GTPase is a key mediator of cell-autonomous innate immunityHis-tagged full-length human MxA (Fig. 1a) was recombinantly expressed in bacteria and purified to homogeneity (Methods, Supp. Fig. 1). In crystallization trials, small needle-shaped protein crystals were obtained which represented proteolytic cleavage products of the MD and GED (Supp. Fig. 2). We solved the phase problem by a single anomalous dispersion protocol and could build and refine a model containing two molecules in the asymmetric unit (Methods, Supp. Table 1 and 2). Each monomer spans nearly the complete MD and the amino(N-)-terminal part of the GED (amino acids 366-633) which together fold into an elongated anti-parallel fourhelical bundle where the MD contributes three helices and the GED one (Fig. 1b, Supp. Fig. 3). This segment corresponds to the stalk region of dynamin 7 , and we refer to it as stalk of MxA. The first visible amino acid, Glu366, is 15 amino acids downstream of the last visible residue of the corresponding G-domain structure in rat dynamin (Supp. Fig. 3) 8 . It marks the start of helix α1 in the MxA stalk which is divided in α1 N and α1 C by a 10 amino acid long loop, L1, introducing a 30° kink. A putative loop L2 (amino acids 438-447) opposite of the deduced position of the G-domain is not visible in our structure. L2 was previously demonstrated to be the target of a functionally neutralising monoclonal antibody 9,10 . Helix α2 runs anti-parallel to α1 back to the G-domain. It ends in a short loop L3 and is followed by helix α3 that extends in parallel to α1. The 40 amino acid long loop L4 (residues 532-572) is at the equivalent sequence position as the PH domain of dynamin (Fig. 1a, Supp. Fig. 3) and is absent in our model. L4 is predicted to be unstructured and was previously shown to be proteinase K sensitive 11 .At the C-terminus, the GED supplies 44 residues to helix α4 which proceeds in parallel to helix α2 back to the G-domain. It is followed by a short helix α5 which directs the polypeptide chain towards the N-terminus of the MD. The carboxy(C-)-terminal 30 highly conserved residues of the GED known to be involved in antiviral specificity 12 are missing in our model. In dynamin, the corresponding residues were shown to directly interact with the G-domain 13 . The stalk of MxA is divergent from the corresponding structures of other dynamin superfamily members, such as GBP1 14 , EHD2 15 and BDLP 16 although some features are shared (Supp. Fig. 4). 4 In the crystal lattice, each MxA stalk is assembled in a criss-cross pattern resulting in a linear oligomer, where each stalk contributes three distinct interfaces (Fig. 1c). Such an arrangement of the stalks is plausible for the Mx oligomer since all G-domains would be located at one side of the oligomer whereas the putative substrate-binding site in L2 and L4 would be located at the opposite side (Fig. 1b, c).The hydrophobic interface-1 covering 1300 Å 2 is conserved among Mx proteins and dynamins and has a two-fold symmetry between the associating monomers (Fig....

Dynamin is a mechanochemical GTPase that oligomerizes around the neck of clathrin-coated pits and catalyses vesicle scission in a GTP-hydrolysis-dependent manner. The molecular details of oligomerization and the mechanism of the mechanochemical coupling are currently unknown. Here we present the crystal structure of human dynamin 1 in the nucleotide-free state with a four-domain architecture comprising the GTPase domain, the bundle signalling element, the stalk and the pleckstrin homology domain. Dynamin 1 oligomerized in the crystals via the stalks, which assemble in a criss-cross fashion. The stalks further interact via conserved surfaces with the pleckstrin homology domain and the bundle signalling element of the neighbouring dynamin molecule. This intricate domain interaction rationalizes a number of disease-related mutations in dynamin 2 and suggests a structural model for the mechanochemical coupling that reconciles previous models of dynamin function.

Mitochondria are double-membrane organelles with varying shapes influenced by metabolic conditions, developmental stage, and environmental stimuli1–4. Their dynamic morphology is realized through regulated and balanced fusion and fission processes5, 6. Fusion is crucial for the health and physiological functions of mitochondria, including complementation of damaged mitochondrial DNAs and maintenance of membrane potential6–8. Mitofusins (Mfns) are dynamin-related GTPases essential for mitochondrial fusion9, 10. They are embedded in the mitochondrial outer membrane and thought to fuse adjacent mitochondria via concerted oligomerization and GTP hydrolysis11–13. However, the molecular mechanisms behind this process remains elusive. Here we present crystal structures of engineered human Mfn1 containing the GTPase domain and a helical domain in different stages of GTP hydrolysis. The helical domain is composed of elements from widely dispersed sequence regions of Mfn1 and resembles the Neck of the bacterial dynamin-like protein. The structures reveal unique features of its catalytic machinery and explain how GTP binding induces conformational changes to promote G domain dimerization in the transition state. Disruption of G domain dimerization abolishes the fusogenic activity of Mfn1. Moreover, a conserved aspartate trigger was found in Mfn1 to affect mitochondrial elongation, likely through a GTP-loading-dependent domain rearrangement. Based on these results, we propose a mechanistic model for Mfn1-mediated mitochondrial tethering. Our study provides important insights in the molecular basis of mitochondrial fusion and mitofusin-related human neuromuscular disorders14.