A small-angle neutron scattering study of the physical mechanism that drives the action of a viral fusion peptide

Highlights

• Mechanism of action of a fusion peptide derived from HIV-1 was studied.

• The peptide distorts the lipid bilayer structure in a conformation-dependent manner.

• Membrane deformation by the peptide in the β-sheet conformation drives concentration-dependent fusion.

Abstract

Viruses have evolved a variety of ways for delivering their genetic cargo to a target cell. One mechanism relies on a short sequence from a protein of the virus that is referred to as a fusion peptide. In some cases, the isolated fusion peptide is also capable of causing membranes to fuse. Infection by HIV-1 involves the 23 amino acid N-terminal sequence of its gp41 envelope protein, which is capable of causing membranes to fuse by itself, but the mechanism by which it does so is not fully understood. Here, a variant of the gp41 fusion peptide that does not strongly promote fusion was studied in the presence of vesicles composed of a mixture of unsaturated lipids and cholesterol by small-angle neutron scattering and circular dichroism spectroscopy to improve the understanding of the mechanism that drives vesicle fusion. The peptide concentration and cholesterol content govern both the peptide conformation and its impact on the bilayer structure. The results indicate that the mechanism that drives vesicle fusion by the peptide is a strong distortion of the bilayer structure by the peptide when it adopts the β-sheet conformation.

Introduction

The gp41 glycoprotein from the surface of the HIV-1 virus interacts directly with the cell membrane to effect fusion of the virus with the target cell (Freed and Martin, 1995). It contains a 23-residue long sequence at its N-terminus that anchors the virus to the cell (Freed and Martin, 1995). The N-terminal sequence is referred to as the fusion peptide (FP) and has the sequence AVGIGALFLGFLGAAGSTMGARS. The FP is rich in hydrophobic amino acids, supporting the idea that it interacts directly with the lipid bilayer. Mutant forms of HIV-1 with a less hydrophobic FP are less able to fuse with a target cell membrane (Felser et al., 1989; Freed et al., 1990; Bergeron et al., 1992; Freed et al., 1992), further supporting the idea that the FP interacts with the cell membrane.

Studies of the mechanism of HIV-1 FP-driven fusion focused on studies of synthetic peptides having the wild-type sequence, as well as point mutations, interacting with model lipid bilayer membranes (Rafalski et al., 1990; Slepushkin et al., 1990; Nieva et al., 1994; Martin et al., 1996; Pereira et al., 1997; Curtain et al., 1999; Mobley et al., 1999; Saez-Cirion and Nieva, 2002; Castano and Desbat, 2005). Neutral phospholipids and lower peptide concentrations display less vesicle fusion and correspond to the peptide being bound to the membranes in an α-helical conformation (Rafalski et al., 1990; Nieva et al., 1994; Martin et al., 1996; Pereira et al., 1997; Heller and Rai, 2017; Heller and Zolnierczuk, 2019). A β-sheet conformation, which is deemed responsible for FP-driven vesicle fusion, is found at higher peptide concentrations and when charged lipids are present (Rafalski et al., 1990; Nieva et al., 1994; Martin et al., 1996; Mobley et al., 1999; Heller and Rai, 2017; Heller and Zolnierczuk, 2019). The presence of divalent cations can also promote this fusogenic FP conformation (Nieva et al., 1994; Pereira et al., 1997; Saez-Cirion and Nieva, 2002). The Ebola FP (Agopian and Castano, 2014) and the FP from Paramyxovirus PIV5 (Yao and Hong, 2013) are two other examples of viral FPs that act similarly, which suggests that a common mechanism drives vesicle fusion.

To better understand the mechanisms that drive the helix-to-sheet transition and fusion, a mutant form of the gp41 FP was developed to be less fusogenic than the wild-type FP. The peptide, referred to as gp41rk, has the sequence RKGIGALFLGFLGAAGSTMKR. This peptide displays a concentration-dependent helix-to-sheet conformational transition in vesicles composed of mixtures of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) with either 1,2-dimyristoyl-sn-glycero-3-phospho-L-serine (DMPS) or 1,2-dimyristoyl-sn-glycero-3-phospho(1’-rac-glycerol) DMPG, but it does not strongly promote vesicle fusion (Heller and Rai, 2017; Heller and Zolnierczuk, 2019). When associated with DMPC:DMPS vesicles, the peptide remains in the α-helical conformation when 1/1000 ≤ P/L ≤ 1/200, but adopts a β-sheet conformation at P/L = 1/50 (Heller and Rai, 2017). The α-helical state of the peptide did not strongly perturb the structure of the bilayer (Heller and Rai, 2017), due in part to the low peptide concentrations. In contrast, the β-sheet conformation thickens the bilayer, on average, which was attributed to regions of the vesicle having more ordered chains (Heller and Rai, 2017). A later study using circular dichroism (CD) spectroscopy in 7:3 M mixtures of DMPC:DMPG revealed that the gp41rk helix-to-sheet transition in this lipid mixture occurs between P/L = 1/150 and P/L = 1/100 (Heller and Zolnierczuk, 2019). Neutron scattering measurements revealed that the β-sheet conformation of gp41rk causes the bilayer to stiffen, while the α-helical conformation did not alter the mechanical properties of the bilayer (Heller and Zolnierczuk, 2019). While these studies provided important insights into how the helix-to-sheet transition alters the bilayer, the relationship between the FP conformational transition and changes to the bilayer structure could not be correlated with vesicle fusion since it was not observed in these lipid mixtures at the peptide concentrations studied.

Here, small-angle neutron scattering (SANS) was used to develop a better understanding of the physical mechanism that drives FP-induced vesicle fusion. The gp41rk peptide was studied associating with vesicles composed of a 7:3 mixture of deuterium labeled 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) at 0, 20 and 30 mol% cholesterol (Chol) is present. Chol is known to impact the conformation and fusogenicity of the gp41 FP (Lai et al., 2012; Lai and Freed, 2014; Yang et al., 2001a; Yang et al., 2001b; Yang and Weliky, 2003; Wasniewski et al., 2004; Sackett and Shai, 2005; Buzón and Cladera, 2006; Qiang et al., 2007; Bodner et al., 2008; Qiang et al., 2008; Qiang and Weliky, 2009; Sackett et al., 2010; Schmick and Weliky, 2010; Sackett et al., 2014; Jia et al., 2015). Circular dichroism confirmed that gp41rk adopts an α-helical or β-sheet conformation in a concentration and Chol-dependent manner. Vesicle fusion was inferred to have taken place from the SANS data collected at the highest peptide concentration studied in the vesicles containing 30 mol% Chol. Analysis of the remaining SANS data demonstrated that gp41rk has little impact on the structure of the bilayer when no Chol is present. However, at 30 mol% Chol, the peptide alters the structure of the bilayer significantly, suggesting that the underlying mechanism that drives vesicle fusion by the peptide is the free energy cost of deforming the structure of the bilayer.