Characterization of a transglycosylase domain of Streptococcus pneumoniae PBP1b

Article abstract of DOI:10.1016/j.bmc.2006.06.058

Inhibitors of transglycosylases may serve as potent antibiotics that are less prone to resistance development in bacterial pathogens. To facilitate the search of such compounds, a transglycosylase (TGase) domain of the membrane integral multidomain Strept

Full text of DOI:10.1016/j.bmc.2006.06.058

Bioorganic & Medicinal Chemistry 14 (2006) 7187–7195
Characterization of a transglycosylase domain
of Streptococcus pneumoniae PBP1b
Haitian Liu and Chi-Huey Wong*
Department of Chemistry, The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, CA 92037, USA
Received 23 May 2006; revised 16 June 2006; accepted 23 June 2006
Available online 25 July 2006
Abstract—Inhibitors of transglycosylases may serve as potent antibiotics that are less prone to resistance development in bacterial
pathogens. To facilitate the search of such compounds, a transglycosylase (TGase) domain of the membrane integral multidomain
Streptococcus pneumoniae PBP1b was cloned and expressed. This TGase domain was characterized by a substrate-dependent fluo-
rescence coupled enzyme assay and an inhibitor-tethered surface plasmon resonance binding assay. Both assays show that the cat-
alytic efficiency of the domain is comparable to that of the monofunctional transglycosylases, and it is fully active in the absence of
other domains. The isolation of the active TGase domain makes it possible to screen for potential antibiotics targeting
transglycosylases.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
carbohydrate-utilizing enzymes in bacteria and evade
the resistance development.
After six decades of widespread antibiotic use, bacterial
pathogens of human and animal origin are becoming
increasingly resistant to many antimicrobial agents.7
Many common bacterial pathogens such as Staphylococ-
cus aureus, Streptococcus pneumoniae, and Enterococcus
faecalis have become resistant. Methicillin-resistant
S. aureus (MRSA), penicillin-resistant S. pneumoniae,
and vancomycin-resistant E. faecalis are now common-
place pathogens that are proving difficult to treat
effectively.21 To date, even vancomycin, the last resort
of antibiotics active against S. aureus and some other
Gram-positive bacteria, has encountered resistance,31
and this new public health crisis has renewed interest in
antibacterial development.5 Current strategies for
tackling the problem of antibiotic resistance generally
involve chemical modification of existing antibiotics,
discovery of new targets, and development of new types
of antibiotics. Our interest in the field is to develop new
antibiotics that target unique carbohydrates or
Bacterial cell-wall (peptidoglycan) biosynthesis is both
essential and unique to bacteria and is a proven target
for identifying compounds as new antibiotics.6 Among
the different steps in the biosynthesis of peptidoglycan,
transglycosylation is the most interesting target for the
development of new antibiotics for two reasons. First,
the transglycosylase is fully exposed to the cell surface
and thus easily accessible to drugs as no transfer across
the plasma membrane is required. Second, the polysac-
charide backbone of peptidoglycan remains intact in
wild-type and resistant strains, whereas the peptide side
chain shows a high frequency of change. Antibiotics tar-
geting the transglycosylation step may therefore be less
prone to resistance development.
In the final step of peptidoglycan biosynthesis, Lipid II,
the substrate for the transglycosylation, is polymerized
and crosslinked into peptidoglycan by transglycosyla-
tion and transpeptidation (Scheme 1).12 Two major
enzymatic activities in the periplasmic space responsible
for the formation of peptidoglycan are the transglyco-
sylase (TGase) and the transpeptidase (TPase). A group
Keywords: Enzyme catalysis; Lipid II; Transglycosylase; Streptococcus
pneumoniae.
of
bifunctional
high-molecular-weight
(HMW)
*
penicillin-binding proteins (PBPs) possessing both
TGase and TPase activity have been identified in both
Corresponding author. Tel.: +1 858 784 2487; fax: +1 858 784
2409; e-mail: wong@scripps.edu
0968-0896/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.06.058

7188
H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
HO
HO
HO
O
O
O
HO
O
AcHN
O
NHAc
O
P
O
P
1. transglycosylation
2. transpeptidation
O
O
O
HN
OH
3
OH
7
O
NH
H
N
NH2
HOOC
O
NH
O
H
N
MurNAc
GlcNAc
COOH
O
Lipid II
Crosslinked
peptide
Scheme 1. Formation of peptidoglycan by transglycosylation and transpeptidation.
Gram-positive and Gram-negative bacteria.22 These
HMW PBPs are multidomain molecules, consisting of
a short N-terminal cytoplasmic region, a single trans-
membrane domain, and a large C-terminal periplasmic
region. The N-terminal domain of the periplasmic
region possesses TGase activity, while the C-terminal
domain contains TPase and penicillin-binding activi-
ty.9,26 Monofunctional enzymes with only TGase or
TPase activity have also been identified. Low molecular
mass PBPs function mostly as the D, D-carboxypepti-
dase and are believed to regulate the amount of cross-
linking within the peptidoglycan mesh.8 Non-PBP-relat-
ed TGase or monofunctional TGase (MTG) has been
reported recently in Escherichia coli, S. aureus, and
Micrococcus luteus.20,10 These proteins have the charac-
teristic fingerprints found in the N-terminal TGase do-
mains of HMW PBPs. In addition, the solubilized
membrane fraction of E. coli cells overexpressing the
mtg gene contains a significantly increased peptidogly-
can activity.10 The roles of PBPs and MTGs in cellular
peptidoglycan synthesis, however, are not known.
advanced cell-wall precursors. These discoveries made
it possible to set up efficient enzymatic assays for
transglycosylases.
We previously reported the characterization and assay
development of bacterial cell-wall stage II enzyme
MurG.15 Now we have extended our program into
transglycosylases, the last enzymes involved in the pep-
tidoglycan biosynthesis. Here we describe the cloning
and expression of a TGase domain of S. pneumoniae
PBP1b and the characterization of the domain by fluo-
rescence coupled enzyme assay and surface plasmon res-
onance (SPR) binding assay. Kinetic characterization of
a MTG from S. aureus by fluorescence coupled enzyme
assay was also performed. The initial reaction rates of
several MTGs from different species were compared.
The enzymatic characterization will be useful for future
structure determination and antibiotic development.
2. Results and discussion
Investigation of the mode of function of transglycosylas-
es, in particular the development of synthetic inhibitors,
has been hampered by difficulties in the development of
an efficient assay system. Some difficulties in studying
transglycosylases are related to problems in obtaining
and handling the substrate Lipid II. Isolating Lipid II
from bacteria is extremely difficult due to its low natural
abundance and its inherent structural complexity. The
total amount of Lipid II present in bacteria is very small
and has been determined to be no higher than 2000 mol-
ecules per cell.28 To separate this miniscule amount of
Lipid II from cell membrane and other cellular lipid
components is tedious and requires enrichment of
radiolabeled precursors.19 Lipid II can be alternatively
prepared by using MurG and its natural substrate Lipid
I, but the yield is low due to the detergent-like property
of the substrate.33 Recently, the development of
improved assay systems for monitoring transglycosyla-
tion has been the subject of many research efforts. The
chemical11 and chemo-enzymatic16 syntheses of the Park
nucleotide as well as several synthetic routes to Lipid I
and Lipid II18,33,30,24 afford milligram quantities of the
To facilitate the mechanistic and structural studies of
transglycosylases, we cloned and overexpressed a TGase
domain of PBP1b from common bacterial pathogen S.
pneumoniae and several MTGs from S. aureus and E.
coli. Full-length Pbp1b from S. pneumoniae which has
both the TGase and the TPase domains contains 821
amino acids (90 kDa). The proposed topology of this
enzyme predicted from a domain software globplot 2
consists of a cytoplasmic tail, a hydrophobic transmem-
brane helix (60–82), and a periplasmic region with a
TGase domain (106–279) and
a TPase domain
(422–696). A plasmid to express the full-length protein
was constructed, but initial attempts to solubilize mem-
brane-associated full-length S. pneumoniae PBP1b using
various detergents including Triton and CHAPS failed.
So a truncated mutant which has only the TGase
domain terminating at amino acid 300 was prepared.
The termination site was chosen to locate in the pro-
posed junction between TGase and TPase. Inclusion of
the transmembrane domain in expression would compli-
cate protein purification due to its hydrophobicity. So
our final TGase domain of PBP1b (amino acid

H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
7189
82–300) does not have the transmembrane domain for
easy purification. After cloning and expression, TGase
domain of the enzyme was obtained with a yield of
1.5 mg/L. Recently the TGase domain of E. coli PBP1b
and PBP2 from S. aureus were cloned and expressed.3,2
To compare the enzymatic activities of this TGase do-
main with other MTGs, we also cloned and expressed
three MTGs including a truncated MTG from S. aure-
us,32 a full-length and a truncated (without the trans-
membrane domain) MTG from E. coli. Their
expression levels are summarized in Table 1.
alkaline conditions and can be conveniently measured at
405 nm on a spectrophotometer. This chromogenic reac-
tion has been utilized for direct characterization of
enzyme activities of, for example, protein phosphatase,17
a-glucosidase,14 and manosidase.4 We adapted it here to
assay transglycosylase activity as shown in Scheme 2. In
this assay, the transglycosylase polymerizes the sugar
part of the Lipid II nitrophenol analogue 1 and releases
the nitrophenol diphosphate. Another enzyme, alkaline
phosphatase, hydrolyzes nitrophenol diphosphate to
give nitrophenol, which absorbs at 405 nm. So the trans-
glycosylase activity can be monitored continuously by
checking absorbance at 405 nm.
To characterize the isolated TGase domain and the
MTGs, we need access to the natural substrate Lipid
II. Lipid II, the final monomeric unit in the bacterial
cell-wall biosynthesis pathway, is a disaccharide an-
chored to the membrane by a 55-carbon undecaprenyl
chain with a cis-allylic pyrophosphate linkage to the su-
gar moiety. The long chain of Lipid II together with the
allylic pyrophosphate linkage complicates the synthesis,
purification, and handling. Therefore, we wish to find an
analogue of Lipid II that is active for this enzyme and
can be manipulated conveniently.
Our synthetic target, Lipid II nitrophenol analogue 1,
consists of three parts: the disaccharide central core 4,
the nitrophenol phosphate group 3, and the pentapep-
tide moiety 5 (Scheme 3). In analogy to the Lipid I nitro-
phenol analogue synthesis,15,30 the first disconnection
reveals disaccharide monophosphate 2 and the nitrophe-
nol-linked phosphate 3. Disaccharide monophosphate 2
could be derived from coupling protected pentapeptide 5
with the disaccharide monophosphate 4. Since we were
successful in synthesizing Lipid I nitrophenol analogue,
the challenge of synthesizing this molecule would be
access to multiple gram quantities of orthogonally
protected N-acetyl-(2-deoxy-2-aminoglycopyranosyl)-
b-[1,4]-N-acetylmuramyl acid 11 (NAG-NAM). The
synthesis of the disaccharide subunit 11 started from
conversion of 1-a-O-benzyl-N-acetyl-4,6-benzylindene
muramic acid 6 to the desired benzoyl-acceptor 8 in
three steps. N-Troc-protected donor 9 was prepared
from glucosamine 7 in three steps for 80% yield. Reac-
tion of donor 9 with acceptor 8 in the presence of N-iod-
osuccinimide and trifluoromethanesulfonic acid gave the
desired b-glycoside 10 in 56% yield after optimization.
At this stage, the disaccharide 10 was unable to be
separated from the acceptor 8. After working up the
reaction, the glycosylation mixture was dissolved in ace-
tic acid and zinc dust was then added to reveal the free
amine, which was acetylated with acetic anhydride in
pyridine to provide glycosylation product 11. Pure gly-
cosylation product 11 can be separated by flash column
(Scheme 4). Next reactions are to introduce the penta-
A Lipid II nitrophenol analogue which can be used to
set up an efficient optical assay for transglycosylase
activity was synthesized. para-Nitrophenyl phosphate
(p-NPP) is a chromogenic substrate for most phospha-
tases such as alkaline phosphatases, acid phosphatases,
protein tyrosine phosphatases, and serine/threonine
phosphatases. The reaction yields para-nitrophenol,
which becomes an intense yellow soluble product under
Table 1. Transglycosylases expressed and purified from Escherichia
coli
Expressed protein
Yield (mg/L)
Full-length PBP1b from S. pneumoniae
Amino acid 82–300 TGase domain
of Pbp1b from S. pneumoniae
n.a.
1.5
Amino acid 51–250 MTG from S. aureus
Amino acid 37–241 MTG from E. coli
Full-length MTG from E. coli,
periplasmic expression
1.4
0.34
2.0
HO
HO
HO
O
O
O
O
HO
O
O
O
-O P
O-
AcHN
phosphatase
-O P O P
O-
O-
NO2
O
O
O
NHAc
O
NO2
O
O P O
OH
NO2
O
P
OH
HN
phosphatase
peptidoglycan
O
NH
transglycosylase
H
N
NH2
HO2C
O
O
NH
H
N
CO2H
-O
NO2
O
1
absorption at 405nm
Scheme 2. Optical assay of transglycosylase with nitrophenol Lipid II analogue 1.

7190
H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
HO
HO
HO
HO
HO
HO
O
O
O
O
O
O
HO
O
HO
O
O
P
OH
AcHN
O
AcHN
AcHN
AcHN
NO2
O
O
P
OH
HO
O
P
+
O
O
P
OH
O
O
O
O
NO2
HN
HN
OH OH
3
O
O
NH
NH
H
H
N
N
NH2
NH2
HOOC
HOOC
O
O
NH
O
O
NH
H
H
N
COOH
N
COOH
O
O
2
1
H2N
O
HO
HO
HO
HO
NH
O
O
H
O
N
O
NHCOCF3
COOMe
+
MeOOC
AcHN
AcHN
O
O
O
NH
OH
H
OH
N
5
O
4
Scheme 3. Retrosynthetic route for Lipid II nitrophenol analogue 1.
peptide part and the nitrophenol phosphate which are
similar with the procedures for synthesizing Lipid I
nitrophenol analogues. The final product was purified
by reverse-phase HPLC.
similar as that of nitrophenol analogue except for the
diphosphate coupling reaction. We utilized phospho-
roimidazolidate method that was exploited in the Lipid
I and Lipid II total syntheses.15,30,29
For comparison, the natural substrate of the enzyme
Lipid II was also synthesized. The synthetic strategy is
Measurements of enzymatic activity were conducted
next with the synthesized Lipid II and TGase domain.
BzO
Ph
O
O
HO
O
O
O
a
O
AcHN
OBn
AcHN
O
OBn
O
COOH
BzO
O
AcO
AcO
6
O
O
SO2Ph
8
O
AcO
c
NH
AcHN
OBn
O
O
AcO
AcO
O
O
O
HO
HO
STol
b
O
AcO
SO2Ph
CCl3
NH
HO
10
O
NH2
OH
O
9
7
CCl3
d
other two acceptors for glycosylation:
BnO
BzO
O
AcO
O
O
AcO
AcO
O
OTBDMS
AcHN
NHAc
O
OBn
O
HO
O
HO
O
O
O
O
AcHN
AcHN
OBn
O
OBn
O
SO2Ph
11
O
SO2Ph
SO2Ph
13
12
Scheme 4. Synthesis of disaccharide subunit 11. Reagents and conditions: (a) i—2-phenolsulfonic acid, DMAP, EDC; ii—H2O, acetic acid;
iii—benzyl chloride, pyridine, 70% for three steps; (b) i—Troc-Cl, NaHCO3; ii—acetic anhydride, pyridine; iii—Tol-SH, BF3ÆEt2O, 80% for three
steps; (c) NIS, triflic acid, molecular sieves AW300, 56%; (d) i—zinc dust, acetic acid; ii—acetic anhydride, pyridine, 40% for two steps.

H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
7191
At the outset, an HPLC-based assay was used to identify
2
1.8
1.6
1.4
1.2
1
the glycosyltransfer reaction.24 Direct detection of poly-
merized Lipid II was confirmed by following the proce-
dures published by Schwartz and coworkers.24 After
positive results were obtained from both the HPLC as-
say and the direct detection of enzymatic product, dan-
3
2
1
0
syl-Lipid II derivative was synthesized and
a
0.8
0.6
0.4
0.2
0
fluorescence coupled enzyme assay was used to evaluate
the isolated TGase domain (Fig. 1).23 This coupled fluo-
rescence assay involves the use of a substrate labeled
with dansyl moiety, whose fluorescence intensity
decreases when the surroundings of the fluorophore
change from a less polar environment to a more polar
one. In this assay, incubation of dansyl Lipid II with
the transglycosylase releases the lipid phosphate and
produces polymerized Lipid II, the uncrosslinked pepti-
doglycan. Muramidase then hydrolyzes the polymer to
produce the dansyl disaccharide derivative. Enzymatic
removal of the lipid phosphate exposes the dansyl fluo-
rophore to a less polar environment resulting in decrease
of fluorescence intensity, which is monitored to measure
the enzyme kinetics. Under the optimized reaction con-
ditions, the TGase obeys Michaelis–Menten kinetics,
and the Km for dansyl-Lipid II was 62 lM and the kcat
was 0.76 sÀ1. The MTG from S. aureus32 was also char-
acterized using this approach and a Km of 2 lM and a
kcat of 0.01 sÀ1 were obtained as shown in Figure 2. A
similar construct of S. aureus MTG was characterized
using a radioactive assay and meso-[14C]A2pm-labeled
C55 Lipid II while this paper was under review.27 The
0
2
4
1 / [Dansyl Lipid II]
0
2
4
6
8
10
12
Dansyl Lipid II (μM)
Figure 2. Michealis–Menten plot of kinetic characterization of Staph-
ylococcus aureus MTG. The inset shows a double reciprocal plot.
results are comparable with ours obtained from the fluo-
rescence coupled enzyme assay.
Using the fluorescence coupled enzyme assay, we char-
acterized and compared the initial reaction rates of
other MTGs we cloned and expressed. Assays were
performed under identical conditions and the results
are shown in Table 2. Although all enzymes are
active, they showed different initial reaction rates
and the TGase domain from S. pneumoniae was the
most active enzyme. The full-length enzyme from
E. coli is less active than the truncated one probably
HO
HO
O
O
HO
O
HO
AcHN
O
AcHN
O
P
O
P
O
O
O
C
55H89
O
HN
OH
OH
O
TGase
NH
H
N
uncross-linked
peptidoglycan
NH
HOOC
O
O
NH
H
O
P
OH
O
P
N
O
COOH
O
C
55H89
HO
O
dansyl Lipid II
OH
flurophore in a less
polar environment
HO
HO
O
O
HO
O
HO
O
AcHN
AcHN
OH
O
O
HN
O
NH
H
muramidase
N+(Me)2
N
=
NH
S
HOOC
O
O
O
NH
H
N
COOH
dansyl labeled
O
disaccharide unit
flurophore in a more
polar environment
Figure 1. Coupled fluorescence assay for transglycosylase.

7192
H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
Table 2. Reaction rates (RFU/min) and the relative reaction rates of
different transglycosylases
taking place at the carbon (1) residue of Lipid II which
is close to the undecaprenol moiety, while the reaction of
MurG is taking place at the carbon (4) residue of Lipid I
which is far from the lipid chain.
Transglycosylases (60 nM)
Reaction Relative
ratesa
rates
(RFU/min) (U/min)
We next used a different method to evaluate TGase
domain that does not rely on a substrate. In this assay,
TGase is bound to a known inhibitor, and potential
inhibitors are evaluated based on their ability to
compete for the binding, which is monitored by surface
plasmon resonance (SPR). Moenomycin, a known
transglycosylase inhibitor which binds to the enzyme
effectively at low concentration, can be used in the
assay.13 It was derivatized with an amine moiety and
S. pneumoniae TGase domain of PBP1 b 17
6.1
4.8
1.8
1
Truncated MTG from E. coli
Full-length MTG from E. coli
Truncated MTG from S. aureus
13
5.0
2.7
a The reaction rates of different enzymes were measured under identical
conditions at a dansyl-Lipid II concentration of 2 lM.
due to its propensity to aggregate in solution. Unfor-
tunately, the Lipid II nitrophenol analogue we synthe-
sized was not active for all enzymes we expressed
including the TGase domain from S. pneumoniae
PBP1b, MTGs from S. aureus and E. coli.
TGase
RU
100
90
80
70
60
50
40
30
20
10
0
We compared the amino acid sequences of MTGs from
S. aureus and E. coli, the TGase domains from S. pneu-
moniae and E. coli PBP1b, and S. aureus PBP2. The
alignment result is shown in Figure 3. It has been pro-
posed the glutamate residue of the first conserved motif,
EDXXFXXH, is highly conserved in all known TGas-
es.27 This residue was found to be essential for catalysis
by site-directed mutagenesis, suggesting it acts as the key
catalytic residue in the active site of the TGase. About
the substrate specificity, we notice that although the
nitrophenol analogue of Lipid I works efficiently with
MurG,15 the nitrophenol analogue of Lipid II is not ac-
tive toward all MTGs we tested. Because Lipid I and
Lipid II share the same lipid chain, these results suggest
MurG and transglycosylases recognize the lipid part of
their substrates differently and the TGase can recognize
at least part of the lipid chain. We think the difference is
partly due to that the catalytic reaction of TGase is
-10
0
70
140
210
280
350
420
490
560
630
700
s
Time
Figure 4. Binding curve for different concentrations of Streptococcus
pneumoniae TGase domain of PBP1b to moenomycin.
TGase domain 1 91 ----------------------------SDLLRTSISSEQISENLKKAIIATEDEHFKEH
MTG 1
MTG 2
1 --MSVAPVPFSAVMVERQVSAWLHGNFRYVAHSDWVSMDQISPWMGLAVIAAEDQKFPEH
72 ------------------------ELRKIENKSSFVSADNMPEYVKGAFISMEDERFYNH
TGase domain 2 181 ATIVNMENNRQFGFFRLDPRLITMISSPNGEQRLFVPRSGFPDLLVDTLLATEDRHFYEH
PBP2 90 ----------------------------NGQRHEHVNLKDVPKSMKDAVLATEDNRFYEH
Conserved motif 1 EDXXFXXH
TGase domain 1 123 KGVVPKAVIRATLGKFVGLGSSSGGSTLTQQLIKQQVVGDAPTLARKAAEIVDALALERA
MTG 1
MTG 2
59 WGFDVASIEKALAHNERNENRIRGASTISQQTAKNLFLWDGRSWVRKGLEAGLTLGIETV
108 HGFDLKGTTRALFSTI-SDRDVQGGSTITQQVVKNYFYDNDRSFTRKVKELFVAHRVEKQ
TGase domain 2 241 DGISLYSIGRAVLANLTAGRTVQGASTLTQQLVKNLFLSSERSYWRKANEAYMALIMDAR
PBP2 122 GALDYKRLFGAIGKNLTGGFGSEGASTLTQQVVKDAFLSQHKSIGRKAQEAYLSYRLEQE
Conserved motif 2 RKXXE
TGase domain 1 183 MNKDEILTTYLNVAPFGRNNKGQN----IAGARQAAEGIFGVDASQLTVPQAAFLAGLPQ
MTG 1
MTG 2
119 WSKKRILTVYLNIAEF-----GDG----VFGVEAAAQRYFHKPASKLTRSEAALLAAVLP
167 YNKNEILSFYLNNIYF-----GDN----QYTLEGAANHYFGTTVNK--------------
TGase domain 2 301 YSKDRILELYMNEVYL-----GQSGDNEIRGFPLASLYYFGRPVEELSLDQQALLVGMVK
PBP2 182 YSKDDIFQVYLNKIYY-----SDG----VTGIKAAAKYYFNKDLKDLNLAEEAYLAGLPQ
Conserved motif 3 KXXXLXXYXN
TGase domain 1 239 SPITYSPYENTGELKS--------------------------------------------
MTG 1
MTG 2
170 NPLRFKVSSPSGYVRSRQAWILRQMYQLGGEPFMQQHQLD--------------------
204 ------------------------------------------------------------
TGase domain 2 356 GASIYELQAKLGDKVK--------------------------------------------
PBP2 233 VPNNYNIYDHPKAAEDRKNTVLYLMHYHKRITDKQWEDAKKIDLKANlVNRTAEERQNID
Figure 3. Alignment of amino acid sequences of TGases and MTGs from different species. The three conservative motifs are highlighted in dark.
TGase domain 1 is TGase domain of S. pneumoniae PBP1b, MTG 1 is MTG from Staphylococcus aureus, MTG 2 is MTG from Escherichia coli,
TGase domain 2 is TGase of E. coli PBP1b, and PBP2 is PBP2 from S. aureus.

H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
7193
then immobilized on the sensor chip CM5.1 When a
solution of the isolated TGase domain passes over the
chip, the enzyme binds to the chip through specific inter-
action with moenomycin. Initial binding experiments
indicated the interaction between the enzyme and immo-
bilized moenomycin was concentration dependent, as
shown in Figure 4. The dissociation constant Kd of
12 lM was obtained from these experiments.
DNA by PCR and inserted into vector pET 26(b) and
pET 16(b) (Novagen). A His6 tag was added to its C-ter-
minus or N-terminus. After expression, enzyme was
purified by Ni2+ affinity chromatography according to
the manufacturer’s protocol (Qiagen).
4.2. Synthesis of benzoylated monosaccharide acceptor 8
Benzyl-N-acetyl muramic acid phenylsulfonyl ester 6
(415 mg, 0.75 mmol) was dissolved in 7 mL CH2Cl2.
To this solution, benzoyl chloride (0.3 mL, 1.5 mmol)
and pyridine (0.27 mL) were added at À40 °C and the
reaction mixture was warmed to 0 °C and stirred for
1 h. The reaction mixture was diluted with CH2Cl2
(50 mL) and washed with ammonia chloride once and
then brine twice. The organic layer was dried and con-
centrated in vacuo. Purification with flash chromatogra-
phy on silica gel eluting with EtOAc/hexanes (4:1)
yielded compound 8 0.369 g (75%) as a white solid:
Rf = 0.2 (EtOAc/hexanes 4:1); 1H NMR (CDCl3,
500 MHz) d 8.09 (d, J = 7.92, 2H), 7.96 (d, J = 7.86,
2H), 7.73 (m, 1H), 7.62–7.54 (m, 3H), 7.48–7.45 (m,
2H), 7.35–7.26 (m, 5H), 5.34 (d, J = 3.30, 1H), 4.91
(dd, J = 2.95, 12.1, 1H), 4.73 (d, J = 11.8, 1H), 4.56–
4.51 (m, 3H), 4.47 (m, 1H), 4.22 (dd, J = 2.20, 12.5,
1H), 3.81–3.76 (m, 2H), 3.64–3.56 (m, 2H), 3.45 (m,
3H), 1.99 (s, 3H), 1.24 (d, J = 6.95, 3H); HR-FTMS
(pos) calcd for C33H37NO11S [M+Na]+ = 678.198,
found 678.1991.
This binding assay can be used to analyze potential
inhibitors by flowing them over the chip to observe the
dissociation of the enzyme. If the inhibitor cannot com-
pete with moenomycin for binding with the enzyme, a
large SPR signal will be observed. If the inhibitor com-
petes with moenomycin for binding with the enzyme, a
small signal will be observed. A mock competition
experiment in solution was performed using moenomy-
cin as a competitive inhibitor. The elution of the enzyme
from the chip was observed. As the concentration of
flowing moenomycin increased, more and more enzyme
was eluted from the chip. We found that when the con-
centration of moenomycin reached at 160 lM, 50% dis-
sociation of TGase domain from the chip occurred. This
result is in good agreement with the value determined
previously using PBP1b from E. coli.25 This SPR bind-
ing assay is now being used to identify new inhibitors.
3. Conclusion
We have characterized several MTGs including a TGase
domain from S. pneumoniae PBP1b. The TGase domain
can be expressed and purified from E. coli with good
yield, and it is fully active in the absence of the trans-
membrane helix and the TPase domain. To set up an
efficient enzyme assay for evaluating inhibitors of trans-
glycosylase, we synthesized a nitrophenol and a dansyl
analogue of Lipid II. The dansyl Lipid II analogue
was used in a coupled fluorescence assay to evaluate
the TGases we expressed. An SPR binding assay using
moenomycin as the tether also works well for the TGase
domain. This assay does not require the substrate which
is difficult to synthesize.
4.3. Synthesis of tert-butyldimethylsilyl monosaccharide
acceptor 13
Benzyl-N-acetyl muramic acid phenylsulfonyl ester 6
(166 mg, 0.30 mmol) was dissolved in 7 mL pyridine.
To this solution, tert-butyl, dimethylsilylchloride
(54.4 mg, 0.36 mmol) was added at 0 °C and the reaction
mixture was warmed to room temperature and stirred
for overnight. The reaction mixture was diluted with
CH2Cl2 (50 mL) and washed with brine twice. The
organic layer was dried and concentrated in vacuo. Puri-
fication with flash chromatography on silica gel eluting
with EtOAc/hexanes (4:1) yielded compound 13
0.369 g (75%) as a white solid: Rf = 0.5 (EtOAc/hexanes
1
The high activity of the isolated TGase domain from the
full-length Pbp1b of S. pneumoniae and its compatibility
with the above two assays should enable us to search for
inhibitors of transglycosylases, which may serve as new
antibiotics that are less prone to resistance development.
In the meantime, we are trying to solve the crystal struc-
ture of the TGase domain to guide our inhibitor search
on a more rational basis.
4:1); H NMR (CDCl3, 600 MHz) d 7.91 (d, J = 7.44,
2H), 7.66 (d, J = 7.44, 2H), 7.56 (t, J = 7.44, 2H), 7.33
(m, 4H), 7.23 (m, 1H), 5.26 (d, J = 3.06, 1H), 4.65 (d,
J = 11.8, 1H), 4.52 (m, 4H), 3.84 (m, 1H), 3.77 (m,
1H), 3.72–3.45 (m, 10H), 1.98 (s, 3H), 1.25 (d,
J = 7.02, 3H), 0.91 (s, 9H), 0.10 (d, J = 3.48, 6H);
HR-FTMS
(pos)
calcd
for
C32H47NO10SSi
[M+Na]+ = 688.2582, found 688.2568.
4.4. Synthesis of benzylated monosaccharide acceptor 12
4. Experimental
Benzyl-N-acetyl-4,6-benzylidine-muramic acid phenyl-
sulfonyl ester 6 (500 mg, 0.78 mmol) was dissolved in
3.1 mL THF with 1 M sodium boroncyanide hydride.
To this solution, 4 N HCl in dioxane was added to make
the solution acidic and the reaction mixture was stirred
for 8 h. The reaction mixture was concentrated in vacuo
and then diluted with CH2Cl2 (50 mL) and washed with
brine twice. The organic layer was dried and
4.1. Generation of transglycosylase expression constructs
The locations of potential transmembrane domains in S.
pneumoniae, S. aureus, and E. coli transglycosylases were
predicted using PSORT and TMpred software analysis
programs. The genes of transglycosylase were cloned
from S. pneumoniae, S. aureus, and E. coli genomic

7194
H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
concentrated in vacuo. Purification with flash chroma-
tography on silica gel eluting with EtOAc/hexanes
(4:1) yielded compound 12 0.426 g (85%) as a white sol-
id: Rf = 0.3 (EtOAc/hexanes 4:1); 1H NMR (CDCl3,
600 MHz) d 7.90 (d, J = 7.44, 2H), 7.65 (t, J = 7.44,
1H), 7.55 (d, J = 7.86, 2H), 7.33–7.24 (m, 11H), 5.28
(d, J = 3.54, 1H), 4.66 (d, J = 11.8, 1H), 4.60 (d,
J = 11.8, 1H), 4.47–4.43 (m, 5H), 3.82 (m, 1H), 3.72
(m, 3H), 3.63 (m, 3H), 3.62 (m, 2H), 3.47 (m, 2H),
3.33 (m, 1H), 1.98 (s, 3H), 1.24 (d, J = 7.02, 3H);
4.7. Disaccharide phosphate
Disaccharide acetate 11 (233 mg, 0.02 mmol) was
hydrogenolyzed in MeOH overnight. Then the reaction
mixture was filtered and concentrated in vacuo. 1H-Tet-
razole (6.5 mg 0.09 mmol) was added and the mixture
was coevaporated with toluene three times. Then the
mixture was dissolved in dried CH2Cl2 (4 mL) and
cooled to À30 °C. To this solution, dibenzyl diet-
hylphosphoramidate (65 uL, 0.03 mmol) was added.
The reaction mixture was warmed to room temperature
and stirred for 30 min. Then the reaction mixture was
cooled to À40 °C and m-CPBA (52 mg, 0.2 mmol) was
added and the reaction was stirred at 0 °C for 30 min.
The reaction mixture was diluted with CH2Cl2 (50 mL)
and washed with sodium sulfite, sodium bicarbonate,
and brine twice, respectively. The organic layer was
dried and concentrated in vacuo. Purification with flash
chromatography on silica gel eluting with EtOAc/hex-
anes (2:3) yielded the titled compound 0.015 g (65%)
as a white solid: Rf = 0.5 (EtOAc/hexanes 3:2); 1H
NMR (CDCl3, 500 MHz): characteristic peaks: d
8.12–7.30 (m, 15H), 5.86 (m, 1H), 4.42 (d, J = 6.60,
2H), 4.33 (m, 1H), 4.09 (d, J = 9.90, 2H), 3.98 (t,
J = 9.15, 1H), 2.16 (s, 3H), 2.09 (s, 3H), 2.07 (s, 3H),
2.03 (s, 3H), 1.57 (d, J = 7.0, 3H), 1.52 (d, J = 6.6,
3H), 1.49 (d, J = 7.0, 3H), 1.46 (d, J = 7.35, 3H);
HR-FTMS
(pos)
calcd
for
C33H39NO10S
[M+Na]+ = 664.2187, found 664.2165.
4.5. Glycosylation
Donor 9 (233 mg, 0.40 mmol), acceptor 8 (130 mg,
0.20 mmol) and molecular sieve AW300 (3 g) were dried
under vacuo overnight. The mixture was dissolved in
dried CH2Cl2 (30 mL) and NIS (141 mg, 0.60 mmol)
was added at À40 °C. To this mixture, triflic acid
(597 lL 0.5 M in ether) was added and the reaction mix-
ture was stirred for 1 h. The reaction mixture was fil-
tered and concentrated in vacuo and then diluted with
CH2Cl2 (50 mL) and washed with brine twice. The
organic layer was dried and concentrated in vacuo. Puri-
fication with flash chromatography on silica gel eluting
with EtOAc/hexanes (3:2) to EtOAc/hexanes (1:3) yield-
ed compound 10 0.194 g (65%) as a white solid: Rf = 0.5
(EtOAc/hexanes 3:1); 1H NMR (CDCl3, 600 MHz) d
8.09 (d, J = 7.50, 2H), 7.96 (d, J = 7.44, 1H), 7.73 (t,
J = 7.44, 2H), 7.66–7.62 (m, 4H), 7.52 (t, J = 7.86,
2H), 7.32 (m, 3H), 6.20 (d, J = 9.66, 1H), 5.33 (d,
J = 3.54, 1H), 5.09 (t, J = 9.6, 1H), 4.98 (t, J = 10.08,
1H), 4.64 (d, J = 12.3, 1H), 4.56–4.50 (m, 4H), 4.47 (q,
J = 6.12, 1H), 4.33–4.28 (m, 3H), 4.12 (q, J = 9.66,
1H), 3.92 (m, 1H), 3.85–3.80 (m, 1H), 3.79 (m, 1H),
3.56–3.47 (m, 2H), 3.35 (m, 1H), 2.03 (s, 3H), 2.01 (s,
3H), 2.00 (s, 3H), 1.99 (s, 3H), 1.98 (s, 3H), 1.21 (d,
J = 7.02, 3H); HR-FTMS (pos) calcd for C47H56N2O19S
[M+Na]+ = 1007.309, found 1007.3059.
HR-FTMS
(pos)
calcd
for
C70H92F3N8O28S
[M+Na]+ = 1603.5603, found 1603.5689.
4.8. Nitrophenol Lipid II analogue 1
The procedure was similar as the preparation of nitro-
1
phenol Lipid I analogue. H NMR (D2O, 600 MHz) d
8.27 (d, J = 8.76, 2H), 7.32 (d, J = 8.76, 2H), 5.41 (m,
1H), 4.50 (t, J = 7.86, 1H), 4.25–4.11 (m, 6H), 4.14 (t,
J = 7.02, 1H), 3.99 (m, 1H), 3.85 (m, 4H), 3.79 (d,
J = 11.88, 1H), 3.68 (m, 5H), 3.48 (t, J = 8.76, 1H),
3.33 (m, 2H), 2.92 (t, J = 6.96, 2H), 2.27 (m, 2H), 1.92
(s, 3H), 1.86 (s, 3H), 1.71 (m, 2H), 1.61 (m, 2H), 1.34
(m, 14H); ESI (neg) calcd for C45H71N9O28P2
[MÀH]À = 1246, found 1246.
4.6. Disaccharide acetate 11
Disaccharide 10 (147 mg, 0.15 mmol) was dissolved in
acetic acid (10 mL) and zinc dust (1.48 g) was added
and the reaction mixture was stirred for 2 h. The reac-
tion mixture was filtered and concentrated in vacuo
and then was dissolved in 5 mL pyridine and 5 mL
acetic anhydride was added. The reaction mixture
was stirred for overnight. The reaction mixture was
quenched by adding MeOH and was concentrated in
vacuo. Purification with flash chromatography on sili-
ca gel eluting with (EtOAc/hexanes 3:1) yielded the
titled compound 11 0.150 g (85%) as a white solid:
Rf = 0.5 (EtOAc/hexanes 3:1); 1H NMR (CDCl3,
600 MHz) d 8.02–7.34 (m, 15H), 5.44 (d, J = 3.96,
1H), 5.09 (t, J = 11.5, 1H), 4.96 (m, 2H), 4.75–4.61
(m, 2H), 4.59–4.53 (m, 2H), 4.50–4.49 (m, 2H),
4.44–4.40 (m, 2H), 4.15–4.05 (m, 1H), 3.99–3.82 (m,
3H), 3.63–3.15 (m, 5H), 3.47 (m, 1H), 2.13 (s, 3H),
2.12 (s, 3H), 2.11 (s, 3H), 2.04 (s, 3H), 1.80 (s, 3H),
1.26 (d, J = 7.02, 3H); HR-FTMS (pos) calcd for
C40H50N2O19S [M+Na]+ = 917.2621, found 917.2616.
4.9. Fluorescence coupled enzyme assay
Reactions were carried out in a Quartz cuvette in 200 lL
volume and the decrease in fluorescence was monitored
at 520 nm using Hitachi F-2000. Each reaction mixture
contained MTG reaction buffer (50 mM PIPES, pH
6.1, 50 mM MgCl2, 10 mM CaCl2, 10% DMSO, and
0.2 mM decyl-PEG), 0.6 lg N-acetylmuramidase, and
an appropriate amount of dansyl-labeled Lipid II ana-
logue and 1 lL enzyme. All of the components except
for the enzyme were premixed in an Eppendorf tube
and incubated at room temperature for 10 min. Then
the enzyme was added and the fluorescence was moni-
tored. The negative control was done by adding 1 lL
buffer instead of 1 lL enzyme to the reaction mixture.
4.10. Surface plasmon resonance assay
The SPR measurements were carried out with Biacore
3000 with research grade sensor chips. (1) Coupling of

H. Liu, C.-H. Wong / Bioorg. Med. Chem. 14 (2006) 7187–7195
10. Hara, H.; Suzuki, H. FEBS Lett. 1984, 168, 155.
7195
moenomycin–amine derivative to a CM5 sensor chip:
The immobilization procedure was carried out accord-
ing to the BIACORE’s protocol. The immobilization
buffer is HBS-Ep buffer biacertified, and moenomycin–
amine derivative at 0.5 mg/mL was used for immobiliza-
tion. Ethanol amine was used as the negative control for
immobilization. (2) After the immobilization, the chip
was rinsed thoroughly with buffer A (10 mM Tris–male-
ate, pH 6.6, 150 mM NaCl, and 1% Triton X-100). A
solution of TGase domain from S. pneumoniae PBP1b
11. Hitchcock, S. A.; Eid, C. N.; Aikins, J. A.; Zia-
Ebrahimi, M.; Blaszczak, L. C. J. Am. Chem. Soc.
1998, 120, 1916.
12. Ho¨ltje, J.-V. Microbiol. Mol. Biol. Rev. 1998, 62, 181.
13. Huber, G.. In Mechanism of Action of Antibacterial
Agents; Springer: New York, 1979; Vol. 5, p 135.
14. Huber, R. E.; Mathison, R. D. Can. J. Biochem. 1976, 54,
153.
15. Liu, H.; Ritter, T. K.; Sadamoto, R.; Sears, P. S.; Wu, M.;
Wong, C. H. ChemBioChem 2003, 4, 603.
16. Liu, H.; Sadamoto, R.; Sears, P. S.; Wong, C.-H. J. Am.
Chem. Soc. 2001, 123, 9916.
17. McClure, J. A.; Korn, E. D. J. Biol. Chem. 1983, 258,
14570.
18. Men, H.; Park, P.; Ge, M.; Walker, S. J. Am. Chem. Soc.
1998, 120, 2484.
19. Nakagawa, J.; Tamaki, S.; Tomioka, S.; Matsuhashi, M.
J. Biol. Chem. 1984, 259, 13937.
20. Park, W.; Matsuhashi, M. J. Bacteriol. 1984, 157, 538.
21. Pfaller, M. A.; Jones, R. N.; Doern, G. V.; Kugler, K.
Antimicrob. Agents Chemother. 1998, 42, 1762.
22. Rodriguez-Tebar, A.; Vazquez, D. Microbiol. Sci. 1984, 1,
211.
in buffer
A at constant concentrations (25 lM,
12.5 lM, 6.24 lM, 3.12 lM, 1.51 lM, 0.76 lM,
0.38 lM, 0.19 lM, 0.095 lM, and 0.048 lM) was inject-
ed and bound to the chip. The interaction was investi-
gated under constant conditions at 125 lL of protein
solution and 25 lL/min flow rate. In the elution experi-
ments, moenomycin was dissolved in buffer A and
injected at different concentrations (1 mM, 500 lM,
30 lM, 3 lM, 300 nM, 30 nM, and 3 nM) to elute the
bound protein.
23. Schwartz, B.; Markwalder, J. A.; Seitz, S. P.; Wang, Y.;
Stein, R. L. Biochemistry 2002, 41, 12552.
Acknowledgments
24. Schwartz, B.; Markwalder, J. A.; Wang, Y. J. Am. Chem.
Soc. 2001, 123, 11638.
25. Stembera, K.; Vogel, S.; Buchynskyy, A.; Ayala, J. A.;
Welzel, P. ChemBioChem 2002, 3, 559.
This research was supported by the National Institutes
of Health (GM 44154). We thank Thomas K. Ritter
for providing moenomycin amine derivative.
26. Terrak, M.; Ghosh, T. K.; van Heijenoort, J.; Van
Beeumen, J.; Lampilas, M.; Aszodi, J.; Ayala, J. A.;
Ghuysen, J. M.; Nguyen-Disteche, M. Mol. Microbiol.
1999, 34, 350.
References and notes
27. Terrak, M.; Nguyen-Disteche, M. J. Bacteriol. 2006, 188,
2528.
1. Anikin, A.; Buchynskyy, A.; Kempin, U.; Stembera, K.;
Welzel, P.; Lantzsch, G. Angew. Chem., Int. Ed. 1999, 38,
3703.
2. Barrett, D.; Leimkuhler, C.; Chen, L.; Walker, D.; Kahne,
D.; Walker, S. J. Bacteriol. 2005, 187, 2215.
3. Barrett, D.; Chen, L.; Litterman, N. K.; Walker, S.
Biochemistry 2004, 43, 12375.
4. Bauer, M. W.; Bylina, E. J.; Swanson, R. V.; Kelly, R. M.
J. Biol. Chem. 1996, 271, 23749.
5. Chu, D. T.; Plattner, J. J.; Katz, L. J. Med. Chem. 1996,
39, 3853.
28. van Heijenoort, Y.; Gomez, M.; Derrien, M.; Ayala, J.;
van Heijenoort, J. J. Bacteriol. 1992, 174, 3549.
29. VanNieuwenhze Michael, S.; Mauldin Scott, C.; Zia-
Ebrahimi, M.; Winger Brian, E.; Hornback William, J.;
Saha Shankar, L.; Aikins James, A.; Blaszczak Larry, C.
J. Am. Chem. Soc. 2002, 124, 3656.
30. VanNieuwenhze, M. S.; Mauldin, S. C.; Zia-Ebrahimi,
M.; Aikins, J. A.; Blaszczak, L. C. J. Am. Chem. Soc.
2001, 123, 6983.
31. Walsh, C. T. Science 1999, 284, 442.
6. Deacon, J. Fungal Biology ; Blackwell: Oxford, 2005.
7. Dychinco, B.; Webb, V. A.; Davies, J.; Mazel, D. Science
1998, 280, 605.
8. Ghuysen, J. M. Biotechnol. Appl. Biochem. 1990, 12, 468.
9. Goffin, C.; Ghuysen, J. M. Microbiol. Mol. Biol. Rev.
(Washington, DC) 1998, 62, 1079.
32. Wang, Q. M.; Peery, R. B.; Johnson, R. B.; Alborn, W.
E.; Yeh, W. K.; Skatrud, P. L. J. Bacteriol. 2001, 183,
4779.
33. Ye, X.-Y.; Lo, M.-C.; Brunner, L.; Walker, D.; Kahne,
D.; Walker, S. J. Am. Chem. Soc. 2001, 123, 3155.