Timing and Reset Mechanism of GTP Hydrolysis-Driven Conformational Changes of Atlastin

Abstract

Summary The endoplasmic reticulum (ER) forms a branched, dynamic membrane tubule network that is vital for cellular function. Branching arises from membrane fusion facilitated by the GTPase atlastin (ATL). Many metazoan genomes encode for three ATL isoforms that appear to fulfill partially redundant function despite differences in their intrinsic GTPase activity and localization within the ER; however, the underlying mechanistic differences between the isoforms are poorly understood. Here, we identify discrete temporal steps in the catalytic cycle for the two most dissimilar isoforms, ATL1 and ATL3, revealing an overall conserved progression of molecular events from nucleotide binding and hydrolysis, to ATL dimerization and phosphate release. A crystal structure of ATL3 suggests a mechanism for the displacement of the catalytic Mg2+ ion following GTP hydrolysis. Together, the data extend the mechanistic framework for how GTP hydrolysis drives conformational changes in ATL and how the cycle is reset for subsequent rounds of catalysis.

Two forms of Opa1 cooperate to complete fusion of the mitochondrial inner-membrane

“…In contrast, with cardiolipin-containing liposomes and bilayers, homotypic l-Opa1:l-Opa1 tethering is enhanced by GTP. Non-hydrolyzable analogues (GMPPCP) or GDP disrupt tethering (Figure 3B), and a hydrolysis-dead mutant (G300E) l-Opa1 shows no tethering (data not shown), supporting a role for the hydrolysis transition-state in tethering, as observed for atlastin (23,24). Bulk liposome light scattering measurements (NTA Nanosight) show l-Opa1-mediated liposome clustering requires the presence of GTP (Supplemental Figure 3).…”

Sensory neuropathy-causing mutations in ATL3 cause aberrant ER membrane tethering

“…ATL1 is predominantly expressed in the brain, whereas ATL2 and ATL3 show a more ubiquitous expression pattern (Rismanchi et al, 2008). In addition, the three ATL proteins appear to have different fusogenic capacities, with ATL1 being the stronger fusogen and ATL3 a much weaker one O'Donnell et al, 2017). In line with this, selective loss of ATL3 has very little effect on ER morphogenesis, whereas ATL1 alone is sufficient to rescue the loss of both ATL2 and ATL3 in COS--7 cells .…”

“…S1), which were hypothesized to correspond to the prefusion and postfusion states of ATL1, respectively (Bian et al, 2011;Byrnes and Sondermann, 2011). In the extended conformation, for which a crystal structure of ATL3 was also recently solved (O'Donnell et al, 2017), the alpha--helical 3HB middle domain packs tightly against the G domain, enforced mainly by hydrophobic interactions (Fig. 4A, S1).…”

“…4A and Fig. S1) (Bian et al, 2011;Byrnes and Sondermann, 2011;Byrnes et al, 2013;Hu and Rapoport, 2016;Liu et al, 2015;O'Donnell et al, 2017;Saini et al, 2014). In vitro, these stable dimers form in the presence of (Table S1) and therefore the possible oligomerization state of a given protein.…”

“…The N--terminal domain consists of a GTPase domain (G domain) connected via a flexible linker to a three--helix bundle (3HB) middle domain (Bian et al, 2011;Byrnes and Sondermann, 2011;Liu et al, 2012). Homotypic ER membrane fusion is mediated by the GTP-dependent homodimerization of ATLs on opposing membranes, promoting tight membrane tethering and eventually membrane fusion (Byrnes and Sondermann, 2011;Byrnes et al, 2013;Hu and Rapoport, 2016;Liu et al, 2015;O'Donnell et al, 2017;Saini et al, 2014)(see Fig. S1).…”