Flotillins play an essential role in Niemann-Pick C1-like 1-mediated cholesterol uptake

Abstract

Dietary absorption is a major way for mammals to obtain cholesterol, which is mediated by Niemann-Pick C1-like 1 (NPC1L1) via vesicular endocytosis. One fundamental question in this process is how free cholesterol is efficiently taken up through the internalization of NPC1L1. Using exogenously expressed NPC1L1-EGFP, we show that the lipid raft proteins flotillins associate with NPC1L1 and their localization is regulated by NPC1L1 during intracellular trafficking. Furthermore, flotillins are essential for NPC1L1-mediated cellular cholesterol uptake, biliary cholesterol reabsorption, and the regulation of lipid levels in mice. Together with NPC1L1, they form cholesterol-enriched membrane microdomains, which function as carriers for bulk of cholesterol. The hypocholesterolemic drug ezetimibe disrupts the association between NPC1L1 and flotillins, which blocks the formation of the cholesterol-enriched microdomains. Our findings reveal a functional role of flotillins in NPC1L1-mediated cholesterol uptake and elucidate the formation of NPC1L1-flotillins-postive cholesterol-enriched membrane microdomains as a mechanism for efficient cholesterol absorption.

Fig. 1.

Flotillins associate with NPC1L1. (A) The NPC1L1–flotillins complex forms constitutively during cholesterol uptake. CRL1601–NPC1L1-EGFP cells were depleted of cholesterol by incubating in cholesterol-depleting medium for 60 min and then replenished with cholesterol by incubating in cholesterol-replenishing medium containing 15 μg/mL of cholesterol for the indicated time durations. IP was performed with anti-EGFP-coupled agarose and indicated proteins were detected by immunoblot (IB). Ctr, control; μ2, AP2 complex subunit μ2; Flot, flotillin; Tnr, transferrin receptor. (B–E) Cholesterol-regulated recycling of flotillin-1 is dependent on NPC1L1. (B and C) Flotillin-1-RFP was coexpressed with NPC1L1-EGFP (B) or EGFP (C) in CRL1601 cells. After 48 h, the cells were depleted of cholesterol for 60 min and then replenished with cholesterol for 60 min as described in Fig. 1A. At different time points, i.e., steady state (SS), cholesterol depletion (CD), and cholesterol replenishment (CR), the cells were fixed and examined by confocal microscopy. (Scale bar: 10 μm or 1 μm, magnified image.) (D and E) The PM-localized flotillin-1 (D) or the overlap coefficient (E) between NPC1L1-EGFP and flotillin-1-RFP shown in B and C was quantified. Error bars represent standard deviations (n≥50). (F and G) FRAP analysis. CRL1601 cells were transfected with indicated plasmids. After 48 h, the cells were depleted of cholesterol by incubating in cholesterol-depleting medium for 60 min. The cells were then maintained in cholesterol-depleting medium without cyclodextrin and FRAP experiments were performed. NPC1L1-EGFP from each sample was bleached in the indicated area on the PM and the fluorescence recovery of NPC1L1-EGFP was monitored (G) and quantified (F). Error bars represent standard deviations (n≥20).

Fig. 2.

Knockdown of flotillins impairs NPC1L1-mediated cholesterol uptake. (A and B) Knockdown efficiency of flotillins measured by Q-PCR and immunoblot (IB). CRL1601–NPC1L1-EGFP cells were transfected with indicated siRNA, respectively. After 96 h, Q-PCR was performed to examine the expression level of flotillin-1 and flotillin-2 (A) or the cells were harvested for IB (B). Error bars represent standard deviations of three experiments. Ctr, control; Flot, flotillin; Tnr, transferrin receptor. (C–F) Flotillins are required for the internalization of NPC1L1-EGFP and cholesterol. CRL1601–NPC1L1-EGFP cells transfected with indicated siRNAs were replenished with cholesterol for 60 min after cholesterol depletion, as described in Fig. 1A. The cells were fixed, stained with filipin, and examined by two-photon confocal microscopy. (Scale bar: 10 μm.) (D) Quantification of intracellular localized NPC1L1 and cholesterol in C. Error bars represent standard deviations (n≥100). (E) CRL1601–NPC1L1-EGFP cells were transfected with indicated siRNAs and depleted of cholesterol, as described in C. Cells were then replenished with cholesterol for the indicated time durations and surface biotinylation experiments were performed to analyze the endocytosis of NPC1L1. Rep, cholesterol replenishment. (F) CRL1601–NPC1L1-EGFP cells transfected with indicated siRNAs were depleted of cholesterol for 60 min (time point 0) and then replenished with cholesterol for 60 min (time point 60), as described in Fig. 1A. The cholesterol levels in time point 0 (Chol#1) and 60 (Chol#2) were measured. The net cholesterol uptake was calculated by subtracting Chol#1 from Chol#2. Error bars represent standard deviations of three experiments.

Fig. 3.

Adenovirus-mediated knockdown of flotillins in mice liver affects NPC1L1-mediated biliary cholesterol reabsorption and plasma lipid levels. Eight-week-old male (five per group) mice were administrated with adenovirus (Ad) expressing NPC1L1-EGFP (A) or indicated combination of adenoviruses (B–L). Four days later, the different tissues (A) or livers (B) were collected and subjected to immunoblot (IB) analysis. Flot, flotillin. Frozen sections of the livers were stained with rabbit polyclonal anti-Mrp2 antibody and examined by confocal microscopy (C). (Scale bar: 10 μm.) Biliary cholesterol (Chol) (D), phospholipids (PL) (E), and bile acids (BA) (F) were measured. Hepatic cholesterol (G) and PL (H) were analyzed. Plasma total cholesterol (I), triglyceride (TG) (J), LDL-c (K), and HDL-c (L) were determined. Statistical analyses were done using two-way ANOVA [Tukey’s HSD (Honestly Significant Differences) post test]. ***p < 0.0001, **p < 0.01, *p < 0.05. Error bars represent standard deviations (n = 5).

Fig. 4.

NPC1L1 increases the distribution of flotillins and cholesterol in LDF in sucrose gradient ultracentrifugation assay. (A) Colocalization of NPCL1-EGFP, Flotillin-1-RFP, and cholesterol. Flotillin-1-RFP and NPC1L1-EGFP plasmids were cotransfected in CRL1601 cells. After 48 h, the cells were depleted of cholesterol followed by replenishment with cholesterol as described in Fig. 1A. The cells were then fixed, stained with filipin, and examined by two-photon confocal microscopy. (Scale bar: 10 μm.) (B–G) Sucrose gradient ultracentrifugation analysis. (B) CRL1601 (control) and CRL1601–NPC1L1-EGFP cells were replenished with cholesterol after cholesterol depletion, as described in Fig. 1A. The plasma-endocytic (P/E) membrane was purified and examined by immunoblot (IB). T, total cell lysate; M, membrane fraction. T, total; M, membrane fraction; Flot-1, flotillin-1; CREB, c-AMP response element binding; Tnr, transferrin receptor; ER, endoplasmic reticulum. (C) The purified P/E membrane was subjected to sucrose gradient ultracentrifugation. Indicated proteins in each fraction were analyzed by IB and GM1 was examined by dot blot. (D) Proteins and GM1 in LDF were quantified as relative to total. Error bars represent standard deviations of three experiments. (E–G) Cholesterol (E), PC (F), and SM (G) in each fraction were determined. (H) The increasing fold of cholesterol, PC, SM, and GM1 in LDF. The fold of increase was calculated using the formula (Lipids _{CRL1601–NPC1L1-EGFP}-Lipids_CRL1601)/Lipids_CRL1601. Error bars represent standard deviations of three experiments. Ctr, control.

Fig. 5.

Ezetimibe inhibits the formation of cholesterol-enriched membrane microdomains by dissociating the association between NPC1L1 and flotillins. (A) Ezetimibe dissociates the association between NPC1L1 and flotillins in cultured cells. Cholesterol-depleted CRL1601–NPC1L1-EGFP cells were incubated without or with ezetimibe and replenished with cholesterol for 30 min. IP was performed with anti-EGFP coupled agarose. Immunoblot (IB) was carried out with indicated antibodies. Ctr, control; μ2, AP2 complex subunit μ2; Flot, flotillin; Tnr, transferrin receptor; rep, replenishment. (B) Ezetimibe dissociates the NPC1L1–flotillins complex in mice intestine. Eight-week old C57/B6 male mice were gavaged with 10 mg/kg ezetimibe suspended in 0.5% methyl cellulose or methyl cellulose (control) per day for 3 d. In the fourth day, 2 h after gavage with ezetimibe, the mice were killed and the mucosa from small intestine was collected. Membrane fractions was purified and lysed in IP buffer containing 2% digitonin followed by high-speed centrifugation to discard debris. The supernatant was incubated with 5 μg immunopurified anti-NPC1L1 antibody for 1 h. Protein A agarose was then added and rotated for 4 h at 4 °C. The agarose was washed with IP buffer containing 0.5% digitonin five times and IB was performed. (C and D) FRAP analysis. CRL1601 cells were transfected with indicated plasmids. After transfection (48 h), the cells were treated without or with ezetimibe and depleted of cholesterol for 60 min. Then NPC1L1-EGFP from each sample was bleached in indicated area on the PM and the fluorescence recovery of NPC1L1-EGFP was monitored (D) and quantified (C). Error bars represent standard deviations (n≥20). (E–G) Ezetimibe impairs the formation of cholesterol-enriched membrane microdomains. Cholesterol-depleted CRL1601–NPC1L1-EGFP cells were treated without or with ezetimibe and replenished with cholesterol. Sucrose gradient ultracentrifugation was performed. Fractions were collected and subject to IB–dot blot (E) and cholesterol assay (F). The relative amount of proteins and GM1 in LDF was quantified (G). Error bars represent standard deviations of three experiments.