Title: Towards an understanding of global patterns of simple sequence repeat-mediated phase variation during host persistence of Campylobacter jejuni and Neisseria meningitidis
1Towards an understanding of global patterns of
simple sequence repeat-mediated phase variation
during host persistence of Campylobacter jejuni
and Neisseria meningitidis
Chris Bayliss RCUK Research Fellow Department of
Genetics University of Leicester
Edinburgh Workshop 29-30th September 2010
2Outline
- Overview of my research areas
- Intro to SSRs and phase variation
- Measuring mutation rates/patterns
- Phase variation of C. jejuni genes in in vitro
and in vivo models - Models of SSR-phase variation
- Issues
3My Research Phase Variation
Experimental models/ Epidemiological samples
In silico models
Impact of phase variation rate on population
structure
Mechanistic studies
Campylobacter jejuni
In vitro models
Colonisation of chickens
Combined model
Carriage samples
Neisseria meningitidis
Disease samples
Selection of phase variants
Hb receptors/reversible selection model
Haemophilus influenzae
R-M systems/Phage infection
4 Consequences of Localised Hypermutation Phase
Variation
SELECTION /MUTATION
SELECTION /MUTATION
MUTATION
OFF
ON
ON
Frequency 10-2 to 10-4
5Streisinger Model
6Streisinger Model
7Streisinger Model
Insertion
8Streisinger Model
9Streisinger Model
Deletion
10In-Frame Repeats
ATG..CAAT(30)..//.TAG ON ATG..CAAT(29)
..TAG OFF ATG..CAAT(28)..TAG OFF ATG..
CAAT(27)..//.TAG ON
Promoter-Located Repeats
-35
-10
ATTATA..TA(10).ATTAAA//ATG ON ATTATA..TA(9
)..ATTAAA//ATG OFF
11Functions of the Products of Repeat-Associated
Genes
Flagella Biosynthetic Enzymes
Iron Acquisition Proteins
Capsule Biosynthetic Enzymes
LOS/LPS Biosynthetic Enzymes
Adhesins
Restriction Enzyme
12Long Tracts of Simple Sequence Repeats in
Bacterial Genomes
Repeat Type (min. no. rpts) G/C (8) A/T (10) Di (6) Tetra (5) Penta (3)
H. influenzae (Rd) 6 2 0 12 2
N. meningitidis (MC58) 26 11 4 2 5
C. Jejuni (NCTC11168) 29 2 0 0 0
E. coli (K12) 12 0 1 0 0
13Length of PolyG/PolyC Repeat Tracts in C. jejuni
Contingency Loci
14Phase Variation of Simple Sequence Contingency
Loci
SELECTION /MUTATION
SELECTION /MUTATION
OFF
ON
ON
What are the mutation rates of SSRs? What are the
determinants of SSR mutation rates? What are the
fitness implications of differing switching
rates? What are the roles of selective and
non-selective bottlenecks? What are the
implications of multiple SSCL?
15Campylobacter jejuni- Phase Variation
Frequencies
16Campylobacter jejuni
Gram ve commensal of gasterointestinal tract
of birds and widespread environmental
contaminant Major agent of foodborne
gasteroenteritis Implicated in autoimmmune
diseases such as Guillain-Barre syndrome
17Reporter Constructs for Detecting Phase
Variation in Campylobacter jejuni
cj1139c
cat
lacZ
G8
G8
lacZ
G11
capA (cj0628/cj0629)
T6-G11
Strain NCTC11168
ON
CapA
a-CapA antibodies
(surface-located autotransporter)
18On-to-off
off variant
Off-to-on
on variant
19Colony Blots of C. jejuni strain 11168 probed
with anti-CapA
ON-to-OFF Freq. -ve 0.03 (filter 1, 9/8/07)
OFF-to-ON Freq. ve 0.03 (filter 4, 23/7/07)
20MHA-VT plates
MHA-VT-XGal plates
21H. influenzae N. meningitidis C. jejuni
GC of Genome 38 51 31
MMR Genes MutS/MutL/ MutH MutS/MutL None
SSR Mutation Frequencies 1x10-3 (AGTC30) 4x10-5 (G12) 4x10-3(G11)
Mutational Pattern 90 1/-1 DeletionsgtInsertions Unknown gt95 1/-1 Short insgtdel Long delgtins
Cis-Acting Factors Repeat Number Repeat Number Repeat Number
Trans-Acting Factors PolI, RNaseH MMR, PolIV Unknown
No environmental factors
22Campylobacter jejuni- In vitro/In vivo Passage
23PCR-Based Measurement of Repeat Tract Length
FAM
GGGGGGGGGG
24Multiple Passages of Growth in MHB Broth
Inoculate 5mL MHB
Inoculate 5mL MHB
Inoculate 5mL MHB
Inoculate 5mL MHB
Inoculate 5mL MHB
Pallet the cells
Suspend inoculum
Plate Dilutions
Plate Dilutions
Day 0
Day 1
Day 2
Day 3
Day 4
Pick 30 colonies
Pick 30 colonies
Colony Blotting
Colony Blotting
PCR Array
PCR Array
25Analysis of Phase Variable Genes and Repeat Tracts
CapA Frequency -ve
Inoculum Output 0.29 0.24-0.36
0.29 0.27-0.36
Constant Inoculum (3.5x108cfu 6 tubes)
Variable Inoculum (from 3.5 x108 to 3.5x103cfu 6
tubes)
26Drift, Bottlenecks, Selection and Hitch-Hiking
6 Genes 64 Genotypes
Selection
Bottleneck
0685-on
Random Drift
Mutation/Bottleneck
Mutation/Selection
0685-on 1139-off
1139-off
Mutation/Bottleneck
Mutation/Selection
0031-on
27Neisseria meningitidisPorA Phase Variation,
Immune Evasion and Variant-Specific Immune
Responses During Carriage
28Escape Assay
- Modified serum bactericidal assay using large
inoculum (1x104-1x107 cfu) and multiple passages - LPS phase variants with switches in expression of
lgtG mediate escape of mAb B5 (translational
switching) - Escape dependent on size of inoculum, amount of
antibody and rate of phase variation
Bayliss et al. 2008 Infect. Immun. 765038
29PV of porA mediates immune escape in vitro
Variants examined had 10C residues in the porA
repeat tract Escape is due to pre-existing
variants
30Correlation of porA PV Expression to Escape
- Repeat tract changes to expression
- Whole cell ELISA and lysate western blotting
Level of PorA expression is highest when 11C
repeat units is present in 8047 3 fold of
reduction in expression of porA
31Week 0
Week 4
Week 12
Week 24
Week -4
32Phase Variation of NadA
Volunteer 1st 2nd 3rd 4th V43 12 - 12 - V51 12 1
2 12 12 V52 12 12 12 - V54 14 14 12 - V58 12 12
- 12 V59 13 12 12 12 V88 11 9 9 9 V138 12 12 1
2 -
OFF 9 and 12 rpts
Number of tetranucleotide repeats All volunteers
colonised with YP1.21,16CC174
33Computer Models
34Multiple simple sequence contingency loci
- Multiple loci multiple potential genotypes
- Haemophilus influenzae strain Rd has 12 genes
containing tetranucleotide repeat tracts, a
potential 4096 genotypes (if two genotypes per
locus, i.e. ON and OFF) - Lic2 locus has three genotypes - ON-Strong,
ON-Weak and OFF (if all 12 loci had 3 genotypes
then there is 531 441 potential genotypes)
35Computer Model 1
- Population founded by single organism which
divides by binary fission - Three phase variable loci
- Switching occurs in both directions at the same
rates - Mutations occur during division giving one
genotype of the parental phenotype and one mutant
36Effect of phase variation rate on the amount of
genetic diversity produced in 20 generations
Mutation rate (repeat number)
1x10-6 (lt 6)
3.6x10-5 (10)
1.24x10-4 (22)
1000
900
800
700
600
Number of populations
500
400
300
200
100
0
1
2
3
4
5
6
7
8
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
Number of genotypes
37Effect of phase variation rate on the production
of genotypes with multiple switches
Solution is when all three loci have switched
from OFF to ON. 30 generations were used. All
cells of the parental genotype were removed at
generation 20. 1000 replicates were performed
Number of populations containing solution
Mutation rate
3.6x10-5
21
1.24x10-4
370
38Model 2 Effect of Interval Between Selective
Environments
Environment A Selection for ON Phenotype
Number of Generations
2,000-100,000
2,000-100,000
Environment B Selection for OFF Phenotype
Variable Repeat Number 17 ON A 18 OFF
B 19 OFF B 20 ON A etc 37 OFF B 38
ON A
Mike Palmer and Marc Lipsitch
39Evolution of Repeat Tracts in the Absence of
Selection
40Evolution of Repeat Tracts with Selection and in
a Fluctuating Environment
Environmental switch period- 20 000
generations Fitness advantage- 0.1
41Environmental switch period- 4 000
generations Fitness advantage- 0.1
42Environmental switch period- 2 000
generations Fitness advantage- 0.1
43Environmental switch period- 100
generations Fitness advantage- 0.1
44Summary Computer Simulation Model
- Selection is required to maintain large numbers
of repeats in the repeat tracts - Repeat number is determined by the frequency of
the environmental switch - Correlation between repeat number and
environmental switch is also influenced by the
conferred fitness advantage and mutational
pattern
45Model 3
- Model phase shifts in multiple loci using known
mutation rates (excludes mutational patterns) - Assumes each locus switches independently of
other loci (can set PV rate for each gene, but
not scalable with tract length changes) - Simple deterministic model, average of multiple
trees from a Monte Carlo simulation, performed in
Excel (maximum of 100 generations)
46Sample from Chicken B9
One Isolate B9.1
Gene cj0045 cj0685 cj1326 capA cj1139 cj0032
Tract 9 9 10 12 9 9
Phenotype OFF ON OFF OFF OFF ON
Binary code 0 1 0 0 0 1
Note- genotype is not directly correlated with
phenotype (i.e. cj0045 is OFF with 9 or 10 repeats
Coded phenotypes of all 30 colonies for B9
010001 010100 010101 110000 110001 110100 110101
10 2 2 3 5 1 7
47Drift, Bottlenecks, Selection and Hitch-Hiking
6 Genes 64 Genotypes
Selection
Bottleneck
0685-on
Random Drift
Mutation/Bottleneck
Mutation/Selection
0685-on 1139-off
1139-off
Mutation/Bottleneck
Mutation/Selection
0031-on
48Modelling Changes in the Distribution of Phase
Variants- no selection
6 Phase variable genes ON/OFF 64 genotypes
0off, 1on Output 100 generations
Output 1 all genes at G9 PV rate
(0.0015) Output 2 varied PV rates
49Scientific Issues
- What factors to include in a model mutation
rate, mutational pattern, population size,
fitness, frequency of environmental switching,
bottlenecks, number of loci, number of
generations - How to model simulation of multiple populations
or deterministic model of average solutions
50Logistical Issues
- Data collection (sample bias)
- Computational power
- Biological and clinical relevance
- Simultaneous data collection and modelling (local
collaborators) - Relevance to systems biology
- Requirement for a modelling community
51Jean-Philipe Gautier Jacques Marlet Fadil
Bidmos Nathalie Ingouf Rebecca Richards Awais
Anjum Vladimir Manchev Richard Haig Julian
Ketley (University of Leicester)
Neil Oldfield Del AlaAldeen Karl
Wooldridge Michael Jones Paul Barrow (University
of Nottingham)
Michael Tretyakov Alexander Gorban (University of
Leicester) Michael Palmer Marc Lipsitch Richard
Moxon