Atlastins (ATLs) are a class of endoplasmic reticulum (ER) GTPases. Invertebrates have a single ATL, while vertebrates possess three ATL proteins (ATL1-3) that are differentially expressed. [1] ATL1 is the predominant paralog of the central nervous system, whereas ATL2 and ATL3 are mainly expressed in tissues outside of the CNS. [1] Loss of all ATLs in mammalian cells dramatically impacts ER structure, including a reduction in tubule three-way junctions. [2]
ATLs maintain the ER tubular network via homotypic fusion. ATLs have a conserved domain structure consisting of a globular G domain, a three-helix bundle, [3] two transmembrane domains, and an amphipathic helix. [4] The ATL fusion cycle consists of two ATL monomers in opposing membranes binding GTP, which induces trans G domain dimerization and a crossing over of the three-helix bundle. [3] [5] [6] Crossover and subsequent insertion of the amphipathic helix into the lipid bilayer triggers lipids to mix for fusion. [4] [7] Lastly, GTP is hydrolyzed driving the dimer to disassembly and resetting the fusion machinery. [8] [9]
While most of the human ATL protein structure is conserved between paralogs, [10] the proteins have non-conserved N- and C-termini with the C-termini of ATL1 and ATL2 being autoinhibitory. [11] ATL1 has been shown to interact with a range of proteins including spastin [1] and REEP1, [12] with spastin enhancing ATL1 fusion activity in vitro. [13] ATL1 and ATL2 have also been observed as interacting with ER protein TMCC3, [14] and ATL3 with nonstructural viral proteins, [15] however it is not currently known how these interactions modulate protein function.
Mutations in ATLs are linked to human disease. Mutations in ATL3 are associated with hereditary sensory neuropathy (HSN), [16] and mutations in ATL1 are linked to HSN [17] and hereditary spastic paraplegia (HSP). [12] Research has identified a number of mutations that correspond to the disease phenotype, including the ATL3 Y192C [18] disease mutation that is equivalent to the Y196C mutation in ATL1. [6] Work to identify disease mutants remains ongoing, with a novel nonsense ATL3 mutation being identified in early 2023. [19] ATL3 HSN mutations affect the protein's fusion cycle by causing aberrant tethering. [20] Similarly, an ATL1 HSP mutation was shown to increase tethering but not impact GTPase activity. [21]