Saccharomyces boulardii CNCM I-745 mitigates antibiotic-induced gut microbiome functional alterations independently of the host

07 - Novembre - 2025

Zhan Huang, Loic Brot, Rand Fatouh, Marius Bredon, Laura Creusot, Antoine Lefèvre, Antonin Lamazière, Jérémie H Lefevre, Patrick Emond, Julien Planchais, Xavier Roux, Harry Sokol, Nathalie Rolhion

Gut Microbes. 2025 Dec 31;17(1):2575924


ABSTRACT:

The probiotic Saccharomyces boulardii CNCM I-745 (Sb) is widely prescribed to alleviate antibiotic-induced diarrhea, yet its mode of action, particularly its potential direct effects on the gut microbiome, remains incompletely defined. This study aimed to evaluate whether Sb can directly mitigate antibiotic-induced gut microbiota dysbiosis and influence downstream host immune response. Using both static (MiPro) and dynamic (SHIME®) in vitro gut microbiota models, we assessed the effects of Sb supplementation under antibiotic treatment with amoxicillin/clavulanic acid (AMC) or vancomycin (Van). Quantitative microbiome profiling integrated with targeted metabolomics showed that Sb helped stabilize bacterial biomass, partially preserved metabolic functions, and restored the production of immunoregulatory metabolites propionate and indole-3-propionic acid under AMC treatment. In addition, ex vivo exposure of primary human immune cells (PBMCs) and intestinal mucosal tissue to microbiota modulated by Sb led to a significant reduction in pro-inflammatory cytokine secretion compared to microbiota not supplemented with Sb. Collectively, these results support a beneficial role for S. boulardii CNCM I-745 in preserving directly gut microbiome function and supporting host immune homeostasis during antibiotic treatment, particularly under AMC exposure. Our findings advance the understanding of probiotic-antibiotic-gut microbiome interactions, thereby guiding future optimization of microbiome-targeted adjuvant therapies.Sb supplementation ameliorates antibiotic-induced alterations in bacterial loads and metabolic outputs in a dose-dependent manner. (a) Schematic representation of in vitro culture of human fecal microbiota in the static MiPro model. Stool samples from eight healthy human donors were treated with AMC or Van (0.1 mg/mL) and supplemented with Sb (2 mg/mL or 4 mg/mL). The control group received no treatments. After 24 h of incubation, samples were collected for further analyses (Methods). (b) Bacterial load across different groups. (c),(d) Absolute abundance of bacteria at the phylum and genus level. Quantitative results were calculated by combining sequencing data with bacterial load measurements (Methods). Data were expressed as the mean ± standard error of the mean for each taxon. (e) Coefficient from MaAsLin analysis representing absolute abundances of genera with significant differences across different comparisons in AMC-treated samples. Only genera with a prevalence more than 49.9% and a q-value below 0.1 were shown. (f) Levels of primary bile acids. Cholic acid (CA) and chenodeoxycholic acid (CDCA) were present in the MiPro culture medium. (g) Levels of short-chain fatty acids. (h) Levels of tryptophan metabolites. Each color (b,f) represents one stool donor in MiPro. Levels of significance (b, f–h) were determined using paired one-way ANOVA followed by Bonferroni's post hoc test. *p < 0.05; **p < 0.01; ****p < 0.0001.Effects of Sb supplementation on the gut microbiome are associated with a reduced pro-inflammatory potential. (a) Levels of TNF-α after stimulation with MiPro supernatants in human PBMCs. Each color represents one stool donor in MiPro. (b) Heatmap visualization of inflammatory cytokine levels, scaled by Z-scores, after stimulation with representative MiPro culture supernatants in PBMCs isolated from four healthy donors. PBS and LPS columns display average cytokines levels from PBMCs stimulated with PBS and LPS (100 ng/ml), while other columns display average data from these PBMCs stimulated with eight stool donors in MiPro. The color in the heatmap represents high (orange) or low (blue) cytokine levels. Results were consistent across PBMC donors and averaged for analysis. Individual donor data are presented in Figure S6. Levels of significance (a) were determined using paired one-way ANOVA followed by Bonferroni's post hoc test. *p < 0.05; **p < 0.01.Sb supplementation induces protective effects against AMC-induced gut microbiome alterations in the dynamic SHIME® model. (a) Schematic representation of the dynamic SHIME® model. Stool samples were cultured in proximal (PC) and distal colon (DC) compartments with continuous feeding (Methods). (b) Schematic representation of long-term intervention of AMC and Sb. Stool samples from 6 independent healthy human donors were sequentially inoculated and stabilized for two weeks. Sb was supplemented in the feed (400 mg TID for 21 d) during and after AMC treatment (50 mg TID for 7 d) (+Sb), while a control group received AMC alone (−Sb). After stabilization, samples were daily collected at a fixed time for further analyses (Methods). (c),(d) Bacterial load in DC compartments throughout the experiment and the area under the curve (d). Each color (d) represents one stool donor in SHIME®. Levels of significance (d) were determined using paired two-tailed Wilcoxon test. *p < 0.05. (e) Absolute abundance of bacteria at the phylum level. Data were expressed as the mean ± standard error of the mean for each taxon. (f) Heatmap visualization of coefficient from MaAsLin analysis representing absolute abundances of species with significant differences between with Sb supplementation (+Sb) and without Sb supplementation (−Sb) at different time points. Only species with a prevalence more than 49.9% and a q-value below 0.1 were shown. (g) Heatmap visualization of coefficient from MaAsLin analysis representing MetaCyc pathways with significant differences between Day 7 with or without Sb supplementation (+Sb or −Sb) and Day 0. Only MetaCyc pathways with a prevalence more than 49.9% and a q-value below 0.25 were shown.Sb supplementation stimulates propionate and IPA production in the dynamic SHIME® model and attenuates pro-inflammatory effects associated with AMC-perturbed gut microbiome. (a) Levels of short-chain fatty acids. (b) Levels of propionate. (c) Levels of tryptophan metabolites. (d) Levels of indole-3-propionic acid (IPA). (e) Levels of antibiotic resistance genes (ARGs) targeting AMC. (f) Heatmap visualization of inflammatory cytokine levels, scaled by Z-scores, after stimulation with representative SHIME® supernatants in human PBMCs. Each column displays average data from 6 stool donors in SHIME®. (g) Levels of Monocyte Chemoattractant Protein-1 (MCP-1) obtained as described in (f). (h) Heatmap visualization of inflammatory cytokine level, scaled by Z-scores, after stimulation with SHIME® culture supernatant at Day 7 in human intestinal mucosa. Each column displays data from one stool donor in SHIME®. The color in the heatmap (f,h) represents high (orange) or low (blue) cytokine levels. PBMCs isolated from four healthy donors and mucosal samples isolated from 2 ADK patients served as technical replicates. Results were consistent across donors and averaged for analysis. Individual donor data are presented in Figure S12 and Figure S13. Each color (b,d,e,g) represents one stool donor in SHIME®. Levels of significance (b,d,e,g) were determined using paired two-way ANOVA with Bonferroni's post hoc test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

 

Hôpital St-Antoine

Bâtiment Kourilsky
34 rue Crozatier
75012 PARIS

Sorbonne Université
27 rue Chaligny
75012 PARIS

Ce site utilise des cookies et vous donne le contrôle sur ce que vous souhaitez activer Tout accepter Personnaliser