Does WGS play a role forFood Safety?

Beatriz Guerra Román

XII IMMEM, Dubrovnik

26.09.2019

© ESCMID eL

ibrary

by au

thor

EFSA Molecular studies for Food Safety

2

Novel molecularbiology methods

Effects chemicalcomponents on living organisms

Geneticpolymorphisms

Novel computationalapproaches to predictdrug metabolism and toxicity

Enzymecharacterization

OMICs

Detection of emerging«high risk» clones

WGS

Metagenomics

AMR surveillance

(EUSR)

Outbreakinvestigation(ROAs)

Characterization/predictionresistance/virulence

Transcriptomics

GMOs

Guidelines applicationsregulated products

Cloud

Novel food

Food/ Feedaditives

Pesticides Technical notes

Metabolomics

Big data

Biomarkers

Proteomics

Source attribution

Health claims

© ESCMID eL

ibrary

by au

thor

What is WGS for?

3

WGS sequencing

…ATTGCATCACTAGATTAGATTATGAAGATATTGATTACACCTCGA….

ACTGAACCTGTACGCCAGAAACAGCAAGCCCATGCTTTTCCGGCGCACAGATTCAAAGTGACGGACCGTT

GTACGTTCGCTGTCGCTGAGATCGTTCCAGCTCCAGCGCTTTCTGATGCTGCGCCGCCCGTTCCTGCTCA

CGATGTGCCTCTATCACGTTCAGCCTTGTGACAACCGGCGCAGCTGCTCGCTCAAACTCTGCACCTGCCT

GCTCAAGCCCGTGACGCGCTCGCTCAGCGCCGCGTTCTCCCGTTGCATAAGACCAGACATCGTCCGCCAT

TCCGCGAAGGCGTTCTCCCACTCGTCCAGCCTTTTCGAGTAGTCCTGCTGTAGCTGCTCTAATGCGCTCA

GCAACTGTTTTTCCAGCTCCGTCATGTGCTATTTCCCCGCCAGTATCATCCAGCTCGAATCCTCTCCGAA

CCGTCCCGCTTCCGGGTTCACCCTCACGCAGCGGCGTCTCTGCTCTCCGCAGGTCGAGAGCTGCACGCCG

CTGTTCCTCTCGGACAGCATGGCCTGCGTGCTCTTCTGCTCCCGGATGATCGTGTAATTGTCGAGTATCT

GCTGCCCCTGCCACCATAGAATGCTTGCGCCCGATGCTGTCAGCAGTGTAGACACCAGGACGATGGTCAG

CCACGTCTGGCTGACCATACGCACCATGCCTTTCGTGCGCTGGCTCATGGCTGAGGACAGCTTCCGGTCG

Resistance traits

Pathogenicity, virulence

Strain characterization

Plasmid characterization

Phylogeny Genetic relatedness

Species identification

Persistence traits publicST121

publicST6

publicST9

publicST8

publicST1

publicST155

publicST2

publicST101

publicST4

publicST87

publicST7

publicST37

publicST3

publicST14

publicST204

publicST31

publicST399

publicST59

publicST5

publicST21

publicST394

publicST20

publicST325

publicST18

“TAG the AGATA’s CAT tail*”

*Iolanda’s Mangone idea! (ISZ, Teramo, out bioinformatician for 1,5 years)

© ESCMID eL

ibrary

by au

thor

4

Foodborne pathogens and zoonotic bacteria

© ESCMID eL

ibrary

by au

thor

EC survey: WGS for food/waterborne pathogens

5

Q1. DO YOU CARRY OUT WGS ACTIVITIES? 28% YES (N=154 respondents)

Status December 2016Current data collected by the EURLs end2017 show a similar picture (EC EURLsWGS WG personal communication)

YesNo

*

*© ESCMID eL

ibrary

by au

thor

General questions section:

6

Q1...DO YOU CARRY OUT WGS… BY RESPONDENT (N=72 respondents)*

*72 respondentscorresponding to a total of 183 replies frompathogen specificlaboratoriesacting as NRLs. Analyses of data by “respondents” or by “Network laboratories” provided a similar picture for thewholequestionnaire.

<50%

© ESCMID eL

ibrary

by au

thor

Examples of WGS for Food Safety

7

WGS for Outbreak investigations

and other activities

Context: examples of work done somewhere else

Our reality: work done in/for EFSA© ESCMID eL

ibrary

by au

thor

Do you remember? E. coli O104 Germany

May 2011

3816 cases

845 HUS cases

54 deaths

Taken from Frank et al. NEJM 2011; 365:1771-1780

Photos kindly provided by Istvan Szabo, National reference Laboratory for Salmonella,

Federal Institute for Risk Assessment, Berlin, Germany

B. Guerra

© ESCMID eL

ibrary

by au

thor

O104:H4 E. coli WGS within days…

9

BGI, Uni Hamburg-Eppendorf

TY2482 (1/2. June)

Health Protection

Agency

H112180280 (10 June)

Lifetech, Uni Münster

LB226692 (1/2. June)

Published on July 27, 2011, N Engl J Med 2011;365:718-24.

Received June 29, 2011; Accepted July 6, 2011; Published July 20, 2011

www.hpa-bioinformatics.org.uk/lgp/genomes

© ESCMID eL

ibrary

by au

thor

Social media Real time WGS Analysis O104:H4

10

Blogs and Wikis!!!!

BacPathGenomicsGenomics and evolution of bacterial pathogens

CrowdSourcingBlogging the German outbreak E.coli 0104:H4

Pathogens: Genes and GenomesA heady mix of bacterial pathogenomics, next-generation sequencing, type-III secretion, bioinformatics and evolution!

EHEC Infection @EHECInfection· http://www.ehecinfection.com

GMO Pundit a.k.a. David TribeHelping readers navigate the confusing myths of modern biology

Figures kindly provided by Kat HoltLondon School of Hygiene & Tropical MedicineMonash University, Melbourne

© ESCMID eL

ibrary

by au

thor

O104, more info...

11

DE isolates, little diversity, 2 SNPs in 4 isolates. FR isolates, 19 SNPs in 7 isolates. DE isolates formed a clade within the more diverse FR strains. Diversity present in the original set, and later reduced in the distribution to Germany?, or original homogeneous populationsubjected to different mutation rates upon their geographicalsegregation?

© ESCMID eL

ibrary

by au

thor

12

Detection methods: real time PCR (based on WGS data)

© ESCMID eL

ibrary

by au

thor

13

O104, more info...

© ESCMID eL

ibrary

by au

thor

Listeriosis outbreak

• Mid-June 2017

• Several clinicians/microbiologists report unusually high numbers of listeriosis cases

• Signaled the start of the outbreak

• 1 January 2017 to 17 July 2018

• 1060 cases

Figures kindly provided by Anthony Smith

Centre for Enteric Diseases (CED),

National Institute for Communicable Diseases (NICD)

South Africa

>1000 cases, the largest L. monocytogenes outbreak

© ESCMID eL

ibrary

by au

thor

15

Outbreak of Listeria monocytogenes in South Africa, 2017-2018

91% of human isolates are

MLST sequence type 6 (ST6)

abcZ bglA cat dapE dat ldh lhkA

3 9 9 3 3 1 5

- WGS data

- minimum spanning tree

- drawn using cgMLST data (1748 genes)

- 695 isolates

Food

Human

Production facility

Figures kindly provided by Anthony Smith

Centre for Enteric Diseases (CED),

National Institute for Communicable Diseases (NICD)

South Africa

L. monocytogenes ST6 Outbreak South Africa

- ST6 cluster

- 91% of human isolates

- <10 SNP differences among isolates

- highly related

• Environmental ST6 isolates

recovered from the food

production facility

• ST6 isolates recovered from

the company food samples

Outbreak of Listeria monocytogenes in South Africa, 2017-2018

- human isolates (n=706)

- WGS data

- maximum likelihood tree (linear tree)

- drawn using SNP profiles

• grey shading links isolates showing ≤7 allele differences

• indicative of highly related isolates

• indicative of epidemiological relatedness

Food, n=22

Human, n=328

Production facility, n=24

- WGS data

- minimum spanning tree

- drawn using cgMLST data

- ST6 isolates only (n=374)

Straincharacterization

Trace back

Genetic relateness

+ Epi and tracing back data

© ESCMID eL

ibrary

by au

thor

Multicountry Rapid Outbreak Assessment (ROAs)

16© ESCMID

eLibr

ary by

autho

r

EFSA WGS Capacity building

17

Online tools and existing platforms: e.g. ENTEROBASE, BIGSdb, CGE, INNUENDO, COMPARE, PATRIC, etc.

© ESCMID eL

ibrary

by au

thor

18

“Technical support to collect and analyse whole genome sequencing (WGS) data in the joint ECDC-EFSA molecular typing database”

at least L. monocytogenes, Salmonella, E.coli

Published May 2019

https://efsa.onlinelibrary.wiley.com/doi/epdf/10.2903/sp.efsa.2019.EN-1337

Other EFSA activities for foodborne pathogens

BIOHAZ PANEL“Self-tasking mandate for scientific opinion on the application and use of

next generation sequencing (including whole genome sequencing) for risk assessment of foodborne microorganisms”

Deadline October 2019

© ESCMID eL

ibrary

by au

thor

19

WGS for AMR:

© ESCMID eL

ibrary

by au

thor

VIM-1 occurrence in the food chain

E. coli ST88 clone

E. coli ST131

2011/2012 2015/2016 2017

Salmonella Infantis

Salmonella rough

Salmonella Infantis

Salmonella Goldcoast

E. coli ST88 clone

Enterobacter cloacae

Salmonella group C

E. coli ST7593

Salmonella?

E. coli ST48

Positive farms (1)Positive farms (5) Positive farms (3)

pRH-R3 like IncHI2 plasmid

Slide kindly provided by Jennie Fischer,

German Federal Institute for Risk Assessment

Roschanski et al. 2019. mSphere 4:e00089-19© ESCMID

eLibr

ary by

autho

r

21

Carbapenemase-produding bacteria in the food chain

© ESCMID eL

ibrary

by au

thor

Larsen et al. (2016) – www.danmap.org

Several human cases in the poultry associated clade reporteddirect contact with raw poultry meat

Figure kindly provided by Anders Rhod Larsen

Statens Serum Institute

Denmark

Livestock-associated MRSA

© ESCMID eL

ibrary

by au

thor

Livestock-associated MRSA at EFSA

© ESCMID eL

ibrary

by au

thor

Harmonized monitoring EU Legislation

24

Decision 2013/652/EU

2014 - 2020

Directive 2003/99/ECArt. 7(3) and 9(1) + Annexes II (B) IV

Zoonotic Bacteria

Salmonella spp.

C. jejuni / C. coli

ESBL-/AmpC-/CP-producing Salmonella

Indicator Bacteria

E. coli

E. faecalis / E. faecium

ESBL-/AmpC-/CP-producing E. coli

Bacterial species

Sampling

AST

MatrixSource

New Decision

2021 - …© ESCMID

eLibr

ary by

autho

r

EUSR-AMR Reference Testing (EURL-AR + MSs)

25

Selection of isolates:

Emerging resistances

Detection of clones

Discrepancies

Confirmation of results

Ask MSs for the isolates

Perform:

WGS, MIC re-testing

Seq. publicly available ENA

WGS analyses to support phenotypical AMR data: around 400 isolates selected sequnced by yearSequences released to ENA (with permission of the MSs)

MSs

© ESCMID eL

ibrary

by au

thor

From WGS testing, detection of…

26

Emerging AMR-Clones/Plasmids?

S. Infantis ESBL in ItalyS. Kentucky Cipro-R

Development of detection methods

Multi-plex PCR

Emerging AMR-Mechanisms?COL: mcr-1, mcr1.6, mcr-3, mcr-4, mcr-5ESBLs, PMQRsAZT: mefC-mphG tándemE. coli OXA-162

Figures kindly provided by and Antonio Battisti (ISZLT, Italy)

Figures kindly provided by Valeria Bortolaia (EURL-AR, DTU, Denmark)

© ESCMID eL

ibrary

by au

thor

27

Request for scientific and technical assistance on harmonised monitoring of antimicrobial resistance (AMR) in bacteria transmitted through food

Published June 2019

https://efsa.onlinelibrary.wiley.com/doi/epdf/10.2903/j.efsa.2019.5709

……..

To address the use of molecular typing methods

o To complement and/or replace the phenotypic methods

o To ensure the comparability between the results of technics

o To integrate molecular data with past/future phenotypicaldata

…This includes the use of WGS

EC Mandate Revision Techn. SpecificationsAMR MON

© ESCMID eL

ibrary

by au

thor

AMR tools? Benchmarking needed

28

Taken from ENGAGE final mtg. 01.2018.Published Final Report

© ESCMID eL

ibrary

by au

thor

EFSA Capacity building AMR benchmarking

29

beta-lactamase genes detection …..

Sometimes, the same tool does not detect all the genes (different input sequences different results)Some tool give you all similar genes….

Sample name File used Trimmed Assembler Pipeline beta-lactams

ERR2019187_EnaSPA ENA_reads no SPADES A blaCTX-M-15,blaTEM-1C

ERR2019187_EnaVelv ENA_reads no Velvet A

blaCTX-M-15,blaTEM-10 (1),blaTEM-10 (2),blaTEM-101 (1),blaTEM-101 (2),blaTEM-102

(1),blaTEM-102 (2),blaTEM-104 (1),blaTEM-104 (2),blaTEM-105 (1),blaTEM-105 (2),blaTEM-

106 (1),blaTEM-106 (2),blaTEM-107 (1),blaTEM-107 (2),blaTEM-108 (1),blaTEM-108 (2),blaTEM-

109 (1),blaTEM-109 (2),blaTEM-11 (1),blaTEM-11 (2),blaTEM-110 (1),blaTEM-110 (2),blaTEM-

111 (1),blaTEM-111 (2),blaTEM-112 (1),blaTEM-112 (2),blaTEM-113 (1),blaTEM-113 (2),blaTEM-

ERR2019187_SRAVelv SRA_reads no Velvet A

blaCTX-M-15,blaTEM-10 (1),blaTEM-10 (2),blaTEM-101 (1),blaTEM-101 (2),blaTEM-102

(1),blaTEM-102 (2),blaTEM-104 (1),blaTEM-104 (2),blaTEM-105 (1),blaTEM-105 (2),blaTEM-

106 (1),blaTEM-106 (2),blaTEM-107 (1),blaTEM-107 (2),blaTEM-108 (1),blaTEM-108 (2),blaTEM-

109 (1),blaTEM-109 (2),blaTEM-11 (1),blaTEM-11 (2),blaTEM-110 (1),blaTEM-110 (2),blaTEM-

111 (1),blaTEM-111 (2),blaTEM-112 (1),blaTEM-112 (2),blaTEM-113 (1),blaTEM-113 (2),blaTEM-

ERR2019187_T_EnaVelvENA_reads yes Velvet A blaCTX-M-15,blaTEM-1C

ERR2019187 ENA_assembly yes SPADES B blaCTX-M-15,TEM-171,TEM-2

ERR2019187 SRA_assembly yes SPADES B blaCTX-M-15,TEM-171,TEM-2

ERR2019187 ENA_reads no no? C blaCTX-M-15,

ERR2019187T ENA_reads yes no? C blaCTX-M-15,blaTEM-1B-like

ERR2019187 ENA_assembly yes SPADES C blaCTX-M-15,blaTEM-1B-like

ERR2019187 ENA_reads no SPADES D

blaCTX-M-15_23,CTX-M-33,OXA-368,TEM-3,TEM-33,TEM-4,blaTEM-102_1,blaTEM-104_1,

blaTEM-112_1,blaTEM-141_1

ERR2019187T ENA_reads yes SPADES D

blaCTX-M-15_23,CTX-M-33,OXA-368,TEM-3,TEM-33,TEM-4,blaTEM-102_1,blaTEM-104_1,

blaTEM-112_1,blaTEM-141_1

ERR2019187 Reftest WGS blaCTX-M-15,blaTEM-1B (100, 741 / 861)

© ESCMID eL

ibrary

by au

thor

30

Aminoglycoside genes detection …..

Nomenclature chaos depending on the databases….

EFSA Capacity building AMR benchmarking

Sample name File used Trimmed Assembler Pipeline aminoglycosidesERR2019166_Ena ENA_reads no SPADES A aac(3)-IIa,aadA1,aadA2,aph(3')-Ia,strA,strB

ERR2019166_Ena ENA_reads no Velvet A

aac(3)-IIa,aadA1 (1),aadA1 (2),aadA1 (3),aadA21 (1),aadA21 (2),aadA22 (1),aadA22 (2),aadA23

(1),aadA23 (2),aadA24 (1),aadA24 (2),aadA3 (1),aadA3 (2),aadA8 (1),aadA8 (2),aadA8b

(1),aadA8b (2),aph(3')-Ia,strA,strB

ERR2019166_SRA SRA_reads no Velvet A aac(3)-IIa,aadA1,aadA2,aph(3')-Ia,strA,strB

ERR2019166_T_Ena ENA_reads yes Velvet A

aac(3)-IIa,aadA1 (1),aadA1 (2),aadA1 (3),aadA15 (1),aadA15 (2),aadA21 (1),aadA21 (2),aadA22

(1),aadA22 (2),aadA23 (1),aadA23 (2),aadA24 (1),aadA24 (2),aadA3 (1),aadA3 (2),aadA8

(1),aadA8 (2),aadA8b (1),aadA8b (2),aph(3')-Ia,strA,strB

ERR2019166 ENA_assembly yes SPADES B AAC(3)-IIc,AAC(6')-Ib7,APH(3'')-Ib (strA),APH(3')-Ia,APH(6)-Id (strB),aadA,aadA3,acrD,kdpE

ERR2019166 SRA_assembly yes SPADES B AAC(3)-IIc,AAC(6')-Ib7,APH(3'')-Ib (strA),APH(3')-Ia,APH(6)-Id (strB),aadA,aadA3,acrD,kdpE

ERR2019166 ENA_reads no C aac(3)-IIa,aadA1-like,aadA2,aph(3')-Ic-like,strA,strB

ERR2019166T ENA_reads yes C aac(3)-IIa,aadA1-like,aadA2,aph(3')-Ia-like,APH(6)-Id (strB),strA

ERR2019166 ENA_assembly yes SPADES C aac(3)-IIa,aadA1-like,aadA2,aph(3')-Ia-like,APH(6)-Id (strB),strA

ERR2019166 ENA_reads no SPADES D

AAC(3)-IIa,AAC(6)-Ib,AAC(6)-Ib7,APH(3)-Ia,APH(3)-Ib,APH(6)-Id (strB),aac(3)-

IIa_1,aadA,aadA15,aadA1_2,aadA1_3,aadA1_4,aadA1_5,aadA2,aadA22,aadA23,aadA24,aadA2

_2,aadA8,aadA9,acrD,ant(3)-Ia_1,aph(3)-Ia_1,aph(3)-Ib_1,APH(6)-Id_1,kdpE,strA_1

ERR2019166T ENA_reads yes SPADES D

AAC(3)-IIa,AAC(6)-Ib,AAC(6)-Ib7,APH(3)-Ia,APH(3)-Ib,APH(6)-Id (strB),aac(3)-

IIa_1,aadA,aadA15,aadA1_2,aadA1_3,aadA1_4,aadA1_5,aadA2,aadA22,aadA23,aadA24,aadA2

_2,aadA8,aadA9,acrD,ant(3)-Ia_1,aph(3)-Ia_1,aph(3)-Ib_1,APH(6)-Id_1,kdpE,strA_1

ERR2019166 Reftest WGS aph(3')-Ic (99.14)/ aac(3)-Iia

© ESCMID eL

ibrary

by au

thor

Tech. specifications AMR MON recommendations

31

Proposed approach to integration of WGS … Incremental approach over 2021-2026

Voluntary use of WGS:

for Specific Monitoring of ESBL/AmpC/CP-producing E. coli (mandatory!)

For Confirmatory Testing using WGS

Harmonised protocol and quality criteria are needed! Several bioninformatic tools/pipelines/sequencing platforms

Various reference AMR gene databases

… can hamper comparative accuracy of WGS results.

-> EURL-AR to provide Harmonised Protocols/Quality Criteria and Training on DNA extraction, library preparation, sequencing in 2019-2020

For the moment...Report genes

Further developments in the future (re-assess regularly 2021-2026)

… To switch: EU-wide WGS data collection, leading to harmonized AMR mon.

Provided a high concordance WGS- genotyping fand AST … technology mature nd fully implemented … and MSs ready.

© ESCMID eL

ibrary

by au

thor

32

Regulated products

© ESCMID eL

ibrary

by au

thor

Regulated products

33

B. Guerra

© ESCMID eL

ibrary

by au

thor

Conclussions

34

WGS is useful for Food Safety…

• YES! A lot of information can be extracted from TAG, AGATA, CAT..

• We need to ask us always three questions • what do we want to know?• how can we achieve it? • which is the context?

• Different aims, similar/different bioinformatic tools needed.

• For surveillance/monitoring, harmonization is extremely important to allow comparability. The use of curated and updated data bases is a must.

• Detailed information on analysis done, tools, versions, parameters, etc. needed.

• Challenge: rapid and constant evolution of the techniques, tools, etc., can make reproducibility of the results difficult…

…reproducibility needs to be ensured© ESCMID

eLibr

ary by

autho

r

35

Beatriz GuerraLead Expert Molecular Epidemiology. Senior Scientific OfficerUnit on Biological Hazards and Contaminants (BIOCONTAM)Department of Risk Assessment and Scientific Assistance (RASA)

e-mail: beatriz.guerra@efsa.europa.eu

Thanks for your attention and thanks to…!

Ernesto Liébana, Valentina Rizzi, Pierre-Alexander Beloeil, Mirko Rossi, Elonora Sarmo, Montserrat Anguita, Jaime Aguilera, Nikoletta Papadopoulous (EFSA)

Iolanda Mangone (ISZ Teramo, EFSA Tasking Grant Bioinformatician)

Raquel Garcia Fierro and Daniel Thomas López (ex-EFSA trainees)

Jennie Fischer and Istvan Szabo (BfR), Lothar Beutin (formerly BfR, Berlin)

Anthony Smith (NICD, South Africa)

Kathryn Holt (London School of Hygiene & Tropical Medicine, Monash Univ., Melbourne)

Ander R Larsen (SSI, Copenhagen)

Valeria Bortolaia (DTU, Copenhagen)

Antonio Battisti (ISZLT, Rome)

© ESCMID eL

ibrary

by au

thor