Direct identification and molecular characterization of zoonotic hazards in raw milk by metagenomics using Brucella as a model pathogen. Issue 5 (4th May 2021)

• Record Type: Journal Article
• Authors: Grützke, Josephine; Gwida, Mayada; Deneke, Carlus; Brendebach, Holger; Projahn, Michaela; Schattschneider, Alexander; Hofreuter, Dirk; El-Ashker, Maged; Malorny, Burkhard; Al Dahouk, Sascha
• Is Part Of: Microbial genomics. Volume 7:Issue 5(2021)
• Publication Date: 2021-05-04
• Subjects: Brucella -- detection limits -- host depletion -- metagenomics -- raw milk; Microbial genomics -- Periodicals; 572.8629
• DOI: 10.1099/mgen.0.000552
• Languages: English
• ISSNs: 2057-5858

Abstract

Abstract : Metagenomics is a valuable diagnostic tool for enhancing microbial food safety because (i) it enables the untargeted detection of pathogens, (ii) it is fast since primary isolation of micro-organisms is not required, and (iii) it has high discriminatory power allowing for a detailed molecular characterization of pathogens. For shotgun metagenomics, total nucleic acids (NAs) are isolated from complex samples such as foodstuff. Along with microbial NAs, high amounts of matrix NAs are extracted that might outcompete microbial NAs during next-generation sequencing and compromise sensitivity for the detection of low abundance micro-organisms. Sensitive laboratory methods are indispensable for detecting highly pathogenic foodborne bacteria like Brucella spp., because a low infectious dose is sufficient to cause human disease through the consumption of contaminated dairy or meat products. In our study, we applied shotgun metagenomic sequencing for the identification and characterization of Brucella spp. in artificially and naturally contaminated raw milk from various ruminant species. With the depletion of eukaryotic cells prior to DNA extraction, Brucella was detectable at 10 bacterial cells ml −1, while at the same time microbiological culture and isolation of the fastidious bacteria commonly failed. Moreover, we were able to retrieve the genotype of a Brucella isolate from a metagenomic dataset, indicating the potential of metagenomics for outbreak investigations using SNPs and core-genome multilocus sequence typing (cgMLST). To improve diagnostic applications, we developed a new bioinformatics approach for strain prediction based on SNPs to identify the correct species and define a certain strain with only low numbers of genus-specific reads per sample. This pipeline turned out to be more sensitive and specific than Mash Screen. In raw milk samples, we simultaneously detected numerous other zoonotic pathogens, antimicrobial resistance genes and virulence factors. Our study showed that metagenomics is a highly sensitive tool for biological risk assessment of foodstuffs, particularly when pathogen isolation is hazardous or challenging.