The fusion of membranes is a central part of the physiological processes involving the intracellular transport and maturation of vesicles and the final release of their contents, such as neurotransmitters and hormones, by exocytosis. Traditionally lipids have been regarded as structural elements playing a relatively minor role in the molecular mechanisms of exocytosis whereas proteins such as SNAREs (soluble N-ethylmaleimide sensitive factor attachment protein receptors) are thought to be the central elements that generate the specificity and force needed for overcoming the repulsion of the negative charges within lipid bilayers that oppose fusion. The effect of sphingosine and synthetic derivatives on the heterologous and homologous fusion of organelles can be considered as a new mechanism of action of sphingolipids influencing important physiological processes, which could underlie therapeutic uses of sphingosine derived lipids in the treatment of neurodegenerative disorders and cancers of neuronal origin such neuroblastoma.
1. Lipids and Exocytosis
The fusion of vesicles with other lipid bilayers is essential for intracellular trafficking and release of neurotransmitters and hormones
[1,2,3,4][1][2][3][4]. Release of neurotransmitters and hormones involves the active transport of the vesicles using cytoskeletal elements such as F-actin and microtubules
[5[5][6],
6], tethering and docking of the vesicles with the target membrane
[7,8][7][8], and finally calcium-induced fusion of the membranes, resulting in the release of vesicular contents to the extracellular media via exocytosis
[9,10][9][10].
Traditionally lipids have been regarded as structural elements playing a relatively minor role in the molecular mechanisms of exocytosis whereas proteins such as SNAREs (soluble N-ethylmaleimide sensitive factor attachment protein receptors) are thought to be the central elements that generate the specificity and force needed for overcoming the repulsion of the negative charges within lipid bilayers that oppose fusion
[11,12,13,14][11][12][13][14]. Even so, membranes need to adopt curved shapes during fusion, which is heavily influenced by the molecular structure of lipid and lysophospholipids in particular facilitating formation of conic shapes that are amenable to fusion
[15,16][15][16].
In addition, some lipid components aggregate to form microdomains that facilitate recruitment of the proteins that catalyse exocytosis. For example, phosphatidyl inositol 4,5-biphosphate (PIP2) microdomains seem to act as beacons for coordinating F-actin bundles involved in recruiting SNARE proteins, and are needed for the translocation of secretory vesicles to their specific docking sites
[17,18,19,20][17][18][19][20]. Further, cholesterol accumulation into lipid rafts can contribute by organizing clusters of secretory proteins such as syntaxin-1
[21].
Furthermore, lipids can be incorporated into secretory proteins such as SNAP-25 via post-translational modification consisting in the acylation of cysteine residues by palmitate, a saturated 16-carbon fatty acid and in that way affect the location and the function of this protein
[22]. It seems that the palmitoylation of 4 residues of this SNARE, increase the clustering of SNAP-25 in cholesterol and sphingomyelin rich lipid rafts, that could enhance the formation of secretory active sites thereby acting as a cohesive factor
[23,24][23][24]. Moreover, some studies suggest that this modification exerts changes that enhance the forces acting on the zippering of the SNARE complex thereby enhancing the fusion of membranes
[25], while others suggest that palmitoylation contributes simply to the anchoring role associated with the insertion of fatty acids in the membranes, as it has been traditionally accepted
[26].
In addition to SNAP-25, other proteins participating in secretion such as synaptobrevin-2 have been demonstrated to be palmitoylated. This protein is modified by palmitoylation during development since the modification is only found in adult rats and not in embryonic cells
[27]. Furthermore, one of the calcium sensors associated with membrane fusion, the protein synaptotagmin 1, incorporates fatty acid acylations in 5 residues close to the membrane anchoring domain
[28]. Finally, the cysteine string protein (CSP), a molecular chaperone involved in protein folding
[29], is heavily palmitoylated with 14 cysteine residues. These lipid residues have been found to be important for exocytosis in neuronal and endocrine cells
[30,31][30][31].
2. The Direct Interaction between Sphingolipids and SNAREs
As mentioned above there are several roles that could be assumed by lipids in order to influence the secretory process ranging from structural determinants to specific “markers” locating the position of exocytotic active sites. Nevertheless more recently it has been highlighted that certain type of lipids acting as intracellular signals, the so called signalling lipids, could interact directly with the SNARE proteins constituting the secretory machinery. Signalling lipids such as sphingosine or arachidonic acid (AA) are released from structural phospholipids by the action of phospholipases
[32[32][33],
33], and upon generation of saturated or polyunsaturated fatty acids (PUFAs), normally present in the sn-2 position, they diffuse freely to interact with either synaptobrevin in the case of sphingosine
[34] or target syntaxin-1 in the case of AA
[35].
These signalling lipids interact with a variety of parts of the secretory machinery such as SNAREs and in that way they modulate exocytosis. In this sense AA interaction with the t-SNARE syntaxin 1 was the first reported evidence of a direct interaction of signalling lipids with SNAREs linked to the exocytotic fusion machinery in 2005
[35]. Herein, the direct application of AA, or alternatively the treatment with PLA2s, was able to promote the formation of SNARE complexes in in vitro experiments using membrane preparations. Furthermore, this interaction occurs even in the “closed” conformation of syntaxin-1 promoted by the presence of Munc-18, suggesting that AA could access the hydrophobic core of syntaxin-1 when this protein forms part of the stable plasma membrane dimers with Munc-18
[35,36][35][36]. This AA interaction seems to be an essential aspect of syntaxins since it has been reported to occur in different syntaxins such as the 1 and 3 forms
[36].
The AA regulation of the secretory machinery was further emphasized when it was reported that the protein α-synuclein accumulating in the neuronal cytoplasm during the pathogenesis of Parkinson disease, was able to sequester AA and in this way impede the activation of exocytosis caused by this signalling lipid
[37]. Therefore, the physiological regulation of syntaxins by AA appears to be essential to sustain the correct levels of exocytosis and might be altered in neuronal disorders.
Recently, by screening the ability of a variety of lipids in changing the formation of the SNARE complex in vitro it was proven that specifically sphingosine and some related lipids were able to induce synaptobrevin-2 binding to syntaxin-1 and SNAP-25 dimers formed in the target membrane
[34] (
Figure 1). Interestingly, L-sphingosine was as efficient as D-sphingosine in promoting SNARE complex formation indicating that this compound may alter the local membrane environment influencing synaptobrevin conformation. From the analysis of the sphingosine-related compounds assayed it was concluded that the length of the carbon chain and the presence of a positive charge in the head of the lipids were two essential features needed to promote SNARE-complex formation and the associated enhancement of the exocytosis
[34]. The involvement of the vesicular SNARE synaptobrevin in the sphingosine effect was further emphasized when in synaptobrevin-2 knockout mice there was no evidence of secretory modulation by sphingosine
[34]. Treatment of control cultured cells of neuronal and neuroendocrine origin with increasing concentrations of sphingosine resulted in dose-dependent enhancement of exocytosis with EC50 values around 10 µM
[34]. A further step was taken to study the sphingosine modulation of SNAREs when the endogenous production of sphingosine was tested by using treatment with external sphingomyelinases (SMases) of nerve terminals
[34] and cultured neuroendocrine chromaffin cells
[38,39][38][39]. The results support the sphingosine role of synaptrobrevin-2 since treatment of these cell preparations with Botulinum Neurotoxin-D, which cleaves synaptobrevin-2
[34], prevented the release of neurotransmitters induced by SMases treatment.
Figure 1. Sphingosine influences different steps of exocytosis. After the production of the signalling lipid sphingosine from sphingolipids, this modulates at least two phases of the exocytotic process, increasing the formation of the SNARE complex which enhance the frequency of vesicle fusions (1), and the transition from the “kiss and run” mode of fusion to the full collapse mode (2), causing an increase in the number of neurotransmitters released per event. Show here also the FTY720 structure.
Another study suggested that sphingosine could additionally act through a different molecular mechanism involving the SNARE syntaxin-1
[40]. In this scenario, sphingosine facilitates the formation of the dimer Munc-18-syntaxin-1, which may reduce the number of the vesicles in a competent state for fast fusion, thus somewhat opposing the conclusions obtained in the articles mentioned above.
Interestingly, a physiological derivative of sphingosine, sphingosine-1P also has been found to alter the secretory response in neuroendocrine chromaffin cells
[15,41][15][41] by a completely different mechanism; i.e., via modulating the amount of intracellular calcium. In this case, and in a recent report
[42], this modified sphingosine appears to have a complex effect on exocytosis at different concentrations. At 1 μmol/L, sphingosine-1P, the Ca2+-activated K+ currents (IK(Ca) were inhibited in electrophysiology experiments whereas at a higher dose of 10 μmol/L the current was activated causing a remarkable increase in intracellular calcium and enhanced exocytosis
[42].
Therefore, it is clear that the direct interaction of sphingosine or physiological derivatives such as sphingosine-1P with different elements of the secretory machinery may cause a complex modulation of the exocytotic response in neuronal and neuroendocrine cellular models, and that the complexity of such interactions deserves further study.
3. Sphingolipids Alter the Single Fusion Properties of Neurotransmitter Release
To study the way sphingolipids affect the secretory process, it is important to appreciate that this is a multi-step process starting with the translocation of the vesicles to the plasma membrane using active transport involving cytoskeletal elements
[43,44][43][44], maturation of the docked vesicles to be competent for membrane fusion
[8[8][45],
45], and finally the fusion process itself that includes the opening of a fusion pore, subsequent dilation, and then the release of active substances that ends in full collapse of the vesicle into the plasma membrane
[46,47,48][46][47][48] (
Figure 2). Therefore, the use of biophysical techniques such as membrane capacitance methods
[49[49][50],
50], and amperometry
[51[51][52],
52], which resolve distinct stages of exocytosis, is essential for better understanding how sphingolipids alter the secretory pathway.
Figure 2. FTY-720 induces drastic changes in vesicles and mitochondria in chromaffin cells. (a) Schematic of the changes induced by 20 μmol/L FTY-720 incubation of cultured bovine chromaffin cells after a variety of incubation times, where the dense vesicles are depicted in red and mitochondria in green. (b) Micrographs showing examples of the changes by using electron microscopy. Three levels of changes are proposed: (1) Rapid formation of microvesicles from dense organelles (dark round organelles) and initiation of the fusion with mitochondria as indicated by arrows, (2) Heterotypic fusion of vesicles with mitochondria to form elongated mixed organelles, and (3) Formation of round mixed macroorganelles consisting of mitochondria incorporating the dense cores of several granules.
Later on, two of the groups participating in this work and using distinct cellular systems studied more deeply the fusion exocytotic steps affected by sphingosine. First, and by using capacitance techniques to distinguish unitary fusion events in pituitary lactrophophs, Robert Zorec’s group demonstrated that sphingosine raises the frequency of fusion of both small vesicles and large dense granules
[53]. In addition, they demonstrated that sphingosine promotes the full fusion mode of the larger vesicles while the smaller ones fused via the “Kiss and run” mode, thus only partially releasing their content. Therefore, it appears that sphingosine modulation depends on the size of the vesicles, and favors a distinct mode of fusion as shown in
Figure 1.
WThe researche
rs reached a similar conclusion using amperometry to measure catecholamines from cultured bovine chromaffin cells
[38].
WThe researche
r used Sphingomyelinase to produce sphingosine and derivatives, which enhanced the number of events released in response to cell depolarization with a high concentration potassium solution. Moreover, the amperometry technique resolved the amount of neurotransmitter released per event and the kinetics of individual vesicle fusion, and additionally demonstrated that enzymatic sphingosine production enhances both parameters, which indicates that sphingosine promotes the full fusion of dense granules and that in control conditions, without added sphingosine, the release of catecholamines is only partial, indicating that it must be happening via the “Kiss and run” mode. Two years later, in collaboration with Dr. Alvarez de Toledo at the University of Seville
we, the researcher performed experiments using the whole cell and on-cell capacitance techniques to resolve the size of the vesicles fused and
wethe researcher found that sphingomyelinase treatment of rat chromaffin cells resulted in a clear increase in the frequency of fusions of both small and large vesicles without affecting the size of the vesicles measured by electron microscopy
[39], in agreement with the results obtained in lactotrophops
[53].
In a recent report, intracellular sphingosine-1P was also found to accelerate the rate of fusion pore expansion in chromaffin cells from mouse
[54], thus indicating that not only sphingosine but also its derivatives may control directly the properties of membrane fusion. In agreement with this, the exocytosis from preparations of cortical granules from oocytes, has been found to be sustained by sphingosine and sphingosine-1P, thus the presence of a critical amounts of these sphingolipids appear essential to maintain normal levels of exocytosis
[55].
It is important to note also that both sphingosine and sphingosine-1P might regulate other steps of the secretory process such us the initial entrance of calcium after cell stimulation
[15[15][41],
41], synapsin levels
[56], and even the rate of membrane retrieval by endocytosis
[57].