An enzyme-initiated Smiles rearrangement enables the development of an assay of MshB, the GlcNAc-Ins deacetylase of mycothiol biosynthesis

Article abstract of DOI:10.1039/c2ob25429h

MshB is the N-acetyl-1-d-myo-inosityl-2-amino-2-deoxy-d-glucopyranoside (GlcNAc-Ins) deacetylase active as one of the enzymes involved in the biosynthesis of mycothiol (MSH), a protective low molecular weight thiol present only in Mycobacterium tuberculosis and other actinomycetes. In this study, structural analogues of GlcNAc-Ins in which the inosityl moiety is replaced by a chromophore were synthesized and evaluated as alternate substrates of MshB, with the goal of identifying a compound that would be useful in high-throughput assays of the enzyme. In an unexpected and surprising finding one of the GlcNAc-Ins analogues is shown to undergo a Smiles rearrangement upon MshB-mediated deacetylation, uncovering a free thiol group. We demonstrate that this chemistry can be exploited for the development of the first continuous assay of MshB activity based on the detection of thiol formation by DTNB (Ellman's reagent); such an assay should be ideally suited for the identification of MshB inhibitors by means of high-throughput screens in microplates. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2ob25429h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 5278
www.rsc.org/obc
PAPER
An enzyme-initiated Smiles rearrangement enables the development of an
assay of MshB, the GlcNAc-Ins deacetylase of mycothiol biosynthesis†
Dirk A. Lamprecht,a Ndivhuwo O. Muneri,a Hayden Eastwood,b,c Kevin J. Naidoo,b,c Erick Strauss*a and
Anwar Jardine*c
Received 27th February 2012, Accepted 17th May 2012
DOI: 10.1039/c2ob25429h
MshB is the N-acetyl-1-D-myo-inosityl-2-amino-2-deoxy-D-glucopyranoside (GlcNAc-Ins) deacetylase
active as one of the enzymes involved in the biosynthesis of mycothiol (MSH), a protective low
molecular weight thiol present only in Mycobacterium tuberculosis and other actinomycetes. In this study,
structural analogues of GlcNAc-Ins in which the inosityl moiety is replaced by a chromophore were
synthesized and evaluated as alternate substrates of MshB, with the goal of identifying a compound that
would be useful in high-throughput assays of the enzyme. In an unexpected and surprising finding one of
the GlcNAc-Ins analogues is shown to undergo a Smiles rearrangement upon MshB-mediated
deacetylation, uncovering a free thiol group. We demonstrate that this chemistry can be exploited for the
development of the first continuous assay of MshB activity based on the detection of thiol formation by
DTNB (Ellman’s reagent); such an assay should be ideally suited for the identification of MshB inhibitors
by means of high-throughput screens in microplates.
for MSH under various stress conditions, led to the identification
Introduction
of the biosynthesis and utilization of this novel thiol as a poten-
tial target for tuberculosis drug development.6
The ability to counteract the effects and consequences of oxi-
dative stress is an important adaptation that ensured the survival
of organisms in an aerobic environment.1 Moreover, many
pathogenic microorganisms also maintain virulence by resisting
the oxidative killing mechanisms of the human immune system’s
defences. Low molecular weight (LMW) thiols play a key role
in these processes by acting as redox buffers, thereby helping to
maintain redox homeostasis within cells.2 In most organisms—
including humans—the principle LMW thiol is glutathione.
However, many Gram-positive bacteria do not contain gluta-
thione, and evidence suggests that they instead rely on the meta-
bolic cofactor coenzyme A (CoA) and/or the recently discovered
bacillithiol (BSH) as redox buffer.3 The actinomycetes, which
include Mycobacterium tuberculosis (Mtb), the causative agent
of tuberculosis, produce mycothiol (MSH) as their principal
LMW thiol.4,5 These differences, and the essential requirement
MSH is biosynthesised from UDP-GlcNAc, L-myo-inositol-1-
phosphate (L-Ins-1-P), L-cysteine and acetyl-CoA through the
sequential action of four enzymes, designated MshA, MshB,
MshC and MshD (Scheme 1).4,7 A fifth enzyme, an as yet un-
identified phosphatase tentatively named MshA2, has also been
implicated in the reaction.8 Two other enzymes also play impor-
tant roles in the maintenance of MSH levels: the NADPH-depen-
dent oxidoreductase mycothiol disulfide reductase (Mtr) is
responsible for the reduction of mycothiol disulfide (MSSM)
formed through the oxidation of MSH,9,10 while the mycothiol-
S-conjugate amidase (Mca) cleaves the cysteine amide of
mycothiol-S-conjugates formed in the MSH-dependent detoxifi-
cation of alkylating agents, free radicals and xenobiotics. In the
process it forms 1-D-myo-inosityl-2-amino-2-deoxy-D-glucopyra-
noside (GlcN-Ins) and a mercapturic acid derivative, which is
excreted.11
The mycothiol biosynthetic enzyme MshB is one of the
enzymes in the pathway that has been identified as an attractive
potential target for drug discovery. This was based on the fact
that MSH levels in knockout mutants of the mycobacterium
M. smegmatis that lacked the mshB gene were reduced up to
95% compared to wild-type (the low residual formation of MSH
was ascribed to the promiscuous nature of Mca).12 MshB is a
metal-dependent hydrolase13 that catalyses the deacetylation of
N-acetyl-1-D-myo-inosityl-2-amino-2-deoxy-D-glucopyranoside
(GlcNAc-Ins, 1) to form the amino sugar GlcN-Ins. Structural
aDepartment of Biochemistry, Stellenbosch University, Private Bag X1,
Matieland 7602, South Africa. E-mail: estrauss@sun.ac.za;
Fax: +2721 808 5863; Tel: +27 21 808 5866
bScientific Computing Research Unit, University of Cape Town, Private
Bag X3, Rondebosch 7701, South Africa
cDepartment of Chemistry, University of Cape Town, Private Bag X3,
Rondebosch 7701, South Africa. E-mail: anwar.jardine@uct.ac.za;
Fax: +2721 650 4010; Tel: +27 21 650 4010
†Electronic supplementary information (ESI) available: Full chromato-
grams and associated mass spectra from the LC-HR-ESI-MS ana-
lyses, and NMR spectra of the thioglycosides 3a–c. See DOI:
10.1039/c2ob25429h
5278 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 1 Biosynthesis and recycling of mycothiol (MSH).
Scheme 2 Proposed catalytic mechanism of MshB, highlighting the active site residues involved in catalysis.17
analysis showed that the active site metal—previously thought to
be zinc, although recent studies have found higher activity with
iron14—is coordinated to two histidine residues (His13 and
His147) and an aspartate (Asp16).15,16 The metal is also bound
to the acetamidocarbonyl oxygen of the substrate and a water
molecule, which was proposed to be activated for nucleophilic
attack through this interaction. This has been confirmed in recent
mechanistic studies, which also showed that MshB uses a
general acid–base mechanism in which Asp15 and His144 act as
the general base and acid catalysts respectively, while Tyr142
plays a dynamic role that includes stabilization of the oxyanion
intermediate (Scheme 2).17
The study of MshB activity and inhibition poses two major
practical challenges. First, its natural substrate can only be
obtained by means of lengthy multi-step synthetic pro-
cedures,18,19 or by isolation of mycothiol-S-conjugates from
natural sources, followed by treatment with Mca to release
GlcN-Ins (Scheme 1).20 Consequently, many recent MshB
studies made use of commercially available GlcNAc (2) as an
alternate minimal substrate instead,14,17,21 even though it has
been shown that MshB is ∼100-fold less active towards this
compound.4,20 Additionally, other groups have invested
significant synthetic effort in the discovery of alternate MshB
substrates that give reasonable levels of activity.22 Second, the
general method for the assay of MshB activity involves derivati-
zation of reaction products followed by lengthy HPLC
analysis, which precludes high-throughput screening.22–24
A
significant advancement has been the recent development of
a fluorescence-based microplate assay that measures MshB
activity based on the reaction of fluorescamine (FSA) with the
free amine of the deacetylation product after the reaction has
been quenched.21 However, its mechanism precludes its use for
the screening of compound libraries for potential MshB inhibi-
tors in which similarly reactive functional groups are
represented.
In this study we sought to address these shortcomings by the
incorporation of chromogenic functional groups into the struc-
tures of potential alternate MshB substrates. Our strategy was to
exploit any changes that occur in the absorbance spectra of such
compounds upon deacetylation for use in a continuous spectro-
photometric assay. In the process we uncovered an unexpected
enzyme-initiated Smiles rearrangement of one of the substrates,
which allowed the development of a new, sensitive microplate-
based assay of MshB activity.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5278–5288 | 5279

View Article Online
Fig. 1 Structures of alternate MshB substrates used in this study.
Results and discussion
Rationale and strategy
Fig. 2 Molecular docking analysis of the MshB active site binding
pocket (enclosure shown in green) identifies two possible binding orien-
tations for the native substrate GlcNAc-Ins (1), shown in red (geometry
1) and blue (geometry 2) respectively. His13 which coordinates zinc,
and the general acid catalyst His144 are shown, as well as the hydro-
phobic residues at the base of the pocket.
Previous studies of MshB and Mca have shown that modification
of the inosityl moiety of GlcNAc-Ins is usually tolerated by
these enzymes, although often to the detriment of the observed
activity.19,22 We therefore chose to investigate whether the re-
placement of the inosityl group by aromatic moieties would
similarly result in compounds accepted as substrates by MshB.
More specifically, we set out to use aromatic groups that are
known strong chromophores, anticipating that deacetylation of
such alternate substrates may result in changes in their absor-
bance maxima. Such changes would allow the continuous spec-
trophotometric assaying of MshB activity, which would be ideal
for high-throughput screening. We decided to introduce such
groups via S-glycoside linkages, since the syntheses of several
α-GlcNAc thioconjugates have been reported.19,22,24–26
Three aromatic groups were chosen as replacements of the
inosityl group of GlcNAc-Ins for this study (Fig. 1). In the first
case a para-nitrophenyl group, a well-known chromophore uti-
lized in several different enzymatic assays, was selected to give
GlcNAc-SPNP (3a). Second, a 2,4-dinitrophenyl group was
chosen to give GlcNAc-SDNP (3b) as a potential alternate sub-
strate. The extra nitro-group was expected to increase the absor-
bance intensity of the chromophore, thereby potentially also
increasing its sensitivity to any changes that might occur in its
absorbance spectrum upon deacetylation. Third, a simple benzyl
group was selected as a reference for the study of the other two
groups, as its absorbance was unlikely to be affected by deacety-
lation due to its distant location from the acetamido group. More-
over, a recent study found that GlcNAc-SPh (the phenyl
thioglycoside) showed ∼30% of the activity of the native
GlcNAc-Ins substrate, suggesting that GlcNAc-SBn (3c) might
similarly act as a good alternate substrate, thereby also acting as
a positive control.22
derivatization of the free amine of the expected product, fol-
lowed by HPLC analysis.22 While this result indicated that 3b
would not be suitable for the purposes of this study, it seemed
curious in light of the good activity observed for the structurally
analogous GlcNAc-SPh using the same assay. Evaluation of the
available structures of recombinant MshB indicated that its
binding pocket is cavernous and that it should be able to accom-
modate a variety of substrates, as indicated by the green enclo-
sure shown in Fig. 2.15,16 However, charged or polar amino
acids predominate in the binding cavity, with few well-defined
hydrophobic regions. This suggested that the hydrophobic aro-
matic groups of the alternate substrates would not be accommo-
dated in the MshB active site, although this could be offset by
the potential of the nitro substituents to act as hydrogen bond
acceptors. To establish whether 3a and/or 3b would be excluded
from MshB’s active site based on such structural considerations,
we set out to perform docking and molecular dynamics studies
using one of the available MshB structures (PDB: 1Q74).16
While the active site zinc ion is clearly visible in this structure, it
does not contain any other bound ligands. The natural substrate
GlcNAc-Ins (1) and minimal substrate GlcNAc (2) were there-
fore also included in the analysis for comparative purposes.
The results of the molecular dynamics studies performed with
the solution-equilibrated MshB structure indicated that the
docked MshB–ligand complexes with good scores have the acet-
amidocarbonyl oxygen in close proximity to the zinc, which is
consistent with the current mechanistic understanding of metal-
dependent deacetylases.13 Moreover, the relatively large active
site cavity volume permits two favourable docked-ligand geome-
tries for each of the ligands, both of which have the acetamido-
carbonyl oxygen placed in a manner allowing catalysis to take
place. For the natural substrate GlcNAc-Ins, the two binding
Docking studies and computational chemistry
One of the alternate substrates that we proposed to use in this
study, the dinitrophenyl derivative 3b, was recently reported to
be inactive in a MshB assay that depended on the fluorescent
5280 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 3 Synthesis of thioglycosides 3a–c. (i) 1-fluoro-4-nitroben-
zene (for a), 1-fluoro-2,4-dinitrobenzene (for b) or benzyl bromide (for
c), Et3N, CH2Cl2, 25 °C. (ii) MeOH, H2O, acetone, Amberlite IRA-400
(OH−). The R-groups are defined in Fig. 1.
pocket orientations can be defined using its coordination to the
zinc ion as pivot, and its displacement relative to His13 and
Asp16 as reference points (Fig. 2). In geometry 1, GlcNAc-Ins
extends into a pocket facing down and away from His13, while
in geometry 2 GlcNAc-Ins stretches across into a binding pocket
that is orthogonal to geometry 1. The results indicate that for
GlcNAc-Ins there may be a small but significant energetic pre-
ference for geometry 1 over geometry 2.
In regards to the alternate substrates, the docking analysis indi-
cates that the minimal substrate GlcNAc (2) prefers binding in
geometry 2, which may partially explain its reduced activity.
Similarly GlcNAc-SPNP (3a) also prefers binding in geometry
2, while GlcNAc-SDNP (3b) prefers geometry 1. The analysis
also shows that regardless of the binding geometry, the aromatic
rings of both proposed alternate substrates can be accommodated
in the same pocket that binds inositol. However, the presence of
the hydrophobic residues Val10, Phe9 and Val134 at the base of
the pocket that surrounds geometry 1 would seem to indicate
that this would be the binding mode most likely to compensate
for the loss of the various hydrogen bonding interactions when
the inosityl moiety of GlcNAc-Ins in replaced by an aromatic
group.
Fig. 3 Characterization of 3a–c as potential alternate MshB substrates.
Panel A, UV-vis spectra of 3a–c at 200 μM. Panel B, HPLC analysis of
100 μM each of 3a, 3b (at 320 nm) and 3c (at 254 nm) incubated in the
presence of MshB followed by heat inactivation at set time intervals,
Taken together, the results of the docking studies suggest that
3a and 3b would most likely act as alternate substrates for
MshB, with activities similar to that previously observed for
GlcNAc-SPh.22
shows
a time-dependent decrease in the concentration of each
compound.
Synthesis of potential chromogenic substrates of MshB
coefficients at ∼325 nm and ∼320 nm respectively, while the
benzyl-containing 3c has a relatively poor absorbance with a
maximum at ∼254 nm (Fig. 3A). Nonetheless, the presence of
the chromophoric groups in all three cases allowed their direct
analysis by HPLC, which provided for a simple method whereby
their potential deacetylation by MshB could be monitored. Sub-
sequent incubation of MshB individually with 100 μM each of
3a–c showed a time-dependent decrease for all three compounds,
confirming that they indeed act as alternate substrates of the
enzyme (Fig. 3B). Integration of the peak area at each time point
allowed for the HPLC results to be converted to time-course data
by comparison to a standard curve prepared independently. In
this manner the specific activity of the enzyme towards each sub-
strate could be calculated (Table 1). These values indicated that
MshB has a similar specific activity for all three the alternate
substrates, which is ∼12% of the value previously determined
for the natural substrate GlcNAc-Ins using HPLC analysis of the
fluorescent derivative of the product.27 However, it is about an
order of magnitude better than the value determined for the
minimal substrate GlcNAc in the same manner, indicating that
The α-GlcNAc-mercaptan (4) has previously been used in the
synthesis of a range of α-thioglycosides—including analogues
of MSH—by simple nucleophilic substitution using an appropri-
ate electrophile.19,26 We therefore decided to expand on this
established chemistry for the synthesis of the α-thioglycosides
3a–c (Scheme 3). Triethylamine-mediated S-alkylation of 4 with
aryl fluorides or benzyl bromide gave the desired O-acetylated
thioglycosides (5a–c). Subsequent O-deacetylation gave the
unprotected thioglycosides (3a–c). As expected, the coupling
constants of the anomeric protons were all consistent with that of
α-coupled thioglycosides.
Confirming MshB activity with alternate substrates by HPLC
With the thioglycosides 3a–c in hand we set out to demonstrate
that these are indeed accepted by MshB as alternate substrates.
The UV-visible spectra of 3a–c showed that, as predicted, the
nitrophenyl-containing 3a and 3b have large extinction
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5278–5288 | 5281

View Article Online
Table 1 MshB activity of various substrates as measured by using the
would allow for a direct comparison of the activity of MshB
towards the respective substrates.
indicated assaysa
Surprisingly, the resulting data (Table 1) gave rise to a very
different activity profile than the one determined by HPLC
analysis. The FSA-based assay highlights GlcNAc-SBn 3c as
being the preferred MshB substrate among the thioglycoside,
showing a specific activity ∼6-fold that of GlcNAc 2, while
GlcNAc-SPNP 3a and GlcNAc-SDNP 3b show no or very low
activities (with large errors). Since the HPLC-based assay clearly
indicated the disappearance of all three thioglycosides, this
finding suggested that the products formed upon deacetylation of
3a and 3b do not contain free amine groups that are available for
reaction with FSA. We next set out to establish a possible
mechanistic basis for such an outcome.
Specific
activity
(nmol min−1 Rate (V/E)b
Assay
Substrate
mg−1
)
(min−1
)
HPLC
GlcNAc-Ins (1)
220 4027 7.21 1.3127
(AccQ-Fluor
deriv.)c
GlcNAc (2)
3
227
0.10 0.0727
HPLC (direct)d
GlcNAc-SPNP (3a) 26.9 1.3
GlcNAc-SDNP (3b) 28.4 3.0
0.88 0.04
0.93 0.10
0.83 0.10
GlcNAc-SBn (3c)
25.4 3.0
Fluorescamine
(FSA)e
GlcNAc (2)
19.1 3.4
0.63 0.11
GlcNAc-SPNP (3a) ∼0
∼0
GlcNAc-SDNP (3b) 4.2 3.7f
0.14 0.12
122.2 17.0f 4.01 0.56
Deacetylation of GlcNAc-SDNP unmasks a free thiol:
an enzyme-initiated Smiles rearrangement
GlcNAc-SBn (3c)
DTNBe
GlcNAc (2)
∼0
∼0
∼0
GlcNAc-SPNP (3a) ∼0
GlcNAc-SDNP (3b) 145.8 15.9 4.78 0.52
GlcNAc-SBn (3c)
We considered that the unmasking of the free amino group of 3a
and 3b could lead to a possible intramolecular rearrangement
that results in this group being blocked again, thereby preventing
their reaction with FSA. Such a case was reported by Kondo
et al., which found that S-(2,4-dinitrophenyl)cysteine undergoes
a facile base-catalyzed S → N rearrangement to form N-(2,4-
dinitrophenyl) cysteine.28 This type of rearrangement, which
falls under a broad category of intramolecular aromatic nucleo-
philic substitution reactions, involves the formation of an anionic
σ complex intermediate, also referred to as a Meisenheimer
complex. Such rearrangements were first reported by Smiles in
1930, and are therefore known by his name.29 In the experiment
by Kondo et al., the reaction was performed by incubating
S-(2,4-dinitrophenyl)cysteine with different organic bases,
including imidazole. This suggested that enzyme-borne bases
could also promote such Smiles rearrangements, and that this
could be the basis for the surprise finding that 3a and 3b are see-
mingly inactive in the FSA-based MshB assay.
∼0
∼0
a Activity values determined in this study represent the average of three
independent measurements, with the standard error shown. b Calculated
from the specific activity data. c Using 100 μM substrate in water, and
0.5 μM MshB. d Using 100 μM substrate in water, and 2 μM MshB.
e Using 5 mM substrate in 5–10% DMSO, and 1 μM MshB. f Based on a
standard curve prepared with GlcN.
the thioglycosides 3a–c may act as suitable alternative substrates
for the assay of MshB activity.
Confirmation of the formation of the expected products upon
deacetylation of 3a–c by MshB
Since synthetic standards of the expected deacetylation products
of 3a–c were not available, the HPLC-based MshB activity
assay described above could only be used to confirm the dis-
appearance of substrate and not the formation of these products.
To confirm that MshB treatment of 3a–c yields products contain-
ing free amino groups, we employed the recently reported MshB
assay in which its substrate (GlcNAc in the original description)
is incubated in the presence of the reactive fluorophore FSA.21
The FSA reacts with the free amine of the deacetylated product
as it is formed to give a fluorescent conjugate, thereby providing
for the monitoring of product formation—and consequently
MshB activity—by fluorimetry.
The activity of MshB towards GlcNAc and the thioglycosides
3a–c (5 mM in 10% DMSO) were monitored in this manner
over a period of 1 h. The observed increase in fluorescence for
each substrate over time was subsequently converted to a change
in concentration over time using a standard curve prepared from
known concentrations of glucosamine (GlcN); in doing so, the
assumption was made that the reactivity of the deacetylated pro-
ducts of 3a–c towards fluorescamine would be similar to GlcN,
and that the rates obtained in this manner (after adjustment for
any background using a control sample containing no substrate)
A Smiles rearrangement of the thioglycoside substrates after
deacetylation by MshB should unmask a free thiol group, as
shown for 3b in Scheme 4. Since the concentration of free thiols
can be established through reaction with 5,5′-dithiobis-(2-nitro-
benzoic acid) (DTNB, or Ellman’s reagent), GlcNAc (acting as
negative control) and 3a–c was individually incubated with
MshB. The reactions were stopped at set time intervals, followed
by treatment with DTNB and measurement of the increase in
absorption at 412 nm to determine the rate of formation of
2-nitro-5-thiobenzoate (NTB, 8), if any.
The results (Table 1) showed an increase only in the case of
GlcNAc-SDNP 3b, with none of the other substrates showing an
increase above the background rate (no substrate control). More-
over, the rate observed for 3b in this assay is very similar to the
rate seen for 3c in the FSA-based assay. This finding confirms
that GlcNAc-SDNP 3b does act as an alternate substrate for
MshB, and that its apparent inactivity (based on the results of
the FSA-based assay and as reported in a previous study that
used AccQ-Fluor derivatization followed by HPLC analysis) is
due to an S → N Smiles rearrangement.
The observation that GlcNAc-SPNP 3a does not show a
similar unmasking of its thiol, while also being unreactive in the
5282 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 4 Proposed Smiles rearrangement of the MshB-catalysed de-
acetylation product (6) of GlcNAc-SDNP 3b. The formation of the
rearrangement product 7 can be followed by derivatisation with DTNB
(Ellman’s reagent) to form the strong chromophore, 2-nitro-5-thiobenzo-
ate (NTB, 8) and the mixed disulfide 9.
FSA assay, may be due to the reversible formation of the corres-
ponding Meisenheimer complex that is less stabilized by the
single nitro group, and which therefore does not complete the
rearrangement. Since the docking studies indicated that 3a and
3b may prefer to take on different binding modes, this might
also play a role in differentiating their activity profiles. However,
no direct evidence in support of either analysis can be provided
at this stage.
Fig. 4 LC-HR-ESI-MS analyses of the product(s) formed in the
MshB-mediated reaction of GlcNAc-SDNP 3b. Panel A, base peak ion
(BPI) chromatogram of a control reaction mixture containing only the
substrate 3b in the presence of the reductant TCEP. Panel B, BPI chro-
matogram of a reaction mixture containing 3b, MshB and TCEP. Panel
C, BPI chromatogram of a reaction mixture containing 3b and MshB
(but no TCEP) that had been treated with DTNB prior to analysis. All
peaks were identified by analysis of the associated mass spectra (the
required and found mass for the major ion in each spectrum is shown)
and observed fragmentation patterns. The full chromatograms and the
associated mass spectra are provided as ESI.†
Identification of the Smiles rearrangement product
We subsequently set out to positively identify and study the
Smiles rearrangement product by its chemical synthesis;
however, all attempts in this regard failed, as all the conditions
used for N- and/or O-deacetylation led to the formation of
N-(2,4-dinitrophenyl) glucosamine and H2S. We therefore
attempted its identification by LC-HR-ESI-MS analysis instead.
Comparison of reaction mixtures that contained 3b in either the
absence or presence of MshB indicated the enzyme-mediated
formation of the deacetylation product 6 but not its rearrange-
ment product 7 (Fig. 4). This identification is based on the analy-
sis of the associated mass spectrum, which showed the formation
of fragments that is in agreement with the structure of 6 but not
7, although it is also possible that a stabilized Meisenheimer
complex would give the same observed fragmentation pattern
(see ESI†); note that all three structures (6, 7, and the complex)
have the same molecular mass. Addition of the reductant tris-
(2-carboxyethyl)phosphine (TCEP) to the reaction mixtures gave
the same result. To establish whether the rearrangement product
is formed, but decomposes prior to or during analysis, DTNB
was added to mixtures after they had been quenched and the pre-
cipitated protein removed. Samples that had been treated in this
manner, and that did not contain TCEP, showed the formation of
a new peak that was identified as the mixed disulfide 9 (Fig. 4);
in the presence of TCEP this peak was not observed.
The results of the LC-HR-ESI-MS analyses confirm the deace-
tylation of GlcNAc-SDNP 3b by MshB to form the deacetyla-
tion product 6, as well as its subsequent rearrangement to form
7. Since the FSA-based assay failed to trap any free amine-con-
taining products, it is most likely that the observation of 6 in
these analyses is due to its formation from the stabilized Meisen-
heimer complex under the acidic conditions used for the MS
analysis; the Meisenheimer complex could also give the same
fragmentation pattern directly (Fig. 4B). Formation of the labile
rearranged product 7 is confirmed by its trapping by DTNB to
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5278–5288 | 5283

View Article Online
form the stable disulfide adduct 9; however, in the absence of a
trapping reagent 7 is not observed, in agreement with the diffi-
culties encountered with its chemical synthesis.
Exploiting the Smiles rearrangement to develop a new MshB
assay amenable to high-throughput screening
The confirmed deacetylation of GlcNAc-SDNP 3b by MshB led
us to consider whether this substrate could be used for the devel-
opment of a new assay for MshB activity, which is also amen-
able to high-throughput screening. Towards this end, we first
determined whether product formation could be measured based
on changes that occur in the absorbance spectrum of 3b upon its
deacetylation (and/or subsequent Smiles rearrangement).
However, incubation of 3b in the presence of MshB showed no
significant time-dependent changes in its absorbance at 320 nm
(its absorbance maximum, see Fig. 3A), making direct spectro-
photometric-based activity analysis impossible and necessitating
an alternative strategy.
The reaction of DTNB with free thiols have been used to
assay the activity of several enzymes (including the mycothiol
disulfide reductase, Mtr9) in continuous fashion. Having
confirmed the reaction of DTNB with 7 and the formation of the
mixed disulfide 9 and the associated NTB (8) by-product, we
therefore set out to determine if MshB activity could also be
determined in this manner. Increasing concentrations of 3b
(between 0 and 5 mM) were incubated in the presence of MshB
and DTNB, and the increase in absorbance at 412 nm monitored
in a time-dependent fashion. For comparison, identical reaction
mixtures were incubated without DTNB, stopped at regular time
intervals, and the amount of product formed determined by
DTNB treatment after the enzyme was precipitated and removed
as before.
The progress curves obtained from the continuous monitoring
of MshB activity through NTB release showed an initial lag
period (presumably due to the slow collapse of the Meisenhei-
mer complex) before giving way to the linear increase in NTB
formation. Importantly, a control reaction that contained 100 μM
substrate but no enzyme showed no increase, confirming that the
observed increase in absorbance was due to enzyme activity.
Regression analysis of the linear portions of the progress curves
allowed for the corresponding activity rates to be calculated,
while the rates from the discontinuous assay were determined
from regression analysis of the data obtained at the set time-
intervals at each concentration. The results (Fig. 5) show that the
data obtained by either method give comparable rates at substrate
concentrations of 1 and 5 mM. However, the continuous moni-
toring of activity is clearly superior since it gave data points with
smaller errors, leading to an excellent linear correlation between
activity and substrate concentration over the whole concentration
range measured. In fact, when the activity of reaction mixtures
with less than 1 mM substrate was determined in a discontinuous
fashion, the variation in the replicates became so large that accu-
rate measurements became impossible (data not shown). The
activity obtained using 5 mM 3b (3.4 0.1 min−1 from the con-
tinuous assay, and 5.0 1.0 min−1 for the discontinuous assay)
also correlate well with those obtained for 5 mM GlcNAc–SBn
3c obtained in the FSA-based assay (Table 1). This suggests that
Fig. 5 Activity profile of MshB with GlcNAc-SDNP 3b as substrate,
measured based on the release of NTB (8) after reaction of DTNB
(Ellman’s reagent) with the rearranged deacetylation product 7 (see
Scheme 4). Closed circles (●) represent data points determined through
the continuous monitoring of product formation by performing the reac-
tion in the presence of DTNB; Open circles (○) are data points obtained
by quenching reaction mixtures are regular intervals, followed by DTNB
treatment. All data points represent the average of two separate measure-
ments, with the errors bars indicating the standard deviation. Linear
regression analysis of the (●) data set gives a line with the equation y =
0.6763x + 0.000138. The y-axis labels on the left and right are associ-
ated with the data points to the left and right of the x-axis break respect-
ively; the units are identical in both cases.
under the assay conditions used, the kinetics of the Smiles
rearrangement do not adversely affect the measurement of
MshB’s deacetylation of 3b by DTNB derivatization.
The finding that MshB’s activity towards 3b as a substrate cor-
relates linearly over the concentration range used (0–5 mM) was
not surprising considering that the KM value determined for
GlcNAc using the FSA assay is 38
4 mM (kcat 46
2.2 min−1).21 However, the poor solubility of 3b in aqueous sol-
utions prevented us from using higher concentrations, and there-
fore the kinetic parameters could not be determined for this
substrate. Nonetheless, the rate obtained for 5 mM GlcNAc-
SDNP using the DTNB-based assay is ∼7–12% of the kcat(Glc-
NAc) obtained in the FSA-(46 2.2 min−1 21
)
and HPLC-based
(29.4
2.4 min−1 27
) assays respectively, suggesting that the
kinetic parameters for 3b do not differ significantly from those
for GlcNAc.
Finally, the result demonstrates that GlcNAc-SDNP 3b acts as
a suitable alternative substrate for MshB, allowing its activity to
be measured continuously using a DTNB-based assay at concen-
trations as low as 50 μM.
Conclusions
In this study we set out to develop a new continuous assay of
MshB activity based on the use of chromogenic substrates. In
the process, we uncovered an unexpected and previously unseen
enzyme-initiated Smiles rearrangement of one of the substrates.
5284 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012

View Article Online
Computational methods
This discovery allowed for the first continuous and sensitive
assay of MshB activity to be developed.
Prior to performing the docking calculations the MshB crystal
structure (PDB ID: 1Q74) was protonated using the Karlsberg
pKa prediction tool.30 The protein was hydrated for molecular
dynamics (MD) simulations using a sphere of 30.0 Å in a
manner similar to one reported previously.31
The study of the mechanistic detail of the Smiles rearrange-
ment of deacetylated 3b would be interesting considering not
only the novelty of the enzyme-based initiation of the reaction,
but would also help relate the rate of deacetylation to the kinetic
and thermodynamic factors that control the formation and sub-
sequent collapse of the associated Meisenheimer complex. More-
over, it is also unclear whether the Smiles rearrangement takes
place in the MshB active site, where interaction between the
unmasked thiol and the catalytic metal ion may result in inhi-
bition, or only after the deacetylated product is released.
However, it is quite likely that the rearrangement takes place
upon collapse of the tetrahedral intermediate, and that adjacent
acidic residues such as His144 plays a role in stabilizing the
Meisenheimer complex intermediate. Conversely, the very rigid
platform provided by the glucosamine skeleton, and the presence
of stabilising electron-withdrawing nitro substituents in 3b,
would promote the rearrangement independent of enzyme invol-
vement. A full understanding of the interplay between such con-
siderations would be helpful in further establishing the utility of
the assay.
MD simulations were performed using the CHARMM32b32
program with the CHARMM27 all atom force field for the
protein33 and CSFF for the carbohydrate and inositol.34 In all
MD simulations a water sphere of 30.0 Å containing 11 322
TIP3P35 water molecules as implemented in CHARMM36 were
used. The system was centred about the glycosidic linkage of the
ligand for the simulations. The water molecules that overlapped
with any of the ligand heavy atoms were removed and a spheri-
cal boundary force was applied to the surface of the water sphere
to maintain a consistent water density of 1.0 g cm−1. A constant
temperature of 298.15 K was maintained throughout the simu-
lations using the Nosé–Hover thermostat.37 A buffer region of
2 Å thickness and a dynamic region with 21 Å radius was used
with Langevin dynamics in all simulations. All hydrogen bond
lengths were kept fixed with the SHAKE algorithm and an inte-
gration step of 1 fs was used.
In conclusion, the serendipitous discovery of GlcNAc-SDNP
3b as an alternate MshB substrate with unique characteristics
provides the opportunity for microplate-based inhibitor screening
of this enzyme, or any other deacetylase reactivity that may
potentiate the Smiles rearrangement by similarly invoking het-
eroatom nucleophiles. This finding may significantly advance
our search of inhibitors of mycothiol biosynthesis as potential
antituberculosis agents.
The crystal structure was subjected to 7 ns explicitly solvated
molecular dynamics and then stripped of water molecules, after
which several ligands were docked in the MD solution equili-
brated MshB structure using Glide.38 A cubic volume of side
9 Å, centred on the central zinc atom was selected. A search was
conducted via a grid generation with no constraints in conjunc-
tion with a Van der Waals scaling factor of 1 and a partial charge
cut-off of 0.25. Conformational searches on each ligand were
performed using the Extra Precision (XP) algorithm, where a
maximum of 2000 conjugate gradient steps, a distance dielectric
cut-off of 2 Å and flexible docking were applied on 1200 poses
for every ligand.
Materials and methods
Of the conformations generated, those where the Coulomb
and van der Waals interactions summed to values greater than 0
were rejected. To ensure that the ligands generated were confor-
mationally distinct an RMS deviation of 0.5 Å was used. The
100 highest scoring geometries for each molecule were examined
to identify structures where the carbonyl carbon is in close proxi-
mity to the zinc.
General reagents and methods
Chemicals were purchased from Sigma-Aldrich. All reactions
were performed at room temperature unless specified otherwise.
Flash chromatography was performed using a 20 : 1 weight ratio
of silica to crude reaction mixture. Thin layer chromatography
(TLC) was performed on Merck Aluminium Silica gel 60 F254
plates. All compounds containing free amines were visualized on
TLC plates by spraying the plates with ninhydrin solution (1.5%
ninhydrin in n-butanol with 3% AcOH) followed by heating the
plates until the colour developed. All chromogenic compounds
were visualized on the TLC plate under UV light. HPLC ana-
lyses were performed on an Agilent 1100 system with an in-line
DAD detector. LC-HR-ESI-MS analyses were performed using a
Waters Aquity UPLC attached to a Waters Synapt G2 QTOF
mass spectrometer. High resolution ESI-MS analyses were per-
formed on a QTOF Ultima API quadrupole mass spectrometer.
1H and 13C NMR was performed on Varian Unity Inova
600 MHz NMR and Varian VXR 300 MHz NMR spectrometers.
Enzyme assays (fluorometric and spectrophotometric analyses)
were conducted using a Varioskan multimode reader (Thermo
Scientific). An extinction coefficient (ε) of 13 600 dm−3 mol−1 cm−1
was used for the absorbance of 2-nitro-5-thiobenzoate (NTB, 8)
at 412 nm.
Synthesis of α-thioglycoside compounds
4-Nitrophenyl 2-acetamido-2-deoxy-1-thio-3,4,6-tri-O-acetyl-
α-D-glucopyranose (5a). To
a
solution of compound 426
(280 mg, 0.773 mmol) in 5 ml dichloromethane stirred under a
N2 atmosphere was added 1-fluoro-4-nitrobenzene (164 μl,
2.8 mmol) and triethylamine (130 μl, 1.15 mmol). The reaction
was allowed to stir for 24 h, and then concentrated. The product
was purified from the crude reaction mixture using flash chrom-
atography (2% MeOH/CH2Cl2), resulting in 360 mg (96% yield)
of pure 5a as a yellow amorphous solid. δH (600 MHz, CDCl3,
Me4Si): 1.974 (s, 3H), 2.026 (s, 3H), 2.061 (s, 3H), 2.081 (s,
3H), 4.070 (dd, 1H), 4.254 (dd, 1H), 4.373 (m, 1H), 4.590 (m,
1H), 5.169 (m, 2H), 5.902 (d, 1H), 6.013 (d, 1H), 7.560 (d, 2H),
8.152 (d, 2H); δC (150 MHz, CD3COCD3): 21.9, 23.9, 54.3,
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5278–5288 | 5285

View Article Online
Benzyl 2-acetamido-2-deoxy-1-thio-α-D-glucopyranose (3c).
Prepared as for 3a from 5c (230 mg, 0.507 mmol), resulting in
76 mg (38% yield) of 3c. m.p. 192–194 °C (dec). δH (300 MHz,
CD3OD, Me4Si): 1.89 (s, 3H), 3.36 (dd, 1H), 3.60 (dd, 1H),
3.76 (m, 4H), 3.99 (m, 2H), 5.34 (d, 1H), 7.20 (t, 1H), 7.27
(t, 2H) and 7.33 (d, 2H). δC (75 MHz, CD3OD, Me4Si): 22.5,
35.0, 55.6, 62.6, 72.6, 74.5, 84.0, 128.0, 129.4, 130.1, 139.6,
173.5 and 173.5. m/z (100% ESI-HRMS): 328.1211 ([M + H]+
C15H22NO5S+ requires 328.1219).
64.0, 71.1, 71.5, 72.3, 87.6, 126.0, 132.0, 145.5, 171.8, 171.9
and 207.4. m/z (100% ESI-MS): 485 [M + H]+, 507 [M + Na]+.
2,4-Dinitrophenyl
2-acetamido-2-deoxy-1-thio-3,4,6-tri-O-
acetyl-α-D-glucopyranose (5b). To a solution of compound 426
(500 mg, 1.37 mmol) in 10 ml dichloromethane stirred under a
N2 atmosphere was added 1-fluoro-2,4-dinitrobenzene (173 μl,
1.37 mmol) and triethylamine (230 μl, 1.65 mmol). The mixture
was allowed to stir under a N2 atmosphere for 24 h. The reaction
mixture was then concentrated and the resulting product was
purified using flash chromatography (2% MeOH/CH2Cl2),
resulting in 570 mg (79% yield) of pure 5b as a yellow solid. δH
(300 MHz, CDCl3, Me4Si): 1.935 (s, 3H), 1.991 (s, 3H), 2.034
(s, 3H), 2.066 (s, 3H), 4.043 (dd, 1H), 4.265 (m, 2H), 4.626 (m,
2H), 5.235 (m, 2H), 6.100 (d, 1H), 8.045 (d, 1H), 8.406 (dd,
1H), 9.026 (d, 1H); m/z (100% ESI-MS): 530 [M + H]+.
MshB preparation
The mshB (Rv1170) gene was amplified from M. tuberculosis
H37Rv genomic DNA using as forward primer 5′-
GGTGCCTC
primer 5′-CCTGGTTGGCC
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
̲
primers respectively contained NcoI and XhoI restriction
enzyme sites (underlined), while the reverse primer also
removed the native stop codon. The resulting PCR product was
treated with NcoI and XhoI and ligated to similarly treated
pET28a(+) to yield the expression vector pET28a(+)-mshB that
encodes recombinant MshB with a C-terminal 6×His-tag. The
plasmid’s sequence was confirmed by DNA sequencing.
For the expression of MshB the pET28a(+)-mshB plasmid
was transformed (heat shock method) into E. coli BL21*(DE3)
cells. LB medium (500 ml) was inoculated with a 5 ml starter
culture of these cells grown in the same medium, followed by
cell growth at 37 °C until the culture reached an OD600 of 0.6.
Expression was induced by the addition of IPTG to a final con-
centration of 0.10 mM, after which the culture was incubated
overnight at 22 °C. Cells were harvested by centrifugation at
17 000 × g for 20 min at 10 °C. The resulting cell pellet was
resuspended in binding buffer (20 mM Tris–HCl, 5 mM imida-
zole, 500 mM NaCl, pH 7.9), and the cells were disrupted using
sonication with external cooling (ice bath). Cell debris were
removed by centrifugation at 75 000 × g for 20 min at 10 °C to
obtain the clarified cell extract used for purification.
The protein was purified using an automated purification
program on an ÄKTAprime system (Amersham Bioscience)
fitted with a 1 ml HiTrap Chelating HP column (GE Healthcare)
pre-loaded with Zn2+. The column-bound MshB was eluted with
100% elution buffer (20 mM Tris-HCl, 500 mM imidazole,
500 mM NaCl, pH 7.9) and desalted using a 5 ml HiTrap desalt-
ing column (GE Healthcare) with gel filtration buffer (25 mM
Tris-HCl, 5 mM MgCl2, 5% glycerol, pH 8) on the ÄKTAprime
system. Protein purity was confirmed with SDS-PAGE analyses
and protein concentration was determined with a Bradford assay.
Pure MshB (in gel filtration buffer with 5% glycerol added) was
aliquoted and stored at −80 °C until needed.
Benzyl 2-acetamido-2-deoxy-1-thio-3,4,6-tri-O-acetyl-α-D-glu-
copyranose (5c). To a solution of compound 426 (220 mg,
0.61 mmol) in 2 ml dichloromethane stirred under a N2 atmos-
phere was added benzylbromide (87 μl, 0.732 mmol) and tri-
ethylamine (170 μl, 1.22 mmol). The reaction mixture was
allowed to stir under a N2 atmosphere for 24 h. Subsequently the
reaction was concentrated and the resulting product was purified
using flash chromatography (2% MeOH/CH2Cl2), resulting in
170 mg (62% yield) of pure 5c as an off white oil. δH
(400 MHz, CD3OD, Me4Si): 1.89 (s, 3H), 1.93 (s, 3H), 2.01
(s, 3H), 2.06 (s, 3H), 3.81 (d, 2H), 3.98 (d, 1H), 4.24 (dd, 1H),
4.38 (m, 3H), 4.97 (dd, 1H), 5.16 (dd, 1H), 5.29 (d, 1H) and
7.30 (m, 5H). δC (100 MHz, CD3OD, Me4Si): 20.4, 22.1, 35.0,
52.9, 63.1, 69.3, 70.6, 71.9, 83.8, 128.1, 129.4, 129.9, 138.9 and
171.2.
4-Nitrophenyl 2-acetamido-2-deoxy-1-thio-α-D-glucopyranose
(3a). To a solution of 5a (230 mg, 0.475 mmol) in 18 ml of
H2O (19%), MeOH (50%) and acetone (31%), was added 2.3 g
of Amberlite IRA-400 (OH−). The reaction mixture was allowed
to stir for 1 h at ambient temperature. The volatile solvents were
removed in vacuo and the residual aqueous product mixture was
subjected to C18 flash chromatography (50% MeCN/H2O).
Aqueous fractions containing the compound were collected and
dried by lyophilization, resulting in 50 mg (29% yield) of pure
3a as an amorphous yellow solid. δH (300 MHz, CD3OD,
Me4Si): 1.99 (s, 3H), 3.44 (t, 1H), 3.74 (m, 3H), 3.97 (m, 1H),
4.15 (dd, 1H), 6.08 (d, 1H), 7.68 (d, 2H) and 8.15 (d, 2H).
δC (75 MHz, CD3OD, Me4Si): 22.5, 56.1, 62.4, 72.3, 72.6, 75.7,
87.3, 124.9, 130.6, 146.3, 147.5 and 173.9. m/z (100%
ESI-HRMS): 359.0929 ([M + H]+ C14H19N2O7S+ requires
359.0913).
2,4-Dinitrophenyl 2-acetamido-2-deoxy-1-thio-α-D-gluco-pyra-
nose (3b). Prepared as for 3a from 5b (230 mg, 0.434 mmol),
resulting in 60 mg (34% yield) of 3b as an amorphous yellow
solid. δH (300 MHz, (CD3)2SO2, Me4Si): 1.829 (s, 3H), 3.264
(m, 1H), 3.516 (q, 1H), 3.588 (m, 2H), 3.674 (m, 1H), 3.983 (m,
1H), 4.554 (t, 1H), 5.125 (d, 1H), 5.272 (d, 1H), 6.051 (d, 1H),
8.150 (d, 1H), 8.273 (d, 1H), 8.426 (dd, 1H), 8.818 (d, 1H). m/z
(100% ESI-HRMS): 404.0781 ([M + H]+ C14H18N3O9S+
requires 404.0764).
HPLC-based activity analysis
All samples used for the direct HPLC analysis of MshB activity
contained one of the thioglycoside substrates (3a, 3b or 3c) at a
final concentration of 0.1 mM, assay buffer (50 mM HEPES,
50 mM NaCl, <1% DMSO, pH 7.4) and MshB (16.7 μg; 2 μM)
in a total volume of 250 μl. Reactions were initiated with the
addition of MshB. Samples of the reaction were stopped at
5286 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012

View Article Online
intervals of 5 min over a period of 20 min by heat inactivation,
followed by removal of the denatured enzyme by centrifugation
at 12 000 × g for 5 min to obtain the clarified samples used for
HPLC analysis.
For the determination of the substrate concentrations in the
samples, standard solutions of 3a–c with a concentration range
of 0–120 μM and a final volume of 250 μl each were given the
same treatment as the enzyme reaction samples before these
were also analysed by HPLC. From these results a standard
curve which correlated peak area with concentration was
obtained for each substrate.
Samples of 5 μl each were injected on to a Synergi 4μ Fusion-
RP 250 × 2.00 mm column (Phenomenex) and eluted using
aqueous acetonitrile as solvent. The amount of acetonitrile used
differed for each substrate: for GlcNAc-SPNP (3a), 30% MeCN/
H2O; for GlcNAc-SDNP (3b), 25% MeCN/H2O; and for
GlcNAc-SBn (3c), 40% MeCN/H2O. Separation was monitored
at 320 nm for 3a and 3b, and at 254 nm for 3c.
Reaction mixtures were analysed on a Waters Aquity UPLC
system with an autosampler using a Synergi 4μ Fusion-RP 250 ×
2.00 mm column (Phenomenex). The sample injection volume
was 20 μl and the autosampler syringe was washed with solvent
A (0.1% formic acid in water) before each injection. A gradient
elution program with a flow-rate of 0.30 ml min−1 was used for
the analysis. The gradient was as follows: 0–6 min, linear
increase to 50% solvent B (0.1% formic acid in acetonitrile);
6–8 min, isocratic 50% B; 8–10 min, linear increase to 95%
solvent B. Detection was performed on a Waters Synapt G2
QTOF mass spectrometer. Analytes were detected in the positive
and negative ion mode using a capillary voltage of 3000 V, cone
voltages of 15 V and a source temperature of 150 °C. The
mass spectrometer was calibrated with sodium formate and
Leu-enkephalin as internal standard to acquire the lock mass for
the HR-ESI-MS analysis. Data processing was done using
MassLynx (Waters).
Continuous DTNB-based MshB assay
FSA- and DTNB-based activity analyses
Reaction mixtures (125 μl) contained GlcNAc-SDNP (3b)
(0–5 mM) in assay buffer (50 mM HEPES, 50 mM NaCl, ∼10%
DMSO, pH 7.5) with DTNB (0.75 mM) and MshB (10 μg;
2.5 μM) were incubated at 30 °C in a 96-well microplate, and
the increase in absorbance at 412 nm was monitored over a
period of 30 min. Reaction rates were determined by linear
regression of the resulting progress curves obtained between
5–30 min. Discontinuous reaction monitoring was performed as
before, using the same reaction mixture without DTNB and with-
drawing 30 μl aliquots at set time intervals for processing and
analysis as described above.
Reaction mixtures (250 μl) containing the respective substrate
(5 mM) in assay buffer (50 mM HEPES, 50 mM NaCl, 1 mM
TCEP, 5–10% DMSO, pH 7.5) and MshB (8.1 μg; 1 μM) were
incubated at 30 °C. Samples (30 μl) of the reaction mixture were
added at 10 min intervals to 10 μl trichloroacetic acid (TCA)
(1.22 M) to quench the reaction, followed by removal of the pre-
cipitated protein by centrifugation at 12 000 × g for 5 min. The
supernatant (25 μl) was subsequently added to 75 μl borate
buffer (1 M, pH 9).
For the fluorescamine (FSA)-based analysis, 49 μl of the
resulting mixture was added to 30 μl of FSA (10 mM) contained
in a 96-well microplate incubated at room temperature, after
which the fluorescence was determined using an excitation wave-
length of 395 nm, and an emission wavelength of 485 nm.
For the 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB)-based
analysis, 49 μl of the resulting mixture was diluted to 100 μl,
and then added to 30 μl DTNB (2.2 mM) contained in a 96-well
microplate. After incubation at room temperature for 10 min, the
absorbance at 412 nm was determined.
Acknowledgements
This work was funded by grants from the Bill and Melinda
Gates Foundation (Grant#52035GA) and the National Research
Foundation (NRF) of South Africa (Grant#65527) to AJ. The
computational work is based upon research supported by the
South African Research Chairs Initiative (SARChI) of the
Department of Science and Technology (DST) and NRF
awarded to KJN. DAL and NOM are grateful recipients of Harry
Crossley and NRF scarce skills bursaries respectively.
LC-HR-ESI-MS analyses
Reaction mixtures (100 μl) contained GlcNAc-SDNP (3b)
(0.7 mM) in assay buffer (50 mM HEPES, 50 mM NaCl, ∼5%
DMSO, pH 7.5) (the enzyme was omitted from a negative
control reaction). One set of reaction mixtures also contained
1 mM TCEP. Mixtures were equilibrated at 30 °C for 10 min
before MshB (1 μM final concentration) was added, followed by
incubation at 30 °C for 1 h. The reaction was quenched by
adding two 45 μl aliquots from each mixture to two separate
microfuge tubes containing 15 μl TCA (6.12 M) each. The preci-
pitated protein was removed by centrifugation at 12 000 × g for
5 min, after which the supernatants from each tube (37.5 μl)
were added to 112.5 μl borate buffer (1 M, pH 9) contained in
two separate microfuge tubes. The tubes subsequently received
either 45 μl acetonitrile or 45 μl DTNB (0.72 mM in acetonitrile)
prior to LC-HR-ESI-MS analysis.
References
1 J. A. Imlay, Annu. Rev. Microbiol., 2003, 57, 395–418; J. A. Imlay, Annu.
Rev. Biochem., 2008, 77, 755–776.
2 P. Zuber, Annu. Rev. Microbiol., 2009, 63, 575–597.
3 G. L. Newton, K. Arnold, M. S. Price, C. Sherrill, S. B. Delcardayre,
Y. Aharonowitz, G. Cohen, J. Davies, R. C. Fahey and C. Davis, J. Bac-
teriol., 1996, 178, 1990–1995; R. C. Fahey, Annu. Rev. Microbiol., 2001,
55, 333–356; G. L. Newton, M. Rawat, C. J. J. La, V. K. Jothivasan,
T. Budiarto, C. J. Hamilton, A. Claiborne, J. D. Helmann and
R. C. Fahey, Nat. Chem. Biol., 2009, 5, 625–627; J. D. Helmann, Anti-
oxid. Redox Signaling, 2011, 15, 123–133.
4 V. K. Jothivasan and C. J. Hamilton, Nat. Prod. Rep., 2008, 25, 1091–
1117.
5 G. L. Newton, N. Buchmeier and R. C. Fahey, Microbiol. Mol. Biol. Rev.,
2008, 72, 471–494.
6 M. Rawat, C. Johnson, V. Cadiz and Y. Av-Gay, Biochem. Biophys. Res.
Commun., 2007, 363, 71–76.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5278–5288 | 5287

View Article Online
7 C. Bornemann, M. A. Jardine, H. S. C. Spies and D. J. Steenkamp,
Biochem. J., 1997, 325, 623–629.
8 G. L. Newton, P. Ta, K. P. Bzymek and R. C. Fahey, J. Biol. Chem.,
2006, 281, 33910–33920.
9 C. J. Hamilton, R. M. J. Finlay, M. J. G. Stewart and A. Bonner, Anal.
Biochem., 2009, 388, 91–96.
10 C. M. Holsclaw, W. B. Muse, III, K. S. Carroll and J. A. Leary,
Int. J. Mass Spectrom., 2011, 305, 151–156; M. P. Patel and
J. S. Blanchard, Biochemistry, 2001, 40, 5119–5126; K. S. E. Ung and
Y. Av-Gay, FEBS Lett., 2006, 580, 2712–2716.
11 M. Steffek, G. L. Newton, Y. Av-Gay and R. C. Fahey, Biochemistry,
2003, 42, 12067–12076.
12 M. Rawat, S. Kovacevic, H. Billman-Jacobe and Y. Av-Gay, Micro-
biology, 2003, 149, 1341–1349; X. Xu, C. Vilcheze, Y. Av-Gay,
A. Gomez-Velasco and W. R. Jacobs, Jr., Antimicrob. Agents Chemother.,
2011, 55, 3133–3139.
25 S. Knapp, B. Amorelli, E. Darout, C. C. Ventocilla, L. M. Goldman,
R. A. Huhn and E. C. Minnihan, J. Carbohydr. Chem., 2005, 24, 103–
130; B. Paul and W. Korytnyk, Carbohydr. Res., 1984, 126, 27–43;
Y. Ding, Y. Miura, J. R. Etchison, H. H. Freeze and O. Hindsgaul, J. Car-
bohydr. Chem., 1999, 18, 471–475.
26 S. Knapp and D. S. Myers, J. Org. Chem., 2002, 67, 2995–2999;
S. Knapp and D. S. Myers, J. Org. Chem., 2001, 66, 3636–3638.
27 G. L. Newton, M. Ko, P. Ta, Y. Av-Gay and R. C. Fahey, Protein
Expression Purif., 2006, 47, 542–550.
28 H. Kondo, F. Moriuchi and J. Sunamoto, J. Org. Chem., 1981, 46, 1333–
1336.
29 W. E. Truce, E. M. Kreider and W. W. Brand, Org. React., 2011, 99–
215.
30 B. Rabenstein and E.-W. Knapp, Biophys. J., 2001, 80, 1141–1150;
G. Kieseritzky and E.-W. Knapp, Proteins: Struct., Funct., Bioinf., 2008,
71, 1335–1348.
13 M. Hernick and C. A. Fierke, Arch. Biochem. Biophys., 2005, 433, 71–
84.
14 X. Huang, E. Kocabas and M. Hernick, J. Biol. Chem., 2011, 286,
20275–20282.
15 A. A. McCarthy, N. A. Peterson, R. Knijff and E. N. Baker, J. Mol. Biol.,
2004, 335, 1131–1141.
16 J. T. Maynes, C. Garen, M. M. Cherney, G. Newton, D. Arad, Y. Av-Gay,
R. C. Fahey and M. N. G. James, J. Biol. Chem., 2003, 278, 47166–
47170.
17 X. Huang and M. Hernick, J. Biol. Chem., 2012, 287, 10424–10434.
18 G. M. Nicholas, L. L. Eckman, P. Kovac, S. Otero-Quintero and
C. A. Bewley, Bioorg. Med. Chem., 2003, 11, 2641–2647; M. A. Jardine,
H. S. C. Spies, C. M. Nkambule, D. W. Gammon and D. J. Steenkamp,
Bioorg. Med. Chem., 2002, 10, 875–881; K. Ajayi, V. V. Thakur,
R. C. Lapo and S. Knapp, Org. Lett., 2010, 12, 2630–2633; S. Lee and
J. P. N. Rosazza, Org. Lett., 2004, 6, 365–368.
19 S. Knapp, S. Gonzalez, D. S. Myers, L. L. Eckman and C. A. Bewley,
Org. Lett., 2002, 4, 4337–4339.
20 G. L. Newton, Y. Av-Gay and R. C. Fahey, J. Bacteriol., 2000, 182,
6958–6963.
21 X. Huang and M. Hernick, Anal. Biochem., 2011, 414, 278–281.
22 D. W. Gammon, D. J. Steenkamp, V. Mavumengwana, M. J. Marakalala,
T. T. Mudzunga, R. Hunter and M. Munyololo, Bioorg. Med. Chem.,
2010, 18, 2501–2514.
23 G. M. Nicholas, L. L. Eckman, G. L. Newton, R. C. Fahey, S. Ray and
C. A. Bewley, Bioorg. Med. Chem., 2003, 11, 601–608.
24 B. B. Metaferia, B. J. Fetterolf, S. Shazad-ul-Hussan, M. Moravec,
J. A. Smith, S. Ray, M.-T. Gutierrez-Lugo and C. A. Bewley, J. Med.
Chem., 2007, 50, 6326–6336.
31 K. J. Naidoo, D. Denysyk and J. W. Brady, Protein Eng., Des. Sel., 1997,
10, 1249–1261; K. J. Naidoo and J. W. Brady, THEOCHEM, 1997,
395–396, 469–475.
32 B. R. Brooks, R. E. Bruccoleri, B. D. Olafson, D. J. States,
S. Swaminathan and M. Karplus, J. Comput. Chem., 1983, 4, 187–217;
B. R. Brooks, C. L. Brooks, A. D. Mackerell, L. Nilsson, R. J. Petrella,
B. Roux, Y. Won, G. Archontis, C. Bartels, S. Boresch, A. Caflisch,
L. Caves, Q. Cui, A. R. Dinner, M. Feig, S. Fischer, J. Gao,
M. Hodoscek, W. Im, K. Kuczera, T. Lazaridis, J. Ma, V. Ovchinnikov,
E. Paci, R. W. Pastor, C. B. Post, J. Z. Pu, M. Schaefer, B. Tidor,
R. M. Venable, H. L. Woodcock, X. Wu, W. Yang, D. M. York and
M. Karplus, J. Comput. Chem., 2009, 30, 1545–1614.
33 A. D. MacKerell, D. Bashford, M. Bellott, R. L. Dunbrack,
J. D. Evanseck, M. J. Field, S. Fischer, J. Gao, H. Guo, S. Ha, D. Joseph-
McCarthy, L. Kuchnir, K. Kuczera, F. T. K. Lau, C. Mattos, S. Michnick,
T. Ngo, D. T. Nguyen, B. Prodhom, W. E. Reiher, B. Roux,
M. Schlenkrich, J. C. Smith, R. Stote, J. Straub, M. Watanabe,
J. Wiórkiewicz-Kuczera, D. Yin and M. Karplus, J. Phys. Chem. B, 1998,
102, 3586–3616.
34 M. Kuttel, J. W. Brady and K. J. Naidoo, J. Comput. Chem., 2002, 23,
1236–1243.
35 W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey and M.
L. Klein, J. Chem. Phys., 1983, 79, 926–935.
36 P. J. Steinbach and B. R. Brooks, Proc. Natl. Acad. Sci. U. S. A., 1993,
90, 9135–9139.
37 S. Nosé, Mol. Phys., 1984, 52, 255–268; W. G. Hoover, Phys. Rev. A: At.,
Mol., Opt. Phys., 1985, 31, 1695–1697.
38 B. Hammouda, J. Polym. Sci., Part B: Polym. Phys., 1992, 30, 1387–
1390.
5288 | Org. Biomol. Chem., 2012, 10, 5278–5288
This journal is © The Royal Society of Chemistry 2012