Calcium sensors in regulated exocytosis

Neurotransmitter release, hormone secretion and a variety of other secretory process are tightly regulated with exocytotic fusion of secretory vesicles being triggered by a rise in cytosolic Ca2+ concentration. A series of proteins that act as part of a conserved core machinery for vesicle docking and fusion throughout the cell have been identified. In regulated exocytosis this core machinery must be controlled by Ca(2+)-sensor proteins that allow rapid activation of the fusion process following elevation of cytosolic Ca2+ concentration. The properties of such Ca2+ sensors are known from physiological studies but their molecular identity remains to be unequivocally established. The multiple Ca(2+)-dependent steps in the exocytotic pathway suggest the likely involvement of several Ca(2+)-binding proteins with distinct properties. Functional evidence for the role of various Ca(2+)-binding proteins and their possible sites of action is accumulating but a definitive identification of the major Ca(2+)-sensor in the final step of Ca(2+)-triggered membrane fusion in different cell types awaits further analysis.     

Early requirement for alpha-SNAP and NSF in the secretory cascade in chromaffin cells.

NSF and alpha-SNAP have been shown to be required for SNARE complex disassembly and exocytosis. However, the exact requirement for NSF and alpha-SNAP in vesicular traffic through the secretory pathway remains controversial. We performed a study on the kinetics of exocytosis from bovine chromaffin cells using high time resolution capacitance measurement and electrochemical amperometry, combined with flash photolysis of caged Ca2+ as a fast stimulus. alpha-SNAP, a C-terminal mutant of alpha-SNAP, and NEM were assayed for their effects on secretion kinetics. Two kinetically distinct components of catecholamine release can be observed upon fast step-like elevation of [Ca2+]i. One is the exocytotic burst, thought to represent the readily releasable pool of vesicles. Following the exocytotic burst, secretion proceeds slowly at maintained high [Ca2+]i, which may represent vesicle maturation/recruitment, i.e. some priming steps after docking. alpha-SNAP increased the amplitude of both the exocytotic burst and the slow component but did not change their kinetics, which we examined with millisecond time resolution. In addition, NEM only partially inhibited the slow component without altering the exocytotic burst, fusion kinetics and the rate of endocytosis. These results suggest a role for alpha-SNAP/NSF in priming granules for release at an early step, but not modifying the fusion of readily releasable granules.     

An antisense oligodeoxynucleotide targeted to chromaffin cell scinderin gene decreased scinderin levels and inhibited depolarization-induced cortical F-actin disassembly and exocytosis

Chromaffin cell secretion requires cortical F-actin disassembly and it has been suggested that scinderin, a Ca2+ dependent F-actin severing protein, controls cortical actin dynamics. An antisense oligodeoxynucleotide targeting the scinderin gene was used to decrease the expression of the protein and access its role in secretion. Treatment with 2 microM scinderin antisense oligodeoxynucleotide for 4 days produced a significant decrease in scinderin expression and its mRNA levels. The expression of gelsolin, another F-actin severing protein, was not affected. Scinderin decrease was accompanied by concomitant and parallel decreases in depolarization-evoked cortical F-actin disassembly and exocytosis. Similar treatment with a mismatched oligodeoxynucleotide produced no effects. Scinderin antisense oligodeoxynucleotide treatment was also a very effective inhibitor of exocytosis in digitonin-permeabilized cells stimulated with increasing concentrations of Ca2+. This ruled out scinderin antisense interference with stimulation-induced depolarization or Ca2+ channel activation. Scinderin antisense treatment decreased the maximum (B(max)) secretory response to Ca2+ without modifying the affinity (K(m)) of the cation for the exocytotic machinery. Moreover, the antisense treatment did not affect norepinephrine uptake or the expression of dopamine ss-hydroxylase, suggesting that the number and function of chromaffin vesicles was not modified. In addition, scinderin antisense treatment did not alter the expression of proteins involved in vesicle-plasma membrane fusion, such as synaptophysin, synaptotagmin or syntaxin, indicating a lack of effects on the fusion machinery components. These observations strongly suggest that scinderin is a key player in the events involved in the secretory process.     

Priming in exocytosis: attaining fusion-competence after vesicle docking

Membrane contact established by tethering or docking mechanisms is not a sufficient condition for membrane fusion. In neural and neuroendocrine cells, only a small fraction of secretory vesicles docked at the plasma membrane are fusion-competent and undergo rapid ATP-independent fusion in response to Ca(2+) elevations. Additional biochemical events termed 'priming' are essential to render vesicles competent for Ca(2+)-triggered fusion. The priming of vesicles is ATP-dependent and a number of ATP-dependent priming reactions have been characterized in permeable neuroendocrine cells. These involve NSF-mediated priming of SNARE protein complexes, the ATP-dependent synthesis of phosphoinositides, and protein kinase-mediated protein phosphorylation. In addition, munc13 is an important protein involved in priming synaptic vesicles. An emphasis in this review is on recent work indicating that priming events identified in the pathways of regulated exocytosis share many features with pre-fusion processes characterized in constitutive fusion pathways.     

Mutational analysis of VAMP domains implicated in Ca2+-induced insulin exocytosis.

Vesicle-associated membrane protein-2 (VAMP-2) and cellubrevin are associated with the membrane of insulin-containing secretory granules and of gamma-aminobutyric acid (GABA)-containing synaptic-like vesicles of pancreatic beta-cells. We found that a point mutation in VAMP-2 preventing targeting to synaptic vesicles also impairs the localization on insulin-containing secretory granules, suggesting a similar requirement for vesicular targeting. Tetanus toxin (TeTx) treatment of permeabilized HIT-T15 cells leads to the proteolytic cleavage of VAMP-2 and cellubrevin and causes the inhibition of Ca2+-triggered insulin exocytosis. Transient transfection of HIT-T15 cells with VAMP-1, VAMP-2 or cellubrevin made resistant to the proteolytic action of TeTx by amino acid replacements in the cleavage site restored Ca2+-stimulated secretion. Wild-type VAMP-2, wild-type cellubrevin or a mutant of VAMP-2 resistant to TeTx but not targeted to secretory granules were unable to rescue Ca2+-evoked insulin release. The transmembrane domain and the N-terminal region of VAMP-2 were not essential for the recovery of stimulated exocytosis, but deletions preventing the binding to SNAP-25 and/or to syntaxin I rendered the protein inactive in the reconstitution assay. Mutations of putative phosphorylation sites or of negatively charged amino acids in the SNARE motif recognized by clostridial toxins had no effect on the ability of VAMP-2 to mediate Ca2+-triggered secretion. We conclude that: (i) both VAMP-2 and cellubrevin can participate in the exocytosis of insulin; (ii) the interaction of VAMP-2 with syntaxin and SNAP-25 is required for docking and/or fusion of secretory granules with the plasma membrane; and (iii) the phosphorylation of VAMP-2 is not essential for Ca2+-stimulated insulin exocytosis.     

Membrane association domains in Ca2+-dependent activator protein for secretion mediate plasma membrane and dense-core vesicle binding required for Ca2+-dependent exocytosis

Ca2+-dependent activator protein for secretion (CAPS) is a cytosolic protein essential for the Ca2+-dependent fusion of dense-core vesicles (DCVs) with the plasma membrane and the regulated secretion of a subset of neurotransmitters. The mechanism by which CAPS functions in exocytosis and the means by which it associates with target membranes are unknown. We identified two domains in CAPS with distinct membrane-binding properties that were each essential for CAPS activity in regulated exocytosis. The first of these, a centrally located pleckstrin homology domain, exhibited three properties: charge-based binding to acidic phospholipids, binding to plasma membrane but not DCV membrane, and stereoselective binding to phosphatidylinositol 4,5-bisphosphate. Mutagenesis studies revealed that the former two properties but not the latter were essential for CAPS function. The central pleckstrin homology domain may mediate transient CAPS interactions with the plasma membrane during Ca2+-triggered exocytosis. The second membrane association domain comprising distal C-terminal sequences mediated CAPS targeting to and association with neuroendocrine DCVs. The CAPS C-terminal domain was also essential for optimal activity in regulated exocytosis. The presence of two membrane association domains with distinct binding specificities may enable CAPS to bind both target membranes to facilitate DCV-plasma membrane fusion.     

Bacterial toxins: useful for studying G-proteins implicated in the mechanism of exocytosis in neuroendocrine cells

In neuroendocrine cells, regulated exocytosis is a multistep process that comprises the recruitment and priming of secretory granules, their docking to the exocytotic sites, and the subsequent fusion of granules with the plasma membrane leading to the release of secretory products into the extracellular space. Using bacterial toxins which specially inactivate subsets of G proteins, we were able to demonstrate that both trimeric and monomeric G proteins directly control the late stages of exocytosis in chromaffin cells. Indeed, in secretagogue-stimulated chromaffin cells, the subplasmalemmal actin cytoskeleton undergoes a specific reorganization that is a prerequisite for exocytosis. Our results suggest that a granule-bound trimeric Go protein controls the actin network surrounding secretory granules through a pathway involving the GTPase RhoA and a downstream phosphatidylinositol 4-kinase. Furthermore, the GTPase Cdc42 plays a active role in exocytosis, most likely by providing specific actin structures to the late docking and/or fusion steps. We propose that G proteins tightly control secretion in neuroendocrine cells by coupling the actin cytoskeleton to the sequential steps underlying membrane trafficking at the site of exocytosis. Our data highlight the use of bacterial toxins, which proved to be powerful tools to dissect the exocytotic machinery at the molecular level.     

CAPS acts at a prefusion step in dense-core vesicle exocytosis as a PIP2 binding protein

CAPS-1 is required for Ca2+-triggered fusion of dense-core vesicles with the plasma membrane, but its site of action and mechanism are unknown. We analyzed the kinetics of Ca2+-triggered exocytosis reconstituted in permeable PC12 cells. CAPS-1 increased the initial rate of Ca2+-triggered vesicle exocytosis by acting at a rate-limiting, Ca2+-dependent prefusion step. CAPS-1 activity depended upon prior ATP-dependent priming during which PIP2 synthesis occurs. CAPS-1 activity and binding to the plasma membrane depended upon PIP2. Ca2+ was ineffective in triggering vesicle fusion in the absence of CAPS-1 but instead promoted desensitization to CAPS-1 resulting from decreased plasma membrane PIP2. We conclude that CAPS-1 functions following ATP-dependent priming as a PIP2 binding protein to enhance Ca2+-dependent DCV exocytosis. Essential prefusion steps in dense-core vesicle exocytosis involve sequential ATP-dependent synthesis of PIP2 and the subsequent PIP2-dependent action of CAPS-1. Regulation of PIP2 levels and CAPS-1 activity would control the secretion of neuropeptides and monoaminergic transmitters.     

Exocytosis in neuroendocrine cells: new tasks for actin

Most secretory cells undergoing calcium-regulated exocytosis in response to cell surface receptor stimulation display a dense subplasmalemmal actin network, which is remodeled during the exocytotic process. This review summarizes new insights into the role of the cortical actin cytoskeleton in exocytosis. Many earlier findings support the actin-physical-barrier model whereby transient depolymerization of cortical actin filaments permits vesicles to gain access to their appropriate docking and fusion sites at the plasma membrane. On the other hand, data from our laboratory and others now indicate that actin polymerization also plays a positive role in the exocytotic process. Here, we discuss the potential functions attributed to the actin cytoskeleton at each major step of the exocytotic process, including recruitment, docking and fusion of secretory granules with the plasma membrane. Moreover, we present actin-binding proteins, which are likely to link actin organization to calcium signals along the exocytotic pathway. The results cited in this review are derived primarily from investigations of the adrenal medullary chromaffin cell, a cell model that is since many years a source of information concerning the molecular machinery underlying exocytosis.     

Synaptotagmin VII as a plasma membrane Ca(2+) sensor in exocytosis

Synaptotagmins I and II are Ca(2+) binding proteins of synaptic vesicles essential for fast Ca(2+)-triggered neurotransmitter release. However, central synapses and neuroendocrine cells lacking these synaptotagmins still exhibit Ca(2+)-evoked exocytosis. We now propose that synaptotagmin VII functions as a plasma membrane Ca(2+) sensor in synaptic exocytosis complementary to vesicular synaptotagmins. We show that alternatively spliced forms of synaptotagmin VII are expressed in a developmentally regulated pattern in brain and are concentrated in presynaptic active zones of central synapses. In neuroendocrine PC12 cells, the C(2)A and C(2)B domains of synaptotagmin VII are potent inhibitors of Ca(2+)-dependent exocytosis, but only when they bind Ca(2+). Our data suggest that in synaptic vesicle exocytosis, distinct synaptotagmins function as independent Ca(2+) sensors on the two fusion partners, the plasma membrane (synaptotagmin VII) versus synaptic vesicles (synaptotagmins I and II).     

Regulation of exocytosis in adrenal chromaffin cells: focus on ARF and Rho GTPases

Neurons and neuroendocrine cells release transmitters and hormones by exocytosis, a highly regulated process in which secretory vesicles or granules fuse with the plasma membrane to release their contents in response to a calcium trigger. Several stages have been recognized in exocytosis. After recruitment and docking at the plasma membrane, vesicles/granules enter a priming step, which is then followed by the fusion process. Cortical actin remodelling accompanies the exocytotic reaction, but the links between actin dynamics and trafficking events remain poorly understood. Here, we review the action of Rho and ADP-ribosylation factor (ARF) GTPases within the exocytotic pathway in adrenal chromaffin cells. Rho proteins are well known for their pivotal role in regulating the actin cytoskeleton. ARFs were originally identified as regulators of vesicle transport within cells. The possible interplay between these two families of GTPases and their downstream effectors provides novel insights into the mechanisms that govern exocytosis.     

Fusion pore dynamics are regulated by synaptotagmin*t-SNARE interactions

Exocytosis involves the formation of a fusion pore that connects the lumen of secretory vesicles with the extracellular space. Exocytosis from neurons and neuroendocrine cells is tightly regulated by intracellular [Ca2+] and occurs rapidly, but the molecular events that mediate the opening and subsequent dilation of fusion pores remain to be determined. A putative Ca2+ sensor for release, synaptotagmin I (syt), binds directly to syntaxin and SNAP-25, which are components of a conserved membrane fusion complex. Here, we show that Ca2+-triggered syt*SNAP-25 interactions occur rapidly. The tandem C2 domains of syt cooperate to mediate binding to syntaxin/SNAP-25; lengthening the linker that connects C2A and C2B selectively disrupts this interaction. Expression of the linker mutants in PC12 cells results in graded reductions in the stability of fusion pores. Thus, the final step of Ca2+-triggered exocytosis is regulated, at least in part, by direct contacts between syt and SNAP-25/syntaxin.     

Intracellular enzyme localization in the epithelium of the bovine glandular stomach

The intracellular localization of calcium adenosine triphosphatase (Ca2(+)-ATPase) was studied ultracytochemically in the pyloric glands of the abomasal mucosa of cattle. A remarkable staining pattern exhibited the Golgi apparatus, as there was a gradation in staining of the interior sides of dictyosomal cisternae from the not or weakly stained cis to the heavily stained trans face. Membranes of Golgi-endoplasmic reticulum lysosome complex-secretory vesicles showed either no or strong enzyme activity. Membranes of secretory vesicles accumulated in the cell apex stained positive for ATPase activity. This accounts also for the apical cortical cytoplasm. From these results it is speculated that Ca2(+)-ATPase may play an important role in the pathway of exocytotic secretion, especially in the process of membrane sorting and biogenesis of secretory vesicles, in the steps of vesicle accumulation and transport to the site of exocytosis as well as in membrane fusion events.     

Proteins are secreted by both constitutive and regulated secretory pathways in lactating mouse mammary epithelial cells

Lactating mammary epithelial cells secrete high levels of caseins and other milk proteins. The extent to which protein secretion from these cells occurs in a regulated fashion was examined in experiments on secretory acini isolated from the mammary glands of lactating mice at 10 d postpartum. Protein synthesis and secretion were assayed by following the incorporation or release, respectively, of [35S]methionine-labeled TCA-precipitable protein. The isolated cells incorporated [35S]methionine into protein linearly for at least 5 h with no discernible lag period. In contrast, protein secretion was only detectable after a lag of approximately 1 h, consistent with exocytotic secretion of proteins immediately after passage through the secretory pathway and package into secretory vesicles. The extent of protein secretion was unaffected by the phorbol ester PMA, 8-bromo-cAMP, or 8-bromo-cGMP but was doubled by the Ca2+ ionophore ionomycin. In a pulse-label protocol in which proteins were prelabeled for 1 h before a chase period, constitutive secretion was unaffected by depletion of cytosolic Ca2+ but ionomycin was found to give a twofold stimulation of the secretion of presynthesized protein in a Ca(2+)-dependent manner. Ionomycin was still able to stimulate protein secretion after constitutive secretion had terminated. These results suggest that lactating mammary cells possess both a Ca(2+)-independent constitutive pathway and a Ca(2+)-activated regulatory pathway for protein secretion. The same proteins were secreted by both pathways. No ultrastructural evidence for apocrine secretion was seen in response to ionomycin and so it appears that regulated casein release involves exocytosis. Ionomycin was unlikely to be acting by disassembling the cortical actin network since cytochalasin D did not mimic its effects on secretion. The regulated pathway may be controlled by Ca2+ acting at a late step such as exocytotic membrane fusion.     

cAMP-GEFII is a direct target of cAMP in regulated exocytosis

Although cAMP is well known to regulate exocytosis in many secretory cells, its direct target in the exocytotic machinery is not known. Here we show that cAMP-GEFII, a cAMP sensor, binds to Rim (Rab3-interacting molecule, Rab3 being a small G protein) and to a new isoform, Rim2, both of which are putative regulators of fusion of vesicles to the plasma membrane. We also show that cAMP-GEFII, through its interaction with Rim2, mediates cAMP-induced, Ca2+-dependent secretion that is not blocked by an inhibitor of cAMP-dependent protein kinase (PKA). Accordingly, cAMP-GEFII is a direct target of cAMP in regulated exocytosis and is responsible for cAMP-dependent, PKA-independent exocytosis.     

A role for soluble NSF attachment proteins (SNAPs) in regulated exocytosis in adrenal chromaffin cells.

Digitonin-permeabilized chromaffin cells secrete catecholamines by exocytosis in response to micromolar Ca2+ concentrations, but lose the ability to secrete in response to Ca2+ as the cells lose soluble proteins through the plasma membrane pores. Such secretory run-down can be retarded by cytosolic fractions, thus providing an assay for proteins potentially involved in the exocytotic process. We have used this assay to investigate the role of N-ethylmaleimide-sensitive fusion protein (NSF) and soluble NSF attachment proteins (SNAPs) in regulated exocytosis. Recombinant alpha- and gamma-SNAP stimulated Ca(2+)-dependent exocytosis, although recombinant NSF was ineffective, despite the fact that NSF and alpha-SNAP leak from the permeabilized cells with similar time courses. However, around one third of cellular NSF was found to be present in a non-cytosolic form and so it is possible that this is sufficient for exocytosis and that exogenous SNAPs stimulate the exocytotic mechanism by acting on the leakage-insensitive NSF. The stimulatory effect of alpha-SNAP displayed a biphasic dose-response curve and was maximal at 20 micrograms/ml. The effect of alpha-SNAP was Ca(2+)- and MgATP-dependent and was inhibited by N-ethylmaleimide and botulinum A neurotoxin, indicating a bona fide action on the exocytotic mechanism. Furthermore, Ca2+ concentrations which trigger catecholamine secretion acted to prevent the leakage of NSF and alpha-SNAP from permeabilized cells. These findings provide functional evidence for a role of SNAPs in regulated exocytosis in chromaffin cells.     

Role of actin in regulated exocytosis and compensatory membrane retrieval: insights from an old acquaintance

This review summarizes new insights into the role of the actin cytoskeleton in exocytosis and compensatory membrane retrieval from mammalian regulated secretory cells. Data from our lab and others now indicate that the actin cytoskeleton is involved in exocytosis both as a negative regulator of membrane fusion under resting conditions and as a facilitator of movement of secretory granules to their site of fusion with the apical plasmalemma. Coating of docked secretory granules with actin filaments correlates with the dissociation of secretory-granule-associated rab3D, pointing out a novel role for rab proteins in modulating the actin cytoskeleton during regulated exocytosis. Compensatory membrane retrieval following regulated exocytosis is also critically dependent on the actin cytoskeleton both in initiating the formation of clathrin-coated retrieval vesicles and subsequent trafficking back into the cell. We propose that insertion of secretory granule membrane into the plasmalemma initiates a trigger for membrane retrieval, possibly by exposing sites where proteins involved in compensatory membrane retrieval are assembled. The results summarized in this review were derived primarily from investigations on the pancreatic acinar cell, an old friend who is providing modern wisdom not attainable in other simpler systems.     

In vitro reconstitution of exocytosis from plasma membrane and isolated secretory vesicles

We describe the reconstitution of exocytotic function through recombination of purified cortical secretory vesicles (CVs) and plasma membrane from sea urchin eggs. CVs were dislodged from a cell surface complex preparation by gentle homogenization in an isotonic dissociation buffer, and purified by differential centrifugation. CV-free plasma membrane fragments were obtained by mechanically dislodging CVs from cortical lawn (CL) preparations with a jet of CL isolation buffer. This procedure produced a "plasma membrane lawn" preparation, consisting of plasma membrane fragments attached via their vitelline layer (an extracellular glycocalyx) to a polylysine-coated microscope slide. When freshly prepared CVs were incubated with plasma membrane lawns, CVs reassociated with the cytoplasmic face of the plasma membrane, forming an exocytotically competent, reconstituted cortical lawn (RL). Exocytosis in RLs was monitored by phase-contrast microscopy, and quantitated with a sensitive microphotometric assay. Half-maximal exocytosis in RLs occurred at 18.5 microM free Ca2+; half-maximal exocytosis in control lawns occurred at 5.7 microM free Ca2+. Greater than 90% of the purified CVs that were not attached to a plasma membrane lawn remained intact when bathed in a buffer containing millimolar Ca2+. This result excluded the possibility that Ca2+-triggered CV lysis was responsible for our observations, and confirmed that the association of CVs with the plasma membrane was required for exocytosis in RLs. Evidence that the Ca2+-stimulated release of CV contents in CLs and RLs is the in vitro equivalent of exocytosis was obtained with an immunofluorescence-based vectorial transport assay, using an antiserum directed against a CV content protein: stimulation of RLs or partially CV-depleted CLs with Ca2+ resulted in fusion of the CV and plasma membranes, and the vectorial transport of CV contents from the cytoplasmic to the extracytoplasmic face of the egg plasma membrane.     

Regulation of transmitter release by Unc-13 and its homologues

Neurotransmitters are released by Ca(2+)-triggered exocytotic fusion of synaptic vesicles. Before fusion, vesicles dock at a specialised presynaptic plasma membrane region, the active zone, where they are primed to a fusion competent state. The nature of this priming reaction has long been enigmatic. Recent evidence demonstrates that priming is an essential and rate-limiting step in secretion from neurons and neuroendocrine cells. Members of the Unc-13 protein family, which are highly conserved during evolution and act as novel targets of the diacylglycerol second-messenger pathway, have been identified to play an essential role in this process.     

Calcium-triggered membrane fusion proceeds independently of specific presynaptic proteins

Complexes of specific presynaptic proteins have been hypothesized to drive or catalyze the membrane fusion steps of exocytosis. Here we use a stage-specific preparation to test the roles of SNAREs, synaptotagmin, and SNARE-binding proteins in the mechanism of Ca2+-triggered membrane fusion. Excess exogenous proteins, sufficient to block SNARE interactions, did not inhibit either the Ca2+ sensitivity, extent, or kinetics of fusion. In contrast, despite a limited effect on SNARE and synaptotagmin densities, treatments with high doses of chymotrypsin markedly inhibited fusion. Conversely, low doses of chymotrypsin had no effect on the Ca2+ sensitivity or extent of fusion but did alter the kinetic profile, indicating a more direct involvement of other proteins in the triggered fusion pathway. SNAREs, synaptotagmin, and their immediate binding partners are critical to exocytosis at a stage other than membrane fusion, although they may still influence the triggered steps.