The SNARE (soluble N-ethylmaleimide-sensitive fusion protein-attachment protein receptor) protein SNAP-25 (25 kDa synaptosome-associated protein) is essential for regulated exocytosis in neuronal and neuroendocrine cells. Whereas the majority of SNARE proteins contain transmembrane domains, SNAP-25 is instead anchored to membranes by the palmitoylation of a central cysteine-rich region. In this review, we discuss the mechanisms of SNAP-25 palmitoylation and how this modification regulates the intracellular trafficking and exocytotic function of this essential protein.
- aspartate-histidine-histidine-cysteine (DHHC) palmitoyltransferase
- 25 kDa synaptosome-associated protein (SNAP-25)
- soluble N-ethylmaleimide-sensitive fusion protein-attachment protein receptor (SNARE)
SNARE (soluble N-ethylmaleimidesensitive fusion protein-attachment protein receptor) proteins and regulated exocytosis
Exocytosis is the process whereby intracellular vesicles fuse with the plasma membrane. This membrane fusion event mediates the delivery of proteins and lipids to the plasma membrane and the secretion of soluble vesicle cargo to the cell exterior. Exocytosis occurs constitutively in all cells where it is essential for the insertion of newly synthesized proteins into the plasma membrane. In addition, a specialized form of exocytosis, termed regulated exocytosis, occurs in certain cell types and mediates the controlled secretion of physiologically important molecules such as neurotransmitters, peptides and hormones. The stimulus for regulated exocytosis in neuronal and neuroendocrine cells is generally an increase in cytoplasmic Ca2+ levels .
Exocytosis is dependent on a number of distinct and overlapping protein–protein interactions . Central to this fusion event is the interaction between SNARE proteins localized on the vesicle and plasma membranes. In neuronal cells, the plasma membrane SNAREs SNAP-25 (25 kDa synaptosome-associated protein) and syntaxin 1 interact with the vesicle SNARE VAMP2 (vesicle-associated membrane protein 2) to form a complex that is thought to drive membrane fusion [3,4]. As SNARE proteins are essential for regulated exocytosis there is considerable interest in how these proteins are regulated by protein–protein and protein–lipid interactions and by post-translational modifications such as phosphorylation [2,5]. Interestingly, SNAP-25, VAMP2 and syntaxin 1 have all been reported to be modified by an additional type of post-translational modification, namely palmitoylation [6–8].
Palmitoylation refers to the post-translational addition of the 16-carbon saturated fatty acid palmitate to proteins in one of two ways: by a cleavable thioester linkage to the thiol group of cysteine residues (known as S-palmitoylation) or, less frequently, by an amide linkage (known as N-palmitoylation) [9,10].
Nearly all palmitoylated proteins are S-palmitoylated. However, it is important to note that referring to a protein as being ‘palmitoylated’ is historically based on the observed incorporation of [3H]palmitic acid into proteins. However, when the lipid profile of palmitoylated proteins has been examined using techniques such as MS, it is clear that other medium- and long-chain fatty acids are also attached to proteins through thioester linkages, such as the saturated fatty acids myristate and stearate, and the unsaturated fatty acids oleate or arachidonate [11,12]. Therefore, the term S-acylation is more appropriate; however, since the majority of S-acylated proteins incorporate palmitate, the term palmitoylation is generally used, and will be used synonymously with acylation hereafter.
Protein palmitoylation has a primary function in allowing otherwise soluble proteins to become stably membrane-bound by increasing the affinity of the modified protein for the lipid bilayer. However, despite the important role that palmitoylation plays in membrane binding, it is becoming clear that palmitoylation does not always function as a membrane anchor. This point is emphasized by the demonstrated palmitoylation of a number of transmembrane proteins, which clearly do not require palmitoylation for membrane anchoring (e.g. VAMP2 and syntaxin 1). Indeed, palmitoylation has now been shown to play diverse roles in addition to membrane tethering: in the targeting of both soluble and transmembrane proteins to distinct membrane compartments and membrane microdomains [13–16]; in promoting protein stability, for example by regulating protein degradation or ensuring correct protein folding [17,18]; and in regulating protein–protein interactions .
Palmitoylation is predominantly an enzyme-mediated event, and the first bona fide palmitoyltransferases were identified in studies in the budding yeast Saccharomyces cerevisiae [20,21]. Two proteins were identified that were required for Ras2p palmitoylation in yeast, Erf2p (effector of Ras function protein 2) and Erf4p. Palmitoyltransferase activity towards Ras2p requires both Erf2p and Erf4p, since neither protein is capable of carrying out palmitoylation reactions individually . Concurrent with the discovery of the Erf2p/Erf4p complex as a palmitoyltransferase, another yeast protein, Akr1p (ankyrin repeat-containing protein 1), was shown to mediate the S-palmitoylation of the type I casein kinase Yck2p . Unlike Erf2p, Akr1p does not need to be part of a complex for palmitoyltransferase activity, as it is able to work independently without accessory subunits . Both Erf2p and Akr1p share a conserved ~50 amino acid domain, known as a DHHC (aspartate-histidine-histidine-cysteine) CRD (cysteine-rich domain). Subsequent analysis of the yeast and mammalian genomes has identified a family of seven DHHC-CRD proteins in yeast; Akr1p, Akr2p, Erf2p, Pfa3p, Pfa4p, Pfa5p and Swf1p; and 23 in mammals; DHHC1–DHHC23 . Genetic and biochemical analyses have demonstrated the palmitoyltransferase activity of several of these DHHC-CRD containing proteins, and accordingly, many substrates have been identified. Interestingly, DHHC proteins are also palmitoylated and the DHHC domain is essential for autopalmitoylation and also for palmitoyltransferase activity, suggesting that palmitoylated DHHC is a key intermediate in the palmitoylation reaction.
DHHC proteins are polytopic membrane proteins that are localized on several intracellular compartments including the Golgi, ER (endoplasmic reticulum), plasma membrane, endosomes/vesicular compartments and the yeast vacuole . To date, it is not clear what elements present within DHHC proteins dictate their respective intracellular localizations.
The versatility of protein palmitoylation as a regulatory module is enhanced by its reversibility, and a number of proteins undergo dynamic cycles of palmitoylation and depalmitoylation (see Figure 1). Compared with the DHHC family of palmitoyltransferases, far less is known about the proteins that catalyse depalmitoylation.
Palmitoylation of SNAP-25
Whereas syntaxin 1 and VAMP2 have membrane-spanning domains, SNAP-25 is a soluble protein that requires palmitoylation on a cluster of cysteine residues to facilitate stable membrane attachment. Deletion of a 12 amino acid region containing the four cysteine residues, or mutation of the cysteine residues within the CRD of SNAP-25, both render the protein cytosolic [6,26–28]. Since palmitoylation of proteins is mediated by membrane-anchored DHHC proteins, mechanisms that target SNAP-25 to its site of palmitoylation must exist. Some studies have suggested that initial membrane binding of SNAP-25 might be mediated by interaction with its SNARE partner syntaxin 1 [26,29]. However, the minimum membrane-binding domain of SNAP-25 has been mapped to amino acids 85–120, a region containing the palmitoylated CRD plus the 28 amino acids immediately downstream of this region ; this domain is not thought to interact with syntaxin. Indeed, we found that membrane binding of SNAP-25 was not affected following siRNA (small interfering RNA)-mediated knockdown of syntaxin 1A in PC12 cells . Furthermore, down-regulation of the SM (Sec1/Munc18) protein Munc18 prevents plasma membrane delivery of syntaxin in PC12 cells, whereas the trafficking of SNAP-25 is unaffected , and SNAP-25 was reported to traffic to the plasma membrane in non-neuronal cells that do not express syntaxin 1 . Thus, it appears that initial membrane targeting and palmitoylation of SNAP-25 may occur independently of syntaxin.
We recently reported results of mutational analyses aimed at enhancing our understanding of how SNAP-25 is targeted to membranes and palmitoylated . In this analysis, we focused on the minimal membrane targeting sequence (amino acids 85–120) present in full-length SNAP-25. Alanine-scanning mutagenesis identified two sets of mutants that blocked membrane association of SNAP-25 in PC12 cells: cysteine mutants and mutations at a highly conserved Gln116/Pro117 motif downstream of the CRD.
Previous work reported that when individual cysteine residues were mutated to alanine, serine or glycine, membrane binding of SNAP-25 was significantly reduced [26–28]. Our results confirmed that cysteine-to-alanine mutations inhibit membrane binding of SNAP-25; interestingly, however, we found that when individual cysteine residues were replaced with the more hydrophobic leucine, membrane binding was largely preserved. These results are consistent with the idea that the hydrophobicity of the CRD, which has the sequence Phe-Cys-Gly-Leu-Cys-Val-Cys-Pro-Cys, is important for the initial membrane interaction of SNAP-25. We proposed that the CRD facilitates a weak membrane interaction of unpalmitoylated SNAP-25, thus allowing transient access to membrane-anchored DHHC proteins. In this model, palmitoylation at cellular membranes increases the membrane affinity of SNAP-25, promoting stable membrane binding and facilitating subsequent sorting; a similar mechanism of palmitoylation and membrane binding has been proposed previously for Ras proteins and for the molecular chaperone cysteine-string protein [34–37]. In support of this model, the inefficient membrane binding of SNAP-25 observed in HEK (human embryonic kidney) -293 cells was significantly enhanced by co-expression of either DHHC3, DHHC7 or DHHC17, proteins previously shown to palmitoylate SNAP-25 . Together these data suggest that SNAP-25 interacts autonomously with cell membranes and requires only sufficient cellular expression of partner DHHC proteins to mediate stable membrane association, allowing subsequent intracellular sorting.
What then is the role of Gln116-Pro117 identified in the alanine scanning screen in PC12 cells? Further analyses in HEK-293 cells found that these residues are essential for interaction with DHHC17, but not DHHC3. Thus, the loss of membrane binding observed with these mutants in PC12 cells is consistent with the conclusion that DHHC17 plays a major role in palmitoylation and membrane binding of SNAP-25 in this cell line; indeed work in Drosophila reported that SNAP-25 was mislocalized in DHHC17 null mutants [39,40].
Regulation of SNAP-25 function by palmitoylation
There is considerable interest in the idea that cellular membranes are composed of distinct microdomains that differ in their lipid and protein content . In particular, the co-clustering of proteins at specific regions of the plasma membrane may be important in ensuring efficient protein–protein interaction. Numerous studies have reported that palmitoylated proteins associate with cholesterol-rich ‘raft’ domains in the plasma membrane [42,43]. Indeed, biochemical and cell imaging approaches have suggested that SNAP-25 may also be partly localized in cholesterol-rich microdomains. Immunofluorescence analyses of plasma membrane fragments revealed that SNAP-25 is clustered in a cholesterol-dependent manner , and work from our group found that SNAP-25 co-purifies with cholesterol-dependent membrane fractions recovered from detergent-solubilized cells . Chemical depalmitoylation of SNAP-25 reduced the interaction with detergent-resistant fractions, demonstrating that palmitoylation of SNAP-25 is important for association with these structures .
Further analyses revealed that the number of palmitoylation sites in SNAP-25 correlated with the level of association with cholesterol-rich fractions. Introducing an additional cysteine into SNAP-25 (F84C mutant) significantly increased association of the protein with cholesterol-rich membranes in vitro . By examining the ability of the F84C SNAP-25 mutant to rescue exocytosis in PC12 cells expressing the light chain of botulinum neurotoxin E, it was concluded that the cysteine mutant had a reduced exocytotic activity . Based on these observations, we proposed that the palmitoylation status of SNAP-25 regulates its association with cholesterol-rich membrane domains, which in turn regulates its ability to support efficient exocytosis. This model poses many questions, which we are currently addressing: (i) how many cysteine residues are actually palmitoylated in SNAP-25? (ii) Is there heterogeneity in the palmitoylation of cellular SNAP-25? (iii) Is SNAP-25 palmitoylation dynamically regulated? (iv) Are there DHHC proteins localized to the plasma membrane that modify SNAP-25? (v) Is DHHC and/or Apt1 (acyl protein thioesterase 1) activity modulated in response to exocytotic stimulation? (vi) What do high-resolution approaches [e.g. STED (stimulated emission depletion) microscopy, electron microscopy] tell us about SNAP-25 membrane microlocalization? Answering these questions will facilitate a more comprehensive understanding of how the regulation of SNAP-25 palmitoylation is coupled to efficient exocytosis.
Our work is funded by the Medical Research Council U.K.
Molecular Mechanisms in Exocytosis and Endocytosis: 7th Junior Academics Meeting, an Independent Meeting held at University of Edinburgh, Edinburgh, U.K., 5–7 April 2009. Organized and Edited by Rolly Wiegand (Edinburgh, U.K.).
Abbreviations: Akr1p, ankyrin repeat-containing protein 1; Apt1, acyl protein thioesterase 1; CRD, cysteine-rich domain; Erf2p, (etc.), effector of Ras function protein 2 (etc.); DHHC, aspartate-histidine-histidine-cysteine; HEK, human embryonic kidney; SNAP-25, 25 kDa synaptosome-associated protein; SNARE, soluble N-ethylmaleimide-sensitive fusion protein-attachment protein receptor; VAMP2, vesicle-associated membrane protein 2
- © The Authors Journal compilation © 2010 Biochemical Society