Reconstituting poly(glycerol phosphate) wall teichoic acid biosynthesis in vitro using authentic substrates

Article abstract of DOI:10.1039/c4sc00802b

Wall teichoic acids (WTAs) are phosphate-rich anionic polymers that constitute a substantial portion of the Gram-positive cell wall. Recent work has demonstrated the importance of WTAs in cell shape, virulence and antibiotic resistance. These findings highlight WTA biosynthetic enzymes as attractive targets for novel antimicrobial agents. Due to challenges involved in the isolation of natural substrates, in vitro studies of the recombinant enzymes have largely employed soluble substrate analogues. Herein we present a semisynthetic approach to obtain the authentic precursor for WTA biosynthesis, Lipid α, complete with its polyisoprenoid lipid moiety. We show that this material can be used to reconstitute the activities of four enzymes involved in poly(glycerol phosphate) WTA biosynthesis in a detergent micelle. This work enables the creation of chemically defined and realistic systems for the study of interfacial catalysis by WTA biosynthetic machinery, which could aid efforts to discover and develop novel agents against WTA biosynthesis.

Full text of DOI:10.1039/c4sc00802b

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
Reconstituting poly(glycerol phosphate) wall
teichoic acid biosynthesis in vitro using authentic
substrates†
Cite this: Chem. Sci., 2014, 5, 3823
a
a
b
a
*
Robert T. Gale, Edward W. Sewell, Teresa A. Garrett and Eric D. Brown
Wall teichoic acids (WTAs) are phosphate-rich anionic polymers that constitute a substantial portion of the
Gram-positive cell wall. Recent work has demonstrated the importance of WTAs in cell shape, virulence and
antibiotic resistance. These findings highlight WTA biosynthetic enzymes as attractive targets for novel
antimicrobial agents. Due to challenges involved in the isolation of natural substrates, in vitro studies of the
recombinant enzymes have largely employed soluble substrate analogues. Herein we present
a
semisynthetic approach to obtain the authentic precursor for WTA biosynthesis, Lipid a, complete with its
polyisoprenoid lipid moiety. We show that this material can be used to reconstitute the activities of four
enzymes involved in poly(glycerol phosphate) WTA biosynthesis in a detergent micelle. This work enables
the creation of chemically defined and realistic systems for the study of interfacial catalysis by WTA
biosynthetic machinery, which could aid efforts to discover and develop novel agents against WTA biosynthesis.
Received 18th March 2014
Accepted 16th June 2014
DOI: 10.1039/c4sc00802b
www.rsc.org/chemicalscience
Despite ready access to recombinant enzymes involved in
WTA biosynthesis, challenges in the isolation of natural
Introduction
substrates from bacterial sources have hindered detailed
studies. Efforts to date have involved the synthesis of soluble
analogues of the substrates to study WTA biosynthesis in
The bacterial cell wall has been targeted with spectacular success
in the development of antibiotics. Indeed, b-lactams and glyco-
peptides that inhibit cell wall biosynthesis are the most widely
used chemotherapeutic agents to treat bacterial infections.1
Nevertheless, increasing resistance to these agents is a serious
threat to their continued use and to public health.2 Thus, the
search is on for new agents of unique chemical class that are
unsusceptible to existing resistance mechanisms and are effec-
tive perturbants of this celebrated and well-validated target.
In addition to peptidoglycan, the Gram-positive bacterial cell
wall is largely composed of long, phosphate-rich, anionic poly-
mers called wall teichoic acids (WTAs).3 These polymers are
typically comprised of repeating polyol phosphate residues
modied with D-alanyl and glycosyl substituents.4,5 WTAs have
emerged as attractive antibacterial targets6 due to their impor-
tant roles in cell shape determination,7 virulence8–10 and anti-
biotic resistance.11,12 While studies of the genetics and
physiology of WTA biosynthesis have provided strong validation
for this biosynthetic pathway as a target for new antibiotics, a
lack of tools and understanding of the biochemistry of WTA
synthesis remains an obstacle to its effective utility in modern
antibiotic drug discovery.
vitro.13,14,17 These soluble analogues substitute
a short
aliphatic or prenyl chain for the C55 undecaprenyl moiety of
the authentic precursor glycolipid, GlcNAc-PP-undecaprenol.
These analogues provided ready access to pure substrates for
the functional characterization of several WTA biosynthetic
enzymes,13–17 and have largely underpinned our current
understanding of the biosynthetic pathway. These soluble
analogues, however, fail to recapitulate the interfacial
nature of catalysis of WTA glycolipids. Indeed, the many
glycosyltransferase enzymes and the ippase transporter
involved in WTA synthesis are membrane bound and thus
the membrane interface is central to synthesis and assembly
of WTA.6
Here we present a semisynthetic strategy to obtain the
authentic WTA glycolipid precursor GlcNAc-PP-undecaprenol,
also known as Lipid a (Table 1 outlines the nomenclature for
WTA intermediates). Further we show that this molecule, in a
detergent micelle, is a capable substrate for the activities of
four consecutive recombinant enzymes – TagA, the ManNAc
transferase, TagB, the poly(glycerol phosphate) primase,
TagF, the poly(glycerol phosphate) polymerase and TagE, the
aMichael G. DeGroote Institute of Infectious Disease Research and Department of
Biochemistry and Biomedical Sciences, McMaster University, Hamilton, Ontario L8N
3Z5, Canada. E-mail: ebrown@mcmaster.ca
bDepartment of Chemistry, Vassar College, Poughkeepsie, New York 12604, USA
† Electronic supplementary information (ESI) available: Supplementary gures,
synthetic protocols and characterization data. See DOI: 10.1039/c4sc00802b
poly(glycerol phosphate) glucosyltransferase
–
in the
synthesis of poly(glycerol phosphate) WTA in vitro. This work
provides an important new avenue for the characterization of
WTA enzymes at interfaces using authentic and chemically
dened substrates.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3823–3830 | 3823

View Article Online
Chemical Science
Edge Article
Table 1 Nomenclature for undecaprenyl-linked WTA intermediatesa
Intermediate
Chemical composition
Lipid a
Lipid b
Lipid f.n
Lipid g
GlcNAc-1-P-P-Und
ManNAc-b-(1–4)-GlcNAc-1-P-P-Und
(GroP)n-ManNAc-b-(1–4)-GlcNAc-1-P-P-Und
(GroP*)n-ManNAc-b-(1–4)-GlcNAc-1-P-P-Und
a
WTA intermediates are named according to the enzyme utilizing the
molecule as a substrate. n represents the number of repeating sn-
glycerol-3-phosphate units. The asterisk (*) denotes oligomeric and
polymeric species bearing a-linked glucose residues. Abbreviations:
GlcNAc, N-acetylglucosamine; ManNAc, N-acetylmannosamine; GroP,
sn-glycerol-3-phosphate; P, phosphate; Und, undecaprenol.
Results and discussion
Semisynthetic preparation of Lipid a
WTA biosynthesis utilizes undecaprenyl phosphate (4) as a lipid
carrier to build glycosyl intermediates (Fig. 1 details the
biosynthetic pathway). This lipid carrier is essential for trans-
port of oligosaccharides across the bacterial membrane during
synthesis of many other important polysaccharides including
lipopolysaccharides,18,19 peptidoglycans20 and capsular poly-
saccharides.21 Undecaprenyl phosphate (4) was required in
quantity for use in chemical transformations to generate Lipid a
(5) and, while this molecule is commercially available, its cost is
prohibitive for purchase in useful quantities. It can be obtained
from bacterial sources, however it represents a relatively minor
species relative to other membrane lipids22 and thus it is chal-
lenging to isolate and purify. Compound 4 is also available
through total synthesis, however the methodology involves a
cumbersome multistep procedure.23 Thus, we elected to obtain
undecaprenyl phosphate (4) through a semisynthetic approach,
whereby the precursor undecaprenol (3) was obtained from a
plant source, namely from Laurus nobilis leaves (bay leaves),
according to published methodology.24–26 Subsequent phos-
phorylation of the polyprenol was achieved with tetra-n-buty-
lammonium dihydrogen phosphate and trichloroacetonitrile
according to published methods,27 to afford undecaprenyl
phosphate (4) in 40% yield over the phosphorylation step.
Lipid a (5) represents a challenging synthetic target due to
the acid sensitivity of the anomeric diphosphate linkage. For
this reason, our synthetic strategy to the undecaprenyl-linked
glycosyl diphosphate involved a late stage introduction of the
diphosphate linkage, and use of base-cleavable protecting
groups that allowed for a nal global deprotection step. Thus,
our strategy involved (a) peracetylation of commercially avail-
able N-acetylglucosamine (1), (b) anomeric acetate deprotection,
(c) hemiacetal a-phosphorylation, (d) undecaprenyl phosphate
(4) coupling, and (e) base-mediated global deprotection (Scheme
1). The synthesis began with the preparation of a-phosphate 2 in
25% yield over ve steps following published methodology.13,14,28
Coupling of undecapenyl phosphate (4) with the a-phosphate 2
was conducted with 1,10-carbonyldiimidazole (CDI) and tin(II)
chloride. Global acetate deprotection using sodium methoxide in
methanol transformed the resulting peracetyl diphosphate
intermediate to desired target Lipid a (5) in 20% yield over
Fig. 1 Poly(glycerol phosphate) WTA biosynthesis in B. subtilis 168.
WTA biosynthesis takes place at the inner leaflet of the cytoplasmic
membrane. The first reaction in the pathway is mediated by the
transmembrane protein TagO, wherein a UDP-GlcNAc precursor is
coupled to a membrane-anchored undecaprenyl lipid carrier to
produce undecaprenylpyrophosphoryl-GlcNAc, also known as Lipid a
(5).49 In a second reaction, TagA catalyzes the transfer of ManNAc from
a UDP-ManNAc precursor to the C(4) hydroxyl group of the lipid-
linked GlcNAc, forming Lipid b.13 TagB catalyzes the transfer of glycerol
phosphate from a CDP-glycerol precursor to the lipid-linked disac-
charide.13,31 An additional 25–35 glycerol phosphate monomers are
added to this product, Lipid f.1, by the polymerase TagF utilizing CDP-
glycerol precursors.14,15,34 The polymer, Lipid f.n, is tailored with a-
linked Glu residues by the glycosyltransferase TagE through use of
UDP-Glu precursors, forming Lipid g.16 The polymer is translocated to
the extracellular leaflet through the action of the two-component ABC
transporter TagGH.50 The last step in B. subtilis 168 poly(glycerol
phosphate) WTA biosynthesis is mediated by currently uncharac-
terized transferases, which are predicted to cleave the polymer from
its lipid anchor and attach it to the C(6) hydroxyl group of peptido-
glycan muramic acid.47 Abbreviations: GlcNAc, N-acetylglucosamine;
ManNAc, N-acetylmannosamine; Glu, glucose; UDP, uridine diphos-
phate; CDP, cytidine diphosphate.
coupling and deprotection steps and in 5% yield over the entire
synthetic procedure. We found that yields of the coupling reac-
tion were higher with the use of tin(II) chloride, which agree with
recent ndings highlighting the effectiveness of this agent for
promoting phosphate–phosphate coupling.29 In addition, our
modest yield is commensurate with those reported for other CDI-
mediated coupling reactions yielding lipid-linked glycosyl
diphosphates.29,30 Detailed synthetic procedures and chemical
characterization can be found in the ESI.†
In vitro reconstitution of TagA and TagB activities
Access to Lipid a (5) allowed us to begin studying WTA
biosynthesis using authentic substrates. We rst reconstituted
the activities of two B. subtilis 168 enzymes involved in poly(-
glycerol phosphate) biosynthesis in vitro. These enzymes, TagA
and TagB, transform Lipid a (GlcNAc-PP-Und) to Lipid b
3824 | Chem. Sci., 2014, 5, 3823–3830
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
Scheme 1 Semisynthetic preparation of Lipid a (5).
Fig. 2 Semisynthetic Lipid a (5) can be used to reconstitute the activities of pure, recombinant WTA biosynthetic enzymes TagA and TagB in vitro.
(a) Reactions catalyzed by TagA and TagB in vivo and in our in vitro system. The asterisk (*) denotes 14C-labelled compounds in assays employing
semisynthetic Lipid a (5). Nomenclature and chemical composition for each WTA intermediate is shown in bold and in brackets respectively. (b)
Product characterization by TLC autoradiography and ESI-MS of WTA intermediates synthesized in vitro. Relevant reaction components are
indicated (+/ꢀ) below the appropriate lane and reaction products confirmed by ESI-MS are indicated to the right of the frame with experimental
m/z values shown (for associated MS spectra, see Fig. S1†). In reactions involving semisynthetic derived Lipid a (5) (lanes 2–4), radioactivity was
incorporated into reaction products from a UDP-[14C]ManNAc precursor (see panel a). Lipid-linked reaction products were extracted according
to published methodology.12 Lane 1, [14C]-Lipid a standard prepared from membranes of Escherichia coli cells expressing recombinant TarO
from Staphylococcus aureus following known methods;12 lane 2, TagA-mediated reaction of semisynthetic Lipid a (5) to Lipid b; lane 3, TagA and
TagB-mediated reaction of semisynthetic Lipid a (5) to Lipid f.1; lane 4, control reaction without either sources of Lipid a (5).
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3823–3830 | 3825

View Article Online
Chemical Science
Edge Article
(ManNAc-GlcNAc-PP-Und) and Lipid b to Lipid f.1 (GroP- corresponds to the ManNAc-GlcNAc-P and undecaprenyl
ManNAc-GlcNAc-PP-Und), respectively.13,31
diphosphate51 respectively from Lipid b (Fig. S2a,† inset).
Incubation of Lipid a mixed-micelles with TagA, TagB, UDP-
We reconstituted Lipid a (5) into detergent micelles con-
taining Triton-X100 to prevent substrate aggregation and to [14C]ManNAc and CDP-glycerol produced a radioactive lipid
provide an interface for catalysis. This approach has previously species more polar than both Lipid a and Lipid b (Fig. 2b). LC/
proved effective for reconstituting functional enzyme activities MS analysis of this species showed a peak at m/z ¼ 742.382
in assays containing undecaprenyl-linked substrates.30,32 We corresponding to the [M ꢀ 2H+]2ꢀ ion of Lipid f.1 (Fig. S1b†).
assessed the ability of TagA and TagB to make their respective MS/MS analysis of this [M ꢀ 2H+]2ꢀ ion (Fig. S2b†) showed the
products in this system through a combination of analytical formation of several predominant product ions consistent with
TLC and high-resolution mass spectrometry (HRMS) (Fig. 2). the structure of Lipid f.1. The ion at m/z 152.996 corresponds to
The radioactive precursor UDP-[14C]ManNAc was used to C3H6O5Pꢀ and is typical of lipids containing glycerol phos-
incorporate radioactivity into products aer TagA-mediated phate.52 The product ions at m/z 639.123, 577.157 and 359.043
catalysis (Fig. 2a). This allowed us to monitor product formation correspond to ions derived from the GroP-ManNAc-GlcNAc-P
following lipid extraction12,33 by analytical TLC autoradiography portion of Lipid f.1 as indicated in Fig. S2b inset.† The ion at m/
(Fig. 2b). Overnight incubation of Lipid a containing micelles z 907.614 is indicative of undecaprenyl diphosphate.51
with TagA and UDP-[14C]ManNAc led to the formation of a
radioactive lipid species more polar than a radiolabelled Lipid a
standard derived from a bacterial membrane source12 as visu-
alized by normal-phase TLC. LC/MS analysis of this species
showed a peak at m/z ¼ 1331.789 corresponding to the [M ꢀ
H+]ꢀ ion of Lipid b (Fig. S1a†). Negative ion collision-induced
dissociation mass spectrometry (MS/MS) analysis of the in vitro
product (m/z 1331.8, Fig. S1a†) showed the formation of two
predominant product ions at m/z 485.116 and 907.601, which
In vitro reconstitution of TagF and TagE activities
The formation of Lipid b and Lipid f.1 intermediates in our in
vitro assays demonstrated that semisynthetic Lipid a was a
capable substrate for the WTA machinery. Therefore, using
Lipid f.1 generated in vitro, we sought to reconstitute the
remaining steps involved in the intracellular biosynthesis of
poly(glycerol phosphate) WTA. To accomplish this, Lipid f.1
was further transformed to Lipid f.n ([GroP]n-ManNAc-GlcNAc-
Fig. 3 Authentic late-stage WTA intermediates are substrates for polymerization and glucosylation. (a) In vivo and in vitro reactions mediated by
the poly(glycerol phosphate) polymerase TagF and the glucosyltransferase TagE. Nomenclature and chemical composition for each late-stage
wall teichoic acid intermediate is shown in bold and in brackets respectively. (b) Product characterization by anion exchange HPLC of late-stage
WTA intermediates synthesized in vitro. In reactions involving TagF (shown in blue), Lipid f.1 was prepared through TagA and TagB-mediated
reactions starting with semisynthetic Lipid a (5) and purified by lipid extraction. Radioactivity was incorporated into polymeric products using a
CDP-[14C]Gro precursor. In reactions involving TagE (shown in red), Lipid f.n polymer was prepared in a non-radioactive assay using Lipid f.1
and TagF. Radioactivity was incorporated into this substrate following incubation with TagE and UDP-[14C]glucose. Radioactive reaction
components were separated by anion exchange HPLC and visualized with in-line scintillation counting. (i) TagF reaction mixture prior to the
addition of enzyme (T ¼ 0); (ii) NaOH-treated (0.5 M NaOH, 37 ꢁC, 25 min) TagF reaction mixture after 4 hours of incubation with enzyme; (iii)
TagE reaction mixture prior to addition of enzyme (T ¼ 0); (iv) NaOH-treated (0.5 M NaOH, 37 ꢁC, 25 min) TagE reaction mixture after 4 hours of
incubation with enzyme. Predicted radioactive components prior to mild alkali treatment are indicated to the right of the panel.
3826 | Chem. Sci., 2014, 5, 3823–3830
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
PP-Und) using the poly(glycerol phosphate) polymerase TagF.34 targets owing to the crucial roles they play in Staphylococcus
We then tailored this polymer with a-linked glucose residues aureus host colonization,8,39 peptidoglycan synthesis co-ordi-
using the glucosyltransferase TagE16 to generate Lipid g nation40 and b-lactam resistance.11,12 These structures have been
([GroP*]n-ManNAc-GlcNAc-PP-Und) (Fig. 3a). We developed a targeted with success recently in both academe12,41 and the
radioactive anion exchange HPLC-based assay to monitor pharmaceutical sector,42 where biosynthetic inhibitors have
product formation during enzyme-mediated reactions (see been shown to restore the efficacy of b-lactams against methi-
Methods). Radioactive precursors CDP-[14C]glycerol and UDP- cillin-resistant S. aureus (MRSA). While progress in the
[14C]GlcNAc were used to incorporate radioactivity into prod- discovery of inhibitors using cell-based approaches has been
ucts following incubation with TagF and TagE respectively. Our promising, many biochemical details of WTA synthesis remain
initial attempts with this approach failed to reveal radioactive obscure.
products that could be detected by HPLC aer incubation of
Biochemical study of WTA enzymes has been hampered by
TagF with non-radioactive Lipid f.1 and CDP-[14C]glycerol (data the limited availability of natural substrates. Recent use of
not shown). We hypothesized that the lipid moiety of the TagF soluble substrate analogues of WTA biosynthetic intermediates
product may have interfered with effective chromatographic has facilitated the study of WTA enzymes in vitro, away from the
separation, likely through aggregation and resulting heteroge- membrane. These studies have culminated in dening the
neity. Accordingly, we treated the reaction products with mild functions of WTA enzymes;13,14,16 obtaining kinetic parameters
alkali conditions (0.5 M NaOH, 37 C, 25 min), well known to for enzyme-mediated reactions;13–17 discerning substrate pref-
ꢁ
liberate poly(glycerol phosphate) polymers from their disac- erences;17 elucidating steps involved in WTA assembly;43,44
charide anchors (Fig. S4a†).35 Anion exchange HPLC analysis of describing mechanism15,16 and mode15 of catalysis; and helping
this sample revealed a broad peak with a greater retention time to unveil WTAs unique role in antibiotic resistance.11 Despite
than its radioactive precursor (Fig. 3b). This result suggested the utility these analogues have displayed in enhancing
that the TagF product was highly anionic. To conrm this, we understanding of WTA biosynthesis, they are unable to mimic
synthesized previously characterized poly(glycerol phosphate) the natural membrane environment where WTA biosynthetic
polymers14,15,36 (Fig. S3a†) of distinct chemical composition and enzymes function.31,34,45 Thus, even basic features of interfacial
compared their elution proles to the alkali-treated TagF reac- WTA synthesis and assembly have eluded characterization.
tion. We found that all polymers had a similar retention time to
In the work presented here, we detail a robust strategy to
the TagF product (Fig. S3b†).
obtain in quantity both undecaprenyl phosphate (4) and a
Treatment of poly(glycerol phosphate) WTAs with mild acid valuable intermediate in WTA biosynthesis, Lipid a (5). We
conditions is known to hydrolyze phosphodiester linkages show that Lipid a is a capable substrate for poly(glycerol
between glycerol phosphate residues (Fig. S4a†).34,35 Thus, we phosphate) WTA by reconstituting the activities of WTA
further subjected the alkali-treated TagF reaction mixture to mild enzymes TagA, TagB, TagF and TagE in vitro. This work
acid conditions (1 N HCl, 100 ꢁC, 3 h) and monitored hydrolysis contributes to other efforts broadly focused on the application
by anion exchange HPLC. Under these conditions, we observed a of authentic undecaprenyl-linked substrates for the in vitro
complete loss of signal from the polyanionic species (Fig. S4b†). study of bacterial cell envelope biosynthesis, namely O-poly-
Hydrolytic products coincided with the retention time of saccharide biosynthesis in Escherichia coli30 and peptidoglycan
[3H]glycerol (Fig. S5†), a well-known product of WTA hydrolysis.34 assembly in pathogenic Streptococcus pneumonia.32 Studies of
Indeed, we subjected a previously characterized poly(glycerol this nature may provide new mechanistic information that is
phosphate) polymer15 to these treatments and observed similar otherwise unobtainable through employment of soluble
degradation (Fig. S4†). Taken together, the characteristic WTA analogues and signicantly enhance our limited understanding
lability patterns of the TagF product demonstrate TagF's ability to of the synthesis and assembly of bacterial cell surface
make Lipid f.n in this interfacial system.
structures.
Access to Lipid a and other WTA intermediates reported here
We have previously shown that recombinant TagE can glu-
cosylate soluble poly(glycerol phosphate) polymers.16 Likewise, lays an important foundation to probe the molecular details of
using the authentic Lipid f.n substrate and UDP-[14C]glucose, interfacial WTA biosynthesis. We envision the employment of
we observed a TagE-dependent signal at a longer retention time these compounds in mixed-micelle or unilamellar vesicle
than the radioactive precursor using anion exchange HPLC systems to properly assess the inuence of the membrane
(Fig. 3b). This retention time was similar to that of other well- interface on catalytic efficiency, substrate binding/dissociation,
characterized WTA polymers and alkali-treated Lipid f.n processivity and length regulation by WTA enzymes. Further,
produced from TagF (Fig. S3†). These ndings are consistent our methods are not limited to the model Gram-positive
with the conclusion that TagE is able to a-glucosylate poly(- bacterium B. subtilis. The synthetic approach described here
glycerol phosphate) polymers assembled on the authentic could be used to study WTA biosynthesis in MRSA and other
undecaprenyl-linked substrate.
pathogens, as Lipid a is a common precursor for the assembly
of WTA. These authentic substrates could also be used in
crystallographic studies to further understand the molecular
structures and interactions between lipid acceptors and the
Conclusions
WTAs constitute a signicant component of the Gram-positive WTA machinery. Indeed, the recent crystal structure of the TagF
bacterial cell wall and have emerged as attractive antibiotic polymerase from Staphylococcus epidermidis was obtained for
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3823–3830 | 3827

View Article Online
Chemical Science
Edge Article
the apo-enzyme.46 Finally, synthetic poly(glycerol phosphate) assays of E. coli cell membranes.12 Briey, TarO-enriched
WTA may well be of utility for the study of currently uncharac- membranes (500 mg protein) were incubated with UDP-GlcNAc
terized enzyme(s), for example, those responsible for anchoring (300 mM), UDP-[14C]GlcNAc (0.1 mCi), 0.1% TritonX-100 and
WTA to the peptidoglycan meshwork.47
reaction buffer (50 mM Tris pH ¼ 8, 10 mM MgCl2, 1 mM EDTA)
for 2 hours at ambient temperature. Lipid-linked products were
extracted following known methods,12 and analyzed by TLC.
Experimental
General methods
High-resolution mass spectrometry
Chemicals and solvents were purchased from Sigma-Aldrich
(Oakville, ON, Canada) or Fisher Scientic (Whitby, ON, Can-
ada) unless otherwise stated. Analytical thin layer chromatog-
raphy (TLC) was carried out on silica gel 60 F254 aluminium-
backed plates from EMD Chemicals (Gibbstown, NJ, USA) or
glass-backed C18 plates from Silicycle (Quebec City, QC, Can-
ada). TLC plates were visualized by exposure to ultraviolet light
and/or exposure to iodine vapour (I2) or an acidic solution of
High-resolution mass spectra and collision-induced dissocia-
tion mass spectra (MS/MS) were collected on an Agilent 6520
quadrupole time-of-ight (Q-TOF) mass spectrometer following
previously described methods.48 All compounds were lyophi-
lized and resuspended in CHCl3/CH3OH (2 : 1) prior to analysis.
TagA and TagB in vitro reconstitution
To reconstitute the activity of TagA, dried semisynthetic Lipid a
(0.22 mM) was resuspended in a buffer containing 0.13% (v/v)
Triton X-100, 50 mM Tris (pH 8.0) and 220 mM NaCl. The
mixture was vortexed vigorously and sonicated for 5 min at
ambient temperature. TagA (11.2 mM), MnaA (6 mM) and UDP-
[14C]GlcNAc (1 mM, 0.1 mCi) were added to initiate the reaction.
The reaction mixture incubated at ambient temperature for 4
p-anisaldehyde. UDP-[14C]GlcNAc (0.1 mCi mLꢀ1
)
was
purchased from American Radiolabeled Chemicals (St. Louis,
MO, USA). Sn-[U-14C]glycerol-3-phosphate (0.05 mCi mLꢀ1),
UDP-[14C]glucose (0.02 mCi mLꢀ1) and Ultima Gold liquid
scintillation cocktail were purchased from PerkinElmer
(Woodbridge, ON, Canada). [2-3H]glycerol (1 mCi mLꢀ1) was
purchased from Amersham Biosciences (Piscataway, NJ, USA).
Recombinant TagA, TagB, TagF, TagE, TarD and MnaA were
prepared as previously described.14,16,31,34,37 Anion exchange
HPLC was conducted on a Waters (Mississauga, ON, Canada)
ACQUITY UPLC H-Class using a ThermoScientic (Waltham,
MA, USA) DNAPac PA 200 column. Radioactive components
were separated using a linear gradient from 0 / 100% 1.25 M
NaCl in 20 mM Tris (pH 8) and visualized with in-line scintil-
lation counting using a PerkinElmer Radiomatic 150TR.
ꢁ
hours then overnight at 5 C. TagB reconstitution assays were
performed identically, with the exception that CDP-glycerol
(2 mM) and TagB (6 mM) were added to the reaction mixture
aer 4 hours of incubation. A control reaction containing no
Lipid a was also set up under these conditions. Reactions were
quenched using CHCl3/CH3OH (3 : 2) and lipid-linked products
were extracted according to published methods.12,33 Reaction
extracts and [14C]-Lipid a standard were separated by normal
phase TLC using a CHCl3/CH3OH/H2O (65 : 25 : 4) solvent
system. The plates were dried and exposed overnight at ambient
temperature to a storage phosphor screen (GE Healthcare).
Radioactive compounds were visualized using a Typhoon Trio
Variable Mode Imager (GE Healthcare). Identical TagA and
TagB reconstitution reactions were set up using nonradioactive
UDP-GlcNAc for MS analysis.
Semisynthetic preparation of Lipid a (5)
Full synthetic procedures and product characterization can be
found in the ESI.†
Compound synthesis
UDP-ManNAc and UDP-[14C]ManNAc were prepared enzymati-
cally from UDP-GlcNAc and UDP-[14C]GlcNAc respectively using
TagF and TagE in vitro reconstitution
MnaA.38 CDP-glycerol and CDP-[U-14C]glycerol were synthesized Lipid f.1 was obtained from lipid extracts of non-radioactive
enzymatically using Staphylococcus aureus TarD following TagB reconstitution reactions. To reconstitute the activity of
previously described methods.37 Lipid f.1 analogue was TagF, dried Lipid f.1 was resuspended in a buffer containing
prepared chemoenzymatically as described previously14 and 0.13% (v/v) Triton X-100, 50 mM Tris (pH 8.0) and 30 mM
subsequently puried over a Waters Sep-Pak C18 Plus Cartridge MgCl2. This solution was vortexed vigorously and sonicated
using 95% MeCN in 0.1% aqueous NH4OH for compound for 5 minutes at ambient temperature. TagF (2.4 mM) and
elution. [14C]Lipid f.10 analogue was prepared through incu- CDP-[14C]glycerol (36 mM, 0.3 mCi) were added to initiate the
bation of Lipid f.1 analogue (3.6 mM) with TagF (2.4 mM) and reaction. The reaction mixture incubated at room temperature
CDP-[U-14C]glycerol (32.4 mM, 0.3 mCi) for 4 hours in 50 mM for 4 hours. An identical reaction was set up using CDP-glycerol
Tris–HCl (pH 8) and 30 mM MgCl2 at ambient temperature.15 in place of CDP-[14C]glycerol to generate non-radioactive Lipid
[14C]Lipid f.40 analogue was prepared in a similar manner f.n for TagE reconstitution assays. To reconstitute the activity of
using 140 mM CDP-[U-14C]glycerol (0.3 mCi). [14C]CDP-glycerol- TagE, UDP-[14C]glucose (36 mM, 0.4 mCi) and TagE (1 mM) were
linked polymer was prepared enzymatically through incubation added to a non-radioactive TagF reaction aer 4 hours of
of TagF (2.4 mM) and CDP-[U-14C]glycerol (36 mM, 0.3 mCi) for 4 incubation. The reaction mixture proceeded for an additional 4
hours in 50 mM Tris–HCl (pH 8), 30 mM MgCl2 and 0.13% (v/v) hours at ambient temperature. TagF and TagE reactions were
Triton X-100 at ambient temperature.36 [14C]-Lipid a standard treated with mild alkali conditions to liberate poly (glycerol
was prepared following previously described TarO activity phosphate) polymers from disacharide anchors35 for anion
3828 | Chem. Sci., 2014, 5, 3823–3830
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
exchange HPLC analysis. Briey, each reaction mixture was 15 E. W. C. Sewell, M. P. Pereira and E. D. Brown, J. Biol. Chem.,
treated to 0.5 M NaOH for 25 min at 37 ꢁC. The reaction
2009, 284, 21132–21138.
mixtures were neutralized with the addition of 0.5 M HCl prior 16 S. E. Allison, M. A. D'Elia, S. Arar, M. A. Monteiro and
to anion exchange HPLC analysis. For WTA lability experiments, E. D. Brown, J. Biol. Chem., 2011, 286, 23708–23716.
these reactions underwent further treatment in 1 N HCl for 3 17 Y.-H. Zhang, C. Ginsberg, Y. Yuan and S. Walker,
hours at 100 ꢁC (ref. 35) prior to anion exchange HPLC analysis.
Biochemistry, 2006, 45, 10895–10904.
18 M. S. Trent, J. Biol. Chem., 2001, 276, 43122–43131.
19 P. D. Rick, H. Mayer, B. A. Neumeyer, S. Wolski and D. Bitter-
Suermann, J. Bacteriol., 1985, 162, 494–503.
Acknowledgements
20 A. Bouhss, A. E. Trunkeld, T. D. H. Bugg and D. Mengin-
Lecreulx, FEMS Microbiol. Rev., 2008, 32, 208–233.
21 L. Masson and B. E. Holbein, J. Bacteriol., 1985, 161, 861–867.
We thank Dr Kalinka Koteva and Dr Mehdi Keramane for
assistance during chemical synthesis. We also thank Soumaya
Zlitni and Dr Sebastian Gehrke for helpful discussions in the
preparation of this manuscript. We are grateful for the guidance
provided by Dr Eean Breukink for the preparation of unde-
caprenyl phosphate. Support to EDB for this work included a
salary award (Canada Research Chair) and an operating grant
from the Canadian Institutes of Health Research (MOP-15496).
This work was also supported by National Science Foundation
Major Research Instrumentation Award 1039659 to TAG.
´
22 H. Barreteau, S. Magnet, M. El Ghachi, T. Touze, M. Arthur,
D. Mengin-Lecreulx and D. Blanot, J. Chromatogr. B, 2009,
877, 213–220.
23 Y. J. Lee, A. Ishiwata and Y. Ito, Tetrahedron, 2009, 65, 6310–
6319.
24 E. Breukink, H. E. van Heusden, P. J. Vollmerhaus,
E. Swiezewska, L. Brunner, S. Walker, A. J. R. Heck and
B. de Kruijff, J. Biol. Chem., 2003, 278, 19898–19903.
25 E. Swiezewska, W. Sasak, T. Mankowski, W. Jankowski,
T. Vogtman, I. Krajewska, J. Hertel, E. Skoczylas and
T. Chojnacki, Acta Biochim. Pol., 1994, 41, 221–260.
26 N. K. Khidyrova and K. M. Shakhidoyatov, Chem. Nat.
Compd., 2002, 38, 107–121.
Notes and references
1 A. L. Koch, Clin. Microbiol. Rev., 2003, 16, 673–687.
2 H. W. Boucher, G. H. Talbot, J. S. Bradley, J. E. Edwards,
D. Gilbert, L. B. Rice, M. Scheld, B. Spellberg and 27 L. L. Danilov, T. N. Druzhinina, N. A. Kalinchuk,
J. Bartlett, Clin. Infect. Dis., 2009, 48, 1–12.
3 M. M. Burger and L. Glaser, J. Biol. Chem., 1964, 239, 3168–
3177.
4 F. C. Neuhaus and J. Baddiley, Microbiol. Mol. Biol. Rev.,
2003, 67, 686–723.
5 L. Glaser and M. M. Burger, J. Biol. Chem., 1964, 239, 3187–
3191.
6 E. W. Sewell and E. D. Brown, J. Antibiot., 2014, 67, 43–51.
7 M. A. D'Elia, K. E. Millar, T. J. Beveridge and E. D. Brown, J.
Bacteriol., 2006, 188, 8313–8316.
S. D. Maltsev and V. N. Shibaev, Chem. Phys. Lipids, 1989,
51, 191–203.
28 M. M. Sim, H. Kondo and C. H. Wong, J. Am. Chem. Soc.,
1993, 115, 2260–2267.
29 A. Holkenbrink, D. C. Koester, J. Kaschel and D. B. Werz, Eur.
J. Org. Chem., 2011, 2011, 6233–6239.
30 R. Woodward, W. Yi, L. Li, G. Zhao, H. Eguchi, P. R. Sridhar,
H. Guo, J. K. Song, E. Motari, L. Cai, P. Kelleher, X. Liu,
W. Han, W. Zhang, Y. Ding, M. Li and P. G. Wang, Nat.
Chem. Biol., 2010, 6, 418–423.
8 C. Weidenmaier, J. F. Kokai-Kun, S. A. Kristian, 31 A. P. Bhavsar, R. Truant and E. D. Brown, J. Biol. Chem., 2005,
T. Chanturiya, H. Kalbacher, M. Gross, G. Nicholson, 280, 36691–36700.
B. Neumeister, J. J. Mond and A. Peschel, Nat. Med., 2004, 32 A. Zapun, J. Philippe, K. A. Abrahams, L. Signor, D. I. Roper,
10, 243–245.
E. Breukink and T. Vernet, ACS Chem. Biol., 2013, 8, 2688–
9 C. Weidenmaier, A. Peschel, Y.-Q. Xiong, S. A. Kristian,
2696.
¨
K. Dietz, M. R. Yeaman and A. S. Bayer, J. Infect. Dis., 2005, 33 C. Schaffer, T. Wugeditsch, P. Messner and C. Whiteld,
191, 1771–1777.
Appl. Environ. Microbiol., 2002, 68, 4722–4730.
10 M. L. Atilano, J. Yates, M. Glittenberg, S. R. Filipe and 34 J. W. Schertzer, J. Biol. Chem., 2003, 278, 18002–18007.
P. Ligoxygakis, PLoS Pathog., 2011, 7, e1002421.
35 N. Kojima, Y. Araki and E. Ito, J. Bacteriol., 1985, 161, 299–306.
11 S. Brown, G. Xia, L. G. Luhachack, J. Campbell, 36 J. W. Schertzer and E. D. Brown, J. Bacteriol., 2008, 190, 6940–
T. C. Meredith, C. Chen, V. Winstel, C. Gekeler, 6947.
J. E. Irazoqui, A. Peschel and S. Walker, Proc. Natl. Acad. 37 D. S. Badurina, M. Zolli-Juran and E. D. Brown, Biochim.
Sci. U. S. A., 2012, 109, 18909–18914. Biophys. Acta, 2003, 1646, 196–206.
12 M. A. Farha, A. Leung, E. W. Sewell, M. A. D'Elia, S. E. Allison, 38 B. Soldo, V. Lazarevic, H. M. Pooley and D. Karamata, J.
L. Ejim, P. M. Pereira, M. G. Pinho, G. D. Wright and
E. D. Brown, ACS Chem. Biol., 2013, 8, 226–233.
13 C. Ginsberg, Y.-H. Zhang, Y. Yuan and S. Walker, ACS Chem.
Biol., 2006, 1, 25–28.
Bacteriol., 2002, 184, 4316–4320.
39 L. V. Collins, S. A. Kristian, C. Weidenmaier, M. Faigle,
K. P. Van Kessel, J. A. Van Strijp, F. Gotz, B. Neumeister
and A. Peschel, J. Infect. Dis., 2002, 186, 214–219.
14 M. P. Pereira, J. W. Schertzer, M. A. D'Elia, K. P. Koteva, 40 M. L. Atilano, P. M. Pereira, J. Yates, P. Reed, H. Veiga,
D. W. Hughes, G. D. Wright and E. D. Brown,
ChemBioChem, 2008, 9, 1385–1390.
M. G. Pinho and S. R. Filipe, Proc. Natl. Acad. Sci. U. S. A.,
2010, 107, 18991–18996.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 3823–3830 | 3829

View Article Online
Chemical Science
Edge Article
41 M. A. Farha, K. Koteva, R. T. Gale, E. W. Sewell, G. D. Wright 47 Y. Kawai, J. Marles-Wright, R. M. Cleverley, R. Emmins,
and E. D. Brown, Bioorg. Med. Chem. Lett., 2014, 24, 905–910.
42 H. Wang, C. J. Gill, S. H. Lee, P. Mann, P. Zuck,
T. C. Meredith, N. Murgolo, X. She, S. Kales, L. Liang,
J. Liu, J. Wu, J. Santa Maria, J. Su, J. Pan, J. Hailey,
S. Ishikawa, M. Kuwano, N. Heinz, N. K. Bui,
C. N. Hoyland, N. Ogasawara, R. J. Lewis, W. Vollmer,
R. A. Daniel and J. Errington, EMBO J., 2011, 30, 4931–
4941.
D. Mcguinness, C. M. Tan, A. Flattery, S. Walker, T. Black 48 E. Bulat and T. A. Garrett, J. Biol. Chem., 2011, 286, 33819–
and T. Roemer, Chem. Biol., 2013, 20, 272–284. 33831.
43 S. Brown, Y. H. Zhang and S. Walker, Chem. Biol., 2008, 15, 49 B. Soldo, V. Lazarevic and D. Karamata, Microbiology, 2002,
12–21. 148, 2079–2087.
44 S. Brown, T. Meredith, J. Swoboda and S. Walker, Chem. 50 V. Lazarevic and D. Karamata, Mol. Microbiol., 1995, 16, 345–
Biol., 2010, 17, 1101–1110. 355.
45 A. P. Bhavsar, M. A. D'Elia, T. D. Sahakian and E. D. Brown, J. 51 Z. Guan, S. D. Breazeale and C. R. H. Raetz, Anal. Biochem.,
Bacteriol., 2007, 189, 6816–6823. 2005, 345, 336–339.
46 A. L. Lovering, L. Y. Lin, E. W. Sewell, T. Spreter, E. D. Brown 52 M. Pulfer and R. C. Murphy, Mass Spectrom. Rev., 2003, 22,
and N. C. Strynadka, Nat. Struct. Mol. Biol., 2010, 17, 582–589. 332–364.
3830 | Chem. Sci., 2014, 5, 3823–3830
This journal is © The Royal Society of Chemistry 2014