Glycobiology vol. 20 no. 6 pp. 787–798, 2010
Advance Access publication on March 3, 2010
Protein tyrosine O-glycosylation—A rather unexplored
prokaryotic glycosylation system
Kristof Zarschler2,4, Bettina Janesch2, Martin Pabst3,
Friedrich Altmann3, Paul Messner1,2,
and Christina Schäffer1,2
2Department of NanoBiotechnology, and3Department of Chemistry, Vienna
Institute of BioTechnology, Universität für Bodenkultur Wien, A-1190 Vienna,
Received on January 18, 2010; revised on February 26, 2010; accepted on
February 26, 2010
Glycosylation is a frequent and heterogeneous posttrans-
lational protein modification occurring in all domains of
life. While protein N-glycosylation at asparagine and O-
glycosylation at serine, threonine or hydroxyproline resi-
dues have been studied in great detail, only few data are
available on O-glycosidic attachment of glycans to the
amino acid tyrosine. In this study, we describe the iden-
tification and characterization of a bacterial protein
tyrosine O-glycosylation system. In the Gram-positive,
mesophilic bacterium Paenibacillus alvei CCM 2051T, a
polysaccharide consisting of [→3)-β-D-Galp-(1[α-D-
Glcp-(1→6)] →4)-β-D-ManpNAc-(1→] repeating units
is O-glycosidically linked via an adaptor with the structure
Galp-(1→ to specific tyrosine residues of the S-layer protein
codes the information necessary for the biosynthesis of this
glycan chain within 18 open reading frames (ORF). The
corresponding translation products are involved in the bio-
synthesis of nucleotide-activated monosaccharides, assembly
andexportaswell as inthe transferofthe completedpolysac-
charide chain to the S-layer target protein. All ORFs of the
cluster, except those encoding the nucleotide sugar biosyn-
thesis enzymes and the ATP binding cassette (ABC)
transporter integral transmembrane proteins, were dis-
rupted by the insertion of the mobile group II intron Ll.
LtrB, and S-layer glycoproteins produced in mutant back-
grounds were analyzed by mass spectrometry. There is
evidence that the glycan chain is synthesized in a process
comparable to the ABC-transporter-dependent pathway
of the lipopolysaccharide O-polysaccharide biosynthesis.
Furthermore, with the protein WsfB, we have identified an
O-oligosaccharyl:protein transferase required for the forma-
tion of the covalent β-D-Gal→Tyr linkage between the
glycan chain and the S-layer protein.
Keywords: glycosylation gene cluster/Paenibacillus alvei/S-
Covalent attachment of glycans to the protein backbone via the
amide nitrogen of an asparagine residue (N-glycosylation) or
via the hydroxyl group of serine, threonine or hydroxyproline
(O-glycosylation) has been reported for many natural glyco-
proteins (Varki et al. 1999). In contrast, as a rare event, in
insect larvae (Chen et al. 1978; Kramer et al. 1980) as well
as in glycogenin of glycogen-containing eukaryotic cells
(Aon and Curtino 1985; Rodriguez and Whelan 1985), an O-
glycosidic linkage between a tyrosine residue and α-D-glucose
has been observed. In prokaryotes, O-glycosidic linkages of
glycans via β-D-galactose or β-D-glucose residues to tyrosine
βwere discovered as completely new types of linkage in the S-
layer glycoproteins of Paenibacillus alvei, Thermoanaerobac-
ter thermohydrosulfuricus and Thermoanaerobacterium
thermosaccharolyticum strains, respectively (Christian et al.
1988; Altman et al. 1991, 1995; Messner et al. 1992, 1995;
Bock et al. 1994; Schäffer et al. 2000). These S-layer glycopro-
teins share the common feature of S-layer proteins to self-
assemble into 2D crystalline arrays on the supporting cell en-
velope layer (Messner et al. 2009; Sleytr and Messner 2009),
covering the bacterium completely. The glycan chains protrude
from the cell surface, comparable to the lipopolysaccharide
(LPS) coating of Gram-negative bacteria (Messner et al.
2008). For several S-layer glycoprotein-carrying bacteria,
polycistronic S-layer glycosylation (slg) gene clusters with a
size of ∼16 to ∼25 kb have been identified and sequenced
(Novotny, Pföstl, et al. 2004). These gene clusters include nu-
cleotide sugar pathway genes that are arranged consecutively,
glycosyltransferase genes, glycan processing genes and trans-
porter genes, all of them exhibiting high homology to
components involved in the biosynthesis of different bacterial
surface polysaccharides (Novotny, Schäffer, et al. 2004).
In the mesophilic, Gram-positive bacterium P. alvei CCM
2051T, the S-layer O-glycan is a polymeric branched polysac-
charide of, on average, 23 [→3)-β-D-Galp-(1[α-D-Glcp-
(1→6)]→4)-β-D-ManpNAc-(1→] repeating units linked via
an adaptor with the structure -[GroA-2→OPO2→4-β-D-Man-
© The Author 2010. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: email@example.com 787
1To whom correspondence should be addressed: Tel: +43-47654 ext. 2202;
Fax: +43-1-4789112; e-mail: firstname.lastname@example.org, and Tel: +43-1-
47654 ext. 2203; Fax: +43-1-4789112; e-mail: email@example.com
4Present address: Institute of Genetics, General Genetics,Technische
Universität Dresden, Zellescher Weg 20b, D-01217 Dresden, Germany
Rhap-(1→3)-β-D-Galp-(1→ to specific tyrosine residues of the
S-layer protein backbone (Altman et al. 1991; Messner et al.
1995) (Figure 1). Two predominant glycosylation sites are pre-
dicted on the S-layer protein, most probably corresponding to
mono- and di-glycosylated S-layer protein as derived from the
periodic acid-Schiff staining behavior of the S-layer glycopro-
tein after separation on a sodium dodecyl sulfate
polyacrylamide (SDS-PA) gel (Zarschler et al. 2009). Recently,
the initiation enzyme of S-layer glycan biosynthesis, WsfP, has
been identified as part of a slg gene cluster of P. alvei CCM
2051T, and a gene disruption system has been developed for
this organism (Zarschler et al. 2009).
In the present study, we report on the determination of the
nucleotide sequence of the complete slg gene cluster of P. alvei
CCM 2051Tand the genetic assignment of these open reading
frames (ORFs). Insertional mutants for nine of the 18 ORFs
located in the slg gene cluster were created, which allows pro-
posing a specific transport mechanism for the glycan across the
cytoplasmic membrane and identifying the O-oligosaccharyl:
protein transferase (O-OTase) catalyzing the transfer of the gly-
can chain to specific tyrosine residues of the S-layer target
protein SpaA. Furthermore, we introduce a working model
for the S-layer glycan biosynthesis route in P. alvei CCM
2051Tbased on the observed similarity between the proteins
encoded by ORFs of the slg gene cluster and database entries,
as well as on the effects of the disruption of selected ORFs of
the slg gene cluster on S-layer glycan composition as deter-
mined by mass spectrometry.
Identification of the P. alvei CCM 2051T slg gene cluster
Since L-rhamnose represents the main constituent of the
adaptor region of the S-layer glycan of P. alvei CCM
2051T, we have chosen the genes responsible for dTDP-L-
rhamnose biosynthesis as suitable candidates for the design
of degenerate primers to obtain entry into a putative slg gene
cluster. Using a genomic DNA bank of P. alvei CCM 2051T
as a template for the PCR amplification reaction and the
Fig. 1. Schematic drawing of the S-layer glycoprotein glycan structure of P. alvei CCM 2051T.
Fig. 2. Genetic organization of the slg gene cluster of P. alvei CCM 2051T. Predicted open reading frames are indicated by horizontal arrows with the respective
gene designations indicated above the arrow. ORFs encoding similar functions in S-layer glycan biosynthesis have a similar gray shading code. ORFs flanking the
slg gene cluster are indicated in black, and ORFs coding for proteins with unknown function are indicated in white. ORFs indicated in light gray encode putative
glycosyltransferases. Wzm and wzt (dark gray) encode the two components of the putative ABC transporter. Squares indicate ORFs encoding proteins involved in
the biosynthesis of nucleotide-activated sugar precursors. The lightest gray indicates the wsfB ORF encoding the putative O-OTase. Putative promoters and
terminators are represented as flags and hairpins, respectively. The location of the Ll.LtrB insertion is indicated by vertical black arrows and by the numbers below,
while positions are given relative to the initial ATG codon. Experimentally identified transcription units are depicted (1). The reverse transcription analysis and
subsequent cDNA amplification are shown, with primer positions indicated as vertical black arrows (2). The nucleotide sequence of the slg gene cluster has been
submitted to the GenBank/EBI Data bank with accession number HM011508.
K Zarschler et al.
primer proof_RmlD_for, a ∼1.4-kb DNA fragment was ob-
tained. After confirmation of the presence of the rmlD gene
coding for a dTDP-4-dehydrorhamnose reductase, sequencing
of upstream and downstream regions by chromosome walk-
ing revealed the presence of 18 ORFs contained in a 24.3-kb slg
gene cluster, encoding components of the putative S-layer pro-
Table I. Predicted gene products encoded by the slg gene cluster of P. alvei CCM 2051Ttogether with database homologies.
Mol. mass Conserved motifs and region
Name/putative function Organism
similarity (%) Accession No.
wsfB 781/87.5 Wzy_C (PF04932)
690-723 O-antigen polymerase
O-antigen polymerase Geobacillus sp. Y412MC10
Paenibacillus larvae subsp. larvae
Geobacillus sp. Y412MC10
Geobacillus sp. Y412MC10
galE 328/36.3 Epimerase (PF01370)3-251 UDP-glucose 4-epimerase
galU 290/33.0 NTP_transferase (PF00483)5-216 UTP-glucose-1-phosphate
Bacillus weihenstephanensis KBAB4 69/82YP_001647518
wzm 232/27.4 ABC2_membrane (PF01061) 1-191Aneurinibacillus thermoaerophilus
Vibrio sp. MED222
Francisella philomiragia subsp.
philomiragia ATCC 25017
Thioalkalivibrio sp. HL-EbGR7
wzt434/48.7 ABC_tran (PF00005)53-22242/62
wsfA 520/60.5 GATase_2 (PF00310)
224-475 Asparagine synthase
5-131 Putative glycerol-3-phosphate
99-279Glycosyl transferase family 2 Anaeromyxobacter sp. Fw109-5
15-264 Transmembrane protein
tagD 139/16.3 CTP_transf_2 (PF01467)Francisella philomiragia subsp.
philomiragia ATCC 25017
Clostridium perfringens str. 13
wsfC 1260/147.3 Glyphos_transf (PF04464)
wsfD 457/53.3 PMT (PF02366)
YP_246702 Rickettsia felis URRWXCal2
Bacillus cereus G9241
Paenibacillus larvae subsp. larvae
Clostridium acetobutylicum ATCC 824 29/48
Pseudomonas fluorescens Pf-5
wsfE 364/42.3Glycos_transf_1 (PF00534)187-342 Glycosyltransferase
group 2 family protein
rmlA 247/54.6NTP_transferase (PF00483) Geobacillus stearothermophilus
Bacillus anthracis str. Ames
rmlC 183/20.8dTDP_sugar_isom (PF00908) 3-177 Geobacillus sp. Y412MC10 70/85ZP_03037628
Geobacillus sp. Y412MC10
rmlB 341/38.5Epimerase (PF01370)76/88 ZP_03037627
Geobacillus tepidamans GS5-97T
Oenococcus oeni ATCC BAA-1163
Geobacillus sp. Y412MC10
rmlD 286/32.2 RmlD_sub_bind (PF04321)60/78 AAR99613
wsfF 314/36.8Glycos_transf_2 (PF00535)
ZP_03037625wsfG 299/34.1Glycos_transf_2 (PF00535) 5-191
Geobacillus tepidamans GS5-97T
Microcystis aeruginosa NIES-843
Arthrospira maxima CS-328
wsfP 468/54.6Bac_transf (PF02397)281-468 WsaP 60/75AAR99615
wsfH 336/38.4 Glycos_transf_2 (PF00535)12-176
Tyrosine O-glycosylation in Paenibacillus alvei
tein glycosylation machinery (Figure 2). Most of the putative
gene products encoded by the assigned ORFs showed high ho-
mology to proteins involved in the biosynthesis of bacterial
surface polysaccharides. Based on these sequence similarities,
putative biological functions have been assigned to almost all of
the ORFs of the slg gene cluster (Table I).
Genetic characterization of the slg gene cluster
The slg gene cluster of P. alvei CCM 2051Tis flanked by
ORFs coding for enzymes involved in lantibiotic biosynthesis
(Bierbaum and Sahl 2009) and lipid/lipoteichoic acid biosyn-
thesis (Chen et al. 1993), respectively. The closed spacings
from ORF wzm to wsfH and their identical transcriptional di-
rection indicate that these ORFs are transcribed as a single
operon (Figure 2). To identify specific mRNA(s) of the gene
cluster, total RNA of P. alvei CCM 2051Twas isolated and re-
verse transcribed into cDNA. The amplification of cDNA
using primer combinations spanning the regions wsfB/galE,
galU/wzm, wzt/wsfA, wsfA, wsfC/wsfD, wsfD/wsfE, rmlB/wsfF
and wsfH/pcrB revealed that the slg gene cluster is transcribed
as a polycistronic unit starting with galU and ending with wsfH
(Figure 3). No PCR products were obtained when primers an-
nealing to wsfB/galE (1f/1r) and cDNA reverse transcribed
with primer 1f or when primers annealing to wsfH/pcrB (9f/
9r) and cDNA reverse transcribed with primer 1f were used
(data not shown). This observation indicates that wsfB and
galE are transcribed separately, implying that they are not part
of the polycistronic slg gene cluster but of the S-layer glyco-
sylation locus. PcrB is most possibly not involved in S-layer
The putative gene products of the slg gene cluster have been
analyzed by extensive database comparison and are discussed
in the order of their appearance within the analyzed region.
wsfB. The gene product of wsfB (Figure S1) contains 12
potential transmembrane domains and a conserved Wzy_C
motif characteristic of O-antigen ligases responsible for the
transfer of undecaprenyl-pyrophosphate-linked sugars to a
target protein (Power et al. 2006). OTases, such as PglL
from Neisseria meningitidis and PilO from Pseudomonas
aeruginosa, show only low levels of amino acid similarity
but possess similar transmembrane topology and short regions
of high homology (Abeyrathne et al. 2005). Both OTases
exhibit relaxed glycan specificity but require the translocation
of the corresponding undecaprenyl-pyrophosphate-linked
oligosaccharide substrates into the periplasm (Faridmoayer et
al. 2007). Like PglL, WsfB possesses in its carboxy-terminal
part a tetratricopeptide repeat (TPR) described as mediator for
protein–protein interactions and the assembly of multiprotein
complexes (D'Andrea and Regan 2003). A similar membrane
spanning topology and the presence of the Wzy_C motif in
WsfB supports the assumption that this enzyme belongs to
the family of O-OTases transferring the S-layer glycan
chain from the lipid carrier to certain tyrosine residues in
galE and galU. The translation products of these two ORFs are
homologous to the uridine diphosphate (UDP)-glucose 4-
epimerase (GalE) and the uridine triphosphate (UTP)-
glucose-1-phosphate uridylyltransferase (GalU), respectively.
GalE catalyzes the interconversion between UDP-glucose
and UDP-galactose, and GalU mediates the transfer of UTP
to glucose-1-phosphate resulting in UDP-glucose. The
presence of galactose and glucose in the S-layer O-glycan of
P. alvei CCM 2051Tsuggests that GalE and GalU are
involved in the biosynthesis of the sugar precursors UDP-
galactose and UDP-glucose.
wzm and wzt. The deduced 232- and 434-amino acid proteins
encoded by wzm and wzt, respectively, reveal high similarity to
proteins of the ABC-2 transporter family, involved in the trans-
port of bacterial surface polysaccharides to the cell surface
(Bronner et al. 1994). The presence of six transmembrane do-
mains in the putative translation product of wzm suggests that
this protein is the integral membrane component of the trans-
porter. An ATP-binding site and an ATP transporter signature
motif identified in the putative translation product of wzt indi-
cate the involvement of this protein in the transport of sugars
across the cytoplasmic membrane (Walker et al. 1982; Roc-
chetta and Lam 1997). Its extended carboxy-terminal part
obviously contains an O-polysaccharide binding domain de-
termining the transporter’s substrate specificity as observed
Fig. 3. RT-PCR analysis of total RNA of P. alvei CCM 2051T. Reverse transcription was performed with the specific primer 4r targeted to wsfA (lanes 2–7) or 8r
annealing to wsfF (lanes 9–20). Subsequent cDNA amplification was carried out with primer pairs 2f/2r annealing to galU/wzm (lanes 2–4), with 3f/3r targeted to
wzt/wsfA (lanes 5–7), with 4f/4r annealing to wsfA (lanes 9–11), with 5f/5r targeted to wsfC/wsfD (lanes 12–14), with 6f/6r annealing to wsfD/wsfE (lanes 15–17)
and with 8f/8r amplifying rmlB/wsfF (lanes 18–20). Lanes (a) show the specific PCR amplification products, using reverse transcribed single-strand cDNA as
template; lanes (b) show control reactions, using DNase I-treated RNAwithout the cDNA-generating step as PCR template; lanes (c) show positive controls using
genomic DNA as template. The 1-kb DNA Plus marker (Invitrogen) was used as DNA size marker (lanes 1, 8 and 20).
K Zarschler et al.
for the polymannan O-antigenic polysaccharides of Escheri-
chia coli O8 and O9a. Furthermore, several amino acids
identified in wzt of E. coli O9a to be critical for binding
and export of O-antigenic polysaccharide were also found
in the homologous protein of P. alvei CCM 2051T(Cuthbert-
son et al. 2005, 2007). For instance, G333 and G387 located
in the carbohydrate-binding pocket of the E. coli protein cor-
respond to G334 and G387 of Wzt of P. alvei CCM 2051T
wsfA. Throughout the whole ORF, wsfA is homologous to
genes coding for asparagine synthetase B (AsnB). This enzyme
acts as a homodimer with each monomer being composed of a
glutaminase domain, hydrolyzing glutamine to glutamic acid
and a combined ammonia and asparagine synthetase domain,
catalyzing the ATP-dependent conversion of aspartate to
asparagine (Milman and Cooney 1979; Scofield et al. 1990).
An asnB mutant of Corynebacterium glutamicum was
isolated as a lysozyme- and temperature-sensitive mutant
(Hirasawa et al. 2000), and in Mycobacterium smegmatis,
AsnB is involved in the natural resistance to rifampin,
erythromycin and novobiocin (Ren and Liu 2006). Up to
date, no specific function could be assigned to the wsfA gene
product of P. alvei CCM 2051T.
tagD. The stop codon of wsfA overlaps with the putative start
codon of tagD. Comparison of the translation product of tagD
with proteins in the database showed a high degree of amino
acid homology to the glycerol-3-phosphate cytidyltransferase
involved in the formation of cytidine diphosphate (CDP)-
glycerol and pyrophosphate from cytidine triphosphate (CTP)
and glycerol-3-phosphate (Mauel et al. 1991; Park et al. 1993).
The presence of a glyceric acid phosphate residue in the
adapter saccharide of the S-layer O-glycan of P. alvei CCM
2051Tsuggests that TagD catalyzes the synthesis of the
building block CDP-glycerol.
gene cluster, coding for a tripartite transferase of 147.34 kDa.
Two glycosyltransferase family 2 motifs were predicted in the
central and carboxy-terminal part, while a single CDP-
motif was found at the amino-terminal part of the protein. The
amino-terminal region shows similarity to the TagB protein
of Bacillus subtilis, catalyzing, there, the incorporation of
a single glycerol phosphate residue from CDP-glycerol to
the nonreducing end of membrane-bound undecaprenyl-
acetylglucosamine-1-phosphate (Bhavsar et al. 2005). The
central part of WsfC is homologous to the glycosyltransferase
LgtD of different Rickettsia strains. In Haemophilus influenzae
and Neisseria gonorrhoeae, LgtD is involved in LPS and
lipooligosaccharide biosynthesis, respectively, possessing acet-
ylgalactosaminyltransferase and galactosyltransferase activity
(Gotschlich 1994; Shao et al. 2002; Randriantsoa et al.
2007). Sequence homology searches for the predicted amino
acid sequences of the carboxy-terminal part of WsfC showed
homology with cyanobacterial and archaeal glycosyltrans-
ferases, with AglG being involved in the N-glycosylation of
the Haloferax volcanii S-layer glycoprotein and showing hex-
uronic acid transferase activity (Yurist-Doutsch et al. 2008).
wsfD. The wsfD gene product contains nine potential trans-
membrane domains and is similar to uncharacterized trans-
membrane proteins of various Gram-positive bacteria. In the
amino-terminal part, a dolichyl-phosphate-mannose-protein
mannosyltransferase domain spanning seven transmembrane
domains was predicted. In fungi, dolichyl-phosphate-
mannose-protein mannosyltransferases (PMTs) are integral
endoplasmic reticulum (ER) membrane proteins responsible
for the initiation of protein O-mannosylation. The PMT of
Saccharomyces cerevisiae, ScPmt1p, is an integral membrane
glycoprotein of 817 amino acids, located in the ER and
catalyzing the transfer of mannose from the lipid carrier Dol-
P-β-D-mannose to serine/threonine residues of specific protein
acceptors (Strahl-Bolsinger and Tanner 1991; Strahl-Bolsinger
et al. 1993; Gentzsch et al. 1995). As ScPmt1p, WsfD possesses
an amino-terminal loop, a large hydrophilic loop and a carboxy-
terminal region, all facing the ER or the external face of the
cytoplasmic membrane, respectively (Strahl-Bolsinger and
Scheinost 1999; Girrbach et al. 2000). We speculate that,
although no mannose residue was found in the S-layer O-
glycan of P. alvei CCM 2051T, WsfD could be involved in the
transfer of a hexose residue from a lipid carrier to the glycan
wsfE. The 364-amino acid translation product of wsfE shows
high similarity to several glycosyltransferases of various
Clostridium, Pseudomonas and Vibrio strains (Nölling et al.
2001; Chen et al. 2003; Gross et al. 2007), and it contains a
potential glycosyltransferase group 1 motif (GT1_wbuB_like).
In E. coli, WbuB is involved in the biosynthesis of the O26 O-
antigen, thereby acting as an N-acetyl-L-fucosamine (L-
FucNAc) transferase (D'Souza et al. 2002).
rmlA, rmlC, rmlB, and rmlD. The rmlACBD gene products
show a high degree of amino acid homology to the RmlACBD
proteins involved in the biosynthesis of dTDP-L-rhamnose in
different Geobacillus strains (Novotny, Schäffer et al. 2004;
Zayni et al. 2007) and in other bacteria (Graninger et al.
2002). Since the adaptor oligosaccharide of the S-layer O-
glycan of P. alvei CCM 2051Tcontains three L-rhamnose
residues, it is conceivable that RmlACBD are providing the
nucleotide-activated building block dTDP-L-rhamnose. The
putative start codon of rmlD overlaps with the stop codon of
wsfF and wsfG. The stop codon of wsfF overlaps with the
putative start codon of wsfG. The protein products of these two
ORFs possess a potential glycosyltransferase family 2 motif in
their amino-terminal parts. The deduced protein sequence of
WsfF shows significant homology to the putative sugar
transferase WsdG of Aneurinibacillus thermoaerophilus
DSM 10155/G+and to different rhamnosyltransferases
(Novotny, Pföstl et al. 2004). The protein encoded by wsfG
is highly similar to the β 1,2-rhamnosyltransferase WsaF of
Geobacillus stearothermophilus NRS 2004/3a, transferring
an L-rhamnose residue to the linkage sugar galactose during
S-layer glycan biosynthesis in this strain (Steiner et al. 2008,
2010). For WsfG, a single transmembrane-spanning domain
was predicted at the carboxy-terminal part of the protein.
Tyrosine O-glycosylation in Paenibacillus alvei
wsfP. The UDP-Gal:phosphoryl-polyprenol Gal-1-phosphate
transferase WsfP has been recently identified as the initiation
enzyme of S-layer glycan biosynthesis in P. alvei CCM 2051T
(Zarschler et al. 2009).
wsfH. The predicted protein WsfH shows high similarity to
several glycosyltransferases found in different cyanobacteria
(Kaneko et al. 2007; Welsh et al. 2008). A cytoplasmic
glycosyltransferase family 2 motif and a single transmembrane-
spanning domain were predicted at the amino- and carboxy-
terminal part of the protein, respectively. Since WsfH also
shares amino acid similarity with several polyprenyl-
phosphate β-D-glucosyltransferases, it is likely that it is
responsible for the intracellular transfer of a glucose residue
to the membrane-associated lipid carrier undecaprenyl-
pcrP. The region downstream of wsfH contains a putative
ρ-independent bacterial terminator followed by a putative
promoter site allowing the transcription of an ORF coding
for PcrB, an enzyme involved in lipid/lipoteichoic acid
biosynthesis (Chen et al. 1993). This protein is not part of
the slg gene cluster.
S-layer protein glycosylation in slg gene mutant backgrounds
To investigate the role of individual ORFs of the P. alvei CCM
2051Tslg gene cluster, insertional knockout mutants in nine
ORFs were constructed. All mutants still produced the S-layer
protein SpaA as detected by sodium dodecyl sulfate polyacryl-
amide gel electrophoresis (SDS-PAGE) followed by
Coomassie staining (Figure 4). However, for SpaA produced
in mutants with an insertion of Ll.LtrB in the wsfA, wsfB, wsfC,
wsfE, wsfF, wsfG or wzt ORF, only a single protein band
corresponding to the non-glycosylated S-layer protein could
be detected on SDS-PA gels. The absence of glycosylation
was corroborated by mass spectrometry (MS) analysis as de-
scribed below. These findings suggest that mutations of the
wsfA, wsfB, wsfC, wsfE, wsfF, wsfG and wzt ORFs had signif-
icant effects on SpaA glycosylation in P. alvei CCM 2051T. In
contrast, knockout mutants in the wsfD and wsfH ORFs re-
sulted in the production of glycosylated SpaA protein of
similar mobility as that produced from wild-type cells. As
described previously, insertion in the wsfP gene resulted in a
glycosylation-deficient phenotype (Zarschler et al. 2009).
To ensure that no polar effects on downstream gene expres-
sion caused by the intron insertion as described by Rodriguez
et al. (2008) had occurred, but rather the inactivation of the
target gene itself is responsible for the lack of S-layer glycosyl-
ation in the wsfA, wsfB, wsfC, wsfE, wsfF, wsfG or wzt mutants,
the mRNA of the wsfE::Ll.LtrB mutant was analyzed for such
effects. Using the reverse transcriptase polymerase chain reac-
tion (RT-PCR) approach described above, no differences
between wild-type and wsfE::Ll.LtrB cells could be observed,
indicating, that despite the intron insertion the polycistronic
mRNA is not interrupted (data not shown).
Structural characterization of S-layer glycans produced in slg
gene mutant backgrounds
The elucidation of the glycan structures linked to the S-layer
protein SpaA was accomplished by mass spectrometry of pep-
tides derived from pronase digestion of S-layer glycoproteins
isolated from different slg gene mutants. While in the S-layer
protein fraction of the wsfA, wsfB, wsfC, wsfE, wsfF, wsfG and
wzt mutants no glycans could be detected, the glycopeptides of
the SpaA protein produced in either wsfD or wsfH mutant
strains yielded tandem mass spectroscopy (MS/MS) results
identical to that of the S-layer glycopeptides from the wild-type
strain (Figure 5). The online MS/MS spectrum of the SpaA
glycopeptides of wild-type P. alvei CCM 2051Tconfirmed
the already known branched structure of the repeating units
(Figure 5A). The three peaks at 528, 1055 and 1582 Da repre-
sent one, two or three repeating units of two hexoses and one
N-acetylhexosamine residue each. Due to the collision energy
chosen for the MS experiment, small fragments consisting of
only a few S-layer glycan repeating units were obtained, which
differs from the native long-chain S-layer glycan. The
corresponding online MS/MS spectra of the wsfD::Ll.LtrB
(Figure 5B) and the wsfH::Ll.LtrB (not shown) mutants
showed four peaks at 366, 731, 1097 and 1462 Da, which
are consistent with one to four repeating units containing one
hexose and one N-acetylhexosamine residue. This observation
indicates the absence of the α1,6-linked glucose residues of the
repeating units in the glycosylated peptides of SpaA protein
produced in wsfD::Ll.LtrB and wsfH::Ll.LtrB mutant strains,
thus indicating the involvement of WsfD and WsfH in the
Fig. 4. Effect of insertional inactivation of ORFs from the P. alvei CCM 2051Tslg gene cluster. An aliquot of biomass (200 µg) from various P. alvei CCM 2051T
mutant strains was analyzed by SDS-PAGE followed by Coomassie Brilliant Blue G 250 staining. The mutant strains carry the Ll.LtrB insertion as indicated. Tri-
banded appearance corresponds to non-glycosylated (N), monoglycosylated (M) and di-glycosylated (D) chimeric SpaA.
K Zarschler et al.
glucosylation of the N-acetylmannosamine residues of the
Gram-positive bacterium P. alvei CCM 2051Tencoding the O-
glycosylation of tyrosine residues of the S-layer protein of this
which the derived Wsf protein sequences show homology to
proteins involved in the biosynthesis of different bacterial sur-
face polysaccharides, such as LPS, exopolysaccharides and
capsule polysaccharides. Both the observed similarity of the
Wsf proteins with database entries of enzymes involved in bac-
terial polysaccharide biosyntheses as well as the disruption of
the corresponding ORFs gave first insights into the S-layer gly-
can biosynthesis pathway of P. alvei CCM 2051T(Figure 6).
Seven ORFs located in the slg gene cluster are involved in
the biosynthesis of nucleotide-activated monosaccharides. Next
to galE converting UDP-glucose to UDP-galactose, galU trans-
ferring UTP to glucose-1-phosphate, resulting in UDP-glucose,
is present. While tagD is involved in the formation of CDP-
glycerol, the four rml genes code for the biosynthesis of
dTDP-L-rhamnose. Since the S-layer glycan contains glucose,
Fig. 5. Mass spectrometry of glycopeptides. Glycopeptides derived from pronase digestion of SpaA protein produced in P. alvei CCM 2051Twild-type (A) and
wsfD::Ll.LtrB mutant strain (B) were analyzed by online MS/MS, showing the constitution of the different repeating units of the wild-type and the mutant strains.
The spectrum of wsfH::Ll.LtrB mutant strain is identical to that observed for the wsfD::Ll.LtrB mutant strain and hence not shown. Please note that, to keep a
constant scale in either mass spectrum, for the mutant strain, one more repeating unit is shown to account for the fact that in each unit one hexose is missing.
Tyrosine O-glycosylation in Paenibacillus alvei
Fig. 6. Working model of S-layer glycan biosynthesis in P. alvei CCM 2051T. The initial transfer of a Gal residue from UDP-α-D-Gal to a lipid carrier is catalyzed by WsfP (A). The adaptor saccharide is formed
by the α1,3-linkage of an L-Rha residue from dTDP-β-L-Rha to the linkage sugar D-Gal possibly performed by WsfG, followed by the transfer of two additional α1,3-linked L-Rha residues possibly by the action
of WsfF (B). The glycan chain would be elongated by the activity of the aminosugar transferase WsfE and the tripartite transferase WsfC. WsfE may form the β1,4-linkage of a ManNAc residue from UDP-
ManNAc to the third rhamnose residue. WsfC putatively adds a single glycerol phosphate from CDP-glycerol to the ManNAc residue of the adaptor oligosaccharide and may form the β1,3-linkage of a ManNAc
residue to the third rhamnose residue as well as the β1,4-linkage of a Gal to the ManNAc residues of the repeating units. The glycan chain would be recognized by the carboxy-terminal part of Wzt and exported by
the ABC transporter system through the cytoplasmic membrane (C). The transfer of cytoplasmic Glc to the lipid carrier would be carried out by WsfH and, after reorientation, is used at the external face of the
cytoplasmic membrane by WsfD for α1,6-linkage of the Glc residues to ManNAc residues of the repeating units (D). The final transfer of the completed S-layer glycan to certain tyrosine residues of the S-layer
protein is predicted to occur co-secretionally upon catalysis of the O-OTase WsfB (E). Eventually, the mature S-layer glycoprotein would be self-assembled at the cell surface (F). Please note that so far only the
WsfP protein has been experimentally verified to perform its predicted role.
K Zarschler et al.
galactose, phosphoglyceric acid and rhamnose, the presence of
these ORFs in the slg gene cluster is not surprising. However,
no ORFs for the biosynthesis of nucleotide-activated N-acetyl-
mannosamine are located in the slg gene cluster. This
observation confirms the assumption that housekeeping genes
are additionally required for S-layer glycan biosynthesis (No-
votny, Schäffer et al. 2004).
As recently described and depicted in Figure 6A, the UDP-
Gal:phosphoryl-polyprenol Gal-1-phosphate transferase WsfP
acts as the initiation enzyme of S-layer glycan biosynthesis
in P. alvei CCM 2051T(Zarschler et al. 2009).
The slg gene cluster encodes two rhamnosyltransferases
(WsfF and WsfG) which are obviously involved in the assem-
bly of the L-rhamnose-containing adaptor saccharide of the S-
layer O-glycan, with WsfG transferring an L-rhamnose residue
onto the linkage sugar galactose and WsfF acting as an α-L-
rhamnose-1,3-α-L-rhamnosyltransferase (Figure 6B). Since
WsfE is related to the aminosugar transferase WbuB, it might
be responsible for the addition of an N-acetylmannosamine res-
idue to the growing glycan chain (Figure 6C). Based on the
observed similarity with the database, the proposed function
of the tripartite transferase WsfC might be the transfer of a sin-
gle glycerol phosphate from CDP-glycerol to the N-
acetylmannosamine residue of the adaptor oligosaccharide
and of a galactose residue to the N-acetylmannosamine resi-
dues of the repeating units. Since this enzyme exhibits a
third transferase domain, it may also catalyze the reaction from
glycerol phosphate to 2-phospho glyceric acid (Figure 6C).
The S-layer glycoproteins of the mutant strains wsfD::Ll.
LtrB and wsfH::Ll.LtrB show identical migration behavior in
SDS-PAGE compared to wild-type cells. MS analysis of the
S-layer O-glycan of both mutants showed the lack of glucose
residues being part of each repeating unit of the mature glycan,
suggesting that both enzymes, WsfD and WsfH, are involved
in the process of glucose addition to the glycan chain. Due to
the similarity of WsfD to fungal Pmts and of WsfH to several
polyprenyl-phosphate β-D-glucosyltransferases, we assume
that WsfH transfers a glucose residue to undecaprenyl-pyro-
phosphate at the inner face of the cytoplasmic membrane,
the lipid carrier is then re-orientated to the external face of
the cytoplasmic membrane, and WsfD adds the glucose residue
to the exported glycan chain (Figure 6D).
The identification of an ABC transporter system (Wzm and
Wzt) and the loss of S-layer glycosylation in the wzt::Ll.LtrB
mutant corroborate the assumption of an ATP hydrolysis-driven
export of the undecaprenyl-pyrophosphate-linked glycan chain
to the external face of the cytoplasmic membrane comparable to
the ABC-transporter-dependent pathway of the LPS O-polysac-
charide biosynthesis. According to this pathway, glycan chain
extension is achieved by processive addition of sugar residues
to the nonreducing terminus of the undecaprenyl-pyrophos-
phate-linked growing chain. Although not yet detected in P.
alvei CCM 2051T, nonreducing terminal modifications, such
as 2-O-methyl groups, were described as chain length termina-
tion signal recognized by the carboxy-terminal domain of Wzt
(Cuthbertson et al. 2007). The polymer is then exported through
the cytoplasmic membrane by the ABC transporter for ligation,
independent of the presence of a Wzx flippase or Wzy polymer-
ase homolog (Whitfield 1995; Raetz and Whitfield 2002). After
the addition of glucose to the N-acetylmannosamine residue of
each repeating unit by WsfD, the completed glycan chain
would be transferred from the lipid carrier to specific tyrosine
residues of SpaA by the O-OTase WsfB, possibly upon export
of the S-layer protein across the cytoplasmic membrane
(Figure 6E). This conclusion is based on the presence of the
conserved Wzy_C motif in WsfB and on the similar transmem-
brane topology of the well characterized OTases WaaL, PglL
and PilO, as well as on the loss of S-layer glycosylation in
the wsfB::Ll.LtrB mutant. Eventually, the mature S-layer glyco-
protein would be self-assembled at the cell surface (Figure 6F).
The role of WsfA in S-layer glycosylation biosynthesis re-
mains still unclear. Since the wsfA::Ll.LtrB mutant shows an
altered S-layer migration in SDS-PAGE due to the loss of
the SpaA-linked glycan chains, its involvement in the glyco-
sylation process is evident but needs to be further investigated.
Although several putative bacterial promoters and termina-
tors have been identified by different prediction programs,
most of the ORFs, namely those coding for GalU, Wzm, Wzt,
WsfA, TagD, WsfCDE, RmlACBD and WsfFGPH, are tran-
scribed by a single polycistronic mRNA. However, the UDP-
glucose 4-epimerase GalE and the OTase WsfB are transcribed
independently. A common phenomenon of bacterial polysac-
charide biosynthesis gene clusters is the low % G + C content
compared to the respective bacterial genome as a whole (No-
votny, Pföstl et al. 2004; Messner et al. 2008). For individual
ORFs of the described slg gene cluster, the % G + C content
ranges between 25% and 43%, whereas it is 44.6% or 46.2%
for the rest of the genome of P. alvei CCM 2051T, depending
on the method of its determination (Claus and Berkeley 1986).
In conclusion, the current report describes the identification,
annotation and characterization of the slg gene cluster of P. al-
vei CCM 2051Tinvolved in tyrosine O-glycosylation of the S-
layer protein of this organism. Considering the documented
chemical stability of this rare O-glycosidic linkage type (Kolbe
1993), these data mark a starting point for further studies and
applications in conjunction with the unique self-assembly fea-
ture of an S-layer protein matrix. This may lead to the future
design of functional glycans and their controlled surface dis-
play for exerting biological activity in various settings.
Materials and methods
Bacterial strains and growth conditions
Bacterial strains, plasmids and primers are listed in Tables SI
and SII. P. alvei CCM 2051Twas obtained from the Czech Col-
lection of Microorganisms (CCM, Brno, Czech Republic) and
was cultivated at 37°C and 200 rpm in Luria–Bertani (LB)
broth or on LB agar plates supplemented with 10 μg/ml chlor-
amphenicol (Cm), when appropriate. E. coli DH5α
(Invitrogen, Lofer, Austria) was grown in LB broth at 37°C
supplemented with 30 μg/ml Cm, when appropriate.
Analytical and general methods
Genomic DNA of P. alvei CCM 2051Twas isolated as de-
scribed recently (Zarschler et al. 2009). Restriction and
cloning enzymes were purchased from Invitrogen. The MinE-
lute gel extraction kit (Qiagen, Vienna, Austria) was used to
purify DNA fragments from agarose gels, and the MinElute re-
action cleanup kit (Qiagen) was used to purify digested
oligonucleotides and plasmids. Plasmid DNA from trans-
formed cells was isolated with the Plasmid Miniprep kit
Tyrosine O-glycosylation in Paenibacillus alvei
(Qiagen). Agarose gel electrophoresis was performed as de-
scribed elsewhere (Sambrook et al. 1989). Transformation of
E. coli DH5 α was done according to the manufacturer’s pro-
tocol (Invitrogen). Transformants were screened by in situ PCR
using RedTaq ReadyMix PCR mix (Sigma-Aldrich, Vienna,
Austria), and recombinant clones were analyzed by restriction
mapping and confirmed by sequencing (Agowa, Berlin, Ger-
many). Transformation of P. alvei CCM 2051Twas
performed as described recently (Zarschler et al. 2009). SDS-
PAGE was carried out according to a standard protocol
(Laemmli 1970) using a Protean II electrophoresis apparatus
(Bio-Rad, Vienna, Austria). Protein bands were visualized with
Coomassie Brilliant Blue G 250 staining reagent. The isolation
and purification of S-layer glycoprotein essentially followed
published methods (Messner and Sleytr 1988).
PCR and DNA sequencing
Primers for PCR and DNA sequencing were purchased from In-
vitrogen, and PCR conditions were optimized for each primer
pair (Table SII). PCR was performed using the Herculase ® II
Fusion DNA Polymerase (Stratagene, La Jolla, CA) and the
thermal cycler My CyclerTM(Bio-Rad). Amplification products
were purified using the MinElute PCR purification kit (Qiagen).
For the identification of the rml genes responsible for dTDP-L-
rhamnose biosynthesis, the highly conserved seven amino acid
tase) was used for the design of the degenerate oligonucleotide
primer proof_RmlD_for. For sequence determination of the slg
gene cluster, chromosome walking was applied as previously
described (Kneidinger et al. 2001; Pilhofer et al. 2007).
Total RNA was extracted from P. alvei CCM 2051Tusing the
RNeasy Protect Bacteria Mini Kit (Qiagen) and subsequently
treated with RNase-free DNase I (Fermentas, St. Leon-Rot,
Germany) to remove DNA contamination. First strand cDNA
was synthesized utilizing the Revert AidTMPremium Reverse
Transcriptase (Fermentas) according to the manufacturer’s in-
structions using a reverse primer specific for wsfB (1f), wsfA
(4r), wsfF (8r) or pcrB (9r) (Table SII). After termination of
the reaction by heating at 85°C for 5 min, one tenth of each
cDNA reaction mixture was used as template for PCR using
the Phusion™ High-Fidelity DNA Polymerase (New England
Biolabs, Frankfurt/Main, Germany). PCR reactions were car-
ried out with primer pairs annealing to wsfB/galE (1f/1r),
galU/wzm (2f/2r), wzt/wsfA (3f/3r), wsfA (4f/4r), wsfC/wsfD
(5f/5r), wsfD/wsfE (6f/6r), rmlB/wsfF (8f/8r) and wsfH/pcrB
(9f/9r) (Table SII). As a positive control, genomic DNA
was used, whereas DNase I-treated RNA without the
cDNA-generating step served as a control for contamination
of total RNA with chromosomal DNA. PCR products were
analyzed by agarose gel electrophoresis.
Nucleotide and protein sequences were analyzed using the
BLASTN and BLASTP sequence homology analysis tools (Na-
tional Center for Biotechnology Information, Bethesda, MD).
Open reading frames in the DNA sequence were identified by
using the Clone Manager Professional Suite (SECentral, Cary,
NC) and the ORF Finder analysis tool (National Center for Bio-
technology Information). The TMHMM Server v. 2.0
transmembrane prediction program and the SignalP 3.0 Server
(Center for Biological Sequence Analysis, Technical Universi-
ty of Denmark, Lyngby, Denmark) were used to identify
putative protein transmembrane-spanning domains and the
presence and location of signal peptide cleavage sites, respec-
tively. The guanine-cytosine (G+C) content of the entire slg
gene cluster was determined using the GC Content and GC
Skew program (Nano+Bio-Center, University of Kaiserslau-
tern, Germany). For in silico reverse translation, the
Sequence Manipulation Suite was used (Stothard 2000). Bac-
terial promoters, transcriptional terminators, operons and ORFs
were predicted by the BProm and FindTerm modules of the
FGenesB gene prediction program in Molquest software (Soft-
Berry Inc., Mount Kisco, NY). The presence of conserved
motifs in a given protein sequence was analyzed by the Pfam
protein families database (Finn et al. 2008). Physical and chem-
ical parameters for a given protein were calculated using the
ProtParam tool (Gasteiger et al. 2005).
Specific disruption of nine ORFs located in the slg gene cluster
of P. alvei CCM 2051Twas performed as described recently
(Zarschler et al. 2009). The Ll.LtrB targetron of pTT_wsfP1176
was retargeted prior to transformation into P. alvei CCM 2051T.
For this purpose, identification of potential insertion sites and
design of PCR primers for the modification of the intron RNA
was accomplished by a computer algorithm (www.Sigma-Al-
drich.com/Targetronaccess). For each ORF, insertion sites
were chosen based on their location and intron insertion effi-
ciency, and modifications of the intron RNA sequences were
introduced via PCR by primer-mediated mutation (Table SIII).
The retargeted Ll.LtrB targetron was subsequently digested
with HindIII and BsrGI and ligated into pTT_wsfP1176 di-
gested with the same restriction enzymes, thereby replacing
the wsfP targetron. Creation of P. alvei CCM 2051Tgene
knockout mutants and confirmation of intron insertion was
achieved as described (Zarschler et al. 2009). All mutant strains
were analyzed for the migration behavior of the SpaA (glyco)
protein on SDS-PA gels (Altman et al. 1995). Glycan structures
linked to the S-layer protein SpaA of P. alvei CCM 2051Twild-
type and mutant cells were analyzed by mass spectrometry.
Protein elution from SDS-PAGE and pronase digest
Aliquots of biomass were run on a 10% PA gel and stained with
Coomassie Brilliant Blue G 250. Relevant SDS-PA gel bands
were excised from the gel and destained (Stadlmann et al.
2008). Destained gel pieces were minced into small particles
from the gel, the protein was dialyzed against 10 mM ammonia
formate buffer and precipitated using five volumes of acetone
(−20°C,1h). Two hundred microliters of0.15M Tris-HCl buff-
pronase for 24 h at 37°C. Subsequently, an additional amount of
1 µg pronase was added, and incubation was continued for 24 h.
Glycopeptides (∼10 μg of gel-eluted material) were enriched
using a PGC-SPE cartridge (10 mg; Fisher Scientific, Vienna,
Austria) according to a published protocol (Packer et al. 1998),
except for the use of 150 mM ammonia formate buffer, pH 9.0,
K Zarschler et al.
in 60% acetonitrile for elution. Prior to liquid chromatography
(LC)-MS analysis, the samples were dried under vacuum and
dissolved in 20 μl of distilled water.
Glycopeptide LC-electrospray ionization-MS/MS
Analysis of glycopeptides was performed by positive-ion LC-
electrospray ionization (ESI)-MS/MS using a 50 × 0.32 mm
porous graphitic carbon column (Thermo) (Pabst and Altmann
2008). A flow rate of 5 μl was maintained with a Dionex Ul-
timate 3000 cap flow system using 150 mM ammonia formate
buffer (pH 9) as solvent A and acetonitrile as solvent B.
Glycopeptides were eluted from the separation column using
a gradient from 5% to 45% of solvent B over 25 min. Analysis
was carried out with a Waters Q-TOF Ultima Global mass spec-
trometer with standard ESI-source and a MassLynx V4.0 SP4
software for evaluation of obtained peaks. For online MS/MS
experiments, the CE was set to 72 with a LM Res of 8 and a
HM Res of 11. Capillary voltage, MS profile, cone voltage, RF
LENS 1 setting and ESI probe adjustment were optimized to
gain maximum signal intensity.
Supplementary data for this article is available online at
Financial support came from the Austrian Science Fund FWF,
project P20745-B11 (to P.M.) and projects P19047-B12 and
P21945-B20 (to C.S.), and the Hochschuljubiläumsstiftung
der Stadt Wien, project H-02229-2007 (to K.Z.).
Conflict of interest statement
ABC, ATP binding casstte; AsnB, asparagine synthetase B;
CDP, cytidine diphosphate; Cm, chloramphenicol; CTP, cyti-
dine triphosphate; ER, endoplasmic reticulum; ESI,
electrospray ionization; G+C, guanine-cytosine; LB, Luria–
Bertani; LPS, lipopolysaccharide; MS, mass spectrometry;
protein transferase; ORF, open reading frame; PMTs,
SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel elec-
trophoresis; TPR, tetratricopeptide repeat; UDP, uridine
diphosphate; UTP, uridine triphosphate.
Abeyrathne PD, Daniels C, Poon KK, Matewish MJ, Lam JS. 2005. Functional
characterization of WaaL, a ligase associated with linking O-antigen poly-
saccharide to the core of Pseudomonas aeruginosa lipopolysaccharide. J
Altman E, Brisson J-R, Messner P, Sleytr UB. 1991. Structure of the glycan
chain from the surface layer glycoprotein of Bacillus alvei CCM 2051.
Biochem Cell Biol. 69:72–78.
Altman E, Schäffer C, Brisson J-R, Messner P. 1995. Characterization of the
glycan structure of a major glycopeptide from the surface layer glycopro-
tein of Clostridium thermosaccharolyticum E207-71. Eur J Biochem.
Aon MA, Curtino JA. 1985. Protein-bound glycogen is linked to tyrosine re-
sidues. Biochem J. 229:269–272.
Bierbaum G, Sahl HG. 2009. Lantibiotics: mode of action, biosynthesis and
bioengineering. Curr Pharm Biotech. 10:2–18.
Bhavsar AP, Truant R, Brown ED. 2005. The TagB protein in Bacillus subtilis
168 is an intracellular peripheral membrane protein that can incorporate
glycerol phosphate onto a membrane-bound acceptor in vitro. J Biol Chem.
Bock K, Schuster-Kolbe J, Altman E, Allmaier G, Stahl B, Christian R, Sleytr
UB, MessnerP. 1994.Primarystructureof the O-glycosidically linked glycan
chain of the crystalline surface layer glycoprotein of Thermoanaerobacter
thermohydrosulfuricus L111-69. Galactosyl tyrosine as a novel linkage unit.
J Biol Chem. 269:7137–7144.
Bronner D, Clarke BR, Whitfield C. 1994. Identification of an ATP-binding
cassette transport system required for translocation of lipopolysaccharide
O-antigen side-chains across the cytoplasmic membrane of Klebsiella pneu-
moniae serotype O1. Mol Microbiol. 14:505–519.
Chen A, Zhang D, Poulter CD. 1993. (S)-geranylgeranylglyceryl phosphate
synthase. Purification and characterization of the first pathway-specific en-
zyme in archaebacterial membrane lipid biosynthesis. J Biol Chem. 268:
Chen CY, Wu KM, Chang YC, Chang CH, Tsai HC, Liao TL, Liu YM, Chen
HJ, Shen AB, Li JC, et al. 2003. Comparative genome analysis of Vibrio
vulnificus, a marine pathogen. Genome Res. 13:2577–2587.
Chen PS, Mitchell HK, Neuweg M. 1978. Tyrosine glucoside in Drosophila
busckii. Insect Biochem. 8:279–286.
Christian R, Messner P, Weiner C, Sleytr UB, Schulz G. 1988. Structure of a
glycan from the surface-layer glycoprotein of Clostridium thermohydrosul-
furicum strain L111-69. Carbohydr Res. 176:160–163.
Claus D, Berkeley RCW. 1986. Bacillus Cohn 1872. In: Sneath PAH, editor.
Bergey’s Manual of Systematic Bacteriology, vol. 2 Baltimore: The Williams
and Wilkins Co. p. 1105–1139.
Cuthbertson L, Kimber MS, Whitfield C. 2007. Substrate binding by a bacterial
ABC transporter involved in polysaccharide export. Proc Natl Acad Sci.
Cuthbertson L, Powers J, Whitfield C. 2005. The C-terminal domain of the
nucleotide-binding domain protein Wzt determines substrate specificity in
the ATP-binding cassette transporter for the lipopolysaccharide O-antigens
in Escherichia coli serotypes O8 and O9a. J Biol Chem. 280:30310–30319.
D'Andrea LD, Regan L. 2003. TPR proteins: the versatile helix. Trends
Biochem Sci. 28:655–662.
D'Souza JM, Wang L, Reeves P. 2002. Sequence of the Escherichia coli O26
O antigen gene cluster and identification of O26 specific genes. Gene.
Faridmoayer A, Fentabil MA, Mills DC, Klassen JS, Feldman MF. 2007.
Functional characterization of bacterial oligosaccharyltransferases involved
in O-linked protein glycosylation. J Bacteriol. 189:8088–8098.
Finn RD, Tate J, Mistry J, Coggill PC, Sammut SJ, Hotz HR, Ceric G, Forslund
K, Eddy SR, Sonnhammer EL, et al. 2008. The Pfam protein families data-
base. Nucleic Acids Res. 36:D281–D288.
Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR, Appel RD,
Bairoch A. 2005. Protein identification and analysis tools on the ExPASy
Server. In: Walker JM, editor. The Proteomics Protocol Handbook. Totowa
(NJ): Humana Press. p. 571–607.
Gentzsch M, Strahl-Bolsinger S, Tanner W. 1995. A new Dol-P-Man:protein
O-D-mannosyltransferase activity from Saccharomyces cerevisiae. Glyco-
GirrbachV, ZellerT, Priesmeier M,Strahl-BolsingerS. 2000.Structure-function
analysis of the dolichyl phosphate-mannose: protein O-mannosyltransferase
ScPmt1p. J Biol Chem. 275:19288–19296.
Gotschlich EC. 1994. Genetic locus for the biosynthesis of the variable por-
tion of Neisseria gonorrhoeae lipooligosaccharide. J Exp Medicine.
Graninger M, Kneidinger B, Bruno K, Scheberl A, Messner O. 2002. Homo-
logs of the Rml enzymes from Salmonella enterica are responsible for
dTDP-β-L-rhamnose biosynthesis in the Gram-positive thermophile Aneur-
inibacillus thermoaerophilus DSM 10155. Appl Environ Microbiol.
Gross H, Stockwell VO, Henkels MD, Nowak-Thompson B, Loper JE, Ger-
wick WH. 2007. The genomisotopic approach: a systematic method to
isolate products of orphan biosynthetic gene clusters. Chem Biol. 14:53–63.
Hirasawa T, Wachi M, Nagai K. 2000. A mutation in the Corynebacterium
growth, and L-glutamate production. J Bacteriol. 182:2696–2701.
Kaneko T, Nakajima N, Okamoto S, Suzuki I, Tanabe Y, Tamaoki M, Naka-
mura Y, Kasai F, Watanabe A, Kawashima K, et al. 2007. Complete genomic
Tyrosine O-glycosylation in Paenibacillus alvei
structure of the bloom-forming toxic cyanobacterium Microcystis aerugino- Download full-text
sa NIES-843. DNA Res. 14:247–256.
Kneidinger B, Graninger M, Messner P. 2001. Chromosome walking by clon-
ing of distinct PCR fragments. Biotechniques. 30:248–249.
Kolbe J. 1993. Charakterisierung von S-Schichtglykoproteinen der Spezies
Clostridium thermosaccharolyticum. Dissertation Universität für Bodenkul-
tur Wien. 146 p.
Kramer KJ, Hopkins TL, Ahmed RF, Mueller D, Lookhart G. 1980. Tyro-
sine metabolism for cuticle tanning in the tobacco hornworm, Manduca
sexta (L.) and other Lepidoptera: identification of β-D-glucopyranosyl-O-
L-tyrosine and other metabolites. Arch Biochem Biophys. 205:146–155.
Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the
head of bacteriophage T4. Nature. 227:680–685.
Mauel C, Young M, Karamata D. 1991. Genes concerned with synthesis of
poly(glycerol phosphate), the essential teichoic acid in Bacillus subtilis
strain 168, are organized in two divergent transcription units. J Gen Microb.
Messner P, Christian R, Kolbe J, Schulz G, Sleytr UB. 1992. Analysis of a
novel linkage unit of O-linked carbohydrates from the crystalline surface
layer glycoprotein of Clostridium thermohydrosulfuricum S102-70. J Bac-
Messner P, Christian R, Neuninger C, Schulz G. 1995. Similarity of “core”
structures in two different glycans of tyrosine-linked eubacterial S-layer gly-
coproteins. J Bacteriol. 177:2188–2193.
Messner P, Egelseer EM, Sleytr UB, Schäffer C. 2009. Surface layer proteins,
secondary cell wall polymers and their glycosylation. In: Moran AP, Bren-
nan PJ, Holst O and von Itzstein M, editors. Microbial Glycobiology:
Structures, Relevance and Applications. San Diego (CA): Academic
Press-Elsevier. p. 109–128.
Messner P, Sleytr UB. 1988. Separation and purification of S-layers from
Gram-positive and Gram-negative bacteria. In: Hancock IC and Poxton
IR, editors. Bacterial Cell Surface Techniques. Chichester: John Wiley &
Sons. p. 97–104.
Messner P, Steiner K, Zarschler K, Schäffer C. 2008. S-layer nanoglycobiology
of bacteria. Carbohydr Res. 343:1934–1951.
Milman HA, Cooney DA. 1979. Partial purification and properties of L-aspar-
agine synthetase from mouse pancreas. Biochem J. 181:51–59.
Novotny R, Pföstl A, Messner P, Schäffer C. 2004. Genetic organization of
chromosomal S-layer glycan biosynthesis loci of Bacillaceae. Glycoconj
Novotny R, Schäffer C, Strauss J, Messner P. 2004. S-layer glycan-specific loci
on the chromosome of Geobacillus stearothermophilus NRS 2004/3a and
dTDP-L-rhamnose biosynthesis potential of G. stearothermophilus strains.
Nölling J, Breton G, Omelchenko MV, Makarova KS, Zeng Q, Gibson R, Lee
HM, Dubois J, Qiu D, Hitti J, et al. 2001. Genome sequence and compar-
ative analysis of the solvent-producing bacterium Clostridium
acetobutylicum. J Bacteriol. 183:4823–4838.
Pabst M, Altmann F. 2008. Influence of electrosorption, solvent, temperature,
and ion polarity on the performance of LC-ESI-MS using graphitic carbon
for acidic oligosaccharides. Anal Chem. 80:7534–7542.
Packer NH, Lawson MA, Jardine DR, Redmond JW. 1998. A general approach
to desalting oligosaccharides released from glycoproteins. Glycoconj J.
Park YS, Sweitzer TD, Dixon JE, Kent C. 1993. Expression, purification, and
characterization of CTP:glycerol-3-phosphate cytidylyltransferase from Ba-
cillus subtilis. J Biol Chem. 268:16648–16654.
Pilhofer M, Bauer AP, Schrallhammer M, Richter L, Ludwig W, Schleifer KH,
Petroni G. 2007. Characterization of bacterial operons consisting of two tu-
bulins and a kinesin-like gene by the novel two-step gene walking method.
Nucleic Acids Res. 35:e135.
Power PM, Seib KL, Jennings MP. 2006. Pilin glycosylation in Neisseria
meningitidis occurs by a similar pathway to wzy-dependent O-antigen
biosynthesis in Escherichia coli. Biochem Biophys Res Commun. 347:
Raetz CRH, Whitfield C. 2002. Lipopolysaccharide endotoxins. Annu Rev Bio-
Randriantsoa M, Drouillard S, Breton C, Samain E. 2007. Synthesis of globo-
pentaose using a novel β1, 3-galactosyltransferase activity of the
Haemophilus influenzae β1, 3-N-acetylgalactosaminyltransferase LgtD.
FEBS Lett. 581:2652–2656.
Ren H, Liu J. 2006. AsnB is involved in natural resistance of Mycobacterium
smegmatis to multiple drugs. Antimicrob Agents Chemother. 50:250–255.
Rocchetta HL, Lam JS. 1997. Identification and functional characterization
of an ABC transport system involved in polysaccharide export of A-
band lipopolysaccharide in Pseudomonas aeruginosa. J Bacteriol.
Rodriguez IR, Whelan WJ. 1985. A novel glycosyl-amino acid linkage: rabbit-
muscle glycogen is covalently linked to a protein via tyrosine. Biochem Bio-
phys Res Commu. 132:829–836.
Rodriguez SA, Yu JJ, Davis G, Arulanandam BP, Klose RE. 2008. Targeted
inactivation of Francisella tularensis genes by group II introns. Appl Envi-
ron Microbiol. 74:2619–2626.
Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
Schäffer C, Dietrich K, Unger B, Scheberl A, Rainey FA, Kählig H, Messner P.
2000. A novel type of carbohydrate-protein linkage region in the tyrosine-
bound S-layer glycan of Thermoanaerobacterium thermosaccharolyticum
D120-70. Eur J Biochem. 267:5482–5492.
Scofield MA, Lewis WS, Schuster SM. 1990. Nucleotide sequence of Escher-
ichia coli asnB and deduced amino acid sequence of asparagine synthetase
B. J Biol Chem. 265:12895–12902.
Shao J, Zhang J, Kowal P, Lu Y, Wang PG. 2002. Overexpression and bio-
chemical characterization of β-1, 3-N-acetylgalactosaminyltransferase
LgtD from Haemophilus influenzae strain Rd. Biochem Biophys Res Comm.
Sleytr UB, Messner P. 2009. Crystalline cell surface layers (S-layers). In:
Schaechter M, editor. Encyclopedia of Microbiology, vol. 1, 3rd ed. San Die-
go (CA): Academic Press/Elsevier Science. p. 89–98.
Stadlmann J, Pabst M, Kolarich D, Kunert R, Altmann F. 2008. Analysis of
immunoglobulin glycosylation by LC-ESI-MS of glycopeptides and oligo-
saccharides. Proteomics. 8:2858–2871.
Steiner K, Hageluken G, Messner P, Schäffer C, Naismith J. 2010. Structural
basis of substrate binding in WsaF, a rhamnosyltransferase from Geobacillus
stearothermophilus. J Mol Biol. 397:436–447.
Steiner K, Novotny R, Werz DB, Zarschler K, Seeberger PH, Hofinger A, Kos-
ma P, Schäffer C, Messner P. 2008. Molecular basis of S-layer glycoprotein
glycan biosynthesis in Geobacillus stearothermophilus. J Biol Chem.
analyzing and formatting protein and DNA sequences. Biotechniques.
Strahl-Bolsinger S, Immervoll T, Deutzmann R, Tanner W. 1993. PMT1, the
gene for a key enzyme of protein O-glycosylation in Saccharomyces cere-
visiae. Proc Natl Acad Sci. 90:8164–8168.
Strahl-Bolsinger S, Scheinost A. 1999. Transmembrane topology of Pmt1p, a
member of an evolutionarily conserved family of protein O-mannosyltrans-
ferases. J Biol Chem. 274:9068–9075.
Strahl-Bolsinger S, Tanner W. 1991. Protein O-glycosylation in Saccharomyces
cerevisiae. Purification and characterization of the dolichyl-phosphate-D-
mannose-protein O-D-mannosyltransferase. Eur J Biochem. 196: 185–190.
Varki A, Cummings R, Esko J, Freeze H, Hart G, Marth J. 1999. Essentials of
Glycobiology. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory
Walker JE, Saraste M, Runswick MJ, Gay NJ. 1982. Distantly related se-
quences in the alpha- and β-subunits of ATP synthase, myosin, kinases
and other ATP-requiring enzymes and a common nucleotide binding fold.
EMBO J. 1:945–951.
Welsh EA, Liberton M, Stockel J, Loh T, Elvitigala T, Wang C, Wollam C,
Fulton RS, Clifton SW, Jacobs JM, et al. 2008. The genome of Cyanothece
51142, a unicellular diazotrophic cyanobacterium important in the marine
nitrogen cycle. Proc Natl Acad Sci. 105:15094–15099.
Whitfield C. 1995. Biosynthesis of lipopolysaccharide O antigens. Trends Mi-
Yurist-Doutsch S, Abu-Qarn M, Battaglia F, Morris HR, Hitchen PG, Dell A,
in the N-glycosylation of the Haloferax volcanii S-layer glycoprotein. Mol
Zarschler K, Janesch B, Zayni S, Schäffer C, Messner P. 2009. Construction of
a gene knockout system for application in Paenibacillus alvei CCM 2051T,
exemplified by the S-layer glycan biosynthesis initiation enzyme WsfP. Appl
Environ Microbiol. 75:3077–3085.
Zayni S, Steiner K, Pföstl A, Hofinger A, Kosma P, Schäffer C, Messner P.
2007. The dTDP-4-dehydro-6-deoxyglucose reductase encoding fcd gene
is part of the surface layer glycoprotein glycosylation gene cluster of Geo-
bacillus tepidamans GS5-97T. Glycobiology. 17:433–443.
K Zarschler et al.