Enzymatic characterization of O-GlcNAcase isoforms using a fluorogenic GlcNAc substrate

Article abstract of DOI:10.1016/j.carres.2006.03.004

A highly sensitive fluorogenic hexosaminidase substrate, fluorescein di(N-acetyl-β-d-glucosaminide) (FDGlcNAc), was prepared essentially as described previously [Chem. Pharm. Bull. 1993, 41, 314] with some modifications. The fluorescent analog is a substrate for a number of hexosaminidases but here we have focused on the cytoplasmic O-GlcNAcase isoforms. Kinetic analysis using purified O-GlcNAcase and its splice variant (v-O-GlcNAcase) expressed in Escherichia coli suggests that FDGlcNAc is a much more efficient substrate (K_m = 84.9 μM) than the conventional substrate, para-nitrophenyl 2-acetamido-2-deoxy-β-d-glucopyranoside (pNP-β-GlcNAc, K_m = 1.1 mM) and a previously developed fluorogenic substrate, 4-methylumbelliferyl 2-acetamido-2-deoxy-β-d-glucopyranoside [MUGlcNAc, K_m = 0.43 mM; J. Biol. Chem. 2005, 280, 25313] for O-GlcNAcase. The variant O-GlcNAcase, a protein lacking the C-terminal third of the full-length O-GlcNAcase, exhibited a K_m of 2.1 mM with respect to FDGlcNAc. This shorter isoform was not previously thought to exhibit O-GlcNAcase activity based on in vitro studies with pNP-β-GlcNAc. However, both O-GlcNAcase isoforms reduced O-GlcNAc protein levels extracted from HeLa and HT-29 cells in vitro, indicating that the splice variant is a bona fide O-GlcNAcase. Fluorescein di-N-acetyl-β-d-galactosaminide (FDGalNAc) is not cleaved by these enzymes, consistent with previous findings that the O-GlcNAcase has substrate specificity toward O-GlcNAc but not O-GalNAc. The enzymatic activity of the shorter isoform of O-GlcNAcase was first detected by using highly sensitive fluorogenic FDGlcNAc substrate. The finding that O-GlcNAcase exists as two distinct isoforms has a number of important implications for the role of O-GlcNAcase in hexosamine signaling.

Full text of DOI:10.1016/j.carres.2006.03.004

Carbohydrate Research 341 (2006) 971–982
Enzymatic characterization of O-GlcNAcase isoforms using a
fluorogenic GlcNAc substrate
a,
b
a
a
*
Eun Ju Kim, Dae Ook Kang, Dona C. Love and John A. Hanover
a
Laboratory of Cell Biochemistry and Biology, NIDDK, National Institutes of Health, MD 20892, USA
Department of Health Science and Biochemistry, College of Natural Sciences, Changwon National University,
Sarim-Dong, Changwon, Kyungnam 641-773, South Korea
b
9
Received 28 October 2005; received in revised form 17 February 2006; accepted 2 March 2006
Available online 11 April 2006
Abstract—A highly sensitive fluorogenic hexosaminidase substrate, fluorescein di(N-acetyl-b-D-glucosaminide) (FDGlcNAc), was
prepared essentially as described previously [Chem. Pharm. Bull. 1993, 41, 314] with some modifications. The fluorescent analog
is a substrate for a number of hexosaminidases but here we have focused on the cytoplasmic O-GlcNAcase isoforms. Kinetic anal-
ysis using purified O-GlcNAcase and its splice variant (v-O-GlcNAcase) expressed in Escherichia coli suggests that FDGlcNAc is a
much more efficient substrate (K
pyranoside (pNP-b-GlcNAc, Km = 1.1 mM) and a previously developed fluorogenic substrate, 4-methylumbelliferyl 2-acetamido-
-deoxy-b-D-glucopyranoside [MUGlcNAc, K = 0.43 mM; J. Biol. Chem. 2005, 280, 25313] for O-GlcNAcase. The variant O-
GlcNAcase, a protein lacking the C-terminal third of the full-length O-GlcNAcase, exhibited a K of 2.1 mM with respect to
m
= 84.9 lM) than the conventional substrate, para-nitrophenyl 2-acetamido-2-deoxy-b-D-gluco-
2
m
m
FDGlcNAc. This shorter isoform was not previously thought to exhibit O-GlcNAcase activity based on in vitro studies with
pNP-b-GlcNAc. However, both O-GlcNAcase isoforms reduced O-GlcNAc protein levels extracted from HeLa and HT-29 cells
in vitro, indicating that the splice variant is a bona fide O-GlcNAcase. Fluorescein di-N-acetyl-b-D-galactosaminide (FDGalNAc)
is not cleaved by these enzymes, consistent with previous findings that the O-GlcNAcase has substrate specificity toward O-GlcNAc
but not O-GalNAc. The enzymatic activity of the shorter isoform of O-GlcNAcase was first detected by using highly sensitive flu-
orogenic FDGlcNAc substrate. The finding that O-GlcNAcase exists as two distinct isoforms has a number of important implica-
tions for the role of O-GlcNAcase in hexosamine signaling.
Published by Elsevier Ltd.
Keywords: O-GlcNAc; OGT; O-GlcNAcase; v-O-GlcNAcase; FDGlcNAc; MUGlcNAc; pNP-b-GlcNAc
Abbreviations: FDGlcNAc, fluorescein di-N-acetyl-b-D-glucosaminide;
1
. Introduction
O-GlcNAcase, a full-length form of peptide O-GlcNAc-b-N-acetyl-
glucosaminidase; v-O-GlcNAcase, shorter isoform of peptide
a
Many nuclear and cytoplasmic proteins such as tran-
scription factors, cytoskeletal proteins, oncogenes,
kinases, phosphatases, and chaperones,
O-GlcNAc-b-N-acetylglucosaminidase; PNP-, para-nitrophenyl-; O-
GlcNAc, O-linked b-N-acetylglucosamine; FDGalNAc, fluorescein
di-N-acetyl-b-D-galactosaminide; O-GalNAc, O-linked b-N-acetylga-
lactosamine; OGT, protein uridine diphospho-N-acetylglucosamine:
polypeptide b-N-acetylglucosaminyl transferase; MUGlcNAc, 4-meth-
ylumbelliferyl 2-acetamido-2-deoxy-b-D-glucopyranoside; UDP, uri-
dine diphosphate; DMSO, dimethyl sulfoxide; NMR, nuclear magnetic
resonance; TLC, thin-layer chromatography; FAB MS, fast atom
bombardment mass spectroscopy; LC–ESIMS, combined liquid chro-
matography and electrospray ionization spectroscopy; HPLC, high
performance liquid chromatography; PBS, phosphate-buffered saline;
DMEM, Dulbecco’s modified eagle medium; PUGNAC, O-(2-acet-
amido-2-deoxy-D-glucopyranosylidene)-amino-N-phenylcarbamate.
1
–3
4,5
6
7
8
9–11
are known
to be post-translationally modified with O-linked b-N-
acetylglucosamine (O-GlcNAc). The O-GlcNAc modifi-
cation is ubiquitous in all higher eukaryotes that have
been examined. In contrast to classical glycosylation,
which occurs on membrane and secretory proteins with
either N- or O-linkage in a more complex fashion, this
modification is added to intracellular proteins through
a b-O-linkage. The O-GlcNAc modification may act in
a manner analogous to protein phosphorylation in that
it is modulated by various extracellular stimuli and
*
Corresponding author. Tel.: +1 301 496 9405; fax: +1 301 496
431; e-mail: eunjuk@intra.niddk.nih.gov
9
0008-6215/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.carres.2006.03.004

972
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
1
2
growth conditions and the turnover rate of the glycan
is much faster than that of the polypeptide backbone.
This implies that O-GlcNAc is dynamically added to
strate. However, neither residues acting as the catalytic
nucleophile or the base, nor the region responsible for
either the O-GlcNAcase activity or the substrate binding
site has been clearly established. Instead, two opposite
1
3–15
and removed from proteins.
Analogous to kinases
3
5,38–40
and phosphatases, two enzymes are involved in the
regulated cycling of O-GlcNAc. Sites modified by O-
GlcNAc are often the same or adjacent to those of phos-
phorylation suggesting a competitive role for O-GlcNAc
and phosphorylation in regulating protein function.
Recent evidence suggests that deregulation of cellular
hypotheses have been proposed.
The N-terminal
‘hyaluronidase’ domain has been proposed to be the cat-
alytic site for O-GlcNAc removal based on the fact that
both isoforms of O-GlcNAcase have the same amino
terminus and this N-terminal region has limited homo-
1
6–18
3
8
logy with other glycosidases. O-GlcNAcase is catego-
rized into the family 84 of glycoside hydrolases (GH)
as defined by the carbohydrate-active enzymes (CAZy)
database. In support of the proposal that the N-terminal
region is responsible for the catalytic O-GlcNAcase
1
9
O-GlcNAc levels might play a role in type II diabetes,
2
0
21
cancer, and neurological disorders. For example,
elevation of O-GlcNAc levels by altering one of the
enzymes of O-GlcNAc metabolism can give rise to insu-
1
9
22
39
lin resistance in transgenic mice and in tissue culture.
activity, Rigden et al. demonstrated that the common
The enzymes responsible for the cycle of O-GlcNAc
addition and removal are the O-GlcNAc transferase
conserved N-terminus region in the glycoside hydrolase
(GH) family 84 is likely to conform to the catalytic
triosephosphate isomerase (TIM) barrel fold. Whereas,
the opposite proposal that the O-GlcNAcase catalytic
activity resides in the C-terminal acetyltransferase do-
(
OGT, EC 2.4.1.94) and the O-GlcNAc specific b-N-
acetylglucosaminidase known as O-GlcNAcase (EC
3
.2.1.52). Both enzymes have been purified, cloned,
7
,23–29
40
and characterized.
Recently, the crystal structure
main was based on the observation that O-GlcNAcase
of the multiple tetratricopeptide repeats (TPR) domain
of OGT has been solved and the TPR domain was
shown to play a role in dimerization of OGT itself.
enzymatic activity has never been detected in the splice
variant of O-GlcNAcase that lacks the putative histone
3
0
36
acetyltransferase domain. To date, the functional role
Compared to OGT, very little is known about the O-
GlcNAcase or its splice variant. Unlike acidic hexosami-
nidases in lysosomes, O-GlcNAcase has a neutral pH
optimum, nucleocytoplasmic distribution, and substrate
specificity toward GlcNAc and not N-acetylgalactos-
of the splice variant is still unclear.
The standard in vitro assay for measuring O-GlcNAc-
ase uses pNP-b-GlcNAc as a substrate. Enzyme activ-
ity is monitored spectrophotometrically by measuring
the absorbance at 400 nm of a p-nitrophenolate ion that
2
9
29
29,41
amine (GalNAc). Because of these distinct characteris-
is enzymatically released from pNP-b-GlcNAc.
tics, O-GlcNAcase is most likely the same enzyme as
However, this assay is relatively insensitive because the
substrate undergoes only a modest change in absor-
bance upon cleavage.
3
1,32
hexosaminidase C.
The primary sequence of O-
GlcNAcase was originally identified as meningioma
expressed antigen 5 (MGEA 5), an autoantigen in men-
To facilitate the study of O-GlcNAcase isoforms, a
highly sensitive method is necessary. Because fluores-
cence is highly sensitive, a great number of fluorogenic
substrates have been widely used in a variety of biologi-
3
3
ingiomas. Interestingly, the O-GlcNAcase gene maps
to chromosome 10q24 and this locus has been impli-
3
3
cated in Alzheimer’s disease. Full-length O-GlcNAc-
ase is a 916-amino acid protein having an acidic
isoelectric point with a predicted molecular weight of
3
7,42–46
cal areas.
To date, a number of hexosaminidases
have been examined by using fluorescent analogues and
2
7,33
1
03 kDa.
The N-terminal and the C-terminal region
the sensitivities and usefulness of fluorogenic assays
3
3,34
37,42–46
of O-GlcNAcase has homology to hyaluronidases
and acetyltransferases, respectively. The C-terminal
have been demonstrated.
For example, fluores-
3
5
cein di-b-D-galactoside (FDG) has been used to measure
the degree of transfection of the b-galactosidase reporter
of O-GlcNAcase has recently been shown to have his-
3
6
43
tone acetyltransferase (HAT) activity. A splice variant
gene. The sensitivity of this fluorometric assay was
of O-GlcNAcase designated as MGEA 5s was also iden-
tified. This isoform is a 677-amino acid protein that
found to be 50-fold higher than that of the spectro-
3
3
photometric assay of b-galactosidase with o-nitrophen-
4
4
lacks C-terminal third of the full-length O-GlcNAcase
yl-b-D-galactopyranoside.
Vocadlo and co-workers
3
3,34
(
the HAT domain).
The crystal structure of either
used a series of 4-methylumbelliferyl 2-N-acetyl-2-deoxy-
b-D-gluocopyranosides (MUGlcNAc) as fluorogenic
substrates for O-GlcNAcase in elucidating the catalytic
O-GlcNAcase or v-O-GlcNAcase has not yet been
solved. Recently, Vocadlo and co-workers proposed
that O-GlcNAcase utilizes a substrate-assisted mecha-
3
7
mechanism of O-GlcNAcase. The fluorescent moiety
in MUGlcNAc is 4-methylumbelliferone. However, it
has been reported that the sensitivity of the assay using
FDG is more than 10-fold higher than that of the assay
3
7
nism of action. In that mechanism, 2-acetamido group
of the substrate is involved in catalytic hydrolysis via
formation and breakdown of a bicyclic oxazoline or
oxazolinium ion intermediate leading to the retention
of the stereochemistry at anomeric center of the sub-
4
5
with 4-methylumbelliferyl-b-D-galactoside.
In other
words, the sensitivity of the assay was higher when fluo-

E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
973
OH
OAc
O
O
HO
HO
CH COCl
AcO
AcO
3
16h, r.t.
AcHN
OH
AcHN
Cl
2
1
HO
O
OH
o
Ag O/CH CN, 40 C
2
3
O
O
48h
cat. pyr.
OR
OR
O
O
RO
RO
RO
RO
O
O
O
O
O
OH
NHAc
+
AcHN
AcHN
O
O
O
O
O
RO
OR
OR
3
5
R = acetyl
R = H
R = acetyl
R = H
4
6
Scheme 1. Synthesis of fluorescein di(N-acetyl-b-D-glucosaminide) (FDGlcNAc). The synthesis of FDGlcNAc was carried out essentially according
47
to the procedure described by Kasai et al. with some modifications; see details in the Experimental.
rescein was chosen as a fluorophore than when 4-methyl-
umbelliferone was chosen. Therefore, we have selected
fluorescein as the backbone of our substrate to achieve
high sensitivity of the assay and prepared fluorescein,
di-N-acetyl-b-D-glucosaminide (FDGlcNAc). The syn-
thesis of FDGlcNAc was carried out according to the
procedure described by Kasai et al. with some modifi-
cations. In this article, we evaluate the usefulness of
FDGlcNAc as a new fluorogenic substrate for the study
of both O-GlcNAcase isoforms by measuring their
kinetic parameters. This fluorogenic substrate becomes
fluorescent only upon cleavage by O-GlcNAcase. Due
to the improved sensitivity of this substrate, we report
that shorter splice variant is an active O-GlcNAcase
gene encoding O-GlcNAcase is associated with diabetes
mellitus and age of onset in the Mexican–American pop-
ulation highlights the importance of the physiological
role of O-GlcNAcase. Growing interest in O-GlcNAcase
stimulated the need for a highly sensitive enzymatic
assay method. A series of 4-methylumbelliferyl 2-
deoxy-2-N-fluoroacetyl b-D-glucopyranoside substrates
for O-GlcNAcase were recently prepared by Vocadlo
4
7
3
7
and co-workers. Because a fluorescein-conjugated sub-
strate has higher sensitivity than the corresponding
4
5
methylumbelliferyl substrate, fluorescein was selected
as the aglycone of the synthetic fluorogenic substrate.
O-GlcNAcase exhibits both stereo and substrate spec-
ificities. This enzyme catalyzes the hydrolysis of b-O-
linked pNP-GlcNAc, but not a-O-linked pNP-GlcNAc.
Neither b-O-linked pNP-GalNAc, nor para-nitrophenyl
gluocopyranoaminide was cleaved by the enzyme (data
not shown). Therefore, fluorescein conjugated to two
O-GlcNAc moieties via b-linkages, 5, was designed as
a substrate for O-GlcNAcase.
(Scheme 1).
2
. Results and discussion
The O-GlcNAc specific b-N-acetylglucosaminidase (O-
GlcNAcase) is involved in O-GlcNAc removal, playing
a key role in O-GlcNAc metabolism. For example, O-
0
0
The synthesis of 2 ,7 -dichlorofluorescein, di-N-acetyl-
b-D-glucosaminide and its use as a substrate for the
colorimetric rate-assay of urinary N-acetyl b-D-gluoco-
saminidase (NAGase, EC 3.2.1.30) was originally
(
2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-
phenylcarbamate (PUGNAC), a potent inhibitor of the
4
7
O-GlcNAcase, blocks O-GlcNAc cycling, and induces
reported by Kasai et al. In the conditions they devel-
oped for glycosidation, acetonitrile was used as a solvent
and either Et N or silver oxide (Ag O) was used as a
2
2,48–51
insulin resistance.
As mentioned before, the hu-
man MGEA5 gene produces alternatively spliced tran-
scripts resulting in two isoforms. One is a full-length
enzyme with a unique C-terminal region considered as
3
2
catalyst. In the optimal reaction system, two GlcNAc
moieties were conjugated into a dichloro derivative of
3
6
47
a putative histone acetyltransferase domain and the
other is a shorter enzyme that does not have this unique
C-terminal region. To date, active O-GlcNAcase has
fluorescein in a moderate yield. However, when fluo-
rescein was used as the aglycone, FDGlcNAc, which is
the product of glycosylation of two hydroxyl groups in
fluorescein, was not formed, possibly due to the
weak reactivity of the secondary hydroxyl group in
5
2
been recognized in only the longer isoform. The report
that a single nucleotide polymorphism in the MGEA5

974
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
fluorescein. To overcome this obstacle, a small amount
of base was used to increase the reactivity of the hydro-
xyl group in fluorescein. Thus, FDGlcNAc was success-
fully synthesized in moderate yield by treatment of
fluorescein with a 3-fold molar quantity of 2-acetam-
ido-3,4,6-tri-O-acetyl-2-deoxy-a-D-glucopyranosyl chlo-
ride (2) as a glycosyl donor using acetonitrile as the
Neither O-GlcNAcase nor v-O-GlcNAcase cleaved b-
O-glycosidic linkage of FDGalNAc. This result is con-
sistent with previous findings that O-GlcNAcase has
substrate specificity for O-GlcNAc but not O-GalNAc.
FDGlcNAc is nonfluorescent until cleavage of the
b-D-O-glycosidic bond by either O-GlcNAcase or
v-O-GlcNAcase. Spontaneous hydrolysis of FDGlcNAc
was not observed, demonstrating that FDGlcNAc is
as stable in aqueous solution (data not shown) as other
substrates such as pNP-b-GlcNAc and MUGlcNAc.
Enzymatic hydrolysis produces fluorescent fluorescein
solvent and silver oxide (Ag O) as a catalyst in the pres-
2
ence of small amounts of pyridine at 40 ꢀC for 48 h. The
two hydrophilic GlcNAc moieties of FDGlcNAc
provide enhanced solubility (>100 mM) compared to
that of the conventional pNP-GlcNAc (<10 mM).
For the purpose of affinity purification and immuno-
logical identification, six consecutive histidine residues
were recombinantly fused to the O-GlcNAcase, and its
variant, at the N-terminus. These recombinant O-
GlcNAcase and v-O-GlcNAcase proteins were expressed
in Escherichia coli and purified on a nickel column via
mono(N-acetyl-b-D-gluocosaminide)
(FMGlcNAc),
which is released from FDGlcNAc, tautomerizes to the
quinoid form and generates fluorescence (Scheme 2).
FMGlcNAc is then further hydrolyzed by the action of
enzyme giving rise to fluorescein at longer time (Scheme
2). The products generated by enzymatic hydrolysis of
FDGlcNAc were examined by LC–ESI (Fig. 2). A single
the His -tag. To obtain relatively pure protein, an opti-
peak representing FDGlcNAc (t = 17.8 min) confirms
6
R
mized washing protocol was employed (see Methods).
The v-O-GlcNAcase eluted with a lower concentration
that it is the only species present prior to addition of
enzyme (Fig. 2A). All of the enzymatic assays performed
were stopped when less than 5% of FDGlcNAc was
utilized. As expected, during the time frame of our kinetic
(
60 mM) of imidazole as compared to full-length O-
GlcNAcase (300 mM). The flow-through fractions
containing protein were assayed with two substrates,
pNP-b-GlcNAc and FDGlcNAc. The fractions contain-
ing enzymatic activity were concentrated with Microcon
centrifugal filter devices (Microcon YM-100 and YM-50
for O-GlcNAcase and v-O-GlcNAcase, respectively)
and the concentrated enzymes were reconstituted in
assay, FMGlcNAc (t = 20.0 min) was the only fluores-
R
cent species generated by the enzymatic hydrolysis and
fluorescence was due to the accumulation of FMGlcNAc
alone (Fig. 2B). When 2-fold more enzyme is used and
incubation time is extended, the fluorescein (t =
R
24.0 min) released upon cleavage of the second N-acet-
yl-b-D-glucosaminyl group from further hydrolysis of
FMGlcNAc was detected (Fig. 2C). However, during
the timeframe of our assay for measuring initial velocities
of enzymatic hydrolysis with varying concentrations of
FDGlcNAc and predetermined amounts of enzyme, only
one of the N-acetyl-b-D-gluocosaminyl groups was
cleaved from the substrate producing fluorescent
FMGlcNAc. Given these data, we used FMGlcNAc as
the standard for kinetics of the O-GlcNAcase enzymatic
reaction. When the fluorometer was set to have excita-
tion and emission wavelengths at 485 and 535 nm,
respectively, with the 4 mm diameter of circular aperture
and 1 s of counting time, a plot of increasing concentra-
tions of FMGlcNAc up to 12 lM versus their fluores-
cence intensities (relative fluorescence units, RFU)
resulted in a linear correlation. From this plot the calcu-
lated molar fluorescence constant of FMGlcNAc was
2
0 mM Tris–HCl, pH 7.5, 1 mM DTT aqueous solution.
Their purity was assessed by Ponceau staining and
immunoblotting (Fig. 1). Recombinant His -tagged-O-
GlcNAcase and -v-O-GlcNAcase migrated at their
apparent molecular weight of ꢀ130 and ꢀ80 kDa,
respectively. Additionally the modified washing proto-
6
col was essential to produce sufficiently pure His -tagged
enzymes.
6
A.
B.
C.
D.
kDa
kDa
1
81.8
1
20
00
1
1
15.5
8
0
8
2.2
6
0
5
0
0
63.2
48.8
1
19,820 units/lM.
4
Utilizing FDGlcNAc as a substrate, kinetic para-
3
7.1
3
0
0
meters including Michaelis constants (K ), specific
m
2
1
5.9
9.4
activities, turnover numbers, and enzymatic efficiencies
of the recombinant O-GlcNAcase were obtained
through Lineweaver–Burk plots (Fig. 3). Surprisingly,
incubation of FDGlcNAc with the v-O-GlcNAcase also
produced fluorescence, indicating that this is an active
enzyme. The kinetic parameters of the variant enzyme
were calculated as well (Fig. 4). For the comparison of
2
Figure 1. Isolation of O-GlcNAcase and v-O-GlcNAcase. The purity
of O-GlcNAcase and v-O-GlcNAcase was confirmed by Ponceau S
staining of the nitrocellulose membranes (A, C) and Western blot
analysis using mouse HRP conjugated 6X His tag antibody [AD1.1.10]
(B, D), respectively.

E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
975
OH
O
HO
HO
O
O
O
NHAc
AcHN
O
O
O
HO
OH
OH
5
O-GlcNAcase or v-O-GlcNAcase
OH
O
OH
HO
HO
O
OH
O
HO
HO
O
O
OH
O
O
O
HO
HO
+
OH
AcHN
AcHN
AcHN
O
O
CO H
2
1
6
O-GlcNAcase or v-O-GlcNAcase
HO
O
OH
+
1
O
O
Scheme 2. Enzymatic hydrolysis reaction of FDGlcNAc (5) by either O-GlcNAcase or v-O-GlcNAcase. FMGlcNAc accumulated at the initial stage
of enzymatic hydrolysis. However, fluorescein can be formed under conditions of higher enzyme concentration or longer reaction times.
these kinetic parameters with FDGlcNAc, enzymes were
assayed with pNP-b-GlcNAc (Fig. 5) and the kinetic
values of both enzymes for FDGlcNAc and pNP-b-Glc-
NAc are summarized in Table 1. Using the FDGlcNAc
While our Km value of O-GlcNAcase for pNP-b-Glc-
4
0
NAc was in good agreement with previous reports, the
specific activity differed. The specific activity of our
recombinant enzyme for pNP-b-GlcNAc was found to
be 2035 nmol/min/mg whereas the reported value of
their recombinant O-GlcNAcase for the pNP-b-GlcNAc
substrate, we established that O-GlcNAcase has a K of
m
8
5 lM and a specific activity of 780 nmol/min/mg with a
4
0
turnover rate of 1.3/s. Catalytic efficiency (K /K ) of
substrate was 652 nmol/min/mg. In contrast to the
cat
m
4
ꢁ1 ꢁ1
O-GlcNAcase for FDGlcNAc was 1.6 · 10 M
s
.
K , specific activity is not a constant value and depends
m
On the other hand, O-GlcNAcase exhibited a Km of
.1 mM and a specific activity of 2035 nmol/min/mg
with a turnover number of approximately 3/s and cata-
on the enzymatic assay conditions as well as the purity
4
0
1
of enzymes. Wells et al. used 50 mM sodium cacodyl-
ate, pH 6.5 as an assay buffer for measuring O-GlcNA-
case activity whereas O-GlcNAcase activity was assayed
in 100 mM citrate phosphate buffer, pH 6.5 in our
kinetic experiments. Furthermore, recombinant O-
GlcNAcase enzymes have different tags. The cloned
human O-GlcNAcase used in our assays has a small
tag composed of six histidine amino acids, while recom-
binant O-GlcNAcase used in their assay has a thiore-
doxin tag. Thioredoxin is an approximately 10 kDa
protein and its presence may reduce O-GlcNAcase cata-
lytic activity. It is likely that a small oligopeptide such as
a six consecutive histidine residue does not significantly
affect the catalytic activity of O-GlcNAcase. Because the
specific activity of our recombinant O-GlcNAcase for
pNP-b-GlcNAc is about 3-fold greater than that
3
ꢁ1 ꢁ1
lytic efficiency of 3.0 · 10 M
s
for pNP-b-GlcNAc.
The low Km and high affinity of O-GlcNAcase for
FDGlcNAc can be explained by the rigidity and hydro-
phobicity of the aglycone fluorescein. The respective in-
crease in both the hydrophobicity and rigidity between
the three substrates pNP-b-GlcNAc, MUGlcNAc, and
FDGlcNAc may reflect an associated decrease in the
K values for these substrates and an increase in appar-
m
ent affinity. The affinity of O-GlcNAcase for FDGlc-
NAc (K = 85 lM) is about 13-fold stronger than it is
m
for pNP-b-GlcNAc (K = 1.1 mM) and 5-fold stronger
m
4
5
than it is for the fluorogenic substrate (K = 430 lM).
m
The most hydrophobic aglycone, fluorescein, is thought
to mimic the rigid serine and threonine residues found in
proteins. O-GlcNAcaes can access this constrained and
hydrophobic substrate better than MUGlcNAc and
the least constrained substrate pNP-b-GlcNAc.
4
0
reported previously, other kinetic parameters such as
turnover number and catalytic efficiency that are depen-
dent on specific activity value are also 3-fold higher.

976
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
C.
B.
A.
24.0
8
7
6
5
4
3
2
1
000000
000000
000000
000000
000000
000000
000000
000000
0
20.0
1
7.8
9
.7
7
.5
10
12.5
15
17.5
20
22.5
25
27.5 min
9
8
7
6
5
4
3
2
1
000000
000000
000000
000000
000000
000000
000000
000000
000000
0
17.8
20.0
9.7
7.5
10
12.5
15
17.5
20
22.5
25
27.5 min
17.8
4
3
2
1
000000
000000
000000
000000
0
7.5
10
12.5
15
17.5
20
22.5
25
27.5 min
Retention time (min)
Figure 2. Enzymatic hydrolysis of FDGlcNAc examined by LC–ESI. LC conditions are as follows: flow rate, 0.2 mL/min; isocratic elution of H
CH CN (v/v), 97:3 for 5 min, then a gradient of H O/CH CN (v/v), from 97:3 to 2:98 for 40 min. (A) A chromatogram of FDGlcNAc
= 17.8 min) in the absence of O-GlcNAcase. (B) New peaks arising after a 10 min incubation with O-GlcNAcase (2 lL) at 37 ꢀC are due to
GlcNAc (t = 9.7 min) and FMGlcNAc (t = 20.0 min), respectively. (C) A chromatogram of FDGlcNAc incubated with O-GlcNAcase (4 lL) at
7 ꢀC for 7 h. Fluorescein is detected and has a retention time of 24 min.
2
O/
3
2
3
(t
R
R
R
3
0
0
0
0
0
0
0
0
0
.8
.7
.6
.5
.4
.3
.2
.1
.0
1
1
1
.6
.4
.2
8
7
6
5
4
3
2
1
1
1
1
1
6
4
2
0
8
6
4
2
1.0
0.8
0.6
0.4
0
0
.2
.0
-
0.02 0.00 0.02 0.04 0.06 0.08 0.10
/[s]
-1
0
1
2
3
4
5
1
1/[s]
0
40
80
120
160
200
240
280
320
0
2
4
6
8
10
12
Substrate concentration (μM)
Substrate concentration (mM)
Figure 3. Enzyme kinetics of full-length O-GlcNAcase. An activity
curve and a Lineweaver–Burk plot were generated for recombinant O-
GlcNAcase by varying concentrations of the substrate, FDGlcNAc.
Figure 4. Enzyme kinetics of variant O-GlcNAcase. An activity curve
and a Lineweaver–Burk plot were generated for recombinant v-O-
GlcNAcase by varying concentrations of the substrate, FDGlcNAc.

E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
977
therefore, the O-GlcNAcase catalytic domain resides
2
1
1
0
0
0
.0
.6
.2
.8
.4
.0
35,40
at the C-terminal end of enzyme.
However, our
in vitro kinetic data suggest that v-O-GlcNAcase is quite
active, although the Km value of v-O-GlcNAcase for
FDGlcNAc was 25-fold higher than that of the full-
length O-GlcNAcase.
1
1
1
0
0
0
.8
.5
.2
.9
.6
.3
To further confirm that the shorter O-GlcNAcase is
active, a mixed population of substrates from both
HT-29 and HeLa cell extracts were incubated with either
O-GlcNAcase or v-O-GlcNAcase as described under
Experimental. As shown in Figure 6A, treatment of
HT-29 cell extracts with either O-GlcNAcase (lane 3
and 4) or its shorter isoform (lane 5) decreased the level
of O-GlcNAc-modified proteins compared to that of cell
extracts treated with no enzyme (lane 2). Given the dif-
ferences in enzymatic kinetics between the two enzymes
-
0.9 -0.6 -0.3 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1
/[s]
1
0
1
2
3
4
5
6
7
8
9
Substrate concentration (mM)
Figure 5. Enzyme kinetics of O-GlcNAcase. An activity curve and a
Lineweaver–Burk plot were generated for recombinant O-GlcNAcase
by varying concentrations of the substrate, pNP-b-GlcNAc.
(
Table 1), it was not surprising that v-O-GlcNAcase did
not lower the levels of O-GlcNAc as well as did full-
length O-GlcNAcase. Similar results were obtained from
HeLa cell extracts (Fig. 6B). These findings suggest that
both the full-length and variant O-GlcNAcases are ac-
tive against cellular O-GlcNAc-modified substrates con-
firming our earlier in vitro data using FDGlcNAc.
Based on these data, v-O-GlcNAcase activity should
be detected using much higher concentrations of pNP-
b-GlcNAc. Given the fact that this shorter isoform is
a bona fide O-GlcNAcase and does not possess the
Interestingly, O-GlcNAcase purified from bovine brain
2
7
was reported
to exhibit
a
specific activity
(
1840 nmol/min/mg) similar to the specific activity re-
ported here.
Previous reports using pNP-b-GlcNAc (up to 8 mM)
as a substrate, did not detect activity of the shorter
4
0
O-GlcNAcase splice variant. However, using FDGlc-
NAc we found that v-O-GlcNAcase has a Km of
2
.1 mM and a specific activity of 1680 nmol/min/mg
with a turnover rate of 2.1. The catalytic efficiency of
this shorter isoform for FDGlcNAc was determined
A.
kDa
B.
4
ꢁ1 ꢁ1
1
2
3
4
5
to be 1.0 · 10 M
s
. The activity detected with
1
2
3
4
5
the shorter O-GlcNAcase was unanticipated. There
kDa
has been controversy regarding the catalytic region
1
70
30
3
5,38–40
1
of O-GlcNAcase.
As previously mentioned, the
170
130
1
100
C-terminal domain of O-GlcNAcase is homologous to
acetyltransferases domain and is absent in v-O-GlcNAc-
ase. Instead, this acetyltransferase domain is replaced
by a novel 15 amino acid tail in the variant. Although
00
72
7
5
2
5
55
3
8
39
Hanover et al. and Rigden et al. proposed that the
catalytic domain of O-GlcNAcase is likely to be located
in the N-terminal region, there have been no empirical
data that support this proposal. On the contrary, the
lack of enzymatic activity in the shorter isoform led to
such a conclusion that v-O-GlcNAcase is not active,
Figure 6. Enzymatic activity of O-GlcNAcase and v-O-GlcNAcase on
mixed substrates. Western blot analysis of HT-29 (A) and HeLa (B) cell
extracts treated with either O-GlcNAcase or v-O-GlcNAcase. Lane 1,
protein marker; lane 2, cell extracts without enzyme (control); lanes 3
and 4, cell extracts treated with 7 and 5 lg of O-GlcNAcase, respec-
tively; and lane 5, cell extracts treated with 5 lg of v-O-GlcNAcase.
Table 1. Kinetic parameters of both recombinant O-GlcNAcase isoforms for FDGlcNAc, 5, and pNP-b-GlcNAc
Kinetic parameter
Substrate
Enzyme
O-GlcNAcase
v-O-GlcNAcase
K
m
(mM)
FDGGlcNAc
pNP-b-GlcNAc
FDGGlcNAc
pNP-b-GlcNAc
FDGGlcNAc
pNP-b-GlcNAc
FDGGlcNAc
pNP-b-GlcNAc
0.085 ± 0.016
1.1 ± 0.1
780 ± 60
2035 ± 161
1.3 ± 0.1
2.1 ± 0.2
—
1680 ± 165
—
Specific activities (nmol/min/mg)
ꢁ
1
Turnover number (Kcat, s
)
2.1 ± 0.2
—
3.4 ± 0.3
ꢁ1
ꢁ1
4
3
K
cat/K
m
(efficiency, M
s
)
1.6 · 10
1.0 · 10
3
3.1 · 10
—

978
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
C-terminal histone acetyl transferase domain, it is likely
that the O-GlcNAcase catalytic domain is located in the
common N-terminal domain that has homology to Mu
Coupling constants, J, are reported in Hertz. Analytical
thin-layer chromatography (TLC) was conducted on
Silica Gel 60-F254 (E. Merck) with detection by UV light
and/or dipping the plates in a cerium sulfate–ammo-
nium molybdate solution followed by heating. Flash col-
umn chromatography was carried out on Merck Kiesel
3
8
toxin. Some evidence suggests that the full-length
O-GlcNAcase might be a bifunctional enzyme capable
of both catalyzing the hydrolysis of b-linked GlcNAc
3
6
and catalyzing acetylation of histones. Because there
has been no tool to measure the enzymatic activity in
v-O-GlcNAcase, it is not clear what function this shorter
isoform serves in a cellular context. One hypothesis is
that it serves as a regulatory tool in concert with the
gel 60 (SiO , 230–400 mesh). High performance liquid
2
chromatography (HPLC) was performed on an Inertsil
ODS-2 column (10 mm i.d. · 250 mm) using a Hewlett
Packard Series 1100 multisolvent delivery system with
a multiple wavelength detector (flow rate, 2 mL/min;
detection, 254, 280 nm; temperature, ambient). Fast
4
0
full-length of O-GlcNAcase.
+
The nature of FDGlcNAc made it an extremely sensi-
tive substrate for O-GlcNAcase enabling the detection
of enzymatic activity in v-O-GlcNAcase for the first
time. In conclusion, FDGlcNAc has several advantages
for the study of both O-GlcNAcase isoforms over the
conventional substrate, pNP-b-GlcNAc and the prece-
atom bombardment (FAB ) MS was obtained at the
National Institutes of Health Mass Spectrometry Labo-
ratory. High resolution-electrospray ionization (HR-
ESI) mass spectrum was obtained on a Waters LCT Pre-
mier time-of-flight (TOF) mass spectrometer. Combined
liquid chromatography and electrospray ionization
(LC–ESI) mass spectra were obtained on Hewlett Pack-
ard Series 1100 MSD equipped with HPLC performed
3
7
dent fluorogenic substrate, MUGlcNAc, in terms of
sensitivity of the assay and the affinity for the enzyme.
In addition, a high-throughput analysis is possible using
a multiwell plate format with this new substrate. To
dissect the role of each O-GlcNAcase isoform, it is
necessary to find specific inhibitors of full-length O-
GlcNAcase over its shorter isoform or vice versa. The
high-throughput assay with this highly sensitive fluoro-
genic substrate, FDGlcNAc will accelerate the discovery
of novel O-GlcNAcase inhibitors. The highly sensitive
substrate FDGlcNAc made it possible to confirm that
the v-O-GlcNAcase as an active enzyme. The functional
role of each O-GlcNAcase isoform is currently under ac-
tive investigation. Although FDGlcNAc is cell perme-
able, its use in intact cells as an imaging agent for the
in vivo analysis of O-GlcNAcase is limited due to the
lack of enzyme specificity between O-GlcNAcase and
lysosomal hexosaminidases. We are currently develop-
ing fluorogenic substrates, which are specific for only
O-GlcNAcase but not lysosomal hexosaminidases.
TM
on an Atlantis dC18 column (2.1 mm i.d. · 100 mm)
at the National Institutes of Health Mass Spectrometry
Laboratory. Elemental analyses were obtained at
Atlantic Microlab, Inc. Fluorometry measurements
were obtained at wavelengths of 485 and 535 nm for
fluorescence excitation and emission, respectively, with
Wallac 1420 multilabel counter (Perkin–Elmer Life
Sciences). Absorbance was measured at 400 nm with a
UV 160U spectrophotometer (Shimadzu).
0
0
3.2. 3 ,6 -Bis[[3,4,6-tri-O-acetyl-2-(acetylamino)-2-deoxy-
b-D-glucopyranosyl]oxy]-spiro[isobenzofuran-1(3H),
9 -[9H]-xanthen]-3-one (3)
0
Fluorescein (1.0 g, 3 mmol) was added to a solution of
N-acetyl 2-deoxy-3,4,6-tri-O-acetyl-a-D-glucopyranosyl
5
3
chloride (2, 3.3 g, 9.0 mmol) in dry CH CN. To this
3
mixture were, subsequently, added silver oxide (Ag O,
2
2
.09 g, 9.0 mmol) and five drops of pyridine. The reac-
tion solution was stirred for 2 days at 40 ꢀC, after which
the remaining solid residue was removed by filtration and
the filtrate was concentrated to yield a dark brown syr-
up. Flash column silica chromatography of the resulting
residue washing with a mixture of hexane/EtOAc (2:1),
then eluting with a gradient of CHCl /CH CN from
3
. Experimental
3
.1. General methods
All chemicals were of reagent grade and purchased from
Sigma Chemical Co. unless otherwise noted. Deuterated
DMSO was from Cambridge Isotope Laboratories.
FDGalNAc was purchased from Marker Gene Technol-
3
3
1:1 to 1:2 (v/v) offered the desired product 3 as a white
1
solid (1.19 g, 1.2 mmol, 40%): R = 0.12 (EtOAc); H
f
NMR (300 MHz, DMSO-d ): d 8.08 (app. dd, 2H,
6
˚
ogies, Inc. Acetonitrile was dried over 4 A molecular
J = 8.7, 2.7 Hz, arom H), 8.03 (d, 2H, J = 7.2 Hz,
sieves overnight prior to use. The Inertsil ODS-2 column
NHCOCH ), 7.80 (dt, 1H, J = 7.6, 1.5 Hz, arom H),
3
1
13
was obtained from GL Science Inc. H NMR and
C
7.74 (dt, 1H, J = 7.2, 1.1 Hz, arom H), 7.31 (d, 1H,
J = 7.6 Hz, arom H), 7.03 (dd, 2H, J = 5.3, 2.0 Hz, xan-
thene H), 6.81–6.72 (m, 4H, xanthene H), 5.44 (d, 2H,
J = 8.4 Hz, GlcNAc H-1), 5.24 (app. dt, 2H, J = 9.7,
3.0 Hz, GlcNAc H-3), 4.94 (t, 2H, J = 9.5 Hz, GlcNAc
H-4), 4.25–3.96 (m, 8H, GlcNAc H-5, H-6a, H-6b, H-
NMR spectra were recorded at 300 and 75 MHz, respec-
tively, with a Varian Mercury 300 spectrometer. Chemi-
cal shifts are reported in parts per million (d) and
referenced to either the residual signal of the solvent
1
used (DMSO: H d 2.50) or internal tetramethylsilane.

E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
979
2
6
), 2.06, 2.03 (s, each 3H, OCOCH ), 2.01, 1.96 (s, each
(300 MHz, DMSO-d ): d 8.02 (d, 1H, J = 7.42 Hz, arom
3
6
1
3
H, OCOCH ), 1.78, 1.77 (s, each 3H, NHCOCH );
C
H), 7.82–7.71 (m, 4H, 2 · arom H, 2 · NHCOCH ),
3
3
3
NMR (75 MHz, DMSO-d ): d 170.02, 169.98, 169.69,
7.30 (d, 1H, J = 7.42 Hz, arom H), 6.96 (app. m, 2H,
xanthene H), 6.71 (app. s, 4H, xanthene H), 5.14–5.09
(m, 6H, GlcNAc H-1, OH-3, OH-4), 4.65 (app. q, 2H,
J = 4.7 Hz, OH-6), 3.75–3.63 (m, 4H, GlcNAc H-2, H-
6
1
1
69.54, 169.34, 168.51 (9C@O), 158.28, 158.22, 152.39,
51.42, 151.39 (5C), 135.84 (arom CH), 130.38 (arom
CH), 129.34 (xanthene, 2CH), 125.65 (1C), 124.86
arom CH), 124.00 (arom CH), 113.86 (xanthene,
(
6a), 3.53–3.15 (m, 8H, GlcNAc H-3, H-4, H-5, H-6b),
3
2
9
CH), 112.81, 112.87 (2C), 103.44 (xanthene, 2CH),
7.50 (2CH, C-1), 81.58 (xanthene, spiro C), 72.24
1.80, 1.79 (s, 6H, 2 · NHCOCH ); C NMR (75 MHz,
3
DMSO-d ): d 169.30 (2NHC@O), 168.57 (C@O),
6
(
(
(
2CH, C-3), 70.96 (2CH, C-5), 68.33 (2CH, C-4), 61.64
2CH, C-6), 53.05 (2CH, C-2), 20.34, 20.44, 20.48
6CH , COCH ), 22.62 (2CH , NHCOCH ). Anal.
159.05, 159.01, 152.45, 151.51, 151.46 (5C), 135.80
(arom CH), 130.33 (arom CH), 129.20 (xanthene,
2CH), 125.74 (1C), 124.81 (arom CH), 123.98 (arom
CH), 113.77 (xanthene, 2CH), 112.31, 112.28 (2C),
103.27, 103.19 (xanthene, CH), 98.78 (2CH, C-1),
3
3
3
3
Calcd for C H N O ÆH O: C, 57.14; H, 5.19; N,
2
4
8
50
2
21
2
.78. Found: C, 56.81; H, 5.18; N, 2.89.
8
1.87 (xanthene, spiro C), 77.32, 73.91, 70.20 (each
0
3
glucopyranosyl]oxy]-6 -hydroxyspiro[isobenzofuran-
1
.3. 3 -[[3,4,6-Tri-O-acetyl-2-(acetylamino)-2-deoxy-b-D-
2CH, C-3, C-4, C-5), 60.71 (2CH, C-6), 55.32 (2CH,
C-2), 23.04 (2CH , NHCOCH ); HR FAB MS: Calcd
for C H N O Cs: m/z 871.1327 [M+Cs] . Found:
36 38 2 15
0
3
3
0
+
(3H),9 -[9H]-xanthen]-3-one (4)
m/z 871.1300. Anal. Calcd for C H N O ÆH O: C,
3
6
38
2
15
2
Silver oxide (Ag O, 0.28 g, 1.2 mmol) and two drops of
pyridine were added to a solution of 2 (0.44 g, 1.2 mmol)
57.14; H, 5.33; N, 3.70. Found: C, 56.89; H, 5.45; N,
3.70.
2
and fluorescein (0.2 g, 1.2 mmol) in dry CH CN. The
3
0
reaction mixture was stirred for 2 days at 40 ꢀC, after
which the remaining solid residue was removed by filtra-
tion. The filtrate was concentrated and subjected to flash
column silica chromatography. Elution with a slow gra-
dient of CH Cl/CH CN from 2:1 to 1:1 (v/v) gave 0.48 g
3.5. 3 -[[2-(Acetylamino)-2-deoxy-b-D-glucopyran-
osyl]oxy]-6 -hydroxyspiro[isobenzofuran-1(3H),9 -[9H]-
xanthen]-3-one (6)
0
0
O-Deacetylation of 4 (150 mg, 0.23 mmol) with aqueous
3
3
of 4 as a yellow colored powder (0.73 mmol, 60%);
NH was carried out as described for 4. Purification of
3
1
R = 0.47 (EtOAc); H NMR (300 MHz, DMSO-d ): d
the crude material by reverse-phase HPLC offered
96 mg of 5 as an orange colored powder (0.18 mmol,
78%): HPLC: t = 11.3 min (H O/CH CN (v/v) from
f
6
8
.10 (app. dd, 1H, J = 9.0, 3.9 Hz, arom H), 8.01 (d,
2
H, J = 6.9 Hz, NHCOCH ), 7.79–7.70 (m, 2H, arom
3
R
2
3
H), 7.29 (app. t, 1H, J = 9.0 Hz, arom H), 7.09–7.07
m, 1H, xanthene H), 6.73–6.69 (m, 3H, xanthene H),
95:5 to 50:50 for 4 min, 50:50 to 27:73 for 11 min,
27:73 to 20:80 for 1 min, 2 mL/min); H NMR
1
(
6
8
9
.60–6.58 (m, 2H, xanthene H), 5.43 (d, 1H, J =
.7 Hz, GlcNAc H-1), 5.22, 5.21 (t, each 1/2H, J =
.8 Hz, GlcNAc H-3), 4.92 (t, 1H, J = 9.7 Hz, GlcNAc
(300 MHz, DMSO-d ): d 8.01 (d, 1H, J = 7.2 Hz, arom
6
H), 7.81–7.69 (m, 3H, 2 · arom H, NHCOCH ), 7.28
3
(app. dd, 1H, J = 7.5, 3.6 Hz arom H), 6.95 (app. s,
1H, xanthene H), 6.71–6.64 (m, 3H, xanthene H), 6.58
(app. s, 2H, xanthene H), 5.20–5.01 (m, 2H, OH), 5.07
(d, 1H, J = 8.7 Hz, GlcNAc H-1), 4.67 (br s, 1H, OH),
3.80–3.16 (m, 6H, GlcNAc H-2, H-3, H-4, H-5, H-6a,
H-4), 4.27–4.15 (m, 2H, GlcNAc H-5, H-6a), 4.15–3.98
m, 2H, GlcNAc H-2, H-6b), 2.06, 2.04 (s, each
(
3
1
/2H, OCOCH ), 2.01, 1.95 (s, each 3H, OCOCH ),
3 3
.77, 1.76 (s, each 3/2H, NHCOCH ); LR LC–ESIMS:
3
+
m/z 684.2 [M+Na ].
H-6b), 1.79 (s, 3H, NHCOCH ); HR-ESIMS: Calcd
3
+
for C H N O Na: m/z 558.1376 [M+Na] . Found:
2
8
25
1
10
0
0
3.4. 3 ,6 -Bis[[2-(acetylamino)-2-deoxy-b-D-glucopyran-
osyl]oxy]-spiro-[isobenzofuran-1(3H),9 -[9H]-xanthen]-3-
m/z 558.1390.
0
one (5)
3.6. Expression and purification of recombinant
O-GlcNAcase and v-O-GlcNAcase
A cooled (ꢁ20 ꢀC) solution of 3 (300 mg, 0.30 mmol) in
CH OH (5 mL) was treated dropwise with aqueous NH
Details describing the cloning of pBAD/HisA human O-
3
3
(
28%, 1 mL, 7.2 mmol). The reaction mixture was
GlcNAcase (MGEA 5) and its splice variant (MGEA5
s) clones will be reported elsewhere. Cultures of TOP
10 cells containing pBAD/HisA expression vector
5
4
warmed to rt over 2 h, stirred for an additional 1 h at
rt, and concentrated in vacuo to yield a pale yellow
solid. This crude material was purified by reverse-phase
HPLC to give 5 (170 mg, 0.23 mmol, 76.7%) as a white
solid: HPLC: t = 16.8 min (H O/CH CN (v/v) from
(Invitrogen) encoding the His -tagged-O-GlcNAcase
6
were grown overnight in Luria–Bertani (LB) broth (Di-
gene) supplemented with ampicillin (50 lg/mL) at 37 ꢀC
at 200 rpm. Induction of O-GlcNAcases was initiated by
adding arabinose solution (0.02%) when the OD600 of
R
2
3
9
6
5:5 to 80:20 for 5 min, 80:20 to 60:40 for 15 min,
1
0:40 to 35:65 for 25 min, 2 mL/min); H NMR

980
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
cultures reached at around 0.5. The cultures were fur-
ther grown for 4 h at 30 ꢀC at 200 rpm. The cells were
harvested by centrifugation at 4000g for 15 min in a
containing the fluorogenic substrate, 5, of which concen-
tration was varied (10–300 lM) in 20 lL of 0.5 M
citrate/phosphate buffer, pH 6.5, in a final volume of
100 lL at the start of the incubation at 37 ꢀC. The
assays were terminated by adding 900 lL of 0.5 M
Na CO solution after 3–5 min incubation. The assay
ꢁ
Sorvall RC 5C Plus (DuPont) centrifuge. The pellet
was subjected to freeze–thaw and suspended in 1/50 of
the original volume in 0.1 mg/mL lysozyme, 20 mM
Tris–HCl, pH 7.5, 1 mM DTT, 0.1% Triton X-100, an
EDTA-free protease inhibitor cocktail tablet (1 tablet/
2
3
solution (200 lL) was transferred into a 96-well plate
and fluorescence was measured at the excitation wave-
length of 485 nm (kex = 485 nm) and at the emission
wavelength of 535 nm (kem = 535 nm) on the Wallac
1
0 mL). The suspension was incubated at rt for 5 min
and sonicated on ice (4 · 10 s, setting 3, Misonix Ultra-
sonic Processor). The supernatant obtained after centri-
fugation at 20,000g for 15 min was stored in aliquots at
TM
1420 fluorometer (Perkin–Elmer Life Sciences). Assays
were performed in triplicate.
ꢁ
80 ꢀC. Expressed His -tagged O-GlcNAcase enzyme in
the supernatant was purified on His-Trap HP column
Amersham Biosciences) according to the manufacture’s
6
3.8. v-O-GlcNAcase assay and enzyme kinetics with
FDGlcNAc
(
protocol with minor variations. In this protocol, phos-
phate buffer (pH 7.4) containing 40 mM imidazole was
used as a binding buffer and the column loaded with
sample was washed successively with the binding buffer
One unit of v-O-GlcNAcase is defined as the amount
that catalyzes the formation of 1.0 lmol of FMGlcNAc
and 1.0 lmol of N-acetyl-D-gluocosamine per minute at
pH 6.5 at 37 ꢀC. The enzymatic reaction was initiated by
adding 4 lL of partially purified recombinant v-O-
GlcNAcase (1.19 lg) to the solution containing the sub-
strate, 5, which is varied (0.2–20 mM) in 20 lL of 0.5 M
citrate/phosphate buffer, pH 6.5, in a final volume of
100 lL. The assays were terminated and quantitated as
described for the O-GlcNAcase. Assays were performed
in triplicate.
(
6
20 mL) and phosphate buffer (pH 7.4) containing
0 mM concentration of imidazole (20 mL). Elution
with phosphate buffer (pH 7.4) containing 300 mM
imidazole was performed at 1.0 mL/min and the eluate
was collected in 1 mL fractions. Fractions were assayed
for protein by measuring absorbance at 280 nm. Frac-
tions containing protein were further assayed for O-
GlcNAcase activity using the protocol described below
in this section. Fractions showing enzyme activity were
pooled and concentrated using a Centricon 100 micro-
concentrator (Amicon). Small amount of concentrated
enzyme fraction was then reconstituted in 20 mM
Tris–HCl, pH 7.5, 1 mM DTT aqueous solution. Pierce
BCA protein assay reagent was used to estimate protein
concentration by using bovine serum albumin (BSA) as
a standard protein. The purified shorter splice variant of
O-GlcNAcase (v-O-GlcNAcase) was prepared by the
same protocol described above except using the phos-
phate buffer containing 10 mM concentration of imid-
azole as a binding buffer and washing thoroughly with
3.9. O-GlcNAcase assay and enzyme kinetics with
pNP-b-GlcNAc
One unit of O-GlcNAcase is defined as the amount that
hydrolyzes 1.0 lmol of pNP-b-GlcNAc to p-nitrophenol
and N-acetyl-D-gluocosamine per minute at pH 6.5 at
37 ꢀC. pNP-b-GlcNAc (14 mM) was prepared in 5%
DMSO in water and used as a stock solution. Purified
recombinant O-GlcNAcase (0.18 lg) was added to
increasing concentrations of pNP-b-GlcNAc (0.3–
8 mM) in 20 lL of 0.5 M citrate/phosphate buffer,
pH 6.5, in a final volume of 100 lL at the start of the
incubation at 37 ꢀC. The assays were terminated by
adding 900 lL of 0.5 M Na CO solution after 5 min
2
6
0 mL of the binding buffer and eluting successively with
0, 100, 300 mM imidazole phosphate buffer (pH 7.4).
Enzyme containing fractions were concentrated by using
a Centricon 50 microconcentrator (Amicon) after frac-
tions showing enzyme activity were pooled. Purity of
the enzymes was confirmed by Ponceau S staining of
the membrane (Sigma) and Western blot analysis using
anti His-Tag antibody (Abcam) (see Fig. 1).
2
3
incubation. The p-phenolate ion generated from the
hydrolysis of pNP-b-GlcNAc by the action of enzyme
was spectrophotometrically measured at 400 nm on
the UV 160U spectrophotometer (Shimadzu). Assays
were performed in triplicate.
3.7. O-GlcNAcase assay and enzyme kinetics with
FDGlcNAc
3.10. Molar fluorescence coefficient of FMGlcNAc
Standard dilutions made for calculating the molar fluo-
rescence coefficient of FMGlcNAc on the Wallac 1420
fluorometer were as follows: 0.01, 0.05, 0.1, 0.5, 1.0,
2.0, 3.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10, and 12 lM in 20 lL
of 0.5 M citrate/phosphate buffer, pH 6.5, 900 lL of
0.5 M Na CO solution in a total volume of 1 mL.
One international unit (1 U) of O-GlcNAcase is defined
as the amount that catalyzes the formation of 1.0 lmol
of FMGlcNAc and 1.0 lmol of N-acetyl-D-glucosamine
per minute at pH 6.5 at 37 ꢀC. Purified recombinant O-
GlcNAcase (0.17 lg, 2 lL) was added to the solution
2
3

E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
981
The assay solution (200 lL) was transferred into a 96-
well plate and fluorescence was measured at kex
85 nm and at kem = 535 nm on the Wallac 1420 fluoro-
fugation at 20,000g for 15 min was stored in aliquots at
ꢁ80 ꢀC. PUGNAC treated HeLa cell extracts were pre-
pared using MPER protein extraction reagent (Pierce)
according to the manufacturer’s protocol from a
10 mL suspension of cultured cells. Protein (20 lg) was
incubated with 0 (control), 7 lg, 5 lg of O-GlcNAcase,
and 5 lg of v-O-GlcNAcase in 0.05 M citrate/phosphate
buffer, pH 6.5, in a final volume of 22.5 lL at 37 ꢀC for
3 h. Reactions were stopped by adding 7.5 lL of Nu-
=
4
meter with a continuous wave lamp set to 5028 in the
stabilized energy mode and with the 4 mm diameter of
circular aperture and 1 s of counting time. The plot of
the concentrations of the standard versus their corre-
sponding fluorescence intensities was constructed and
the line-of-best-fit through the experimental points was
obtained. The slope of this straight line is the molar
fluorescence coefficient.
ꢁ
PAGE LDS sample buffer (4·). Samples were loaded
onto a 10% NuPAGE pre-cast gel and proteins were
separated by electrophoresis and then transferred to a
nitrocellulose membrane for immunoblotting. The blot
was probed with O-GlcNAc specific antibody RL2
(mouse IgG1, Affinity Bioreagents, Inc.) at 1:1000 dilu-
tion. Peroxidase-conjugated goat anti-mouse IgG (Jack-
son Immunoresearch Laboratories, Inc.) was used as the
secondary antibody. The blot was visualized with
The calculated molar fluorescence constant of
FMGlcNAc is 119,820 units/lM.
3
.11. Molar absorptivity of pNP-b-GlcNAc
pNP-b-GlcNAc (10 mM) was prepared in 5% DMSO in
water and used as a stock solution. 3, 7, 10, 20, 30, 40,
ꢁ
5
0, 65, 80, 100, 110, and 120 lM of pNP-b-GlcNAc were
enhanced SuperSignal West Pico Chemiluminescent
made in 20 lL of 0.5 M citrate/phosphate buffer,
pH 6.5, 900 lL of 0.5 M Na CO solution in a total vol-
Substrate (Pierce) (see Fig. 6).
2
3
ume of 1 mL. Absorbance of para-phenolate ion in a
series of dilution standards was measured at 400 nm at
Acknowledgments
2
5 ꢀC on the UV 160U spectrophotometer (Shimadzu).
Assays were performed in triplicate. The plot of the con-
centrations of the standard versus their corresponding
absorbance was constructed and the line-of-best-fit
through the experimental points was obtained. From
the slope of this straight line, calculated molar absorp-
This research was financially supported by National
Institute of Health Grants (VFJA009594). We thank
Mr. Victor Livengood who is an operator of the mass
spectrometers in the NIDDK, NIH for obtaining the
mass spectral data and allowing us to use LC–ESI
instrument for monitoring the time-course of the en-
zyme reaction.
ꢁ
1
tivity of pNP-b-GlcNAc is 18.8 mM cm
.
3
.12. Substrate specificity
Either 2 lL of purified recombinant O-GlcNAcase
0.17 lg) or 4 lL of partially purified recombinant
(
References
v-O-GlcNAcase (1.19 lg) was added to the solution
containing 30 lM of FDGalNAc (Marker Gene Tech-
nologies, Inc.) in 20 lL of 0.5 M citrate/phosphate,
pH 6.5, in a final volume of 100 lL and incubated at
1
2
. Jackson, S. P.; Tjian, R. Cell 1988, 55, 125–133.
. Jackson, S. P.; Tjian, R. Proc. Natl. Acad. Sci. U.S.A.
1
989, 86, 1781–1785.
. Kelly, W. G.; Dahmus, M. E.; Hart, G. W. J. Biol. Chem.
993, 268, 10416–10424.
3
3
9
7 ꢀC for 30 min. The assays were terminated by adding
00 lL of 0.5 M Na CO solution. The assay solution
1
4. Holt, G. W.; Haltiwanger, R. S.; Torres, C.-R.; Hart,
2
3
G. W. J. Biol. Chem. 1987, 262, 14847–14850.
(
200 lL) was transferred into a 96-well plate and fluores-
5
. Dong, L.-Y.; Xu, Z.-S.; Chevrier, M. R.; Cotter, R. J.;
cence was measured at kex = 485 nm and at kem
=
Cleveland, D. W.; Hart, G. W. J. Biol. Chem. 1993, 268,
5
35 nm on the Wallac 1420 fluorometer.
1
6679–16687.
6
7
8
9
. Chou, T.-Y.; Dang, C. V.; Hart, G. W. Proc. Natl. Acad.
Sci. U.S.A. 1995, 92, 4417–4421.
. Lubas, W. A.; Hanover, J. A. J. Biol. Chem. 2000, 275,
3.13. Preparation of HT-29 and HeLa cell extracts,
enzyme activity assay, and Western blotting
1
0983–10988.
. Meikrantz, W.; Smith, D. M.; Sladicka, M. M.; Schlegel,
R. A. J. Cell Sci. 1991, 98, 303–307.
. Walgren, J. L.; Vincent, T. S.; Schey, K. L.; Buse, M. G.
Am. J. Physiol.: Endocrinol. Metab. 2003, 284, E424–
E434.
The suspension of 10 mL culture of PUGNAC treated
HT-29 cells was centrifuged at 2500g for 10 min. The
pellet was suspended in 0.3 mL of lysis buffer containing
2
0 mM Tris–HCl, pH 8.0, 1 mM DTT, 0.1% Triton
1
1
0. Roquemore, E. P.; Chevrier, M. R.; Cotter, R. J.; Hart, G.
W. Biochemistry 1996, 35, 3578–3586.
1. Lefebvre, T.; Cieniewsi, C.; Lemoine, J.; Guerardel, Y.;
Leroy, Y.; Zanetta, J. P.; Michalski, J. C. Biochem. J.
2001, 360, 179–188.
X-100, an EDTA-free protease inhibitor cocktail (1 tab-
let/10 mL). The suspension was incubated at rt for 5 min
and sonicated on ice (4 · 10 s, setting 3, Misonix Ultra-
sonic Processor). The supernatant obtained after centri-

982
E. J. Kim et al. / Carbohydrate Research 341 (2006) 971–982
1
1
1
1
1
1
1
1
2. Kearse, K. P.; Hart, G. W. Proc. Natl. Acad. Sci. U.S.A.
991, 88, 1701–1705.
3. Chou, C. F.; Smith, A. J.; Omary, M. B. J. Biol. Chem.
992, 267, 3901–3906.
4. Ku, N. O.; Omary, M. B. Exp. Cell Res. 1994, 211, 24–
5.
5. Favreau, C.; Worman, H. J.; Wozniak, R. W.; Frappier,
T.; Courvalin, J. C. Biochemistry 1996, 35, 8035–8044.
6. Cheng, X.; Cole, R. N.; Zaia, J.; Hart, G. W. Biochemistry
34. Comtesse, N.; Maldener, E.; Meese, E. Biochem. Biophys.
1
Res. Commun. 2001, 283, 634–640.
35. Schultz, J.; Pils, B. FEBS Lett. 2002, 529, 179–182.
36. Toleman, C.; Paterson, A. J.; Whisenhunt, T. R.; Kudlow,
J. E. J. Biol. Chem. 2004, 279, 53665–53673.
37. Macauley, M. S.; Whitworth, G. E.; Debowski, A.; Chin,
D.; Vocadlo, D. J. J. Biol. Chem. 2005, 280, 25313–25322.
38. Hanover, J. A. FASEB J. 2001, 15, 1865–1876.
39. Rigden, D. J.; Jedrzejas, M. J.; de Mello, L. V. FEBS Lett.
2003, 544, 103–111.
40. Wells, L.; Gao, Y.; Mahoney, J. A.; Voseller, K.; Chen,
C.; Rosen, A.; Hart, G. W. J. Biol. Chem. 2002, 277, 1755–
1761.
41. Glasgow, L. R.; Paulson, J. A.; Hill, R. L. J. Biol. Chem.
1977, 252, 8615–8623.
1
3
2
000, 39, 11609–11620.
7. Comer, F. I.; Hart, G. W. Biochemistry 2001, 40, 7845–
852.
8. Chou, T. Y.; Hart, G. W.; Dang, C. V. J. Biol. Chem.
995, 270, 18961–18965.
7
1
9. McClain, D. A.; Lubas, W. A.; Cooksey, R. C.; Hazel, M.;
Parker, G. J.; Love, D. C.; Hanover, J. A. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 10695–10699.
42. Yegorov, A. M.; Markaryan, A. N.; Vozniy, Y. V.;
Cherednikova, T. V.; Demcheva, M. V.; Berezin, I. V.
Anal. Lett. 1988, 21, 193–209.
43. Rakhmanova, V. A.; MacNonald, R. C. Anal. Biochem.
1998, 257, 234–237.
2
2
0. Shaw, P.; Freeman, J.; Bovey, R.; Iggo, R. Oncogene 1996,
12, 921–930.
1. Arnold, C. S.; Johnson, G. V.; Cole, R. N.; Dong, D. L.;
Lee, M.; Hart, G. W. J. Biol. Chem. 1996, 271, 28741–
44. MacGregor, G. R.; Nolan, G. P.; Fiering, S.; Roederer,
M.; Herzenberg, L. A. In Gene Expression in Vivo;
Murray, E. J., Walker, J. M., Eds.; Humana Press:
Clifton, NJ, 1990; pp 217–236.
28744.
2
2
2
2
2
2
2
2
3
2. Vosseller, K.; Wells, L.; Lane, M. D.; Hart, G. W. Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 5313–5318.
3. Haltiwanger, R. S.; Blomberg, M. A.; Hart, G. W. J. Biol.
Chem. 1992, 267, 9005–9013.
4. Hartweck, L. M.; Scott, C. L.; Olszewski, N. E. Genetics
45. Young, D. C.; Kingsley, S. D.; Ryan, K. A.; Dutko, F. J.
Anal. Biochem. 1993, 215, 24–30.
46. Berg, M.; Undisz, K.; Thiericke, R.; Moore, T.; Posten, C.
J. Biomol. Screen. 2000, 5, 71–76.
47. Kasai, K.; Uchida, R.; Yamaji, N. Chem. Pharm. Bull.
1993, 41, 314–318.
48. Arias, E. B.; Kim, J.; Cartee, G. D. Diabetes 2004, 53,
921–930.
49. Park, S. Y.; Ryu, J.; Lee, W. Exp. Mol. Med. 2005, 37,
220–229.
50. Arias, E. B.; Cartee, G. D. Acta Physiol. Scand. 2005, 183,
281–289.
51. Perreira, M.; Kim, E. J.; Thomas, C. J.; Hanover, J. A.
Bioorg. Med. Chem. 2006, 14, 837–846.
52. Lehman, D. M.; Fu, D. J.; Freeman, A. B.; Hunt, K. J.;
Leach, R. J.; Johnson-Pais, T.; Hamlington, J.; Dyer, T.
D.; Arya, R.; Abboud, H.; Goring, H. H.; Duggirala, R.;
Blangero, J.; Konrad, R. J.; Stern, M. P. Diabetes 2005,
54, 1214–1221.
53. Horton, D. In Methods in Carbohydrate Chemistry;
Whistler, R. L., BeMiller, J. N., Eds.; Academic: New
York, 1972; pp 282–285.
2002, 161, 1279–1291.
5. Lubas, W. A.; Frank, D. W.; Krause, M.; Hanover, J. A.
J. Biol. Chem. 1997, 272, 9316–9324.
6. Kreppel, L. K.; Hart, G. W. J. Biol. Chem. 1999, 274,
32015–32022.
7. Gao, Y.; Wells, L.; Comer, F. I.; Parker, G. J.; Hart,
G. W. J. Biol. Chem. 2001, 276, 9838–9845.
8. Wrabl, J. O.; Grishin, N. V. J. Mol. Biol. 2001, 314, 365–
3
74.
9. Dong, D. L.; Hart, G. W. J. Biol. Chem. 1994, 269, 19321–
9330.
1
0. Jinek, M.; Rehwinkel, J.; Lazarus, B. D.; Izaurralde, E.;
Hanover, J. A.; Conti, E. Nat. Struct. Mol. Biol. 2004, 11,
1001–1007.
3
1. Braidman, I.; Carroll, M.; Dance, N.; Robinson, D.;
Poenaru, L.; Weber, A.; Dreyfus, J. C.; Overdijk, B.;
Hooghwinkel, G. J. FEBS Lett. 1974, 41, 181–184.
3
2. Overdijk, B.; Van der Kroef, W. M.; Van Steijn, G. J.;
Lisman, J. J. Biochim. Biophys. Acta 1981, 659, 255–266.
3. Heckel, D.; Comtesse, N.; Brass, N.; Blin, N.; Zang, K.
D.; Meese, E. Hum. Mol. Genet. 1998, 7, 1859–1872.
3
54. Kang, D. O.; Lee, G.; Kim, E. J.; Hanover, J. A., in
preparation.