Dysbiosis, a new medical entity ?

Knowledge of the "healthy" microbiota and its intra/interindividual variability is an essential prerequisite for defining a "pathological/pathogenic" microbiota. This definition must be made on the scale of the microbiota of each colonized site (intestine, skin, oral cavity, respiratory tract, vagina), as intersite differences on an individual scale are far more important than intrasite differences on an inter-individual scale. The parameters for defining dysbiosis are therefore difficult to define, and the field is constantly evolving. The most visible aspect is the reduction in diversity.

Furthermore, understanding the composition-function relationships of the intestinal microbiota is a key issue. Studies focusing exclusively on the genus/species composition of microbiota (16S rRNA) cannot provide unambiguous information on function. Microbiota of very different composition can be functionally close, with some taxonomically distinct species sharing common functions (particularly metabolic).

Experimental validation of the pathogenic properties of flora is difficult in humans. The correlative value of the data is enhanced when working with twins (homozygotes). Human-animal microbiota transfer is possible. A pathogenic effect is often observed, but extending the results to humans remains problematic. Experiments involving the contamination of mice by dysbiotic flora through co-housing are easy in mice (coprophagia), but in humans, cases of transmission are rare and debatable (as far as studied).

There's a big difference between an experiment in gnotoxic animals, where a candidate species will appear pathogenic (pathobionte), and one that isn't when it's part of a complex flora. The threshold and significance of the imbalance are difficult to determine.

These studies nevertheless offer hope of identifying dysbiotic "signatures" or "biomarkers" of pathological situations. To achieve this, we need to step up our efforts in metagenomic analysis "at the patient's bedside", at the genomic level, using deep sequencing. The difficulties highlighted in identifying a microbial species or genus signature could lead to the emergence of biomarkers that are not microbiological, but metabolic and/or immunological.

Despite these reservations, "pharmacobiotics" is developing with the aim of manipulating the intestinal microbiota for therapeutic purposes. Encouraging data exist in experimental models, but there is little evidence of success in the human clinic. Four approaches are being considered:

• probiotics, which are live microorganisms that, when ingested in adequate quantities, are expected to confer a health benefit on the host(Lactobacillus, Streptococcus, Bifidobacterium and others);
• prebiotics: food ingredients, essentially non-absorbable sugars (oligofructose, inulin, galacto-oligosaccharides, lactulose, breast milk oligosides or HMOs), which are thought to selectively promote the growth of certain probiotic bacteria or the activity of the microbiota;
• postbiotics : new developments in the metagenomics and metabolomics of the intestinal microbiome and host response mean that these new approaches can identify postbiotics, i.e. molecules that act as effectors of any beneficial effects;
• fecal transplantation, which is currently the only approach with satisfactory results, particularly in preventing recurrences of Clostridium difficile pseudomembranous colitis .