Structure and gene cluster of the O-antigen of Escherichia coli O41

Article abstract of DOI:10.1016/j.carres.2011.12.009

The acidic O-polysaccharide (O-antigen) of Escherichia coli O41 was studied by sugar analysis along with 1D and 2D ¹H and ¹³C NMR spectroscopy, and the following structure of the branched hexasaccharide repeating unit was established: This structure is unique among the known structures of bacterial polysaccharides. The O-antigen gene cluster of E. coli O41 was sequenced. The gene functions were tentatively assigned by a comparison with sequences in the available databases and found to be in full agreement with the E. coli O41 O-polysaccharide structure.

Full text of DOI:10.1016/j.carres.2011.12.009

Carbohydrate Research 349 (2012) 86–89
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Note
Structure and gene cluster of the O-antigen of Escherichia coli O41
b
b
a
Hongfei Zhu a, Andrei V. Perepelov b, , Sof’ya N. Senchenkova , Alexander S. Shashkov , Lei Wang ,
⇑
Yuriy A. Knirel b
a TEDA School of Biological Sciences and Biotechnology, Nankai University, TEDA, Tianjin, China
b N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 119991 Moscow, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
The acidic O-polysaccharide (O-antigen) of Escherichia coli O41 was studied by sugar analysis along with
1D and 2D 1H and 13C NMR spectroscopy, and the following structure of the branched hexasaccharide
repeating unit was established:
Received 11 November 2011
Received in revised form 7 December 2011
Accepted 9 December 2011
Available online 17 December 2011
Keywords:
Escherichia coli
Bacterial polysaccharide structure
O-antigen
Lipopolysaccharide
O-antigen gene cluster
This structure is unique among the known structures of bacterial polysaccharides. The O-antigen gene clus-
ter of E. coli O41 was sequenced. The gene functions were tentatively assigned by a comparison with
sequences in the available databases and found to be in full agreement with the E. coli O41 O-polysaccharide
structure.
Ó 2011 Elsevier Ltd. All rights reserved.
Escherichia coli clones, both commensal and pathogenic, are
normally identified by a combination of their somatic (O), flagellar
(H), and sometimes capsular (K) antigens.1 The O-antigen (O-poly-
saccharide) consisting of a number of oligosaccharide repeats is an
essential component of the lipopolysaccharide on the cell surface
of Gram-negative bacteria and one of the most variable cell con-
stituents. Up to now, more than 180 O-antigen forms have been
recognized in E. coli, and a number of E. coli O-polysaccharide
structures have been elucidated.2
The diversity of the O-antigen forms is almost entirely due to
genetic variations in the O-antigen gene cluster, which is located
between galF and gnd on the chromosome of E. coli. The following
three groups of genes are identified in the cluster: (1) nucleotide
sugar biosynthesis genes; (2) sugar transferase genes; and (3) O-
antigen processing genes including those for O-antigen flippase
(wzx) and polymerase (wzy).1
the lipopolysaccharide isolated from dried bacterial cells by the
phenol–water procedure. Sugar analysis by GLC of the alditol ace-
tates derived after full acid hydrolysis of the polysaccharide re-
vealed Fuc, Gal, and GlcN in a ratio of ꢀ2:1:1. GLC analysis of the
acetylated (S)-2-octyl glycosides demonstrated the
D
configuration
of Gal, GlcN, and GlcA (see below) and the
L configuration of Fuc.
The 13C NMR spectrum of the polysaccharide (Fig. 1) showed
signals for six anomeric carbons in the region d 100.1–104.9, three
C-CH2OH groups (C-6 of hexoses) at d 61.2–63.0, two C-CH3 groups
(C-6 of two Fuc residues) at d 16.5 and 16.6, two nitrogen-bearing
carbons (C-2 of two GlcN residues) at d 56.7 and 57.7, 22 oxygen-
bearing non-anomeric sugar ring carbons in the region d 67.6–81.7,
one C-CO2H group (C-6 of GlcA, see below) at d 174.8 and two N-
acetyl groups at d 23.6, 23.7 (both CH3), 176.1 and 176.3 (both
CO). Accordingly, the 1H NMR spectrum contained signals for six
anomeric protons at d 4.59–5.18, other sugar protons in the region
d 3.43–4.44, two C-CH3 groups (H-6 of two Fuc residues) at d 1.16
and 1.27 and two N-acetyl groups at d 2.00 and 2.04. Therefore, the
polysaccharide has a hexasaccharide repeat (O-unit) containing
In this work, we established the O-polysaccharide structure of
E. coli O41, which had been unknown earlier, and characterized
the O-antigen gene cluster of this bacterium.
Structure elucidation of the O-polysaccharide. A high-molecular
mass O-polysaccharide was obtained by mild acid degradation of
one residue each of
two residues of -GlcNAc (units C and E) and two residues of
-Fuc (units B and D).
D-Gal (denoted as unit A) and D-GlcA (unit F),
D
L
The 1H and 13C NMR spectra of the polysaccharide were as-
signed using 2D homonuclear 1H,1H COSY, TOCSY, ROESY, and
heteronuclear 1H,13C HSQC and HMBC experiments (Table 1).
⇑
Corresponding author. Tel.: +7 499 1376148; fax: +7 499 1355328.
E-mail address: perepel@ioc.ac.ru (A.V. Perepelov).
0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2011.12.009

H. Zhu et al. / Carbohydrate Research 349 (2012) 86–89
87
Figure 1. 13C NMR spectrum of the O-polysaccharide of E. coli O41. Numbers refer to carbons in sugar residues denoted as shown in Table 1.
Table 1
1H and 13C NMR chemical shifts (d, ppm) of the O-polysaccharide of E. coli O41
Sugar residue
H-1
C-1
H-2
C-2
H-3
C-3
H-4
C-4
H-5
C-5
H-6 (6a, 6b) C-6
NAc
CH3
CO
?3)-
A
?3,4)-
B
?4)-b-
C
?3)-
D
a
-
D
-Galp-(1?
-Fucp-(1?
-GlcpNAc-(1?
-Fucp-(1?
-GlcpNAc-(1?
5.18
102.1
5.00
100.8
4.68
100.1
5.04
101.0
4.85
103.0
4.59
104.9
3.81
69.1
4.04
69.3
3.74
57.7
3.80
67.6
3.89
56.7
3.43
74.7
3.85
80.8
3.98
76.5
3.68
73.8
3.99
78.2
3.69
81.7
3.54
76.5
4.11
79.2
4.15
81.5
3.60
78.2
3.88
70.6
3.50
70.2
3.63
73.2
4.18
72.1
4.44
69.0
3.54
76.7
4.29
68.1
3.64
76.9
3.77
78.0
3.66; 3.71
63.0
1.27
16.6
3.84; 3.99
61.2
1.16
16.5
3.75; 3.97
62.5
a
-
L
D
2.04
23.6
176.1
176.3
a-
L
?3)-b-
E
D
2.00
23.7
b-
D-GlcpA-(1?
F
174.8
Based on intra-residue H,H and H,C correlations and coupling
constant values, spin systems were revealed for residues of A–F,
all being in the pyranose form. A relatively large J1,2 coupling con-
stant value of ꢀ7 Hz showed that units C, E, and F are b-linked,
The signals for C-3 of units A, D and E, C-4 of unit C, and both C-
3 and C-4 of unit B were shifted downfield as compared with their
positions in the corresponding non-substituted monosaccharides.3
These data demonstrated a branching character of the polysaccha-
ride chain and defined the glycosylation pattern in the O-unit.
The ROESY spectrum of the polysaccharide showed the follow-
ing correlations between anomeric protons and protons at the
linkage carbons: A H-1, B H-3; B H-1, C H-4; C H-1, D H-3; D H-
1, E H-3, E H-1, A H-3 and F H-1, B H-4 at d 5.18/3.98; 5.00/3.60,
4.68/3.99; 5.04/3.69; 4.85/3.85, and 4.59/4.15, respectively. The
monosaccharide sequence thus determined was confirmed by a
heteronuclear 1H,13C HMBC experiment, which showed correla-
tions between the anomeric protons and linkage carbons and vice
whereas significantly smaller values of <4 Hz indicated the a-link-
age of units A, B, and D.
The spin systems for units C and E were distinguished by a cor-
relation between proton at the nitrogen-bearing carbons (H-2) and
the corresponding carbons (C-2) at d 3.74/57.7 and 3.89/56.7,
respectively. Unit F was identified as GlcA based on correlations
of H-1 with H-2–H-5 in the TOCSY spectrum and H-5 with C-6 at
d 3.77/174.8 in the HMBC spectrum.

88
H. Zhu et al. / Carbohydrate Research 349 (2012) 86–89
versa (data not shown).Therefore, the O-polysaccharide of E. coli
O41 has the structure shown above. To our knowledge, this struc-
ture is unique among the known bacterial polysaccharide
structures.
fcl, gmm, manC, and manB have been identified to be responsible
for the biosynthesis of GDP-
-Fuc.6,7 Therefore, orf7-10 and orf12
were concluded to encode the enzymes for the synthesis of GDP-
L
L-Fuc and named accordingly.
Characterization of the O-antigen gene cluster. The O-antigen
gene cluster of E. coli O41 was found between the housekeeping
genes galF and gnd. A DNA sequence of 15,358 bp was obtained,
which contains 12 open reading frames (ORFs) with transcription
direction from galF to gnd (Fig. 2). All genes were assigned func-
tions based on their similarities to genes from the available dat-
abases (Table 2). The gene wecA that is responsible for the
transfer of GlcNAc-1-phosphate to an undecaprenol phosphate
(UndP) carrier to initiate the O-unit synthesis is located outside
the O-antigen gene cluster.4 Genes involved in the biosynthesis
Orf2, 3, 5, and 11 share 28–69% identities to glycosyltransfer-
ases of other origins and belonged to glycosyltransferase family 1
(Pf00534) or family 2 (PF00535). Therefore, orf2, 3, 5, and 11 were
proposed to encode glycosyltransferases and named wfcV, wfcW,
wfcX, and wfcZ, respectively.
Orf1 and 4 are the only two proteins with predicted membrane
segments. Orf1 has 10 predicted transmembrane segments, which
is a typical topological character of Wzx proteins.8 It also shares
32% identity to the Wzx protein of Yersinia enterocolitica O8. Orf4
has 9 predicted transmembrane segments and a periplasmic loop
of 52 amino acid residues, which is a typical topological character
of Wzy proteins.9 Therefore, orf1 and orf4 were assigned functions
of the O-antigen flippase and polymerase genes, respectively, and
named accordingly.
of common sugar nucleotide precursors such as UDP-
D-GlcNAc
and UDP- -Gal are not in the cluster too, and the gene ugd respon-
D
sible for the synthesis of UDP-D-GlcA is located between gnd and
hisI near the O-antigen gene cluster.5
Proteins encoded by orf7-10 and orf12 share high level of iden-
tity to many known Gmd, Fcl, Gmm, ManC, and ManB proteins
(53–90%), respectively. The corresponding genes including gmd,
No function of Orf6 was revealed by searching available dat-
abases. However, as the O-antigen contains six sugars but only four
possible glycosyltransferase genes were found in the O-antigen
Figure 2. Schematic of the E. coli O41 O-antigen gene cluster.
Table 2
Characteristics of ORFs in the E. coli O41 O-antigen gene cluster
Orf
No.
Gene
name
Position of
gene
G+C
content
(%)
Conserved domain(s)
Similar protein(s), strain(s)
(Genbank Accession No.)
%Identical /
%similar
(No. of aa
overlap)
Putative function of
protein
1
2
wzx
1114..2433
2433..3362
32.0
30.6
Polysaccharide biosynthesis protein
(PF01943)
Wzx
32/56 (426)
O-antigen flippase
Glycosyl transferase
Y. enterocolitica type 0:8
(AAC60766)
Glycosyl transferase, group 1/2
family protein
E value = 6 Â eÀ25
wfcV
28/52 (274)
G. sulfurreducens PCA
(AAR36415)
3
wfcW
3374..4630
32.8
Glycosyl transferases group 1 (PF00534) Putative glycosyltransferase
33/54 (416)
Glycosyltransferase
E value = 1.4 Â eÀ27
B. thetaiotaomicron VPI-5482
(AAO78052)
4
5
wzy
4649..5845
5842..6591
28.1
28.5
NADH dehydrogenase subunit 2
D. simulans (AAF77291)
Putative glycosyl transferase
B. fragilis (AAD40722)
26/47 (215)
31/53 (218)
O-antigen polymerase
Glycosyltransferase
wfcX
Glycosyl transferase group 2 (PF00535)
E value = 2.1 Â eÀ23
6
7
wfcY
gmd
6588..7445
7447..8565
28.6
42.1
Glycosyltransferase
GDP-mannose-4,6-
dehydratase
NAD dependent epimerase/dehydratase Gmd
84/93 (271)
76/85 (320)
53/68 (148)
81/91(468)
67/81 (241)
90/94 (473)
family (PF01370)
Y. enterocolitica
(type 0:8) (AAC60773)
-fucose synthetase
E value = 4.9 Â eÀ18
8
fcl
8569..9534
9537..9998
33.3
37.2
NAD dependent epimerase/dehydratase GDP-
L
GDP-L-fucose synthetase
family (PF01370)
Y. pseudotuberculosis (type O:1b)
(CAB63301)
GDP-mannose mannosyl
hydrolase
S. typhimurium (AAG24815)
GDP-mannose
pyrophosphorylase
E value = 1.3 Â eÀ06
NUDIX domain (PF00293)
E value = 2.3 Â eÀ19
9
gmm
manC
wfcZ
manB
GDP-mannose
mannosyl hydrolase
10
11
12
10004..11410 42.1
11410..12154 34.3
12163..13587 38.3
Nucleotidyl transferase (PF00483)
GDP-mannose
pyrophosphorylase
E value = 6.9 Â eÀ127
E. coli (AAG41753)
Glycosyl transferase group 2
(PF00535)
Glycosyltransferase-like protein
Y. pseudotuberculosis (type O:1b)
(CAB63303.1)
Phosphomannomutase
S. boydii (AAL27334)
Glycosyltransferase
E value = 2.3 Â eÀ33
Phosphoglucomutase/phospho-
mannomutase (PF02879)
E value = 2.9 Â eÀ35
Phosphomannomutase

H. Zhu et al. / Carbohydrate Research 349 (2012) 86–89
89
gene cluster, Orf6 was proposed to be a glycosyltransferase gene
and named wfcY.
Therefore, the functions assigned tentatively to the genes in the
O-antigen gene cluster of E. coli O41 correspond well with the O-
antigen structure established in this work.
99.95% D2O. NMR spectra were recorded on a Bruker Avance II
600 spectrometer (Germany) at 30 °C using internal TSP (dH 0)
and acetone (dC 31.45) as references. 2D NMR spectra were ob-
tained using standard Bruker software, and Bruker TopSpin 2.1
program was used to acquire and process the NMR data. A mixing
time of 100 and 150 ms was used in TOCSY and ROESY experi-
ments, respectively.
1. Experimental
1.1. Bacterial strain and isolation of the lipopolysaccharide
1.5. Sequencing and analysis of genes
E. coli O41 type strain (laboratory stock number G3080) was
obtained from the Institute of Medical and Veterinary Science
(Adelaide, Australia). Bacteria were grown to late log phase in 8 L
of Luria–Bertani broth using a 10-L BIOSTAT C-10 fermentor (B.
Braun Biotech Int., Germany) under constant aeration at 37 °C
and pH 7.0. Bacterial cells were washed and dried as described.10
The lipopolysaccharide was isolated in a yield 6.1% from dried
cells by the phenol–water method,11 the crude extract was dia-
lyzed without separation of the layers and freed from nucleic acids
and proteins by treatment with 50% aq CCl3CO2H to pH 2 at 4 °C.
The supernatant was dialyzed and lyophilized.
Chromosomal DNA was prepared as described previously.13 The
primers #1523 and #1524 based on the housekeeping genes galF
and gnd,14 respectively, were used to amplify the O-antigen gene
cluster. The PCR cycles used were as follows: denaturation at
94 °C for 10 s, annealing at 60 °C for 30 s, and extension at 68 °C
for 15 min. A shotgun bank was constructed as described.15
Sequencing was carried out using an ABI 3730 automated DNA se-
quencer (Applied Biosystems, Foster City, CA), and sequence data
were analyzed using computer programs as described.16
Acknowledgments
1.2. Isolation of the O-polysaccharide
This work was supported by the Russian Foundation for Basic
Research (projects 11-04-91173-a and 11-04-01020-a), the Na-
tional Science Foundation of China (30900255 and 31111120026),
Tianjin Research Program of Application Foundation and Advanced
Technology (10JCYBJC10100).
Delipidation of the lipopolysaccharide (100 mg) was performed
with 2% aq HOAc at 100 °C until precipitation of lipid A. The precip-
itate was removed by centrifugation (13,000Âg, 20 min), and the
supernatant was fractionated by GPC on a column (56 Â 2.6 cm)
of Sephadex G-50 Superfine (Amersham Biosciences, Sweden) in
0.05 M pyridinium acetate buffer pH 5.5, monitored with a differ-
ential refractometer (Knauer, Germany). A high-molecular-mass
polysaccharide was obtained in a yield of 33% of the lipopolysac-
charide mass.
References
1. Reeves, P. P.; Wang, L. Curr. Top. Microbiol. Immunol. 2002, 264, 109–135.
2. Stenutz, R.; Weintraub, A.; Widmalm, G. FEMS Microbiol. Rev. 2006, 30, 382–
403.
1.3. Monosaccharide analysis
3. Jansson, P.-E.; Kenne, L.; Widmalm, G. Carbohydr. Res. 1989, 188, 169–191.
4. Alexander, D. C.; Valvano, M. A. J. Bacteriol. 1994, 176, 7079–7084.
5. Samuel, G.; Reeves, P. R. Carbohydr. Res. 2003, 338, 2503–2519.
6. Frick, D. N.; Townsend, B. D.; Bessman, M. J. J. Biol. Chem. 1995, 270, 24086–
24091.
A polysaccharide sample (0.5 mg) was hydrolyzed with 2 M
CF3CO2H (120 °C, 2 h). Monosaccharides were identified by GLC
of the alditol acetates on an Agilent 7820A chromatograph (USA)
equipped with an HP-5 column (0.32 mm  30 m) and a tempera-
ture program of 160 (1 min) to 290 °C at 7 °C minÀ1. The absolute
configurations of the monosaccharides were determined by GLC
of the acetylated (S)-2-octyl glycosides as described.12
7. Ginsburg, V. J. Biol. Chem. 1961, 236, 2389–2393.
8. Liu, D.; Cole, R. A.; Reeves, P. R. J. Bacteriol. 1996, 178, 2102–2107.
9. Daniels, C.; Vindurampulle, C.; Morona, R. Mol. Microbiol. 1998, 28, 1211–
1222.
10. Robbins, P. W.; Uchida, T. Biochemistry 1962, 1, 323–335.
11. Westphal, O.; Jann, K. Methods Carbohydr. Chem. 1965, 5, 83–91.
12. Leontein, K.; Lönngren, J. Methods Carbohydr. Chem. 1993, 9, 87–89.
13. Bastin, D. A.; Reeves, P. R. Gene 1995, 164, 17–23.
14. Feng, L.; Wang, W.; Tao, J.; Guo, H.; Krause, G.; Beutin, L.; Wang, L. J. Clin.
Microbiol. 2004, 42, 3799–3804.
1.4. NMR spectroscopy
15. Wang, L.; Reeves, P. R. Infect. Immun. 1998, 66, 3545–3551.
16. Feng, L.; Senchenkova, S. N.; Yang, J.; Shashkov, A. S.; Tao, J.; Guo, H.; Zhao, G.;
Knirel, Y. A.; Reeves, P.; Wang, L. J. Bacteriol. 2004, 182, 383–392.
A polysaccharide sample was deuterium-exchanged by freeze-
drying from 99.9% D2O and then examined as a solution in