APOA2 increases cholesterol efflux capacity to plasma HDL by displacing the C-terminus of resident APOA1

Abstract

The ability of high-density lipoprotein (HDL) to promote cellular cholesterol efflux is a more robust predictor of cardiovascular disease protection than HDL-cholesterol levels in plasma. Previously, we found that lipidated HDL containing both apolipoprotein A-I (APOA1) and A-II (APOA2) promotes cholesterol efflux via the ATP-binding cassette transporter (ABCA1). In the current study, we directly added purified, lipid-free APOA2 to human plasma and found a dose-dependent increase in whole plasma cholesterol efflux capacity. APOA2 likewise increased the cholesterol efflux capacity of isolated HDL with the maximum effect occurring when equal masses of APOA1 and APOA2 coexisted on the particles. Follow-up experiments with reconstituted HDL corroborated that the presence of both APOA1 and APOA2 were necessary for the increased efflux. Using limited proteolysis and chemical cross-linking mass spectrometry, we found that APOA2 induced a conformational change in the N- and C-terminal helices of APOA1. Using reconstituted HDL with APOA1 deletion mutants, we further showed that APOA2 lost its ability to stimulate ABCA1 efflux to HDL if the C-terminal domain of APOA1 was absent, but retained this ability when the N-terminal domain was absent. Based on these findings, we propose a model in which APOA2 displaces the C-terminal helix of APOA1 from the HDL surface which can then interact with ABCA1-much like it does in lipid-poor APOA1. These findings suggest APOA2 may be a novel therapeutic target given this ability to open a large, high-capacity pool of HDL particles to enhance ABCA1-mediated cholesterol efflux.

Keywords: ABCA1; Apolipoprotein A-I; Apolipoprotein A-II; Apolipoprotein(s); Cholesterol Efflux; High Density Lipoproteins; Lipoprotein metabolism; Lipoproteins; Structure.

Fig. 1

Effect of APOA2 addition on cholesterol efflux and lipoprotein distribution in human plasma. A: ABCA1-mediated cholesterol efflux from J774 cells to either 2% human plasma or 2% human plasma with APOB-containing lipoproteins depleted by PEG precipitation. Plasma and PEG-depleted plasma samples were incubated with increasing amounts of exogenous APOA2, before the cholesterol efflux assay wherein ABCA1 expression was induced by cAMP. Data points and error bars represent the mean and SD from three technical replicates, respectively. B: phospholipid profile of human plasma incubated with increasing amounts of APOA2 from panel A. Plasma was incubated with purified human APOA2 for 1 h at 37°C and then applied to tandem Superose 6 size-exclusion columns. Each fraction was analyzed for choline-containing phospholipids using a colorimetric assay.

Fig. 2

Effect of APOA2 on ABCA1-mediated cholesterol efflux to isolated HDL. A: size-exclusion profiles of UC-isolated HDL incubated with increasing concentration of exogenous APOA2 for 1 h and immediately separated on tandem Superdex 200 columns. Peaks eluting between 20-25 ml and 25–30 ml were HDL-bound (lipidated) and displaced (lipid-free) protein fractions, respectively. B: SDS-PAGE-based densitometry analysis of HDL-bound fractions shows the change in APOA1 and APOA2 content with increasing addition of exogenous APOA2. The APOA2:APOA1 band intensity ratios are indicated above each APOA2 data point. C: cholesterol efflux ability of HDL-bound fractions to mediate cholesterol efflux in BHK cells with and without ABCA1 expression. Data points and error bars represent the mean and SDs from three technical replicates, respectively.

Fig. 3

Ability of different scaffold proteins in reconstituted HDL particles to promote ABCA1-mediated cholesterol efflux. A: cholesterol efflux from BHK cells with ABCA1 expression for reconstituted HDL (rHDL) containing APOA1 alone, APOA2 alone, and APOA1/APOA2 showing that both APOA1 and APOA2 are required for enhanced ABCA1-mediated cholesterol efflux. Particles were compared at equal phospholipid concentration. B: native PAGE analysis of reconstituted HDL particles produced with POPC. Lanes 1, 2, and 3 show reconstituted HDL particles containing APOA1, APOA2, and APOA1/APOA2, respectively. C: SDS-PAGE analysis (nonreducing) of the rHDL particles showing APOA1 (28 kDa) and/or APOA2 (14 kDa). Lanes 1, 2, and 3 show reconstituted HDL particles containing APOA1, APOA2, and APOA1/APOA2 particles. All gels were loaded with equal protein and stained with Coomassie blue. Bars and error bars represent the mean and SDs from three technical replicates, respectively. Star (∗) represents a difference from the APOA2-only rHDL particle by Kruskal-Wallis and Dunn’s post-hoc test at P < 0.05.

Fig. 4

Structural changes in APOA1 in the presence of exogenous APOA2. Top bar shows cartoon with the 10 amphipathic alpha-helical repeats as reported by Segrest et al. as a reference for the structural barcodes. Residue-level structural changes in APOA1 in the presence of 20% (middle) and 40% (bottom) exogenous APOA2 are represented as 2-D structural barcodes of APOA1. The log2 fold changes of significant semi-tryptic peptides from five biological replicates (P-value < 0.01) are distributed across the encompassing residues. Regions in red get more exposed to the nonspecific protease and the regions shown in the blue are shielded from the nonspecific protease upon the exogenous addition of APOA2.

Fig. 5

Cholesterol efflux to rHDL particles produced with various mutants of APOA1. Cholesterol efflux from BHK cells expressing ABCA1 is shown. All the particles were reconstituted with POPC at an approximate wt:wt ratio of 2.5 mg POPC/1.0 mg WT APOA1 or mutant. Gray bars show the rHDL particles containing the WT APOA1 or mutant indicated alone. Black bars show the same particle with added APOA2 (at about 20% of APOA1 mass). Data and error bars represent the mean and SDs from six technical replicates. The data from individual replicates is shown along each bar. ∗∗P < 0.01 by Mann-Whitney test compared to corresponding non-APOA2 containing rHDL particle.

Fig. 6

Cartoon showing mechanism of increase in ABCA-1-mediated cholesterol efflux induced by exogenous APOA2. Lipid-free APOA1 interacts with ABCA1 through its exposed C-termini and acts as an efficient acceptor of cholesterol. However, in mature HDL, the APOA1 termini are in close contact with the lipoprotein surface, hindering interaction in ABCA1. Driven by the structural and functional evidence at hand, we propose that APOA2 leads to the dislodgement of Helix 10 away from the lipid surface, making it available to form auxiliary interactions with ABCA1, thereby increasing the rate of apolipoprotein-mediated cholesterol efflux.