Characterization of apoA-I-dependent lipid efflux from adipocytes and role of ABCA1

Abstract

Adipose tissue is a major reservoir of cholesterol and, as such, it may play a significant role in cholesterol homeostasis. The aims of this study were to obtain a quantitative characterization of apolipoprotein A-I (apoA-I)-dependent lipid efflux from adipocytes and examine the role of ATP-binding cassette transporter A1 (ABCA1) in this process. The rates of apoA-I-induced cholesterol and phospholipid efflux were determined and normalized by cellular protein or ABCA1 levels. In order to allow a comparative analysis, parallel experiments were also performed in macrophages. These studies showed that apoA-I induces cholesterol efflux from adipocytes at similar rates as from macrophages. Enhancement of the expression of ABCA1 increased the rates of cholesterol efflux from both adipocytes and macrophages. The results also suggested that a non-ABCA1-dependent mechanism could make significant contributions to the rate of apoA-I-dependent cholesterol efflux when the expression levels of ABCA1 are low. Furthermore, the study of the effect of inhibitors of lipid efflux showed that glyburide and brefeldin A, which affect ABCA1 function, exerted strong and similar inhibitory effects on lipid efflux from both adipocytes and macrophages, whereas BLT1, an SRB-I inhibitor, only exerted a moderate inhibition. Overall these studies suggest that ABCA1 plays a major role in apoA-I-dependent lipid efflux from adipocytes and showed high similarities between the abilities of adipocytes and macrophages to release cholesterol in an apoA-I-dependent fashion.

Fig. 1

Lipid efflux from adipocytes and macrophages. a Cholesterol efflux: [³H]-cholesterol-labeled 3T3 L-1 adipocytes or J774 macrophages were incubated with apoA-I (75 µg/ml) or with buffer as described in “Methods” section “Cholesterol efflux assays”. The data points representing apoA-I-induced CL efflux from adipocytes (□) and macrophages (■) were calculated by subtracting background CL efflux (no apoA-I) from the CL efflux of wells containing apoA-I. The mean efflux and SD, for each time point, include data from 25 wells for adipocytes and from 10 wells for macrophages. b Phospholipid efflux: Cells were labeled with [³H]-choline and used to determine apoA-I-induced PL (PC + SM) efflux as indicated in Section “Phospholipid efflux assays”. apoA-I-induced PL efflux was calculated by subtracting the background PL efflux, determined in the absence of apoA-I, from the PL efflux determined in wells containing apoA-I. At each time point, the mean efflux and SD were calculated with data from nine wells for adipocytes (□) and seven wells for macrophages (■)

Fig. 2

ABCA1 protein expression levels in adipocytes and macrophages. a Western blot: Aliquots of the cell homogenates from control and treated adipocytes and macrophages were loaded in the same gels and simultaneously blotted to allow semi-quantitatively estimation of the relative levels of ABCA1 for different treatments and cells. Each lane was loaded with 70 µg of homogenate protein. The Western blot shows the results of one experiment. Three separate experiments provided similar results. b Relative levels of ABCA1 protein: The level of ABCA1 protein was estimated by densitometric scanning of the western blots as explained in methods. The figure shows the relative average levels of ABCA1 normalized by cellular protein (mean ± SD). A total of seven samples from three separate experiments were included in the calculations of the means and corresponding SD. The differences between the means of control and LXR-treated samples of adipocytes, and also between control and cAMP-treated macrophages were estimated to be significant (P < 0.001)

Fig. 3

Effect of ABCA1 expression levels in apoA-I-dependent CL Efflux from adipocytes or macrophages. a Rates of CL efflux normalized by total cellular protein content. The experiments were performed as indicated in “Materials and methods” section. The cells were mock-treated (CONT) or treated with either the LXR-agonist, 2 µM GW3965 (LXR), or 180 µM 8-Br-cAMP (cAMP) for 24 h prior to assays. The rates of apoA-I-induced CL efflux normalized by protein content were determined from the slopes of the time courses (% of cellular CL released into medium vs time). The mean values ± SD for adipocytes were obtained from three independent experiments and include 19 data points. The data for macrophages were obtained from four experiments (n = 15). The differences between the means of control and treated cells were significant (P < 0.001) for both adipocytes and macrophages. b Rates of CL efflux normalized by ABCA1 protein levels. The average rates of apoA-I-induced CL efflux shown in Fig. 3a were converted into nmoles of CL/h-ABCA1 protein using the average relative levels of ABCA1 expression shown in the Fig. 2b and the cholesterol contents of the cells. The lipid compositions of control and treated homogenates showed no significant differences. Each bar represents the mean ± SD. The SDs were calculated using the propagation of errors equation and the SD of ABCA1 levels and rates of efflux. The differences between the means of control and treated cells were significant (P < 0.001) for both adipocytes and macrophages

Fig. 4

Effect of ABCA1-inhibiting drugs on apoA-I-dependent efflux from adipocytes compared to macrophages. The effect of GLYB (500 µM), BFA (36 µM), BLT1 (10 µM), and BLT4 (150 µM) on apoA-I-dependent lipid efflux from macrophages and adipocytes was determined as described in “Methods” section Lipid efflux inhibition assays. a Inhibition of apoA-I-induced CL efflux from adipocytes and macrophages. The bars represent the mean ± SD of 2–4 independent experiments for adipocytes(n = 6–14) and one experiment for macrophages (n = 3). b Effects of the same drugs on inhibition of apoA-I-dependent phospholipid efflux from adipocytes and macrophages. The bars represent the mean ± SD of one experiment with n = 3–4