Structural studies on membrane-embedded influenza hemagglutinin and its fragments

A new in vitro hemagglutinin inhibitor screening system based on a single-vesicle fusion assay

Hemagglutinin (HA) from the influenza virus plays a pivotal role in the infection of host mammalian cells and is, therefore, a druggable target, similar to neuraminidase. However, research involving the influenza virus must be conducted in facilities certified at or above Biosafety Level 2 because of the potential threat of the contagiousness of this virus. To develop a new HA inhibitor screening system without intact influenza virus, we conceived a single-vesicle fusion assay using full-length recombinant HA. In this study, we first showed that full-length recombinant HA can mediate membrane fusion in ensemble and single-vesicle fusion assays. The fluorescence resonance energy transfer (FRET) frequency pattern of single-vesicle complexes completely differed when the inhibitors targeted the HA1 or HA2 domain of HA. This result indicates that analysing the FRET patterns in this assay can provide information regarding the domains of HA inhibited by compounds and compounds' inhibitory activities. Therefore, our results suggest that the assay developed here is a promising tool for the discovery of anti-influenza virus drug candidates as a new in vitro inhibitor screening system against HA from the influenza virus.

Membrane orientation of Gα(i)β(1)γ(2) and Gβ(1)γ(2) determined via combined vibrational spectroscopic studies

The manner in which the heterotrimeric G protein complexes Gβ1γ2 and Gαiβ1γ2 interact with membranes is likely related to their biological function. We combined complementary measurements from sum frequency generation (SFG) vibrational and attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy to determine the possible membrane orientations of Gβ1γ2 and the Gαiβ1γ2 heterotrimer more precisely than could be achieved using SFG alone. The most likely orientations of Gβ1γ2 and the Gαiβ1γ2 heterotrimer were both determined to fall within a similar narrow range of twist and tilt angles, suggesting that Gβ1γ2 may bind to Gαi without a significant change in orientation. This "basal" orientation seems to depend primarily on the geranylgeranylated C-terminus of Gγ2 along with basic residues at the N-terminus of Gαi, and suggests that activated G protein-coupled receptors (GPCRs) must reorient G protein heterotrimers at lipid bilayers to catalyze nucleotide exchange. The innovative methodologies developed in this paper can be widely applied to study the membrane orientation of other proteins in situ.

Attenuated total reflection IR spectroscopy as a tool to investigate the orientation and tertiary structure changes in fusion proteins

Membrane fusion proceeds via a merging of two lipid bilayers and a redistribution of aqueous contents and bilayer components. It involves transition states in which the phospholipids are not arranged in bilayers and in which the monolayers are highly curved. Such transition states are energetically unfavourable since biological membranes are submitted to strong repulsive hydration electrostatic and steric barriers. Viral membrane proteins can help to overcome these barriers. Viral proteins involved in membrane fusion are membrane associated and the presence of lipids restricts drastically the potential of methods (RMN, X-ray crystallography) that have been used successfully to determine the tertiary structure of soluble proteins. We describe here how IR spectroscopy allows to solve some of the problems related to the lipid environment. The principles of the method, the experimental setup and the preparation of the samples are briefly described. A few examples illustrate how attenuated total reflection Fourier-transform IR (ATR-FTIR) spectroscopy can be used to gain information on the orientation and the accessibility to the water phase of the fusogenic domain of viral proteins. Recent developments suggest that the method could also be used to detect changes located in the membrane domains and to identify intermediate structural states involved in the fusion process.

Early steps of the conformational change of influenza virus hemagglutinin to a fusion active state: stability and energetics of the hemagglutinin

A conformational change of the homotrimeric glycoprotein hemagglutinin (HA) of influenza virus mediates fusion between the viral envelope and the endosome membrane. The conformational change of the HA ectodomain is triggered by the acidic pH of the endosome lumen. An essential step of the conformational change is the formation of an extended coiled-coil motif exposing the hydrophobic fusion peptide toward the target membrane. The structures of the neutral-pH, non-fusion active conformation of the HA ectodomain and of a fragment of the ectodomain containing the coiled-coil motif are known. However, it is not known by which mechanism protonation triggers the conformational change of the stable neutral-pH conformation of the ectodomain. Here, recent studies on the stability of the HA ectodomain at neutral pH, the energetics of the conformational change toward the fusion-active state and of the unfolding of the HA ectodomain are summarised. A model for the early steps of the conformational change of the HA ectodomain is presented. The model implicates that protonation leads to a partial dissociation of the distal domains of the HA monomers that is driven by electrostatic repulsion. The opening of the ectodomain enables water to enter the ectodomain. The interaction of water with respective sequences originally shielded from contact with water drives the formation of the coiled-coil structure.

Hypothesis: spring-loaded boomerang mechanism of influenza hemagglutinin-mediated membrane fusion

Substantial progress has been made in recent years to augment the current understanding of structures and interactions that promote viral membrane fusion. This progress is reviewed with a particular emphasis on recently determined structures of viral fusion domains and their interactions with lipid membranes. The results from the different structural and thermodynamic experimental approaches are synthesized into a new proposed mechanism, termed the "spring-loaded boomerang" mechanism of membrane fusion, which is presented here as a hypothesis.

Reversible stages of the low-pH-triggered conformational change in influenza virus hemagglutinin

The refolding of the prototypic fusogenic protein hemagglutinin (HA) at the pH of fusion is considered to be a concerted and irreversible discharge of a loaded spring, with no distinct intermediates between the initial and final conformations. Here, we show that HA refolding involves reversible conformations with a lifetime of minutes. After reneutralization, low pH-activated HA returns from the conformations wherein both the fusion peptide and the kinked loop of the HA2 subunit are exposed, but the HA1 subunits have not yet dissociated, to a structure indistinguishable from the initial one in functional, biochemical and immunological characteristics. The rate of the transition from reversible conformations to irreversible refolding depends on the pH and on the presence of target membrane. Importantly, recovery of the initial conformation is blocked by the interactions between adjacent HA trimers. The existence of the identified reversible stage of refolding can be crucial for allowing multiple copies of HA to synchronize their release of conformational energy, as required for fusion.

Targeting influenza virosomes to ovarian carcinoma cells

Reconstituted influenza virus envelopes (virosomes) containing the viral hemagglutinin (HA) have attracted attention as delivery vesicles for cytosolic drug delivery as they possess membrane fusion activity. Here, we show that influenza virosomes can be targeted towards ovarian carcinoma cells (OVCAR-3) with preservation of fusion activity. This was achieved by incorporating poly(ethylene glycol) (PEG)-derivatized lipids into the virosome membrane. This PEG layer serves as shield to prevent interaction of HA with ubiquitous sialic acid residues and as spatial anchor for antibody attachment. Coupling of Fab' fragments of mAb 323/A3 (anti-epithelial glycoprotein-2) to the distal ends of PEG lipids resulted in specific binding of virosomes to OVCAR-3 cells. These antibody-redirected virosomes fused with membranes of OVCAR-3 cells in a pH-dependent fashion.

Attenuated total reflection IR spectroscopy as a tool to investigate the structure, orientation and tertiary structure changes in peptides and membrane proteins

During the last few years, attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) has become one of the most powerful methods to determine the structure of biological materials and in particular of components of biological membranes, like proteins that cannot be studied by x-ray crystallography and NMR. ATR-FTIR requires a little amount of material (1-100 microg) and spectra are recorded in a matter of minutes. The environment of the molecules can be modulated so that their conformation can be studied as a function of temperature, pressure, pH, as well as in the presence of specific ligands. For instance, replacement of amide hydrogen by deuterium is extremely sensitive to environmental changes and the kinetics of exchange can be used to detect tertiary conformational changes in the protein structure. Moreover, in addition to the conformational parameters that can be deduced from the shape of the infrared spectra, the orientation of various parts of the molecule can be estimated with polarized IR. This allows more precise analysis of the general architecture of the membrane molecules within the biological membranes. The present review focuses on ATR-IR as an experimental approach of special interest for the study of the structure, orientation, and tertiary structure changes in peptides and membrane proteins.

pH-dependent self-association of influenza hemagglutinin fusion peptides in lipid bilayers

We have recently designed a host-guest peptide system that allows us to quantitatively measure the energetics of interaction of viral fusion peptides with lipid bilayers. Here, we show that fusion peptides of influenza hemagglutinin reversibly associate with one another at membrane surfaces above critical surface concentrations, which range from one to five peptides per 1000 lipids in the systems that we investigated. It is further demonstrated by using circular dichroism and Fourier transform infrared spectroscopy that monomeric peptides insert into the bilayers in a predominantly alpha-helical conformation, whereas self-associated fusion peptides adopt predominantly antiparallel beta-sheet structures at the membrane surface. The two forms are readily interconvertible and the equilibrium between them is determined by the pH and ionic strength of the surrounding solution. Lowering the pH favors the monomeric alpha-helical conformation, whereas increasing the ionic strength shifts the equilibrium towards the membrane-associated beta-aggregates. The binding data are interpreted in terms of a cooperative binding model that yields free energies of insertion and free energies of self-association for each of the peptides studied at pH 7.4 and pH 5. At pH 5 and 35 mM ionic strength, the insertion energy of the 20 residue influenza hemagglutinin fusion peptide is -7.2 kcal/mol and the self-association energy is -1.9 kcal/mol. We propose that self-association of fusion peptides could be a major driving force for recruiting a small number of hemagglutinin trimers into a fusion site.

Membrane fusion mechanisms: the influenza hemagglutinin paradigm and its implications for intracellular fusion

The mechanism of membrane fusion induced by the influenza virus hemagglutinin (HA) has been extensively characterized. Fusion is triggered by low pH, which induces conformational changes in the protein, leading to insertion of a hydrophobic 'fusion peptide' into the viral membrane and the target membrane for fusion. Insertion perturbs the target membrane, and hour glass-shaped lipidic fusion intermediates, called stalks, fusing the outer monolayers of the two membranes, are formed. Stalk formation is followed by complete fusion of the two membranes. Structures similar to those formed by HA at the pH of fusion are found not only in many other viral fusion proteins, but are also formed by SNAREs, proteins involved in intracellular fusion. Substances that inhibit or promote HA-induced fusion because they affect stalk formation, also inhibit or promote intracellular fusion, cell-cell fusion and even intracellular fission similarly. Therefore, the mechanism of influenza HA-induced fusion may be a paradigm for many intracellular fusion events.

Membrane fusion mediated by coiled coils: a hypothesis

A molecular model of the low-pH-induced membrane fusion by influenza hemagglutinin (HA) is proposed based upon the hypothesis that the conformational change to the extended coiled coil creates a high-energy hydrophobic membrane defect in the viral envelope or HA expressing cell. It is known that 1) an aggregate of at least eight HAs is required at the fusion site, yet only two or three of these HAs need to undergo the "essential" conformational change for the first fusion pore to form (Bentz, J. 2000. Biophys. J. 78:000-000); 2) the formation of the first fusion pore signifies a stage of restricted lipid flow into the nascent fusion site; and 3) some HAs can partially insert their fusion peptides into their own viral envelopes at low pH. This suggests that the committed step for HA-mediated fusion begins with a tightly packed aggregate of HAs whose fusion peptides are inserted into their own viral envelope, which causes restricted lateral lipid flow within the HA aggregate. The transition of two or three HAs in the center of the aggregate to the extended coiled coil extracts the fusion peptide and creates a hydrophobic defect in the outer monolayer of the virion, which is stabilized by the closely packed HAs. These HAs are inhibited from diffusing away from the site to admit lateral lipid flow, in part because that would initially increase the surface area of hydrophobic exposure. The other obvious pathway to heal this hydrophobic defect, or some descendent, is recruitment of lipids from the outer monolayer of the apposed target membrane, i.e., fusion. Other viral fusion proteins and the SNARE fusion protein complex appear to fit within this hypothesis.

Minimal aggregate size and minimal fusion unit for the first fusion pore of influenza hemagglutinin-mediated membrane fusion

The data of Melikyan et al. (J. Gen. Physiol. 106:783, 1995) for the time required for the first measurable step of fusion, the formation of the first flickering conductivity pore between influenza hemagglutinin (HA) expressing cells and planar bilayers, has been analyzed using a new mass action kinetic model. The analysis incorporates a rigorous distinction between the minimum number of HA trimers aggregated at the nascent fusion site (which is denoted the minimal aggregate size) and the number of those trimers that must to undergo a slow essential conformational change before the first fusion pore could form (which is denoted the minimal fusion unit). At least eight (and likely more) HA trimers aggregated at the nascent fusion site. Remarkably, of these eight (or more) HAs, only two or three must undergo the essential conformational change slowly before the first fusion pore can form. Whether the conformational change of these first two or three HAs are sufficient for the first fusion pore to form or whether the remaining HAs within the aggregate must rapidly transform in a cooperative manner cannot be determined kinetically. Remarkably, the fitted halftime for the essential HA conformational change is roughly 10(4) s, which is two orders of magnitude slower than the observed halftime for fusion. This is because the HAs refold with distributed kinetics and because the conductance assay monitored the very first aggregate to succeed in forming a first fusion pore from an ensemble of hundreds or thousands (depending upon the cell line) of fusogenic HA aggregates within the area of apposition between the cell and the planar bilayer. Furthermore, the average rate constant for this essential conformational change was at least 10(7) times slower than expected for a simple coiled coil conformational change, suggesting that there is either a high free energy barrier to fusion and/or very many nonfusogenic conformations in the refolding landscape. Current models for HA-mediated fusion are examined in light of these new constraints on the early structure and evolution of the nascent fusion site. None completely comply with the data.

Structure of influenza haemagglutinin at neutral and at fusogenic pH by electron cryo-microscopy

The three-dimensional structures of the complete haemagglutinin (HA) of influenza virus A/Japan/305/57 (H2N2) in its native (neutral pH) and membrane fusion-competent (low pH) form by electron cryo-microscopy at a resolution of 10 A and 14 A, respectively, have been determined. In the fusion-competent form the subunits remain closely associated preserving typical overall features of the trimeric ectodomain at neutral pH. Rearrangements of the tertiary structure in the distal and the stem parts are associated with the formation of a central cavity through the entire ectodomain. We suggest that the cavity is essential for relocation of the so-called fusion sequence of HA towards the target membrane.

Poly(ethylene glycol) (PEG)-mediated fusion between pure lipid bilayers: a mechanism in common with viral fusion and secretory vesicle release?

Membrane fusion is fundamental to the life of eukaryotic cells. Cellular trafficking and compartmentalization, import of food stuffs and export of waste, inter-cellular communication, sexual reproduction, and cell division are all dependent on this basic process. Yet, little is known about the molecular mechanism(s) by which fusion occurs. It is known that fusing membranes must somehow be docked and brought into close contact. Specific proteins, many of which have been identified within the past decade, accomplish this. An electrical connection or 'fusion pore' is established between compartments surrounded by the fusing membranes. Three primary views of the mechanism of pore formation during secretory and viral fusion have been proposed within the past decade. In one view, a protein ring forms an initial transient connection that expands slowly by recruiting lipid so as to form a lipidic junction. In another view, the initial fusion pore consists of a protein-lipid complex that transforms slowly until the fusion proteins dissociate from the complex to form an irreversible lipidic pore. In a third view, the initial pore is a transient lipid pore that fluctuates between open and closed states before either expanding irreversibly or closing. Recent work has helped define the mechanism by which poly(ethylene glycol) (PEG) mediates fusion of highly curved model membranes composed only of synthetic phospholipids. PEG is a highly hydrated polymer that can bring vesicle membranes to near molecular contact by making water between them thermodynamically unfavourable. Disrupted packing in the contacting monolayers of these vesicle membranes is necessary to induce fusion. The time course and sequence of molecular events of the ensuing fusion process have also been defined. This sequence of events involves the formation of an initial, transient intermediate in which outer leaflet lipids have mixed and small transient pores join fusing compartments ('stalk'). The transient intermediate transforms in 1-3 min to a fusion-committed, second intermediate ('septum') that then 'pops' to form the fusion pore. Inner leaflet mixing, which is shown to be distinct from outer leaflet mixing, accompanies contents mixing that marks formation of the fusion pore. Both the sequence of events and the activation energies of these events correspond well to those observed in viral membrane fusion and secretory granule fusion. These results strongly support the contention that both viral and secretory fusion events occur by lipid molecule rearrangements that can be studied and defined through the use of PEG-mediated vesicle fusion as a model system. A possible mechanism by which fusion proteins might mediate this lipidic process is described.

To fuse or not to fuse: the effects of electrostatic interactions, hydrophobic forces, and structural amphiphilicity on protein-mediated membrane destabilization

The development of lipid-based delivery vehicles for therapeutic molecules has become a topic of intense research. Recently, much of this effort has been directed towards mimicking the characteristics of viruses that give them an advantage for the delivery of genetic medicines. One of the most desirable properties of viral-based vectors is the ability to promote the destabilization of the host cell membrane to allow the entry of the genetic medicine into the target cell. This has been found to be largely controlled by the coat proteins on the surface of enveloped viruses. Although the exact mechanism by which proteins involved in the fusion process are able to promote the destabilization of membranes has yet to be elucidated, much understanding based upon information gained from a wide variety of studies is advancing the state of knowledge in this area. Parameters such as hydrophobic and electrostatic interactions as well as structural amphiphilicity, control to a large extent, the nature of the interaction of proteins with membranes. Thus, membrane fusion is mediated primarily by these forces acting in concert with one another. Ultimately, the knowledge gained from these studies will help to develop the ideal delivery system for the next generation of therapeutics.

pH-induced conformational changes of membrane-bound influenza hemagglutinin and its effect on target lipid bilayers

Influenza virus hemagglutinin (HA) has served as a paradigm for both pH-dependent and -independent viral membrane fusion. Although large conformational changes were observed by X-ray crystallography when soluble fragments of HA were subjected to fusion-pH conditions, it is not clear whether the same changes occur in membrane-bound HA, what the spatial relationship is between the conformationally changed HA and the target and viral membranes, and in what way HA perturbs the target membrane at low pH. We have taken a spectroscopic approach using an array of recently developed FTIR techniques to address these questions. Difference attenuated total reflection FTIR spectroscopy was employed to reveal reversible and irreversible components of the pH-induced conformational change of the membrane-bound bromelain fragment of HA, BHA. Additional proteolytic fragments of BHA were produced which permitted a tentative assignment of the observed changes to the HA1 and HA2 subunits, respectively. The membrane-bound HA1 subunit undergoes a reversible conformational change, which most likely involves the loss of a small proportion of beta-sheet at low pH. BHA was found to undergo a partially reversible tilting motion relative to the target membrane upon exposure to pH 5, indicating a previously undescribed hinge near the anchoring point to the target membrane. Time-resolved amide H/D exchange experiments revealed a more dynamic (tertiary) structure of membrane-bound BHA and its HA2, but not its HA1, subunit. Finally BHA and, to a lesser degree, HA1 perturbed the lipid bilayer of the target membrane at the interface, as assessed by spectral changes of the lipid ester carbonyl groups. These results are discussed in the context of a complementary study of HA that was bound to viral membranes through its transmembrane peptide (Gray C, Tamm LK, 1997, Protein Sci 6:1993-2006). A distinctive role for the HA1 subunit in the conformational change of HA becomes apparent from these combined studies.

Specific single or double proline substitutions in the "spring-loaded" coiled-coil region of the influenza hemagglutinin impair or abolish membrane fusion activity

We tested the role of the "spring-loaded" conformational change in the fusion mechanism of the influenza hemagglutinin (HA) by assessing the effects of 10 point mutants in the region of high coiled-coil propensity, HA2 54-81. The mutants included proline substitutions at HA2 55, 71, and 80, as well as a double proline substitution at residues 55 and 71. Mutants were expressed in COS or 293T cells and assayed for cell surface expression and structural features as well as for their ability to change conformation and induce fusion at low pH. We found the following: Specific mutations affected the precise carbohydrate structure and folding of the HA trimer. All of the mutants, however, formed trimers that could be expressed at the cell surface in a form that could be proteolytically cleaved from the precursor, HA0, to the fusion-permissive form, HA1-S-S-HA2. All mutants reacted with an antibody against the major antigenic site and bound red blood cells. Seven out of ten mutants displayed a wild-type (wt) or moderately elevated pH dependence for the conformational change. V55P displayed a substantial reduction (approximately 60- 80%) in the initial rate of lipid mixing. The other single mutants displayed efficient fusion with the same pH dependence as wt-HA. The double proline mutant V55P/ S71P displayed no fusion activity despite being well expressed at the cell surface as a proteolytically cleaved trimer that could bind red blood cells and change conformation at low pH. The impairment in fusion for both V55P and V55P/S71P was at the level of outer leaflet lipid mixing. We interpret our results in support of the hypothesis that the spring-loaded conformational change is required for fusion. An alternate model is discussed.