Transcript Slide 1

Host specificity of pathogenic
Escherichia coli
Eliora Z. Ron, Tel-Aviv University
‫תודה‬
References
Ideses, D., U. Gophna, et al. ( 2005). A Degenerate Type-III Secretion System from Septicemic Escherichia
coli Contributes to Pathogenesis. J Bacteriol in press.
Mokady, D., U. Gophna, et al. (2005). Virulence factor of septicemic E. coli strains. IJMM in press.
Mokady, D., U. Gophna, et al. (2005). Extensive gene diversity in septicemic Escherichia coli strains. J Clin
Microbiol 43: 66-73.
Ideses, D., D. Biran, et al. (2005). The lpf operon of invasive Escherichia coli. Int J Med Microbiol 295:227236.
Gophna, U., D. Ideses, et al. (2004). OmpA of a septicemic Escherichia coli O78--secretion and convergent
evolution. Int J Med Microbiol 294: 373-381.
Gophna, U., E. Z. Ron, et al. (2003). Bacterial type III secretion systems are ancient and evolved by multiple
horizontal-transfer events. Gene 312: 151-163.
Gophna, U., A. Parket, et al. (2003). A novel ColV plasmid encoding type IV pili. Microbiology 149:177-84.
Gophna, U. and E. Z. Ron (2003). Virulence and the heat shock response. Int J Med Microbiol 292:453-61.
Adiri, R. S., U. Gophna, et al. (2003). Multilocus sequence typing (MLST) of Escherichia coli O78 strains.
FEMS Microbiol Lett 222: 199-203.
Gophna, U., T. A. Oelschlaeger, et al. (2002). Role of fibronectin in curli-mediated internalization. FEMS
MICROBIOLOGY LETTERS 212: 55-58.
Gophna, U., T. A. Oelschlaeger, et al. (2001). Yersinia HPI in septicemic Escherichia coli strains isolated
from diverse hosts. FEMS Microbiol. Let 196: 57-60.
Gophna, U., M. Barlev, et al. (2001). Curli Fibers Mediate Internalization of Escherichia coli by Eukaryotic
Cells. Infect Immun 69: 2659-65.
Dobrindt, U., G. Blum-Oehler, et al. (2001). S-Fimbria-Encoding Determinant sfaI Is Located on
Pathogenicity Island III536 of Uropathogenic Escherichia coli Strain 536. Infect. Immun. 69: 4248-4256.
Babai, R., B. E. Stern, et al. (2000). New Fimbrial Gene Cluster of S-Fimbrial Adhesin Family. Infect Immun
68: 5901-5907.
Babai, R., G. Blum-Oehler, et al. (1997). Virulence patterns from septicemic Escherichia coli O78 strains.
FEMS Microbiol Lett 149: 99-105.
Yerushalmi, Z., N. I. Smorodinsky, et al. (1990). Adherence pili of avian strains of Escherichia coli O78."
Infect Immun 58: 1129-31.
Virulent E. coli strains
• Most of the E. coli strains are commensal, but
a small number are pathogenic
• Pathogenic E. coli strains are divided into two
groups:
– Intestinal strains. These produce
enterotoxins and constitute a major
problem, especially in young children and
travellers (Montesumu’s revenge)
– Extraintestinal strains – ExPEC
(Extraintestinal Pathogenic E. coli)
Extraintestinal diseases caused by E. coli
• Urinary
tract infections (UTI) (pyeolonephritis, kidney
failure, productivity loss)
• UTIs are responsible for > seven million patient visits
and one million hospital admissions (due to
complications) per year in the United States only. 80 90% of the cases are caused by E. coli
• Neonatal meningitis: bacterial meningitis
• 0.25 per 1000 live births in industrialized countries
(2.66 per 1000 in developing countries). ~30% caused
by E. coli , ~10% mortality
•Intra-abdominal infections, respiratory tract infections,
wound and surgical infections
•Septicemia
Septicemia (colibacillosis)
• Colisepticemia is the major causes of
mortality from community and hospitalacquired infections (more than 80%)
• Main cause of mortality in immunosupressed patients (HIV, chemotherapy,
old age)
• Colisepticemia is an emerging disease –
83% increase 1980 – 1992, over 40% of
the bacteremia cases in community
acquired infections
Colisepticemia in farm animals
• A lethal disease in newborn lambs – bacteria carry
K99 fimbriae and colV plasmid. Colonize lambs on
birth, death within two weeks.
• Avian colisepticemia - serious disease of chickens
and turkeys. Heavy direct losses due to high
morbidity and mortality, as well as indirect losses
due to intensification of other respiratory diseases
caused by viruses or mycoplasma
• Losses of several million dollars a year reported
by DelMarVa industries alone
Avian Colisepticemia
• Bacteria enter the host by adherence and
colonization of upper respiratory tract
• Later, the bacteria cross epithelial barriers
and invade deeper tissues
• Finally the bacteria enter bloodstream and
proceed to the vital organs of the host
• A good model system for ExPEC infections
• 80% of the cases of colisepticemia, world –
wide, are caused by E. coli serotypes O2 and
O78
ExPEC strains
• Known virulence factors include genes for
efficient iron uptake, serum resistance and
adherence to host tissues. Do not produce
toxins
• What makes them virulent?
• Are the virulence factors host dependent ?
• Is there a clonal relationship? (host
dependent?)
• Can we predict an outbreak of ExPEC (early
warning) ?
Goals:
• Define virulence-essential ExPEC-specific
genes
• Profile strains involved in UTI, NBM and
sepsis using these ExPEC-specific genes
• Look for virulence genes which determine
host specificity
• Use the data to define potential targets
for development of vaccines and/or
antibacterial drugs.
What is a virulence factor?
•
Encoded by a gene present only in pathogenic
strains
•
Without this factor virulence is decreased
without a decrease in growth rate
•
Example: toxins
•
Virulence factors can determine host
specificity
•
Example: adherence fimbriae (pili)
E. coli O78 – septicemic in humans and
birds
AC/I pili –
contributes to
specific
adherence
only in avian
septicemic
strains
S-fimbrial adhesin family
• Found in human
pathogenic E. coli
strains
major subunits
• Composed of
around 1000
protein units,
major and minor
subunits
minor subunits
Degree of identity between AC/I
(Fac) orfs to SfaI, SfaII and Foc
Fac orfs
Sfa I
Sfa II
Foc
FacA
FacD
FacE
FacF
FacG
FacS
FacH
66
98
99
98
100
69
?
99
97
98
60
(major)
(minor)
(minor)
(minor)
(UTI)
100
72
80
(NBM)
?
?
?
99
71
82
(UTI)
96
Adherence of strain 781 to avian
epithelial cells
- adherence of strain
781 producing AC/I
compared with
- 781 strain not
expressing AC/I
(grown in 180C)
- Unpiliated strain
Preferential adherence of strain
781 to avian epithelial cells
Unpiliated strain
CFA/I expressing
strain
AC/I expressing
strain
AC/I fimbriae appear to be a virulence factor,
probably avian specific
Identification of virulence related
sequences in septicemic strains
• Whole genome sequencing
• Subtractive hybridization
Subtractive hybridization
• Obtain pathogen specific sequences,
absent from non-pathogenic K12 strain
• Excellent chance of “hitting”
pathogenicity islands which are pathogen
specific and very large
• Faster (and much cheaper) than whole
genome sequencing
Subtractive hybridization
A way to study comparative genomics
with organisms which have not been
sequenced
Pathogen
NonPathogen
Pathogen
Specific
Library of
pathogen specific
genes
O78-9 sequences
known
functions
putative virulence-associated
putative and known
virulence factors
unknown
functions
phage related
mobility-related
Search for unique “septicemic”
sequences
• Using suppression subtractive hybridization
(SSH) we identified sequences unique to strain
O78-9 and absent from the non-pathogenic
strain K-12
• Over 80 O78-specific open reading frames were
found (91 to 1473 bp in length)
• The same experiment was repeated with
another septicemic strain O2-1772
• 117 unique O2 sequences were identified
O2-1772 sequences
unknown
functions
known functions
putative
virulence-associated
putative and known
virulence factors
phage associated
mobility-related
• Both libraries contain many sequences
associated with genomic plasticity evolution by horizontal gene transfer
• Many sequences of O2 and O78 are
homologous to virulence related
sequences of human ExPEC strains
• The virulence related genes identified by
SSH include iron uptake systems,
adhesins autotransporters and secretion
genes, including a new T3SS
T3SS
• Type three secretion systems (T3SS) of E. coli
O157 and other invasive bacteria deliver
effectors into the cytosol of the host cells
• A novel T3SS was discovered in genomic studies
of E. coli O157 and others – called ETT2 (E. coli
type-three secretion system 2)
• So far if is not clear if ETT2 has a role of in
pathogenesis
• Our results of SSH indicated that septicemic
strains have the ETT2 type of TTSS – first
demonstration in septicemic strains
However • The cluster contains a large deletion and
several stop codons and appears to be “dead”
ETT2sepsis
has several premature stop
codons and a large (five Kb) deletion
• These genetic modifications result in
an inability to produce the “needle”
• The 5 kb deletion is conserved in
eleven E. coli strains from septicemia
and newborn meningitis.
• ETT2sepsis
Is ETT2sepsis really dead?
• Although the ETT2sepsis is degenerate,
the gene cluster is transcribed
• A null mutant deleted for ETT2sepsis was
constructed and grows as well as the wild
type
• However, the deletion of ETT2sepsis
results in modification of bacterial
surface properties which could affect
interaction with host cells and immune
system
Turbidity (Klett Units)
Growth of strain 789 & the ETT2 deletion
100
O789, 37oC
prg, 37oC
O789, 42oC
prg, 42oC
10
0
1
2
3
4
Time (hours)
5
6
7
250C
370C
epr/p
789 Δ
epr
789 Δ
789
789
789 Δ
epr
EPR
Null mutation in ETT2sepsis affects
surface properties but only above 370C
(host conditions?)
ETT2sepsis null mutants are not virulent
100
Survival (%)
80
wild type 789
prgHIJK null mutant
60
40
20
0
0
2
4
Days after injection
6
8
T3SS
• SSH indicated that septicemic E. coli
strains have the ETT2 type of TTSS –
ETT2sepsis
• ETT2sepsis is degenerate but important for
virulence
• These results are the first demonstration
of the importance of ETT2 in pathogenesis
• The biological role of ETT2sepsis probably
does not involve classical secretion of
effectors
Some virulence factors are missed
by the genomic approaches
• Genomics and proteomics give
information about sequences of proteins
which are unique to virulent strains
• By the classical definition, a virulence
factor is encoded by a gene present only
in pathogenic strains
• This is not always so – example: curli
fibers
E. coli O78 are internalized and
replicate within cells
6)
human (Hela, T24) and avian (T24)
3
.
6
e
+
5
3
.
0
e
+
5
Numberofintraceluarbcteria(cfu/10
2
.
4
e
+
5
1
.
8
e
+
5
1
.
2
e
+
5
0
5
0 1
0
0 1
5
0 2
0
0 2
5
0 3
0
0
T
i
m
e
a
f
t
e
r
t
h
e
a
d
d
i
t
i
o
n
o
f
p
o
l
y
m
i
x
i
n
(
m
i
n
)
Confocal laser scanning microscopy of internalized bacteria
Cells are visualized using anti-actin or anti-tobulin antibodies and the bacteria express GFP
Transmission electron microscopy of internalized bacteria
A clone from E. coli O78 cosmid library
mediates internalization
Invasion of HeLa cells by E. coli clones
Strain
No. of intracellular bacteria/5x10 5
cellsa
VCS257(pMMB33)
VCS257(pMMB33Inv)
C600 (pMMB33)
C600(pMMB33Inv)
100
2 x 105
12
3.5x105
The clone carries the csg gene cluster encoding curli
Curli fibers
Bian, Z., Branuer, A., Li, Y. and Normark, S., 2000. JID 181: 602
Curli - thin coiled fibers with high
affinity binding for several host
proteins:
Plasminogen and plasminogen activator
MHC Class I molecules
Laminin
Fibronectin
However...
• Non pathogenic E. coli strains also
contain the csg cluster encoding
curli...
Is curli of O78 different than this
of K-12?
Is Curli a virulence factor?
A high level of curli expression promotes
internalization by eukaryotic cells
2.0E+05
1.5E+05
1.0E+05
5.0E+04
0.0E+00
C600
(pCKcsg)
C600
(pCLInv)
C600
Internalized bacteria (cfu/ml)
2.5E+05
When expressed in
multicopy, the
presence of the csg
gene cluster of
pathogenic and non
pathogenic strains
promotes
internalization.
The csg clone from
E. coli O78 is more
effective than that
of the E. coli K-12
clone
Is it a better curli
or more curli?
E. coli O78 expresses high levels of
curli at host conditions
Curli expression at 42°C
Expression of O78 curli is differnet
from K-12 curli
• Curli of O78 are expressed at higher levels
• Curli of O78 are expressed under host
conditions (high temperature, high osmolarity)
• Sequencing indicated that the CsgD - activator
required for transcription curli operons – of
O78 is different from K-12 and similar to that
of Salmonella. Could explain the differences in
expression of curli.
Recent support for the role of curli
in virulence
• Isolates from human sepsis constitutively
express high levels of curli.
• O157:H7 isolates with increased curli
expression are more invasive to cells and
more virulent in mice.
• Curli mutants of avian septicemic E. coli
serotype O78 are attenuated in vivo
•
Virulence factors are encoded by a
gene present only in pathogenic strains
•
Other virulence factors encoded by
genes which are presnet in virulent and
non virulent strains, but have a
different expression or activity in
virulent strains
•
These virulence factors are missed in
genomic studies
Analysis of the unique “septicemic”
sequences
• Using subtractive hybridization (SSH) of
septicemic strains and K-12, we identified
over 80 sequences unique to strain O78-9
and over 110 sequences unique to another
septicemic strain, O2-1772
• Are the unique sequences of O78 similar
to the unique sequences of O2?
Screening of additional septicemic
strains of E. coli O78 and O2
serotypes
• Comparison of the two subtractive
hybridiation libraries indicated a large
diversity between the O2 and the O78
strains
• Is this diversity serotype specific?
• To determine this we screened additional
septicemic strains of the same serotypes
the presence of each of the unique
sequences
PCR of septicemic E. coli O2 and O78
strains with virulence specific sequences
Kb
5
3
2
O 78 - 18 ( 1369 bp )
1
0 .5
Large variability, not serotype related
O 78 - 55 ( 570 bp )
O 2 - 210
( 419 bp )
O 2 - 334
( 264 bp )
O2 strains
sequence
O2-1
hypothetical protein
O2-14
no homology
O2-32
unknown
O2-38
predicted ATP-binding protein
O2-42
hypothetical protein
O2-87
Screening of additional
septicemic strains of
serogroups O2 and O78 for
presence of specific
sequences
unknown
O2-108
iron transport proteins
O2-125
no homology
O2-132
no homology
O2-138
unknown
O2-144
no homology
O2-154
P pili assembly chaperon
O2-157
no homology
O2-158
unknown
O2-164
no homology
O2-165
unknown
O2-193
no homology
O2-207
hypothetical protein
O2-210
hypothetical protein
O2-280
no homology
O2-311
unknown
O2-319
no homology
O2-334
no homology
O2-345
unknown
O2-349
putative protein
O2-353
hypothetical protein
O2-355
hypothetical protein
O78-13/O78-74
minor subunits of AC/I pili
O78-17
hypothetical protein
O78-18
hypothetical protein
O78-24
no homology
O78-27a
unknown
O78-44/O78-102
enterobactin receptor iroN
O78-55
no homology
O78-63
hypothetical protein
O78-69
no homology
O78-79
hypothetical protein
O78-95
Autotransporter
O78-96
yersiniabactin synthesis
O78-138
no homology
O78-161
hypothetical protein
O78-163/O2-106
TTSS proteins
1772
avian
YN
avian
MAN
avian
SA
avian
O78 strains
BEN
U33
B18
285
286
287
63-1
avian human human human human human sheep
786
avian
787
avian
789
avian
K12
Comparison of unique sequences of
septicemic strains of serotype O2 and O78
• high level of genome plasticity
• there is a high diversity between the
SSH libraries of O2 and O78 strains,
with only a few shared genes coding for
virulence factors
• unexpected for two strains causing the
same disease
• Septicemic strains of serogroups O2
and O78 contain a large pool of virulence
genes which are used in a “mix and
match” fashion
Byproduct:
- we found one sequence which is present in all
O78 strains and in none of the O2 strains
- can be used for detecting O78, especially in
food
1 2 3
4
5 6 7
8 9 10 11 12 13 14 15 16
Kb
5
3
2
O78-95 (955bp)
0.5
O2 strains
O78
strains
• There is large diversity in the profiles
of virulence specific genes
– this is in contrast to the results of O157
• The profile of virulence specific genes
is independent of the host
• Is there a virulence-associated or hostdependent clonal relationship between
the strains?
• Clonal relationship was determined using
MLST
MLST of O78 strains
• Multi Locus Sequence Typing
• 450 – 500 bp of 7 “housekeeping”
genes
• Criteria for chosing genes:
– 97-98% homology to E. coli K-12 (from
blast data)
– appear in pathogenic and non pathogenic
strains
– map at considerable distance from each
other
– several allels in the population
Genes chosen for use in MLST
Gene
adenylate kinase
glyoxylate carboligase
glucose-6-phosphate
dehydrogenase
Symbol # of
alleles
adk
8
gcl
9
gdh
8
malate dehydrogenase
mdh
7
homoserine transsuccinylase
polyphosphate kinase
metA
8
ppk
8
Genes selected for MLST
Bacteria used for MLST
• O78 strains, pathogens (ExPEC) and non
pathogens (28 strains)
– Human
– Avian
– Sheep - cattle
Neighbor-joining MLST phylogenetic tree of E. coli strains and virulence to 1-dayold chick .
H: Human; C: cattle; A: avian pathogen; -: no mortality; +: less than 25% mortality; ++: less
than 25–49% mortality; +++: 50–74% mortality; ++++: 75–100% mortality
• There is a positive correlation between
virulence, invasiveness and clonal origin
• Clonal division in E. coli O78 strains is
host independent - closely related
clones reside in different hosts
• The MLST results are compatible
with the results from subtractive
hybridization and sequencing
• The profile of virulence factors in
ExPEC strains is independent of the
host and independent of the serotype
• Is there host specificity in ExPEC
strains??
• E. coli strains isolated from avian
septicemia are more virulent to chicks
than strains isolated from NBM
• This result was unexpected in view of
the finding that the virulence genes and
the clonal profile of virulence factors in
ExPEC strains is independent of the host
Open Questions
• Which genes are involved in the
higher virulence to chicks?
• Are there strains that are more
virulent to mammals than to chicks
(“human specific”)?
• What is the zoonotic risk of human
infection with avian colisepticemic
strains?
Virulence factors and host specificity of
ExPEC strains – beyond the “omics”
•
•
•
•
•
It is possible to identify unique genetic
sequences, present only in virulent strains
There is a high variability in these sequences,
which is host independent
There are virulence factors which are encoded
by genes present also in non virulent strains
(i.e., curli)
“Dead” gene clusters can nevertheless be
important virulence factors (ETT2)
There is a clear host specificity, but its basis
is not obvious from examining “virulence
genes”
Support
•
•
•
•
European Community project COLIRISK
Center for Emerging Diseases, Israel
GIF – German Israeli Science Foundation
Manja and Morris Leigh Chair of
Biophysics and Biotechnology
Tel Aviv University
Collaborations
Eliora Z. Ron
U. Würzburg
Joerg Hacker
Reuven Babai
Uri Gophna
Diana Ideses
Daphna Mokady
Roni Segal-Adiri
Zohar Yerushalmi
Michael Naveh
Dvora Biran
U. Helsinki –
Timo Korhonen
INRA - Centre de Tours
M. Moulin-Schouleur