The intestinal microbiota regulates host cholesterol homeostasis

Abstract

Background: Management of blood cholesterol is a major focus of efforts to prevent cardiovascular diseases. The objective of this study was to investigate how the gut microbiota affects host cholesterol homeostasis at the organism scale.

Results: We depleted the intestinal microbiota of hypercholesterolemic female Apoe^-/- mice using broad-spectrum antibiotics. Measurement of plasma cholesterol levels as well as cholesterol synthesis and fluxes by complementary approaches showed that the intestinal microbiota strongly regulates plasma cholesterol level, hepatic cholesterol synthesis, and enterohepatic circulation. Moreover, transplant of the microbiota from humans harboring elevated plasma cholesterol levels to recipient mice induced a phenotype of high plasma cholesterol levels in association with a low hepatic cholesterol synthesis and high intestinal absorption pattern. Recipient mice phenotypes correlated with several specific bacterial phylotypes affiliated to Betaproteobacteria, Alistipes, Bacteroides, and Barnesiella taxa.

Conclusions: These results indicate that the intestinal microbiota determines the circulating cholesterol level and may thus represent a novel therapeutic target in the management of dyslipidemia and cardiovascular diseases.

Keywords: Antibiotics; Cholesterol metabolism; Dyslipidemia; Enterohepatic cycle; Intestinal microbiota; Microbiome; Microbiota transfer.

Fig. 1

Intestinal microbiota depletion raises plasma cholesterol levels and intestinal cholesterol absorption. a Experimental design. See also Additional file 2: Figure S1. b Plasma cholesterol, phospholipids and triglycerides levels in conventionally raised (Conv-R) and microbiota-depleted mice (AB-Mdpl). c Cholesterol distribution across the VLDL, LDL, and HDL lipoprotein classes analyzed by fast protein liquid chromatography. d Plasma radioactivity 2 h after gavage with [³H]-cholesterol. e Relative expression of genes related to cholesterol absorption in the jejunum. f Relative expression of genes related to lipoprotein secretion in the jejunum. Data are represented as mean ± SEM, n = 5–10 mice/group (d, e) or as dots with median (b–f). Data were analyzed with Mann–Whitney test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 2

Intestinal microbiota depletion increases hepatic cholesterol uptake and hepatic cholesterol synthesis. a Liver radioactivity 2 h after gavage with [³H]-cholesterol in conventionally raised (Conv-R) and microbiota-depleted mice (AB-Mdpl). b Hepatic relative expression of cholesterol transporters. c Plasma cholesterol increase in microbiota-depleted mice in comparison to control mice in Apoe (○) and LDLr (□)^−/− mice. d Hepatic relative expression of genes related to cholesterol synthesis. See also Additional file 5: Figure S3. e Cholesterol and lathosterol concentration analyzed by GC-MS in the liver. Data are represented as mean ± SEM, n = 6–9 mice/group (b–d) or as dots with median (a, c, e). Data were analyzed with Mann–Whitney test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 3

Enterohepatic cycle of cholesterol and bile acids in conventionally raised and microbiota-depleted mice. a Bile volume collected in 1 h of gallbladder cannulation in conventionally raised (Conv-R) and microbiota-depleted mice (AB-Mdpl). b Quantity of cholesterol secreted in the bile during 1 h of gallbladder cannulation. c Hepatic gene expression of enzymes involved in bile acid biosynthesis and of transporters of cholesterol and bile acids in conventionally raised (Conv-R) and microbiota-depleted mice (AB-Mdpl). d Fecal excretion of ¹⁴C bile acids (water-soluble fraction) and ¹⁴C cholesterol (cyclohexane soluble fraction) during 72 h after oral gavage with ¹⁴C cholesterol. e ¹⁴C bile acids excreted in the feces expressed as percent of total radioactivity (cholesterol + bile acids). f Relative expression of fgf15 in the distal ileum. g Plasma radioactivity 2 h after gavage with [³H]-taurocholic acid. h Relative gene expression of bile acid transporters in the distal ileum. Data are represented as mean ± SEM (c, f, h) or as dots with median (a, b, g), n = 5–8 mice/group. Data were analyzed with Mann–Whitney test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 4

Plasma cholesterol levels are transferable from humans to mice by intestinal microbiota transplantation. a Donors’ characteristics and experimental design. b Plasma cholesterol, phospholipids, and triglycerides levels in mice colonized with the microbiota from normocholesterolemic donors (NorChol-r1 and r2, pictured cyan and dark cyan) and high-cholesterol donors (HiChol-r1 and r2, pictured in red and dark red). Data are represented as dots with median (a, b), n = 8–12 mice/group. Recipient groups were analyzed using Kruskal–Wallis test followed by Dunn’s pairwise multiple comparison procedure. *q < 0.05, **q < 0.01, ***q < 0.001

Fig. 5

Intestinal microbiota regulates cholesterol absorption/synthesis balance. a Relative expression of genes related to cholesterol absorption and lipoprotein secretion in the jejunum in mice colonized with the microbiota from normocholesterolemic donors (NorChol-r1 and r2, pictured cyan and dark cyan) and high-cholesterol donors (HiChol-r1 and r2, pictured in red and dark red). b Relative expression of enzymes involved in cholesterol synthesis in the liver. See also Additional file 9: Figure S4. c Cholesterol and lathosterol concentration analyzed by GC-MS in the liver. d Triglycerides and phospholipids analyzed by biochemic assay in the liver. e Hepatic relative expression of LDLr. f Hepatic relative expression of Cyp7a1 in the liver. g Relative expression of fgf15 in the distal ileum. Data are represented as mean ± SEM (a, b, e, f, g) or as dots with median (c, d), n = 8–12 mice/group. Recipient groups were analyzed using Kruskal–Wallis test followed by Dunn’s pairwise multiple comparison procedure. *q < 0.05, **q < 0.01, ***q < 0.001

Fig. 6

Mice colonized by the microbiota of normocholesterolemic and high-cholesterol human donors harbor specific intestinal microbiota composition. a Interclass principal component analysis performed based on ASVsabundance. Mice microbiota were clustered and the center of gravity computed for each group. The p value of the link between recipient groups and ASV abundance was calculated using a Monte Carlo test (999 replicates). b Cladogram generated using GraPhlAn [38] representing recipients’ microbiota with cyan clade-markers highlighting bacterial groups significantly more abundant in NorChol recipients and red clade-markers highlighting bacterial groups significantly more abundant in HiChol recipients as assessed by Kruskal–Wallis test followed by Dunn’s pairwise multiple comparison procedure. Circular heatmap represents normalized abundance of all ASV in each recipient group, with the darkest color corresponding to the group having the highest percentage of the given ASV. Black bars represent the mean abundance of the ASVs in the whole data set. c Bacterial ASVs statistically more abundant in both HiChol recipients’ groups than in both NorChol recipients’ groups. n = 9–12 mice/group. d Spearman correlations between ASV-level microbial populations and cholesterol metabolism-associated parameters. Strong correlations are indicated by large circles, whereas weaker correlations are indicated by small circles. The colors of the circles denote the nature of the correlation with dark blue indicating strong positive correlation and dark red indicating a strong negative correlation. ^¤q < 0.05, ^¤¤q < 0.01, ^¤¤¤q < 0.001 after FDR correction

Fig. 7

Microbial regulation of whole-body cholesterol fluxes and enterohepatic cycle. Microbiota depletion in Apoe^−/− mice raises plasma VLDL and LDL cholesterol. Microbiota-depleted mice have increased intestinal cholesterol absorption, hepatic cholesterol uptake, and hepatic cholesterol and bile acid synthesis. Bile secretion is also increased in microbiota-depleted mice, which is associated with increased fecal excretion of bile acids. Microbiota depletion is associated with a decrease in fgf15 expression in the distal ileum, thus alleviating feedback inhibition of hepatic bile acid synthesis