Microbiota prevents cholesterol loss from the body by regulating host gene expression in mice

Abstract

We have previously observed that knockout of Niemann-Pick C1-Like 1 (NPC1L1), a cholesterol transporter essential for intestinal cholesterol absorption, reduces the output of dry stool in mice. As the food intake remains unaltered in NPC1L1-knockout (L1-KO) mice, we hypothesized that NPC1L1 deficiency may alter the gut microbiome to reduce stool output. Consistently, here we demonstrate that the phyla of fecal microbiota differ substantially between L1-KO mice and their wild-type controls. Germ-free (GF) mice have reduced stool output. Inhibition of NPC1L1 by its inhibitor ezetimibe reduces stool output in specific pathogen-free (SPF), but not GF mice. In addition, we show that GF versus SPF mice have reduced intestinal absorption and increased fecal excretion of cholesterol, particularly after treatment with ezetimibe. This negative balance of cholesterol in GF mice is associated with reduced plasma and hepatic cholesterol, and likely caused by reduced expression of NPC1L1 and increased expression of ABCG5 and ABCG8 in small intestine. Expression levels of other genes in intestine and liver largely reflect a state of cholesterol depletion and a decrease in intestinal sensing of bile acids. Altogether, our findings reveal a broad role of microbiota in regulating whole-body cholesterol homeostasis and its response to a cholesterol-lowering drug, ezetimibe.

Figure 1. NPC1L1 and microbiota.

(A) Average wet weight of ceca and cecum-to-body weight (BW) ratios in SPF and GF mice treated with or without ezetimibe (n = 10(11). (B) Relative abundance of bacterial phyla in feces of L1-KO mice and their wild-type (WT) controls on a chow diet or a high fat diet (HFD) for 5 weeks (n = 4 for L1-KO/Chow, WT/Chow, and WT/HFD; n = 3 for L1-KO/HFD). Analysis was done by pyrosequencing of 16S rRNA genes. (C) Stool output in SPF and GF mice treated with or without ezetimibe (n = 10–11). The values with different small letters differ significantly (One-way ANOVA, p < 0.05). WD, Western diet; Ezet, ezetimibe.

Figure 2. Cholesterol balance is altered in GF mice.

(A) Intestinal cholesterol absorption and (B) fecal neutral sterol excretion in SPF and GF mice on Western diet (n = 10-11). The values with different small letters differ significantly (One-way ANOVA, p < 0.05).

Figure 3. GF mice display reduced concentrations, but unaltered molar ratios of lipids in gallbladder.

Gallbladder concentrations and molar ratios of cholesterol (A), phospholipids (B) and bile acids (C) were analyzed in the mice fed a Western diet (WD) with or without ezetimibe (Ezet) for 38 days (n = 10-11). The values with different small letters differ significantly (One-way ANOVA, p < 0.05).

Figure 4. Plasma lipids are decreased in GF mice relative to SPF mice.

Plasma concentrations of cholesterol (A-C), triglycerides (D) and phospholipids (E) were measured in the mice fed a Western diet (WD) with or without ezetimibe (Ezet) for 38 days (n = 10-11). Cholesterol ester concentrations were calculated by multiplying the mass difference between total and free cholesterol by 1.67. The values with different small letters differ significantly (One-way ANOVA, p < 0.05).

Figure 5. Hepatic lipids are reduced in GF mice compared to SPF mice.

Hepatic contents of cholesterol (A-C), triglycerides (D) and phospholipids (E) were measured in the mice fed a Western diet (WD) with or without ezetimibe (Ezet) for 38 days (n = 10-11). The content of cholesterol ester was calculated by multiplying the mass difference between total and free cholesterol by 1.67. The values with different small letters differ significantly (One-way ANOVA, p < 0.05).

Figure 6. Intestinal expression levels of genes involved in lipid metabolism and bile acid sensing.

The mice were fed a Western diet (WD) with or without ezetimibe (Ezet) for 38 days, followed by necropsy for tissue collection. (A) Relative mRNA levels analyzed by qPCR in the proximal second segment of small intestine that was divided into 5 equal segments (n = 6-7). GAPDH was used as an invariant control. The values with different small letters differ significantly (One-way ANOVA, p < 0.05). (B) Western blots of protein levels in the middle (third) segment of small intestine.

Figure 7. Hepatic expression levels of genes involved in lipid metabolism and bile acid metabolism.

The mice were fed a Western diet (WD) with or without ezetimibe (Ezet) for 38 days, followed by necropsy for tissue collection. (A) Relative mRNA levels analyzed by qPCR (n = 6-7). GAPDH was used as an invariant control. The values with different small letters differ significantly (One-way ANOVA, p < 0.05). (B) Western blots of hepatic proteins.