Effect of Bifidobacterium crudilactis and 3'-sialyllactose on the toddler microbiota using the SHIME® model.

The important changes in diet during the first years of life strongly modulate the intestinal microbiota of young children. Among in vitro digestive models, the simulator of human intestinal microbial ecosystem (SHIME®) model, seems particularly adapted to study the effects of prebiotics and/or probiotics on the dynamic microbiota of toddlers. The main purpose of this study was to investigate different formulations with prebiotic (3'-sialyllactose: 3'SL) and probiotic (Bifidobacterium crudilactis FR/62/b/3) effects on young child microbiota using the SHIME® model. The ascending (AC), transverse (TC) and descending (DC) colons of the SHIME® model were inoculated with feces from 3 donors aged between 1 and 2 years, in three separate vessels. After two weeks of microbiota stabilization, four treatments of one week (prebiotic, probiotic, synbiotic and cell-free spent media from the synbiotic) were administered. In all the colon vessels, the short chain fatty acid analyses, determined using high-performance liquid chromatography highlighted a ratio acetate/propionate/butyrate proportion of 5:19:6, situated between infant and adult normal values. As already observed in other validated studies focusing on the SHIME® model, the 16S rDNA sequencing highlighted a low richness and diversity in the AC, while the microbial communities in the TC and the DC remained similar to each other. Although some bacteria involved in biofilm development have been identified (Stenotrophomonas, Megasphera and Enterobacter), specific bacterial populations, proper to each colon were developed. Some bacteria associated to the upper intestinal tract, such as Lactobacillus and Veillonella genera, seemed to grow easily in the AC. The quantitative polymerase chain reaction (qPCR) targeting the hsp60 gene confirmed the ability of bifidobacteria to survive in this toddler model. In addition, the synbiotic treatment tended to a bifidogenic effect (P < 0.1). On the other hand, the feces of the donors and the content of the three colon vessels were filtered and placed in contact with Escherichia coli O157:H7 ATCC 43890 to evaluate the modulation of virulence gene expression using reverse transcription PCR. Finally, filtered supernatants from donor feces significantly up-regulated the expression of the luxS gene of E. coli O157:H7 (P = 0.013). In conclusion, despite the presence of biofilms, the toddler SHIME® model used in his study shared characteristics found both in adults and infants. Although additional investigations should be performed, combining 3'SL and B. crudilactis FR/62/b/3 could lead to a beneficial effect on infant microbiota by favoring bifidobacterial presence. Finally, the filtrated supernatant from young child feces could be able to modulate the quorum sensing mechanism for E. coli O157:H7.