|Home | About | Journals | Submit | Contact Us | Français|
The sequential transit of cargo in the secretory pathway depends upon membrane budding and membrane fusion reactions. The final delivery of cargo to the plasma membrane involves exocytic vesicle fusion, which occurs in several steps involving the tethering/docking of vesicles at the plasma membrane, priming events that enable fusion readiness, and the final fusion step. The final fusion step for Ca2+-dependent vesicle exocytosis in neuroendocrine cells is mediated by topologically ordered SNARE protein interactions with VAMP2 on the vesicle zippered into heterodimer complexes of SNAP-25 and syntaxin-1 on the plasma membrane to form a twisted four helix bundle. A tightly assembled SNARE complex in trans across the bilayers promotes their close apposition and membrane merger (Malsam et al., 2008). The Ca2+-dependent triggering of SNARE-dependent fusion is mediated by several synaptotagmin isoforms on the vesicle that engage SNARE complexes and the plasma membrane. A key issue for understanding vesicle exocytosis is elucidating the pathway for SNARE protein assembly prior to triggered fusion.
The events responsible for vesicle tethering/docking at the plasma membrane are incompletely understood but appear to involve the plasma membrane SNAREs SNAP-25 and syntaxin-1 interacting with the vesicle proteins rabphilin and Slp4a, respectively (Verhage & Sorensen, 2008). Munc-18-1 also plays an essential role in the tethering/docking process presumably by interacting with syntaxin-1. Not all tethered/docked vesicles are competent for Ca2+-triggered fusion in neuroendocrine cells (Wojcik & Brose, 2007). The ~10% of the plasma membrane-localized vesicles that fuse in response to Ca2+ elevations are referred to as the release ready pool of vesicles. The conversion of vesicles into this pool is considered to correspond to priming, a set of reactions that occurs after vesicle tethering/docking. Priming reactions are also responsible for maintaining secretion during prolonged stimulation following the depletion of release ready vesicle pools. It is unclear how SNAREs re-organize into trans SNARE complexes between the tethered/docked state and the primed state (release ready pool).
A number of lipids and proteins are essential for priming in neuroendocrine cells. Munc13-1, a multi-domain protein with C2, C1 and MH(munc13 homology) domains, was shown to be essential for the priming of synaptic vesicles in central synapses in mouse knockout studies where the phenotype was an intact docked pool of vesicles that failed to undergo Ca2+-triggered exocytosis (Wojcik & Brose, 2007). Indirect measurements suggested that a release ready pool was absent. Under conditions of strong stimulation, the DAG-binding C1 domain of Munc13-1 is required, indicating a role for DAG in priming reactions. DAG may also function to activate protein kinase C isoforms, which have been implicated in the regulation of Munc18-1 function in priming. The possible mechanisms underlying Munc13-1 function in priming have been reviewed (Wojcik & Brose, 2007).
More recently, the roles of PI(4,5)P2 and CAPS in vesicle priming have become evident. PI(4,5)P2 was originally identified as essential for Ca2+-triggered vesicle exocytosis in permeable neuroendocrine cells by the identification of phosphatidylinositol transfer protein (Hay & Martin, 1993) and PI(4)P 5-kinase (Hay et al., 1995) as components required for ATP-dependent priming reactions. More recent work established that PI(4,5)P2 is required for the priming of vesicles into a release ready pool (Olsen et al., 2003; Milosevic et al., 2005) and for ongoing rates of stimulated secretion (Aikawa & Martin, 2003). A number of PI(4,5)P2-binding proteins have been proposed for mediating the essential role of this lipid in vesicle priming reactions including synaptotagmin, CAPS, rabphilin, Munc18-1, SCAMP2 and actin-binding proteins (Martin, 1998). Our recent studies (James et al., 2008) reconstituted PI(4,5)P2 function in a SNARE-dependent liposome fusion assay and identified CAPS as a PI(4,5)P2-dependent regulator of SNARE protein function.
CAPS is a multi-domain protein that contains C2, PH and MH domains. It was discovered as a brain cytosolic factor that reconstituted Ca2+-dependent secretion from permeable neuroendocrine cells (Walent et al., 1992). Studies in permeable cells found that the process restored by CAPS exhibited the kinetics of vesicle priming and was highly dependent upon the ongoing synthesis of PI(4,5)P2 (Grishanin et al., 2004). Recent mouse knockout studies confirmed the essential role of CAPS in vesicle priming in adrenal chromaffin cells and in central synapses (Jockusch et al., 2007; Liu et al., 2008). CAPS directly binds PI(4,5)P2 and undergoes conformation changes, and we proposed that CAPS is an essential effector for PI(4,5)P2 in regulated vesicle exocytosis that mediates the role of this lipid in priming. The studies presented here were directed at reconstituting PI(4,5)P2 and CAPS function in the SNARE-dependent liposome fusion assay. These in vitro studies (James et al., 2008; James et al., 2009) provide important clues on the mechanisms by which PI(4,5)P2 and CAPS function in vesicle exocytosis to organize SNARE complexes for fusion.
To determine the location and size of the PI(4,5)P2 pools potentially required for regulated vesicle exocytosis, we incubated plasma membrane sheets prepared from PC12 cells with the monovalent PI(4,5)P2-binding PLCδ1-PH-GFP protein (Fig. 1A–D). Strong punctuate labeling on the membrane corresponded to high concentration microdomains of PI(4,5)P2 and not to membrane invaginations. Calibration of the fluorescent signal (Fig. 1E,F) indicated that microdomains consisting of ≥ 6 mol% PI(4,5)P2 were present on the plasma membrane of these neuroendocrine cells. Triple co-localization studies revealed that ~50% of CAPS on these membrane sheets co-localized with PI(4,5)P2–containing microdomains and docked dense-core vesicles (Fig. 1G,H). Overall ~20% of the docked vesicles co-localized to sites containing PI(4,5)P2 and CAPS. Because PI(4,5)P2 and CAPS are essential for priming vesicles, these vesicles may correspond to a release ready pool. Aoyagi et al. (Aoyagi et al., 2005) provided evidence for preferential vesicle exocytosis at microdomains enriched in PI(4,5)P2 and syntaxin-1. We suggest that ~20% of the vesicles that undergo preferential release may localize to microdomains enriched for PI(4,5)P2 and CAPS.
As an inverted cone-shaped lipid in the cytoplasmic leaflet, PI(4,5)P2 would be anticipated to inhibit membrane fusion by antagonizing the high negative membrane curvature required for the formation of a hemifusion intermediate that transitions to full fusion (Chernomordik & Zimmerberg, 1995). In a SNARE-dependent liposome fusion assay, we found that concentrations of PI(4,5)P2 in the range of 6 mol% strongly inhibited fusion (Fig. 2A,B). This appeared to correspond to the expected curvature effects of an inverted cone-shaped lipid because it was non-additive with the similar effects of lysoPC (Fig. 2C). As expected for this mechanism, similar inhibitory effects were observed with PI(3,4)P2 but lesser inhibition was observed for PI(4)P (James et al., 2008). In addition, similar inhibition was observed for inclusion of the PI(4,5)P2 in either donor or acceptor liposomes (James et al., 2008).
Syntaxin-1 and VAMP2 each contain juxtamembrane stretches of basic amino acids, which exhibit high affinity interactions with negatively charged PI(4,5)P2. To examine the possible role of these interactions in the inhibition of fusion by PI(4,5)P2, we generated K to A substitutions in the juxtamembrane domain of syntaxin-1. When tested in the liposome fusion assay, the K to A substitution mutants functioned normally in PC/PS liposomes (Fig. 2D). However, in liposomes containing PI(4,5)P2, the K to A substitution mutants exhibited more strongly inhibited fusion. This suggests that syntaxin-1, via its basic membrane-proximal domain, counteracts the inhibitory effects of PI(4,5)P2. A ring of SNARE complexes during fusion could serve to sequester PI(4,5)P2 away from the central fusion site.
The preceding results on the inhibition of fusion by PI(4,5)P2 in reconstituted liposome fusion fail to account for the well-characterized positive essential role of PI(4,5)P2 in regulated vesicle exocytosis (Hay et al., 1995). We considered the possibility that additional factors missing from the minimal fusion assay may be needed to reconstitute this positive role of PI(4,5)P2. We found that CAPS addition to the minimal system dramatically accelerated rates of fusion (Fig. 3A). CAPS stimulation of fusion was highly dependent on PI(4,5)P2, exhibiting maximal effects at 5 mol% (Fig. 3B). In addition, the strongly synergistic effect of CAPS and PI(4,5)P2 was only evident when PI(4,5)P2 was present in the SNAP-25/syntaxin-1-containing acceptor liposomes (Fig. 3C). Lastly, mutations in the PH domain of CAPS that confer loss of PI(4,5)P2 binding were found to eliminate the ability of CAPS to stimulate SNARE-dependent fusion (Fig. 3D). This modified SNARE-dependent liposome fusion assay successfully reconstitutes many of the features that were defined by the permeable PC12 cell assay for vesicle priming (Grishanin et al., 2004; James et al., 2008). Both exhibit a strong dependence on PI(4,5)P2 and CAPS. These results indicate that CAPS functions as an effector for PI(4,5)P2 in its positive essential role for vesicle exocytosis.
It was unlikely that CAPS stimulated SNARE-dependent liposome fusion solely through interactions with PI(4,5)P2. A CAPS PH domain fusion protein not only failed to stimulate SNARE-dependent fusion but prevented CAPS from doing so (James et al., 2008). In addition, full length CAPS did not promote aggregation of the PI(4,5)P2, SNARE-containing liposomes, which excluded an aggregation mechanism for the increased membrane fusion stimulated by CAPS. The asymmetric requirement for PI(4,5)P2 in the SNAP-25/syntaxin-1-containing acceptor liposomes for CAPS stimulation suggested that CAPS is preferentially recruited to these membranes and interacts directly with the SNARE proteins. Studies of CAPS binding to SNARE-containing liposomes were undertaken (James et al., 2009) and revealed the ability of CAPS to directly interact with membrane-integrated syntaxin-1 (Fig. 4A–D). Previous attempts to demonstrate CAPS interactions with SNARE proteins were unsuccessful because truncated soluble SNARE proteins were utilized. Our current work indicates that CAPS exhibits high affinity interactions with membrane-integrated SNAREs because of concomitant CAPS-SNARE and CAPS-membrane binding. Extensive characterization of the SNARE-binding properties of CAPS and that of its MH domain will be described elsewhere (N. Daily et al., in preparation; C. Khodthong et al., in preparation).
Fusion is mediated through the formation of heterotrimeric trans SNARE complexes across the membrane bilayers. Heterotrimeric cis SNARE complexes have been characterized as highly stable and can be detected by SDS-PAGE. While it is not known whether trans SNARE complexes possess the same intrinsic thermostability as cis complexes, trans complexes would collapse into cis complexes at the time of liposome solubilization in detergent. We analyzed liposome fusion reactions by SDS-PAGE and found that CAPS strongly stimulated the formation of ~165kD complexes that contained each of the three SNAREs (shown for syntaxin-1 in Fig. 4E). These complexes likely represented the accumulation of cis SNARE complexes as a consequence of SNARE-dependent fusion. However, we found that blocking fusion with lysoPC (Fig. 4F) or by holding reactions at 0–4°C (not shown) where fusion fails to occur, reduced but did not eliminate CAPS-dependent formation of ~165kD SNARE complexes. The characteristics of these SNARE complexes, and independent confirmation of their formation (James et al., 2009), indicated that CAPS stimulates the formation of trans SNARE complexes on the liposomes, which is consistent with the CAPS stimulation of fusion that ensues under permissive conditions. It is noteworthy that SNARE complex formation promoted by CAPS is dependent on PI(4,5)P2 and syntaxin-1 binding by CAPS (James et al., 2008; James et al., 2009). For the latter, we demonstrated that a C-terminal portion of CAPS, which contains a Munc-13 homology domain (MHD-1), competed for CAPS binding to t-SNAREs and inhibited CAPS-stimulated SNARE-dependent fusion. In summary, we find that CAPS requires both PI(4,5)P2 binding and SNARE interactions for efficient stimulation of SNARE-dependent membrane fusion.
Herein we summarized a number of key experiments that motivate a model (Fig. 5) for describing the roles of PI(4,5)P2 and CAPS in priming vesicles for exocytosis. Local high concentrations of PI(4,5)P2 on the plasma membrane, which we measure as ≥ 6 mol% (James et al., 2008), may represent microdomains for the organization of vesicle fusion sites, as also proposed by Aoyagi et al. (Aoyagi et al., 2005). Strong interactions between PI(4,5)P2 and the basic linker domain of syntaxin-1 may recruit SNAP-25/syntaxin-1 heterodimers to specific sites on the plasma membrane. PI(4,5)P2 at 1–5 mol% was recently shown to promote disassembly of syntaxin-1 homo-oligomers (Murray & Tamm, 2009), which may drive SNAP-25/syntaxin-1 associations. The high local concentrations of PI(4,5)P2 might serve to inhibit spontaneous vesicle fusion prior to Ca2+-triggered exocytosis. However, neutralization of the basic linker domain of syntaxin-1 was found to enhance the inhibitory effects of PI(4,5)P2 on fusion (James et al., 2008), which suggests that syntaxin-1-mediated sequestration counteracts the inhibitory effects of PI(4,5)P2. It is noteworthy that neutralization of the basic linker domain of syntaxin-1 was recently reported to compromise stimulated secretion in PC12 cells (Lam et al., 2008).
A key feature of the model (Fig. 5) envisions a major effector role for CAPS in the essential functions of PI(4,5)P2 in regulated exocytosis. Previous studies indicated that CAPS activity in promoting vesicle exocytosis was PI(4,5)P2-dependent and that CAPS was recruited to the plasma membrane in part through interactions of its PH domain with PI(4,5)P2 (Grishanin et al., 2002; Grishanin et al., 2004). The current work showed that CAPS was effective in promoting SNARE-dependent liposome fusion only when PI(4,5)P2 was included in the plasma membrane-like, SNAP-25/syntaxin-1-containing acceptor liposomes. Consistent with this, mutagenesis of the CAPS PH domain was found to abrogate its stimulation of SNARE-dependent fusion (James et al., 2008). Thus, a central role for PI(4,5)P2 in vesicle fusion may be to recruit effectors like CAPS to localized fusion sites on the plasma membrane.
The effector function for CAPS in fusion resides outside of its PH domain and involves direct SNARE binding (James et al., 2009). Here we describe the syntaxin-1-binding properties of CAPS and the ability of CAPS to promote trans SNARE complex formation on the liposomes (James et al., 2009). It is noteworthy that CAPS-SNARE interactions do not absolutely require PI(4,5)P2 but do require membrane. However, CAPS stimulation of trans SNARE complex formation and of SNARE-dependent fusion do require PI(4,5)P2. While the precise details of how CAPS promotes trans SNARE complex assembly are under investigation, it seems clear that CAPS is an effector protein that couples membrane PI(4,5)P2 interactions and SNARE interactions into the productive reassembly of SNARE proteins for priming vesicle exocytosis.
Ca2+-triggered vesicle exocytosis in neuro endocrine cells requires priming reactions that follow vesicle tethering/docking and precede triggered fusion. Priming requires PI(4,5)P2 and priming factors, and likely involves SNARE protein complex assembly. In studies with proteoliposomes, the priming factor CAPS interacts with PI(4,5)P2, binds the SNARE protein syntaxin-1, promotes trans SNARE complex formation, and stimulates PI(4,5)P2- and SNARE-dependent liposome fusion. We propose that CAPS functions in priming vesicle exocytosis by coupling membrane binding to SNARE complex assembly.
This work was supported by an NIH grant (DK40428) to T.F.J.M. and by an AHA fellowship to D.J.J.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ca2+-dependent membrane fusion
Regulated vesicle exocytosis