User:Kinkreet/Protein Science/Protein Membrane Fusion

Eukaroytic cells contain many intracellular compartments and organelles, and requires vesicles to transport cargo as well as to communicate with each other. Vesicles must be able to 'bud' from one compartment and fuse with another.

Membrane fusion is defined as two separate lipid bilayers completely merging into a single continuous bilayer.

Due to the presence of bound water molecules on the lipid head groups, there is a strong repulsion between the bilayers when they are < 3nm apart; therefore, close proximity is not the only thing required for membrane fusion, a 'defect' must also be introduced. The 'defect' must expose the hydrophobic interior of the membrane of one bilayer to that of the other in order to allow for membrane fusion. These 'defects' can be in the form of small hydrophobic peptides, a special lipid composition, or any insertions (e.g VSV glycoprotein G, synaptotagmin-3, EEA1) which causes an unfavourable curve in the membrane.

Membrane fusion is mediated by fusion proteins. Some fusion proteins have only one functional domain (e.g. paraxyvirus fusion protein), whereas others have other functional domains as well (e.g. haemaglutinin) A common model to study membrane fusion and fusion proteins is in virus, because it must fuse with the membrane of the host cell in order to gain entry. Haemagglutinin of influenza is a model system for type I viral-cellular membrane fusion. Soluble E ectodomain (sE) of dengue fever is a model system for type II viral-cellular membrane fusion. Fusion proteins are also classified into two groups: class I is exemplified by HA2 of influenza, and class II is exemplified by Semliki Forest virus fusion protein.

Intracellular proteins have also been extensively studied, and common mechanisms can be found between the two. The fusion proteins first tether onto the opposing membrane and bring them into close proximity to each other. A curvature of the membranes, or a defect causes the formation of a stalk (hemifusion ) then fusion pore forms, which opens and dilates, combining with nearby fusion pores until the two membranes are fully integrated. The bigger the defect or a larger curvature of the membrane will lead to a bigger likelihood of fusion.

Haemagglutinin
There are two essential molecules on the surface of influenza virus that characterize its subtype: neuraminidase and haemagglutinin. Neuraminidase cleaves glycosidic linkages of sialic acids after virions are produced and exported to the cell surface, to allow the virus to be detached from the cell and infect other cells. Haemagglutinin is responsible for binding of virus to host cell.

Receptors on the host cell surface binds to surface molecules (e.g. HA1) on the virus, and endocytose the virus. Under the acidic conditions of the endosome, the virus is able to break free from the endosome and enter into the cytosol. The virus uncoats and releases the viral RNA. The viral RNA enters the nucleus and is replicated. Some of the RNA is used by ribosomes to synthesize new protein components (mostly membrane proteins) to make up a new virion, while some are packaged as viral RNA.

Haemagglutinin is initially synthesized at the endoplasmic reticulum (ER) as a single-chain polypeptide precursor, HA0. HA0 is transported to the Golgi and eventually released from the host cell as part of a new viral particle. Between the trans-Golgi and being released, HA0 is cleaved into HA1 and HA2. HA1 and HA2 subsequently associates with each other to form a complex stabilized by cysteine disulphide bonds. Each complex contains a trimer of the HA1-HA2 complex. The HA1 contains a receptor site that binds to receptors, and so the whole complex contains 3 receptor sites. There are also 4 major antigen variable regions, which determines the specificity of the binding.

HA2 contains a fusion peptide sequence as well as the membrane anchor. The fusion peptide sequence is highly conserved (GLFGAIAGFIENGWEGMIDGWYGF) and located at the N-terminus of HA2, this sequence is highly hydrophobic and thus is buried inside the alpha helices of HA2 to minimize the entropic penalty induced by hydrophobic effects. When in aqueous solution of normal pH, it adopt a random coil conformation; but when endocytosed, the low pH of the endosome induces an irreversible conformational change that causes the fusion peptide to take on a coiled-coil conformation that extrude for 100Å. The fusion peptide can now insert into the membrane of the endosome. Because the fusion peptide is anchored onto the viral membrane using its C-terminus' antiparallel coiled-coil, this insertion brings the endosomal and viral membranes into close proximity.

There are two hypothesis of how this conformational change occurs. The spring-loaded mechanism assumes that the N-terminus is most favourable when in an alpha-helical conformation, but only prevented from being so because of being bound by HA1. When the HA complex is placed at low pH, the interaction between HA1 and HA2 becomes weaker, and HA2 'spring' into its helical conformation.

Alternative models exists where the free energy captured in the metastable form of HA (HAM) is released at low pH and results in HA1 moving away from the fusion peptide to enable its insertion into the target membrane, and work done to pull the two membranes together.

HIV
In human and simian immunodeficiency viruses (HIV and SIV), the surface molecule which allows them to bind are the envelope glycoproteins (Env). Env consists of a transmembrane glycoprotein (gp41) and a surface glycoprotein (gp41), which are the equivalents of HA2 and HA1. gp41 has its C-terminus inserted into the viral membrane and anchors a trimer of gp120 onto the viral membrane.

In the native state (with no CD4 bound), the fusion peptide is hidden within the hydrophobic pocket within the envelope protein. When gp120 binds to CD4 and co-receptors, it causes a conformational change in the trimer (outward rotation and displacement of each monomer of gp120) that leads to CD4 binding more 'perpendicularly'; this brings Env closer to the target membrane. It also frees the V3 loop and fusion peptide from an unexposed location. The V3 loop binds to a chemokine receptor (CXCR4 or CCR5; both expressed on lymphocytes and mononuclear cells) using a conserved site, and the fusion peptide is responsible for insertion. After insertion, energy is used to pull the two membranes together to form a hairpin.

The major difference between influenza virus and HIV is that HIV virus contains a reverse transcriptase which produce viral DNA from its RNA genome. This viral DNA is then inserted into the host chromosomal DNA. It is from the integrated DNA that new viral DNA is synthesized from. Furthermore, HIV virus does not rely on pH changes to function, and its release is triggered by cellular receptors.

Drugs against HIV focuses on inhibitors of three types: against fusion, against reverse transcription and against proteases. Reverse transcriptase inhibitors (RTIs) and protease inhibitors were the first available, but their uses have been limited due to drug-resistant mutations accumulated in the virus. Therefore, a third inhibitor (fusion inhibition) was developed.

Enfuvirtide (Acetyl-YTSLIHSLIEESQNQ QEKNEQELLELDKWASLWNWF-amide) is the first fusion inhibitor which binds to gp41. It has an amino acid sequence derived from residues 638–673 within the HR2 domain of the HIV-1HXB2 gp41, which is critical to the correct folding of gp41. This peptide is able to bind to gp41 and stop its folding into the correct conformation, thus blocking its function.

5-helix is a protein designed to bind to the C-peptide of gp41 and inhibits fusion. For it to function, the C-peptide must be exposed. It is found that native gp41 is does not bind 5-helix well, whereas CD4- or receptor-bound gp41 did bind 5-helix. This provides a possible drug which inhibits fusion, and also provide structural information about the pre-fusion intermediates, namely that the C-peptide is exposed.

Other inhibitors exists.

Dengue
Dengue virus causes dengue fever, an infectious tropical disease that causes fever, headache, muscle and joint pains and a characteristic skin rash. It can develop into a life-threatening dengue hemorrhagic fever, which results in bleeding, thrombocytopenia, vomitting, severe abdominal pain, and death. It is transmitted by several species of mosquito within the genus Aedes, principally A. aegypti. There is no vaccine. Envelope glycoprotein, E, a class II fusion protein on the surface of the dengue virus, can bind to a receptor which is endocytosed into the host cell. Within the endosome, the decreasing pH causes E to undergo conformational change from a dimer to a trimer which ultimately leads to its fusion with the host membrane.

Membrane fusion can be split into several stages: recognition (to determine specificity), energy utilization (fusion is often energetically unfavourable), fusion pore formation, and regulation.

The immature virions are produced in the ER (pH 7.2); it contains an auxiliary precursor membrane protein, prM, and an envelop protein, E. prM prevents virus from maturing. In the ER, the prM subunits aggregate together. As the virion moves through the secretory pathway, the decreasing pH allows prM to be cleaved off by furin in the trans-Golgi network. This allows the E proteins to form a homodimer and flattens itself so it makes a better contact with the viral membrane. The prM remains associated at the low pH, only when the virion is released into a neutral extracellular environment does prM dissociates.

At normal pH, the E protein fusion peptide interact with the C-terminus of another E protein to form a dimer. At low pH, the E protein changes conformation so that C-terminus moves in between the N-terminus and the fusion peptide by turning on its hinges. In this conformation, the C- and N-termini of three E proteins can aggregate together, 'releasing' the fusion peptide so the fusion loop can insert into the membrane. In this conformational change, there are no dramamtic refolding, simply a rotation of domains around hinges.

Intracellular membrane fusion
There are two types of membrane fusion - heterotypical (vesical-target; ER → Golgi) and homotypcial (target-target; Golgi → Golgi). The life cycle of a vesicle is a repeat of formation and exocytosis, fusion followed by endocytosis, and recycling.

NSP, p97 and their co-factors and SNARES

===SNARE === SNAREs (soluble N-ethylmaleimide-sensitive factor [NSF]-attachment protein receptors) are involved in the fusion of neurotransmitter-containing vesicles to the pre-synaptic terminal membrane of a nerve cell.

Structure
SNARE proteins are characterised by homologous 60 to 70-residue sequence called the SNARE motif, each motif consists of eight heptad repeats. The first, third and last residues of this repeat are hydrophobic. When the SNARE motif folds into an α-helix, it creates a hydrophobic face on one side of the α-helix, allowing it to interact with the hydrophobic faces of other SNARE proteins.

However, if the only the hydrophobic faces interact, then the bundled-coiled coils will be able to slide against each other and the structure will not be very strong, and will be unable to cause fusion. Instead, in the centre of the bundle (a.k.a. zero layer) is a hydrophilic region, consisting of one arginine and three glutamines, one on each of the four SNARE motifs. These residues hydrogen bond with each other on the inside of the bundle. Either side of this hydrophilic region the four motifs again form hydrogen bonds with its neighbouring bundle. These hydrogen bonds holds the helices at specific places and prevents the helices from sliding, and ensure the four helices are packed correctly.

General Mechanism
SNARES are present on both the fusing membranes. There are four types of SNARE motifs - R, Qa, Qb and Qc; to form a core complex, all four motifs must be present and assembled together to form a four-helical bundle. Each SNARE protein has 1-2 SNARE motifs, and so 2-4 SNARE proteins are required for each core complex. The core complex formation forces the fusing membranes close to each other, encouraging fusion. But it does not necessarily cause fusion; the SNARE complex might form and accumulate without fusing, only when a defect is caused or the temperature is raised will fusion be more likely.

v-SNARE was prepared into a liposome made up of a fluorescent lipid which is quenched at high concentrations; a t-SNARE liposome is made with non-fluorescent lipids. At 37°C, after incubation, the two liposomes fuse and the fluorescent lipid gets diluted and fluoresces; at 4°C, incubation of the two liposomes did not show fluorescence and thus the liposomes might have docked, but have not fused. Using the same apparatus, it is shown that SNARE proteins can fuse within miliseconds, but some can take minutes to hours.

Pre-synaptic fusion
There are three SNARE proteins involved in synaptic vesicular fusion: synaptobrevin (a.k.a. vesicle-associated membrane protein, VAMP) on synaptic vesicles provides the R motif, syntaxin 1 on the plasma membrane provides the Qa motif, and SNAP-25 (synaptosomal-associated protein of 25 kDa) on the presynaptic plasma membrane provides the Qb and Qc motif.

The three peptides come together to form the core complex. Complexins bind to the core complex and promotes the action of synaptotagmin 1. Synaptotagmins 1 associates with the complexin/SNARE core complex in the absence of Ca2+. Upon binding of Ca2+, the C2 domains associates with the phospholipid membrane and destabilizes the intermediate and enable pore formation. This is the mechanism in which calcium influx triggers vesicle fusion.

Regulation
Synaptophysin is a single-pass membrane protein, and the most abundant vesicle membrane protein; it binds to synaptobrevin (also a single-pass membrane protein; synaptobrevin can also bind VAP33). When bound, it prevents synaptobrevin from associating with other SNARE proteins, and thus is thought to negatively regulate membrane fusion and thus neurotransmitter release. This is supported by evidence that chronic blockade of glutamate receptors caused an increase in neurotransmitter release but a decrease in the synaptobrevin/synaptophysin complex.

Syntaxins have a SNARE motif (H3) domain which binds to synaptobrevin and SNAP-25 in the complex. Syntaxin 1 is initially bound to nSec1 (neuronal homolog of the yeast Sec1 protein, a.k.a. Munc18), which stabilizes a closed conformation of syntaxin 1, this decreases its binding affinity to other SNAREs, thus slowing SNARE complex formation. Thus nSec1 can be viewed as a regulatory protein for syntaxin 1, much like synaptophysin is to synaptobrevin.

Dissociation
Once the SNARE complex is formed, it is stable for up to 90°C. And thus energy must be used to dissociate them. note that sometimes the SNARE complex is not dissociated so the vesicle remains docked; this allow the vesicle to be loaded again and fuse with the membrane.

Disassembly after heterotypical fusion
α-SNAP and N-ethylmaleimide-sensitive factor (NSF, a homohexamer) bind to the SNARE complex for dissociation by ATP hydrolysis. Upon ATP hydrolysis, the α-SNAP move to the side of the NSF 20 particle and this pulls the SNARE proteins apart. Some have suggested that the polar residues of the zero layer enable NSF-mediated SNARE complex disassembly, although mutation of a glutamine to an arginine does not affect the rate of NSF-mediated disassembly.

Disassembly after homotypical fusion
p97/VCP/CDC48p is a homologous (to NSF) Mg2+-dependent AAA ATPase (ATPase Associated with various cellular activities) which are responsible for core disassembly. While NSF is involved in dissassembly in heterotypical fusion events such as vesicle trafficking, neurotrasmitter release and Golgi reassembly (after break down during mitosis), p97 is involved in many unrelated homotypical cellular events (membrane trafficking, protein degradation and nuclear envelop formation.

It can be shown using a cell-free system that p97 action requires an adaptor protein, p47. Using a pull-down assay using GST-syntaxin V, it is found that p97 interacts with syntaxin only through both binding to the adaptor p47.