In the 1980s, Rothman’s laboratory identified a protein that could revive Golgi transport following the inactivation of the Golgi membrane by N-ethylmaleimide (NEM). Owing to its sensitivity to NEM, this protein was designated as the N-ethylmaleimide-sensitive factor, or NSF (Block et al., 1988). NSF possesses ATPase activity and the ability to dissociate the SNARE complex (Tagaya et al., 1993), a capability essential for facilitating membrane fusion events (May et al., 1999; Yu et al., 1999). NSF, which shares homology with the Sec18 protein in Saccharomyces cerevisiae, is vital for key biological processes such as membrane fusion, intracellular transport, and vesicle trafficking (Wilson et al., 1989). This homology highlights the conserved importance of these proteins across different species. Previous studies have successfully identified a conditional mutant gene of NSF that may lead to defects in membrane transport under restrictive temperature conditions (May et al., 1999; Yu et al., 1999).

N-ethylmaleimide-sensitive factor is a member of the AAA+ (ATPases Associated with diverse cellular Activities) superfamily. AAA+ ATPases are characterized by a conserved ATPase domain that includes several key regions: the Walker A and B motifs for ATP binding and hydrolysis; the Sensor 1 and Sensor 2 regions for nucleotide sensing; the Second Region of Homology (SRH) essential for oligomerization, and the Nucleotide Binding Domain (NBD), where ATP hydrolysis occurs. The flexible linker within the AAA+ domain allows for the conformational changes required for the enzyme’s mechanical work (Khan et al., 2022). NSF contains a core structure with an ATP binding site region composed of 185 conserved amino acids (Kunau et al., 1993), along with two Type II ATPase loops (Sollner et al., 1993). NSF exists as a homohexamer with a central hole, resulting in a total molecular weight of 500 kDa. Each subunit comprises three domains: the amino-terminal N domain (NSF-N, 1–205), which primarily interacts with α-soluble NSF attachment protein (SNAP) and SNARE complexes, a low-affinity ATP-integrated type Walker ATP structure domain (NSF-D1, 206–477), and a second Walker type ATP binding domain (NSF-D2, 478–744) with high ATP affinity (Matveeva et al., 1997; Tagaya et al., 1993; Wilson et al., 1989; Yu et al., 1998). The coordinated activity of these domains allows NSF to execute its cellular functions (Figure 1).

NSF AAA+ is composed of different structural domains with distinct features, each serving different functions. NSF’s D1 domain, part of its AAA+ structure, demonstrates high ATPase activity that is crucial for its overall function (Whiteheart et al., 1994). The 20S supercomplex, a structure formed by NSF, SNAPs, and SNAREs, relies on the D1 domain to facilitate its dissociation (Liu et al., 2013; Nagiec et al., 1995). Upon specific stimulation of SNAP-SNARE, the D1 domain demonstrates ATPase activity, facilitating the dissociation and release of SNARE complexes, thereby playing a crucial role in the membrane fusion process (Matveeva and Whiteheart, 1998; Nagiec et al., 1995). The D1 domain exhibits a relatively weaker ATP-binding activity compared to the D2 domain, as indicated by their dissociation constants (KD) for ATP, which are approximately 15–20 μmol for the D1 domain and 30–40 nmol for the D2 domain (Matveeva et al., 1997). Mutations in the D1 domain, including K266A and E329Q, lead to a substantial reduction in ATPase activity—approximately 70–80% lower than that of the wild type—preventing normal NSF function without SNAP-SNARE complex stimulation (Whiteheart et al., 1994). Additionally, specific mutations, such as R385A, R388A, and D359K, impair NSF’s capacity to assemble into the 20S supercomplex, which is crucial for its function (Nagiec et al., 1995). Arginine fingers within the D1 domain of NSF, specifically Arg385 and Arg388, are essential for its ATPase activity and for the dissociation of the SNAP-SNARE complexes (Zhao et al., 2010).

The D2 domain is widely recognized as the prototype of the AAA+ ATPase domain (Lenzen et al., 1998). It exhibits the highest level of sequence conservation and shares a similar high-level structure with other domains. The D2 domain is composed of two distinct subdomains, known as the α/β subdomain and the α subdomain, which together are characteristic features of all AAA+ ATPase domains (Lenzen et al., 1998; May et al., 1999; Yu et al., 1998, 1999). The primary function of the D2 domain is to facilitate hexamer formation (Matveeva et al., 1997; Whiteheart et al., 2001). In the NSF hexamer, the K631 residue on one D2 subunit interacts with the γ-phosphate of ATP on neighboring D2 subunits within the ring, displacing water molecules that are typically involved in the hydrolysis of the γ-phosphate (King, 2000). The D2 domain comprises a nucleotide-binding subdomain and a C-terminal subdomain with a unique structure among nucleotide-binding proteins, these features are essential for ATP-dependent oligomerization (Yu et al., 1998). The absence of functional arginine fingers in the D2 domain leads to significantly slower ATP hydrolysis (Morgan et al., 1994).

The N-terminal domain of NSF is essential for its binding to the SNAP-SNARE complex, a vital step in cell membrane fusion (Nagiec et al., 1995; Whiteheart et al., 2001; Zhao et al., 2010). The N-terminal domain shares a nearly identical structure with the yeast homolog Sec18 (Babor and Fass, 1999) and the homologous VAT protein from the archaebacterium Thermoplasma acidophilum (Coles et al., 1999), indicating structural conservation within the AAA+ ATPase families. The N-terminal domain comprises two subdomains, NA and NB (Lenzen et al., 1998). The domain interface, formed by a groove, may interact with the alpha C-terminal of the SNAP protein (Yu et al., 1999). Tyr83, situated on the loop that connects substructures NA and NB within the N-terminal domain, stabilizes the loop’s structure through hydrogen bonds with adjacent amino acids Gln90 and Lys87 (Zhao et al., 2007). The N domain can recognize various SNAP proteins, which in turn can bind to a range of SNARE complexes (Zhao et al., 2015). Notably, while the N domain only interacts with the C-terminal domain of α-SNAP, it does not directly engage with the SNARE complex itself, suggesting a specific selectivity in its molecular interactions (Wimmer et al., 2001). Moreover, NSF can dismantle “dual” SNARE complexes formed by connecting two SNARE complexes head-to-tail (Cipriano et al., 2013). Mutations in the N domain, such as R67, S73, and Q76, impact binding to the SNAP-SNARE complex (Matveeva et al., 2002). The N domain’s structure, while not dispensable for the initial assembly of the NSF hexamer, is crucial for endowing it with functionality; its removal results in the assembly of a non-functional hexameric complex (Whiteheart et al., 1994). The N domain, when isolated, fails to engage with the SNAP-SNARE complex, implying that a multitude of interaction points are indispensable for the effective binding between NSF and its target complex (Nagiec et al., 1995) (Table 1).

The NSF is an ATPase with a highly conserved AAA+ domain structure, existing as a homohexamer with each subunit comprising three domains: an N domain and two domains D1 and D2. These domains have differential ATP-binding capabilities. The N domain is responsible for binding to SNAP and SNARE complexes. The D1 domain has a low affinity for ATP binding and primarily functions in ATP hydrolysis for energy generation. Conversely, the D2 domain displays a high ATP-binding capacity and is crucial for maintaining the hexameric structure of NSF. NSF’s precise recognition and efficient hydrolysis of ATP are instrumental in facilitating the depolymerization of SNARE complexes, thereby exerting a crucial regulatory effect on intracellular membrane transport.

Research indicates that the activity of NSF in cells is regulated through post-translational modifications, which allow for temporal and spatial control of membrane fusion (Huynh et al., 2004). Various cell types utilize specific post-translational modifications of NSF to achieve localized regulation of membrane fusion. Moreover, NSF can undergo reversible inactivation through phosphorylation and S-nitrosylation (Morgan and Burgoyne, 2004).

N-ethylmaleimide-sensitive factor is a vital protein involved in intracellular membrane fusion processes in eukaryotes (Block et al., 1988; Malhotra et al., 1988). It interacts with SNARE proteins to facilitate the fusion of intracellular vesicles with the cell membrane, allowing for the transport of substances within and between cells. NSF is actively recruited to the membrane through its binding with adapter proteins SNAPs (Wilson et al., 1992). During membrane fusion, SNARE proteins assemble into α-helical trans-SNARE complexes that bring the membranes into proximity, thereby facilitating the merging of the lipid bilayers (Sudhof and Rizo, 2011). Recent findings demonstrate that an increase in SNARE complex assembly not only augments the rate of neurotransmitter release through single pores but also enables the release of larger cargos, suggesting a direct correlation between SNARE complex number and fusion pore efficiency (Bao et al., 2018). NSF, in synergy with proteins like Munc18-1, Munc13-1, and complexin-1, ensures the stability of these complexes, preventing their disassembly by NSF-αSNAP and thus maintaining vesicle priming (Prinslow et al., 2019). Moreover, NSF’s ATPase activity is pivotal for the disassembly of SNARE complexes post-fusion, allowing for their recycling and the continuation of vesicular transport (Littleton et al., 2001). The recent study reveals that NSF, in conjunction with syt1, drives the formation of committed trans-SNARE complexes that form large, stable fusion pores. These pores, once opened, can only be closed through the action of the ATPase NSF, which mediates pore closure via a complex “stuttering” mechanism (Das et al., 2020). This discovery highlights the dynamic interplay between NSF’s role in fusion pore opening and its subsequent disassembly activity, which is critical for the regulated assembly and disassembly of fusion pores.

N-ethylmaleimide-sensitive factor also plays a crucial role in other intracellular fusion events, such as the transport of secretory vesicles (Sztul et al., 1993), platelet granule secretion (Huynh et al., 2004; Polgár and Reed, 1999), and the release of pro-inflammatory mediators in endothelial cells (Lowenstein and Tsuda, 2006). Research has explored strategies to target NSF, such as designing specific peptides and antibody interventions, to inhibit its ATPase activity and impact secretion in human platelets and endothelial cells (Matsushita et al., 2003). Additionally, NSF is involved in the release of pro-inflammatory and prothrombotic mediators in endothelial cells by functioning as a redox sensor. In endothelial cells, the activity of NSF is regulated by nitric oxide and hydrogen peroxide through S-nitrosylation and oxidation, respectively. This regulation plays a role in influencing the mechanisms of exocytosis (Lowenstein and Tsuda, 2006). In summary, NSF plays a crucial role in various intracellular membrane fusion processes, regulating synaptic transmission and other cellular fusion events.

The NSF protein plays a critical role in regulating neurotransmitter release. Within the cytoplasm, NSF facilitates the fusion between the postsynaptic membrane and the SV membrane by disassembling the SNARE complex through the process of ATP hydrolysis. This fusion enables the orderly release of neurotransmitters (Rizo, 2022; Zinsmaier and Bronk, 2001). Studies employing a temperature-sensitive mutation in Drosophila dNSF1 known as comatose (comt) have demonstrated that under temperature-sensitive conditions, comt significantly increases the number of docked vesicles at synaptic terminals. This finding suggests that the NSF protein is involved in the maturation process of docked vesicles, ensuring the sustained release capability of neurotransmitters (Kawasaki et al., 1998).

Furthermore, NSF exerts a pivotal influence on the dynamics of neurotransmitter release, meticulously tuning both the volumetric output and the velocity of release. Recent findings highlight the intricate relationship between SNARE complex assembly and fusion pore properties, showing that altering the number of SNARE complexes can modulate the release rate through individual fusion pores, potentially impacting NSF’s role in neurotransmitter release (Bao et al., 2018). Another study utilized a photosensitive NSF inhibitory peptide and demonstrated that the NSF protein regulates neurotransmitter release through two distinct reaction rate steps. The slower step involves the decomposition of cis-SNARE complexes and regulates the quantity of neurotransmitter release. At the same time, the rapid action within 0.22 s affects postsynaptic currents, modulating the dynamic speed of neurotransmitter release (Kuner et al., 2008).

Moreover, research indicates that the phosphorylation level of the NSF protein, regulated by PKC, directly influences its binding affinity to SNAP-SNARE complexes. Consequently, this phosphorylation inhibits NSF activity, leading to a reduction in neurotransmitter release at synaptic terminals. This discovery elucidates a direct link between specific phosphorylation events and the modulation of neurotransmission (Matveeva et al., 2001). Another molecule called neurexin (NRX), primarily localized at the presynaptic terminal, directly interacts with the NSF protein. By influencing the distribution of the NSF protein at the presynaptic terminal and its ability to disassemble the SNARE complex, NRX regulates NSF and consequently impacts neurotransmitter release (Li et al., 2015). Additionally, the NSF protein undergoes ubiquitin modification by the mitochondrial E3 ubiquitin ligase Ariadne-1 (Ari-1). In mouse models lacking Ari-1, there is a decrease in the frequency of spontaneous neurotransmitter release but an enhancement in neurotransmitter release induced by stimulation. This suggests that Ari-1, through its modification of NSF, can selectively regulate spontaneous and stimulus-induced neurotransmitter release pathways (Ramírez et al., 2021). These research findings reveal multiple vital regulatory mechanisms of the NSF protein in modulating neurotransmitter release processes.

More reports suggest that the NSF protein plays a critical role in synaptic reorganization and plasticity. The function of NSF is essential for high-dynamic responses before SV fusion with the presynaptic membrane to release neurotransmitters (Kuner et al., 2008; Li et al., 2019). Synaptic plasticity describes molecular and morphological changes associated with learning and memory, representing an experience-dependent adaptation process that involves the adjustment of synaptic strength (Hanley, 2018). NMDARs are ionotropic receptors involved in synaptic transmission and regulating neural signals. They modulate the stability and efficacy of synaptic connections and impact synaptic plasticity (Diering and Huganir, 2018). The interaction between NSF and NMDAR allows NSF to regulate NMDAR activity, thereby influencing synaptic plasticity, which is crucial for cognitive functions such as learning and memory (Morrell et al., 2005).

Additionally, NSF relies on binding to GluR2 subunits to achieve dynamic regulation of AMPAR, which is involved in the induction and expression of LTD and LTP, thereby modulating synaptic plasticity (Gardner et al., 2005; Lim and Isaac, 2005). The stable interaction between NSF and GluR2 is also involved in mediating LTP induced by protein kinase Mf (PKMf), as well as in sustaining contextual memory within the hippocampus (Migues et al., 2014). In the lateral amygdala (LA), the interaction between NSF and GluR2 is crucial for fear memory formation and consolidation (Joels and Lamprecht, 2010). Furthermore, the schizophrenia risk gene MIR137 influences the expression of the NSF gene, thereby affecting the SV recycling process. This disruption leads to abnormal SV distribution, reduced synaptic release efficiency, and impacts the formation of synaptic plasticity (Siegert et al., 2015). Moreover, NSF is involved in the Dysbindin-dependent synaptic plasticity regulatory pathway, playing a crucial role in synaptic plasticity (Gokhale et al., 2015).

N-ethylmaleimide-sensitive factor is involved in the processes of autophagosome formation and vesicle fusion. Alongside other conventional membrane transport factors such as SNAP and SNARE, NSF is responsible for the formation of autophagosomes. It facilitates the fusion of heterotypic membranes between autophagosomes and lysosomes by interacting with SNARE proteins. In yeast mutants lacking Sec18 (the yeast equivalent of NSF), autophagosomes can form but fail to fuse with vesicles, thereby preventing the completion of the autophagic process (Ishihara et al., 2001). While the dissociation of the SNARE complex is crucial for the fusion between autophagosomes and lysosomal membranes, it is relatively less critical for the formation of autophagosomes (Abada et al., 2017). However, whether the ATPase activity of NSF plays a role in the later stages of mammalian cell autophagy remains unclear (Mehrpour et al., 2010). Studies have indicated that under nutrient-deficient conditions, the activity of NSF protein is inhibited, resulting in weakened binding and fusion between autophagosomes and lysosomes, consequently decreasing the rate of autophagy (Fass et al., 2006). In Drosophila, mutations in dNSF1 lead to various neurodegenerative phenotypes, including motor dysfunction, a shortened lifespan, and progressive neurodegeneration, particularly the loss of dopaminergic neurons. The neurodegenerative phenotypes are linked to the maintenance of autophagy, which is compromised by defects in lysosomal protease transport. This connection arises because the NSF protein is instrumental in the disassembly and reorganization of the SNARE complex, processes that are essential for the continuity of autophagy. The overexpression of dNSF1 can alleviate α-synuclein-induced toxicity in a PD model of dopaminergic neurons, highlighting the neuroprotective role of dNSF1 in autophagy and degradation pathways (Babcock et al., 2015).

The C-terminal WD40 domain of the PD-associated gene LRRK2 can interact with various proteins involved in neuronal vesicle transport, including the NSF protein. Experimental findings suggest that overexpression of the WD40 domain of LRRK2 can affect the transport and distribution of cycling SV, potentially through interactions with proteins like NSF, resulting in the onset of PD (Piccoli et al., 2014). Further research has demonstrated that LRRK2 can phosphorylate the Thr645 residue of NSF, enhancing NSF’s ATPase activity and accelerating the disassembly of its catalytic SNARE complex. T645, situated within the D2 domain of NSF, is recognized as a critical site that facilitates NSF oligomerization upon ATP binding (Matveeva et al., 1997). LRRK2 phosphorylates NSF on T645, enhances its ATPase activity, and ultimately augments its ability to dissociate the SNARE complex. The common LRRK2 mutation G2019S enhances NSF kinase activity, potentially disrupting SV dynamics through abnormal phosphorylation of NSF and contributing to the manifestation of Parkinson’s symptoms (Belluzzi et al., 2016). Studies have discovered that the LRRK2 G2019S mutation can induce the aggregation of NSF protein in brain regions such as the striatum and hippocampus. The aggregation of NSF may exert toxicity on neurons, potentially contributing to the development of PD. Additionally, with the aging of individuals, the accumulation of NSF aggregates increases, while the proteasomal and autophagic functions in the brain gradually decline, suggesting that the combined impact may be a significant mechanism underlying neuronal cell death. In mouse models expressing the LRRK2 G2019S variant, the induction of autophagy has been found to clear NSF aggregates in the nervous system, leading to improvements in motor and cognitive impairments (Pischedda et al., 2021). While no clinically relevant drugs have been identified for LRRK2-associated PD, strategies that enhance autophagy offer promising therapeutic potential.

Traditional proteomic studies on human AD tissues and AD mouse models have shown that the NSF protein is present in neurofibrillary tangles, a neuropathological characteristic of AD. This suggests a potential involvement of NSF in the disease’s pathogenic mechanisms (Hashimoto et al., 2019). Research indicates that reduced levels of NSF hinder the fusion of autophagosomes with mature lysosomes, thus inhibiting autophagy. Consequently, this leads to the accumulation of pathological proteins such as β-amyloid and hyperphosphorylated tau protein, ultimately triggering AD-like pathology (Zhu et al., 2022). Proximity labeling techniques have confirmed a direct interaction between the tau protein and NSF, which leads to a dose-dependent reduction in NSF activity. NSF contributes to the stability of AMPAR through its ATPase activity, promoting the surface expression of AMPAR. Tau protein, on the other hand, affects AMPAR by inhibiting NSF activity. As AMPAR plays a crucial role in synaptic plasticity and memory formation, the dysregulation of NSF is likely to result in early synaptic dysfunction and cognitive decline in AD. Mouse models suggest that inhibiting NSF activity leads to memory loss in mice, and this memory impairment can be rescued by the absence of tau protein (Prikas et al., 2022).

Screening for epilepsy-related genes has revealed a correlation between the NSF gene and the occurrence of spontaneous seizures (Guan et al., 2001; Yin et al., 2005; Yu et al., 2002). Researchers have identified mutations in the NSF gene as a cause of early infantile epileptic encephalopathy (DEE) (Suzuki et al., 2019). In an animal model of epilepsy induced by kainic acid, researchers observed a downregulation of NSF mRNA and protein expression in the hippocampal CA1 region following spontaneous seizures (Yin et al., 2005). NSF plays a role in regulating neurotransmitter release and influences signal transduction. Elevated levels of NSF can disrupt synaptic signal transduction, leading to network dysfunction among postsynaptic neurons, and providing a neurobiological basis for seizure onset. NSF alters the function and distribution of different receptors, thereby influencing the activity of various neurotransmitter systems, which may contribute to the molecular basis of epilepsy. Studies have observed an association between elevated levels of NSF and AMPAR in synapses with spontaneous sharp waves in temporal lobe epilepsy (Herold et al., 2018). Genetic variations in the NSF gene are associated with developmental DEE. While the molecular mechanisms underlying NSF-related DEE are not yet fully understood, ongoing research continues to shed light on this complex condition. NSF mutations can disrupt intracellular membrane fusion processes through dysregulation of protein structure, leading to neurodegeneration via the mTOR pathway. Inhibiting mTOR could be a therapeutic approach for NSF-related DEE (Hayashi et al., 2023).

Transient cerebral ischemia can lead to increased proteins such as NSF within the PSD, suggesting that cerebral ischemia may initiate protein assembly mechanisms related to NSF within the PSD. Such changes can impact the composition of postsynaptic proteins, resulting in changes in neural transmission function and selective neuronal cell damage (Hu et al., 1998). Inhibition of NSF ATPase activity can lead to the inactivation and deposition of NSF proteins, halting intracellular membrane transport activities and causing cell death in cell cultures and Drosophila models (Lu et al., 2014). Similarly, complete inactivation of NSF ATPase may interrupt intracellular membrane transport activities, resulting in delayed neuronal death after transient cerebral ischemia. Studies indicate that the injury events induced by the inactivation of NSF represent a novel and common mechanism in cerebral ischemia–reperfusion injury. Transient cerebral ischemia leads to NSF inactivation, triggering a series of lethal releases and delayed neuronal death (Yuan et al., 2018a,b). Specifically, cerebral ischemia can result in the inactivation of NSF ATPase activity, initiating a cascade of events: extensive accumulation of Golgi fragments, transport vesicles, and late endosomes, disrupting the Golgi-endosome-lysosome pathway and releasing cathepsin B (CTSB). The release of CTSB can induce changes in mitochondrial outer membrane permeability and result in cell death, ultimately contributing to cerebral ischemia–reperfusion injury (Yuan et al., 2021). Following the release of CTSB, transient cerebral ischemia triggers delayed neuronal death. During this process, NSF ATPase becomes irreversibly trapped in aggregates of inactive proteins within ischemic neurons destined to die (Yuan et al., 2018a,b). These findings suggest that the inactivation of NSF ATPase plays a crucial role in the subsequent neuronal death following transient cerebral ischemia and stroke.