Microbiome-based environmental monitoring of a dairy processing facility highlights the challenges associated with low microbial-load samples

Abstract

Efficient and accurate identification of microorganisms throughout the food chain can potentially allow the identification of sources of contamination and the timely implementation of control measures. High throughput DNA sequencing represents a potential means through which microbial monitoring can be enhanced. While Illumina sequencing platforms are most typically used, newer portable platforms, such as the Oxford Nanopore Technologies (ONT) MinION, offer the potential for rapid analysis of food chain microbiomes. Initial assessment of the ability of rapid MinION-based sequencing to identify microbes within a simple mock metagenomic mixture is performed. Subsequently, we compare the performance of both ONT and Illumina sequencing for environmental monitoring of an active food processing facility. Overall, ONT MinION sequencing provides accurate classification to species level, comparable to Illumina-derived outputs. However, while the MinION-based approach provides a means of easy library preparations and portability, the high concentrations of DNA needed is a limiting factor.

Fig. 1. Mock community analysis.

MinION sequencing followed by MEGAN taxonomic classification of a simple mock community. a Taxonomic classification following 16S and whole metagenome (WMGS) sequencing. Also shown are expected relative abundances. b De novo assembly of genomes by the canu assembler, followed by mapping back to original known genomes, to illustrate coverage at 97% identity. 4 genomes, with 6 plasmids illustrated, of which 4 genomes and 5 plasmids had sequences aligned at 97% identity.

Fig. 2. Schematic of dairy processing facility sampling areas.

Dairy processing facility schematic includes the 8 areas sampled in each of October, November, and December 2018. Areas were sampled post CIP and prior to the recommencement of processing.

Fig. 3. Species level classification of MinION and NextSeq sequenced environmental samples.

Taxonomic assignment of MinION and NextSeq sequenced samples generated following the use of different pre-processing and sequencing methods. Pre-processing methods include MDA amplification, no pre-processing (NPP), and spread plating on BHI before washing colonies, pelleting, and treating as a metagenomic sample (Plate). Species level classification was performed using LAST (for MinION) and Diamond (for NextSeq) alignment of reads against the NR database and classification with MEGAN (LR for MinION). Species present in at least 5% in at least one sample are shown.

Fig. 4. Diversity analysis.

a Shannon and Simpson alpha diversity analysis. b Bray Curtis multidimensional scaling (MDS) beta diversity analysis. Boxplot centre line, median; box limits, upper and lower quartiles; whiskers, ×1.5 interquartile range; points, outliers (*** = p < 0.001, ** = p < 0.01, * = p < 0.05). Controls are excluded from these calculations and figures. MDA MinION, refers to samples whose DNA was subject to MDA before sequenced on an Oxford Nanopore MinION sequencer. MDA NextSeq, refers to samples whose DNA was subject to MDA before sequenced on a Illumina NextSeq sequencer. NPP NextSeq, refers to samples whose DNA had no pre-processing (NPP) in terms of amplification before sequencing on an Illumina NextSeq sequencer. Plate NextSeq, refers to samples whose metagenomic DNA was extracted from easily cultured microorganisms before sequencing on an Illumina NextSeq sequencer.

Fig. 5. Genus level classification of environmental samples and controls following different pre-processing methods and sequencing methods.

MEGAN LCA based genera level classification of MinION and NextSeq sequences. Also depicted are Sanger results to genus level for morphologically different colonies from each sample (TBC) along with thermophilic sporeformer enriched (ST) and mesophilic sporeformer enriched (SM) counts. Also included are CFU/swab counts for each culturing type. Sanger results represent relative abundance of a subset of morphologically distinct isolates rather than total isolates.

Fig. 6. Significant differences in the relative abundance of taxa based on processing and sequencing method.

a Significant species level differences due to sequencing and processing methods on environmental samples. Controls are excluded from these calculations and figures. b Significant genera level differences in relative abundance due to sequencing and processing methods on environmental samples. Controls are excluded from these calculations and figures. Boxplot centre line, median; box limits, upper and lower quartiles; whiskers, 1.5x interquartile range; points, outliers (*** = p < 0.001, ** = p < 0.01, * = p < 0.05).