Fishing in the Soup - Pathogen Detection in Food Safety Using Metabarcoding and Metagenomic Sequencing.

Anne Busch,Herbert Tomaso,Simon H Tausch,Jens Andre Hammerl,Burkhard Malorny,Carlus Deneke,Josephine Grützke

doi:10.3389/fmicb.2019.01805

Abstract

In food safety the detection of food contaminations with pathogenic microorganisms is a race against time and often outpaced by error-prone epidemiological approaches. For evidence-based outbreak investigations fast and reliable techniques and procedures are required to identify the source of infection. Metagenomics has the potential to become a powerful tool in the field of modern food safety, since it allows the detection, identification and characterization of a broad range of pathogens in a single experiment without pre-cultivation within a couple of days. Nevertheless, sample handling, sequencing and data analysis are challenging and can introduce errors and biases into the analysis. In order to evaluate the potential of metagenomics in food safety, we generated a mock community containing DNA of foodborne bacteria. Herewith, we compare the aptitude of the two prevalent approaches – 16S rDNA amplicon sequencing and whole genome shotgun sequencing – for the detection of foodborne bacteria using different parameters during sample preparation, sequencing and data analysis. 16S rDNA sequencing did not only result in high deviations from the expected sample composition on genus and species level, but more importantly lacked the detection of several pathogenic species. While shotgun sequencing is more suitable for species detection, abundance estimation, genome assembly and species characterization, the performance can vary depending on the library preparation kit, which was confirmed for a naturally Francisella tularensis contaminated game meat sample. The application of the Nextera XT DNA Library Preparation Kit for shotgun sequencing did not only result in lower reference genome recovery and coverage, but also in distortions of the mock community composition. For data analysis, we propose a publicly available workflow for pathogen detection and characterization and demonstrate its benefits on the usability of metagenomic sequencing in food safety by analyzing an authentic metagenomic sample.

Highlights

During foodborne outbreaks reliable techniques are required to identify the source of infection as fast as possible to prevent further infections with the causative agent
In addition to obligate human pathogenic bacteria that cause foodborne illness when ingested, closely related opportunistic or nonpathogenic bacteria were chosen in order to analyze the ability to dissect pathogenic and nonpathogenic species from one another during the analysis (Figure 1)
The predominant metagenomic sequencing methods 16S rDNA amplicon and shotgun sequencing as well as several parameters that can distort the analysis including variable regions of the 16S rRNA gene, library preparation protocols, sequencing platform, sequencing depths, clustering, taxonomic classification tools and sequence databases were tested. 16S rDNA amplicon data is generated and analyzed in many studies and it can be useful for the detection of pathogenic bacteria in foodstuffs because the dominant eukaryotic DNA originating from the food matrix is excluded

Summary

Introduction

During foodborne outbreaks reliable techniques are required to identify the source of infection as fast as possible to prevent further infections with the causative agent. Targeted screening methods as PCR or ELISA can be directly applied without a cultivation step and are very fast but carry the risk of missing atypical strains that are not covered by the applied method These methods do not resolve the affiliation of the detected pathogen to an ongoing outbreak due to a low resolution on molecular level. Whole genome sequencing (WGS) requires a cultivation step in order to receive an isolate from the patient and the contaminated food that are sequenced by next-generation sequencing (NGS). This method has a high discriminatory power and can be used to decipher between outbreak relevant and -irrelevant strains. The usage of sequencing-based metagenomics allows the simultaneous identification and typing of the causative agent as well as antimicrobial resistance (AMR) or virulence genes and promises to be a very powerful tool for the surveillance of food and drinking water

Objectives

Methods

Results

Conclusion