ABCA1 overexpression leads to hyperalphalipoproteinemia and increased biliary cholesterol excretion in transgenic mice

Abstract

The discovery of the ABCA1 lipid transporter has generated interest in modulating human plasma HDL levels and atherogenic risk by enhancing ABCA1 gene expression. To determine if increased ABCA1 expression modulates HDL metabolism in vivo, we generated transgenic mice that overexpress human ABCA1 (hABCA1-Tg). Hepatic and macrophage expression of hABCA1 enhanced macrophage cholesterol efflux to apoA-I; increased plasma cholesterol, cholesteryl esters (CEs), free cholesterol, phospholipids, HDL cholesterol, and apoA-I and apoB levels; and led to the accumulation of apoE-rich HDL1. ABCA1 transgene expression delayed 125I-apoA-I catabolism in both liver and kidney, leading to increased plasma apoA-I levels, but had no effect on apoB secretion after infusion of Triton WR1339. Although the plasma clearance of HDL-CE was not significantly altered in hABCA1-Tg mice, the net hepatic delivery of exogenous 3H-CEt-HDL, which is dependent on the HDL pool size, was increased 1.5-fold. In addition, the cholesterol and phospholipid concentrations in hABCA1-Tg bile were increased 1.8-fold. These studies show that steady-state overexpression of ABCA1 in vivo (a) raises plasma apoB levels without altering apoB secretion and (b) raises plasma HDL-C and apoA-I levels, facilitating hepatic reverse cholesterol transport and biliary cholesterol excretion. Similar metabolic changes may modify atherogenic risk in humans.

Figure 1

Hepatic and macrophage expression of human ABCA1 and plasma lipoprotein analysis in control and hABCA1-Tg mice. (a) Northern blot analysis of total RNA (20 μg) from representative control (C) and hABCA1-A Tg male mice. Human ABCA1 and β-actin cDNAs were utilized as probes (top and middle). Immunoblot analysis of mouse liver and macrophage homogenates (25 μg each) was performed using an anti-human ABCA1 polyclonal antibody (bottom). (b) Two-dimensional gel electrophoresis analysis of 15 μl of pooled plasma (n = 5; each group) from control and hABCA1-A Tg mice. (c) FPLC elution profile of pooled plasma (60 μl) from control (filled diamonds) and hABCA1-A Tg (open squares) mice. Ten microliters of FPLC fractions 1, 2, 3, and 4 was analyzed by immunoblotting as described in Methods (c, inset).

Figure 2

Kinetic analysis of apoA-I and CEt HDL catabolism. ¹²⁵I apoA-I and ³H-CEt HDLs were injected into male hABCA1-B Tg and C57BL/6 mice (n = 4 mice per group). Values indicate the percent remaining counts in plasma compared with the 1-minute value expressed as mean ± SEM. The FCRs were determined from the area under the curve using an exponential curve fitting technique on the WINSAAM program using data points through 24 hours of study. ApoA-I HDL FCRs calculated using data points through 48 hours of study were similar (controls = 2.0 ± 0.1 d^–1 and hABCA1-Tg = 1.30 ± 0.04 d^–1; P < 0.01).

Figure 3

VLDL-apoB production after Triton WR1399 infusion. Control (n = 5) and hABCA1-A Tg (n = 5) mice were injected with Triton WR1399 to block lipolysis and ³⁵S-methionine to label newly synthesized proteins. The mice were bled 3 hours after Triton infusion and VLDL isolated by ultracentrifugation (d = 1.006) was analyzed by either immunoblotting with anti-apoB antibodies or quantification of ³⁵S-methionine incorporation into apoB-100 and apoB-48 (see below). The mean apoB value for control mice (n = 5) was calculated. The value of hABCA1-A Tg mice (n = 5) expressed as a percentage relative to the mean apoB value for control mice was: for apoB-100, 95 ± 11%, and for apoB-48, 103 ± 7% (P > 0.05). STD, standard.