O-GlcNAcase catalyzes cleavage of thioglycosides without general acid catalysis

Article abstract of DOI:10.1021/ja0567687

O-GlcNAcase catalyzes the removal of N-acetylglucosamine residues from serine and threonine residues of post-translationally modified proteins using a catalytic mechanism involving substrate-assisted catalysis and general acid/base catalysis. Since thioglycosides are widely perceived as resistant to hydrolysis by glycosidases, it was surprising to find that O-GlcNAcase also catalyzes the efficient hydrolysis of S-glycosides. Bronsted analyses and pH-activity studies of the O-GlcNAcase-catalyzed hydrolysis of a series of aryl S- and O-glycosides reveal that O-GlcNAcase effects hydrolysis of thioglycosides without the assistance of general acid catalysis. α-Deuterium kinetic isotope effects for O- and S-glycosides, as well as Taft-like analyses using N-fluoroacetyl-β-glycosides, suggest that O-GlcNAcase accomplishes hydrolysis of thioglycosides by stabilizing late transition states. For S-glycosides this transition state shows greater nucleophilic participation from the 2-acetamido group than for O-glycosides. The rate constants governing the O-GlcNAcase-catalyzed hydrolysis of O- and S-glycosides as compared to those previously determined for the spontaneous hydrolysis of structurally similar O,O- and O,S-acetals show a similar ratio. O-GlcNAcase therefore demonstrates similar catalytic proficiency toward both O- and S-glycosides. We conclude that O-GlcNAcase is a bifunctional catalyst capable of efficiently cleaving thioglycosides without general acid catalysis, an observation that may have biological implications. Copyright

Full text of DOI:10.1021/ja0567687

Published on Web 11/16/2005
O-GlcNAcase Catalyzes Cleavage of Thioglycosides without General Acid Catalysis
Matthew S. Macauley, Keith A. Stubbs, and David J. Vocadlo*
Department of Chemistry, Simon Fraser UniVersity, 8888 UniVersity DriVe, Burnaby, BC, Canada V5A 1S6
Received October 10, 2005; E-mail: dvocadlo@sfu.ca
Thioglycosides are widely perceived as being impervious to
glycosidase-catalyzed hydrolysis.1 Accordingly, thioglycosides have
gained widespread use as inhibitors of these enzymes,2 ligands for
affinity chromatography,3 and have also been suggested for use as
therapeutics.4 The stability of thioglycosides to nearly all glycosi-
Figure 1. O-GlcNAcase uses a catalytic mechanism involving substrate-
assisted catalysis from the 2-acetamido group of the substrate to form a
dases is so marked that a number of co-crystal structures in which
the thioglycosidic linkage is bound intact across the active site have
been obtained.5 The inability of most glycosidases to catalyze
hydrolysis of thioglycosides6 has been postulated to stem, in part,
from the poor hydrogen bonding ability of the sulfur atom, which
renders S,O-acetals unable to benefit significantly from general acid
catalysis.7a-c
transient oxazoline intermediate. Departure of oxygen leaving groups (X
) O) is aided by a general acid catalytic residue within the enzyme. For
leaving groups bearing sulfur (X ) S), there is no acid catalysis.
In Nature, thioglycosidase activity has been identified in extracts
from several organisms.8 Nevertheless, the only glycosidases with
well-defined thioglycosidase activity having been identified are the
plant myrosinases.9 Elegant studies have shown that, unlike all other
glycosidases which have a highly conserved general acid catalytic
residue to facilitate departure of an alkoxide leaving group,10 the
family 1 myrosinases lack such a group.11 Accordingly, myrosinases
cleave activated S-glycosides bearing an O-sulfated thiohydroxamic
acid aglycone for which a general acid cannot be expected to assist
in catalysis.
Recently, we have shown that family 84 human O-GlcNAcase
uses a catalytic mechanism involving anchimeric assistance (Figure
1).12 This enzyme cleaves GlcNAc from post-translationally modi-
fied intracellular proteins.13 In light of the apparent stability of thio-
glycosides to glycosidases,6 we were surprised to observe that
human O-GlcNAcase cleaved pNP-S-GlcNAc with comparable effi-
ciency as pNP-O-GlcNAc.14 Indeed, the ratio of the second-order
rate constants [(Vmax/[E]oKM)O/(Vmax/[E]oKM)S] for the O-GlcNA-
case-catalyzed hydrolysis of these substrates is only ≈20 (Figure
2A). To address whether this large enzyme might contain a second
active site, we used NAG-thiazoline, a competitive inhibitor of the
O-glycoside activity of O-GlcNAcase,12 and found identical KI
values for the cleavage of both pNP-S-GlcNAc and pNP-O-GlcNAc.
Furthermore, we assayed the O-GlcNAcase-catalyzed hydrolysis
of 4-methylthioumbelliferyl 2-deoxy-2-N-fluoroacetyl-â-D-glucoside
(Figure 2B). The steep negative slope in the Taft-like linear free
energy analysis shows that the 2-acetamido group of the substrate
participates in catalysis to a greater extent for the S-glycosides than
found earlier for the corresponding O-glycosides.12 Together, these
results reveal that O-GlcNAcase uses the same active site and
general catalytic mechanism to cleave both S- and O-glycosides.
Studies on the spontaneous hydrolysis of benzaldehyde O-ethyl
S-phenyl acetal have previously shown that cleavage of the C-S
bond in S,O-acetals does not benefit significantly from general acid
catalysis, while the corresponding O-acetals do.7 To evaluate the
extent of acid catalysis provided by O-GlcNAcase for S- and
O-glycosides, we assayed two series of substrates bearing different
substituted thiophenol or phenol leaving groups. The resulting
Brønsted analyses provide insight into the development of effective
charge at the exocyclic heteroatom in the transition state. For both
series, strong linear relationships between the second-order rate
Figure 2. O-GlcNAcase uses substrate-assisted catalysis to effect S- and
O-glycoside cleavage. (A) Initial velocities from the cleavage of pNP-S-
GlcNAc, X ) S (b), and pNP-O-GlcNAc, X ) O (0). Inset: Detail of the
region showing cleavage of the pNP-S-GlcNAc. (B) Taft-like analyses of
(Vmax/[E]oKM) versus the Taft parameter (σ*) for O-GlcNAcase-catalyzed
hydrolysis of pNP-O-GlcNAc (0, data taken from ref 12) or pNP-S-GlcNAc
(b) substrates bearing different N-fluoroacetyl groups. O-GlcNAcase does
not use acid catalysis to facilitate cleavage of S-glycosides but does for
O-glycosides. (C) Brønsted plots for O-linked (0) and S-linked (b)
substrates. (B) pH dependence for catalysis of pNP-O-GlcNAc (0) and pNP-
S-GlcNAc (b).
constant for each substrate and the pKa value of its corresponding
leaving group are observed (Figure 2C). The shallow slope (âlg-
(V/K) ) -0.11 ( 0.01) observed with the O-glycosides indicates
little negative charge accumulation on the exocyclic oxygen and
suggests two possible mechanistic scenarios. The first is a late
transition state wherein both cleavage of the glycosidic bond and
proton donation are significantly advanced. The second possibility
is an early transition state, with both little C-O bond cleavage
and little proton donation. The first scenario is most likely since it
is consistent with the general structures of transition states for the
glycosidase superfamily,10 including family 20 hexosaminidases,15
and agrees with the large R-D(V)-KIE values we describe below.
In marked contrast, for the series of S-glycosides a steep slope is
seen (âlg(V/K) ) -0.93 ( 0.06), indicating pronounced negative
charge accumulation on the exocyclic sulfur atom. Such a steep
negative slope suggests that the O-GlcNAcase-catalyzed hydrolysis
of thioglycosides involves little, if any, general acid catalysis, and
that the reaction proceeds through a transition state in which the
9
17202
J. AM. CHEM. SOC. 2005, 127, 17202-17203
10.1021/ja0567687 CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
leaving group has essentially dissociated from the anomeric center.
Interestingly, such large negative âlg(V/K) values (≈ -1) have been
observed for wild-type retaining glycosidases as well as for mutants
in which the acid/base carboxyl group has been deleted.10
There have been reports identifying S-glycoside-modified pro-
teins in mammals.18 Although it remains to be determined if the
O-GlcNAc transferase (OGTase) is capable of modifying a cysteine
residue with GlcNAc, it cannot be ruled out, especially considering
that several O-glycosyl transferases catalyze glycosyl transfer to
thiol acceptors with only approximately 100-fold lower catalytic
efficiency.19 These observations open an interesting possibility.
O-GlcNAcase may have evolved some catalytic efficiency toward
S-glycosides as a means of slowly degrading potentially toxic
S-GlcNAc-modified proteins or peptides that may arise from
catalytic promiscuity of OGTase. Alternatively, this thioglycosidase
activity could be serendipitous, and future studies may reveal similar
activities in other glycosidase families.
Acknowledgment. We thank NSERC, PENCE, and the Canada
Research Chairs program for support, NSERC and MSFHR for
fellowships to M.S.M., S. G. Withers for (1-2H)-GlcNAc and
(1-2H)-pNP-O-GlcNAc, J. Hanover for DNA encoding O-GlcNA-
case DNA, and A. J. Bennet for discussions and access to
equipment.
Two further sets of data support the view that no acid catalysis
is used by O-GlcNAcase for cleavage of S-glycosides. The first
set is the pH profiles for the cleavage of pNP-S-GlcNAc and
pNP-O-GlcNAc (Figure 2D). The bell-shaped curve for cleavage
of pNP-O-GlcNAc is characteristic of the very large majority of
glycosidases and arises from the ionization of two key catalytic
residues within the active site of O-GlcNAcase (Figure 1).10 The
pH profile for pNP-S-GlcNAc, on the other hand, reveals that only
one ionization is important for catalysis. Such an asymmetric pH
profile resembles those of some glycosidases in which the acid/
base residue has been mutated to a nonionizable group.10 Second,
a mutant of O-GlcNAcase lacking the general acid/base residue
catalyzes the hydrolysis of S- and O-glycosides at nearly identical
rates.16 Taken together, these data provide strong evidence that O-
GlcNAcase uses acid catalysis to facilitate cleavage of O-glycosides
but not S-glycosides, an observation entirely in accord with previous
studies showing that nonenzymatic hydrolysis of S,O-acetals also
does not benefit significantly from general acid catalysis.7
Supporting Information Available: Details of assays, preparation
of substrates, and tables of kinetic parameters. This material is available
free of charge via the Internet at http://pubs.acs.org.
This lack of general acid catalysis toward sulfur is likely common
for all glycosidases, yet this deficiency cannot, on its own, account
for the apparent inability of most glycosidases to cleave S-
glycosides. Because most glycosidases cannot cleave even highly
activated thioglycosides,6 these enzymes must generally be unable
to effectively stabilize a transition state bearing a sulfur atom in
the exocyclic position. This situation does not apply to O-
GlcNAcase. Indeed, the [(Vmax/[E]oKM)O/(Vmax/[E]oKM)S] ratio for
the hydrolysis of phenyl-O-GlcNAc versus that calculated for
phenyl-S-GlcNAc is only 900. This value is comparable with a ratio
of 700 previously measured for the spontaneous hydrolysis of
benzaldehyde O-ethyl-O-phenyl acetal versus that for the related
O-ethyl S-phenyl acetal.7c Together, these results reveal that
O-GlcNAcase is a proficient bifunctional catalyst, with likely similar
(kcat/KM)/kuncat values for both S- and O-glycosides.
References
(1) Driguez, H. Top. Curr. Chem. 1997, 187, 85-116.
(2) Barr, B. K.; Holewinski, R. J. Biochemistry 2002, 41, 4447-4452.
(3) Piyachomkwan, K.; Penner, M. H. Anal. Biochem. 1998, 255, 223-235.
(4) Witczak, Z. J. Curr. Med. Chem. 1999, 6, 165-178.
(5) (a) Czjzek, M.; Cicek, M.; Zamboni, V.; Burmeister, W. P.; Bevan, D.
R.; Henrissat, B.; Esen, A. Biochem. J. 2001, 354, 37-46. (b) Sulzen-
bacher, G.; Driguez, H.; Henrissat, B.; Schulein, M.; Davies, G. J.
Biochemistry 1996, 35, 15280-15287.
(6) For the large majority of cases, including the functionally related human
HexB (see Supporting Information), glycosidases have not been found to
catalyze hydrolysis of any thioglycoside. For other examples, see: (a)
Jones, C. S.; Kosman, D. J. J. Biol. Chem. 1980, 255, 11861-11869. (b)
Chitlaru, E.; Roseman, S. J. Biol. Chem. 1996, 271, 33433-33439. (c)
Parry, N. J.; Beever, D. E.; Owen, E.; Vandenberghe, I.; Van Beeumen,
J.; Bhat, M. K. Biochem. J. 2001, 353, 117-127, and refs 2, 3, and 5.
Interestingly, â-D-glucosidases likely related to family 1 myrosinases show
500-25 000-fold decreases in (Vmax/[E]oKM) for pNP-S-Glc versus pNP-
O-Glc: (d) Grover, A. K.; Cushley, R. J. Biochim. Biophys. Acta 1977,
482, 109-124. (e) Mega, T.; Matsushima, Y. J. Biochem. 1983, 94, 1637-
1647. (f) De Gussem, R. L.; Aerts, G. M.; Claeyssens, M.; De Bruyne,
C. K. Biochim. Biophys. Acta 1978, 525, 142-153.
An interesting question is what makes O-GlcNAcase an efficient
catalyst toward both S- and O-glycosides, with comparable rate
accelerations, while most glycosidases appear unable to cleave
S-glycosides at all.1,6 Previously, large R-D KIE values have been
observed for the pH-independent spontaneous hydrolysis of ben-
zaldehyde O-ethyl S-phenyl thioacetals (1.13 ( 0.02) and for the
acetic acid catalyzed hydrolysis of the corresponding O-acetal (1.17
( 0.03), suggesting that both C-S and C-O bond cleavage is
significantly advanced in these transition states.7b,d We therefore
measured R-D(V)-KIE values for the O-GlcNAcase-catalyzed
hydrolysis of (1-2H)-pNP-O-GlcNAc and (1-2H)-pNP-S-GlcNAc.
The kH/kD ratio was a very large 1.14 ( 0.02, for the O-glycoside
and 1.20 ( 0.02 for the S-glycoside. This latter value is the largest
measured to date for the glycosylation step of any glycosidase.10
These large KIE’s reveal Very dissociative transition states and
suggest that the transition state for S-glycoside cleavage may be
later than that for O-glycoside cleavage. Such a view is consistent
with the increased participation of the 2-acetamido group, as shown
in the Taft analyses. However, direct comparisons between these
KIE’s are difficult since they likely report on different steps of the
mechanism. As well, realization of the true intrinsic KIE’s may be
obscured by binding isotope effects17 or by contributions from
external commitments. Regardless, these values are large, and so
it appears that O-GlcNAcase has evolved to stabilize unusually late
transition states suited to facilitating hydrolysis of S-glycosides
without acid catalysis and O-glycosides with acid catalysis.
(7) (7) (a) Jensen, J. L.; Jencks, W. P. J. Am. Chem. Soc. 1979, 101, 1476-
1488. (b) Fife, T. H.; Przystas, T. J. J. Am. Chem. Soc. 1980, 102, 292-
299. (c) Ferraz, J. P.; Cordes, E. H. J. Am. Chem. Soc. 1979, 101, 1488-
1491. (d) Lamaty, G.; Menut, C.; Nguyen, M. Recl. TraV. Chim. Pays-
Bas 1984, 103, 75-77.
(8) (a) Reese, E. T.; Clapp, R. C.; Mandels, M. Arch. Biochem. Biophys. 1958,
75, 228-242. (b) Meulenbeld, G. H.; Hartmans, S. Appl. Microbiol.
Biotechnol. 2001, 56, 700-706. (c) Goodman, I.; Fouts, J. R.; Bresnick,
E.; Menegas, R.; Hitchings, G. H. Science 1959, 130, 450-451.
(9) Family 4 enzymes also show thioglycosidase activity, although they use
an elimination mechanism: (a) Varrot, A. et al. J. Mol. Biol. 2005, 346,
423-435. Thioglycosidase activity found for a sialidase from T. rangeli
may also be due to an elimination mechanism: (b) Amaya, M. F.; Busch-
iazzo, A.; Nguyen, T.; Alzari, P. M. J. Mol. Biol. 2003, 325, 773-784.
(10) Davies, G. J.; Sinnott, M. L.; Withers, S. G. In ComprehensiVe Biological
Catalysis: A Mechanistic Reference; Sinnott, M. L., Ed.; Academic
Press: San Diego, 1998; Vol. 1, pp 119-208.
(11) Burmeister, W. P.; Cottaz, S.; Rollin, P.; Vasella, A.; Henrissat, B. J.
Biol. Chem. 2000, 275, 39385-39393.
(12) Macauley, M. S.; Whitworth, G. E.; Debowski, A. W.; Chin, D.; Vocadlo,
D. J. J. Biol. Chem. 2005, 280, 25313-25322.
(13) Dong, D. L.; Hart, G. W. J. Biol. Chem. 1994, 269, 19321-19330.
(14) O-GlcNAcase could not be purified on a p-aminothiophenyl affinity
column, and attempts at purification degraded the affinity column:
Overdijk, B.; Van der Kroef, W. M.; Van Steijn, G. J.; Lisman, J. J.
Biochim. Biophys. Acta 1981, 659, 255-266. Others have suggested that
O-GlcNAcase does not cleave pNP-S-GlcNAc at all.13
(15) Vocadlo, D. J.; Withers, S. G. Biochemistry 2005, 44, 12809-12818.
(16) Cetinbas, N.; Macauley, M. S.; Stubbs, K. A.; Drapala, R.; Vocadlo, D.
J. Manuscript submitted to J. Biol. Chem.
(17) Lewis, B. E.; Schramm, V. L. J. Am. Chem. Soc. 2003, 125, 4785-4798.
(18) (a) Lote, C. J.; Weiss, J. B. FEBS Lett. 1971, 16, 81-85. (b) Weiss, J.
B.; Lote, C. J.; Bobinski, H. Nat. New Biol. 1971, 234, 25-26.
(19) Rich, J. R.; Szpacenko, A.; Palcic, M. M.; Bundle, D. R. Angew. Chem.,
Int. Ed. 2004, 43, 613-615.
JA0567687
9
J. AM. CHEM. SOC. VOL. 127, NO. 49, 2005 17203