MARCKS Effector Domain, a reversible lipid ligand, illuminates late stages of membrane fusion

After their membrane assembly, Sec18 (NSF) and Sec17 (SNAP) promote membrane fusion

The energy that drives membrane fusion can come from either complete SNARE zippering, from Sec17 and Sec18, or both. Sec17 and Sec18 initially form a complex which binds membranes. Sec17, Sec18, and the apolarity of a loop on the N-domain of Sec17 are required for their interdependent membrane association. To determine whether Sec18 and the Sec17 loop apolarity are still required for fusion after their membrane arrival, a hydrophobic transmembrane (TM) anchor was affixed to the N-terminus of Sec17, forming TM-Sec17. Fusion without energy from complete SNARE zippering requires Sec18 as well as either Sec17 or TM-Sec17. Even without the need for membrane targeting, the TM-Sec17 apolar loop strongly stimulates Sec17/18-driven fusion. Thus, Sec18 and the Sec17 apolar loop are first required for membrane targeting, and once bound, drive rapid fusion. Each of these variables-the absence or presence of Sec17, its N-loop apolarity, addition or omission of Sec18, and unimpeded or diminished energy from SNARE zippering-has almost no effect on the amount of trans-SNARE complex, but instead regulates the capacity of docked membranes to fuse.

Membrane fusion reactions limited by defective SNARE zippering or stiff lipid fatty acyl composition have distinct requirements for Sec17, Sec18, and adenine nucleotide

Intracellular membrane fusion is catalyzed by SNAREs, Rab GTPases, SM proteins, tethers, Sec18/NSF and Sec17/SNAP. Membrane fusion has been reconstituted with purified vacuolar proteins and lipids to address 3 salient questions: whether ATP hydrolysis by Sec18 affects its promotion of fusion, whether fusion promotion by Sec17 and Sec18 is only seen with mutant SNAREs or can also be seen with wild-type SNAREs, and whether Sec17 and Sec18 only promote fusion when they work together or whether they can each work separately. Fusion is driven by two engines, completion of SNARE zippering (which does not need Sec17/Sec18) and Sec17/Sec18-mediated fusion (needing SNAREs but not the energy from their complete zippering). Sec17 is required to rescue fusion that is blocked by incomplete zippering, though optimal rescue also needs the ATPase Sec18. ATP is an essential Sec18 ligand, but at limiting Sec17 levels Sec18 ATP hydrolysis also drives release of Sec17 without concomitant trans-SNARE complex disassembly. At higher (physiological) Sec17 levels, or without ATP hydrolysis, fusion prevails over Sec17 release. Stiff 16:0, 18:1 fatty acyl chain lipids provide an alternative route to suppressing fusion, with entirely wild-type SNAREs and without impediment to zippering. In this case, Sec17 and Sec18 restore comparable fusion with either ATP or a nonhydrolyzable analog. Fusion blocked by impaired zippering can be restored by concentrated Sec17 alone (but not by Sec18), while fusion inhibited by stiff fatty acyl chains is partially restored by Sec18 alone (but not by Sec17). With distinct fusion impediments, Sec18 and Sec17 have both shared roles and independent roles in promoting fusion.

Sec18 binds the tethering/SM complex HOPS to engage the Qc-SNARE for membrane fusion

Membrane fusion is regulated by Rab GTPases, their tethering effectors such as HOPS, SNARE proteins on each fusion partner, SM proteins to catalyze SNARE assembly, Sec17 (SNAP), and Sec18 (NSF). Though concentrated HOPS can support fusion without Sec18, we now report that fusion falls off sharply at lower HOPS levels, where direct Sec18 binding to HOPS restores fusion. This Sec18-dependent fusion needs adenine nucleotide but neither ATP hydrolysis nor Sec17. Sec18 enhances HOPS recognition of the Qc-SNARE. With high levels of HOPS, Qc has a Km for fusion of a few nM. Either lower HOPS levels, or substitution of a synthetic tether for HOPS, strikingly increases the Km for Qc to several hundred nM. With dilute HOPS, Sec18 returns the Km for Qc to low nM. In contrast, HOPS concentration and Sec18 have no effect on Qb-SNARE recognition. Just as Qc is required for fusion but not for the initial assembly of SNAREs in trans, impaired Qc recognition by limiting HOPS without Sec18 still allows substantial trans-SNARE assembly. Thus, in addition to the known Sec18 functions of disassembling SNARE complexes, oligomerizing Sec17 for membrane association, and allowing Sec17 to drive fusion without complete SNARE zippering, we report a fourth Sec18 function, the Sec17-independent binding of Sec18 to HOPS to enhance functional Qc-SNARE engagement.