Campylobacter jejuni
Epidemiology and Evolution
Veterinary Training Research Initiative
Food-borne zoonotic pathogens:
Transmission, pathogen evolution and control

Lancaster University


University of Liverpool


Howard Leatherbarrow, Tony Hart, Malcolm Bennett
Health Protection Agency



Paul Fearnhead, Edith Gabriel, Peter Diggle
Andrew Fox, Steve Gee, Sam James
John Cheesbrough, Eric Bolton
Funding: DEFRA

(Department for Environment, Food and Rural Affairs, UK
Government)
Characterisation of
C. jejuni infection
• Epidemiology
• Evolutionary history
• Population genetics
• Source of human cases
Foodborne illness in the UK
Food Standards Agency figures for 2000
Salmonella
16,987
20.9%
Campylobacter
62,867
77.3%
1,147
1.4%
Clostridium perf.
166
0.2%
Listeria
113
0.1%
E.coli O157
Total
81,280
$8bn
Annual cost to US economy
Buzby et al. JID (1997)
Cases and controls
Poisson point process
Significance
Risk
Seasonal patterns
25.4% reduction
year-on-year
Harmonic regression
MLST: Multi-locus sequence typing
477bp aspA
477bp glnA
1.6 million bp genome
MLST: 3300 bp in total
(0.2%)
402bp gltA
498bp pgm
tkt 459bp
489bp uncA
Multi-locus sequence typing (MLST)
glyA 507bp
st
cc
aspA
glnA
gltA
glyA
pgm
tkt
uncA
freq
21
21
2
1
1
3
2
1
5
116
104
21
2
1
1
3
7
1
5
54
53
21
2
1
21
3
2
1
5
40
50
21
2
1
12
3
2
1
5
38
19
21
2
1
5
3
2
1
5
24
140
120
21
257
100
Abundance
48
80
104
45
53
60
50
19
40
61
574
20
0
21
257
48
104
Multi-locus sequence typing (MLST)
45
53
Se que nce Type
50
19
61
574
ST 50: 2 1 12 3 2 1 5
ST 104: 2 1 1 3 7 1 5
ST 21: 2 1 1 3 2 1 5
ST 21: 2 1 1 3 2 1 5
Genetic inhomogeneity in Campylobacter jejuni
Figure 1. Number of nucleotide diff erences betwee n each pairs of alleles at the seven loci. When C. coliderived alleles are removed, the red portion of the histogram disappears.
Common ancestor
Campylobacter
jejuni
Campylobacter
coli
gene flow
Putatative
Putatative
isolates
isolates
Campylobacter jejuni
ST
Campylobacter coli
aspA glnA gltA glyA pgm
tkt
uncA
ST aspA glnA
gltA glyA pgm
tkt uncA
21
2
1
1
3
2
1
5
825
33
39
30
82
113
47
17
257
9
2
4
62
4
5
6
826
33
39
30
114
104
35
17
66
2
4
5
2
7
1
5
832
33
39
30
79
113
43
17
61
1
4
2
2
6
3
17
868
81
104
81
113
143
119
67
ST aspA glnA gltA glyA pgm
257
825
21
61
233
9
1
139
2
4
130
4
2
62
382
2
113
2
4
6
tkt
147
5
3
uncA
17
517
6
Allele frequency
C.
C.jejuni
jejuni 0.13
0.33
0.07
0.01
0.22
0.01
0.14
0.08
0.01
0.15
0.05
0.01
0.18
0.2
0.23
0.05
0.06
0
0.29
0.07
0.18
0
0.29
0.17
0.04
C.
C. coli
coli 0.64
0
0.49
0
0.62
0.01
0
0.46
0.01
0
0.18
0
0.15
0.01
0
0.48
0
Population Structure: identifying hybrids
C. jejuni – C. coli hybrids
3 sequence types, 20/881 isolates (2.2%)
Population Structure: identifying hybrids
Common ancestor
Campylobacter
jejuni
Campylobacter
coli
gene flow
aspA-33
pgm-93
uncA-17
QuickTime™ and a
decompressor
are needed to see this picture.
Coalescent modelling with ABC
QuickTime™ and a
decompressor
are needed to see this picture.
1592
(1772
-1305)
MRCA
uncA-17
1966
pgm-93
1998
2000
2002
aspA-33
Campylobacter
jejuni
Campylobacter
coli
Common ancestor
4 500 BC
(1 500 10 000 BC)
Domestic pig domestication in the
Paris Basin, circa 4000 BC
Larson et al. PNAS (2007)
gene flow
present
Campylobacter jejuni
aspA-33
pgm-93
uncA-17
Campylobacter coli
9.9 m
24 m
6.6 m
32 m
9.8 m
51 m
7.5 m years
BIRD
SHEEP
CATTLE
CHICKEN
ENVIRONMENT
PIG
BIRD
SHEEP
CATTLE
CHICKEN
ENVIRONT
PIG
HUMAN
CATTLE
CHICKEN
BIRD
ENVIRONMENT
SHEEP
PIG
cattle
beef offal or meat
calf
cows milk
calf faeces
cattle faeces
chicken
chicken offal or meat
chick
wild bird
starling
goose faeces
turkey
goose
duck
starling faeces
sand (bathing beach)
environmental waters
soil
potable/drinking water
sheep
lamb offal or meat
lamb
sheep faeces
pig
pork offal or meat
piglet
212
47
12
11
1
1
222
153
17
172
71
25
22
12
7
4
52
28
3
2
84
74
10
2
35
10
1
284
392
313
85
170
46
broiler environment
cat
dog
farm slurry
gazelle
giraffe
goat
horse
human blood culture
human stool
human unspecified
marmoset
ostrich
other animal
rabbit
unspecified
17
3
5
10
1
1
5
1
57
1684
102
2
1
25
3
144
Haplotype structure in sequences of known origin
from pubMLST
origin
PIG
SHEEP
ENVIRONMENT
BIRD
CHICKEN
CATTLE
Haplotype structure in human isolates
key
NOVEL
PIG
SHEEP
ENVIRONMENT
BIRD
CHICKEN
CATTLE
Haplotype structure in human isolates
key
NOVEL
PIG
SHEEP
ENVIRONMENT
BIRD
CHICKEN
CATTLE
Attributing novel genotypes

ST 574: 7 53 2 10 11 3 3

Human-specific, but similar to...





ST
ST
ST
ST
305: 9 53 2 10 11
713: 12 53 2 10 11
728: 4 53 2 10 10
2585: 7 2 3 10 11
3
3
3
3
3
3
3
3
All these found in chicken, so the likely
source is chicken
BIRD
SHEEP
CATTLE
CHICKEN
ENVIRONMENT
PIG
BIRD
SHEEP
CATTLE
CHICKEN
ENVIRONT
PIG
HUMAN
Model C: unlinked
Does it work?
Empirical cross-validation

Split sequences of known origin into two groups.
Treat one group as having unknown origin (pseudohuman cases)

Infer the proportion of pseudo-human cases drawn
from each source population

Repeat 100 times to study the performance of the
method
Simulation and empirical cross-validation
Results: Linked model
Predicted Correct
Actual Correct
Coverage
Bias
CATTLE
CHICKEN
BIRD
ENVIRONMENT
SHEEP
PIG
Combined
CATTLE
CHICKEN
BIRD
ENVIRONMENT
SHEEP
PIG
Unlinked model
0.86
0.56
22
85
81
63
16
89
18
-0.13
-0.01
-0.02
-0.01
0.16
0.01
Linked model
0.66
0.58
100
82
100
99
98
94
95
0.00
-0.05
-0.02
0.00
0.06
0.00
Case-by-case: probability of source
PIG
CATTLE
SHEEP
ENVIRONMENT
WILD BIRD
CHICKEN
Gene flow between source populations
Tracing the source of infection: results
Tracing the host species of human
food-borne infections

Infected meat principal source (97%)

No evidence for environment or wild birds
as a major transmission route

Prevention strategies:


Enhanced on-farm biosecurity
Interrupting transmission chain