N-Ethylmaleimide-sensitive Fusion Protein
Created by Ben Iredell
N-Ethylmaleimide-sensitive fusion protein (NSF), found in eukaryotic cytosol, is a hexamer involved in the intracellular fusion of vesicle membranes.1 This is an important process for the cell because many cells secrete proteins and other products into intercellular space through a process called exocytosis. The Golgi apparatus packs the products into secretory vesicles, or lipid membrane-bound portions of the cytosol. As the vesicles reach the cell membrane, they fuse with the cell membrane, emptying the contents of the vesicle into extracellular space. NSF is involved in both constitutive secretion (continuous) and regulatory secretion (controlled by cell signals).1
Before membrane fusion, the NSF molecule is bound to a soluble NSF attachment protein (SNAP).1 The SNAP molecule is also bound to an integral membrane called a SNARE (SNAP receptor).1 When NSF hydrolyzes ATP to ADP, the NSF-SNAP-SNARE complex breaks apart, causing the NSF and SNAP molecules to be released into the cytosol.1 Consequently, the SNARE protein undergoes a conformational change, which acts to fuse the membranes together for exocytosis.1 This makes NSF a molecular chaperone because it transfers the ATP hydrolysis to a conformational change in another protein.1
NSF is a homohexamer, meaning the protein's natural form in organisms consists of six identical copies of the protein (called subunits or protomers) that are bound together by weak forces to fulfill the protein's function.3 Each complex consists of three domains: the N domain (on the N terminal), the D1 domain, and the
D2 domain.3 The N domain is responsible for binding to a soluble NSF attachment protein (SNAP). The D1 domain binds ATP to allow the N domain to bind the SNAP protein. Finally, the D2 domain is responsible for hexamerization.7
The D2 subunit (
secondary structure) can be further categorized into
two subdomains: a nucleotide-binding subdomain and a helical subdomain.6 The nucleotide-binding subdomain takes on the shape of a
triangular wedge, which is common in nucleotide-binding proteins.1 This subdomain consists of a central 5-stranded parallel β-sheet flanked by α helices.1 The
helical subdomain, however, consists of four α helices.1
The D2 domain contains two Walker motifs: the A motif and the B motif.1 The A motif consists of a phosphate-binding
P-loop between β1 and α2, which forms strategically-positioned "knob-like" projections, coupling ATP binding to conformation changes.4 The B motif consists of hydrophobic residues and a conserved aspartate.1
NSF is in the family of AAA proteins (ATPases associated with various cellular activites), which are ATPases that are Mg2+ dependent.1 AAA proteins have diverse functions, from proteolysis to membrane trafficking.1 Each AAA protein contains 1-2 copies of a 230-250-residue cassette that contains Walker motifs A and B.1
Although it has been theorized that NSF undergoes conformational changes, no complete structure has been isolated. It is still unclear whether D1, D2, or both are responsible for conformational changes.4 However, it is known that the conformation change involves movement of the P-loop, and that the change is sensitive to the state of the bound nucleotide.1
The nucleotide bound to the D2 domain is the ATP analog
5'-adenyl-imido-triphosphate (AMPPNP).1 NSF differs from other AAA proteins in how the protein
binds to the nucleotide.4 First, the adenine group is syn, instead of the usual anti.1 Ile-515 and Ile-715
pack on either side of the adenine ring.1 Ile-516 and Leu-560
contract the ring.1 Ala-559 and the aliphatic portion of Lys-716
interact with the ribose group through Van der Waals interactions.1
N6 and N7 of the adenine form hydrogen bonds with the backbone at Ile-516.1 N1 forms a hydrogen bond with a
water molecule positioned by Trp-518 and Ser-555.1
NSF forms bonds with the
phosphate group in the same manner as other AAA proteins.1 The P-loop forms several hydrogen bonds with the phosphate groups. Lys-557 forms hydrogen bonds with the β and γ-phosphate oxygen atoms, and Thr-558 coordinates the magnesium cation.1 The charge of the magnesium ion is countered by the phosphate groups.6 The ion serves to stabilize the transition state and anionic products of ATP hydrolysis.6
There is also a highly conserved
"DEAD" box near the nucleotide (residues 611-614).1 In domain D1, the sequence is DEID, but in domain D2, the sequence is DDIE.1 D2 also has a much higher affinity for binding ATP and ADP. The Kd values are 40 nM and 2 µM for binding ATP and ADP respectively, while the values are 40 µM and 140 µM for D1.1 This is caused both by the DEAD sequence, and because D2 has
Ala-562 packed against Ile-511 where most AAA proteins have a basic residue.1
The D2 binding of a nucleotide is thought to enable hexamerization, in part because NSF is only active as a hexamer.6 Not only does the D2 domain of each protomer form a
hexamer, but the D1 domain also forms a hexamer, stacked on top of the D2 hexamer.1 The D2 hexamer forms when the
loop between β3 and α4 participates in interactions with both adjacent protomers.1 This forms a gap in the middle of the six-protomer structure.1
Protomers have extensive and specific interfaces. Each protomer burys 2,500Å2, or 20% of the protomer's water-soluble area.1 Among these interactions are hydrophobic interactions, hydrogen bonds, salt bridges, and water-mediated hydrogen bonds.1
Each protomer also has a charge distribution. The N-terminus is basic, and the C-terminus is acidic.1 Therefore, the D1 domain, which is similar in structure to the D2 domain, can easily fit on the top of the D1 domain with the help of favorable ionic interactions.1 However, electron micrographs show that the two domains aren't in direct contact.1 This further supports the theory that conformational changes occur. If changes didn't occur, the interactions would be hydrophobic and therefore more permanent.1
The crystallization of the D2 Domain presented as 1D2N is from the ovary of Cricetulus griseus, or the Chinese Hamster. There are currently no drugs that target NSF.
Metalloprotease FtsH (2CEA) was concluded to have a similar tertiary structure based on Dali results (Z score of 16.3).5 FtsH is an integral membrane protein that functions to breakdown proteins.5 Being an AAA protein, FtsH binds ATP, but in addition, FtsH is an oligomer composed of multiple layers, like NSF.5 Therefore, both FtsH and NSF have residues that enable the protein to form stacked oligomers. The similarity can be seen from a
superimposition of the two proteins.