Novel application of metagenomics for the strain-level detection of bacterial contaminants within non-sterile industrial products

The home and personal care (HPC) industry generally relies on initial cultivation and subsequent biochemical testing for the identification of microorganisms in contaminated products. This process is slow (several days for growth), labour intensive, and may miss organisms which fail to revive from the harsh environment of preserved consumer products. Since manufacture within the HPC industry is high-throughput, the process of identification of microbial contamination could benefit from the multiple cultivation-independent methodologies that have developed for the detection and analysis of microbes. We describe a novel workflow, starting with automated DNA extraction directly from a HPC product, which applies metagenomic methodologies for species and strain-level identification of bacteria. The workflow was validated by application to a historic microbial contamination of a general-purpose cleaner (GPC). A single strain of Pseudomonas oleovorans was detected metagenomically within the product, and this mirrored the genome of the contaminant isolated in parallel by a traditional cultivation-based approach. Using a dilution series of the incident sample, we also provide evidence to show that the workflow enables detection of contaminant organisms down to 100 bacterial cells per mL of product. To our knowledge, this is the first validated example of metagenomics analysis providing confirmatory evidence of a traditionally-isolated contaminant organism, in a HPC product.