GD presentation 13112014

Report
A framework for the future:
Building molecular tools to understand
the epidemiology of Clostridium
difficile in Scotland
Talk Plan
• Highlight extent of problem in 2009
• Progress made – first SIRN grant
– Evaluation of Molecular tools
• Conclusions reached regarding application
• Progress new grant
– Aims to link epidemiological data generated from whole genomic
sequencing with patient healthcare informatics
Scale of problem
Scottish C. difficile reference
laboratory set up 2007
Typing methods
• In 2007, numerous typing methods
– Toxinotyping, Pulse Field Electrophoresis – North Americans
(NAP), Restriction Endonuclease activity (REA)
– Ribotyping
• Based on amplification of 16s-23S ribosomal RNA intergenic
regions
Comparison
1. 3.
PCR
amplification
with
BioNumerics®
database
Ribotype 001 –
represented first
banding pattern
observed, 002.....
2. Image analysis
O’Neill GL et al Anaerobe 1996; 2; 205-209
Ribotyping is informative
• Problems with ribotyping
• Lack of discriminatory power
– Difficulties - band calling
• Size of band /not sequence
– Exchangeability between labs
Prevalence
a percentage
of ten most frequently
– Basis of as
variation
/epidemiology
recovered ribotypes Pre/Post 2009
30
20
10
001
027
106
002
005
014
015
020
023
078
Funding enabled
• Determine strength and weaknesses of several typing
methods
– Multi-locus sequence analysis (MLST)
• Amplification and sequencing of 7 amplicons
– Multi-locus Variable Number Tandem Repeat Analysis (MLVA)
• Amplification and sizing of 7 unique regions encoding tandem
repeat sequences
– Whole Genome Single nucleotide polymorpism (SNP) analysis
• Sequencing and analysis of entire genome
MLST
• Involves amplification of 7 amplicons
– C.difficile
• adk, atpA, dxr, glyA, recA, sodA, tpi
– Sequence and determine varients in each amplicon
adk
atpA
dxr
glyA
recA
sodA tpi
0
0
1
0
0
0
0
1
0
0
1
0
1
0
0
2
Determine population based on relative
similarities/ differences in these alleles
type
96 isolates typed using MLST
• Sequence based/epidemiologically accurate
• No significant advantage over ribotyping
– Lacked sufficient discrimination for use in outbreaks
Multilocus Variable Number
Tandem Repeat Analysis (MLVA)
VNTR = Variable-Number Tandem Repeat
Locus 1
• Strain A: VNTR array 4x3
• atgggtaatccgtcgACgCACgCACgCgccaatcgatacgat
• Strain B: VNTR array 4x5
• atgggtaatccgtcgACgCACgCACgCACgCACgCgccaatcgatacgat
Locus 2
• Strain A: VNTR array 3x4
• ggtaccggtaaagcgcACCACCACCACCttgacactgccggttg
• Strain B: VNTR array 3x6
• ggtaccggtaaagcgcACCACCACCACCACCACCttgacactgccggttg
Data analysis – relative relatedness
Using this approach – determine minimum total
distance between strains
MLVA analysis performed on the top 10
ribotypes in Scotland (n = 748)
MLVA analysis of 027 isolates reveals
strong geographical clustering over time
Diversity of MLVA loci in the ten most common
ribotypes in Scotland.
Ribotype
MLVA Locus (Van den Berg et al, 2007)
A6
B7
C6
E7
F3
G8
H9
001
0.893
0.729
0.910
0.119
0.230
0.506
0.119
002
0.898
0.827
0.916
0.587
0.436
0.711
0.240
005
0.921
0.882
0.930
0.785
0.518
0.760
0.038
014
0.888
0.816
0.878
0.816
0.582
0.551
0.255
015
0.922
0.909
0.936
0.082
0.160
0.806
0.161
020
0.934
0.920
0.927
0.656
0.299
0.865
0.156
023
-
0.790
0.907
0.821
0.457
0.889
0.296
027
0.929
0.773
0.870
0.173
0.497
0.796
0.024
078
-
-
-
0.549
0.128
0.790
0.128
106
0.889
0.836
0.923
0.513
0.096
0.367
0.083
Simpson’s Index of Diversity for the seven MLVA loci in the ten most common ribotypes in
Scotland. High CDI values indicate accurate measurement of a highly variable locus.
Whole Genome
Sequence
Analysis
Using SNP (single
nucleotide
polymorphism)
Scottish Isolates included in global
analysis of 027 ribotypes
He et al. Nature genetics 2013
Correlation between whole genome and
MVLA analysis
GLA
010
GLA
GLA 012
018
GLA
014
GLA
GLA
008
GLA
004
013/019
GLA
002/009
Gla015
Gla004
2.00
Gla010
1.00
Gla020
2.00
Gla021
Gla022
2.00
2.00
2.00
2.00
1.00
2.00
Gla005
2.00
Gla008
Gla007
GLA
007
GLA
017
1.00
Gla017
Gla018
1.00
3.00
1.00
Gla013
GLA
003
GLA
021
GLA
022
GLA
020
GLA
006
Gla003
GLA
015
3.00
Gla016
Gla002, Gla009
2.00
3.00
4.00
Gla019
2.00
GLA
005
Gla006
GLA
001
Gla014
Gla012
Gla001
GLA
016
MLVA sufficiently discriminating to allow tracking within an outbreak
situation
Conclusions
• PCR-ribotyping valuable typing tool
– lacks sufficient discriminatory power to analyse outbreaks.
• MLVA as a supplemental subtyping method
– Cost effective
– is highly discriminatory/discern outbreaks.
– Application limited for some ribotypes
• Whole Genome SNP analysis
– validated and supported data generated by MLVA
– Opportunity to use in limited analysis of 078 outbreaks
• Prediction of phenotypes (antibiotic resistance)
IMPORTANCE OF DATA
LINKAGE
2009 Antibiotic policy change
• Reduction in use of
–
–
–
–
Fluoroquinolones
Third generation cephalosporins
Clindamycin
Co-amoxyclav
30
20
10
001
027
106
002
005
014
015
020
023
078
Modelling impact of policy change
Resistant vs sensitive
Moxifloxacin
Levofloxacin
Erythromycin
Policy Change
0.25
0.50
0.31
Epidemic
167.89
25.38
54.48
Non-epidemic
0.38
0.5
0.49
• Unclear what driving this enhanced resistance
• Possible treatment of a particular population of
patients
• Identifiable if we could link with patient data
Going forward
Project Objectives
• To develop a fully automated pipeline to allow rapid SNP
analysis of genomic DNA from C. difficile.
• To use this pipeline to evaluate
– Relationship between community- and hospital-associated C.
difficile strains
– Further differentiate epidemiology of ribotype 078 in Scotland
• To link patient health data and C. difficile WGS
– To identify risk factors that could be reduced through modification
of clinical practice
Where are we?
• Have identified 500 strains to be sequenced
– 100 078 strains
– 134 community associated strains isolated pre 2009
– 270 matched community/hospital associated strains isolated post
2009
• 350/500 have been sequenced
• Automated SNP analysis pipeline generated
• Permissions have been sort and granted to allow patient
linkage
• Plan in next 12 months to begin to link the data together
• Long term aim:
•
To develop a unique Scottish electronic resource that can be
interrogated by researchers focussed on this infection
Acknowledgments
SSSCDRL
John Coia
Derek Brown
Health Protection Scotland
Camilla Wuiff
A-Lan Banks
University of Glasgow
Jan Lindstrom
Cosmika Goswami
Umer Ijaz
Chris Quince
University of Dundee
Charis Marwick
Peter Doonan
Peter Davey
Nichosa De Souza
University of Strathclyde
Marion Bennie
The Wellcome Trust Sanger
Institute
Trevor Lawley
Miao He
Sequencing undertaken Glasgow
Polyomics facility

similar documents