Insights into the assessment of highly pathogenic Shiga toxin-producing Escherichia coli in raw milk and raw milk cheeses by High Throughput Real-time PCR

High throughput qPCR analyses suggest that Enterobacterales of French sheep and cow cheese rarely carry genes conferring resistances to critically important antibiotics for human medicine

Bacteria present in raw milk can carry acquired or intrinsic antimicrobial resistance genes (ARGs) and mobile genetic elements (MGEs). However, only a few studies have evaluated raw milk cheese as a potential reservoir of ARGs. This study thus aimed at providing new data regarding resistance markers present in raw milk cheese. Sheep (n = 360) and cow (n = 360) cheese samples produced in France were incubated in buffered peptone water supplemented with acriflavin or novobiocin; as corroborated by 16S metabarcoding, samples were enriched in Gram-negative bacteria since Escherichia coli and Hafnia alvei respectively accounted for 40 % and 20 % of the samples' microbiota. Screening of the samples for the presence of 30 ARGs and 16 MGEs by high throughput qPCR array showed that nine ARGs conferring resistances to 1st-generation beta-lactams, aminoglycosides, trimethoprim/sulfonamides and tetracyclines occurred in >75 % of both sheep and cow samples. This is neither surprising nor alarming since these resistance genes are widely spread across the One Health human, animal and environmental sectors. Conversely, genes conferring resistances to last-generations cephalosporins were rarely identified, while those conferring resistances to carbapenems or amikacin, which are restricted to human use, were never detected. Multiple MGEs were detected, the most frequent ones being IncF plasmids, confirming the potential transmission of ARGs. Our results are in line with the few studies of the resistome of milk or milk cheese showing that genes conferring resistances to 1st-generation beta-lactams, aminoglycosides and tetracyclines families are widespread, while those conferring resistances to critically important antibiotics are rare or absent.

Combination of whole genome sequencing and supervised machine learning provides unambiguous identification of eae-positive Shiga toxin-producing Escherichia coli

Introduction

The objective of this study was to develop, using a genome wide machine learning approach, an unambiguous model to predict the presence of highly pathogenic STEC in E. coli reads assemblies derived from complex samples containing potentially multiple E. coli strains. Our approach has taken into account the high genomic plasticity of E. coli and utilized the stratification of STEC and E. coli pathogroups classification based on the serotype and virulence factors to identify specific combinations of biomarkers for improved characterization of eae-positive STEC (also named EHEC for enterohemorrhagic E.coli) which are associated with bloody diarrhea and hemolytic uremic syndrome (HUS) in human.

Methods

The Machine Learning (ML) approach was used in this study on a large curated dataset composed of 1,493 E. coli genome sequences and 1,178 Coding Sequences (CDS). Feature selection has been performed using eight classification algorithms, resulting in a reduction of the number of CDS to six. From this reduced dataset, the eight ML models were trained with hyper-parameter tuning and cross-validation steps.

Results and discussion

It is remarkable that only using these six genes, EHEC can be clearly identified from E. coli read assemblies obtained from in silico mixtures and complex samples such as milk metagenomes. These various combinations of discriminative biomarkers can be implemented as novel marker genes for the unambiguous EHEC characterization from different E. coli strains mixtures as well as from raw milk metagenomes.

A step forward for Shiga toxin-producing Escherichia coli identification and characterization in raw milk using long-read metagenomics

Shiga toxin-producing Escherichia coli (STEC) are a cause of severe human illness and are frequently associated with haemolytic uraemic syndrome (HUS) in children. It remains difficult to identify virulence factors for STEC that absolutely predict the potential to cause human disease. In addition to the Shiga-toxin (stx genes), many additional factors have been reported, such as intimin (eae gene), which is clearly an aggravating factor for developing HUS. Current STEC detection methods classically rely on real-time PCR (qPCR) to detect the presence of the key virulence markers (stx and eae). Although qPCR gives an insight into the presence of these virulence markers, it is not appropriate for confirming their presence in the same strain. Therefore, isolation steps are necessary to confirm STEC viability and characterize STEC genomes. While STEC isolation is laborious and time-consuming, metagenomics has the potential to accelerate the STEC characterization process in an isolation-free manner. Recently, short-read sequencing metagenomics have been applied for this purpose, but assembly quality and contiguity suffer from the high proportion of mobile genetic elements occurring in STEC strains. To circumvent this problem, we used long-read sequencing metagenomics for identifying eae-positive STEC strains using raw cow's milk as a causative matrix for STEC food-borne outbreaks. By comparing enrichment conditions, optimizing library preparation for MinION sequencing and generating an easy-to-use STEC characterization pipeline, the direct identification of an eae-positive STEC strain was successful after enrichment of artificially contaminated raw cow's milk samples at a contamination level as low as 5 c.f.u. ml^-1. Our newly developed method combines optimized enrichment conditions of STEC in raw milk in combination with a complete STEC analysis pipeline from long-read sequencing metagenomics data. This study shows the potential of the innovative methodology for characterizing STEC strains from complex matrices. Further developments will nonetheless be necessary for this method to be applied in STEC surveillance.