Efflux of intracellular versus plasma membrane cholesterol in HepG2 cells: different availability and regulation by apolipoprotein A-I

We have compared the efflux of cholesterol from different cellular pools of human hepatoma cells HepG2 using intact cells or isolated membrane fractions. To label different pools, cells were incubated with either unesterified 14C]cholesterol that had been incorporated into high density lipoproteins (14C]FC-HDL), low density lipoproteins (14C]FC-LDL), or phosphatidylcholine liposomes (14C]FC-PC), or with 14C]acetate. Cell fractionation revealed that labeling of cells with 14C]FC-PC resulted in the incorporation of 14C]cholesterol almost exclusively into the plasma membrane (PM), while incubation with 14C]FC-HDL resulted in the majority of 14C]cholesterol incorporation into the PM, but with a smaller component associated with lysosomes. Labeling with 14C]FC-LDL or 14C]acetate led to an accumulation of 14C]cholesterol predominantly in lysosomes or the endoplasmic reticulum (ER), respectively. When the kinetics of 14C]cholesterol efflux was analyzed after pulse-labeling of different cellular pools, half-times of cholesterol efflux from lysosomes and ER were significantly longer than that from PM. In another set of experiments, when both labeling and efflux times varied, efflux of 14C]cholesterol from the PM to human serum after 1.5 h pulse and chase incubations was double that from lysosomes and 8-fold that from ER. Extension of the incubation times from 1.5 to 3 h diminished the difference in cholesterol efflux from different membranes. Further incubation to 6 h almost abolished the different responses. Cell-free preparations of membranes, obtained from cells labeled with 14C]cholesterol, showed no differences in cholesterol efflux. No differences in the distribution of 14C]cholesterol released into serum among lipoprotein subfractions was observed. Pretreatment of the serum with Fab fragments of polyclonal rabbit anti-human apolipoprotein A-I antibodies reduced its ability to promote efflux of cholesterol from the ER by 77%, but had no effect on cholesterol efflux from the PM. Fab fragments of non-immune IgG had no effect on the efflux of both ER and PM cholesterol. We conclude that the availability of cellular cholesterol for efflux from HepG2 cells is strongly influenced by its subcellular location, and is regulated by apolipoprotein A-I.