Identifying and counting: these two objectives are therefore met or at least almost met because it is not yet possible to count every bacteria to the nearest number but rather it is only possible to give them assessable proportions in relation to each other. However, it is important to precisely know these quantities to the nearest number so that this objective becomes the subject of future research work by Prof. Georges Daube. “The idea is to bring about an end to the typical reflex on the part of the food industry, which tends to consider that the bacteria present in quantities in excess of one million are bad and cause the food to degrade. This is not true however. Some bacteria, although present in large proportions in a food are neutral or could contribute to its taste or smell instead of causing it to deteriorate”.

Practical applications

The first example of an application: “Targeted metagenomics can be used to validate the use-by dates for consumption of food products”, explains George Daube enthusiastically. Surprisingly perhaps, this has remained a very empirical method until today: the product is allowed to age, is tasted daily and the use-by date is fixed at a few days before the first signs of alteration. Given that, at a bacteriological level, it was only possible to detect the presence or absence of certain bacteria or the extent of poorly-defined populations of bacteria in a given food, the food manufacturer tends to systematically lower the quantity below the threshold of one million (beyond which it is generally acknowledged that a bacteria could potentially cause the degradation of a food), with direct consequences for the use-by date: “reducing a given type of germ means leaving a space for other germs to multiply. In a food, the bacteria are somewhat in competition with each other”. Take the example of the Leuconostoc bacteria, which gives a starchy quality to foods and is also used in certain yoghurts to make them creamy. However, this same bacteria on slices of cooked ham placed in the fridge can multiply in cold conditions and cause a viscous coating to appear on the slices and this is often the first sign that the ham has been altered. In the first case, Leuconostoc is therefore a bacteria that is to be favored, and in the second, to be inhibited. “We are therefore far from the “everything sterile” argument. Some foods have to be able to contain an important number of micro-organisms because they protect the product. Targeted metagenomics will make it possible to favor bioprotection of foods in the laboratory”.

Another application: monitoring fermentation. Fermented products are products in which bacteria but also yeast or even molds can develop. Cheese, sausage, and choucroute: these are all products which would not exist if bacteria were not freely allowed to bring about their fermentation. “But food manufacturing takes place today without the manufacturer really knowing which bacteria multiply during fabrication and which bacteria will be in the finished product. In other words, the manufacturer has been the inheritor of a trial and error process over the years. His fabrication process is not checked simply because until today there was no checking method in existence”. Targeted metagenomics today makes it possible to objectify the knowledge of the cheese-maker. “Today’s cheese-maker knows that he is potentially no longer the only one to know his products. Being able to better identify the composition of food enables him to modify his recipe or even — and a certain interest in this activity was communicated to us — to buy cheese from his competitors to determine its composition”.