High fat intake sustains sorbitol intolerance after antibiotic-mediated Clostridia depletion from the gut microbiota

Abstract

Carbohydrate intolerance, commonly linked to the consumption of lactose, fructose, or sorbitol, affects up to 30% of the population in high-income countries. Although sorbitol intolerance is attributed to malabsorption, the underlying mechanism remains unresolved. Here, we show that a history of antibiotic exposure combined with high fat intake triggered long-lasting sorbitol intolerance in mice by reducing Clostridia abundance, which impaired microbial sorbitol catabolism. The restoration of sorbitol catabolism by inoculation with probiotic Escherichia coli protected mice against sorbitol intolerance but did not restore Clostridia abundance. Inoculation with the butyrate producer Anaerostipes caccae restored a normal Clostridia abundance, which protected mice against sorbitol-induced diarrhea even when the probiotic was cleared. Butyrate restored Clostridia abundance by stimulating epithelial peroxisome proliferator-activated receptor-gamma (PPAR-γ) signaling to restore epithelial hypoxia in the colon. Collectively, these mechanistic insights identify microbial sorbitol catabolism as a potential target for approaches for the diagnosis, treatment, and prevention of sorbitol intolerance.

Keywords: Clostridia; antibiotics; carbohydrate intolerance; gut microbiota; high-fat diet; polyol; sorbitol intolerance.

Fig. 1:. Mouse model of prolonged sorbitol intolerance.

Mice maintained on a low-fat diet (LF) or a high-fat diet (HF) for 14 days were mock-treated or received a single dose of streptomycin (Str) by oral gavage. After maintaining mice for four more weeks on the same diet, mice received drinking water supplemented with 5 % sorbitol for three days. (A) Change in body weight for each group (n = 14) over time. (B) Fecal water content after 3 days of sorbitol supplementation. (C) Sorbitol concentration after 3 days of sorbitol supplementation. LOD, limit of detection. (D) Sorbitol dehydrogenase activity in cecal contents. (B-D) Group sizes (n) are indicated by the number of symbols. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001. (A-B, F) Analysis with two-way (A) or one-way (B-E) ANOVA followed by Tukey’s multiple-comparison test (A), or student’s t test (F).

Fig. 2:. Clostridia are a main source of sorbitol dehydrogenase genes during homeostasis.

Mice were maintained on a low-fat diet (LF) or a high-fat diet (HF) for 14 days. Fecal samples were collected (“LF” or “HF before Str”) and mice were mock-treated or received a single dose of streptomycin (Str) by oral gavage. After maintaining mice for four more weeks on the same diet, a second fecal sample was collected for analysis (“LF + 4 weeks” or “HF 4 weeks after Str”). (A-C) Microbiota profiling of fecal DNA collected at the indicated time points. (A and B) Relative abundance of amplicon sequence variants (ASVs) belonging to the class Clostridia (A) or the genus Enterococcus (B). Box plots represent the first to third quartiles, and lines indicates median values. *, P < 0.05. Analysis with Kruskal-Wallis test. (C) The cladogram shows differences in taxa composition between samples collected before streptomycin treatment (HF before Str) and four weeks after streptomycin treatment (HF 4 weeks after Str). Taxa that are significantly (Kruskal-Wallis test, LEfSe) more abundant (green) or less abundant (red) before compared to 4 weeks after streptomycin treatment are shown. (D-F) DNA isolated from the indicated samples were subjected to metagenomic analysis. (D) Volcano plot of genes involved in carbohydrate metabolism. Negative values indicate genes with decreased abundance four weeks after streptomycin treatment compared to prior to streptomycin treatment. (E and F) Abundance of genes encoding sorbitol dehydrogenase (E) or sorbitol-6-phosphate 2-dehydrogenase (F) in samples collected prior to or four weeks after streptomycin treatment.

Fig. 3:. Abundant probiotics protect against transient sorbitol intolerance.

(A-D and F-K) Mice were mock-treated or treated with a single dose of streptomycin. One day later, drinking water was supplemented with 5% sorbitol and mice were inoculated with different doses of E. coli Nissle 1917 (EcN WT), a E. coli Nissle 1917 srlABE mutant (EcN srlABE) or A. caccae. Samples were collected after 2 days of sorbitol supplementation. (A) Colony-forming units (cfu) recovered from feces of animals (y axis) 2 days after inoculation with the indicated doses of EcN WT (x axes). (B) The abundance of E. coli in fecal samples was determined by real-time PCR using Enterobacterales-specific primers. (C and J) Fecal water content in fecal pellets (y axis) collected from animals two days after inoculation with the indicated doses (x axes) of E. coli Nissle 1917 (C) or A. caccae (J). (D) Sorbitol dehydrogenase activity in cecal contents of mice inoculated with the indicated doses of E. coli Nissle 1917. (E) Minimal medium containing sorbitol as a sole carbon source was inoculated with the indicated E. coli Nissle 1917 strains carrying either no plasmid, a plasmid encoding the srlAEB genes (pAWLR169) or the empty plasmid vector (pWSK29). After overnight culture, the sorbitol concentration in culture supernatants was measured. (F) Fecal water content in fecal pellets (y axis) collected from animals two days after inoculation with the indicated doses of EcN srlABE (x axes). (G) Cfu recovered from feces of animals (y axis) 2 days after inoculation with the indicated doses of EcN srlABE (x axes). (H) In vitro anaerobic growth of A. caccae in no-carbon defined medium (NCDM) supplemented with glucose, sorbitol or without supplementation (no sugar). Enterocloster asparagiformis, a Clostridia species that does not ferment sorbitol, was used as a negative control. (I) The abundance of A. caccae in fecal samples was determined by real-time PCR using genus specific primers. (K) Sorbitol dehydrogenase activity in cecal contents of mice inoculated with the indicated doses of A. caccae. (A, B and I) A grey bar indicates the threshold of colonization required for protection against sorbitol-induced diarrhea. (A, B, G and I) LOD, limit of detection. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001. (C-F, H, J, K) Analysis by one-way ANOVA followed by Tukey’s multiple-comparison tests.

Fig. 4:. Sorbitol-catabolizing probiotics protect against prolonged sorbitol intolerance.

Mice reared and maintained throughout the experiment on a low-fat or a high-fat (HF) diet were mock-treated or received a single dose of streptomycin (Str), respectively. Four weeks later, mice received drinking water supplemented with 5 % sorbitol and were inoculated with 10⁹ colony-forming units (cfu) of E. coli Nissle 1917 (EcN), A. caccae (AC), or Lactiplantibacillus plantarum (LP). Samples were collected after three or seven days of sorbitol supplementation. (A) Schematic of experimental groups and time points. (B) Fecal water content in fecal pellets. (C) Sorbitol dehydrogenase activity in cecal contents. (D) Cfu of EcN or LP. (E) Absolute abundance of AC in feces was determined by real-time PCR using genus specific primers. (F) Absolute abundance of Clostridia in feces determined by real-time PCR using class specific primers. (G). Numbers of amplified sequence variants (ASVs) belonging to the class Clostridia at the indicated time points. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001. Analisis by one-way ANOVA followed by Tukey’s multiple-comparison tests (B-D, F) or student’s t test (E). (C, E, and F) LOD, limit of detection.

Fig. 5:. A. caccae restores butyrate levels and epithelial hypoxia in mice with prolonged sorbitol intolerance.

Mice reared and maintained throughout the experiment on a low-fat or a high-fat (HF) diet were mock-treated or received a single dose of streptomycin (Str), respectively. Four weeks later, mice received drinking water supplemented with 5 % sorbitol and were inoculated with 10⁹ colony-forming units (cfu) of E. coli Nissle 1917 (EcN), A. caccae (AC), or Lactiplantibacillus plantarum (LP). Samples were collected after three or seven days of sorbitol supplementation. (A) Butyrate concentrations in cecal contents. (B) Relative abundance of Clostridia families containing gene sequence involved in butyrate metabolism. (C-D) Mice were injected with pimonidazole (PMDZ) before euthanasia. PMDZ was detected using hypoxyprobe-1 primary antibody and a Cy-3 conjugated goat anti-mouse secondary antibody (red fluorescence) in colonic sections counter stained with nuclear stain (blue fluorescence). (C) Representative images for each group seven day after inoculation with probiotics. L, intestinal lumen. The graph shows PMDZ intensity from the lumen across the epithelial layer (distance in arbitrary units). (D) The graph shows the average peak PMDZ intensity. (A and D) Each symbol represent data from one animal. **, P < 0.01; ***, P < 0.005. Analysis by one-way ANOVA followed by Tukey’s multiple-comparison tests (A) or Kruskal–Wallis test (D).

Fig. 6:. A. caccae and butyrate stimulate epithelial PPAR-γ signaling to promote microbiota recovery.

Mice reared and maintained throughout the experiment on a low-fat or a high-fat (HF) diet were mock-treated or received a single dose of streptomycin (Str), respectively. Four weeks later, mice received drinking water supplemented with 5 % sorbitol and were inoculated with A. caccae (AC) or received supplementation with 5 aminosalicylic acid (5-ASA) or tributyrin (TB). Samples were collected after three or seven days of sorbitol supplementation. (A-F) Experiments with C57BL/6J mice. (A) Butyrate concentrations in cecal contents. (B) Fecal water content in fecal pellets. (C) Sorbitol concentration. (D and I) Mice were injected with pimonidazole before euthanasia. Binding of pimonidazole (PMDZ) was detected using hypoxyprobe-1 primary antibody and a Cy-3 conjugated goat anti-mouse secondary antibody. (D) Average PMDZ peak intensity. (E) Absolute abundance of Clostridia in fecal samples was determined by real-time PCR using class specific primers. (F) Relative abundance of Clostridia families containing gene sequence involved in butyrate metabolism. (G-J) Experiments were performed with Pparg^fl/flVillin^cre/− mice (Pparg) or Pparg^fl/flVillin^−/− littermate controls (WT). (G) Fecal water content. (H) Sorbitol concentration. (I) Average PMDZ peak intensity. (J) Butyrate concentrations in cecal contents. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001. Analysis by one-way ANOVA followed by Tukey’s multiple-comparison tests (A-C, E G-J) or Kruskal–Wallis test (D). (C and H) LOD, limit of detection.

Fig. 7:. 5-ASA treatment prevents development of prolonged sorbitol intolerance.

Mice maintained on a low-fat diet (LF) or a high-fat diet (HF) for 14 days were mock-treated or received a single dose of streptomycin (Str) by oral gavage. (A-B) Four weeks later, mice received drinking water supplemented with 5 % sorbitol and were treated by supplementing chow with 5-ASA. (A) the relative abundance of Clostridia in fecal samples collected from C57BL/6J mice. (B) Relative abundance of Clostridia families containing gene sequence involved in butyrate metabolism. (C-F) At the time of streptomycin treatment mice were switched to chow supplemented with 5-ASA. Four weeks later, mice received drinking water supplemented with 5 % sorbitol to assess sorbitol tolerance. (C and D) Absolute abundance of Clostridia in fecal samples of C57BL/6J mice (C), Pparg^fl/flVillin^cre/− mice (Pparg) or Pparg^fl/flVillin^−/− littermate control mice (WT) (D) was determined by real-time PCR using class specific primers. (E) Sorbitol concentration in C57BL/6J mice. LOD, limit of detection. (F) Fecal water content in samples from C57BL/6J mice. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ****, P < 0.001. (C-F) Analysis by one-way ANOVA followed by Tukey’s multiple-comparison tests.