Influence of C-terminal α-helix hydrophobicity and aromatic amino acid content on apolipoprotein A-I functionality

Abstract

The apoA-I molecule adopts a two-domain tertiary structure and the properties of these domains modulate the ability to form HDL particles. Thus, human apoA-I differs from mouse apoA-I in that it can form smaller HDL particles; the C-terminal α-helix is important in this process and human apoA-I is unusual in containing aromatic amino acids in the non-polar face of this amphipathic α-helix. To understand the influence of these aromatic amino acids and the associated high hydrophobicity, apoA-I variants were engineered in which aliphatic amino acids were substituted with or without causing a decrease in overall hydrophobicity. The variants human apoA-I (F225L/F229A/Y236A) and apoA-I (F225L/F229L/A232L/Y236L) were compared to wild-type (WT) apoA-I for their abilities to (1) solubilize phospholipid vesicles and form HDL particles of different sizes, and (2) mediate cellular cholesterol efflux and create nascent HDL particles via ABCA1. The loss of aromatic residues and concomitant decrease in hydrophobicity in apoA-I (F225L/F229A/Y236A) has no effect on protein stability, but reduces by a factor of about three the catalytic efficiencies (V(max)/K(m)) of vesicle solubilization and cholesterol efflux; also, relatively large HDL particles are formed. With apoA-I (F225L/F229L/A232L/Y236L) where the hydrophobicity is restored by the presence of only leucine residues in the helix non-polar face, the catalytic efficiencies of vesicle solubilization and cholesterol efflux are similar to those of WT apoA-I; this variant forms smaller HDL particles. Overall, the results show that the hydrophobicity of the non-polar face of the C-terminal amphipathic α-helix plays a critical role in determining apoA-I functionality but aromatic amino acids are not required. This article is part of a Special Issue entitled Advances in High Density Lipoprotein Formation and Metabolism: A Tribute to John F. Oram (1945-2010).

Figure 1

Helical wheel projections of the C-terminal amphipathic α-helical region of apoA-I variants. A. Residues 220–241 of WT human apoA-I; aromatic residues in the nonpolar face are shaded grey. B. Residues 220–241 of human apoA-I (F225L/F229A/Y236A); the mutated residues in the nonpolar face are grey-hatched. C. Residues 220–241 of human apoA-I (F225L/F229L/A232L/Y236L); the mutated residues in the nonpolar face are grey-hatched. D. Residues 217–238 of WT mouse apoA-I (this is the equivalent segment to residues 220–241 in human apoA-I because the mouse protein is three residues shorter [23]). The helical wheels are drawn with the Wheel program [54].

Figure 2

Thermal unfolding of WT human apoA-I and C-terminal α-helix variants monitored by the molar ellipticity at 222nm. The protein concentrations were 50μg/ml.

Figure 3

GdnHCl-induced (A) and urea-induced (B) denaturation of human apoA-I variants monitored by Trp fluorescence. WT apoA-I (o, solid line); apoA-I (F225L/F229A/Y236A) (Δ, dashed line); apoA-I (F225L/F229L/A232L/Y236L) (□, dotted line).

Figure 4

Fluorescence spectra of ANS (250 μM) in the presence of 50 μg/ml human apoA-I variants. (a) apoA-I (F225L/F229L/A232L/Y236L), (b) WT apoA-I, (c) apoA-I (F225L/F229A/Y236A), (d) free ANS.

Figure 5

Effect of C-terminal α-helix nonpolar face hydrophobicity on the ability of apoA-I to solubilize DMPC MLV. The fractional decrease in absorbance of the MLV suspension (0.25mg DMPC/ml) at 325nm in 10min by different concentrations of apoA-I was measured as described in Materials and Methods. The data were fitted to the Michaelis-Menten equation to obtain the V_max and K_m values. WT apoA-I (o, solid line); apoA-I (F225L/F229A/Y236A) (Δ, dashed line); apoA-I (F225L/F229L/A232L/Y236L) (□, dotted line).

Figure 6

Native polyacrylamide 4–12% gradient gels stained with Coomassie Brilliant Blue comparing the sizes of discoidal complexes formed by incubation of DMPC/5 mol % cholesterol MLV with human apoA-I variants. The migration positions of standard proteins of known hydrodynamic diameters are indicated on the left: Lane 1, WT apoA-I; Lane 2, apoA-I (F225L/F229A/Y236A); Lane 3, apoA-I (F225L/F229L/A232L/Y236L). Optical scans of lanes 1–3 are shown at the bottom of the figure.

Figure 7

Effects of apoA-I C-terminal α-helix nonpolar face hydrophobicity on the concentration-dependence of cholesterol efflux via ABCA1 from BHK cells. The efflux of [³H]cholesterol in 4h to different concentrations of apoA-I variants was measured as described in Materials and Methods. The V_max and K_m values were calculated by fitting the data to the Michaelis-Menten equation. WT apoA-I (o, solid line); apoA-I (F225L/F229A/Y236A) (Δ, dashed line); apoA-I (F225L/F229L/A232L/Y236L) (□, dotted line).

Figure 8

Blots stained with anti-apoA-I of native polyacrylamide 4–12% gradient gels of medium collected after incubation of BHK cells with apoA-I variants. The cholesterol efflux experiments were performed as described in Fig. 7 and the concentrated medium containing nascent HDL particles was subjected to electrophoresis in the gradient gel. The migration positions of standard proteins of known hydrodynamic diameters are indicated on the left: Lane 1, WT human apoA-I; Lane 2, human apoA-I (F225L/F229A/Y236A); Lane 3, human apoA-I (F225L/F229L/A232L/Y236L).