Synthesis and use of mechanism-based protein-profiling probes for retaining β-D-glucosaminidases facilitate identification of Pseudomonas aeruginosa NagZ

Article abstract of DOI:10.1021/ja0763605

The NagZ class of retaining exo-glucosaminidases play a critical role in peptidoglycan recycling in Gram-negative bacteria and the induction of resistance to β-lactams. Here we describe the concise synthesis of 2-azidoacetyl-2-deoxy-5-fluoro-β-D-glucopyranosyl fluoride as an activity-based proteomics probe for profiling these exo-glycosidases. This active-site directed reagent covalently inactivates this class of retaining N-acetylglucosaminidases with exquisite selectivity by stabilizing the glycosyl-enzyme intermediate. Inactivated Vibrio cholerae NagZ can be elaborated with biotin or a FLAG-peptide epitope using the Staudinger ligation or the Sharpless-Meldal click reaction and detected at nanogram levels. This ABPP enabled the profiling of the Pseudomonas aeruginosa proteome and identification at endogenous levels of a tagged protein with properties consistent with those of PA3005. Cloning of the gene encoding this hypothetical protein and biochemical characterization enabled unambiguous assignment of this hypothetical protein as a NagZ. The identification and cloning of this NagZ may facilitate the development of strategies to circumvent resistance to β-lactams in this human pathogen. As well, this general strategy, involving such 5-fluoro inactivators, may prove to be of general use for profiling proteomes and identifying glycoside hydrolases of medical importance or having desirable properties for biotechnology.

Full text of DOI:10.1021/ja0763605

Published on Web 12/08/2007
Synthesis and Use of Mechanism-Based Protein-Profiling
Probes for Retaining â-D-Glucosaminidases Facilitate
Identification of Pseudomonas aeruginosa NagZ
Keith A. Stubbs,† Adrian Scaffidi,‡ Aleksandra W. Debowski,† Brian L. Mark,§
Robert V. Stick,‡ and David J. Vocadlo*,†
Contribution from the Department of Chemistry, Simon Fraser UniVersity, 8888 UniVersity
DriVe, Burnaby, B.C. Canada, V5A 1S6, School of Biomedical, Biomolecular and
Chemical Sciences, UniVersity of Western Australia, 35 Stirling Highway, Crawley,
Western Australia, Australia, 6009, and Department of Microbiology, UniVersity
of Manitoba, Rm 418 Buller Building, Winnipeg, Manitoba, Canada, R3T 2N2
Received August 23, 2007; E-mail: dvocadlo@sfu.ca
Abstract: The NagZ class of retaining exo-glucosaminidases play a critical role in peptidoglycan recycling
in Gram-negative bacteria and the induction of resistance to â-lactams. Here we describe the concise
synthesis of 2-azidoacetyl-2-deoxy-5-fluoro-â-D-glucopyranosyl fluoride as an activity-based proteomics
probe for profiling these exo-glycosidases. This active-site directed reagent covalently inactivates this class
of retaining N-acetylglucosaminidases with exquisite selectivity by stabilizing the glycosyl-enzyme intermedi-
ate. Inactivated Vibrio cholerae NagZ can be elaborated with biotin or a FLAG-peptide epitope using the
Staudinger ligation or the Sharpless-Meldal click reaction and detected at nanogram levels. This ABPP
enabled the profiling of the Pseudomonas aeruginosa proteome and identification at endogenous levels of
a tagged protein with properties consistent with those of PA3005. Cloning of the gene encoding this
hypothetical protein and biochemical characterization enabled unambiguous assignment of this hypothetical
protein as a NagZ. The identification and cloning of this NagZ may facilitate the development of strategies
to circumvent resistance to â-lactams in this human pathogen. As well, this general strategy, involving
such 5-fluoro inactivators, may prove to be of general use for profiling proteomes and identifying glycoside
hydrolases of medical importance or having desirable properties for biotechnology.
Introduction
Several activity-based proteomics probes (ABPPs) directed
toward glycosidases have been developed8-13 but only those
Glycoside hydrolases are an extremely large class of enzymes
that act to cleave carbohydrate motifs. They are found through-
out nature in organisms ranging from bacteria to humans and
have been classified into over 100 families on the basis of
structural similarity.1 In light of their biological significance,
the rapid detection of proteins involved in processing glyco-
conjugates is of considerable interest.2 However, compared to
activity-based methods for profiling protease and esterase
activities,3-7 which have met with considerable success, the
methods used to profile the activities of glycan-processing
enzymes within proteomes and the discovery of new families
of these enzymes is lagging.2
incorporating a 2-deoxy-2-fluorosugar have proven useful for
probing complex proteomes such as cell lysates. The first of
these 2-deoxy-2-fluorosugar probes, directed against exo-
glycosidases, inactivates the enzyme quite slowly owing to lost
enzyme-carbohydrate interactions.12 The second, directed
against one class of endo-glycosidases, is not suitable for
labeling exo-glycosidases owing to the structure of the AB-
PP.13,14
Indeed, a major complexity in the development of probes
directed against glycoconjugate-degrading enzymes is that,
unlike proteases, which commonly have a cleft-like active site,
most carbohydrate processing enzymes are exo-acting and cleave
only the terminal ends of glycoconjugates. These enzymes
† Simon Fraser University.
‡ University of Western Australia.
§ University of Manitoba.
(8) Tsai, C. S.; Li, Y. K.; Lo, L. C. Org. Lett. 2002, 4, 3607-3610.
(1) Henrissat, B.; Davies, G. Curr. Opin. Struct. Biol. 1997, 7, 637-644.
(2) Adam, G. C.; Sorensen, E. J.; Cravatt, B. F. Mol. Cell. Proteomics 2002,
1, 781-790.
(9) Kurogochi, M.; Nishimura, S.; Lee, Y. C. J. Biol. Chem. 2004, 279, 44704-
44712.
(10) Hinou, H.; Kurogochi, M.; Shimizu, H.; Nishimura, S. Biochemistry 2005,
(3) Phillips, C. I.; Bogyo, M. Cell. Microbiol. 2005, 7, 1061-1076.
(4) Speers, A. E.; Cravatt, B. F. Chem. Biol. 2004, 11, 535-546.
(5) Ovaa, H.; van Swieten, P. F.; Kessler, B. M.; Leeuwenburgh, M. A.;
Fiebiger, E.; van den Nieuwendijk, A. M.; Galardy, P. J.; van der Marel,
G. A.; Ploegh, H. L.; Overkleeft, H. S. Angew. Chem., Int. Ed. 2003, 42,
3626-3629.
44, 11669-11675.
(11) Romaniouk, A. V.; Silva, A.; Feng, J.; Vijay, I. K. Glycobiology 2004, 14,
301-310.
(12) Vocadlo, D. J.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2004, 43, 5338-
5342.
(13) Williams, S. J.; Hekmat, O.; Withers, S. G. ChemBioChem 2006, 7, 116-
(6) Speers, A. E.; Cravatt, B. F. ChemBioChem 2004, 5, 41-47.
(7) Speers, A. E.; Adam, G. C.; Cravatt, B. F. J. Am. Chem. Soc. 2003, 125,
4686-4687.
124.
(14) Hekmat, O.; Kim, Y. W.; Williams, S. J.; He, S.; Withers, S. G. J. Biol.
Chem. 2005, 280, 35126-35135.
9
10.1021/ja0763605 CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 327-335
327

A R T I C L E S
Stubbs et al.
commonly have pocket-shaped active site architectures that
hinder the development of useful probes since there is little space
to accommodate pendent biochemical reporter groups such as
biotin. An exception are the endo-glycosidases that have a
canyon-like active site that can tolerate bulky reporter groups
attached to the probe.13,14
In previous work, the small pendant azide group attached to
the 2-fluoro sugar probe diminished, but did not prevent, binding
of the saccharide moiety by exo-galactosidases.12 In that study,
the target enzyme could be covalently inactivated by the affinity
probe bearing the azide group and then elaborated with a reporter
group by using a ligation reaction.12 As alluded to, however,
this 2-fluorosugar probe, while highly selective, was slow to
inactivate the target enzymes thereby limiting its use with
unstable enzymes or proteomes rich in proteases.
One distinct advantage in common with the serine and
cysteine proteases, however, is that most retaining exo-glycosi-
dases use a catalytic mechanism involving the transient forma-
tion of a covalent intermediate. These glycosyl-enzyme inter-
mediates can be stabilized so that they persist by using either
deoxy sugars15 or, more commonly, deoxy-fluorosugars as
mechanism-based inactivators or probes of enzyme mechan-
ism.16-18 The inclusion of a fluorine atom at C2 or C5 of the
pyranose ring has the effect of destabilizing the two transition
states that flank the covalent glycosyl-enzyme intermediate,
thereby slowing both enzyme-catalyzed steps. A key functional
requirement is that a very good leaving group be incorporated
at the anomeric center. In this way, the intermediate is kinetically
accessible since the leaving group compensates for the desta-
bilizing effect of the fluorine, yet the breakdown of the
intermediate is very slow since the excellent leaving group is
no longer present. The net effect is that these fluorosugars act
as both a recognition element and reactive group, since a stable
covalent bond is formed between the enzymic nucleophile and
fluorosugar.16,17,19,20
A key challenge for developing ABPPs for glycoside hydro-
lases is the point of attachment of the reporter group. As
described earlier, exo-glycosidases have a pocket-shaped active
site in which extensive contacts are formed between the protein
and all substrate hydroxyl groups, making it impractical to
incorporate a bulky biochemical reporter group such as biotin.
One solution to this problem has been demonstrated for several
families of retaining â-galactosidases by using a 6-azido-2,6-
dideoxy-2-fluoro-â-D-galactopyranosyl fluoride ABPP where the
pendant azide group could later be ligated to a reporter group.12
Several chemoselective ligations have been developed that
make use of the small unobtrusive azide functionality as one
reactant.21-24 The reactive partner for the azide group varies
for these different reactions; for the Staudinger ligation it is a
phosphine, for the Sharpless-Meldal click reaction it is an
alkyne, and for the strained Huisgen ligation it is a cyclooctyne
derivative.21-24 These chemoselective ligations can be carried
out under physiological conditions and, owing to the entirely
bioorthogonal nature of both partners, are so exquisitely selective
that the large concentrations of biologically relevant functional
groups do not interfere. Any of these reactive functional group
pairs can be appended to various affinity probes or ligands as
desired.
Here we present a significantly improved strategy using an
azide-derivatized 5-fluorosugar that quickly inactivates the target
exo-glycosidases and forms a stable species (Figure 1A). A
biochemical reporter group can be appended to this inactivated
enzyme by using the Staudinger ligation or Sharpless-Meldal
click reaction, enabling subsequent proteomic profiling and
purification. Recent studies in our laboratory have focused on
the enzyme NagZ,25 which is involved in the murein recycling
pathway in Gram-negative bacteria. This enzyme has been
shown to regulate the inducible expression of AmpC â-lacta-
mase in Gram-negative bacteria.26,27 NagZ and other sequence
related glucosaminidases are members of CAZy family 328,29
of glycoside hydrolases that also contains glycosidases active
on a variety of other saccharides. NagZ is an exo-glycosidase
that uses a catalytic mechanism involving the formation and
subsequent breakdown of a covalent glycosyl-enzyme inter-
mediate.20 Based on previous mechanistic studies of Vibrio
furnisii â-glucosaminidase using fluorinated saccharides20 and
structural studies of NagZ from Vibrio cholerae,25 we envisioned
that 2-azidoacetamido-2-deoxy-5-fluoro-â-D-glucopyranosyl fluo-
ride 1 (2AA5FGF) could be an efficient ABPP that would be
useful for functional proteomic analysis of organisms that
produce mechanistically related glucosaminidases.
Results and Discussion
A high-resolution crystal structure of V. cholerae NagZ
(VCNagZ) in complex with an inhibitor has recently been
reported and revealed the presence of a large pocket surrounding
the 2-acetamido group.25 This observation suggested to us that
substrates and inhibitors bearing extensions at the 2-acetamido
group should be well tolerated by the enzyme. This site therefore
seemed suitable, and synthetically accessible, for the installation
of an azide functionality.
Before embarking on a lengthy synthesis we decided to
investigate whether an azide group attached to the 2-acetamido
moiety of an N-acetyl-D-glucosaminyl substrate would be
tolerated by the enzyme. We therefore prepared the known
chromogenic substrate 4-nitrophenyl 2-azidoacetamido-2-deoxy-
â-D-glucopyranoside (pNP-GlcNAz).30 We found this compound
to be a good substrate for the â-D-glucosaminidase VCNagZ,
an enzyme that we had previously cloned.25 Compared to the
parent 4-nitrophenyl 2-acetamido-2-deoxy-â-D-glucopyranoside
(15) Roeser, K. R.; Legler, G. Biochim. Biophys. Acta 1981, 657, 321-333.
(16) Wicki, J.; Rose, D. R.; Withers, S. G. Methods Enzymol. 2002, 354, 84-
(pNP-GlcNAc; kcat/KM ) 8.6 ( 0.66 µmol mM-1 min-1 mg-1
,
105.
(17) Withers, S. G.; Rupitz, K.; Street, I. P. J. Biol. Chem. 1988, 263, 7929-
7932.
(18) Hartman, M. C. T.; Coward, J. K. J. Am. Chem. Soc. 2002, 124, 10036-
(24) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004, 126,
15046-15047.
10053.
(19) Vocadlo, D. J.; Davies, G. J.; Laine, R.; Withers, S. G. Nature 2001, 412,
835-838.
(25) Stubbs, K. A.; Balcewich, M.; Mark, B. L.; Vocadlo, D. J. J. Biol. Chem.
2007, 282, 21382-21391.
(20) Vocadlo, D. J.; Mayer, C.; He, S.; Withers, S. G. Biochemistry 2000, 39,
117-126.
(26) Votsch, W.; Templin, M. F. J. Biol. Chem. 2000, 275, 39032-39038.
(27) Cheng, Q.; Li, H.; Merdek, K.; Park, J. T. J. Bacteriol. 2000, 182, 4836-
4840.
(21) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007-2010.
(22) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596-2599.
(28) Henrissat, B.; Bairoch, A. Biochem. J. 1996, 316, 695-696.
(29) Henrissat, B.; Bairoch, A. Biochem. J. 1993, 293, 781-788.
(30) Vocadlo, D. J.; Hang, H. C.; Kim, E. J.; Hanover, J. A.; Bertozzi, C. R.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9116-9121.
(23) Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057-
3064.
9
328 J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008

ABPPs for Retaining
â
-D-Glucosaminidases
A R T I C L E S
Figure 1. (A) The activity-based detection and purification strategy using 2-azidoacetamido-2-deoxy-5-fluoro-â-D-glucopyranosyl fluoride (2AA5FGF).
Incubation of the proteome with 2AA5FGF results in rapid inactivation and covalent labeling of the catalytic nucleophile of the â-glucosaminidase. Denaturation
under acidic conditions and conjugation using either the Staudinger ligation or by the Sharpless-Meldal click reaction facilitates either profiling in lysates
or purification and identification using an immobilized streptavidin stationary phase. (B and C) Reporter groups used in this study. (B) Phosphine-FLAG and
(C) Alkyne-biotin.
kcat ) 4.4 ( 0.08 µmol min-1 mg-1, KM ) 0.51 ( 0.03 mM),
the value of the second-order rate constant for the VCNagZ-
catalyzed hydrolysis of pNP-GlcNAz is only 2.8 fold higher
harming either anomeric fluorine. Subsequent acetylation of the
intermediate amine afforded the amide 8, after which depro-
tection yielded the desired 2-acetamido-2-deoxy-5-fluoro-â-D-
glucopyranosyl fluoride 2. The 2AA5FGF ABPP 1 was prepared
similarly starting from the common intermediate 7. This time,
however, a one-pot procedure was used in which removal of
the phthalimido group was followed by chloroacetylation.
Installation of the azide moiety was effected by displacement
of the R-chloride to yield the probe 1, in good yield. It is worth
noting that previous efforts to generate such 2-acetamido 1,5-
difluoro compounds failed to yield the D-gluco epimer but rather
afforded the L-ido analogue in very modest yield.20 Furthermore,
unlike the previous study, the synthesis presented here enables
facile installation of an azide group, which is a prerequisite to
develop this ABPP. Although the yield of the two-step photo-
bromination/fluorination proceedure is modest, the overall
synthesis is facile and readily furnishes more than adequate
quantities of ABPP 1 since very small quantities of material
are required for proteomic profiling.
(pNP-GlcNAz; kcat/KM ) 24 ( 5.7 µmol mM-1 min-1 mg-1
,
kcat ) 0.48 ( 0.02 µmol min-1 mg-1, KM ) 0.02 ( 0.004 mM),
indicating that the azide moiety only modestly affects enzyme
activity (see Supporting Information).
Having established that the azide group is well tolerated by
this representative member of the family 3 exo-â-D-glu-
cosaminidases, we synthesized both 2-acetamido-2-deoxy-5-
fluoro-â-D-glucopyranosyl fluoride 2 (2A5FGF) and the azide
analogue 2AA5FGF 1 as potential inactivators (Scheme 1).
Using the hydrochloride 331 as a starting point, the liberated
amine was protected using a phthalimido group to give 4.
Selective removal of the anomeric acetyl group was effected
using benzylamine, and the resulting hemiacetal 532 was treated
with (diethylamino)sulfur trifluoride (DAST) to yield the
â-fluoride 6 in excellent yield. Photobromination of 6 gave the
presumed 5-bromo intermediate that was immediately treated
with silver tetrafluoroborate in diethyl ether to furnish exclu-
sively the 5-fluoro D-gluco derivative 7. Surprisingly, removal
of the phthalimido group could be effected by hydrazine without
With the putative mechanism-based inhibitors 1 and 2 in hand,
we turned our attention to evaluating their ability to inactivate
VCNagZ. Incubation of VCNagZ with 2A5FGF 2 results in
rapid time-dependent inactivation of the enzyme relative to a
control (Figure 2A). This result differs somewhat from a report
(31) Cunha, A. C.; Pereira, L. O. R.; de Souza, M. C. B. V.; Ferreira, V. F. J.
Chem. Ed. 1999, 76, 79-80.
(32) Pozsgay, V.; Jennings, H. J. J. Carbohydr. Chem. 1990, 9, 333-343.
9
J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008 329

A R T I C L E S
Stubbs et al.
Scheme 1. Synthesis of the 2-Acetamido-2-deoxy-5-fluoro-â-D-glucopyranosyl Fluoride 2 (2A5FGF) and the Azide Analogue 2AA5FGF 1
by Vocadlo et al. who saw incomplete inactivation of a family
3 â-N-acetylglucosaminidase from V. furnissi when using the
C-5 epimer, the L-ido inactivator.20 The more natural D-gluco
inactivator studied here may therefore offer a more appropriate
scaffold for generation of ABPPs for â-N-acetylglucosamini-
dases. We carried out inactivation kinetics at several inactivator
concentrations, and the kinetic parameters governing inactivation
(kinact and Ki) were determined by plotting kobs versus [I] (Figure
2B). Inactivation of VCNagZ by 2A5FGF 2 revealed a Ki value
of 86 ( 15 µM and a kinact value of 0.74 ( 0.09 s-1. In all
cases the inactivation data fitted well to a single exponential
decay equation.
Gratifyingly, 2AA5FGF 1 acts as an even more efficient time-
dependent inactivator of VCNagZ; in only 7 minutes complete
inactivation of VCNagZ is observed at concentrations as low
as 40 µM (Figures 2C and 2D). The value of the second-order
rate constant (kinact/Ki ) 0.011 ( 0.002 µM-1 s-1) governing
inactivation compares favorably with that of the parent inacti-
vator 2A5FGF 2 (kinact/Ki ) 0.009 ( 0.003 µM-1 s-1), consistent
with the studies described above using the azide derivatized
substrate. Consequently, the azide moiety has little effect on
the inactivation process. 2AA5FGF 1 therefore inactivates
family 3 â-D-glucosaminidases on a time scale appropriate for
practical applications and suggests that this compound could
be used to detect â-D-glucosaminidase activity within complex
proteomes at endogenous levels of target protein.
and the azide moiety remained intact. Treatment of the sample
with phosphine-FLAG21 followed by SDS-PAGE and Western
blot analysis revealed no signal if the acrylamide gel was
subjected to electrophoresis for the normal time (90 V for 20
minutes and then 180 V for 65 minutes). This result was
presumably owing to the acylal ester linkage of the glycosyl-
enzyme intermediate breaking down under the basic conditions
(pH 8.8) of the buffered gel. Decreasing the pH of the gel
resulted in diffuse bands that were poorly resolved (data not
shown).
After some optimization a short electrophoresis protocol (90
V for 20 minutes and then 200 V for 8 minutes) was developed
that when used in conjunction with Western blot analysis
showed the enzyme was indeed labeled (Figure 3). To confirm
that the phosphine-FLAG was reacting specifically with the
azide of NagZ inactivated with 1 and not with the protein itself,
various controls were performed (Figure 3A). The results clearly
reveal that both the inactivation of NagZ using 1 and the
subsequent Staudinger ligation are highly specific, and neither
reagent on its own results in labeling of the enzyme.
To validate the selectivity of our approach toward those â-D-
glucosaminidases that form a glycosyl-enzyme intermediate, we
tested two enzymes from family 84 of glycoside hydrolases.
These â-D-glucosaminidases differ in their catalytic mechanism
in that they use substrate-assisted catalysis involving the
2-acetamido group as a nucleophile in place of an enzymatic
nucleophile.33-37 These enzymes therefore do not form a
covalent glycosyl-enzyme intermediate, and as a result, they
The reporter group chosen for initial studies was the phos-
phine-FLAG ligand (Figure 1B).21 After incubation of purified
VCNagZ (0.84 mg mL-1) with 1 (333 µM) for 10 minutes no
residual enzyme activity could be detected. To stop the turnover
of the presumed glycosyl-enzyme intermediate the protein was
denatured under mild reducing conditions (pH 3.5 in the
presence of urea). In this way both the labile acylal ester linkage
(33) Tews, I.; Perrakis, A.; Oppenheim, A.; Dauter, Z.; Wilson, K. S.; Vorgias,
C. E. Nat. Struct. Biol. 1996, 3, 638-648.
(34) Knapp, S.; Vocadlo, D.; Gao, Z. N.; Kirk, B.; Lou, J. P.; Withers, S. G. J.
Am. Chem. Soc. 1996, 118, 6804-6805.
(35) Macauley, M. S.; Whitworth, G. E.; Debowski, A. W.; Chin, D.; Vocadlo,
D. J. J. Biol. Chem. 2005, 280, 25313-25322.
9
330 J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008

ABPPs for Retaining
â
-D-Glucosaminidases
A R T I C L E S
Figure 2. Time-dependent inactivation of VCNagZ. (A) Experimental data represent the VCNagZ activity of control samples (without 2) over time. [2]
used with VCNagZ: 0 (b), 10 (0), 20 (9), 30 (]), 50 (2), 100 ([) µM. Curves are the nonlinear fits of data to a single-exponential decay equation. (B)
Plot of the inactivation rate constants (kobs) as a function of [2] for VCNagZ. (C) Time-dependent inactivation of VCNagZ by 1: Experimental data represent
the VCNagZ activity of control samples (without 1) over time. [1] used with VCNagZ: 0 (b), 10 (0), 20 (9), 30 (]), 40 (2), 50 ([) µM. Curves are the
nonlinear fits of data to a single-exponential decay equation. (D) Plot of the inactivation rate constants (kobs) as a function of [1] for VCNagZ.
should not be labeled using the approach described here.
Repeating the experiment with VCNagZ along with the family
84 human O-GlcNAcase and a close bacterial homologue of
O-GlcNAcase, BtGH84, demonstrated the selectivity of the
labeling strategy (Figure 3B). The family 3 â-D-glucosaminidase,
VCNagZ, is labeled, but as predicted, the â-D-glucosaminidases
from family 84 are not.
Having shown the probe is selective for family 3 â-D-
glucosaminidases we determined the lower detection limit of
our strategy. Incubating purified VCNagZ (0.03, 0.09, 0.3, and
0.84 mg mL-1) with 1 revealed the lower limit for the detection
of VCNagZ to be approximately 80 ng (Figure 4A). Even though
it was obvious to us that we were losing signal the longer the
gel was subjected to electrophoresis, the approach is still quite
sensitive.
Having established that VCNagZ could be successfully
labeled using 1, our attention turned to extending this strategy
to the analysis of â-D-glucosaminidase activity from cell lysates.
The problem we faced, however, was that owing to the short
electrophoresis protocol, a gel could not be obtained that would
give sufficiently good resolution of cell lysates to resolve protein
bands clearly. We therefore turned our attention to using a
biotin-alkyne reporter group, which has the benefit of having a
biotin moiety that can be used to immobilize labeled protein to
avidin resin even under the acidic conditions required to stabilize
the acylal ester linkage. The avidin resin should immobilize only
those proteins bearing the glycosyl-enzyme intermediate con-
jugated to the reporter group. The washed resin, free of
unlabeled proteins, can be mixed with SDS-PAGE loading
buffer and boiled to cleave the acylal ester linkage, thereby
releasing the enzyme. After electrophoresis of the supernatant,
gels can be stained with SYPRO Orange to identify the affinity
purified proteins. We decided to use the proven alkyne-linked
biotin (Figure 1C).13,38-41
With this new reporter group in hand, we turned our attention
to labeling VCNagZ. Samples of VCNagZ (0.027, 0.09, 0.16,
and 0.3 mg mL-1) were treated with 1 in the manner described
above. Again, to prevent the turnover of the glycosyl-enzyme
intermediate, the protein was denatured under mild reducing
conditions (pH 3.5) by adding a solution of TCEP, the biotin
tag in DMSO, and the triazole ligand39 in a mixture of DMSO
and tert-butanol. After incubation of the sample for five hours,
followed by treatment with avidin resin, the sample was
centrifuged. The pelleted beads were washed with PBS buffer,
and SDS-PAGE loading buffer was then added and the sample
boiled. After electrophoresis (10% acrylamide), proteins were
visualized using SYPRO Orange. Gratifyingly, we observed a
protein band corresponding to the molecular weight of VCNagZ
(Figure 4B). These results indicate that both the Sharpless-
Meldal click cycloaddition and the avidin binding assay are
highly specific, but what is more encouraging was that this
capture and release approach is highly sensitive and allows
longer electrophoresis times. The lower limit for the detection
of the VCNagZ after all sample handling is approximately 25
(38) Link, A. J.; Vink, M. K. S.; Agard, N. J.; Prescher, J. A.; Bertozzi, C. R.;
Tirrell, D. A. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 10180-10185.
(39) Link, A. J.; Tirrell, D. A. J. Am. Chem. Soc. 2003, 125, 11164-11165.
(40) Wilbur, D. S.; Hamlin, D. K.; Vessella, R. L.; Stray, J. E.; Buhler, K. R.;
Stayton, P. S.; Klumb, L. A.; Pathare, P. M.; Weerawarna, S. A.
Bioconjugate Chem. 1996, 7, 689-702.
(36) Dennis, R. J.; Taylor, E. J.; Macauley, M. S.; Stubbs, K. A.; Turkenburg,
J. P.; Hart, S. J.; Black, G. N.; Vocadlo, D. J.; Davies, G. J. Nat. Struct.
Mol. Biol. 2006, 13, 365-371.
(37) Rao, F. V.; Dorfmueller, H. C.; Villa, F.; Allwood, M.; Eggleston, I. M.;
van Aalten, D. M. EMBO J. 2006, 25, 1569-1578.
(41) Gayo, L. M.; Suto, M. J. Tetrahedron Lett. 1996, 37, 4915-4918.
9
J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008 331

A R T I C L E S
Stubbs et al.
Figure 4. (A) Western blot showing the detection limit of 1, used in
conjunction with Staudinger ligation: amounts of VCNagZ were 775, 280,
80, 30 ng. (B) SYPRO-stained 10% SDS-PAGE showing the detection limit
of 1, used in conjunction with copper-catalyzed Sharpless-Meldal click
reaction. Lanes shown (left to right); no enzyme, no inhibitor 1, no biotin
label, flow through, wash 1, 280 ng of VCNagZ, 150 ng of VCNagZ, 80
ng VCNagZ, 25 ng VCNagZ.
Figure 5. SYPRO-stained 10% SDS-PAGE analysis of a cell lysate (240
µg) from a culture of P. aeruginosa PAO1. The cell lysate was treated
with 1 and then added to avidin resin. (A) Flow through, (B) wash 1, (C)
wash 2, (D) wash 3, (E) wash 4, (F) elution of bound PANagZ, (G) standard
of purified recombinant PANagZ (60 ng).
Figure 3. Visualization of 1-labeled VCNagZ by denaturing SDS/PAGE-
Western blot. After inactivation, the samples were labeled with phosphine-
FLAG and analyzed by the Western blot technique using anti-FLAG-HRP
(HRP, horseradish peroxidase): (A) SDS/PAGE-Western blot with the
appropriate controls. Coomassie blue-stained 10% SDS-PAGE gel, elec-
trophoresed fully, showing equal loading of lanes. The protein concentration
used in each lane was approximately 1 µg. Western blot analysis shows all
appropriate species must be present for a successful result. (B) Demonstrates
the selectivity of the technique for those â-D-glucosaminidases that have a
catalytic mechanism involving a glycosyl-enzyme intermediate. Lane (1)
VCNagZ (2.8 µg), (2) family 84 human O-GlcNAcase (1.2 µg), (3) family
84 BtGH84 (2 µg). The Coomassie blue-stained 10% SDS-PAGE gel,
electrophoresed fully, shows the different molecular weights of the three
proteins.
of strategies to combat this antibiotic resistance mechanism.25
A small culture of P. aeruginosa was grown to exponential
phase, harvested, and then lysed. The clarified lysate of P.
aeruginosa was incubated with activity-based probe 1 (5.6 µg),
treated with the biotin-alkyne reporter group (Figure 5), and
incubated with streptavidin beads as described above for purified
VCNagZ. Analysis of the beads by SDS-PAGE followed by
SYPRO Orange staining revealed a band with a molecular
weight consistent with that predicted for PA3005, a putative P.
aeruginosa family 3 â-D-glucosaminidase homologue (predicted
MW ) 33 kDa). We therefore cloned this putative NagZ,
PA3005, and after overexpression and purification we subjected
it to the same protocol as that used with the cell lysate (Figure
5, Lane G). Gratifyingly, pure PA3005 protein product is labeled
by 1 and migrates during electrophoresis to the same position
as the labeled protein in the cell lysate. 2AA5FGF 1 therefore
effectively labels this putative NagZ in a complex proteome
with exquisite selectivity, allowing it to be both detected and
purified. To verify the efficiency of this ABPP with this putative
enzyme, termed PANagZ, we analyzed the purified enzyme as
we did VCNagZ. Again, we find that selectivity is extremely
high, and the detection limit using Western blot is similar (≈50
ng) (Figures 6A and B).
ng, which compares favorably with the approximate 80 ng
detection limit obtained using the FLAG-phosphine method.
We next evaluated the proteome from P. aeruginosa, a Gram-
negative bacterium that harbors a chromosomally inducible
AmpC â-lactamase.42 This opportunistic pathogen is problematic
for patients suffering from cystic fibrosis, chronic obstructive
pulmonary disease, and severe burns.43-45 The genome sequence
of P. aeruginosa is known, and one gene (PA3005) has been
annotated as a family 3 â-D-glucosaminidase homologue28,29
(For more information on the classification of glycoside
hydrolases, see the CAZY Database available at afmb.cnrs-
mrs.fr/CAZY/). By analogy to other inducible AmpC systems,
this putative NagZ could be important in the induction of AmpC
â-lactamase and could therefore be a target for the development
(42) Sanders, C. C.; Sanders, W. E., Jr. Clin. Infect. Dis. 1992, 15, 824-839.
(43) Govan, J. R.; Deretic, V. Microbiol. ReV. 1996, 60, 539-574.
(44) Lyczak, J. B.; Cannon, C. L.; Pier, G. B. Clin. Microbiol. ReV. 2002, 15,
194-222.
(45) Nagaki, M.; Shimura, S.; Tanno, Y.; Ishibashi, T.; Sasaki, H.; Takishima,
T. Chest 1992, 102, 1464-1469.
The protein product of PA3005 from P. aeruginosa bears
48% amino acid sequence identity with VCNagZ. To verify its
activity as a â-N-acetylglucosmainidase we assayed the recom-
binant enzyme against a series of 4-nitrophenyl glycosides and
9
332 J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008

ABPPs for Retaining
â
-D-Glucosaminidases
A R T I C L E S
Conclusion
In summary, this ABPP (2AA5FGF 1) is the first highly
efficient probe for analysis of retaining exo-glycosidases at
endogenous levels from complex proteomes. This probe is useful
for profiling and affinity-based purification of â-D-glucosamini-
dases that use a catalytic mechanism involving a covalent
glycosyl-enzyme intermediate. The exquisite selectivity of both
the fluorosugar inactivation and the Sharpless-Meldal click
reaction also enables purification of this class of enzyme from
complex mixtures containing the enzymes of interest at low
native levels, as found within cell lysates. In this case we used
ABPP 2AA5FGF 1 to validate a bioinformatics assignment of
PA3005 as a â-D-glucosaminidase and verified this biochemi-
cally by cloning and characterization of P. aeruginosa NagZ
as a potential target for inhibiting the chromosomally inducible
AmpC â-lactamase pathway. We anticipate that this general
strategy will find utility in analyses of glycoside hydrolases from
various proteomes for both biotechnological and perhaps
therapeutic purposes.
Figure 6. Characterization of PANagZ: (A) Visualization of 1-labeled
PANagZ by denaturing SDS/PAGE-Western blot with appropriate controls.
After inactivation, the samples were treated as for VCNagZ SDS/PAGE-
Western blot analysis. (B) Western blot showing the detection limit of 1,
used in conjunction with Staudinger ligation. For PANagZ amounts used
were 920, 460, 210, and 63 ng.
Experimental Section
General. 1H and 13C NMR spectra were obtained on a BrukerAV600
(600 MHz for 1H and 150.8 MHz for 13C) spectrometer. Unless stated
otherwise, deuterated chloroform (CDCl3) was used as the solvent with
CHCl3 (δH 7.26) or CDCl3 (δC 77.0) being employed as internal
standards. NMR spectra run in D2O used internal CH3OH (δH 3.34, δC
49.0) as the standard. Melting points were determined on a Reichert
hot stage melting point apparatus. Optical rotations were performed
with a Perkin-Elmer 141 Polarimeter in a microcell (1 mL, 10 cm path
length) in CHCl3 at room temperature, unless stated otherwise. Mass
spectra were recorded with a VG-Autospec spectrometer using the fast
atom bombardment (FAB) technique, with 3-nitrobenzyl alcohol as a
matrix, unless otherwise stated. Elemental analyses of all synthesized
compounds used in enzyme assays were performed at the Simon Fraser
University. Flash chromatography was performed on BDH silica gel
or Geduran silica gel 60 with the specified solvents. Thin-layer
chromatography (TLC) was effected on Merck silica gel 60 F254
aluminum-backed plates that were stained by heating (g200 °C) with
5% sulfuric acid in EtOH. Percentage yields for chemical reactions as
described are quoted only for those compounds that were purified by
recrystallization or by column chromatography, and the purity was
found no detectable activity under the conditions of our assays
(1 mM substrate, 25 µM enzyme) with pNP R-L-fucopyranoside,
pNP â-D-glucopyranoside, pNP â-D-galactopyranoside, pNP
â-D-glucuronide, pNP R-D-mannopyranoside, and pNP R-D-
GlcNAc. A full kinetic analysis of the enzyme using pNP-
GlcNAc and pNP-GlcNAz30 revealed kinetic parameters (see
Supporting Information; pNP-GlcNAc: kcat/KM ) 6.6 ( 0.57
µmol mM-1 min-1 mg-1, kcat ) 2.97 ( 0.06 µmol min-1 mg-1
KM ) 0.45 ( 0.03 mM; pNP-GlcNAz: kcat/KM ) 4.3 ( 0.42
µmol mM-1 min-1 mg-1, kcat ) 0.38 ( 0.007 µmol min-1 mg-1
,
,
KM ) 0.089 ( 0.007 mM) similar to those observed for
VCNagZ (pNP-GlcNAc: kcat/KM ) 8.6 ( 0.66 µmol mM-1
min-1 mg-1, kcat ) 4.4 ( 0.08 µmol min-1 mg-1, KM ) 0.51
( 0.03 mM; pNP-GlcNAz: kcat/KM ) 24 ( 5.7 µmol mM-1
min-1 mg-1, kcat ) 0.48 ( 0.02 µmol min-1 mg-1, KM ) 0.02
( 0.004 mM) providing good support for assignment of this
enzyme as a NagZ. These values also reveal that the azide group
has a similarly small effect on catalysis for both of the enzymes,
and the values obtained for pNP-GlcNAc are consistent with
those measured for previously cloned family 3 N-acetyl-â-D-
glucosaminidases from V. furnisii and E. coli.46,47 The inactiva-
tion of PANagZ with 2A5FGF 2 and 2AA5FGF 1 also reveals
the kinetic similarity of PANagZ and VCNagZ (see Supporting
Information), consistent with our ability to inactivate and capture
this enzyme. The second-order rate constants governing inac-
tivation using both probes 2 (kinact/Ki ) 0.013 ( 0.004 µM-1
s-1, Ki ) 75 ( 14 µM, kinact ) 1.0 ( 0.1 s-1) and 1 (kinact/Ki )
0.010 ( 0.003 µM-1 s-1) are also very similar (see Supporting
Information) to those measured for VCNagZ. Altogether, this
data strongly support the bioinformatic assignment of PA3005
as an N-acetyl-â-D-glucosaminidase and as an NagZ.
1
assessed by TLC or H NMR spectroscopy. All solvents except DMF
and MeCN were distilled before use and dried according to the methods
of Burfield and Smithers.48 “Usual workup” refers to dilution with water,
repeated extraction into an organic solvent, sequential washing of the
combined extracts with hydrochloric acid (1 M, where appropriate),
saturated aqueous sodium bicarbonate and brine solutions, followed
by drying over anhydrous magnesium sulfate, filtration, and evaporation
of the solvent by means of a rotary evaporator at reduced pressure. V.
cholerae NagZ was expressed as previously described.25
3,4,6-Tri-O-acetyl-2-deoxy-2-phthalimido-D-glucose (5). To 1,3,4,6-
tetra-O-acetyl-2-deoxy-2-phthalimido-â-D-glucose 4 (5.5 g, 12 mmol)
dissolved in THF (30 mL) was added benzylamine (3.0 mL, 28 mmol),
and the mixture kept (1 h). Concentration of the reaction mixture
followed by flash chromatography (EtOAc/petrol, 1:1) furnished 5 as
a colorless powder (3.8 g, 76%). The spectral data were consistent with
those previously reported.32
3,4,6-Tri-O-acetyl-2-deoxy-2-phthalimido-â-D-glucosyl Fluoride
(6). (Diethylamino)sulfur trifluoride (0.80 mL, 6.0 mmol) was added
to 5 (1.5 g, 3.7 mmol) in CH2Cl2 (25 mL) at 0 °C, and the solution
kept (0.5 h). The reaction mixture was quenched with aqueous NaHCO3
(46) Chitlaru, E.; Roseman, S. J. Biol. Chem. 1996, 271, 33433-33439.
(47) Tsujibo, H.; Hatano, N.; Mikami, T.; Hirasawa, A.; Miyamoto, K.; Inamori,
Y. Appl. EnViron. Microbiol. 1998, 64, 2920-2924.
(48) Burfield, D. R.; Smithers, R. H. J. Org. Chem. 1983, 48, 2420-2422.
9
J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008 333

A R T I C L E S
Stubbs et al.
followed by usual workup and removal of the solvent. Crystallization
of the residue (Et2O) yielded 6 as a colorless powder (1.2 g, 80%).
The spectral data were consistent with those previously reported.49
3,4,6-Tri-O-acetyl-2-deoxy-5-fluoro-2-phthalimido-â-D-glucosyl
Fluoride (7). N-Bromosuccinimide (370 mg, 2.10 mmol) was added
to 6 (300 mg, 0.70 mmol) dissolved in CCl4 (10 mL), and the mixture
was irradiated using a 200 W bulb (3 h). Concentration of the reaction
mixture followed by flash chromatography (EtOAc/toluene, 1:19) gave
the presumed bromide as a gum (150 mg) that was immediately treated
with AgBF4 (64 mg, 0.33 mmol) in dry Et2O (1 mL). The mixture was
stirred (0.3 h), and upon consumption of the bromide the mixture was
cooled and quenched with Et3N (0.2 mL). The mixture was then filtered
through silica and concentrated. Flash chromatography (EtOAc/toluene,
1:9) followed by crystallization (MeOH) furnished 7 as fine needles
(50 mg, 16%), mp 145-148 °C, [R]D20 +21.8°. δH (600 MHz) 7.90-
7.75 (4H, m, Ar), 6.44 (dd, 1H, J1,2 ) 8.1 Hz, J1,F1 ) 53.3 Hz, H1),
mixture followed by flash chromatography (MeOH/CHCl3, 3:17) gave
a colorless solid that was dissolved in DMF (2 mL) and treated with
NaN3 (20 mg, 0.31 mmol), and the resulting mixture was stirred at 80
°C (2 h). Concentration of the mixture followed by flash chromatog-
raphy (MeOH/CHCl3, 3:17) furnished 1 as colorless flakes (15 mg,
32%), mp 161-163 °C, Vmax(film)/cm-1 2070 (N3). δH (600 MHz, CD3-
OD) 5.53 (dd, J1,F1 ) 54.0 Hz, H1), 3.98-3.93 (m, 3H, H2, CH2N3),
3.86 (dd, 1H, J2,3 ) 10.2 Hz, H3), 3.82 (dd, 1H, J6,F5 ) 9.3 Hz, H6),
3.73 (dd, 1H, J3,4 ) 9.4 Hz, J4,F5 ) 24.1 Hz, H4), 3.68 (dd, 1H, J6,F5
) 4.4 Hz, J6,6 ) 11.9 Hz, H6). δC (150.9 MHz, CD3OD) 170.85
(COCH2N3), 113.99 (dd, J5,F1 ) 8.1 Hz, J5,F5 ) 227.1 Hz, C5), 106.91
(dd, J1,F5 ) 3.9 Hz, J1,F1 ) 215.3 Hz, C1), 71.53 (d, J4,F5 ) 24.8 Hz,
C4), 70.47 (d, J3,F1 ) 9.8 Hz, C3), 62.62 (d, J6,F5 ) 33.8 Hz, C6),
56.94 (d, J2,F1 ) 19.9 Hz, C2), 53.03 (CH2N3). HR-MS m/z (FAB)
283.0849; [M + H]+ requires 283.0854. Anal. calcd for C8H12F2N4O5:
C, 34.05; H, 4.29. Found: C, 34.31; H, 4.23%.
6.04 (dd, 1H, J2,3 ) 10.4 Hz, J3,4 ) 9.6 Hz, H3), 5.43 (dd, 1H, J4,F5
)
Inactivation Kinetics. The time-dependent inactivation of each
enzyme by 1 or 2 was monitored by measuring the residual enzyme
activity over time. This was accomplished by incubation of the enzyme
(0.053 mg mL-1 for VCNagZ and 0.082 mg mL-1 for PANagZ) in
500 µL of PBS buffer solutions containing either inactivator 1 or 2.
For VCNagZ and PANagZ with 2, five inactivation concentrations (0,
10, 20, 30, 50, and 100 µM) were used. For the inactivation of VCNagZ
with 1, different concentrations (0, 10, 20, 30, 40, and 50 µM) were
used. The control mixture contained the same concentration of the
appropriate enzyme but no inactivator. Both inactivation and control
samples were incubated at room temperature (25 °C), and at various
time intervals an aliquot (5 µL) of each inactivation mixture was added
to a solution of the substrate, 4-nitrophenyl 2-acetamido-2-deoxy-â-
D-glucopyranoside in PBS (50 mM NaPi, 100 mM NaCl, 0.1% BSA,
pH 7.4), so that the final assay contained VCNagZ (active + inactive)
at a concentration of 3.3 µg mL-1 and 0.5 mM substrate in PBS buffer;
the total volume was 80 µL (for PANagZ the final concentration of
enzyme was 5.1 µg mL-1). The initial rates were then measured under
steady-state conditions by spectrophotometric monitoring of the release
of 4-nitrophenolate ion at 400 nm. Pseudo-first-order rate constants,
kobs ) ki[I]/(Ki + [I]), for the decay of activity at each inactivator
concentration were determined from fitting the decay curve to a single-
exponential decay equation using the nonlinear regression analysis
package of the program Grafit.
23.2 Hz, H4), 4.58 (ddd, 1H, J2,F1 ) 14.0 Hz, H2), 4.42 (dd, 1H, J6,F5
) 7.6 Hz, J6,6 ) 11.9 Hz, H6), 4.15 (dd, 1H, J6,F5 ) 4.2 Hz, H6), 2.87,
2.15, 2.10 (9H, 3s, CH3). δC (150.9 MHz) 169.65, 169.52, 169.20,
167.11 (CdO), 134.54, 131.11, 123.79 (Ar), 109.48 (dd J5,F1 ) 9.2
Hz, J5,F5 ) 230.7 Hz, C5), 102.12 (dd, J1,F1 ) 219.1 Hz, J1,F5 ) 5.1
Hz, C1), 68.27 (d, J4,F5 ) 24.0 Hz, C4), 66.61 (d, J3,F1 ) 9.9 Hz, C3),
61.48 (d, J6,F5 ) 36.9 Hz, C6), 53.89 (d, J2,F1 ) 22.5 Hz, C2), 20.46,
20.31, 20.16 (3C, COCH3). HR-MS m/z (ES) 494.0668; [M + K]+
requires 494.0665.
3,4,6-Tri-O-acetyl-2-acetamido-2-deoxy-5-fluoro-â-D-glucosyl Fluo-
ride (8). Hydrazine hydrate (0.5 mL) was added to 7 (30 mg, 0.07
mmol) in MeOH (2 mL), and the mixture kept (5 h). The mixture was
concentrated, and the residue dissolved in pyridine (3 mL), then Ac2O
(0.5 mL, 5.3 mmol) and DMAP (2 mg) were added, and the mixture
was again kept (2 h). The mixture was quenched with MeOH and
subjected to a usual workup followed by flash chromatography (EtOAc/
petrol, 3:2) to yield 8 as a colorless powder (20 mg, 82%), mp 221-
223 °C, [R]D20 -328°. δH (600 MHz) 5.70 (dd, 1H, J1,2 ) 7.5 Hz, J1,F1
) 53.3 Hz, H1), 5.45-5.35 (m, 2H, H3,H4), 4.40 (dd, 1H, J6,F5 ) 8.8
Hz, H6), 4.25-4.17 (m, 1H, H2), 4.11 (dd, 1H, J6,F5 ) 5.2 Hz, J6,6
)
12.0 Hz, H6), 1.94, 2.00, 2.05, 2.08 (4s, 12H, COCH3). δC (150.9 MHz)
170.91, 171.28, 171.42, 173.64 (4C, COCH3), 111.14 (dd, J5,F1 ) 8.1
Hz, J5,F5 ) 228.0 Hz, C5), 106.29 (dd, J1,F5 ) 4.2 Hz, J1,F1 ) 219.1
Hz, C1), 69.68 (d, J3,F1 ) 9.6 Hz, C3), 69.54 (d, J4,F5 ) 23.9 Hz, C4),
62.76 (d, J6,F5 ) 36.7 Hz, C6), 54.78 (d, J2,F1 ) 23.4 Hz, C2), 20.34,
20.38, 22.59 (COCH3). HRMS m/z (ES) 406.0722; [M + K]+ requires
406.0716.
Chemoselective Ligation Experiments Using Phosphine-FLAG.
An aliquot of enzyme sample (50 µL) was treated with a solution (10
µL) of 1 (2.0 mM, final [1] ) 333 µM) in PBS and incubated at room
temperature for 10 min. An aliquot of this mixture (50 µL) was taken,
and 1 M pH 2.0 sodium phosphate (20 µL), saturated urea (one-third
of the total volume), and water (30 µL) were added, in that order, to
yield a solution with pH ≈ 3.5. An aliquot of this mixture (25 µL) was
then taken and added to an equivalent volume of a solution of FLAG-
phosphine in water (500 µM, 250 µM final concentration), and the
mixture was allowed to incubate at room temperature for 6 h. The
sample was then mixed with SDS-PAGE loading buffer and without
heating the sample (20 µL) was loaded onto precast 10% Tris-HCl
polyacrylamide gels. After electrophoresis, the samples were electrob-
lotted to nitrocellulose membrane (0.45 µm, Biorad). Transfer was
verified by visual inspection of the transfer of prestained markers
(Biorad, product 161-0318). The membrane was blocked using PBS
containing 5% bovine serum albumin and 0.1% Tween-20 (blocking
buffer) with rocking for 1 h at room temperature. The blocking solution
was then decanted, and a solution of blocking buffer containing anti-
FLAG-HRP mAb conjugate (1:3500, Sigma) was added. Membranes
were incubated with rocking at room temperature for 1 h after which
the blocking solution containing antibody was decanted and the
membrane rinsed with PBS, containing 0.1% Tween-20 (wash buffer).
Membranes were then washed for 2 × 5 min and 2 × 20 min with
wash buffer. Detection of membrane-bound FLAG-HRP conjugates was
2-Acetamido-2-deoxy-5-fluoro-â-D-glucopyranosyl Fluoride (2).
Triethylamine (0.5 mL) was added to 8 (15 mg) in MeOH (2 mL), and
the mixture was stirred at 30 °C (3 h). Concentration of the reaction
mixture gave 2 as a colorless powder (10 mg), mp 153-155 °C. δH
(600 MHz, CD3OD) 5.46 (dd, 1H, J1,2 ) 7.9 Hz, J1,F1 ) 53.9 Hz, H1),
3.93-3.87 (m, 1H, H2), 3.83-3.77 (m, 2H, H3,H6), 3.73 (dd, 1H, J3,4
) 9.4 Hz, J4,F5 ) 23.9 Hz, H4), 3.66 (dd, 1H, J6,F5 ) 4.4 Hz, J6,6
)
11.9 Hz, H6), 2.00 (s, 3H, COCH3). δC (150.9 MHz, CD3OD) 173.89
(COCH3), 113.99 (dd, J5,F1 ) 8.0 Hz, J5,F5 ) 226.8 Hz, C5), 107.15
(dd, J1,F5 ) 3.9 Hz, J1,F1 ) 215.5 Hz, C1), 71.52 (d, J4,F5 ) 24.8 Hz,
C4), 70.47 (d, J3,F1 ) 9.8 Hz, C3), 62.64 (d, J6,F5 ) 33.9 Hz, C6),
56.89 (d, J2,F1 ) 20.0 Hz, C2), 22.84 (COCH3). HR-MS m/z (ES)
505.1427; [M2 + Na]+ requires 505.1421. Anal. calcd for C8H13F2-
NO5: C, 39.84; H, 5.43. Found: C, 39.67; H, 5.61%.
2-[(Azidoacetyl)amino]-2-deoxy-5-fluoro-â-D-glucopyranosyl Fluo-
ride (1). Hydrazine hydrate (1.0 mL) was added to 7 (70 mg, 0.20
mmol) in MeOH (5 mL), and the mixture kept (5 h). The mixture was
concentrated to yield a gum that was dissolved in MeOH (5 mL) and
then treated with Et3N (1 mL) and (ClCH2CO)2O (∼20 mol equiv)
until the reaction was judged complete (TLC). Concentration of the
(49) Blomberg, L.; Norberg, T. E. J. Carbohydr. Chem. 1992, 11, 751-760.
9
334 J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008

ABPPs for Retaining
â
-D-Glucosaminidases
A R T I C L E S
accomplished by chemiluminescent detection using a SuperSignal West
Pico Chemiluminescent Detection Kit (Pierce) and film (Pierce CL-
XPosure).
BamHI) were carried out on both the PCR product and empty pET28a
vector. The reaction volumes were 50 µL and consisted of ap-
proximately 5 µg of DNA and 1.0 U of each restriction enzyme
(Fermentas) and were allowed to proceed for 3-4 h at 37 °C. The
reaction mixtures were then subjected to electrophoresis through a 1.4%
agarose gel, and the desired DNA was gel purified as described above.
1 µL of the digested plasmid DNA was then combined with 5 µL of
the digested PCR product and ligated together using T4 DNA Ligase
(New England Biolabs) for 1 h at 18 °C. 2 µL of this ligation mixture
were used to transform E. coli ultracompetent cells as described above
except plates containing 50 µg/mL of kanamycin were used. A colony
was picked and cultured, and the plasmid DNA was harvested the
following day as described above. The entire insert was sequenced to
ensure accuracy. DNA encoding NagZ in the pET28a expression vector
was used to transform Tuner(λDE3) cells (Novagen). A colony was
selected and cultured to exponential phase in LB media (5 mL)
containing 50 µg/mL of kanamycin. This culture was used to inoculate
a 750 mL culture that was grown to an OD600 ≈ 0.8 at 37 °C. At this
point protein expression was induced using 1 mM IPTG (Bioshop) for
2.5 h at 25 °C. Postinduction cells were harvested by centrifugation
(Beckman J2-HS) for 10 min at 5000 rpm and resuspended in 20 mL
of nickel-column binding buffer (50 mM Na2PO4, 500 mM NaCl, 5
mM imidazole; pH 7.4) for each liter of cell culture. The resuspended
cells were incubated on ice for 30 min with 1 mg/mL of lysozyme
(USB) and 1 mM PMSF followed by sonication (6 × 20 s at 60%
power, Fisher Scientific, model 500). The cell debris was then removed
by centrifugation at 14 000 rpm for 75 min, and the supernatant was
loaded Via syringe onto a 5 mL FF HisTrap column (Amersham
Biosciences). The column was washed with 100 mL of wash buffer
(same as binding buffer but containing 60 mM imidazole) and eluted
with 30 mL of elution buffer (same as binding buffer but containing
250 mM imidazole). Purified enzyme was subsequently dialyzed
overnight against 4 L of PBS (pH 7.4) containing 10% (w/v) glycerol,
and the protein concentration of the resulting solution was quantified
by a Bradford assay using BSA as a standard.
Using Biotin-Avidin. The experiment involving recombinant
VCNagZ was similar to that described above up to the point just after
the addition of the inactivator 1. An aliquot of the inactivation mixture
(44 µL) was added to a solution (1 µL) of the biotin tag (120 × stock
in DMSO), followed by 10 µL of a 5 mM solution of TCEP (60 ×
stock in water) and 5 µL of a 100 µM solution of ligand (15 × stock
in DMSO:t-butanol 1:4) solution, giving a final tert-butanol concentra-
tion of 5%. Samples were gently vortexed, and 1 mM CuSO4 (60 ×
stock in water, 1 µL) was added to each sample, making the total
reaction volume 60 µL. Samples were gently vortexed again and
allowed to react at room temperature for 5 h, at which time 50 µL of
avidin resin (Promega) were added and allowed to stand for 4 h at 4
°C. The suspension was then centrifuged (Eppendorf 5417C) for 3 min
at 5000 rpm. The supernatant was removed and a fresh 50 µL of PBS
buffer was added, and after mixing, the suspension was kept for 1 h.
The sample was washed in this way three more times, and then 2 ×
SDS-PAGE loading buffer (reducing) was added (50 µL). After boiling
the sample, an aliquot (25 µL) was loaded onto a precast 10% Tris-
HCl polyacrylamide gel. After electrophoresis (0.05% SDS running
buffer), the gel was rinsed with water and treated with SYPRO-Orange
stain in 7.5% aqueous acetic acid for 30 min. The liquid was decanted,
and the gel was washed with 7.5% aqueous acetic acid for 20 min and
then rinsed with water. The gel was then scanned with a TYPHOON
imager using a 488 nm blue laser and 580 nm filter.
Probing P. aeruginosa Cell Lysates. Cultures (5 mL) of P.
aeruginosa PAO1 were grown at 37 °C overnight to exponential phase,
and the cells were harvested by centrifugation for 3 min at 13 000 rpm
(3500 × g). The cells were washed with 250 µL of PBS buffer, pelleted,
and then resuspended in fresh PBS buffer again. This suspension was
then sonicated (4 × 15 s at 20% power, Fisher Scientific, model 500)
after which the cell debris was removed by centrifugation at 15 000
rpm for 60 min. A sample of the supernatant (50 µL) was treated with
a solution (10 µL) of 1 (2.0 mM, final1 ) 333 µM, 5.6 µg total) in
PBS and incubated at room temperature for 10 min. An aliquot of this
mixture (44 µL) was treated as described above for pure VcNagZ.
Cloning of P. aeruginosa NagZ. Genomic DNA from P. aeruginosa
PAO1 was used as the template for cloning. The PCR was accomplished
using the following primers 5′-GATATACATATGCAAGGCTCTCT-
GATGCTC-3′ (NdeI cut site shown in bold) and 5′-GATATAG-
GATCCATCAATCAGTTGCGCAGC-3′ (BamHI cut site shown in
bold). The conditions used for the PCR were as follows: 28 cycles of
denaturation at 95 °C for 30 s followed by annealing of primers for 30
s at 60 °C; extension for 5 min at 72 °C using pfu DNA polymerase.
The reaction volume was 50 µL and contained 0.25 mM NTPs, 0.4
µM of each primer, 50 ng of template DNA, and 1 µL of pfu DNA
polymerase. The PCR reaction mixture was subjected to electrophoresis
through a 1.4% agarose gel for 45 min at 90 V, and the product was
excised and purified using a gel extraction kit (Qiagen). Double digest
reaction mixtures using both restriction endonucleases (NdeI and
Acknowledgment. Ethan D. Goddard-Borger is thanked for
technical assistance. D.J.V. is a scholar of the MSFHR and a
Tier II Canada Research Chair in Chemical Glycobiology. We
thank the Canadian Cystic Fibrosis Foundation and the Canadian
Institutes of Health research for funding.
Supporting Information Available: Details of kinetic analysis
of VCNagZ and PANagZ with the substrates 4-nitrophenyl
2-acetamido-2-deoxy-â-D-glucopyranoside and pNP-GlcNAz as
well as details of inactivation kinetic analysis of PANagZ with
1 and 2. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA0763605
9
J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008 335