Identification and Characteristics of Fusion Peptides Derived From Enveloped Viruses

Abstract

Membrane fusion events allow enveloped viruses to enter and infect cells. The study of these processes has led to the identification of a number of proteins that mediate this process. These proteins are classified according to their structure, which vary according to the viral genealogy. To date, three classes of fusion proteins have been defined, but current evidence points to the existence of additional classes. Despite their structural differences, viral fusion processes follow a common mechanism through which they exert their actions. Additional studies of the viral fusion proteins have demonstrated the key role of specific proteinogenic subsequences within these proteins, termed fusion peptides. Such peptides are able to interact and insert into membranes for which they hold interest from a pharmacological or therapeutic viewpoint. Here, the different characteristics of fusion peptides derived from viral fusion proteins are described. These criteria are useful to identify new fusion peptides. Moreover, this review describes the requirements of synthetic fusion peptides derived from fusion proteins to induce fusion by themselves. Several sequences of the viral glycoproteins E1 and E2 of HCV were, for example, identified to be able to induce fusion, which are reviewed here.

Keywords: enveloped viruses; fusion; membranotropic; peptides; secondary structures.

FIGURE 1

Structural and functional comparison between viral HA and eukaryotic SNARE in membrane fusion events. A key difference between these processes is that heterodimerization of the SNAREs is a prerequisite for the vesicular process whereas viral fusion proteins insert and destabilize the lipid bilayers to induce fusion in a more general manner. (A) Viral fusion events are mediated by the binding of fusion proteins to the membrane of the host cell. The proteins undergo a conformational change that allows the membranes to merge. (B) v-SNARE on the vesicle and the t-SNARE on the target membrane bind to one another leading to the formation of the trans-SNARE complex. A more detailed description of this process and its structural features can be found below.

FIGURE 2

General mechanism of fusion process employed by Class I fusion proteins. An environmental trigger such as an acidic pH or the binding to a coreceptor (represented figuratively by the scissors) induces a conformational change that exposes the fusion peptide (A). The fusion peptide then inserts into the host cell membrane causing the fusion proteins to fold back on themselves, inducing the bending of apposed membranes (B). The folding creates a contact between the membranes, leading first to hemifusion (C). Finally, the refolding leads to the formation of the fusion pore (D) and subsequent mixing of contents.

FIGURE 3

Major structural features of membrane fusion processes across the three canonical classes of fusion protein. (Column 1) Membrane fusion driven by viral class I proteins. The model of class I protein-mediated hemifusion and fusion depicts the progression through an extended prehairpin followed by breaking the threefold symmetry and dissociation of the C-heptad repeat domains. In the native prefusion conformation, the paramyxovirus fusion (F) protein consists of globular head domain attached to the transmembrane (TM) domains and short luminal tails through the TM domain-proximal heptad repeat sequences. When fusion commences, major structural rearrangements lead to assembly of the head-domain HR segments into a central, trimeric alpha-helical coiled-coil structure, displacing the fusion peptides in the direction of the host-cell membrane (A). Subsequent hairpin-like refolding (B) then positions the heptad repeat domains into the grooves of the central triple helix, resulting in the formation of the stable six-helix bundle (6HB) post-fusion structure (C), in which the fusion peptide-proximal core coiled-coil structure is surrounded by the three TMD-proximal HR helices; this is a defining feature of fusion proteins of this class. (Column 2) Class II viral protein-mediated fusion. Homo-/hetero-di-/trimers of E1/E2 (D) engage a target membrane with their fusion loops following low pH-induced conformational changes. Subsequent conformational changes involve the reorientation of domain III around a hinge region (E) that positions the TM region and the fusion loop (yellow) proximal to one another. Subsequent refolding of the extended trimeric conformation into a hairpin structure promotes hemifusion and fusion pore formation (F). Despite their marked structural differences, proteins of class I and class II are able to progress through a similar refolding pathway. (Column 3) Membrane merger induced by class III viral fusogens. The native trimer of class III fusion proteins folds into a tripod-like arrangement, on which the fusion loops are positioned at the tip of each leg and are therefore directed into the viral envelope. Low pH conditions result in the protonation of a specific cluster of histidine residues that exert a major destabilizing effect on the prefusion structure. Once triggered, the tripod legs are proposed to swing upward (G), driving the fusion loops toward the target membrane. Repositioning of the domains with conserved tertiary structure relative to each other through secondary structure reorganization in hinge regions and major changes of the trimerization domain (H) then ultimately result in a classic hairpin post-fusion conformation (I).

FIGURE 4

Examples of analytical techniques to characterize peptide’s membranotropic properties. (A) Fusogenic properties: 1) Membrane leakage experiments. After pore formation upon peptide addition, dilution of the self-quenching dye from labeled vesicles to unlabeled vesicles results in fluorescence increase, 2) Lipid mixing with FRET. The increasing distance between two fluorophores composing a FRET system upon lipid mixing can be visualized by monitoring the energy transfer efficiency decrease, 3) Lipid mixing can also be highlighted by dynamic light scattering (DLS) monitoring the vesicle size increase, (B) Insertion or internalization propensity: 1) Peptide insertion can be monitored by following Trp fluorescence. Upon membrane insertion, the hydrophobic environment around Trp results in a blue-shift in fluorescence, 2) NBD/sodium dithionite experiments. NBD-labeled sequences are incubated with liposomes and a reducing agent, sodium dithionite, is added. After reduction, the remaining fluorescence indicate the degree of membrane insertion/internalization, 3) Imaging experiments using fluorescence microscopy, such as confocal or total internal reflection fluorescent (TIRF) microscopy, with labeled peptide and liposomes or cells (C) Structural characterization in contact with model or cellular membranes to visualize conformational changes upon membrane interaction.

FIGURE 5

Design of fusogenic peptides containing different ratios of helix promoting, β-sheet-promoting or helix destabilizing residues (in bold) adapted from Hofmann et al. The first arrow indicates the increase of Val ratio. The second arrow indicates the increase of helix destabilization (Hofmann et al., 2004).

FIGURE 6

Schematic representation of HCV glycoproteins E1 and E2. E1 is composed of an N-terminal domain (NTD, yellow), a putative fusion peptide (PFP, red), a conserved region (CR, blue), and a transmembrane domain (TMD, black). E2 is composed of two hypervariable regions (HVR1 and HVR2, orange) and a transmembrane region (TMD, black). The glycosylation site N250 is specific to genotype 1b/6.

FIGURE 7

Structure of E1 pre-transmembrane region determined by NOE experiments (Spadaccini et al., 2010). Two alpha helices MAWDM (cyan) and AALVVAQLL (orange) separated by a bend illustrated by Trp326 (yellow). PDB code: 2KNU.

FIGURE 8

Schematic representation of HCV E1 and E2 fusion proteins with selected membrane active sequences, and their corresponding properties according to literature. Upper panel: HCV E1 protein with N-terminal domain (green), putative fusion peptide (red), pre-transmembrane domain (blue) and its structure determined by Spadaccini et al. (PDB code 2KNU) and C-terminal region including the transmembrane domain (black). Lower panel: HCV E2 protein and its partial ectodomain structure (determined by Kong et al. (Kong et al., 2013), PDB code 4MWF) with N-terminal domain flanked by hypervariable regions (orange), hydrophobic core region (purple and gold), pre-transmembrane domain (cyan) and C-terminal region transmembrane domain (black).