Molecular Biology Branch, Center for Food
Safety and Applied Nutrition, Food and Drug Administration,
Washington, D.C. 20204
 |
TEXT |
How Escherichia coli
O157:H7 evolved is a subject whose interest spans clinical medicine,
food safety research, and evolutionary biology. This enterohemorrhagic
E. coli strain was first recognized as a human pathogen in
1982 (22). It has since risen to prominence as a cause of
major outbreaks of food-related disease worldwide. In a recent example,
a single outbreak in Japan in 1996 (23) accounted for over
5,000 illnesses and six deaths. In the United States, E. coli O157:H7 is responsible for an estimated 20,000 cases of
food-borne disease annually (7). The recognition that beef
products were sources of E. coli O157:H7 contamination
(22) and the identification of healthy dairy cattle as one
reservoir for the organism (15) implicated a cattle-to-human
zoonosis. Clonal analysis supports the thesis that O157:H7 evolved
recently from a lineage of E. coli that lived as a commensal
organism in animals (25). This versatile enteric organism
has adapted to other environments as well, as it is known to survive
conditions of low pH (1) or high salt and temperature
(12) that formerly safeguarded the food supply from E. coli contamination.
The availability of the entire nucleotide sequence of E. coli MG1655 (4), a representative of
laboratory-attenuated E. coli K-12 strains, makes
possible a formal comparison with E. coli O157:H7 sequences
as they become available. The quest of a comparative genomic approach
is to identify the types and sources of genetic variability and to
delimit unique sequences that contribute to important phenotypes within
a particular organism. Such information should be useful in
understanding the pathogenicity of O157:H7 and its ability to adapt to
unconventional environments. Moreover, evolutionary artifacts, the
remnants of sequences left within the genome from past genetic
transgressions, provide valuable insights into the genesis and
evolution of E. coli O157:H7.
E. coli O157:H7 mutS-rpoS intergenic
region.
In earlier work (14), long PCR analysis of the
intergenic region between the mutS and rpoS genes
of E. coli O157:H7 isolate EC536 indicated an apparent
deletion of 3.4 kb relative to the expected 6.9-kb intergenic region
found in the K-12 strain W3110. To identify the nature of the shortened
intergenic region in O157:H7, nucleotide sequencing was performed on
PCR products made from primers that spanned the region from the 3' end
of the mutS gene (5'-TGCATCTCGATGCACTGGAG) to the
middle of the rpoS gene (5'-CTCAACATACGCAACCTGG) in EC536. The results showed a mutS-rpoS intergenic
region of 3,737 bp, of which 2,930 bp had no apparent similarity to the K-12 sequence and replaced 6,098 bp of the K-12 intergenic region. The
apparent replacement extended from position 32719 to position 38818 in
the 61- to 62-min region of the K-12 chromosome (GenBank accession no.
U29579). The results are diagrammed in Fig. 1A and
B and show that the deletion from K-12
removes six of the seven open reading frames (ORFs) between the
mutS and rpoS genes of the K-12 sequence. The
mutS-proximal ORF, which remains intact in O157:H7, shows
95.0% similarity in encoded amino acid sequence and 96.4% identity at
the nucleotide level with the recently identified PrpB phosphatase gene
sequence (17) (GenBank accession no. U51682, identified as
pphB in GenBank accession no. AE000357). Of particular interest, the position of a prpB termination codon in the
K-12 sequence is conserved, but the TAG sequence is altered to TAA at
the beginning of the novel O157:H7 sequence. The
rpoS-proximal end of the deletion encompasses a hairpin
structure characteristic of a termination site for transcription of the
rpoS gene (9), whose polarity is opposite that of
the mutS and prpB genes.

View larger version (11K):
[in this window]
[in a new window]
|
FIG. 1.
(A) Genomic map of the mutS-rpoS region of
the E. coli K-12 chromosome based on a published sequence
(4) (GenBank accession no. U29579). Seven ORFs between the
mutS and rpoS genes are indicated by gray boxes,
the first of which has been identified as encoding PrpB phosphatase
(17). (B) In E. coli O157:H7, a novel sequence of
2,930 bp replaces six of the seven ORFs found in the E. coli
K-12 chromosome and abuts the prpB gene. (C) In S. dysenteriae type 1, a 768-bp IS1 element replaces 713 bp of the prpB gene and abuts the 2,930-bp novel
sequence.
|
|
In order to determine if the shortened intergenic region is a common
feature of the E. coli O157:H7 serotype and to establish how
widely this sequence is distributed among pathogenic strains, we used
three types of analyses to characterize several independent isolates of
O157:H7 outbreak strains, other E. coli strains, and a
collection of diverse enteric bacteria. First, colony hybridization experiments (6) were carried out with oligonucleotide probes specific for a site either within the 2,930-bp sequence of O157:H7 (5'-GACATATTCGGCAACTGCAC) or at the rpoS-proximal
border of the region (5'-GGCCTTTTTCTTTTGTTTGGG), which
contains both the novel O157:H7 sequence and a sequence like that
in K-12. Probe-positive strains were then assessed by sizing PCR
amplification products from the mutS-rpoS intergenic region.
Finally, these PCR products were used for limited DNA sequencing around
the border regions of the novel sequence. The results of the colony
hybridization experiments, summarized in Table
1, confirmed that all O157:H7 isolates
tested, including an O157:H
strain, contained the novel
sequence; these encompassed an array of early (1983) to recent (1995)
isolates of E. coli O157:H7 outbreak strains. We also
identified the novel sequence in strains from two reference collections
assembled to represent the genetic diversity of E. coli in
nature: diarrheagenic DEC5 strains of the E. coli O55:H7
serotype (the closest known siblings to O157:H7) (25) and
strains ECOR37 and ECOR42 from the E. coli reference
collection (20), two ECOR strains that were found to be
clustered with O157:H7 by the criteria of genetic distance measured by
multilocus enzyme electrophoresis and similarity of malate
dehydrogenase sequences (21). PCR products from the
mutS-rpoS intergenic regions of the ECOR37, ECOR42, and DEC5
strains were the same size as those from O157:H7 isolates, and the
nucleotide sequences at border regions around the novel sequence were
identical to those in the EC536 isolate. Negative colony hybridization
results were obtained when other E. coli strains of the
enterohemorrhagic disease class, and other classes, were tested. A
diverse group of enteric pathogens also yielded negative colony
hybridization results, with the notable exception of strains of
Shigella dysenteriae type 1, which were probe positive for
oligonucleotides specific to both internal and rpoS-proximal
border regions of the novel sequence. Probe and PCR analyses of
non-type 1 S. dysenteriae strains and isolates of
Shigella boydii, Shigella flexneri, and
Shigella sonnei showed much larger mutS-rpoS
intergenic regions than that found in E. coli O157:H7,
ranging from 8 to 12 kb, and the absence of the novel sequence element
(data not shown).
S. dysenteriae mutS-rpoS intergenic region.
When
filters for colony hybridization were probed with an
oligonucleotide (5'-CGGCCTCATTACTTTATTTTAT) encompassing the
sequence around the mutS-proximal border of the novel
sequence found in EC536, O157:H7 isolates were probe positive while
S. dysenteriae type 1 isolates were probe negative (Table
1). We therefore sequenced the mutS-rpoS intergenic region
with PCR product prepared from a strain of S. dysenteriae
type 1, SD567. PCR product amplified from this region of SD567 was
detectably larger on agarose gels than product from EC536. A summary of
the nucleotide sequence information is diagrammed in Fig. 1B and C,
which illustrate the genomic structures of the region in the two
strains. The results showed the novel sequence of 2,930 bp in S. dysenteriae to be 99.5% similar to that in O157:H7 and also
showed an identical flanking sequence in the direction of the
rpoS gene. On the mutS-proximal side of the novel
sequence, 768 bp of unique S. dysenteriae sequence was found
in place of 713 bp in K-12 and O157:H7; this sequence element showed
99.3% identity with the mobile insertion sequence IS1 from
S. dysenteriae (GenBank accession no. J01731), of which
S. dysenteriae contains multiple copies (18). The
IS1 element replaces the entire prpB gene
sequence, in effect making S. dysenteriae a null mutant of
prpB.
Figure 2 shows sequence comparisons at
the borders of elements in the region between the mutS and
rpoS genes in E. coli K-12, E. coli
O157:H7, and S. dysenteriae. A hypothetical crossover between the K-12 and S. dysenteriae sequences, which could
give rise to the novel sequence abutting the prpB gene of
O157:H7, is illustrated. Sequences that are identical to that in K-12
at the endpoint of the novel sequence (Fig. 2, underlined) and
distinctive nucleotide substitutions downstream from the
mutS gene, present in both O157:H7 and S. dysenteriae, suggest that multiple crossovers have occurred in the
region. We have examined the surrounding sequence for companion
IS1 elements indicative of a specific translocation event;
no other elements were revealed by short and long PCR analyses of over
20 kb to either side of the mutS-rpoS intergenic region in
strains SD567 and EC536.

View larger version (12K):
[in this window]
[in a new window]
|
FIG. 2.
Nucleotide sequences at border regions of elements in
the intergenic region between the mutS and rpoS
genes in E. coli K-12 (position no. 32708 to 32731, GenBank
accession no. U29579), E. coli O157:H7 (EC536), and S. dysenteriae type 1 (SD567). Elements are boxed, and sequences are
shown for the 3' end of the prpB genes in E. coli
K-12 and O157:H7, the 5' and 3' ends of the IS1 element in
S. dysenteriae, and the border regions of the novel sequence
in E. coli O157:H7 and S. dysenteriae.
Single-base-pair changes from the K-12 sequence in the
mutS-proximal region are underlined. In
rpoS-proximal sequences, the dyad symmetry of the
rpoS transcription terminator in K-12 is represented and the
sequence at the endpoint of the novel sequence (AGAAAA) that
is identical to that in K-12 is underlined.
|
|
Conclusions.
The finding that the lineage of a unique segment
of the E. coli O157:H7 chromosome can be traced to
S. dysenteriae implicates a role for horizontal DNA
transfer in the evolution of the O157:H7 chromosome. This conclusion is
underscored by the presence of a mobile insertion element
(IS1) in S. dysenteriae that abuts the same
sequence as that identified in O157:H7. The occurrence of the
transposable element offers, in part, a mechanism for genetic exchange
in an ancestral cell of the O157:H7 lineage. We infer that the DNA
exchange occurred between an E. coli O157:H7 ancestor and
S. dysenteriae, or a S. dysenteriae-like
organism, because the O157:H7 mutS-rpoS intergenic region
comprises a sequence identical to that found associated with
IS1 in S. dysenteriae (2,930 bp) adjoined to an
E. coli K-12-like prpB sequence (713 bp) not
present in S. dysenteriae (Fig. 1). The O157:H7 sequence
would seem to be the product of a precise crossover, making independent
acquisition of the novel sequence in S. dysenteriae and
O157:H7 unlikely. Since Shigella species are thought to have
evolved from E. coli relatively recently (19), we
surmise that the yet more recent O157:H7 strain acquired the
mutS-rpoS chromosomal segment from an ancestral lineage that
involved horizontal transfer from S. dysenteriae.
It is striking that other gene similarities indicate a flow of genetic
information from S. dysenteriae to O157:H7. The O157:H7 genes for shiga-like toxins (stx) show close genetic
identity with the shiga toxin of S. dysenteriae (5, 8,
10, 11); they are prophage encoded, indicating the likely mode of
DNA transfer from one organism to the other by bacteriophage. Another
link is suggested by DNA homology between genes in the two organisms for heme iron transport, genes that are absent from other
Shigella species and laboratory strains of E. coli but that show 99.5% sequence identity, namely
shuA (S. dysenteriae) and chuA
(E. coli O157:H7) (24). As these examples
accumulate, the extent to which the evolution of the O157:H7 chromosome
involved blending with the S. dysenteriae genome becomes an
intriguing question. But we also note evidence that a perosamine
synthetase specifying part of the side chain for O antigen from O157
strains is derived from a Vibrio cholerae-like organism
(2). The results of genomic sequencing of O157:H7 portray an
E. coli genome interspersed with sequences not found in
K-12, creating a genome 20% larger than that of K-12 (3).
These lines of evidence may be indicative of a more general promiscuous
behavior during the evolution of the O157:H7 chromosome that accounts
for its chimeric structure.
The mutS-rpoS intergenic sequences characterized here occupy
3.7 kb in E. coli O157:H7 and S. dysenteriae type
1 strains, in contrast to regions with 6.9 kb in E. coli
K-12 (4) and 8 to 12 kb in natural isolates of E. coli and Shigella species. In Salmonella
typhimurium, an extensive region with 12.6 kb comprises multiple
sequence elements (13), yet the neighboring mutS
and rpoS genes show respective sequence identities of 83 and
91% in S. typhimurium and E. coli K-12,
typical of homologous genes in the two genera. Such local sequence
variation earmarks the mutS-rpoS intergenic region as
another "bastion of polymorphism" (16), which is
indicative of horizontal transfer events. This serves notice that, akin
to mutational hot spots, there exist regions of the chromosome that may
be hot or cold for recombination.
Nucleotide sequence accession numbers.
GenBank accession no.
AF054420 and AF055472 have been assigned to the sequences for the
mutS-rpoS intergenic regions of E. coli O157:H7
and S. dysenteriae type 1, respectively.
We thank F. R. Blattner for discussing results prior to
publication, Philip Tarr and Tom Whittam for generously providing clinical isolates of E. coli O157:H7 and O55:H7 strains, and
Howard Ochman for the ECOR collection.
| 1.
|
Besser, R. E.,
S. M. Lett,
J. T. Weber,
M. P. Doyle,
T. J. Barrett,
J. G. Wells, and P. M. Griffin.
1993.
An outbreak of diarrhea and hemolytic uremic syndrome from Escherichia coli O157:H7 in fresh-pressed apple cider.
JAMA
269:2217-2220[Abstract].
|
| 2.
|
Bilge, S. S.,
J. C. Vary, Jr.,
S. F. Dowell, and P. I. Tarr.
1996.
Role of the Escherichia coli O157:H7 O side chain in adherence and analysis of an rfb locus.
Infect. Immun.
64:4795-4801[Abstract].
|
| 3.
| Blattner, F. R. Personal communication.
|
| 4.
|
Blattner, F. R.,
G. Plunkett III,
C. A. Bloch,
N. T. Perna,
V. Burland,
M. Riley,
J. Collado-Vides,
J. D. Glasner,
C. K. Rode,
G. Mayhew,
J. Gregor,
N. W. Davis,
H. A. Kirkpatrick,
M. A. Goeden,
D. Rose,
B. Mau, and Y. Shao.
1997.
The complete genome sequence of Escherichia coli K-12.
Science
277:1453-1462[Abstract/Free Full Text].
|
| 5.
|
Calderwood, S. B.,
F. Auclair,
A. Donohue-Rolfe,
G. T. Keusch, and J. J. Mekalanos.
1987.
Nucleotide sequence of the shiga-like toxin genes of Escherichia coli.
Proc. Natl. Acad. Sci. USA
84:4364-4368[Abstract/Free Full Text].
|
| 6.
|
Cebula, T. A., and W. H. Koch.
1990.
Analysis of spontaneous and psoralen-induced Salmonella typhimurium hisG46 revertants by oligodeoxyribonucleotide colony hybridization: use of psoralens to cross-link probes to target sequences.
Mutat. Res.
229:79-87[Medline].
|
| 7.
|
Centers for Disease Control and Prevention.
1995.
Escherichia coli O157:H7 outbreak linked to commercially distributed dry-cured salami Washington and California.
Morbid. Mortal. Weekly Rep.
44:157[Medline].
|
| 8.
|
De Grandis, S.,
J. Ginsberg,
M. Toone,
S. Climie,
J. Frisen, and J. Brunton.
1987.
Nucleotide sequence and promoter mapping of the Escherichia coli shiga-like toxin operon of bacteriophage H19B.
J. Bacteriol.
169:4313-4319[Abstract/Free Full Text].
|
| 9.
|
Iriarte, M.,
I. Stainier, and G. R. Cornelis.
1995.
The rpoS gene from Yersinia enterocolitica and its influence on expression of virulence factors.
Infect. Immun.
63:1840-1847[Abstract].
|
| 10.
|
Jackson, M. P.,
R. J. Neill,
A. D. O'Brien,
R. K. Holmes, and J. W. Newland.
1987.
Nucleotide sequence analysis and comparison of the structural genes for shiga-like toxin I and shiga-like toxin II encoded by bacteriophages from Escherichia coli 933.
FEMS Microbiol. Lett.
44:109-114.
|
| 11.
|
Jackson, M. P.,
J. W. Newland,
R. K. Holmes, and A. D. O'Brien.
1987.
Nucleotide sequence analysis of the structural genes for shiga-like toxin I encoded by bacteriophage 933J from Escherichia coli.
Microb. Pathog.
2:147-153[Medline].
|
| 12.
|
Keene, W. E.,
E. Sazie,
J. Kok,
D. H. Rice,
D. D. Hancock,
V. K. Balan,
T. Zhao, and M. P. Doyle.
1997.
An outbreak of Escherichia coli O157:H7 infections traced to jerky made from deer meat.
JAMA
277:1229-1231[Abstract].
|
| 13.
| Kotewicz, M. L., B. Li, D. D. Levy, J. E. LeClerc, and T. A. Cebula. Unpublished data.
|
| 14.
|
LeClerc, J. E.,
B. Li,
W. L. Payne, and T. A. Cebula.
1996.
High mutation frequencies among Escherichia coli and Salmonella pathogens.
Science
274:1208-1211[Abstract/Free Full Text].
|
| 15.
|
Martin, M. L.,
L. D. Shipman,
J. G. Wells,
M. E. Potter,
K. Hedberg,
I. K. Wachsmuth,
R. V. Tauxe,
J. P. Davis,
J. Arnoldi, and J. Tilleli.
1986.
Isolation of Escherichia coli O157:H7 from dairy-cattle associated with two cases of hemolytic uremic syndrome.
Lancet
ii:1043.
|
| 16.
|
Milkman, R.
1997.
Recombination and population structure in Escherichia coli.
Genetics
146:745-750[Medline].
|
| 17.
|
Missiakas, D., and S. Raina.
1997.
Signal transduction pathways in response to protein misfolding in the extracytoplasmic compartments of E. coli: role of two new phosphoprotein phosphatases PrpA and PrpB.
EMBO J.
16:1670-1685[Medline].
|
| 18.
|
Nyman, K.,
K. Nakamura,
H. Ohtsubo, and E. Ohtsubo.
1981.
Distribution of the insertion sequence IS1 in gram-negative bacteria.
Nature (London)
289:609-612[Medline].
|
| 19.
|
Ochman, H., and E. A. Groisman.
1995.
The evolution of invasion by enteric bacteria.
Can. J. Microbiol.
41:555-561[Medline].
|
| 20.
|
Ochman, H., and R. K. Selander.
1984.
Standard reference strains of Escherichia coli from natural populations.
J. Bacteriol.
157:690-693[Abstract/Free Full Text].
|
| 21.
|
Pupo, G. M.,
D. K. R. Karaolis,
R. Lan, and P. R. Reeves.
1997.
Evolutionary relationships among pathogenic and nonpathogenic Escherichia coli strains inferred from multilocus enzyme electrophoresis and mdh sequence studies.
Infect. Immun.
65:2685-2692[Abstract].
|
| 22.
|
Riley, L. W.,
R. S. Remis,
S. D. Helgerson,
H. B. McGee,
J. G. Wells,
B. R. Davis,
R. J. Hebert,
E. S. Olcott,
L. M. Johnson,
P. A. Hargrett,
P. A. Blake, and M. L. Cohen.
1982.
Hemorrhagic colitis associated with a rare Escherichia coli serotype.
N. Engl. J. Med.
308:681-685[Abstract].
|
| 23.
|
Takeda, Y.
1997.
Enterohaemorrhagic Escherichia coli.
World Health Stat. Q.
50:74-80[Medline].
|
| 24.
|
Torres, A. G., and S. M. Payne.
1997.
Haem iron-transport system in enterohaemorrhagic Escherichia coli O157:H7.
Mol. Microbiol.
23:825-833[Medline].
|
| 25.
|
Whittam, T. S.,
M. L. Wolfe,
I. K. Wachsmuth,
F. Ørskov,
I. Ørskov, and R. A. Wilson.
1993.
Clonal relationships among Escherichia coli strains that cause hemorrhagic colitis and infantile diarrhea.
Infect. Immun.
61:1619-1629[Abstract/Free Full Text].
|