Stimulus–Secretion Coupling In Excitable Cells: A Central Role For Calcium

The release of neurotransmitters and hormones by exocytosis is probably the most extensively studied process for establishing the crucial role of Ca²⁺ as the triggering and controlling event. However, although the concept of Ca²⁺ control of exocytosis dates back over 20 years (Douglas, 1968), only recently developed methodological approaches have provided a proper quantitative evaluation of the role Ca²⁺ plays in this process. For example, in adrenal chromaffin cells, the use of video-imaging techniques to visualise the stimulus-induced changes in the concentration of cytosolic free Ca²⁺ ([Ca²⁺]_i) in Fura-2-loaded cells (see reviews by Burgoyne, 1991; Cheek, 1991), and of whole-cell patch-clamp (Augustine and Neher, 1992a; Neher and Augustine, 1992) coupled with flash photolysis (Neher and Zucker, 1993) techniques to manipulate [Ca²⁺]_i directly at the single-cell level, have indicated that the entry of external Ca²⁺, rather than the release of Ca²⁺ from intracellular stores, is the vital trigger for secretion from these cells, and that this is likely to be because only entry of Ca²⁺ from the outside medium delivers Ca²⁺ in sufficient amounts to the subplasmalemmal exocytotic sites.

The mechanisms by which local large increases in [Ca²⁺]_i beneath the plasma membrane activate exocytosis and the nature of the components of the exocytotic machinery are now beginning to be resolved, but we still only have limited insight into these aspects. It seems likely that the intracellular Ca²⁺ receptor mechanism and the exocytotic fusion machinery may involve several cytosolic proteins and membrane proteins, perhaps with varying Ca²⁺ affinities, acting in concert (Burgoyne and Morgan, 1993).

This chapter will deal with studies on changes in [Ca²⁺]_i related to exocytosis in electrically excitable cells, and the nature of the Ca²⁺ signalling machinery (particularly Ca²⁺ channels) involved in the triggering of this process. Considerable emphasis will be given to studies using bovine adrenal chromaffin cells and their clonal counterpart, rat pheochromocytoma PC12 cells, but reference will also be made to relevant studies using other excitable cells.

The classical physiological stimulus for catecholamine secretion from bovine adrenal chromaffin cells is acetylcholine acting through cholinergic nicotinic receptors. During the 1980s, however, it became clear that activation of a number of other cell surface receptors either stimulates catecholamine secretion or modulates nicotine-induced secretion (Marley, 1987). Receptor activation in chromaffin cells results either in depolarisation and the opening of voltage-dependent channels in the plasma membrane or in direct activation of phosphoinositidase C, with subsequent generation of the Ca²⁺-mobilising signal inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃]. Nicotine and γ-aminobutyric acid (GABA) lead to depolarisation, whereas muscarinic agonists, bradykinin, angiotensin II, histamine, prostaglandin E₂, vasoactive intestinal peptide (VIP) and ATP receptors are all linked to Ins(1,4,5)P₃ production via activation of phophoinositidase C. These signalling pathways are not necessarily mutually exclusive because nicotinic stimulation and depolarisation with high [K⁺] have been reported to lead to Ins(1,4,5)P₃ production through Ca²⁺-dependent activation of phosphoinositidase C (Eberhard and Holz, 1988), and Ins(1,4,5)P₃ may promote Ca²⁺ entry through non-voltage-dependent Ca²⁺ channels in the plasma membrane (Mochizuki-Oda et al. 1991).

The various receptor agonists differ in their abilities to trigger secretion, although nicotinic stimulation consistently gives the largest response. Apart from nicotinic agonists, only GABA and (after a longer time course) histamine produce a substantial secretory response. Reduced, or indeed no, secretion has been reported in response to the remaining agonists. Irrespective of the nature of the stimulating agent, secretion from intact cells is abolished by the removal of external Ca²⁺ (see Burgoyne, 1991), a simple demonstration that Ca²⁺ entry and not the release of internally stored Ca²⁺ is pivotal in triggering exocytosis from these cells. Indeed, it is by reducing Ca²⁺ entry that D₂ dopamine receptors are thought to exert an inhibitory control on chromaffin cell secretion (Bigornia et al. 1988). The importance of Ca²⁺ entry versus internal Ca²⁺ release in triggering secretion is discussed in more detail in the following section.

Many laboratories have used Ca²⁺-sensitive fluorescent indicator dyes in intact cells to follow the stimulus-induced changes in [Ca²⁺]_i in chromaffin cell populations (e.g. Knight and Kesteven, 1983; Kao and Schneider, 1985; O’Sullivan and Burgoyne, 1989; Stauderman and Pruss, 1989) or single cells (O’Sullivan et al. 1989; Cheek et al. 1989a,b; Kim and Westhead, 1989; Stauderman et al. 1990; Ito et al. 1991). From these studies, resting [Ca²⁺]_i in the chromaffin cell appears to be in the range 10–100nmol l⁻¹and, after stimulation, this rises to 300–1000nmol l⁻¹. An assessment of the relationship between the Ca²⁺ response and the triggering of secretion has resulted in the consensus that the major signal for secretion is a rise in [Ca²⁺]_i due to Ca²⁺ entry from the extracellular medium (Cheek et al. 1989a,b; Bunn and Marley, 1989; Kim and Westhead, 1989; Ballesta et al. 1989; Stauderman et al. 1990; Yamagami et al. 1991). The key observation supporting this being that agents which release internally stored Ca²⁺, such as Ins(1,4,5)P₃-mobilising hormones or less-specific Ca²⁺-mobilising compounds such as caffeine and thapsigargin, all fail to promote secretion in the absence of external Ca²⁺. Thus, the efficacy of a given agonist reflects the degree to which it is able to promote Ca²⁺ entry through either voltage-dependent or voltage-independent pathways rather than its ability to elevate [Ca²⁺]_iper se. By using video-imaging techniques to visualise stimulus-induced changes in [Ca²⁺]_i (Burgoyne, 1991; Cheek, 1991), and whole-cell patch-clamp (Augustine and Neher, 1992a; Neher and Augustine, 1992) and flash photolysis techniques (Neher and Zucker, 1993) to manipulate [Ca²⁺]_i directly at the single-cell level, considerable light has been shed on why Ca²⁺ influx is so crucial for secretion: it seems likely that only entry of Ca²⁺ from the external medium delivers Ca²⁺ in sufficient amounts to the subplasmalemmal exocytotic sites to activate the fusion process (Fig. 1). The difference between Ca²⁺ entry and Ca²⁺ mobilisation, as far as exocytosis is concerned, is due to the differing Ca²⁺ gradients generated, which can lead to a similar mean [Ca²⁺]_i within the cell but to a quite different [Ca²⁺]_i at the plasma membrane.

Nicotinic or GABA stimulation of chromaffin cells results in membrane depolarisation: nicotine via Na⁺ influx (Kilpatrick et al. 1982) and GABA via Cl⁻efflux (Bormann and Clapham, 1985). As a result, Ca²⁺ enters the cell from the outside medium through voltage-dependent Ca²⁺ channels, or via the acetylcholine receptor itself in the case of nicotine (Holz et al. 1982; Kilpatrick et al. 1982). Chromaffin cells possess both L-type dihydropyridine-sensitive Ca²⁺ channels (Artalejo et al. 1987; Cena et al. 1989) and dihydropyridine-insensitive ω-conotoxin-sensitive Ca²⁺ channels (Rosario et al. 1989) as well as other pharmacologically distinct Ca²⁺ channels (Artalejo et al. 1991). L-type channels are not the only channels involved in voltage-dependent Ca²⁺ entry leading to secretion because dihydropyridine antagonists do not completely block the rise in [Ca²⁺]_i or the secretion elicited by either nicotine or GABA (Kataoka et al. 1988; Kitayama et al. 1990). Electrophysiological studies on single-channel kinetics have shown that chromaffin cells also possess a class of dihydropyridine-sensitive Ca²⁺ channel that is recruited in response to depolarisation to form a facilitation current whose role in exocytosis has yet to be determined (Artalejo et al. 1991).

Spatial aspects of the Ca²⁺ signal in single bovine chromaffin cells have been determined using video-imaging of Fura-2-loaded intact cells (O’Sullivan et al. 1989; Cheek et al. 1989a,b, 1990; Burgoyne et al. 1989; Ito et al. 1991) and Fura-2 in whole-cell patch-clamp recording (Neher and Augustine, 1992). Following stimulation with nicotine or electrical depolarisation, the rise in [Ca²⁺]_i due to Ca²⁺ influx was initially restricted to the sub-plasmalemmal region of the cell (O’Sullivan et al. 1989; Cheek et al. 1989b; Neher and Augustine, 1992). An increase in [Ca²⁺]_i then spread into the interior of the cell to give a more uniform elevation of [Ca²⁺]_i throughout the cytoplasm. Subsequently, a further increase in [Ca²⁺]_i occurred that was sometimes localised to one area of the cell (O’Sullivan et al. 1989). This latter response may be due to the release of Ca²⁺ from internal stores (see below). The early (<1s) subplasmalemmal rise in [Ca²⁺]_i was calculated to be 40–50nmol l⁻¹above the resting level (O’Sullivan et al. 1989). The temporal resolution of the imaging system used (5ratioimages s⁻¹: see Moreton, 1991) is such that a Ca²⁺ wave spreading at a rate of about 10 μms⁻¹has to diffuse about 2 μm from its point source (the channel in the plasma membrane) before it can be detected. As Ca²⁺ spreads this distance from its point source, its concentration has been predicted to decay exponentially (Fig. 2) from a concentration in the vicinity of the mouth of the Ca²⁺ channel that is probably in the range 10–100 μmol l⁻¹(Smith and Augustine, 1988). Recent studies on chromaffin cells using whole-cell-attached patches have more directly indicated that these high levels of Ca²⁺ can exist at the sites of exocytosis in stimulated cells: when raising [Ca²⁺]_i by dialysis, even 100 μmol l⁻¹[Ca²⁺] did not achieve the same rates of secretion as those seen in response to depolarisation, as judged by measuring the change in capacitance across the plasma membrane (Augustine and Neher, 1992a). Unfortunately, such spatially restricted high levels of [Ca²⁺]_i cannot reliably be visualised with current imaging technology, not only because of the limited temporal and spatial resolution of the imaging systems but also because the majority of dyes currently available to monitor [Ca²⁺]_i saturate in the range 1–2 μmoll⁻¹.

The voltage-operated Ca²⁺ channels responsible for delivering this high [Ca²⁺]_i to the sites of exocytosis are probably located over the entire cell surface: by visualising the sites of exocytosis using an antibody raised against the chromaffin granule membrane protein dopamine-β-hydroxylase, it has been shown that nicotine-induced depolarisation triggers a secretory response that originates from the entire surface of the cell (Cheek et al. 1989a).

An interesting variation on this theme appears to exist in fast synapses, where Ca²⁺ entry through voltage-dependent channels is also the key trigger for exocytosis. Various experimental approaches have led to the idea that rapid physiological transmitter release may occur only from synaptic vesicles very close to Ca²⁺ channels (Augustine et al. 1991). Recent work using a low-affinity form of the Ca²⁺-sensitive photoprotein aequorin has shown the presence of microdomains of Ca²⁺ in the squid giant synapse where [Ca²⁺]_i can reach 200–300 μmol l⁻¹during stimulation (Llinas et al. 1992), which may explain why exocytosis can occur within a few hundred microseconds in these very specialised cells. Voltage-dependent Ca²⁺ entry may also be localised in small secretory cells such as rat pituitary gonadotropes (Rawlings et al. 1991), but whether this triggers localised exocytotic secretion is not known.

A secondary rise in [Ca²⁺]_i, that follows Ca²⁺ influx, is observed on continued depolarisation of chromaffin cells (O’Sullivan et al. 1989), but not after stimulation with a single action potential (Neher and Augustine, 1992). An interesting possibility is that this elevation in [Ca²⁺]_i is due to Ca²⁺ release from internal stores that only become sensitised following the greater Ca²⁺ influx (Hua et al. 1993) resulting from activation of the facilitation Ca²⁺ current, because this current only occurs with repetitive depolarisation (see above; Artalejo et al. 1991). Ca²⁺ mobilisation could occur either from stores that are sensitive to Ins(1,4,5)P₃ or by a process of Ca²⁺-induced Ca²⁺ release (CICR) from stores containing caffeine-sensitive ryanodine receptors. In favour of the first mechanism, it has been shown that both nicotine and high [K⁺] result in the generation of Ins(1,4,5)P₃ in bovine chromaffin cells due to Ca²⁺-dependent activation of phosphoinositidase C (Eberhard and Holz, 1987, 1988). However, a process of CICR similar to that characterised in muscle cells has been proposed to occur in non-muscle cells (Berridge, 1993), and this could be responsible for the secondary rise in [Ca²⁺]_i. CICR in muscle cells is mediated by a ryanodine receptor that is caffeine-sensitive.

Chromaffin cells contain caffeine-sensitive Ca²⁺ stores (Liu et al. 1991) evenly distributed throughout the cytoplasm (Burgoyne et al. 1989; Cheek et al. 1990) and these are, at least in part, Ins(1,4,5)P₃-insensitive (Robinson and Burgoyne, 1991; Cheek et al. 1991a; Stauderman et al. 1991). Several lines of evidence indicate that these stores may contribute to the Ca²⁺ signal. Emptying caffeine-sensitive stores reduces the magnitude of the histamine-induced rise in [Ca²⁺]_i in bovine chromaffin cells (Stauderman and Murawsky, 1991), and the sensitivity of the store to caffeine has been reported to be regulated by the level of luminal free Ca²⁺, as is the corresponding store in muscle cells (Cheek et al. 1991a, 1993a). In rat chromaffin cells, spontaneous fluctuations in [Ca²⁺]_i were generated by an Ins(1,4,5)P₃-insensitive but caffeine-sensitive store (Malgaroli et al. 1990). Caffeine is unable to trigger significant secretion from intact bovine chromaffin cells, but the state of filling of the store has been shown to influence the decay time of a depolarisation-induced Ca²⁺ signal (Cheek et al. 1990). The role of the caffeine-sensitive store in Ca²⁺ signalling and bovine chromaffin cell function still has to be fully elucidated but it may be that the store is involved more in Ca²⁺ homeostasis than in the direct triggering of exocytosis. Whether the putative messenger molecule cyclic ADP–ribose (Galione, 1993) is able to promote Ca²⁺ release from the caffeine-sensitive store in chromaffin cells has yet to be investigated.

Muscarinic agonists raise [Ca²⁺]_i in chromaffin cells by stimulating release from internal stores (Kao and Schneider, 1985; Cheek and Burgoyne, 1985), but this does not promote secretion from bovine cells (Fisher et al. 1981; Cheek and Burgoyne, 1985; Ballesta et al. 1989; Yamagami et al. 1991). Within a population, not all chromaffin cells respond to muscarinic receptor activation and, in those that do, the early subplasmalemmal rise in [Ca²⁺]_i seen in response to depolarisation (Fig. 1) does not occur. Instead, the rise in [Ca²⁺]_i initiates in a more localised region (arrow, Fig. 3C) that appears to co-localise with a particular domain of the endoplasmic reticulum/nuclear envelope organelle (asterisk, Fig. 3A). One possibility is that the local area of initiation of the Ca²⁺ signal represents a subdomain of the organelle containing the most sensitive Ins(1,4,5)P₃ receptors, as has been suggested in exocrine secretory cells (Thorn et al. 1993). It is likely that in intact cells such a local release of Ca²⁺ does not result in a sufficient elevation in [Ca²⁺]_i at the exocytotic sites to activate the fusion process, hence the lack of secretory response. This idea is supported by the findings that non-specific release of Ca²⁺ from the Ins(1,4,5)P₃-sensitive store in response to thapsigargin was also incapable of triggering release, unless there was a secondary influx of external Ca²⁺ (Cheek and Thastrup, 1989), and that caffeine, while considerably elevating average [Ca²⁺]_i by releasing internal stores, was also ineffective at triggering secretion (Cheek et al. 1990). All these data indicate that the optimal Ca²⁺ signal required for secretion is a large (>10 μmol l⁻¹) rise in [Ca²⁺]_i immediately beneath the plasma membrane and that, in intact cells, this is most effectively achieved by Ca²⁺ entry and not by Ca²⁺ release from internal stores. Interestingly, in whole-cell-attached patched cells, elevating [Ca²⁺]_i to 10 μmol l⁻¹by releasing Ca²⁺ from intracellular stores has been found to trigger secretion (Augustine and Neher, 1992). Presumably under experimental conditions in which the Ca²⁺-buffering capacity of the cell is reduced, sufficient Ca²⁺ is able to reach the plasma membrane from the internal stores to activate the exocytotic sites.

After initiation in a localised region of the cell (see Fig. 3), the Ca²⁺ signal in response to the Ins(1,4,5)P₃-mobilising stimulus histamine has been found to spread across the cell in the form of a wave (Fig. 4), in a manner similar to that reported for other cell types (Berridge, 1993). As mentioned above, pretreatment of cells with caffeine attenuated the histamine-induced Ca²⁺ response (Stauderman and Murawsky, 1991), implicating the caffeine-sensitive store in mediating the propagation phase of the signal, perhaps by a process of CICR as already described. It should be noted, however, that caffeine has subsequently been reported to be able to inhibit Ins(1,4,5)P₃-mediated Ca²⁺ release in certain tissues (Brown et al. 1992), so whether the Ins(1,4,5)P₃-sensitive store or the caffeine-sensitive store is responsible for propagating the Ca²⁺ wave in chromaffin cells remains to be clarified.

Histamine and angiotensin II both trigger secretion that depends upon external Ca²⁺ (Noble et al. 1988; O’Sullivan et al. 1989; Stauderman et al. 1990), with histamine being much more potent than angiotensin II. This may be because the angiotensin receptor, but not the histamine receptor, desensitises rapidly, thereby shutting off Ca²⁺ entry (Stauderman et al. 1990). Additionally, the localisation of Ca²⁺ entry in response to the two stimuli may differ. Angiotensin II causes an initial local rise in [Ca²⁺]_i and a polarised entry of external Ca²⁺, as judged by visualising the quenching of Fura-2 fluorescence by the Ca²⁺ surrogate Mn²⁺ (Cheek et al. 1989a, 1993b). This results in a low level of catecholamine release (O’Sullivan and Burgoyne, 1989; Bunn and Marley, 1989) that is also polarised (Cheek et al. 1989a). The more potent secretagogue histamine, at concentrations optimal for secretion, does not open voltage-dependent Ca²⁺ channels (Artalejo et al. 1993), but nevertheless stimulates Ca²⁺ entry around the whole plasma membrane, explaining its efficacy in stimulating secretion (Cheek et al. 1993b).

The nature of the channels through which angiotensin II and histamine stimulate Ca²⁺ entry and the mechanism of their opening are unknown. Mn²⁺ entry has been reported to be activated by Ca²⁺-ATPase inhibitors such as thapsigargin in the absence of surface receptor stimulation (Robinson et al. 1992), indicating the existence of a capacitative entry mechanism in these cells (Putney, 1990). Whether such a mechanism is responsible for all of the entry observed after receptor activation, however, remains to be elucidated. Ca²⁺ entry in response to angiotensin II was observed after prior depletion of the internal Ca²⁺ store with ionomycin, suggesting that, even with depleted stores, entry could still be stimulated by receptor activation (Stauderman and Pruss, 1989). One possibility is that hormones directly open a receptor-operated channel (ROC) in the plasma membrane, as does ADP in platelets (Sage et al. 1989) and ATP in PC12 cells (V. A. Barry and T. R. Cheek, in preparation). An alternative possibility is that an intracellular messenger, such as Ins(1,4,5)P₃ (Kuno and Gardner, 1987) or inositol 1,4,5-tetrakisphosphate [Ins(1,3,4,5)P₄] (Irvine, 1992), may promote Ca²⁺ entry through a second-messenger-operated channel. Angiotensin II and histamine both result in a transient generation of Ins(1,4,5)P₃ within 10s in chromaffin cells. The response to histamine, but not to angiotensin II, was biphasic with a secondary generation of Ins(1,4,5)P₃ occurring at 20 s (Stauderman and Pruss, 1990). The secondary increase in Ins(1,4,5)P₃ in response to histamine was paralleled by an increase in Ins(1,3,4)P₃ that was absent after stimulation with angiotensin II. This Ins(1,3,4)P₃ could only have originated from dephosphorylation of Ins(1,3,4,5)P₄, suggesting that histamine produces more Ins(1,3,4,5)P₄, as well as more Ins(1,4,5)P₃, than does angiotensin II. With respect to eliciting a secretory response, it may be significant that histamine is more potent than angiotensin II in generating both Ins(1,4,5)P₃ and Ins(1,3,4,5)P₄ because both of these inositol phosphates have been proposed to trigger Ca²⁺ entry into cells (see Irvine, 1992). There is as yet no direct evidence supporting a role for Ins(1,3,4,5)P₄ in gating Ca²⁺ entry into chromaffin cells, but Mochizuki-Oda et al. (1991) have demonstrated that Ins(1,4,5)P₃ is capable of directly opening a plasma membrane channel that gates Ca²⁺ in these cells. With reference to this last point, a recent study has indicated that chromaffin granules may be Ins(1,4,5)P₃-sensitive (Yoo and Albanesi, 1990). Exocytosis of these granules, with the resulting incorporation of the Ins(1,4,5)P₃ receptor into the cytosolic surface of the plasma membrane, may be a means of targeting the Ins(1,4,5)P₃ receptor to the plasma membrane.

Other possible intracellular messenger activators of plasma membrane Ca²⁺ channels, yet to be explored in chromaffin cells, include Ca²⁺ itself (von Tscharner et al. 1986) and GTP (Mullaney et al. 1988).

A remarkable demonstration of the physiological importance of targeting the elevation in [Ca²⁺]_i to the strategic area of the cell responsible for exocytosis is shown in Fig. 5. Video-imaging experiments using PC12 cells have revealed that, in the absence of extracellular Ca²⁺, approximately 40% of single cells responded to extracellular ATP with a transient elevation in [Ca²⁺]_i and that sometimes this was in the form of an initially localised Ca²⁺ signal propagated across the cell in the form of a wave (Fig. 5A). Conversely, in the presence of extracellular Ca²⁺, more than 90% of single cells responded with a transient elevation in [Ca²⁺]_i to an application of ATP, with Ca²⁺ often initially localised to the subplasmalemmal region before infilling into the cell (Fig. 5B). Further experiments using the Ca²⁺ surrogate Mn²⁺, and the nucleotide UTP to desensitise the release of intracellularly stored Ca²⁺, have indicated that this subplasmalemmal elevation in [Ca²⁺]_i is due to ATP triggering Ca²⁺ entry independently of intracellular Ca²⁺ mobilisation (V. A. Barry and T. R. Cheek, in preparation), presumably through a putative receptor-operated channel (Sela et al. 1991).

Interestingly, [³H]dopamine secretion from populations of PC12 cells was only triggered by ATP in the presence of external Ca²⁺ (Fig. 5C), indicating that a subplasmalemmal elevation in [Ca²⁺]_i, and not a release of intracellularly stored Ca²⁺ in the form of a wave, is most effective at activating exocytotic fusion. These experiments indicate that the same stimulus can deliver Ca²⁺ signals with different temporal and spatial profiles into the same cell type by activating different signalling pathways and, most importantly, that these Ca²⁺ signals can be functionally distinct.

The main conclusion to emerge from studies in which stimulus-induced changes in [Ca²⁺]_i in chromaffin cells have been monitored at the single-cell level is that Ca²⁺ entry is essential for the activation of exocytosis, and this is likely to be because only entry delivers Ca²⁺ in sufficient amounts to the subplasmalemmal exocytotic sites (Cheek et al. 1989a,b, 1991b, 1993b; Neher and Augustine, 1992; Augustine and Neher, 1992b; Neher and Zucker, 1993).

As already discussed, because of technical limitations, the [Ca²⁺]_i existing within a few nanometres of exocytotic sites cannot yet be directly visualised (see also Augustine and Neher, 1992b), but both theoretical calculations (Smith and Augustine, 1988) and patch-clamp experiments (Neher and Augustine, 1992; Augustine and Neher, 1992a) indicate that such intracellular Ca²⁺ concentrations probably lies in the range 10–100 μmol l⁻¹before dissipating very rapidly (approximately 100–200ms) into the cytosol. Using the technique of flash photolysis and caged Ca²⁺, Neher and Zucker (1993) have elevated average [Ca²⁺]_i in chromaffin cells to these predicted levels while simultaneously monitoring secretion by measuring plasma membrane capacitance. Elevating [Ca²⁺] to above 10 μmol l⁻¹was found to trigger ultrafast (<50ms) exocytosis, possibly resulting from the release of granules already docked at the plasma membrane, while elevating [Ca²⁺] to the range 70–150 μmol l⁻¹was found, in addition, to trigger both a fast (1–2s) and a slow (10–30s) exocytotic response, perhaps representing the release of granules restrained within the subplasmalemmal cytoskeleton or recruited from deep within the cell. Similar findings have been made for pituitary melanotrophs (Thomas et al. 1993).

Clearly, experiments such as these, in which [Ca²⁺]_i can be changed rapidly and uniformly in a known manner, will be central to unravelling the multiplicity of Ca²⁺-dependent processes that are related to the secretory response in chromaffin and other excitable cell types.

Grateful thanks are due to Mike Berridge, Roger Moreton, Bob Burgoyne, Alan Morgan and Tony O’Sullivan for their contributions to the studies outlined here and to Gordon Koch (MRC Centre, Hills Road, Cambridge, UK) for the gift of anti-endoplasmin antibody. V.A.B. is in receipt of an MRC Studentship. T.R.C. is a Royal Society University Research Fellow.