Metagenomic analysis as a tool to better characterize the bacterial content of food and food preparations.

Metagenomic analysis is a new culture-independent approach for assigning a taxonomic, genic or functional identity to bacterial DNA fragments of unknown origin. Its power and utility is increasingly rising thanks to the next generation sequencing techniques. It is now mature and cheap enough to be transposed to more applied fields like the food microbiology. We demonstrated in several studies the extraordinary potential of the targeted metagenomic analysis to different problematics related to food products. First, this approach is highly useful for the validation of the shelf life of food products. We analyzed standardized pork minced meat and meat product samples packaged either under modified atmosphere (MAP - 30% CO2, 70% O2) or under permeable atmosphere packaging, stored at different temperature (4°C, 4-8°C and 12°C) until the end of shelf-life. The metagenomic analysis allowed to identify species of all the sub-dominant bacterial populations. This approach showed why MAP can improve meat quality by favoring certain species rather than others. As a second example, we sought to identify the potential spoiling bacteria in several food products like raw fish, rind cheese or vacuum packed beef meats in order to illustrate the usefulness of metagenomics for the quality control of food preparations. Samples from various food matrices were screened to identify the bacterial contaminants. We combined the bioinformatics analysis with a classical approach to generate effective quantitative data for the various bacterial populations detected. This analysis characterizes the samples both on the identity of the potential spoiling bacteria present and on the quantification level of the contaminants. Finally, the metagenomic analysis reveals the presence of numerous uncultured and uncharacterized bacteria. The use of a carefully designed analysis pipeline has been used to ensure to label the bacterial population with a precise taxonomic identity and to determine whether the targeted population corresponds to a known species or not. This way, even if the nearest known homologous sequence is an environmental sample, its relatedness to known species can be deduced. This represents a new tool to trace yet uncharacterized food spoiling bacteria.