APHL 2020 POSTER ABSTRACTS
Food Safety
PulseNet WGS workshops are hosted at CDC up to three times per year and , while standardized , are continually improved upon to meet the needs of the PHLs as WGS technology evolves and PHLs adapt their workflows . Surveys are provided to each attendee at the end of the workshop and the feedback is used to improve subsequent trainings . Training checklists , job aids and handouts have been developed to aid in both the administration of these courses and to assist in trainee learning . To date , PulseNet has trained over four hundred laboratory scientists on WGS methods and certified 65 PulseNet labs to perform WGS . In this way , CDC and PulseNet are ensuring that public health laboratories are utilizing the latest technology and state-of-art knowledge to improve their ability to detect , investigate , and stop foodborne illness .
Presenter : Angela Poates , Centers for Disease Control and Prevention , Atlanta , GA , nir9 @ cdc . gov
Targeted Enrichment of Pathogenic Escherichia coli DNA from Stool Samples for Metagenomic Subtyping Analysis
Y . Gao , M . Thakur , B . Aspinwall , K . Dillon , H . Carleton , A . J . Williams and A . Huang , Centers for Disease Control and Prevention , Atlanta , GA
Culture-independent diagnostic tests allow for efficient diagnoses of foodborne pathogens , but do not yield isolates critical for outbreak surveillance and investigation . Shotgun metagenomic sequencing allows pathogen subtyping directly from stool , but its practicality for routine public health surveillance activities is limited by the costs associated with deep sequencing and complex analytical pipelines required to resolve low pathogen signal to high stool background noise . For pathogens such as Shiga-toxin producing Escherichia coli ( STEC ), this is further complicated by the similarities between pathogenic STEC and commensal E . coli genomes . Shotgun metagenomics is further limited by challenges with phasing , or the ability to differentiate between , pathogenic and commensal E . coli sequences .
We tested the Roche HyperCap bait capture system to enrich for STEC DNA . The baits were designed by tiling across two target STEC O157 : H7 genomes ( F7353 and K4623 ), excluding the 16S rRNA genes . Enrichment was tested using mock disease stool , prepared by spiking health stool DNA with 2 % target STEC isolate DNA . To characterize phasing issues , we applied baits to DNA samples that were equal mixes of two different E . coli isolates that include the target STECs , a non-target STEC , and two commensal E . coli strains . To distinguish the different components , DNA libraries from stool and isolates were indexed separately before mixing . Samples were sequenced before and after bait capture using an Illumina MiSeq . The sequences were aligned against the reference genome F7353 using Bowtie2 and coverage was assessed using R . High quality SNP analysis was performed using lyve-SET ( https :// github . com / lskatz / lyve-SET ).
The proportion of STEC DNA in the mock disease sample increased from 2 % to 93 % after bait capture , and 97 % of the enriched STEC reads mapped to the reference STEC genome , indicating high target specificity . Bait captured STEC sequences also showed high breadth of coverage across the whole genome . The baits did not preferentially enrich for any one E . coli strain , but bias from the bait design was observed : coverage maps show that regions in the genome encoding mobile genetic elements were over-enriched . How this regional bias contributes to subtyping strains is currently being analyzed .
The bait capture system showed promising results in enriching STEC DNA from a complex stool background . From low levels of pathogen input , the enriched DNA showed high breadth whole genome coverage that is necessary for subtyping using SNP analysis . However , because the baits cannot distinguish different strains of E . coli , they can also enrich for commensal E . coli , making subtyping analyses more difficult . Ongoing work includes performing SNP analysis on data generated from bait captured samples containing two different E . coli strains and understanding how bait bias may affect analysis .
Presenter : Yang Gao , Centers for Disease Control and Prevention , Atlanta , GA , nrj0 @ cdc . gov
An Outbreak of E . coli O103 Traced to Sprouts : Implications for the FSMA Produce Safety Rule
O . Oni 1 , K . Torkelson 1 , A . Garvey 1 , G . Kline 2 , R . Jepson 2 , C . Lord 2 , N . Hall 2 , M . Pentella 2 , M . Speltz 1 , T . Nguyen 1 ; 1 Iowa Department of Public Health , Des Moines , IA , 2 State Hygienic Laboratory at the University of Iowa , Iowa City , IA
Sprouts are often linked to foodborne outbreaks because of the warm humid conditions necessary to produce them . In this outbreak , Iowans became ill with E . coli O103 after eating at a restaurant chain in central and eastern Iowa . This outbreak was identified through routine surveillance on December 18 , 2019 . After notification , Iowa ’ s food retail and manufacturing agency , the Iowa Department of Inspection and Appeals ( DIA ), notified the franchise owners and sprout grower . Both parties voluntarily suspended the sales of clover sprouts .
A cohort study was conducted . Phone interviews were performed by Iowa Department of Public Health ( DPH ) epidemiologists and local public health partners using a standard E . coli questionnaire and an outbreak-specific supplemental questionnaire . Spent irrigation water ( SIW ) and 16 packages of clover sprout samples were collected and sent to Iowa ’ s State Hygienic Laboratory ( SHL ) for testing .
Twenty-three illnesses were identified . Of these , 22 were laboratory confirmed and one probable . A confirmed case was defined as a person with E . coli 0103 infection , whose clinical isolate is related within 0-2 SNP ’ s , and with an isolate collection date between 11 / 26 / 2019 – 12 / 21 / 2019 . A total of 22 ( 96 %) cases met this case definition from 10 Iowa counties . Of these , 14 ( 64 %) were females . The median age among cases was 29 ( range 18-50 years ). One case was hospitalized . No deaths were reported .
Statistical analysis was performed to identify the specific food that caused the illness . Clover sprouts were found to be significant with 9 / 20 ( 45 %) of the cases reporting consuming sprouts . This is significantly higher than the average of 4 % of people that mentioned eating sprouts based on the 2006 – 2007 CDC FoodNet Population Survey .
All clinical and environmental isolates submitted to SHL were identified as carrying the stx1a toxin gene using real-time PCR , serotyped by conventional agglutination , and sequenced using PulseNet procedures . SHL identified E . coli O103 from 12 / 16 clover sprouts packages and corresponding SIW . Both the environmental samples were highly related to the 22 human clinical isolates by whole genome sequencing ( 0-2 SNP ’ s ).
The FSMA Produce Safety rule establishes science-based minimum standards for the safe growing , harvesting , packing and holding of fruits and vegetables grown for human consumption which includes sprout . Sprout growing operations are subject to sprout-