characterization of escherichia coli o86 o-antigen gene cluster and identification of o86-specific...

www.elsevier.com/locate/vetmic

Veterinary Microbiology 106 (2005) 241–248

Abstract

Escherichia coli O86 belongs to the enteropathogenic E. coli (EPEC) group, some strains of which are pathogens of humans,

wild birds and farm animals. The O-antigen gene cluster of E. coli O86 was amplified by long-range PCR using primers based on

the housekeeping genes galF and gnd, and then sequenced. Genes involved in GDP-Fuc and N-acetyl-galactosamine (GalNAc)

synthesis and genes encoding glycosyltransferases, O-unit flippase and O-antigen polymerase were identified on the basis of

homology. By screening against 186 E. coli and Shigella-type strains, two genes specific to E. coli O86 were identified. A

polymerase chain reaction (PCR) assay, based on the specific O-antigen genes identified here, could be used for the rapid

detection of E. coli O86 in environmental and clinical samples. The relationship between E. coli O86 and O127 was also

determined by comparing the two O-antigen gene clusters.

# 2005 Elsevier B.V. All rights reserved.

Keywords: E. coli O86; PCR assay; Molecular typing; EPEC; E. coli O127

1. Introduction

Escherichia coli is a clonal species, including both

commensals and pathogens, which are normally

identified by the combination of their O- and H-

* Corresponding authors. Tel.: +86 22 6622 9592;

fax: +86 22 6622 9596.

E-mail addresses: [email protected] (L. Feng),

[email protected] (L. Wang).

0378-1135/$ – see front matter # 2005 Elsevier B.V. All rights reserved

doi:10.1016/j.vetmic.2004.12.021

(and sometimes K-) antigens. O-antigens are present

in all E. coli strains, and at least 166 O-antigen forms

have been recognized in E. coli, of which only some

are commonly found in pathogenic strains (Nataro and

Kaper, 1998).

Genes involved in O-antigen synthesis are nor-

mally clustered between two housekeeping genes,

galF and gnd, on the E. coli chromosome. They are

classified into three main classes: (i) genes for

synthesis of nucleotide sugar precursors, (ii) genes

Characterization of Escherichia coli O86 O-antigen gene

cluster and identification of O86-specific genes

Lu Fenga,b,*, Weiqing Hana, Quan Wanga, David A. Bastina,b,c, Lei Wanga,b,c,*

aTeda School of Biological Sciences and Biotechnology, Nankai University, Teda College, 23 HongDa Street,

Teda, Tianjin 300457, ChinabTianjin State Laboratory of Microbial Functional Genomics, TEDA College, Nankai University, 23 HongDa Street,

Teda, Tianjin 300457, ChinacTianjin Biochip Technology Corporation, 23 HongDa Street, Teda, Tianjin 300457, China

Received 4 July 2004; received in revised form 29 November 2004; accepted 4 December 2004

.

L. Feng et al. / Veterinary Microbiology 106 (2005) 241–248242

encoding glycosyltransferases for transfer of sugars

from their respective nucleotide sugar precursors to

build up the O-unit, and (iii) genes for carrying out

specific processing and assembly steps in conversion

of the O-unit to the O-antigen as part of the complete

lipopolysaccharide (LPS), including the flippase

(Wzx) and polymerase (Wzy) genes (Reeves and

Wang, 2002). The difference in O-antigen forms is

almost entirely due to genetic variations in their

respective O-antigen gene clusters. It has been

suggested that inter- and intra-species lateral transfer

of O-antigen genes play important roles in expanding

O-antigen polymorphic forms (Reeves and Wang,

2002). Horizontal gene transfer may permit the

transition of some E. coli strains from commensal

to pathogen (Lan and Reeves, 2002).

E. coli O86 belongs to the enteropathogenic E. coli

(EPEC) group (Lior, 1994). EPEC has been implicated

as a major cause of acute and persistent infantile

diarrhoea in developing countries (Levine, 1987).

Some E. coli O86 serotypes, such as E. coli O86:K61,

have been associated with mortality in wild birds

(Pennycott et al., 1998), and it has also been isolated

from diarrhoeic calves (Blanco et al., 1998), pigs

(Alexa et al., 1995) and horses (Holland et al., 1996).

The E. coli O86 O-antigen consists of pentasac-

charide O-units with three different sugars: L-fucose

(Fuc), D-galactose (Gal) and N-acetyl-galactosamine

(GalNAc) (Fig. 1).

In this study, the O-antigen gene cluster of E. coli

O86 was sequenced, and functions of the genes were

identified on the basis of homology. Two genes,

specific for E. coli O86, were identified by the PCR

screening of 186 E. coli and Shigella-type strains. The

O-antigen gene clusters of E. coli O86 and O127,

which is also an EPEC, were compared and shown to

be closely related.

Fig. 1. Structure of the E. coli O86 (Andersson et al., 1989) and O127

modification, but the position is not clear.

2. Materials and methods

2.1. Bacterial strains and plasmids

The E. coli O86 O-antigen-type strain G1275 and

three H-antigen-type strains G1394 (O86:H36),

G1397 (O86:H47), G1474 (O86:H34) were obtained

from the Institute of Medical and Veterinary Science

(IMVS), Adelaide, Australia. Other E. coli and

Shigella-type strains used are listed in Table 1. The

plasmid pGEM-T easy was purchased from Promega.

2.2. Construction of DNaseI shot gun bank

Chromosomal DNA was prepared as previously

described (Bastin and Reeves, 1995). Primers #1523

and #1524 (Wang et al., 2001), based on the galF and

gnd genes, respectively, were used to amplify the DNA

of E. coli O86 O-antigen gene cluster, using the

expand long-template PCR system from Roche. The

PCR cycles used were as follows: denaturation at

94 8C for 10 s, annealing at 60 8C for 30 s and

extension at 68 8C for 15 min. The PCR products were

digested with DnaseI, and the resulting DNA

fragments were cloned into pGEM-T easy to produce

a bank using the method described previously (Wang

and Reeves, 1998).

2.3. Sequencing and analysis

Sequencing was carried out using an ABI 3730

automated DNA sequencer from Applied Biosystems.

Sequence data were assembled using the Staden

package (Staden, 1996). The ARTEMIS program

(Rutherford et al., 2000) was used to identify open

reading frames (ORFs) and annotations. The BLOCK-

MAKER program was used to search conserved

(Widmalm and Leontein, 1993) O-antigen. O127 has an O-acetyl

L. Feng et al. / Veterinary Microbiology 106 (2005) 241–248 243

Table 1

E. coli and Shigella-type strains and PCR pools used for testing of E. coli O86-specific primers

Pool no. Strains of which chromosomal DNA included in the pool Source

1 E. coli-type strains for O serotypes IMVSa

1, 2, 5, 7, 12, 13, 14, 15, 16, 17, 19ab, 20, 21, 22, 23, 24, 59, 3, 11

2 E. coli-type strains for O serotypes IMVS

25, 26, 27, 28, 29, 30, 32, 31, 33, 35, 36, 37, 38, 40, 41, 42, 43, 39, 59


44, 45, 46, 48, 49, 50, 51, 52, 54, 55, 56, 57, 58, 60, 61, 62, 64, 73


63, 65, 66, 69, 70, 71, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 96, 95


84, 85, 86, 87, 88, 89, 91, 92, 98, 99, 101, 102, 103, 104, 105, 106, 100, 151


107, 108, 109, 110, 111, 112ab, 112ac, 113, 115, 116, 118, 120, 123, 125, 126, 128


129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145

8 E. coli-type strains for O serotypes IMVSb

146, 147, 148, 150, 152, 154, 156, 157, 158, 159, 160, 161, 163, 164, 165, 166

9 E. coli-type strains for O serotypes IMVSc

168, 169, 170, 171, 172, 173, 155, 124

And S. dysenteriae-type strains for O serotypes –d

D1, D2, D3, D4, D5, D6, D7, D8, D9, D10, D11, D12

10 S. boydii-type strains for O serotypes –d

B1, B2, B3, B4, B6, B7, B8, B9, B10, B11, B12, B13, B14, B15, B16, B17, B18

11 S. flexneri-type strains for O serotypes –d

F1a, F1b, F2a, F2b, F3, F4b, F5 (v: 4), F5 (v: 7), F6, FX variation, FY variation

And S. sonnei-type strains for O serotypes DS, DR

12 E. coli-type strains for O serotypes IMVSe

3, 11, 39, 59, 64, 73, 96, 95, 100, 114, 151, 167, 162, 121, 127, 149, 119

13 As pool 5, but lacks E. coli O86, used as control IMVS

a Institute of Medical and Veterinary Science, Adelaide, Australia.b O165 and O166 from Statens Serum Institute, Copenhagen, Denmark; the rest from IMVS.c O155 and O124 from IMVS; the rest from Statens Serum Institute, Copenhagen, Denmark.d Institute of Epidemiology and Microbiology, Chinese Academy of Preventive Medicine, Beijing, PR China.e O167 from Statens Serum Institute, Copenhagen, Denmark; the rest from IMVS.

motifs. BLAST and PSI-BLAST (Altschul et al., 1997) were

used to search databases, including GenBank, COG

and Pfam protein motif databases (Bateman et al.,

2002; Tatusov et al., 2001). The algorithm, described

by Eisenberg et al. (1984), was used to identify

potential transmembrane segments. Sequence align-

ment and comparisons were performed using the

CLUSTALW program. (http://www.ebi.ac.uk/clustalw).

2.4. Specificity assay by PCR

Chromosomal DNA, prepared from each of the 186

representative strains of all E. coli O-serotypes, was

examined for quality by PCR amplification of the mdh

gene (coding for malate dehydrogenase and present as

a housekeeping gene in E. coli) using primers

described previously (Wang and Reeves, 1998). A

total of 13 pools of DNA were made, each containing

DNA from 12 to 19 strains (Table 1). Pools were

screened using primers based on the E. coli O86-

specific genes wzx (wl-2378 (50-TTATGATGCTA-

CAGGTAAAACG-30)/wl-2379 (30-GCAGTTCAGG-

TACAATTATTG-50); wl-2380 (50-TTAACTGC-

TAACTACAGAGGGGT-30)/wl-2381 (30-GTAGGT-

TTGGAAATCTTATGGGTCC-50)) and wzy (wl-

2382 (50-GAGTTATTTTGGTTCACCCTT-30)/wl-

2383 (30-CGAGATAAGTATCCACCCGAT-50); wl-

2384 (50-CTCAACCCAATGACTTTCT-30)/wl-2385

(30-CCATATCTTCATTTATCGCCT-50)). The PCR

cycles used were as follows: denaturing at 95 8C for

15 s, annealing for 30 s (at 60 8C for the first three

pairs of primers and at 55 8C for the last pair of

genbank:AAC60777


primers) and extension at 72 8C for 1 min. PCR was

carried out in a total volume of 25 ml, of which 10 ml

was run on an agarose gel to check for amplified DNA.

2.5. Nucleotide sequence accession number

The DNA sequence of the E. coli O86 O-antigen

gene cluster has been deposited in GenBank under the

accession number AY670704.

3. Results

3.1. Sequencing

A sequence of 14 156 bases from galF (positions

1–765) to gnd (positions 13 628–14 156) was

obtained. In addition to galF and gnd, 12 open

reading frames were identified, all of which were

presumed to be transcribed from galF to gnd (Fig. 2),

as in other O-antigen clusters (Reeves and Wang,

2002). On the basis of their similarity to those from

available databases, all the ORFs were assigned

functions as summarised in Table 2.

3.2. O-antigen gene cluster of E. coli O86

The O-unit of E. coli O86 consists of five sugar

residues: two of N-acetyl-galactosamine, two of

galactose, and one of fucose (Fig. 1). Only genes

involved in the synthesis of GDP-Fuc and UDP-

GalNAc were expected in the O-antigen gene cluster.

Genes for the synthesis of UDP-Gal, which are also

Fig. 2. Comparison of O-antigen gene clusters of E. coli O86 (this study)

All the genes are presumed to be transcribed in a direction from galF to gnd

coli O86 and O127. In the E. coli O86 manB gene, there are two differe

involved in housekeeping functions, are located

elsewhere on the chromosome (Reeves and Wang,

2002). Sugar transferases are specific to sugar donors,

sugar acceptors and the linkages between them.

Therefore, five transferases are required for the

synthesis of E. coli O86 O-antigen. When a GalNAc

is used as the first sugar of an O-unit, as in the case of

E. coli O86, it is transferred by the product of the wecA

gene, which is located outside of the O-antigen gene

cluster (Alexander and Valvano, 1994). Therefore,

only four transferase genes were expected in the O-

antigen gene cluster of E. coli O86. Also expected here

are the O-unit flippase gene (wzx) and the O-antigen

polymerase gene (wzy), both of which are specific to

O-antigens.

3.2.1. Genes of sugar biosynthesis pathways

Proteins encoded by orf3, orf4, orf5, orf6 and orf7

shared 96, 99, 98, 58, and 93% identity to Gmd, Fcl,

Gmm, ManC and ManB, respectively, of the E. coli

O157 O-antigen gene cluster (Table 2). These are

enzymes of the GDP–L-Fuc biosynthesis pathway and

their functions have been biochemically examined

(Frick et al., 1995; Ginsburg, 1961). Therefore, orf3, 4,

5, 6 and 7 were confidently identified as encoding the

same enzymes for the biosynthesis of GDP-L-Fuc of

the E. coli O86 O-antigen and are named gmd, fcl,

gmm, manC and manB, respectively.

orf1 shared 57% identity or 72% similarity to gne

of Yersinia enteritica O:8. In Y. enteritica O:8

(Table 2), gne encodes UDP-GlcNAc C4 epimerase,

which catalyzes the conversion of GlcNAc to GalNAc

(Bengoechea et al., 2002). orf1 was assigned the same

and O127 (http://www.sanger.ac.uk/Projects/Escherichia_Shigella/).

. Dotted lines align homologous genes or DNA segments between E.

nt levels of sequence identity with O127.

genbank:CAA87705

genbank:AAC60777


Table 2

Putative genes in E. coli O86 O-antigen gene cluster

Gene

name

Location in

sequence

G + C

content (%)

Similar protein(s), strain(s)

(Genbank accession no.)

% Identical aa/%

similar aa

(no. of aa overlap)

Putative function of protein

gne 1075–2094 33.2 Gne, Yersinia enterocolitica

(type 0:8) (AAC60777)

57/72 (335) UDP-N-acetylglucosamine-4-

epimerase

wcmA 2113–3129 34.2 wbcQ, Yersinia enterocolitica

(CAA87705)

47/68 (337) Glycosyltransferase

gmd 3156–4277 52.5 GDP-D-mannose dehydratase,

Escherichia coli O157:H7

(AAG57113)

96/98 (373) GDP-D-mannose dehydratase

fcl 4280–5245 56.2 nucleotide di-P-sugar epimerase

or dehydratase, Escherichia coli

O157:H7 (AAG57112)

99/100 (321) GDP-L-fucose synthetase

gmm 5245–5748 52.2 GDP-mannose mannosyl hydrolase,


(AAG57111)

98/98 (149) GDP-mannose

mannosyl hydrolase

manC 5741–7189 36.4 GDP-mannose pyrophosphorylase,


(AAG57109)

58/77 (478) GDP-mannose

pyrophosphorylase

manB 7193–8581 50.5 Phosphomannomutase,


(AAG57108)

93/95 (455) Phosphomannomutase

wzx 8601–9803 27.8 flippase, Escherichia coli

(BAA77727)

21/42 (263) O-unit flippase

wcmB 9806–10510 28.2 Galactosylaminyltransferase,

Homo sapiens (AAL37338)

25/40 (204) Glycosyltransferase

wzy 10491–11831 28.6 O-antigen polymerase,

Shigella boydii (AAL27318)

21/42 (414) O-antigen polymerase

wcmC 11843–12586 31.3 WbsK, Escherichia coli (AAO37699) 38/60 (212) Glycosyl transferase

wcmD 12598–13506 32.3 WbsJ, Escherichia coli (AAO37698) 35/56 (295) Glycosyl transferase

function in the synthesis of GalNAc of E. coli O86 O-

antigen, and is named gne. The substrate of Gne, UDP-

GlcNAc, is a common precursor of a variety of surface

carbohydrates in bacteria, which is synthesized by

genes located elsewhere in the chromosome.

3.2.2. Genes encoding sugar transferases

The protein encoded by orf9 belongs to the

glycosyltransferase family 6 (PF03414 E-value =

1.4 � e�3). It also shared 24% identity or 44%

similarity with a galactosylaminyltransferase of Homo

sapiens (Table 2), which is responsible for the

formation of a-D-Gal-(1 ! 3)-b-D-Gal linkage

(Yamamoto et al., 1990). The same linkage is also

present in the O-antigen of E. coli O86. Therefore,

orf9 is proposed to encode a glycosyltransferase for

the a-D-Gal-(1 ! 3)-b-D-Gal linkage in E. coli O86,

and is named wcmB.




similarity with WbsJ of E. coli O128 (Table 2), which

is a sugar transferase responsible for the formation of

a-L-Fuc-(1 ! 2)-b-D-Gal linkage (Shao et al., 2003).

The same linkage is also present in the O-antigen of E.

coli O86. orf12 is proposed to encode a fucose

transferase for the same linkage in E. coli O86, and is

named wcmD.




similarity with WbsK of E. coli O128 (Table 2), which

is a putative glycosyltransferase (Shao et al., 2003).

The b-D-Gal-(1 ! 3)-a-D-GalNAc and a-L-Fuc-

(1 ! 2)-b-D-Gal linkages are the two identical

linkages present in both O-antigens of E. coli O86

genbank:AAG57113

genbank:AAG57112

genbank:AAG57111

genbank:AAC60777

genbank:CAA87705

genbank:AAG57113

genbank:AAG57112

genbank:AAG57111

genbank:PF00534

genbank:AAK81005

genbank:AAO68867

http://www.sanger.ac.uk/Projects/Escherichia_Shigella/

genbank:AAL27318

genbank:AAO37699

genbank:AAO37698


and E. coli O128. As already described, orf12 is

proposed to encode a glycosyltransferase for the a-L-

Fuc-(1 ! 2)-b-D-Gal linkages for both strains, while

orf11 is proposed to encode a glycosyltransferase for

the b-D-Gal-(1 ! 3)-a-D-GalNAc linkage in both E.

coli O86 and O128, and is named wcmC.


glycosyltransferase family 1 (PF00534, E-value =


similarity with WbcQ of Yersinia enterocolitica O:3

(Table 2), which is a putative glycosyltransferase

involved in lipopolysaccharide core synthesis (Skur-

nik et al., 1995). However, by direct comparison of the

O-antigen gene clusters of E. coli O86 and O127

(described below), a homologue of orf2 was found.

orf2 is proposed to be the glycosyltransferase

responsible for the a-D-GalNAc-(1 ! 3)-b-D-GalNAc

linkage and is named wcmD.

3.2.3. Genes for O-unit processing

The protein encoded by orf8 was found to have 12

predicted transmembrane segments, which is a feature

of Wzx proteins. It also shared 21% identity to the

putative Wzx proteins of E. coli O157 (Table 2) and

Clostridium acetobutylicum ATCC 824 (AAK81005).

When Orf8 and the two putative Wzx proteins were

analyzed using the BLOCKMAKER program, five con-

served motifs were revealed (36, 24, 9, 32, and 22

amino acids, respectively). The consensus sequences

of those motifs were used to run the PSI-BLAST program

to search the Genpept database, and other distantly

related Wzx proteins were retrieved (E-value =

2 � e�22) after three iterations. orf8 was, therefore,

proposed to be an O-unit flippase gene, and is named

accordingly.

The protein encoded by orf10 was found to have 11

predicted transmembrane segments with a large

periplasmic loop of 75 amino acid residues, which

is the typical topological character of Wzy proteins. It

also shared 21 and 20% identity to Wzy proteins of

Salmonella boydii O4 (Table 2) and Salmonella

enterica Typhi Ty2 (AAO68867), respectively. When

Orf10 and the two Wzy proteins were analyzed using

the BLOCKMAKER program, five conserved motifs were

revealed (23, 9, 14, 31 and 12 amino acids,

respectively). The consensus sequences of those

motifs were used to run the PSI-BLAST program to

search the Genpept database, and other distantly

related Wzy proteins were also retrieved (E-value =

2 � e�27) after three iterations. Therefore, orf10 is

proposed to be an O-antigen polymerase gene and is

named accordingly.

3.3. Identification of E. coli O86-specific genes

A total of four pairs of primers based on O-unit

processing genes wzx and wzy (two pairs for each

gene) were designed and used to screen DNA pools

containing representatives of the 186 known O-

antigen forms of E. coli and Shigella strains. With

either primer pair, except for the pools containing E.

coli O86, which gave the expected PCR products, no

PCR products were detected. The four primer pairs

were also tested on three E. coli O86 H-antigen-type

strains O86:H36 (G1394), O86:H47 (G1397) and

O86:H34 (G1474), and the expected PCR products

with correct size were also detected in all three strains

tested (data not shown). Therefore, all of the four

primer pairs are specific to E. coli O86 and can be used

for the development of a PCR assay for the

identification and detection of E. coli O86 strains.

3.4. Comparison of O-antigen gene clusters

between E. coli O86 and O127

From the genome sequence data of E. coli

O127 (E2348/69) (http://www.sanger.ac.uk/Projects/

Escherichia_Shigella/), whose O-antigen structure is

similar to that of E. coli O86 (Fig. 1), we retrieved the

sequence of O-antigen gene cluster and analyzed it.

When the O-antigen gene clusters of E. coli O86 and

O127 were compared, it was found that the first six

genes downstream of galF (ORF1–6) shared DNA

identity between 96 and 99% or protein identity

between 98 and 100%. The first 1130 bases of orf7

(manB) also shared 94% DNA identity or 99% protein

identity, but the remaining 259 bases shared only 58%

DNA identity or 56% protein identity between the two

strains. The remaining five genes upstream of gnd

(ORF8–12), which encode Wzx, O-acetyl transferase,

Wzy, glycosyltransferase, and glycosyltransferase in

E. coli O127, respectively (data not shown), also

shared a DNA identity of between 47 and 66%. In

addition, the non-coding region upstream of gnd

shared 56% DNA identity, and the first 528 bases of

gnd shared 97% DNA identity. The first 1130 bases of

genbank:PF00534

genbank:AAK81005

genbank:AAO68867




orf7 (manB) seem to be one junction of a homologous

recombination.

Among the first seven genes, orf2, encoding a

glycosyltransferase, shared 99% DNA identity or

100% protein identity between the two strains. The a-

D-GalNAc-(1 ! 3)-b-D-GalNAc linkage is the only

identical linkage in the O-antigens of the two strains

for which the respective glycosyltransferase gene had

not been proposed by previous homology search.

Therefore, orf2 was proposed to be responsible for the

a-D-GalNAc-(1 ! 3)-b-D-GalNAc linkage in both E.

coli O86 and O127.

4. Discussion

Characteristically, O-antigen gene clusters of E.

coli are located between the galF and gnd genes on the

chromosome (Reeves and Wang, 2002). This is also

the case for E. coli O86. All of the genes in the O-

antigen gene cluster were identified or confidently

proposed. Characterization of the E. coli O86 O-

antigen gene cluster showed good agreement to the O-

antigen structure.

Glycosyltransferases, O-unit flippase and O-anti-

gen polymerase usually display a low level of

similarity in terms of their primary sequences (Wang

et al., 1998). Both Wzx and Wzy are membrane

proteins, Wzx proteins characteristically containing

12 transmembrane segments (Liu et al., 1996), and

Wzy proteins containing 10–13 transmembrane

segments and a large periplasmic loop (Daniels

et al., 1998). These characteristics of Wzx and Wzy

proteins, plus the conserved motif search, aid their

identification in E. coli O86. From both sequence and

structure information, we have confidently proposed

the functions for all of the four glycosyltransferase

genes in the E. coli O86 O-antigen gene cluster,

providing a good starting point for future biochemical

studies of these genes.

Two specific genes to E. coli O86 were identified

and showed high specificity against all 186 E. coli and

Shigella-type strains. The developed PCR assay, based

on the specific genes, can be potentially used for

identification and detection of E. coli O86 cells. This

may be of particular importance, since some E. coli

O86 strains, such as O86:K61, are pathogens for

humans and animals (La Ragione et al., 2002). In

contrast to traditional serological testing, which is

slow, labor intensive and may be impeded by cross-

reactions, PCR-based methods prove to be rapid, cost

saving, accurate, and O-serotype-specific. PCR assays

have been developed in many pathogenic E. coli

strains, such as E. coli O157 and O111 (Wang and

Reeves, 1998; Wang et al., 1998). However, specific

genes, based on H- and K-antigen genes, are needed,

in combination with O-specific genes, to distinguish

between different O86 strains.

There are only two differences in the O-antigen

between E. coli O86 and O127. One of them, the

b(1 ! 4) and a(1 ! 2) linkage between O-units of

O86 and O127, respectively, is due to the O-unit

polymerase, Wzy. The other, the presence of the

branched a-D-Gal-(1 ! 3)-b-D-Gal linkage and O-

acetyl group of E. coli O86 and O127, respectively, is

due to the Orf9 proteins.

A close evolutionary relationship is present

between the O-antigen gene clusters of E. coli O86

and O127. The regions of vastly different levels of

DNA similarity suggested that transfer of O-antigen

genes, by homologous recombination, occurred. We

propose, therefore, that one of the O-antigen gene

clusters of E. coli O86 and O127 evolved from another

by gaining O-antigen genes from other strains through

homologous recombination. However, we are unable

to indicate the origin of the genes at this stage.

Although we proposed that orf11and orf12 of E. coli

O86 and O127 are responsible for the two identical

linkages, they had lower identity than orf2. Perhaps,

the reason is that these two glycosyltransferase genes

evolved from different origins to orf2. The O-antigen

present in a strain can change by recombination

involving the DNA flanking the locus (Reeves and

Wang, 2002). In a clonal species, like E. coli, this has

the effect of a strain gaining a new O-antigen by lateral

transfer from another clone within the species, as in

the case of the E. coli O157:H7 and O55:H7 (Reeves

and Wang, 2002). Our results provide further evidence

for the movement of O-antigen genes.

Acknowledgements

This work was supported by the Chinese National

Science Fund for Distinguished Young Scholars

(30125001), NSFC General Program (30270029),


the 863 Program (2002AA2Z2051), and funding from

the Science and Technology Committee of Tianjin

City (013181711) to LW.

References

Alexa, P., Salajka, E., Hamrik, J., Nejezchleb, A., 1995. Pathogenic

strains of Escherichia coli in weaned piglets in the Czech

Republic. Pig News Inform. 16, 81N–84N.

Alexander, D.C., Valvano, M.A., 1994. Role of the rfe gene in the

biosynthesis of the Escherichia coli O7-specific lipopolysac-

charide and other O-specific polysaccharides containing N-

acetylglucosamine. J. Bacteriol. 176, 7079–7084.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z.,

Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a

new generation of protein database search programs. Nucl.

Acids Res. 25, 3398–3402.

Andersson, M., Carlin, N., Leontein, K., Lindquist, U., Slettengren,

K., 1989. Structural studies of the O-antigenic polysaccharide of

Escherichia coli O86, which possesses blood-group B activity.

Carbohydr. Res. 185, 211–223.

Bastin, D.A., Reeves, P.R., 1995. Sequence and analysis of the O-

antigen gene (rfb) cluster of Escherichia coli O111. Gene 164,

17–23.

Bateman, A., Birney, E., Cerruti, L., Durbin, R., Etwiller, L., Eddy,

S.R., Griffiths-Jones, S., Howe, K.L., Marshall, M., Sonnham-

mer, E.L., 2002. The Pfam protein families database. Nucl.

Acids Res. 30, 276–280.

Bengoechea, J.A., Pinta, E., Salminen, T., Oertelt, C., Holst, O.,

Radziejewska-Lebrecht, J., Piotrowska-Seget, Z., Venho, R.,

Skurnik, M., 2002. Functional characterization of Gne (UDP-

N-acetylglucosamine-4-epimerase), Wzz (chain length determi-

nant), and Wzy (O-antigen polymerase) of Yersinia enterocoli-

tica serotype O:8. J. Bacteriol. 184, 4277–4287.

Blanco, M., Blanco, J.E., Mora, A., Blanco, J., 1998. Distribution

and characterization of faecal necrotoxigenic Escherichia coli

CNF1+ and CNF2+ isolated from healthy cows and calves. Vet.

Microbiol. 59, 183–192.

Daniels, C., Vindurampulle, C., Morona, R., 1998. Overexpression

and topology of the Shigella flexneri O-antigen polymerase (Rfc/

Wzy). Mol. Microbiol. 28, 1211–1222.

Eisenberg, D., Schwarz, E., Komaromy, M., Wall, R., 1984. Ana-

lysis of membrane and surface protein sequences with the

hydrophobic moment plot. J. Mol. Biol. 179, 125–142.

Frick, D.N., Townsend, B.D., Bessman, M.J., 1995. A novel GDP-

mannose mannosyl hydrolase shares homology with the MutT

family of enzymes. J. Biol. Chem. 270, 24086–24091.

Ginsburg, V., 1961. Studies of the biosynthesis of guanosine dipho-

sphate L-fucose. J. Biol. Chem. 236, 2389–2393.

Holland, R.E., Schmidt, A., Sriranganathan, N., Grimes, S.D.,

Wilson, R.A., Brown, C.M., Walker, R.D., 1996. Characteriza-

tion of Escherichia coli isolated from foals. Vet. Microbiol. 48,

243–255.

La Ragione, R.M., McLaren, I.M., Foster, G., Cooley, W.A., Wood-

ward, M.J., 2002. Phenotypic and genotypic characterization of

avian Escherichia coli O86:K61 isolates possessing a

gamma-like intimin. Appl. Environ. Microbiol. 68, 4932–

4942.

Lan, R., Reeves, P.R., 2002. Escherichia coli in disguise: molecular

origins of Shigella. Microbes Infect. 4, 1125–1132.

Levine, M.M., 1987. Escherichia coli that cause diarrhea: enter-

otoxigenic, enteropathogenic, enteroinvasive, enterohemorrha-

gic and enteroadherent. J. Infect. Dis. 155, 377–389.

Lior, H., 1994. Classification of Escherichia coli. In: Gyles, C.L.

(Ed.), Escherichia coli in Domestic Animals and Humans. CAB

International, Wallingford, UK, pp. 31–72.

Liu, D., Cole, R., Reeves, P.R., 1996. An O-antigen processing

function for Wzx (RfbX): a promising candidate for O-unit

flippase. J. Bacteriol. 178, 2102–2107.

Nataro, J.P., Kaper, J.B., 1998. Diarrheagenic Escherichia coli. Clin.

Microbiol. Rev. 11, 142–201.

Pennycott, T.W., Ross, H.M., McLaren, I.M., Park, A., Hopkins,

G.F., Foster, G., 1998. Causes of death of wild birds of the family

Fringillidae in Britain. Vet. Rec. 143, 155–158.

Reeves, P.R., Wang, L., 2002. Genomic organization of LPS-specific

loci. Curr. Top. Microbiol. Immunol. 264, 109–135.

Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajan-

dream, M.A., Barrell, B., 2000. Artemis: sequence visualisation

and annotation. Bioinformatics 16, 944–945.

Shao, J., Li, M., Jia, Q., Lu, Y., Wang, P.G., 2003. Sequence of

Escherichia coli O128 antigen biosynthesis cluster and func-

tional identification of an alpha-1,2-fucosyltransferase. FEBS

Lett. 553, 99–103.

Skurnik, M., Venho, R., Toivanen, P., Alhendy, A., 1995. A novel

locus of Yersinia enterocolitica serotype O-3 involved in lipo-

polysaccharide outer core biosynthesis. Mol. Microbiol. 17,

575–594.

Staden, R., 1996. The Staden sequence analysis package. Mol.

Biotechnol. 5, 233–241.

Tatusov, R.L., Natale, D.A., Garkavtsev, I.V., Tatusova, T.A., Shan-

kavaram, U.T., Rao, B.S., Kiryutin, B., Galperin, M.Y., Fedor-

ova, N.D., Koonin, E.V., 2001. The COG database: new

developments in phylogenetic classification of proteins from

complete genomes. Nucl. Acids Res. 29, 22–28.

Wang, L., Curd, H., Qu, W., Reeves, P.R., 1998. Sequencing

of Escherichia coli O111 O-antigen gene cluster and identi-

fication of O111-specific genes. J. Clin. Microbiol. 36, 3182–

3187.

Wang, L., Qu, W., Reeves, P.R., 2001. Sequence analysis of four

Shigella boydii O-antigen loci: implication for Escherichia coli

and Shigella relationships. Infect. Immun. 69, 6923–6930.

Wang, L., Reeves, P.R., 1998. Organization of Escherichia coli

O157 O-antigen gene cluster and identification of its specific

genes. Infect. Immun. 66, 3545–3551.

Widmalm, G., Leontein, K., 1993. Structural studies of the Escher-

ichia coli O127 O-antigen polysaccharide. Carbohydr. Res. 247,

255–262.

Yamamoto, F., Marken, J., Tsuji, T., White, T., Clausen, H., Hako-

mori, S., 1990. Cloning and characterization of DNA comple-

mentary to human UDP-GalNAc: Fuc alpha 1–2Gal alpha 1–

3GalNAc transferase (histo-blood group A transferase) mRNA.

J. Biol. Chem. 265, 1146–1151.

characterization of escherichia coli o86 o-antigen gene cluster and identification of o86-specific...

Documents