Oleoylethanolamide treatment affects gut microbiota composition and the expression of intestinal cytokines in Peyer's patches of mice

Abstract

The lipid sensor oleoylethanolamide (OEA), an endogenous high-affinity agonist of peroxisome proliferator-activated receptor-α (PPAR-α) secreted in the proximal intestine, is endowed with several distinctive homeostatic properties, such as control of appetite, anti-inflammatory activity, stimulation of lipolysis and fatty acid oxidation. When administered exogenously, OEA has beneficial effects in several cognitive paradigms; therefore, in all respects, OEA can be considered a hormone of the gut-brain axis. Here we report an unexplored modulatory effect of OEA on the intestinal microbiota and on immune response. Our study shows for the first time that sub-chronic OEA administration to mice fed a normal chow pellet diet, changes the faecal microbiota profile, shifting the Firmicutes:Bacteroidetes ratio in favour of Bacteroidetes (in particular Bacteroides genus) and decreasing Firmicutes (Lactobacillus), and reduces intestinal cytokines expression by immune cells isolated from Peyer's patches. Our results suggest that sub-chronic OEA treatment modulates gut microbiota composition towards a "lean-like phenotype", and polarises gut-specific immune responses mimicking the effect of a diet low in fat and high in polysaccharides content.

Figure 1

Effect of sub-chronic OEA administration on body weight and food consumption. (A) Time course of the effects of vehicle (VEH) or OEA (10 mg/kg, i.p.) on body weight change in mice. Two-way ANOVA with repeated measures and Bonferroni’s post hoc test; significantly different from vehicle treated mice, *P < 0.05; **P < 0.01; (B) Time course of the daily and (C) Cumulative food intake. Student’s T test, *P < 0.05; **P < 0.01; n = 9–13 per group.

Figure 2

Alpha and Beta diversity. (A–D) Alpha diversity measures. Box plots of (A) observed OTUs, (B) Shannon index, (C) Dominance, and (E) Equitability calculated for mice groups at T0 = before the treatment, after 11 days of OEA treatment (OEA_T11) and after 11 days of vehicle treatment (Vehicle_ T11). Pairwise comparisons by using the Wilcoxon rank sum test were not significant. (E-F) Beta diversity measure. (E) PCoA, and (F) NMDS based on Bray Curtis dissimilarities. Samples at T0, samples of OEA_T11 and Vehicle_T11 are indicated with different colours. P = 0.035 PERMANOVA using the Adonis function with 999 permutations. For NMDS, stress value=0.09 indicated a good representation.

Figure 3

Barplot representation of microbial profiles. Relative abundances, on average, at the phylum and genus level of the microbiota in mice before the treatment (T0), after OEA and vehicle treatment at T11.

Figure 4

Metagenomic biomarker discovery by LEfSe analysis. Differential enriched bacterial taxa at the phylum level in gut microbiota of mice (A) before (T0) and after OEA treatment (T11) and (B) of mice treated for 11 days with OEA or vehicle- (Vehicle_T11). Differentially enriched bacterial taxa at the genus level in the gut microbiota of mice (C) before (T0) and after OEA treatment (T11) and (D) 11 days after OEA (T11) or vehicle treatment (Vehicle_T11). In the lower panels, cladograms show the most discriminative bacterial clades. Coloured regions/branches indicate differences in the bacterial population structure between the different groups. Statistically significant taxa enrichment among groups was obtained with Kruskal-Wallis test among classes (Alpha value = 0.05). The threshold for the logarithmic LDA score was 2.0.

Figure 5

PICRUSt functional analyses. Extended error bar plots representing (A) Microbial pathways predicted to be differentially enriched in microbiomes of mice at T0 and after OEA treatment (OEA_T11). (B) Microbial pathways predicted to be differentially enriched in microbiomes of mice after OEA treatment (OEA_T11) versus controls (Veh_T11). Each extended error bar plot indicates the p-value along with the effect size and the associated difference in mean proportion and confidence interval for each predicted KEGG function. Each bar plot indicates the mean proportion of sequences assigned to the KEGG categories in each group. P-values by White’s nonparametric t-test, Storey FDR correction.

Figure 6

Changes of cytokines and chemokines secretion in intestinal Peyer’s Patches. Protein production was determined by Luminex technology on cells supernatant of OEA- (black column) and vehicle-treated mice (withe column). Cell were activated by either PHA (5 µg/ml) (A) or LPS (1 µg/ml) (B). Each histogram represents mean value ± SEM of protein concentration (pg/ml) of 12 samples for each group of PHA condition from 3 independent experiments and 4 samples for each group of LPS from 1 experiment. Statistical significance of the differences between VEH- and OEA-treated mice was analysed using Student’s t-test; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7

Ex vivo analysis of cells isolated from intestinal Peyer’s Patches. Relative expression of transcription factors by real time PCR. The graph shows mean values ± SD of 4 vehicle-treated (dark column) and 4 OEA-treated mice (light column) of two independent experiments. Expression of transcription factors is reported as ratio to ubiquitin. Student’s t-test, **p = 0.03; ***p < 0.001.