ApoE and ApoE Nascent-Like HDL Particles at Model Cellular Membranes: Effect of Protein Isoform and Membrane Composition

Abstract

Apolipoprotein E (ApoE), an important mediator of lipid transportation in plasma and the nervous system, plays a large role in diseases such as atherosclerosis and Alzheimer's. The major allele variants ApoE3 and ApoE4 differ only by one amino acid. However, this difference has major consequences for the physiological behaviour of each variant. In this paper, we follow (i) the initial interaction of lipid-free ApoE variants with model membranes as a function of lipid saturation, (ii) the formation of reconstituted High-Density Lipoprotein-like particles (rHDL) and their structural characterisation, and (iii) the rHDL ability to exchange lipids with model membranes made of saturated lipids in the presence and absence of cholesterol [1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) or 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) with and without 20 mol% cholesterol]. Our neutron reflection results demonstrate that the protein variants interact differently with the model membranes, adopting different protein conformations. Moreover, the ApoE3 structure at the model membrane is sensitive to the level of lipid unsaturation. Small-angle neutron scattering shows that the ApoE containing lipid particles form elliptical disc-like structures, similar in shape but larger than nascent or discoidal HDL based on Apolipoprotein A1 (ApoA1). Neutron reflection shows that ApoE-rHDL do not remove cholesterol but rather exchange saturated lipids, as occurs in the brain. In contrast, ApoA1-containing particles remove and exchange lipids to a greater extent as occurs elsewhere in the body.

Keywords: ApoE isoforms; lipid exchange; model membranes; neutron reflection; reconstituted HDL; small-angle neutron scattering.

Figure 1

Schematic representation of protein incorporation into the phospholipid bilayer (A). NR cannot distinguish between the conformation of the protein upon binding to the model membrane: whether there are two individual protein molecules adsorbing either in the core or the headgroup region vs. a single individual protein molecule bending across the membrane (B). Schematic representation for lipid exchange between the rHDL and the phospholipid bilayer (C). Lipid exchange represents the combination of lipid removal from the model membrane and lipid deposited by the rHDL particles (D). The light grey colour represents deuterated lipids while the black colour represents non-deuterated lipids. Both DMPC and POPC lipids were used above their melting temperature, giving fluid supported lipid bilayers. Addition of 20 mol% cholesterol to either POPC or DMPC give SLBs with fluid properties (Waldie et al., 2020).

Figure 2

ApoE interaction with saturated (dDMPC) or unsaturated (dPOPC) model membranes measured at 37°C, in Tris-buffer at pH 7.4: kinetics of lipid replacement in terms of the relevant change in solvent penetration of the lipid core, taking into account SLD change also, are given for ApoE3 and ApoE4 (A). Net lipid removal calculated as the difference in solvent coverage within the bilayers before and after 6 h incubation with apolipoproteins (B) and volume fractions of protein binding within the core (C) or on top of (D) the SLBs upon 6 h of incubation and rinsing with Tris buffer. The NR profiles and best fits are shown in the Supplementary Figures 1, 2. *Statistically different assuming p = 0.1; **Statistically different assuming p = 0.05.

Figure 3

Size exclusion chromatograms for ApoE3 and ApoE4 discs with inset of model used for fitting and negative stained TEM images for ApoE3-rHDL (A) SANS data and best fits in three contrasts for the ApoE3- (B) and ApoE4-rHDL (C) measured at 25°C in Tris buffer pH 7.4. The parameters used for the best fits shown are listed in Table 2.

Figure 4

Lipid removal (A) and lipid exchange (B) across model membranes for ApoE3- and ApoE4-rHDL. Three model membranes were used: saturated lipids (dDMPC) in the absence or presence of hydrogenous or deuterated cholesterol. Data for mature HDL and ApoA1-rHDL against the cholesterol containing bilayers (Waldie et al., 2020) are also included. The asterisk indicates an incubation time of 8 h compared to 6 h for those without. Kinetics of lipid exchange for rHDL containing ApoE3 (C) and ApoE4 (D) in terms of the relative change in SLD of the solvated lipid core over time. HDL data is replotted from Haertlein et al. (2016). The NR profiles and best fits are given in the Supplementary Figures 4, 5. All statistically different assuming p = 0.05 apart from dDMPC + hchol for lipid exchange (B).