Membrane fusion proteins (not to be confused with chimeric or fusion proteins) are proteins that cause fusion of biological membranes. Membrane fusion is critical for many biological processes, especially in eukaryotic development and viral entry. Fusion proteins can originate from genes encoded by infectious enveloped viruses, ancient retroviruses integrated into the host genome,[1] or solely by the host genome.[2]Post-transcriptional modifications made to the fusion proteins by the host, namely addition and modification of glycans and acetyl groups, can drastically affect fusogenicity (the ability to fuse).[3]
The SNARE family include bona fide eukaryotic fusion proteins. They are only found in eukaryotes and their closest archaeal relatives like Heimdallarchaeota.[4]
ERV3 is not fusogenic in humans. Still plays a role in helping the placenta evade immune response.[5]
HAP2 family
HAP2 is a fusexin (similar to viral class II) found in diverse eukaryotes including Toxoplasma, vascular plants, and fruit flies. This protein is essential for gamete fusion in these organisms.[6] Its origin is unclear, as the broader grouping of fusexins could be older than the viral class II with the discovery of archaeal homologs.[7]
Pathogenic viral fusion
Enveloped viruses readily overcome the thermodynamic barrier of merging two plasma membranes by storing kinetic energy in fusion (F) proteins. F proteins can be independently expressed on host cell surfaces which can either (1) drive the infected cell to fuse with neighboring cells, forming a syncytium, or (2) be incorporated into a budding virion from the infected cell which leads to the full emancipation of plasma membrane from the host cell. Some F components solely drive fusion while a subset of F proteins can interact with host factors. There are four groups of fusion proteins categorized by their structure and mechanism of fusion.[8]
Despite their very different structure and presumably different origins, classes I, II, and III all work by forming a trimer of hairpins.[9]
Class I
Class I fusion proteins resemble influenzavirushemagglutinin in their structure. Post-fusion, the active site has a trimer of α-helical coiled-coils. The binding domain is rich in α-helices and hydrophobic fusion peptides located near the N-terminus (some examples show internal fusion peptides, however).[9] Fusion conformation change can often be controlled by pH.[10]
Class II
Class II proteins are dominant in β-sheets and the catalytic sites are localized in the core region. The peptide regions required to drive fusion are formed from the turns between the β-sheets.x[9] They usually start as dimers, becoming a trimer as fusion happens.[11][10]
Class III
Class III fusion proteins are distinct from I and II. They typically consist of 5 structural domains, where domain 1 and 2 localized to the C-terminal end often contain more β-sheets and domains 2-5 closer to the N-terminal side are richer in α-helices. In the pre-fusion state, the later domains nest and protect domain 1 (i.e. domain 1 is protected by domain 2, which is nested in domain 3, which is protected by domain 4). Domain 1 contains the catalytic site for membrane fusion.[10][9]
Others
A number of fusion proteins belong to none of the three main classes.[9]
Poxviruses employ a multiprotein system of 11 different genes and their relatives in the broader group of Nucleocytoviricota appear to do likewise.[12] The structure of the fusion complex is not yet resolved. Scientists have produced some information on what each of the components bind to, but still not enough to produce a full picture.[13][14]
Hepadnaviridae, which includes the Hep B virus, uses different forms of the surface antigen (HBsAg - S, M and L) to fuse.[9] It was found in 2021 that it has a fusion peptide in preS1, which is found in the L form.[15]
FAST
Fusion-associated small transmembrane proteins (FAST) are the smallest type of fusion protein. They are found in reoviruses, which are non-enveloped viruses and are specialized for cell-cell rather than virus-cell fusion, forming syncytia. They are the only known membrane fusion proteins found in non-enveloped viruses.[16] They exploit the cell-cell adhesion machinery to achieve initial attachment. They might encourage fusion by inducing membrane curvature using a variety of hydrophobic motifs and modified residues.[17]
The fusexin family consists of eukaryotic HAP2/GCS1, eukaryotic EFF-1, viral "class II", and haloarchaeal Fsx1. They all share a common fold and fuse membranes.[7] In an unrooted phylogenetic tree from 2021, HAP2/GCS1 and EFF-1/AFF-1 occupy two ends of the tree, the middle being occupied by viral sequences; this suggests that they may have been acquired separately.[11] The latest structure-based unrooted phylogenetic tree of Brukman et al. (2022), which takes into account the newly-discovered archaeal sequences, shows that Fsx1 groups with HAP2/GCS1, and that they are separated from EFF-1 by a number of viral sequences. Based on where the root is placed, a number of different hypotheses regarding the history of these families – their horizontal transfer and vertical inheritance – can be generated.[7] Older comparisons excluding archaeal sequences would strongly favor an interpretation where HAP2/GCS1 is acquired from a virus,[6] but the grouping of Fsx1 with HAP2/GCS1 has allowed the possibility of a much more ancient source.[7]
^ abBaquero, Eduard; Albertini, Aurélie AV; Gaudin, Yves (August 2015). "Recent mechanistic and structural insights on class III viral fusion glycoproteins". Current Opinion in Structural Biology. 33: 52–60. doi:10.1016/j.sbi.2015.07.011. PMID26277251.
