Gut Microbiota and Phenotypic Changes Induced by Ablation of Liver- and Intestinal-Type Fatty Acid-Binding Proteins

Abstract

Intestinal fatty acid-binding protein (IFABP; FABP2) and liver fatty acid-binding protein (LFABP; FABP1) are small intracellular lipid-binding proteins. Deficiency of either of these proteins in mice leads to differential changes in intestinal lipid transport and metabolism, and to markedly divergent changes in whole-body energy homeostasis. The gut microbiota has been reported to play a pivotal role in metabolic process in the host and can be affected by host genetic factors. Here, we examined the phenotypes of wild-type (WT), LFABP^-/-, and IFABP^-/- mice before and after high-fat diet (HFD) feeding and applied 16S rRNA gene V4 sequencing to explore guild-level changes in the gut microbiota and their associations with the phenotypes. The results show that, compared with WT and IFABP^-/- mice, LFABP^-/- mice gained more weight, had longer intestinal transit time, less fecal output, and more guilds containing bacteria associated with obesity, such as members in family Desulfovibrionaceae. By contrast, IFABP^-/- mice gained the least weight, had the shortest intestinal transit time, the most fecal output, and the highest abundance of potentially beneficial guilds such as those including members from Akkermansia, Lactobacillus, and Bifidobacterium. Twelve out of the eighteen genotype-related bacterial guilds were associated with body weight. Interestingly, compared with WT mice, the levels of short-chain fatty acids in feces were significantly higher in LFABP^-/- and IFABP^-/- mice under both diets. Collectively, these studies show that the ablation of LFABP or IFABP induced marked changes in the gut microbiota, and these were associated with HFD-induced phenotypic changes in these mice.

Keywords: gut microbiota; intestinal fatty acid-binding protein; liver fatty acid-binding protein.

Figure 1

Effect of IFABP and LFABP knockout on body weight (A), body weight change (B), intestinal transit time (C), and total fecal output (D). Repeated-measures ANOVA with Tukey’s post hoc was applied in (A). One-way ANOVA with Tukey’s post hoc was applied in (B–D). * p < 0.05, *** p < 0.001. (B,C) were from a separate group of mice with the same genotypes and fed the same HFD. N = 6 for each group. LFABP; FABP1: Intestinal fatty acid-binding protein (IFABP; FABP2) and liver fatty acid-binding protein.

Figure 2

Effect of IFABP and LFABP knockout and a HF diet on the gut microbiota. (A) Shannon Index; (B) ASV number; (C) Principal coordinate plot based on weighted UniFrac distance; (D) Weighted UniFrac distance from IFABP^−/− and LFABP^−/− to WT at each time point. Data at different timepoints within the same genotype group were compared using the Wilcoxon matched-pairs signed-ranks test (two-tailed) and data at the same timepoint between the groups were compared using the Mann–Whitney test (two-tailed). * p < 0.05, ** p < 0.01. Boxes show the medians and the interquartile ranges (IQRs), and the whiskers denote the lowest and highest values within 1.5 times the IQR from the 1st and 3rd quartiles. N = 6 for each group.

Figure 3

Differences and changes in the guilds of the three different genotypes. The heatmap shows the log10-transformed relative abundance of each guild. At each time point, guilds were compared among the groups using the Kruskal–Wallis test and post hoc Dunn’s test. Values not sharing common letters are significantly different from one another (p < 0.05). Wilcoxon matched-pairs signed-ranks test (two-tailed) was used to test the same guild between week 0 and week 11 within each genotype. p < 0.05 was considered as significant. N = 6 for each group.

Figure 4

The association between the gut microbiota and body weight at week 0, prior to HF feeding (chow fed from weaning until 8 weeks of age). Random Forest (RF) model regressing body weight on the guild abundance at week 0. (A) shows the number of variables and mean squared error of the corresponding model. (B) The RF assigns a mean error rate, or feature-importance score, to each feature; this value indicates the extent to which each predictor contributes to the accuracy of the model. (C) Significantly positive correlation between the measured body weight and the predicted values from leave-one-out cross-validation based on RF model. (D) Significantly positive correlation between the measured body weight and the predicted values from guild abundance at week 11 based on the model trained in (A). Pearson correlation was applied.

Figure 5

The association between the gut microbiota and body weight following 11 weeks of the HF diet. Random Forest (RF) model regressing body weight on the guild abundance at week 11. (A) shows the number of variables and mean squared error of the corresponding model. (B) The RF assigns a mean error rate, or feature-importance score, to each feature; this value indicates the extent to which each predictor contributes to the accuracy of the model. (C) Scatter plot of the measured body weight and the predicted values from leave-one-out cross-validation. Pearson correlation was applied.

Figure 6

Analysis of SCFAs in WT, IFABP^−/−, and LFABP^−/− mice at week 0 (chow diet from weaning until 8 weeks of age) and after 11 weeks of the HF diet. (A) Acetate; (B) Propionate; (C) Isobutyrate; (D) Butyrate; (E) Isovalerate; (F) Valerate. Feces were pooled from six mice in each genotype. Two-way ANOVA with Tukey’s post hoc was applied. * p < 0.05, **p < 0.01, *** p < 0.001. The error bars are from technical replicates.