NaHCO3 to
bring the pH to 5.0, and the solution was set at 4 ЊC overnight.
The precipitate was collected by filtration and washed with
acetone.
J2,3a 6.40 and J2,3b 6.40, 2-H), 7.83 (2H, d, J2Ј,3Ј 8.55, 2 × 3Ј-H),
8.17 (2H, d, J2Ј,3Ј 8.55, 2 × 2Ј-H); δC(75.5 MHz; (CD3)2SO–
NaOD; dimethyl sulfoxide at 39.50 ppm) 38.24 (C-3), 50.30
(C-2), 118.71 (2 × C-2Ј), 124.85 (2 × C-3Ј), 141.96 (C-4Ј), 145.42
(C-1Ј), 169.49 (C-4), 169.97 (C-1).
N4-(4Ј-Methoxyphenyl)-L-asparagine 6. 4-Methoxyaniline
(1.5 g, 12.2 mmol) was added and the reaction was stirred for
4 h to give an approximate equimolar mixture of 6 and N 1-(4Ј-
methoxyphenyl)--asparagine (7) (0.96 g) as a white powder.
The powder was dissolved in water (20 ml) and the solution was
adjusted to pH 11 with 6 M NaOH. The solution was adjusted
to pH 9.0 with 5 M HCl, at which point 7 precipitated and it
was collected by filtration. The filtrate was adjusted to pH 3.0
with 5 M HCl, at which point 6 precipitated. The precipitate
was collected by filtration and dried in an oven at 60 ЊC (2 h) to
give 6 (0.36 g, 12%) as a white powder. Mp 225–226 ЊC (lit.,16
265 ЊC (, mixture)) (Found: C, 55.19; H, 6.09; N, 11.64. Calc.
for C11H14N2O4: C, 55.46; H, 5.88; N, 11.76%); δH(300.2 MHz;
(CD3)2SO–NaOD; dimethyl sulfoxide at 2.50 ppm) 2.03 (1H,
dd, J2,3a 8.55 and J3a,3b Ϫ14.60, 3a-H), 2.31 (1H, dd, J2,3b 4.27
and J3a,3b Ϫ14.60, 3b-H), 3.42 (1H, dd, J2,3a 8.55 and J2,3b 4.27,
2-H), 3.68 (3H, s, 5Ј-H), 6.71 (2H, d, J2Ј,3Ј 9.15, 2 × 2Ј-H), 7.39
(2H, d, J2Ј,3Ј 9.15, 2 × 3Ј-H); δC(75.5 MHz; (CD3)2SO–NaOD;
dimethyl sulfoxide at 39.50 ppm) 44.80 (C-3), 54.24 (C-2), 55.18
(C-5Ј), 113.43 (2 × C-3Ј), 121.94 (2 × C-2Ј), 138.21 (C-1Ј),
153.67 (C-4Ј), 174.40 (C-4), 176.41 (C-1).
N4-Phenyl-L-asparagine 2. Aniline (1.1 g, 12.2 mmol) was
added and the reaction was stirred for 12 h to give 2 (0.81 g,
33%) as a white powder. Mp 247–248 ЊC (lit.,9 251–252 ЊC; lit.,10
250–252 ЊC; lit.,11 241–242 ЊC; lit.,13 248–249 ЊC; lit.,16 255 ЊC
(, mixture)) (Found: C, 57.75; H, 5.87; N, 13.48. Calc. for
C10H12N2O3: C, 57.69; H, 5.77; N, 13.46%); δH(300.2 MHz;
(CD3)2SO–NaOD; dimethyl sulfoxide at 2.50 ppm) 2.17
(1H, dd, J2,3a 7.33 and J3a,3b Ϫ15.80, 3a-H), 2.36 (1H, dd,
J2,3b 5.50 and J3a,3b Ϫ15.80, 3b-H), 3.56 (1H, dd, J2,3a 7.33
and J2,3b 5.50, 2-H), 7.00 (1H, t, J3Ј,4Ј 7.93, 4Ј-H), 7.26 (2H, dd,
J2Ј,3Ј 7.93 and J3Ј,4Ј 7.93, 2 × 3Ј-H), 7.57 (2H, d, J2Ј,3Ј 7.93, 2 ×
2Ј-H); δC(75.5 MHz; (CD3)2SO–NaOD; dimethyl sulfoxide at
39.50 ppm) 43.80 (C-3), 53.60 (C-2), 119.45 (2 × C-2Ј), 123.40
(C-4Ј), 129.03 (2 × C-3Ј), 139.41 (C-1Ј), 174.69 (C-4), 175.92
(C-1).
Kinetic studies
Human glycosylasparaginase (dimeric form) was isolated from
human amniotic fluid as described earlier4,34 utilizing the
exceptional resistance to sodium dodecyl sulfate (SDS):35
solubilization and centrifugation of amniotic fluid, ammonium
sulfate precipitation, 3% SDS–35% ammonium sulfate, con-
canavalin A affinity chromatography, gel-exclusion chromato-
graphy. To inhibit any protease activity present in the amniotic
fluid, disodium ethylenediaminetetraacetic acid (EDTA) and
phenylmethylsulfonyl fluoride (PMSF), as well as sodium azide,
were included in the solubilizing buffer; these chemicals have no
affect on GA activity. Only GA activity was measured in the
protein sample at the end; we are also not aware of another
enzyme capable of catalyzing the hydrolysis of N 4-(4Ј-substi-
tuted phenyl)--asparagines. GA activity for (GlcNAc-)Asn
was assayed overnight in 0.09 M citrate–0.06 M phosphate
buffer at pH 5.6, 37 ЊC, where optimal activity was observed,
and the N-acetyl--glucosamine released during the reaction
was measured using the Morgan–Elson reaction.36 Protein con-
centration was measured with the Coomassie blue (Bradford)
protein assay, using bovine serum albumin as the standard.37
The kinetic parameters used for the natural substrate, (Glc-
N4-(4Ј-Ethylphenyl)-L-asparagine 3. 4-Ethylaniline (1.5 g,
12.2 mmol) was added and the reaction was stirred for 9 h to
give 3 (0.62 g, 22%) as a white powder. Mp 225–226 ЊC (Found:
C, 60.84; H, 6.79; N, 11.77. Calc. for C12H16N2O3: C, 61.02; H,
6.78; N, 11.86%); δH(300.2 MHz; (CD3)2SO–NaOD; dimethyl
sulfoxide at 2.50 ppm) 1.10 (3H, t, J5Ј,6Ј 7.93, 6Ј-H), 2.04 (1H,
dd, J2,3a 9.76 and J3a,3b Ϫ14.60, 3a-H), 2.34 (1H, dd, J2,3b 4.27
and J3a,3b Ϫ14.60, 3b-H), 2.48 (2H, q, J5Ј,6Ј 7.93, 5Ј-H), 3.45 (1H,
dd, J2,3a 9.76 and J2,3b 4.27, 2-H), 6.95 (2H, d, J2Ј,3Ј 8.00, 2 × 2Ј-
H), 7.30 (2H, d, J2Ј,3Ј 8.00, 2 × 3Ј-H); δC(75.5 MHz; (CD3)2SO–
NaOD; dimethyl sulfoxide at 39.50 ppm) 16.32 (C-6Ј),
28.03 (C-5Ј), 45.07 (C-3), 54.78 (C-2), 122.11 (2 × C-2Ј), 127.52
(2 × C-3Ј), 135.93 (C-4Ј), 144.69 (C-1Ј), 175.13 (C-4), 177.62
(C-1).
N4-(4Ј-Bromophenyl)-L-asparagine 4. 4-Bromoaniline (2.1 g,
12.2 mmol) was added and the reaction was stirred for 24 h to
give 4 (0.95 g, 27%) as a white powder. Mp 250–251 ЊC (Found:
C, 41.87; H, 3.80; Br, 27.82; N, 9.75. Calc. for C10H11BrN2O3: C,
41.83; H, 3.83; Br, 27.85; N, 9.76%); δH(300.2 MHz; (CD3)2SO–
NaOD; dimethyl sulfoxide at 2.50 ppm) 1.95 (1H, dd, J2,3a 10.37
and J3a,3b Ϫ14.04, 3a-H), 2.37 (1H, dd, J2,3b 3.05 and J3a,3b
Ϫ14.04, 3b-H), 3.40 (1H, dd, J2,3a 10.37 and J2,3b 3.05, 2-H),
7.07 (2H, d, J2Ј,3Ј 8.55, 2 × 3Ј-H), 7.37 (2H, d, J2Ј,3Ј 8.55, 2 × 2Ј-
H); δC(75.5 MHz; (CD3)2SO–NaOD; dimethyl sulfoxide at
39.50 ppm) 45.81 (C-3), 55.64 (C-2), 108.62 (C-4Ј), 125.86
(2 × C-2Ј), 129.93 (2 × C-3Ј), 152.77 (C-1Ј), 176.89 (C-4), 178.90
(C-1).
NAc-)Asn, were Km 215 ( 27) µM and kcat 5.0 ( 0.9) sϪ1
.
The concentrations of the N 4-(4Ј-substituted phenyl)--
asparagines in 0.09 M citrate–0.06 M phosphate buffer at pH
5.6, 37 ЊC ranged from approx. 0.1 to 7 times Km, and a single
quantity of enzyme was used in each assay. Following overnight
incubation, the reaction was quenched with 40% trichloroacetic
acid. The release of the substituted aniline was measured color-
imetrically by a diazotization reaction as previously described.19
The time and temperature for color development, wavelength
for measurement of the product, and the molar absorptivity
value calculated by a linear regression analysis (with a corre-
lation coefficient, r ≥ 0.9994) from triplicate measurements for
each aniline were: 4-nitroaniline 10 min, 25 ЊC, 548 nm, 55.7
mMϪ1 cmϪ1; 4-bromoaniline 10 min, 25 ЊC, 562 nm, 44.45 mMϪ1
cmϪ1; aniline 15 min, 25 ЊC, 558 nm, 38.9 mMϪ1 cmϪ1; 4-ethyl-
aniline 60 min, 25 ЊC, 572 nm, 42.6 mMϪ1 cmϪ1; 4-methoxy-
aniline 60 min, 37 ЊC, 588 nm, 31.9 mMϪ1 cmϪ1. The average
of triplicate measurements was used to calculate the kinetic
parameters by regression analysis.
N4-(4Ј-Nitrophenyl)-L-asparagine 5. 4-Nitroaniline (1.7 g,
12.2 mmol) and glacial acetic acid (10 ml) were added, and the
reaction was stirred for 8 h to give 5 (0.31 g, 10%) as a slightly
yellow powder. Mp 244–245 ЊC (Found: C, 44.73; H, 4.75; N,
15.35. Calc. for C10H11N3O5ؒ0.9H2O: C, 44.52; H, 4.76; N,
15.58%); δH(300.2 MHz; (CD3)2SO–NaOD; dimethyl sulfoxide
at 2.50 ppm) 2.69 (1H, dd, J2,3a 6.40 and J3a,3b Ϫ16.50, 3a-H),
3.05 (1H, dd, J2,3b 6.40 and J3a,3b Ϫ16.50, 3b-H), 3.68 (1H, dd,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
1904

View Online
19 A. Menard, R. Castonguay, C. Lherbet, C. Rivard, Y. Roupioz and
Acknowledgements
J. W. Keillor, Biochemistry, 2001, 40, 12678–12685.
20 J. F. G. Vliegenthart, L. Dorland and H. van Halbeek,
Adv. Carbohydr. Chem. Biochem., 1983, 41, 209–374.
21 J. Bennett, M. Delmas and J. C. Maire, Org. Magn. Reson., 1969, 1,
319–327.
This research was supported, in part, by a Thomas D. Walsh
Fellowship (to W. D.) and by funds provided by the University
of North Carolina at Charlotte. The authors thank the
Chemistry Laboratory of the Carolinas Medical Center
(Charlotte, North Carolina) for the human amniotic fluid.
22 C. J. O’Connor, D. J. McLennan, D. J. Calvert, T. D. Lomax,
A. J. Porter and D. A. Rogers, Aust. J. Chem., 1984, 37, 497–510.
23 Unlike the other four anilines, color development for 4-methoxy-
aniline increased up to at least 72 h at 37 ЊC. We found that plots of
absorbance vs. concentration at any time, e.g., after 60 min, 120 min,
180 min, etc., were linear with excellent correlation coefficients.
Therefore, it was important to measure the absorbance immediately
at a fixed incubation time. We chose a 60 min incubation time that
may account for the major difference between the molar absorptivity
value reported earlier for a 30 min incubation19 and our value. It is
not clear that the same phenomenon may account for the differences
in molar absorptivity values for 4-ethylaniline, which we did not
pursue experimentally. Whether this phenomenon may be true for
anilines containing electron-donating groups, in general, is also not
clear from these data.
Notes and references
1 The IUPAC-IUBMB Joint Commission on Biochemical
Nomenclature: Nomenclature of Carbohydrates (Recommend-
ations 1996) (www.chem.qmul.ac.uk/iupac/2carb/) states that the
term “N-glycoside” is improper and does not recommend its use, but
“glycosyl” is recommended. Therefore, “glycosylic” is used in this
paper, rather than “glycosidic”.
2 C. M. Smith, Tetrahedron, 1998, 54, 11317–11362.
3 N. N. Aronson, Jr., Biochim. Biophys. Acta, 1999, 1455, 139–154.
4 J. M. Risley, D. H. Huang, J. J. Kaylor, J. J. Malik, Y.-Q. Xia and
W. M. York, Arch. Biochem. Biophys., 2001, 391, 165–170.
5 C. Oinonen, R. Tikkanen, J. Rouvinen and L. Peltonen,
Nature Struct. Biol., 1995, 2, 1102–1108.
24 Y. Liu, C. Guan and N. N. Aronson, Jr., J. Biol. Chem., 1998, 273,
9688–9694.
25 C. Hansch and A. Leo, Substituent Constants for Correlation
Analysis in Chemistry and Biology, John Wiley & Sons, New York,
1979.
26 X. Huang, N. Day, X. Luo, Y. Roupioz, M. Seid and J. W. Keillor,
J. Pept. Res., 1999, 53, 126–133.
6 J. Xuan, A. L. Tarentino, B. G. Grimwood, T. H. Plummer, Jr.,
T. Cui, C. Guan and P. Van Roey, Protein Sci., 1998, 7, 774–781.
7 H.-C. Guo, Q. Xu, D. Buckley and C. Guan, J. Biol. Chem., 1998,
273, 20205–20212.
8 C. Oinonen and J. Rouvinen, Protein Sci., 2000, 9, 2329–2337.
9 F. E. King and D. A. A. Kidd, J. Chem. Soc., 1951, 2976–2978.
10 F. Weyand, P. Klinke and I. Eigen, Chem. Ber., 1957, 90, 1896–1905.
11 J. Kovacs, H. Nagy Kovacs, I. Konyves, J. Csaszar, T. Vajda and
H. Mix, J. Org. Chem., 1961, 26, 1084–1091.
12 M. Tanaka and I. Yamashina, Carbohydr. Res., 1973, 27, 175–183.
13 M. J. Murphy and J. F. Stubbins, J. Pharm. Sci., 1980, 69, 553–555.
14 W. E. Adams, E. W. Snook and T. J. Curphey, J. Labelled Compd.
Radiopharm., 1980, 18, 991–997.
27 W. P. Jencks, Cold Spring Harbor Sym. Quant. Biol., 1971, 36,
1–11.
28 H. F. Bundy and C. L. Moore, Biochemistry, 1966, 5, 808–811.
29 T. Inagami, A. Patchornik and S. S. York, J. Biochem., 1969, 65,
809–819.
30 D. Petkov, E. Christova, I. Pojarlieff and N. Stambolieva,
Eur. J. Biochem., 1975, 51, 25–32.
31 L. Parker and J. H. Wang, J. Biol. Chem., 1968, 243, 3729–3734.
32 M. Caplow, J. Am. Chem. Soc., 1969, 91, 3639–3645.
33 E. C. Lucas and M. Caplow, J. Am. Chem. Soc., 1972, 94, 960–963.
34 J. M. Risley, D. H. Huang, J. J. Kaylor, J. J. Malik and Y.-Q. Xia,
J. Enzyme Inhib., 2001, 16, 269–274.
15 D. Nisato, J. Wagnon, G. Callet, D. Mettefeu, J.-L. Assens,
C. Plouzane, B. Tonnerre, V. Pliska and J.-L. Fauchere, J. Med.
Chem., 1987, 30, 2287–2291.
16 Y. Liwschitz, A. Zilkha and E. Nemes, Bull. Res. Counc. Isr., Sect. A,
1961, 10, 107–110.
35 T. Heiskanen, O. K. Tollersrud, M. Zhao and L. Peltonen, Protein
Expression Purif., 1994, 5, 205–210.
36 J. L. Reissig, J. L. Strominger and L. F. Leloir, J. Biol. Chem., 1955,
217, 959–966.
37 C. M. Stoscheck, Methods Enzymol., 1990, 182, 50–68.
17 BACHEM, #L-1690.
18 X. Huang, X. Luo, Y. Roupioz and J. W. Keillor, J. Org. Chem.,
1997, 62, 8821–8825.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 1 9 0 0 – 1 9 0 5
1905