In suckling mammals, the onset of solid food ingestion is coincident with the maturation of the gut barrier. This ontogenic process is driven by the colonization of the intestine by the microbiota. However, the mechanisms underlying the microbial regulation of the intestinal development in early life are not fully understood. Here, we studied the co-maturation of the microbiota (composition and metabolic activity) and of the gut barrier at the suckling-to-weaning transition by using a combination of experiments in vivo (suckling rabbit model), ex vivo (Ussing chambers) and in vitro (epithelial cell lines and organoids). The microbiota composition, its metabolic activity, para-cellular epithelial permeability and the gene expression of key components of the gut barrier shifted sharply at the onset of solid food ingestion in vivo, despite milk was still predominant in the diet at that time. We found that cecal content sterile supernatant (i.e. containing a mixture of metabolites) obtained after the onset of solid food ingestion accelerated the formation of the epithelial barrier in Caco-2 cells in vitro and our results suggested that these effects were driven by the bacterial metabolite butyrate. Moreover, the treatment of organoids with cecal content sterile supernatant partially replicated in vitro the effects of solid food ingestion on the epithelial barrier in vivo. Altogether, our results show that the metabolites produced by the microbiota at the onset of solid food ingestion contribute to the maturation of the gut barrier at the suckling-toweaning transition. Targeting the gut microbiota metabolic activity during this key developmental window might therefore be a promising strategy to promote intestinal homeostasis.

Metabarcoding of the 16S rRNA gene is commonly used to characterize microbial communities, by estimating the relative abundance of microbes. Here, we present a method to retrieve the concentrations of the 16S rRNA gene per gram of any environmental sample using a synthetic standard in minuscule amounts (100 ppm to 1% of the 16S rRNA sequences) that is added to the sample before DNA extraction and quantified by two quantitative polymerase chain reaction (qPCR) reactions. This allows normalizing by the initial microbial density, taking into account the DNA recovery yield. We quantified the internal standard and the total load of 16S rRNA genes by qPCR. The qPCR for the latter uses the exact same primers as those used for Illumina sequencing of the V3‐V4 hypervariable regions of the 16S rRNA gene to increase accuracy. We are able to calculate the absolute concentration of the species per gram of sample, taking into account the DNA recovery yield. This is crucial for an accurate estimate as the yield varied between 40% and 84%. This method avoids sacrificing a high proportion of the sequencing effort to quantify the internal standard. If sacrificing a part of the sequencing effort to the internal standard is acceptable, we however recommend that the internal standard accounts for 30% of the environmental 16S rRNA genes to avoid the PCR bias associated with rare phylotypes. The method proposed here was tested on a feces sample but can be applied more broadly on any environmental sample. This method offers a real improvement of metabarcoding of microbial communities since it makes the method quantitative with limited efforts.