Staphylococcus aureus and bacillus subtilis W23 make polyribitol wall teichoic acids using different enzymatic pathways

Article abstract of DOI:10.1016/j.chembiol.2010.07.017

Wall teichoic acids (WTAs) are anionic polymers that play key roles in bacterial cell shape, cell division, envelope integrity, biofilm formation, and pathogenesis. B. subtilis W23 and S. aureus both make polyribitol-phosphate (RboP) WTAs and contain similar sets of biosynthetic genes. We use in vitro reconstitution combined with genetics to show that the pathways for WTA biosynthesis in B. subtilis W23 and S. aureus are different. S. aureus requires a glycerol-phosphate primase called TarF in order to make RboP-WTAs; B. subtilis W23 contains a TarF homolog, but this enzyme makes glycerol-phosphate polymers and is not involved in RboP-WTA synthesis. Instead, B. subtilis TarK functions in place of TarF to prime the WTA intermediate for chain extension by TarL. This work highlights the enzymatic diversity of the poorly characterized family of phosphotransferases involved in WTA biosynthesis in Gram-positive organisms.

Full text of DOI:10.1016/j.chembiol.2010.07.017

Chemistry & Biology
Article
Staphylococcus aureus and Bacillus subtilis W23
Make Polyribitol Wall Teichoic Acids Using
Different Enzymatic Pathways
Stephanie Brown, Timothy Meredith, Jonathan Swoboda,1,3 and Suzanne Walker1,*
1
1,2
1
Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, MA 02115, USA
2Present address: Merck Frosst Research Laboratories, Merck & Co., Montr e´ al, QC H2X 3Y8, Canada
3Present address: The Skaggs Institute of Chemical Biology and Department of Chemistry, The Scripps Research Institute, La Jolla,
CA 92037, USA
*Correspondence: suzanne_walker@hms.harvard.edu
DOI 10.1016/j.chembiol.2010.07.017
SUMMARY
dependent on the expression of WTAs in S. aureus (Weidenmaier
et al., 2004, 2005). A detailed understanding of WTA biosyn-
Wall teichoic acids (WTAs) are anionic polymers that thesis is important for exploring their roles in bacterial physiology
play key roles in bacterial cell shape, cell division, and assessing their potential as antibacterial targets (May et al.,
2
005; Swoboda et al., 2009).
WTAs are attached via a phosphodiester linkage to the
envelope integrity, biofilm formation, and pathogen-
esis. B. subtilis W23 and S. aureus both make polyri-
bitol-phosphate (RboP) WTAs and contain similar
sets of biosynthetic genes. We use in vitro reconsti-
tution combined with genetics to show that the path-
ways for WTA biosynthesis in B. subtilis W23 and
S. aureus are different. S. aureus requires a glyc-
erol-phosphate primase called TarF in order to
N-acetyl muramic acid sugars of peptidoglycan. WTAs typically
consist of a disaccharide linkage unit followed by a polymeric
main chain (Figure 1B). The B. subtilis W23 main chain is
structurally identical to the main chain in S. aureus and contains
ribitol-5-phosphate (RboP) repeats (Swoboda et al., 2010).
As shown in Figure 1C, a pathway for polyribitol phosphate
WTA synthesis was proposed many years ago by Lazarevic
make RboP-WTAs; B. subtilis W23 contains a TarF et al. (2002). This model was based on comparing the genes
homolog, but this enzyme makes glycerol-phos- for WTA biosynthesis in B. subtilis W23 to the genes in B. subtilis
phate polymers and is not involved in RboP-WTA 168, which makes polyglycerol-phosphate WTAs (Neuhaus and
Baddiley, 2003; Ward, 1981). Previous studies have confirmed
synthesis. Instead, B. subtilis TarK functions in place
of TarF to prime the WTA intermediate for chain
extension by TarL. This work highlights the enzy-
matic diversity of the poorly characterized family of
phosphotransferases involved in WTA biosynthesis
in Gram-positive organisms.
the proposed functions of the first three steps in the RboP-WTA
biosynthetic pathway. The first enzyme in this pathway, TagO, is
an integral membrane protein that transfers phospho-GlcNAc
from UDP-GlcNAc to an undecaprenyl carrier lipid embedded
in the cytoplasmic surface of the bacterial membrane (D’Elia
et al., 2006b; Weidenmaier et al., 2004). The lipid-linked
monosaccharide is then converted to disaccharide 4 by the
INTRODUCTION
UDP-ManNAc transferase TagA (Brown et al., 2008; D’Elia
et al., 2009; Zhang et al., 2006). A primase, TagB, then attaches
Teichoic acids are highly abundant anionic polymers found in a single glycerol-phosphate (GroP) unit to the nonreducing end
Gram-positive bacteria. There are two types of teichoic acids: of the disaccharide (Brown et al., 2008). Following assembly of
the lipoteichoic acids (LTAs), which are embedded in the bacte- the disaccharide linkage unit, the pathway for polyRboP-WTAs
rial membrane and extend into the peptidoglycan layers; and the was proposed to require three enzymes (TarF, TarK, and TarL)
wall teichoic acids (WTAs), which are covalently attached to the to complete the polymeric main chain (Lazarevic et al., 2002).
peptidoglycan layers and extend beyond them (Figure 1A) The proposed functions of these three enzymes are shown in
(
Neuhaus and Baddiley, 2003). Teichoic acids play important Figure 1C. Once WTA synthesis is complete, the RboP polymers,
but as yet poorly understood roles in cell shape determination still attached to the undecaprenyl carrier lipid, are flipped to the
D’Elia et al., 2006a; Pollack and Neuhaus, 1994; Soldo et al., external surface of the membrane where they are attached to
002), cell division (Grundling and Schneewind, 2007; Oku peptidoglycan (Swoboda et al., 2010).
et al., 2009; Schirner et al., 2009), biofilm formation (Fabretti Recent studies have shown that polyRboP-WTA polymer
(
2
et al., 2006; Fedtke et al., 2007; Vergara-Irigaray et al., 2008), synthesis in S. aureus differs from the proposed pathway in
cell adhesion (Gross et al., 2001; Weidenmaier et al., 2004), Figure 1C in that only two enzymes are required to complete the
and other aspects of Gram-positive physiology (Swoboda polyRboP main chain (Brown et al., 2008; Meredith et al., 2008;
et al., 2010; Xia et al., 2009). Although neither type of teichoic Pereiraetal., 2008a). OneenzymeisTarF (TarFSa), whichtransfers
acid is strictly essential for survival in vitro, organisms lacking a single GroP to the linkage unit. The other enzyme is TarL, which
either one display growth defects that range from mild to severe. combines the proposed functions of TarK and TarL shown in
Furthermore, it has been shown that pathogenicity is critically Figure 1C. That is, S. aureus TarL (TarLSa) is a ribitol-phosphate
Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved 1101

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Figure 1. Teichoic Acids Are
a Major
Component of the Gram-Positive Cell
Wall, and the Pathway for Ribitol-Phos-
phate WTAs Has Been Proposed
(
A) A schematic of the Gram-positive cell wall de-
picting membrane-anchored LTA and peptido-
glycan-anchored WTA polymers.
(
B) The chemical structure of ribitol-phosphate
WTA as found in S. aureus and B. subtilis W23
y = 1–2, z = 20–40). The tailoring modifications
on the main chain hydroxyls are omitted for clarity.
C) Lazarevic et al. (2002) proposed the biosyn-
(
(
thetic pathway depicted above for polyribitol-
phosphate WTAs. The numbers in parentheses
correspond to the substrates utilized in our
in vitro experiments (see Figure 2 for chemical
structures).
substrates for several WTA enzymes
in vitro (Brown et al., 2008; Ginsberg
et al., 2006; Zhang et al., 2006). Here,
we use the short chain substrate analogs
shown in Figure 3 to characterize the late
polymerase that can act directly on the TarF product without steps of polyribitol-phosphate WTA biosynthesis in B. subtilis
requiring a RboP-primed substrate. It functions to prime and W23. As described later, analysis of the products formed
extend the elaborated linkage unit (5) (Brown et al., 2008). in vitro by B. subtilis TarF (TarFBs), combined with the in vitro
Although we have shown that S. aureus contains two copies of substrate preferences for B. subtilis TarK (TarKBs), suggested
tarL (one of which was the originally proposed TarK), only one of a revised pathway for polyRboP-WTA biosynthesis in B. subtilis
these copies is essential in cells (Meredith et al., 2008; Swoboda W23. Using genetic approaches, we have confirmed the revised
et al., 2010). These discovered differences between the proposed B. subtilis W23 WTA biosynthetic pathway.
and demonstrated S. aureus WTA pathway prompted us to inves-
tigate the WTA biosynthetic pathway for B. subtilis W23, the Biochemical Reconstitution of B. subtilis W23 TarF
organismonwhichthepathway shown inFigure1C forpolyRboP- and Comparison to S. aureus TarF
WTA biosynthesis is based. We show here that the B. subtilis The first three steps in the pathways for WTA synthesis in
pathway differs in several ways from the S. aureus pathway, B. subtilis W23 and S. aureus are identical and lead to the
even though both organisms make a polyribitol-phosphate WTA production of lipid-pp-GlcNAc-ManNAc-GroP (compound 4)
polymer. Furthermore, it differs from the previously proposed B. (Figure 1C). We have shown that S. aureus TarF is a primase
subtilis pathway.
that adds a single glycerol phosphate unit to 4 to produce
lipid-pp-GlcNAc-ManNAc-GroP (compound 5) (Brown et al.,
2
RESULTS
2008). The function of TarF in B. subtilis W23 has not previously
been studied, but it was proposed, like its S. aureus ortholog, to
act as a GroP primase (Lazarevic et al., 2002), which catalyzes
Description of the Approach
As described in the Introduction, WTA precursors are synthe- the reaction of 4 with CDP-glycerol (1) to form WTA intermediate
sized in the cytoplasm on an undecaprenyl phosphate-carrier 5. To test this prediction, we cloned and expressed B. subtilis
lipid embedded in the membrane (Swoboda et al., 2010). W23 TarF (TarFBs) as a C-terminal 6-His fusion from genomic
2+
The first enzyme in the WTA biosynthetic pathway, TagO, is an DNA and purified it via Ni chromatography. We incubated
integral membrane protein, but many of the other enzymes purified TarFBs with radiolabeled 4b in the presence of
1
4
involved in the assembly of the WTA polymer do not have pre-
[ C]-CDP-glycerol (1) and analyzed the results by polyacryl-
dicted membrane-spanning regions, although it is presumed amide gel electrophoresis. PAGE analysis showed that
that they are membrane-associated either directly or indirectly CDP-glycerol (1) reacted in the presence of TarFBs and lipid-
(
Bhavsar et al., 2007; Formstone et al., 2008). Membrane- pp-GlcNAc-ManNAc-GroP (4b) to generate a new radioactive
associated enzymes in other lipid carrier-mediated biosynthetic spot (see Figure S1A available online); see below for product
pathways (e.g., peptidoglycan biosynthesis) will accept water- identification. TarFBs under the same conditions did not react
soluble substrates containing truncated lipid chains in the with 1 and 3b, or with CDP-ribitol (2) and 4b. These results sug-
absence of biological membranes or membrane mimics (Chen gested that TarFBs, like its S. aureus counterpart, catalyzes the
et al., 2002, 2007; Faridmoayer et al., 2008; May et al., 2009; addition of GroP, but not ribitol-phosphate, to primed substrate
Men et al., 1998; Ye et al., 2001). Therefore, we synthesized 4, but not to the unprimed precursor 3.
WTA pathway intermediates containing short lipid chains and
We next analyzed the products formed by the reaction of
have previously shown that these intermediates are functional radiolabeled 1 and 4b in the presence of TarFBs. For comparison
1102 Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Figure 2. Structures of WTA Substrates and Intermediates
The structures of the CDP-glycerol (1) and CDP-ribitol (2) donor and lipid acceptor substrates 3, 4, 5, 6, and 7 are shown. The natural acceptor substrates contain
an undecaprenyl chain. Substrates having the ‘‘a’’ suffix contain a neryl chain and were used for LC/MS analysis; substrates having the ‘‘b’’ suffix contain a farnesyl
1
4
chain and were used for PAGE analysis. Synthesis of substrates and incorporation of [ C] radiolabels for PAGE analysis were carried out as described previously
Brown et al., 2008).
(
the same substrates were also incubated with S. aureus TarF conditions (high enzyme and substrate concentrations for
TarFSa). Products were analyzed via PAGE (Figures 3B and prolonged reaction times), an additional product containing
C). TarFSa produces a single well-resolved product, identified three GroP units (compound 6) can also be observed (Fig-
by exact mass analysis of a nonradioactive reaction as lipid- ure S2A). In marked contrast, under all conditions examined
pp-GlcNAc-ManNAc-GroP (compound 5); this product corre- (low and high enzyme concentrations; short or long reaction
sponds to the transfer of a single GroP unit to 4. Under forcing times), the TarFBs reaction produces a smear of radioactivity
(
3
2
Figure 3. TarF Is a Polymerase in B. subtilis W23 and a Primase in S. aureus
(
(
A) Reaction carried out by TarF (n = 3–9 for TarFBs; n = 2 for TarFSa).
B and C) PAGE analysis of TarFBs (B) and TarFSa (C) reactions using 1 and substrate 4b in the presence of heat-treated (À) or active enzyme (+). Tabulated masses
of purified TarF reaction products using substrate 4a are shown.
Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved 1103

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
without a distinct banding pattern. These results suggested the appended to the ManNAc sugar. Reactions were analyzed for
formation of multiple products. Reactions with TarFBs and the appearance of polymeric product (Figure 4A). The TarKBs
/
substrates 1 and 4a were carried out using nonradiolabeled TarLBs tandem reaction converted substrate 4b to polymeric
substrates and subjected to Q-TOF MS analysis. Discrete prod- product. Tandem reactions using substrates containing more
ucts with increasing numbers of GroPs up to 9 units were than 1 GroP unit (e.g., 5b or 6b) resulted in little to no WTA
detected. Thus, whereas TarFSa is a primase that preferentially polymer.
adds a single GroP unit to 4, TarFBs behaves as a polymerase,
To confirm that the resulting polymeric products observed in
adding GroP units to 4 in rapid succession up to a detectable the TarKBs/TarLBs reactions were due to a tandem reaction of
length of at least 9 units.
TarKBs acting as a primase and then TarLBs acting as a poly-
merase, we tested the ability of B. subtilis W23 TarL to catalyze
the addition of ribitol-phosphate units to different WTA interme-
Reconstitution of B. subtilis W23 TarK and TarL
Lazarevic et al. (2002) proposed that TarK (TarKBs) and TarL diates. TarLBs was incubated with 4b, 5b, 6b, or 7b in the pres-
TarLBs) from B. subtilis W23 form a primase-polymerase pair, ence of CDP-ribitol (2). The reaction products were analyzed by
(
with the former transferring a single ribitol-phosphate unit to PAGE (Figure 4B; Figure S4A). No radioactive products were
the TarFBs product and the latter adding multiple ribitol-phos- observed for reactions containing lipid substrates 4b, 5b, or
phate units to build the polymer chain (Figure 1C). Using a 6b, showing that TarLBs is unable to use any of the GroP primed
heterologous complementation approach, we have previously intermediates to make polymer. In contrast the enzymatic reac-
provided evidence that B. subtilis W23 TarK and TarL function tion with compound 7b showed the disappearance of starting
as a primase-polymerase pair because both are required to material and the appearance of a smear of radioactivity corre-
complement the deletion of S. aureus TarL, which combines sponding to higher molecular weight products. The gel analysis
both functions (Meredith et al., 2008; Pereira et al., 2008a). shows that TarLBs reacts with the ribitol-phosphate-primed
However, the genetic complementation experiments did not substrate and is consistent with the addition of multiple RboP
provide detailed information on the enzymatic reactions carried subunits. These data confirm that TarLBs is a ribitol-phosphate
out by TarKBs and TarLBs. Here, we have reconstituted the polymerase and that it requires for reaction a ribitol-phos-
in vitro activities of these enzymes to characterize their substrate phate-primed substrate produced by TarKBs
.
preferences and verify their proposed products.
The TarKBs/TarLBs tandem reactions showed that TarKBs
TarKBs and TarLBs were cloned from genomic B. subtilis W23 has a strong preference for substrate 4, but it was not
DNA, expressed as C-terminal 6-His fusion proteins and purified clear whether the small amount of polymeric product ob-
by nickel chromatography. WTAs extracted from B. subtilis W23 served in the middle panel of Figure 4A resulted from the reac-
cells are reported to contain 1–2 GroP units, and because TarKBs tion of TarKBs with 5b or with contaminating 4b because the
is a proposed ribitol-phosphate phosphotransferase, we first starting material was estimated to be only 90% pure, and the
tested the activity of TarKBs with CDP-ribitol (2) and WTA impurity could not easily be removed. Therefore, we incubated
intermediate 4a, which contains 1 GroP unit. The reaction was TarKBs with CDP-ribitol (2) and a mixture of 4a and 5a and
analyzed by HPLC because products could not be resolved by analyzed the products by Q-TOF MS. The major peak proved
2
polyacrylamide gel electrophoresis (Figure S3A). A decrease in to be unreacted lipid-pp-GlcNAc-ManNAc-GroP (compound
-
the CDP-ribitol peak, accompanied by the appearance of 5a) (m/z [M-H] expected: 1027.2261, actual: 1027.2264).
-
CMP, showed that TarKBs reacts with CDP-ribitol in the Another peak corresponding to compound 7a (m/z [M-H]
presence of 4a. We also tested the activity of TarKBs with 4a expected: 1087.2472, actual: 1087.2468) was detected, result-
and CDP-glycerol (1). An unchanged peak area for CDP-glycerol ing from TarKBs utilizing 4a as a substrate. We were unable to
with no appearance of CMP showed that TarKBs does not utilize detect any product peak corresponding to the addition of
1
as a donor.
RboP to 5a. These results imply that TarKBs is only able to
Exact mass analysis of a TarKBs reaction of 2 with 4a showed accept lipid-pp-GlcNAc-ManNAc-GroP (4), the TagB product,
that the predominant product, compound 7a, contained 1 ribitol- as a substrate and that the small amount of polymeric
-
phosphate unit (m/z [M-H] expected: 1087.2472, actual: product observed in the middle panel of Figure 4A is due to
1
087.2509); a small amount of product (mass abundance less a reaction with the small amount of substrate 4 in the starting
than 0.05% compared with compound 7a) containing 2 ribitol- material.
-
phosphate units was also detected (m/z [M-H] expected:
301.2715, actual: 1301.2591). The MS data and HPLC traces TarKBs and TarLBs comprise a RboP primase-polymerase pair
The results described in the previous paragraph confirm that
1
confirm that TarK from B. subtilis W23 is a ribitol-phosphate and carry out the reactions shown in Figure 4C. These reactions
primase that catalyzes the addition of a ribitol-phosphate unit are not consistent with the model shown in Figure 1C (Lazarevic
from CDP-ribitol to WTA intermediate 4 to form lipid-pp- et al., 2002). In that model, TarFBs acts as a primase to synthe-
GlcNAc-ManNAc-GroP-RboP (7).
2
size a lipid-GlcNAc-ManNAc-GroP substrate for TarKBs, but
To evaluate the substrate specificity of TarKBs in more detail, our results show that TarKBs acts only on the TagB product,
we carried out tandem reactions of B. subtilis TarK and TarL. which contains a single GroP unit. Because TarKBs is unable to
Tandem reactions were used to enable product analysis by accept WTA intermediates containing more than a single GroP
PAGE because the TarK products cannot be resolved from start- unit, and because TarFBs is a polymerase that adds multiple
ing material. TarKBs and TarLBs were incubated with CDP-ribitol GroP units to 4, we surmised that TarFBs could not be on the
(
2) and radiolabeled lipid intermediates 4b, 5b, and 6b. These pathway to polyribitol WTA formation in B. subtilis W23. If this
intermediates differ in the number of glycerol phosphates supposition is correct, then one would predict that B. subtilis
1104 Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Figure 4. TarK and TarL React Sequentially on the TarB Product (4)
(
A) PAGE analysis of a tandem TarKBs/TarLBs reaction using either heat-treated (À) or active (+) enzymes incubated with CDP-ribitol and 4b, 5b, or 6b. The poly-
meric material in the middle panel is due to reaction with contaminating 4b. Polymer product only forms when 4b is used as the starting material (see text).
(
B) PAGE analysis of TarLBs heat-treated (À) or active (+) enzyme incubated with 4b or 7b and CDP-ribitol (2). TarLBs reacts only with 7b. See also Figure S4.
C) Reactions carried out by TarKBs and TarLBs with CDP-ribitol (2) and substrates 4 and 7, respectively.
(
tarF can be deleted, and polyribitol-phosphate WTAs will still be allele was almost exclusively observed upon resolving the inte-
produced in strains containing the B. subtilis TarK-TarL primase- grated tarF deletion vector, and the doubling time of the DtarF
polymerase pair.
hybrid strain was faster than that of the hybrid parent strain,
suggesting that the removal of TarF actually increases strain
fitness. We confirmed that WTAs extracted from the DtarF hybrid
strain are identical in length and banding pattern to those ex-
TarF Is Dispensable for Polyribitol-Phosphate WTA
Synthesis in Pathways Containing a TarK Primase
To test the predicted dispensability of B. subtilis TarF in strains tracted from the parent strain (Figure 5C). Thus, replacing
expressing B. subtilis TarK and TarL, we deleted tarF from two S. aureus TarL with TarKBs-TarLBs renders S. aureus TarF
different strains expressing these B. subtilis enzymes. The first dispensable for polyRboP-WTA synthesis in an S. aureus back-
strain was an S. aureus hybrid strain (JT215) containing B. subtilis ground.
TarK-TarL in place of the native bifunctional S. aureus enzymes
TarL or TarL’) (Meredith et al., 2008). This hybrid strain contains subtilis W23 strain containing only the native B. subtilis enzymes.
We next constructed a marked tarF deletion in a wild-type B.
(
the native S. aureus tarF, which is essential in the wild-type The DtarF strain (DM1) had a similar growth rate as the parent
S. aureus background (D’Elia et al., 2006b). Despite its essenti- strain, and the colony morphology was unchanged (Figures 5A
ality in the wild-type background, we were able to delete tarF and 5B). There were no apparent differences in the amounts of
from the hybrid S. aureus strain. In fact the DtarF deletion extracted WTAs in the DtarF and parent strains. Furthermore,
Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved 1105

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Figure 5. TarKBs and TarLBs Make polyRboP-WTAs
In Vivo in the Absence of TarFBs
(
A) Growth curves of B. subtilis W23 wild-type (WT) or DtarF
strains.
B) Photograph of bacterial colonies of B. subtilis W23 wild-
type (WT) or DtarF strains.
C and D) WTAs of the parent wild-type strain or the DtarF
(
(
strain extracted from the B. subtilis/S. aureus hybrid strain
JT215 (C) or B. subtilis W23 (D) analyzed by PAGE analysis
and Alcian blue/silver staining.
(
E) Agarose gel analysis of PCR-amplified fragments from B.
subtilis W23 cDNA or genomic DNA (positive control). (+)
and (À) indicate the presence or absence of reverse transcrip-
tase during cDNA synthesis. The sizes of the amplified frag-
ments are 170 bp (tarK), 159 bp (tarF), and 68 bp (16S rRNA)
(
internal control).
capable of making polyGroP-WTAs in a DtarK
background. All efforts to delete tarK were unsuc-
cessful, suggesting that this gene, unlike tarF, is
essential in B. subtilis W23. Therefore, we exam-
ined the relative levels of tarF and tarK gene
expression using RT-PCR. Under normal labora-
tory conditions, the tarF transcript is barely detect-
able in cells harvested during late exponential
growth, whereas the tarK transcript is abundant
(
Figure 5E). Increased amounts of template cDNA
in the PCR reaction did not change the levels of
detectable transcript. Although TarFBs is a func-
tional enzyme, it is not expressed under normal
WTAs extracted from the two strains were indistinguishable by laboratory growth conditions. Its cellular function, if any, remains
PAGE analysis (Figure 5D). Degradation of extracted WTAs a mystery, but it is not involved in polyRboP-WTA biosynthesis.
from each strain followed by Q-TOF MS analysis confirmed
that in both strains the subunits of the main chain polymer are DISCUSSION
composed of ribitol-phosphates covalently modified with
-
glucose (m/z [M-H] expected: 393.0804; actual wild-type: WTAs comprise a large fraction of the Gram-positive cell wall.
3
93.0813, DtarF: 393.0811) (Figure S5). These deletion experi- They play important roles in cell envelope integrity, and there
ments are fully consistent with the in vitro experiments showing is increasing evidence that they are involved in fundamental
that TarKBs reacts preferentially with the TagB product, and not aspects of cell growth and division (Xia et al., 2009). Under-
with intermediates containing additional GroP units. Thus, standing how these polymers are made is a necessary step in
genetic studies in two different strain backgrounds confirm the dissecting their cellular functions and in exploring their potential
biochemical work showing that TarFBs is not on the pathway to as antimicrobial targets in pathogens such as S. aureus.
polyRboP WTAs in B. subtilis W23.
Although pathways for polyglycerol-phosphate and polyribitol-
phosphate WTA biosynthesis in B. subtilis were proposed
many years ago (Lazarevic et al., 2002; Ward, 1981), verification
and detailed characterization of the enzymatic steps were not
B. subtilis tarF Is Not Expressed during
Exponential Growth
We have shown that tarFBs, which is located within the WTA possible until recently. The development of chemoenzymatic
biosynthetic gene cluster, encodes a functional enzyme that approaches to obtain functional WTA intermediates has made
makes short GroP-WTA polymers, but it is not involved in the in vitro reconstitution of both individual enzymes and entire
polyRboP WTA synthesis. Because it has been reported that pathways for WTA biosynthesis feasible (Brown et al., 2008;
strains in the W23 group (Nakamura et al., 1999), S. xylosus, Ginsberg et al., 2006; Sewell et al., 2009; Zhang et al., 2006).
S. saprophyticus (Endl et al., 1983, 1984), and an S. aureus In this paper we have carried out an extensive biochemical
biofilm-producing strain (Vinogradov et al., 2006) produce both and genetic analysis of polyRboP-WTA synthesis in B. subtilis
ribitol-phosphate and GroP WTAs, we considered the possibility W23. Although some steps of the proposed model for poly-
that B. subtilis W23 might contain a small proportion of poly- RboP-WTA biosynthesis were confirmed, we have revised
GroP-WTAs. We were unable to confirm the presence of GroP- others.
WTAs by MS analysis of WTAs extracted from B. subtilis W23
the detection limit is 5 pmol), so we attempted to generate dele- the late steps of the B. subtilis polyribitol-phosphate WTA
tions of tarK in B. subtilis W23 to determine whether TarFBs was pathway revealed two unexpected results. First, unlike S. aureus
Characterization of the enzymes proposed to be involved in
(
1106 Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Figure 6. Two Pathways for Polyribitol-
Phosphate WTA Formation in Bacteria
A schematic of the verified late-stage intracellular
steps in WTA biosynthesis in S. aureus and B. sub-
tilis W23.
ManNAc-phosphoglycerol unit (4) (Bhav-
sar et al., 2005; Ginsberg et al., 2006;
Soldo et al., 2002; Zhang et al., 2006).
Following the third step, catalyzed by
TagB, the B. subtilis W23 and S. aureus
pathways diverge. B. subtilis W23 uses
a RboP primase (TarK) followed by an
RboP polymerase (TarLBs) to complete
the WTA main chain; S. aureus uses a
GroP primase (TarF) followed by a bifunc-
tional RboP primase/polymerase (TarLSa
)
to complete the chain. The revised path-
ways, shown in Figure 6, are in line with
previously reported data by Yokoyama
TarF, B. subtilis TarF is not a primase. Rather than adding a single et al. (1986) on the isolation of WTA intermediates from B. subtilis
glycerol phosphate unit, it adds several units in rapid succession. W23 and S. aureus.
Therefore, it is a polymerase that makes short GroP WTA
In addition to establishing the pathways for polyribitol-phos-
polymers. Second, B. subtilis TarK, although it is indeed a phos- phate WTA polymer synthesis, we note that this work highlights
phoribitol primase, acts only on the TagB product. It cannot use the enzymatic diversity of the large family of CDP-phosphotrans-
substrates made by B. subtilis TarF, which adds multiple GroP ferases involved in WTA biosynthesis. These enzymes include
units. This finding implies that B. subtilis TarF is not on the both GroP and RboP primases and polymerases, and it does
pathway to polyRboP-WTAs. not yet appear possible to predict their functions or substrate
We confirmed this surprising conclusion by first deleting preferences accurately based on protein sequences (Figure S6).
S. aureus tarF from an S. aureus hybrid strain (Meredith et al., For example, S. aureus TarF is a GroP primase that adds 1 GroP
2
008) expressing the primase-polymerase pair from B. subtilis, subunit to the linkage unit, whereas B. subtilis 168 TagF is a GroP
TarKBs/TarLBs. Although S. aureus TarF is essential in a back- polymerase that adds 45–60 GroP units (Pereira et al., 2008b).
ground containing S. aureus TarL (D’Elia et al., 2006b), we B. subtilis W23 TarF makes short GroP polymers in vitro, adding
showed that it is dispensable in a strain containing the comple- about eight GroPs to the linkage unit. The RboP transferases are
menting B. subtilis enzymes. In fact its deletion in the hybrid similarly diverse, with the primase TarKBs having stringent spec-
strain appears to confer a modest fitness advantage. A plausible ificity for singly GroP-primed substrates and the sequence-
explanation for the growth advantage of the DtarF S. aureus related polymerase TarLBs having similarly stringent specificity
strain is that TarF activity impedes the maturation of WTAs for RboP-primed substrates, whereas S. aureus TarL functions
because the product it makes is not a preferred substrate for both as an RboP primase and polymerase and is selective for
the complementing B. subtilis enzymes. We were also able to doubly GroP-primed substrates. The first structure of a member
delete tarFBs from the B. subtilis W23 wild-type strain without of this diverse family of phosphotransferases was recently
any apparent effects on growth, colony morphology, or the reported (Staphylococcus epidermidis TagF) (Lovering et al.,
production of WTAs. In contrast all similar attempts to delete 2010), and it may, when combined with accurate information
tarKBs from B. subtilis W23 failed, suggesting that this gene about enzymatic function, facilitate the identification of the
cannot be deleted without affecting viability. This result is consis- features that determine which substrates are selected or
tent with other studies showing that enzymes required to whether a particular enzyme functions as a primase or a poly-
complete the WTA main chain are conditionally essential merase. In closing we note that there is a strong argument to
(
D’Elia et al., 2006a, 2006b; Swoboda et al., 2010). The ability be made for a new systematic nomenclature to designate the
to delete tarFBs, but not tarKBs, supports the hypothesis that CDP-phosphotransferases involved in WTA biosynthesis now
the latter is required for RboP-WTA synthesis, but the former is that the diversity of enzymatic function has been recognized,
not. RT-PCR analysis further supports this hypothesis because and the first ascribed gene names do not accurately describe
it shows that tarK is expressed at high levels during exponential these functions.
growth, whereas tarF transcription is almost undetectable.
The pathways for polyribitol-phosphate WTA biosynthesis in SIGNIFICANCE
B. subtilis W23 and S. aureus are now established, as shown in
Figure 6. These pathways proceed through the same three initial WTAs are important cell surface polymers in many Gram-
steps and result in the formation of a lipid-diphospho-GlcNAc- positive organisms, both pathogens and nonpathogens
Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved 1107

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
alike. They play central roles in cell growth, division,
morphology, adhesion, and envelope integrity. Under-
standing how they are made is important for conducting
detailed studies of their cellular functions (Swoboda et al.,
details, see Supplemental Experimental Procedures. WTA intermediates 3,
4
, and 5 were prepared as described previously (Brown et al., 2008).
To prepared radiolabeled compound 6b, an enzymatic reaction (30 ml) contain-
1
4
ing 2 mM TagB, 2 mM TarFSa, 20 mM of purified 3b, and 80 mM [ C]-CDP-glyc-
erol were incubated in 20 mM Tris (pH 7.5), 100 mM NaCl, and 10 mM MgCl
for 30 min at room temperature. This allows formation of intermediate 5b.
2
2
010). We have developed chemoenzymatic approaches to
1
4
obtain WTA intermediates to enable the detailed enzymatic
characterization of WTA biosynthesis. Here, we have used
an extensive set of purified WTA intermediates to compare
the pathways for polyribitol-phosphate WTA synthesis in
B. subtilis W23 and S. aureus. We show that these organisms
use different sets of enzymes to make similar WTA
polymers. Hence, bacteria employ two distinct pathways
to make polyribitol-phosphate WTAs as outlined in Figure 6.
This work should facilitate continued efforts to predict tei-
choic acid gene functions and determine pathway order in
sequenced genomes, will enable well-grounded biological
studies to establish the roles of WTA enzymes in the growth
and division of S. aureus and B. subtilis, and may facilitate
antibiotic discovery.
TarFBs (2 mM) was then added to utilize the remaining [ C]-CDP-glycerol to
form compound 6b. 6b was purified using a Phenomenex Strata X 33 mm poly-
meric reversed phase (30 mg/1 ml) column. To prepare radiolabeled
compound 7b, a 30 ml reaction containing 500 nM TagB, 500 nM TarFSa
,
1
4
5 mM TarKBs, 50 mM of purified 3b, 100 mM of [ C]-CDP-glycerol, and
00 mM of CDP-glycerol was incubated in 20 mM Tris (pH 7.5), 100 mM
1
2
NaCl, and 30 mM MgCl for 2 hr at room temperature. The reaction was puri-
fied as described for compound 6b. Cold reactions performed simultaneously
and analyzed by LC/MS confirmed the formation of compounds 6 and 7.
HPLC Analysis of TarK Reactions
For HPLC analysis a Phenomenex Phenosphere SAX 80 A, 250 3 4.6 mm 5 m
strong anion exchange column was used on an Agilent 1100 HPLC at flow rate
of 1 ml/min with a linear gradient of 0%–20% B over 15 min (buffer A: 5 mM
ammonium acetate [pH 3.8]; buffer B: 750 mM ammonium dihydrogen phos-
phate [pH 3.7]). The UV signal was monitored at 271 nM. TarKBs (30 ml) reac-
tions contained 200 mM 4a, 200 mM 1 or 2, and 1 mM enzyme in 20 mM Tris
EXPERIMENTAL PROCEDURES
2
(pH 7.5), 100 mM NaCl, and 10mM MgCl . After 30 min at room temperature,
the reactions were quenched with 30 ml methanol and analyzed by HPLC.
Materials and Methods
Reactions containing heat-treated enzymes served as negative controls.
B. subtilis W23 is from ATCC (ATCC 23059). Restriction enzymes were from
New England Biolabs. The plasmid used for cloning, pET-24b(+), and His-Bind
Resin are available from EMD Chemicals. PCR was performed using KOD Hot-
Start (EMD Chemicals). The pKM074a vector was a generous gift from David
PAGE Analysis of Enzymatic Reactions
To assay TarFBs substrate specificity, reactions (3 ml) contained 500 nM or
1
4
5 mM enzyme, 1 mM 4b, and 6 mM of either 2 or [ C]-1 in a buffer of 100 mM
14
Rudner (Harvard Medical School). Radiolabeled L-[ C]-glycerol-3-phosphate
was purchased from GE Healthcare. All other reagents and chemicals used
were from Sigma. The WTA extraction and analysis by polyacrylamide gel
electrophoresis and Alcian blue/silver staining have been described previously
NaCl, 20 mM Tris (pH 7.5), and 10 mM MgCl . A reaction was also set up
2
1
4
with 2 mM 3b and 2 mM [ C]-1. Heat-treated enzyme served as a negative
control. Reactions were quenched after 3 hr with 3 ml DMF. Two microliters
of 43 loading dye (80% glycerol, 0.02% bromophenol blue) was added, and
4 ml of the mixture was loaded on a 7 3 8 3 0.1 cm, 0.25 M TBE/20% acryl-
amide gel (preparation described previously [Brown et al., 2008]). The gel
was electrophoresed at 100 V for 110 min, dried, exposed to a phosphor
screen, and analyzed by radiometry.
(Meredith et al., 2008).
Cloning of B. subtilis W23 Genes
The S. aureus enzymes used in this study were cloned, overexpressed, and
purified as described previously (Brown et al., 2008). All primers and strains
are listed in Supplemental Information . The tarF, tarK, and tarL genes were
PCR amplified using B. subtilis W23 genomic DNA as a template and the
primers tarFp24bF and tarFp24bR, tarKp24bF and tarKp24bR, and tarLp24bF
and tarLp24bR. The PCR products were digested with the appropriate restric-
tion enzymes and ligated into a similarly digested pET-24b(+) vector. The con-
structed plasmids were verified by restriction digest and DNA sequencing
TarF reactions (3 ml) were performed using 500 nM enzyme (S. aureus or B.
1
4
subtilis W23) with 3 mM 4b, 18 mM [ C]-1, and 18 mM 1 in 100 mM NaCl, 20 mM
Tris (pH 7.5), and 10 mM MgCl . Reactions were quenched with 3 ml DMF after
2
30 min or 12 hr at room temperature. Reactions containing the same compo-
sition, but with 2 mM enzyme, were allowed to proceed for 12 hr at room
temperature. Identical reactions using heat-treated enzyme served as nega-
tive controls. Two microliters of 43 loading dye was added, and 8 ml of the
mixture was loaded on a 16 3 16 3 0.1 cm 0.25 M TBE/20% acrylamide
gel. After running for 260 min at 225 V, the gel was dried, exposed to phosphor
screen, and analyzed by radiometry.
(Dana-Farber/Harvard Cancer Center DNA Resource Core).
Protein Overexpression and Purification
S. aureus proteins used in this study were overexpressed and purified as
described previously (Brown et al., 2008). TagB from B. subtilis 168 was
used to construct various WTA intermediates and was overexpressed and
purified as described previously (Ginsberg et al., 2006). All constructed Tar
protein-pET-24b plasmids were transformed into Rosetta2(DE3)pLysS cells
for overexpression with IPTG. The cell pellets were lysed with rLysozyme
and benzonase in 100 mM Tris HCl (pH 7.5), 500 mM NaCl, 0.6% CHAPS,
and 5% glycerol. The clarified lysate was purified over charged nickel His-
Bind resin. More detailed procedures can be found in Supplemental
Experimental Procedures. The proteins were stored as 50% glycerol stocks
TarLBs reactions (3 ml) were performed using 5 mM enzyme with 2 mM 4b, 5b,
6b, or 7b and 200 mM 2. TarKBs reacting in tandem with TarLBs (3 ml reactions)
contained 1 mM of each enzyme with 2 mM 4b, 5b, or 6b and 200 mM 2. Reac-
2
tion buffer was 100 mM NaCl, 20 mM Tris (pH 7.5), and 30 mM MgCl . Each
substrate combination was also incubated with a heat-treated enzyme to
serve as a negative control. After 2 hr at room temperature, the reactions
were quenched with 3ml DMF and analyzed as described for TarFBs substrate
specificity reactions.
Q-TOF LC/MS Analysis
ꢀ
ꢀ
at either À20 C or À80 C. The yield is approximately 5 mg/l for TarFBs
,
All analyses were performed on an Agilent 6520 Q-TOF LC/MS using
a Phenomenex Gemini 5 micron C18 100 A column (50 3 4.6 mm) at a flow
rate of 1 ml/min (solvent A: water; solvent B: methanol both with 0.1% ammo-
nium hydroxide as a solvent modifier), with a linear gradient from 0% to 100%
B over 10 min. Detailed reaction conditions and purification procedures are
located in Supplemental Experimental Procedures.
1
0 mg/l for TarKBs, and 4 mg/l for TarLBs
.
Construction of WTA Intermediates
1
4
Radiolabeled [ C]-CDP-glycerol (1) was prepared as described previously
using TarD (Badurina et al., 2003). CDP-ribitol (2) was prepared as described
previously (Brown et al., 2008; Pereira and Brown, 2004). Due to issues with
degradation upon long-term storage, CDP-ribitol was purified by HPLC after
chemoenzymatic synthesis using a strong anion exchange column from
Phenomenex (Phenosphere SAX 80A, 250 3 4.6 mm 5 m). For purification
Construction of S. aureus tarF Deletion Strain
The tarF deletion cassette was obtained by assembly PCR of two amplicons
encoding approximately 1 kB DNA regions flanking the tarF open reading
1108 Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
frame (P1tarF-P2tarF and P3tarF-P4tarF), digested with ApaI, and ligated into
pKOR1 (digested with ApaI/EcoRV) to generate pKOR1-223. The deletion
vector was introduced into strain JT215 by electroporation (Schenk and
Laddaga, 1992), and the tarF deletion strain JT230 was subsequently isolated
by double crossover as described (Meredith et al., 2008).
funded by the NIH (1P01AI083214 and 5R01M078477 to S.W., and
F3178727 to J.S.).
Received: January 29, 2010
Revised: May 17, 2010
Accepted: July 16, 2010
Published: October 28, 2010
Construction of B. subtilis W23 tarF Deletion Strain
The marked tarF gene deletion in B. subtilis W23 was made using a protocol
developed for B. subtilis 168 gene deletion (Jarmer et al., 2002). Assembly
of the linear fragment, Z, containing the CAT cassette flanked by 1 kB DNA
flanking of B. subtilis W23 tarF is described in Supplemental Experimental
Procedures. The B. subtilis W23 strain to be transformed was streaked on
REFERENCES
Badurina, D.S., Zolli-Juran, M., and Brown, E.D. (2003). CTP:glycerol 3-phos-
phate cytidylyltransferase (TarD) from Staphylococcus aureus catalyzes the
cytidylyl transfer via an ordered Bi-Bi reaction mechanism with micromolar K
(m) values. Biochim. Biophys. Acta 1646, 196–206.
ꢀ
LB/agar and grown overnight at 37 C. A single colony was used to inoculate
ꢀ
a 5 ml culture of B. subtilis W23 and grown overnight at 37 C in modified
Spizizen minimal salt medium with glutamate (Jarmer et al., 2002). At OD450
Bernal, P., Zloh, M., and Taylor, P.W. (2009). Disruption of D-alanyl esterifica-
tion of Staphylococcus aureus cell wall teichoic acid by the {beta}-lactam
resistance modifier (-)-epicatechin gallate. J. Antimicrob. Chemother. 63,
ꢁ
1, 50 ml of cells was diluted into 5 ml of fresh minimal media and grown at
ꢀ
3
7 C. At OD450 ꢁ0.6, 200 ml of cells was incubated with at least 2 mg of linear
ꢀ
DNA Z. After 2 hr shaking at 37 C, the cells were plated on LB/agar containing
1
156–1162.
ꢀ
5
mg/ml chloramphenicol and incubated overnight at 37 C. A marked tarF gene
Bhavsar, A.P., Truant, R., and Brown, E.D. (2005). The TagB protein in Bacillus
subtilis 168 is an intracellular peripheral membrane protein that can incorpo-
rate glycerol phosphate onto a membrane-bound acceptor in vitro. J. Biol.
Chem. 280, 36691–36700.
deletion was confirmed by PCR analysis using primers that anneal outside the
deletion region.
WTA Extraction and Degradation
Bhavsar, A.P., D’Elia, M.A., Sahakian, T.D., and Brown, E.D. (2007). The Amino
terminus of Bacillus subtilis TagB possesses separable localization and func-
tional properties. J. Bacteriol. 189, 6816–6823.
WTAs were extracted and broken into monomeric units following a combina-
tion of previously published protocols (Bernal et al., 2009; Sadovskaya et al.,
2004; Tomita et al., 2009; Wickham et al., 2009). A detailed protocol can be
found in Supplemental Experimental Procedures. Briefly, the cell pellet was
disrupted by sonication. Noncovalently bound components were removed
from the cell wall by incubation with SDS, and covalently bound protein and
nucleic acids were removed by treatment with trypsin, DNase, and RNase.
The WTA was extracted by 10% trichloroacetic acid. Following centrifugation
to remove peptidoglycan, the WTAs were precipitated in 95% ethanol.
Brown, S., Zhang, Y.H., and Walker, S. (2008). A revised pathway proposed for
Staphylococcus aureus wall teichoic acid biosynthesis based on in vitro
reconstitution of the intracellular steps. Chem. Biol. 15, 12–21.
Chen, L., Men, H., Ha, S., Ye, X.Y., Brunner, L., Hu, Y., and Walker, S. (2002).
Intrinsic lipid preferences and kinetic mechanism of Escherichia coli MurG.
Biochemistry 41, 6824–6833.
To prepare monomeric units, the WTAs were treated with 1 N NaOH at
ꢀ
Chen, M.M., Weerapana, E., Ciepichal, E., Stupak, J., Reid, C.W.,
Swiezewska, E., and Imperiali, B. (2007). Polyisoprenol specificity in the
Campylobacter jejuni N-linked glycosylation pathway. Biochemistry 46,
1
00 C for 3 hr. The yield was between 5 and 9 mg from 500 ml cultures. The
monomeric units were analyzed by Q-TOF MS using direct inject with Solvent
A (described previously) at a flow rate of 0.2 ml/min.
1
4342–14348.
D’Elia, M.A., Millar, K.E., Beveridge, T.J., and Brown, E.D. (2006a). Wall
teichoic acid polymers are dispensable for cell viability in Bacillus subtilis.
J. Bacteriol. 188, 8313–8316.
RT-PCR
Three milliliters of B. subtilis W23 culture at an OD600 of ꢁ0.8 was harvested by
centrifugation, resuspended in RNALater reagent (QIAGEN), and stored
ꢀ
D’Elia, M.A., Pereira, M.P., Chung, Y.S., Zhao, W., Chau, A., Kenney, T.J.,
Sulavik, M.C., Black, T.A., and Brown, E.D. (2006b). Lesions in teichoic acid
biosynthesis in Staphylococcus aureus lead to a lethal gain of function in the
otherwise dispensable pathway. J. Bacteriol. 188, 4183–4189.
at À20 C until further use. RNA was extracted from the cell pellets using the
QIAGEN RNeasy Mini Kit with an initial treatment of 0.3 mg/ml lysostaphin
and 100 mg/ml Proteinase K. RQ1 DNase (Promega) was used to degrade
DNA. Four hundred nanograms of RNA was reverse transcribed using Super-
Script III Reverse Transcriptase (Invitrogen) using random hexamer primers.
RNA without reverse transcriptase treatment was used to detect potential
DNA contamination. Of this reaction, 0.3–5 ml was used as template for PCR
using KOD Hot Start (Novagen). Amplification of a fragment of the 16 s rRNA
transcript was used as an internal reference to ensure equal amounts of
cDNA template in each PCR reaction. RT-PCR was performed in both biolog-
ical and analytical duplicate. PCR of B. subtilis W23 genomic DNA was
performed to confirm that the primers were able to amplify the desired frag-
ment of cDNA. PCR products were analyzed by electrophoresis on an ethidium
bromide stained 2% agarose gel. The primers used are listed in Supplemental
Experimental Procedures.
D’Elia, M.A., Henderson, J.A., Beveridge, T.J., Heinrichs, D.E., and Brown,
E.D. (2009). The N-acetylmannosamine transferase catalyzes the first
committed step of teichoic acid assembly in Bacillus subtilis and Staphylo-
coccus aureus. J. Bacteriol. 191, 4030–4034.
Endl, J., Seidl, H.P., Fiedler, F., and Schleifer, K.H. (1983). Chemical composi-
tion and structure of cell wall teichoic acids of staphylococci. Arch. Microbiol.
1
35, 215–223.
Endl, J., Seidl, P.H., Fiedler, F., and Schleifer, K.H. (1984). Determination of cell
wall teichoic acid structure of staphylococci by rapid chemical and serological
screening methods. Arch. Microbiol. 137, 272–280.
Fabretti, F., Theilacker, C., Baldassarri, L., Kaczynski, Z., Kropec, A., Holst, O.,
and Huebner, J. (2006). Alanine esters of enterococcal lipoteichoic acid play
a role in biofilm formation and resistance to antimicrobial peptides. Infect.
Immun. 74, 4164–4171.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures,
six figures, and two tables and can be found with this article online at doi:10.
Faridmoayer, A., Fentabil, M.A., Haurat, M.F., Yi, W., Woodward, R., Wang,
P.G., and Feldman, M.F. (2008). Extreme substrate promiscuity of the Neisse-
ria oligosaccharyl transferase involved in protein O-glycosylation. J. Biol.
Chem. 283, 34596–34604.
1016/j.chembiol.2010.07.017.
ACKNOWLEDGMENTS
Fedtke, I., Mader, D., Kohler, T., Moll, H., Nicholson, G., Biswas, R., Henseler,
K., Gotz, F., Zahringer, U., and Peschel, A. (2007). A Staphylococcus aureus
ypfP mutant with strongly reduced lipoteichoic acid (LTA) content: LTA
governs bacterial surface properties and autolysin activity. Mol. Microbiol.
65, 1078–1091.
We would like to thank Yuriy Rebets and David Rudner for helpful discussions
on gene deletions; Emma Doud for assistance with mass spectrometry anal-
ysis; Jennifer Campbell for allowing us to modify her WTA schematic; and
Deborah Perlstein for a critical reading of this manuscript. This research was
Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved 1109

Chemistry & Biology
Pathway for B. subtilis W23 WTA Biosynthesis
Formstone, A., Carballido-Lopez, R., Noirot, P., Errington, J., and Scheffers,
D.J. (2008). Localization and interactions of teichoic acid synthetic enzymes
in Bacillus subtilis. J. Bacteriol. 190, 1812–1821.
Sadovskaya, I., Vinogradov, E., Li, J., and Jabbouri, S. (2004). Structural eluci-
dation of the extracellular and cell-wall teichoic acids of Staphylococcus epi-
dermidis RP62A, a reference biofilm-positive strain. Carbohydr. Res. 339,
1
467–1473.
Ginsberg, C., Zhang, Y.H., Yuan, Y., and Walker, S. (2006). In vitro reconstitu-
tion of two essential steps in wall teichoic acid biosynthesis. ACS Chem. Biol.
Schenk, S., and Laddaga, R.A. (1992). Improved method for electroporation of
1
, 25–28.
Staphylococcus aureus. FEMS Microbiol. Lett. 73, 133–138.
Gross, M., Cramton, S.E., Gotz, F., and Peschel, A. (2001). Key role of teichoic
acid net charge in Staphylococcus aureus colonization of artificial surfaces.
Infect. Immun. 69, 3423–3426.
Schirner, K., Marles-Wright, J., Lewis, R.J., and Errington, J. (2009). Distinct
and essential morphogenic functions for wall- and lipo-teichoic acids in
Bacillus subtilis. EMBO J. 28, 830–842.
Grundling, A., and Schneewind, O. (2007). Genes required for glycolipid
synthesis and lipoteichoic acid anchoring in Staphylococcus aureus. J. Bac-
teriol. 189, 2521–2530.
Sewell, E.W., Pereira, M.P., and Brown, E.D. (2009). The wall teichoic acid
polymerase TagF is non-processive in vitro and amenable to study using
steady state kinetic analysis. J. Biol. Chem. 284, 21132–21138.
Jarmer, H., Berka, R., Knudsen, S., and Saxild, H.H. (2002). Transcriptome
analysis documents induced competence of Bacillus subtilis during nitrogen
limiting conditions. FEMS Microbiol. Lett. 206, 197–200.
Soldo, B., Lazarevic, V., and Karamata, D. (2002). tagO is involved in the
synthesis of all anionic cell-wall polymers in Bacillus subtilis 168. Microbiology
148, 2079–2087.
Lazarevic, V., Abellan, F.X., Moller, S.B., Karamata, D., and Mauel, C. (2002).
Comparison of ribitol and glycerol teichoic acid genes in Bacillus subtilis
W23 and 168: identical function, similar divergent organization, but different
regulation. Microbiology 148, 815–824.
Swoboda, J.G., Meredith, T.C., Campbell, J., Brown, S., Suzuki, T.,
Bollenbach, T., Malhowski, A.J., Kishony, R., Gilmore, M.S., and Walker, S.
(2009). Discovery of a small molecule that blocks wall teichoic acid biosyn-
thesis in Staphylococcus aureus. ACS Chem. Biol. 4, 875–883.
Lovering, A.L., Lin, L.Y., Sewell, E.W., Spreter, T., Brown, E.D., and Strynadka,
N.C. (2010). Structure of the bacterial teichoic acid polymerase TagF provides
insights into membrane association and catalysis. Nat. Struct. Mol. Biol. 17,
Swoboda, J.G., Campbell, J., Meredith, T.C., and Walker, S. (2010). Wall
teichoic acid function, biosynthesis, and inhibition. Chembiochem 11, 35–45.
Tomita, S., Furihata, K., Nukada, T., Satoh, E., Uchimura, T., and Okada, S.
582–589.
(
2009). Structures of two monomeric units of teichoic acid prepared from the
May, J.J., Finking, R., Wiegeshoff, F., Weber, T.T., Bandur, N., Koert, U., and
Marahiel, M.A. (2005). Inhibition of the D-alanine:D-alanyl carrier protein ligase
from Bacillus subtilis increases the bacterium’s susceptibility to antibiotics that
target the cell wall. FEBS J. 272, 2993–3003.
cell wall of Lactobacillus plantarum NRIC 1068. Biosci. Biotechnol. Biochem.
73, 530–535.
Vergara-Irigaray, M., Maira-Litran, T., Merino, N., Pier, G.B., Penades, J.R.,
and Lasa, I. (2008). Wall teichoic acids are dispensable for anchoring the
PNAG exopolysaccharide to the Staphylococcus aureus cell surface. Microbi-
ology 154, 865–877.
May, J.F., Splain, R.A., Brotschi, C., and Kiessling, L.L. (2009). A tethering
mechanism for length control in a processive carbohydrate polymerization.
Proc. Natl. Acad. Sci. USA 106, 11851–11856.
Vinogradov, E., Sadovskaya, I., Li, J., and Jabbouri, S. (2006). Structural
elucidation of the extracellular and cell-wall teichoic acids of Staphylococcus
aureus MN8m, a biofilm forming strain. Carbohydr. Res. 341, 738–743.
Men, H., Park, P., Ge, M., and Walker, S. (1998). Substrate synthesis and
activity assay for MurG. J. Am. Chem. Soc. 120, 2484–2485.
Meredith, T.C., Swoboda, J.G., and Walker, S. (2008). Late-stage polyribitol
phosphate wall teichoic acid biosynthesis in Staphylococcus aureus. J. Bac-
teriol. 190, 3046–3056.
Ward, J.B. (1981). Teichoic and teichuronic acids: biosynthesis, assembly, and
location. Microbiol. Rev. 45, 211–243.
Weidenmaier, C., Kokai-Kun, J.F., Kristian, S.A., Chanturiya, T., Kalbacher, H.,
Gross, M., Nicholson, G., Neumeister, B., Mond, J.J., and Peschel, A. (2004).
Role of teichoic acids in Staphylococcus aureus nasal colonization, a major
risk factor in nosocomial infections. Nat. Med. 10, 243–245.
Nakamura, L.K., Roberts, M.S., and Cohan, F.M. (1999). Relationship of
Bacillus subtilis clades associated with strains 168 and W23: a proposal for
Bacillus subtilis subsp. subtilis subsp. nov. and Bacillus subtilis subsp. spizi-
zenii subsp. nov. Int. J. Syst. Bacteriol. 49, 1211–1215.
Weidenmaier, C., Peschel, A., Xiong, Y.Q., Kristian, S.A., Dietz, K., Yeaman,
M.R., and Bayer, A.S. (2005). Lack of wall teichoic acids in Staphylococcus
aureus leads to reduced interactions with endothelial cells and to attenuated
virulence in a rabbit model of endocarditis. J. Infect. Dis. 191, 1771–1777.
Neuhaus, F.C., and Baddiley, J. (2003). A continuum of anionic charge: struc-
tures and functions of D-alanyl-teichoic acids in gram-positive bacteria. Micro-
biol. Mol. Biol. Rev. 67, 686–723.
Oku, Y., Kurokawa, K., Matsuo, M., Yamada, S., Lee, B.L., and Sekimizu, K.
Wickham, J.R., Halye, J.L., Kashtanov, S., Khandogin, J., and Rice, C.V.
(2009). Revisiting magnesium chelation by teichoic acid with phosphorus
(2009). Pleiotropic roles of polyglycerolphosphate synthase of lipoteichoic
acid in growth of Staphylococcus aureus cells. J. Bacteriol. 191, 141–151.
solid-state NMR and theoretical calculations. J. Phys. Chem.
177–2183.
B 113,
2
Pereira, M.P., and Brown, E.D. (2004). Bifunctional catalysis by CDP-ribitol
synthase: convergent recruitment of reductase and cytidylyltransferase activ-
ities in Haemophilus influenzae and Staphylococcus aureus. Biochemistry 43,
Xia, G., Kohler, T., and Peschel, A. (2009). The wall teichoic acid and lipotei-
choic acid polymers of Staphylococcus aureus. Int. J. Med. Microbiol. 300,
148–154.
11802–11812.
Pereira, M.P., D’Elia, M.A., Troczynska, J., and Brown, E.D. (2008a). Duplica-
tion of teichoic acid biosynthetic genes in Staphylococcus aureus leads to
functionally redundant poly(ribitol phosphate) polymerases. J. Bacteriol. 190,
Ye, X.Y., Lo, M.C., Brunner, L., Walker, D., Kahne, D., and Walker, S. (2001).
Better substrates for bacterial transglycosylases. J. Am. Chem. Soc. 123,
3155–3156.
5
642–5649.
Yokoyama, K., Miyashita, T., Araki, Y., and Ito, E. (1986). Structure and func-
tions of linkage unit intermediates in the biosynthesis of ribitol teichoic acids
in Staphylococcus aureus H and Bacillus subtilis W23. Eur. J. Biochem. 161,
479–489.
Pereira, M.P., Schertzer, J.W., D’Elia, M.A., Koteva, K.P., Hughes, D.W.,
Wright, G.D., and Brown, E.D. (2008b). The wall teichoic acid polymerase
TagF efficiently synthesizes poly(glycerol phosphate) on the TagB product lipid
III. Chembiochem 9, 1385–1390.
Zhang, Y.H., Ginsberg, C., Yuan, Y., and Walker, S. (2006). Acceptor substrate
selectivity and kinetic mechanism of Bacillus subtilis TagA. Biochemistry 45,
10895–10904.
Pollack, J.H., and Neuhaus, F.C. (1994). Changes in wall teichoic acid during
the rod-sphere transition of Bacillus subtilis 168. J. Bacteriol. 176, 7252–7259.
1110 Chemistry & Biology 17, 1101–1110, October 29 , 2010 ª2010 Elsevier Ltd All rights reserved