CaM kinase II-dependent mobilization of secretory granules underlies acetylcholine-induced stimulation of exocytosis in mouse pancreatic B-cells

Abstract

1. Measurements of cell capacitance were used to investigate the mechanisms by which acetylcholine (ACh) stimulates Ca²⁺-induced exocytosis in single insulin-secreting mouse pancreatic B-cells.

2. ACh (250 μM) increased exocytotic responses elicited by voltage-clamp depolarizations 2.3-fold. This effect was mediated by activation of muscarinic receptors and dependent on elevation of the cytoplasmic Ca²⁺ concentration ([Ca²⁺]_i) attributable to mobilization of Ca²⁺ from intracellular stores. The latter action involved interference with the buffering of [Ca²⁺]_i and the time constant (τ) for the recovery of [Ca²⁺]_i following a voltage-clamp depolarization increased 5-fold. As a result, Ca²⁺ was present at concentrations sufficient to promote the replenishment of the readily releasable pool of granules (RRP; > 0.2 μM) for much longer periods in the presence than in the absence of the agonist.

3. The effect of Ca²⁺ on exocytosis was mediated by activation of CaM kinase II, but not protein kinase C, and involved both an increased size of the RRP from 40 to 140 granules and a decrease in τ for the refilling of the RRP from 31 to 19 s.

4. Collectively, the effects of ACh on the RRP and τ result in a > 10-fold stimulation of the rate at which granules are supplied for release.

Acetylcholine is the classical neurotransmitter of the parasympathetic nervous system and activation of cholinergic nerves during feeding is important for nutrient-induced insulin secretion in vivo (see review by Rasmussen et al. 1990). Acetylcholine (ACh) is released within the islets of Langerhans and potentiates glucose-stimulated insulin secretion (Zawalich et al. 1989). Its actions on the B-cell are mediated by muscarinic receptors of the M₃-subtype (Henquin & Nenquin, 1988) and are believed to involve several mechanisms. Firstly, it stimulates phosphoinositide breakdown leading to the production of inositol 1,4,5-trisphosphate (InsP₃) and mobilization of Ca²⁺ from intracellular stores (Biden et al. 1987). Secondly, the associated production of diacylglycerol activates protein kinase C (PKC) and thus enhances Ca²⁺-induced secretion (Weng et al. 1993; Ämmäläet al. 1994). Thirdly, ACh increases Na⁺ influx through the plasma membrane resulting in membrane depolarization (Gilon & Henquin, 1993), which results in the opening of voltage-dependent Ca²⁺ channels and culminates in the initiation of Ca²⁺-dependent secretion.

The B-cell contains about 13000 secretory granules (Dean, 1973) of which only a fraction are accessible for release during stimulation (Eliasson et al. 1997). Studies on both chromaffin and pituitary cells have indicated that the granules pass a series of functional states before undergoing exocytosis and a similar situation exists in the B-cell. The bulk of granules (> 95 %) thus belong to a reserve pool and are not immediately available for release (Neher & Zucker, 1993). In the pancreatic B-cell, the number of granules that belong to the readily releasable pool (RRP) is small and has been estimated as 100 granules per cell (Eliasson et al. 1997). The process by which granules are transferred from the reserve pool into the RRP is referred to as mobilization. The B-cell may represent a suitable system for investigating the mechanisms controlling granule mobilization given the smallness of its RRP. The regulation of the transfer of granules between the reserve pool and the RRP remains obscure but there is evidence suggesting the involvement of cytoskeletal components (Li et al. 1994).

Studies in a variety of (neuro)endocrine cells, including chromaffin cells and pancreatic B-cells, have indicated that mobilization requires lower cytoplasmic free Ca²⁺ ([Ca²⁺]_i) levels than those required for exocytosis (Neher & Zucker, 1993; von Rüden & Neher, 1993; Renström et al. 1997) and that it is fuelled by hydrolysis of Mg-ATP (Parsons et al. 1995; Eliasson et al. 1997). We demonstrate here, using high-resolution capacitance measurements of exocytosis, that ACh promotes exocytosis in mouse B-cells. This results from mobilization of intracellular Ca²⁺ from the endoplasmic reticulum and increased availability of secretory granules for release. This mobilization of secretory vesicles from a reserve pool to the RRP was observed at [Ca²⁺]_i only slightly higher than the resting concentration (300 nM as compared with 200 nM) and involved activation of calmodulin-dependent kinase II (CaM kinase II).

METHODS

Preparation and culture of mouse B-cells

Pancreatic B-cells isolated from NMRI mice (Bomholtgård, Ry, Denmark) were used throughout this study. Briefly, the mice were stunned by a blow against the head and killed by cervical dislocation. The pancreas was quickly removed and pancreatic islets were isolated by collagenase (3 mg ml⁻¹, type XI from Sigma) digestion. The islets were dispersed into single cells by shaking in a Ca²⁺-free solution and the resulting cell suspension was plated on Nunc Petri dishes and maintained for up to 6 days in RPMI 1640 tissue-culture medium (Gibco BRL, Life Technologies Ltd, Paisley, UK) supplemented with 10 % (v/v) heat-inactivated fetal calf serum, 100 i.u. ml⁻¹ penicillin, 100 μg ml⁻¹ streptomycin; no changes of the exocytotic and electrophysiological properties were observed during the period of culture.

Electrophysiology

Patch pipettes were pulled from borosilicate glass (tip resistance 3-4 MΩ when filled with the pipette solution), coated with Sylgard and fire-polished before use. The zero-current potential was adjusted before establishment of the seal with the pipette in the bath. The holding potential in all experiments was -70 mV.

Exocytosis was monitored in single B-cells as changes in cell membrane capacitance using either the standard or the perforated-patch whole-cell configuration. An EPC-7 patch-clamp amplifier (List Electronic) was used and exocytosis was elicited by voltage-clamp depolarizations. Unless otherwise indicated, the depolarizations were 500 ms long and went from -70 to 0 mV. Changes in cell capacitance were detected using in-house software written in Axobasic (Axon Instruments). Briefly, a 28 mV peak-to-peak 800 Hz sine wave was added to the holding potential (-70 mV) and 10 cycles were averaged for each data point. The resulting current was analysed at two orthogonal phase angles with a resolution of 100 ms per point. The phase angle was determined before each depolarization by varying the series conductance (G_series) and cell capacitance (C_slow) settings of the patch-clamp amplifier until a change in G_series did not influence the measured C_slow. During the experiments the cells were situated in an experimental chamber with a volume of 0.4 ml which was continuously superfused at a rate of 1.5 ml min⁻¹ to maintain the temperature at +33°C.

Measurements of [Ca²⁺]_i

The [Ca²⁺]_i measurements were made using an Axiovert 135 inverted microscope equipped with a Plan-Neofluar × 100/1.30 objective (Carl Zeiss, Oberkochen, Germany) and an Ionoptix (Milton, MA, USA) fluorescence imaging system as described elsewhere (Bokvist et al. 1995). Excitation was effected at 340 and 380 nm and emitted light recorded at 510 nm with a video camera synchronized to the excitation light source and a computer interface. The experiments were conducted using the perforated-patch whole-cell configuration using the pipette-filling solution specified above. Prior to the experiments, the cells were loaded with 0.2 μM fura-2 AM (Molecular Probes) for 16-18 min. Calibration of the fluorescence ratios was performed by using the standard whole-cell configuration to infuse fura-2 with different mixtures of Ca²⁺ and EGTA of known [Ca²⁺]_i.

Solutions

The pipette solution for standard whole-cell experiments (Figs 4B–D and 8) contained (mM): 125 caesium glutamate, 10 CsCl, 10 NaCl, 1 MgCl₂, 5 Hepes, 0.05 EGTA, 3 Mg-ATP, 0.1 cAMP and 0.01 GTP (pH 7.15 with CsOH). Calmodulin binding domain (290-309) (H-LKKFNARRKLKGAILTTMLA-NH₂) was from Calbiochem (La Jolla, CA, USA). Scrambled calmodulin binding domain (290-309) (H-HKRKALFGIMRKALATKNLLT-NH₂) was synthesized by Schafer-N (Copenhagen, Denmark). The molecular mass of the calmodulin binding domain is < 3000 Da. Experiments using dextran-conjugated fura-2 with a molecular mass of 3000 Da (Molecular Probes) suggest that solution exchange between the pipette and the cell interior is nearly complete (> 95 %) in < 2 min. In perforated-patch whole-cell experiments (Figs 1–3, 4E-F and 5–7), the pipette solution consisted of (mM): 76 Cs₂SO₄, 10 NaCl, 10 KCl, 1 MgCl₂ and 5 Hepes (pH 7.35 with CsOH). Electrical contact with the cell interior was established by adding 0.24 mg ml⁻¹ amphotericin B to the pipette solution (Ämmäläet al. 1993). Perforation required a few minutes and the voltage clamp was considered satisfactory when the G_series was constant and > 35-40 nS. The extracellular medium consisted of (mM): 118 NaCl, 20 mM tetraethylammonium-Cl (TEA-Cl), 5.6 KCl, 1.2 MgCl₂, 2.6 CaCl₂, 5 Hepes (pH 7.40 using NaOH) and 5 D-glucose. TEA-Cl was included to block outwardly rectifying K⁺ currents which persist even after replacement of intracellular K⁺ with Cs⁺. In some of the perforated-patch whole-cell recordings, forskolin was included as indicated in the extracellular solution to increase the exocytotic capacity. All chemicals were, unless otherwise indicated, purchased from Sigma.

Figure 4. A small but sustained elevation of [Ca²⁺]_i amplifies the secretory response by a CaM kinase II-dependent mechanism.

Effects of sustained voltage-clamp depolarizations from -70 to -50 mV for 2 min (A) on exocytosis (ΔC_m) elicited by 500 ms depolarizations from -70 to 0 mV before, 10 s and 2 min after the sustained membrane depolarization using the standard whole-cell configuration of the patch-clamp technique under control conditions (B), in the presence of 100 μM calmodulin binding domain (C) or scrambled calmodulin binding domain (D). The peptides were allowed to diffuse into the cell interior for 2-3 min before stimulation commenced. E and F, simultaneous recordings of [Ca²⁺]_i (E) and cell capacitance (F) using the perforated-patch whole-cell recordings. In E, the change in membrane voltage is shown schematically above the [Ca²⁺]_i recording and the horizontal line indicates the basal [Ca²⁺]_i. In B-D and F, the dotted lines correspond to the prestimulatory capacitance level.

Figure 8. Inhibition of CaM kinase II abolishes ACh-stimulated exocytosis.

Ca²⁺ currents (I_Ca) and changes in cell capacitance (ΔC_m) elicited by 500 ms voltage-clamp depolarizations going from -70 to 0 mV (V_m) were recorded in the absence and presence of ACh under control conditions (A), when the pipette solution was supplemented with 100 μM of the CaM binding domain (290-309) (C) or the scrambled form of the peptide (E). B, D and F, histograms summarizing the effects of ACh on depolarization-induced increases in cell capacitance (ΔC_m) and peak Ca²⁺ current (I_Ca) under control conditions (B), in the presence of CaM binding domain (D) or the scrambled peptide (F). In A, C and E the dotted lines correspond to the zero-current level (I_Ca) or the prestimulatory capacitance level (ΔC_m). Data are mean values ±s.e.m. of 9 (B) and 5 experiments (D and F). *P < 0.05, **P < 0.01.

Figure 1. Acetylcholine stimulates exocytosis in mouse pancreatic B-cells.

A, effects of ACh (250 μM) on whole-cell Ca²⁺ currents (I_Ca) and cell capacitance (ΔC_m) elicited by 200 ms depolarizations from -70 to 0 mV as indicated (V_m). The experiment was performed in the continuous presence of 2 μM of the adenylate cyclase activator forskolin and the currents and the exocytotic responses were recorded before and 2 min after application of ACh. Note that the Ca²⁺ currents are displayed on an expanded time scale. B, histograms summarizing the effects of ACh on the peak Ca²⁺ current (I_ca) and exocytosis (ΔC_m) in a series of 11 experiments. *P < 0.05. C, as in A but the experiment conducted in the presence of the muscarinic receptor antagonist atropine (25 μM). D, as in B but experiments conducted in the presence of atropine. In these experiments and those of Fig. 3, the depolarizations were 200 ms long (rather than the standard 500 ms) to avoid extensive degranulation of the cells. In A and C, the dotted lines indicate the zero-current level and the prestimulatory capacitance level, respectively. Data are mean values ±s.e.m. of six experiments.

Figure 3. ACh-induced stimulation of exocytosis requires mobilization of intracellular Ca²⁺ stores.

A, effects of 250 μM ACh and 100 nM PMA on cell capacitance (ΔC_m) and whole-cell Ca²⁺ currents (I_Ca) elicited by 200 ms voltage-clamp depolarizations from -70 to 0 mV (V_m). The cells were pre-treated with 0.5 μM thapsigargin for > 20 min and the experiments were conducted in the continuous presence of thapsigargin and 2 μM forskolin. Note that whereas ACh failed to stimulate exocytosis under these experimental conditions, PMA remained effective. B and C, histograms showing average changes in cell capacitance (ΔC_m, B) and peak Ca²⁺ current (I_Ca, C) in the absence and presence of ACh and PMA. In A, the dotted lines indicate the zero-current level and the prestimulatory capacitance level, respectively. Data are mean values ±s.e.m. of 7 experiments. * P < 0.05.

Figure 5. ACh elicits spontaneous changes in cell capacitance.

Effects of 250 μM ACh on cell capacitance (ΔC_m), [Ca²⁺]_i and Ca²⁺ currents (I_Ca) in a cell voltage clamped at -70 mV except during the depolarizations, which lasted 500 ms and went to 0 mV (V_m). The dotted vertical lines indicate the temporal correlation between the components of capacitance increase and the underlying changes of [Ca²⁺]_i. The inset compares the depolarization-evoked [Ca²⁺]_i transients observed in the absence (a) and presence (b) of ACh. The traces are typical for a total of 6 experiments.

Figure 7. ACh increases the rate of recovery of the readily releasable pool.

Measurements of cell capacitance (ΔC_m) in the perforated-patch configuration before (A) and 2 min after addition of 250 μM ACh (B). Changes in cell capacitance were evoked by a single 200 ms depolarizations from the holding potential of -70 mV to 0 mV 45 s before and at different times (3-80 s) after a train of 500 ms voltage-clamp depolarizations applied at 1 Hz (trace labelled V_m). C and D, average changes in cell capacitance (ΔC_m, C) and peak Ca²⁺ current (I_Ca, D) under control conditions (^) and in the presence of ACh (•) in response to the 200 ms depolarizations. The shaded areas in C indicate the capacitance increases elicited by the 200 ms depolarization applied before the train (mean values ±s.e.m.). In C, all values obtained in the presence of ACh at times later than 3 s are significantly (P < 0.01) larger than those observed under control conditions.

Data analysis

Results are presented as mean values ±s.e.m. for the indicated number of experiments. All current amplitudes are given without compensation for leak conductance. Statistical significance was evaluated using Student's t test for paired data. Increases in cell capacitance are quoted as the percentage change in excess of the control level. Accordingly, stimulation of exocytosis from a basal value of 10 fF to a new value of 50 fF corresponds to an increase of 400 %. Experiments commenced when two successive depolarizations applied at a 2 min interval elicited exocytotic responses of the same amplitude (± 10 %) to ascertain that the observed changes are not simply attributable to spontaneous long-term changes of the secretory capacity.

RESULTS

Acetylcholine stimulates exocytosis in mouse B-cells

Figure 1A shows Ca²⁺ currents and associated changes in cell membrane capacitance elicited by 200 ms voltage-clamp depolarizations going from -70 to 0 mV before and after addition of 250 μM ACh. Under control conditions, the membrane depolarization evoked a capacitance increase of 40 fF corresponding to the release of about 20 secretory granules as each granule contributes ∼2 fF of capacitance upon exocytosis (Ämmäläet al. 1993). Two minutes after the addition of ACh (250 μM), the same membrane depolarization evoked a capacitance increase of 130 fF. On average (Fig. 1B), ACh (250 μM) produced a 230 ± 70 % (P < 0.05; n = 11) stimulation of exocytosis. The latter effect was accompanied by a small but significant decrease in the amplitude of the peak Ca²⁺ current (7 ± 2 %, P < 0.05; cf. Gilon et al. 1997). A similar small reduction of the integrated Ca²⁺ current was observed: from 9.1 ± 0.2 to 8.5 ± 0.3 pC in the presence of ACh (P≡ 0.05). The stimulatory effect of ACh on exocytosis was reversible, and 6 min following washout of the neurotransmitter from the bath solution, the depolarization-evoked capacitance increase had decreased from 99 ± 28 to 36 ± 12 fF. The action of ACh on exocytosis was abolished by atropine (25 μM) as expected for an effect mediated by activation of muscarinic receptors (Fig. 1C-D). When applied at 25 μM, ACh failed to enhance depolarization-evoked secretion (43 ± 10 fF under control vs. 38 ± 12 fF in the presence of ACh). The high concentrations of ACh required for stimulation of exocytosis are in agreement with recent data on insulin secretion reported by Niwa et al. (1998). Such high concentrations of ACh may develop in the immediate vicinity of the cholinergic nerve endings within the islet.

Acetylcholine stimulates exocytosis following inhibition of PKC

It is well documented that long-term exposure of pancreatic B-cells to ACh leads to activation of PKC (Persaud et al. 1991; Weng et al. 1993). To test for the involvement of PKC in the stimulatory action of ACh on exocytosis, the B-cells were pre-treated with calphostin C (1.5 μM for > 15 min). This selective PKC inhibitor did not influence the exocytotic capacity in the absence of ACh and failed to reduce the stimulatory action of the neurotransmitter (Fig. 2A and B). By contrast, these manoeuvres suppressed the stimulatory action of the PKC activator phorbol 12-myristate 13-acetate (PMA, 100 nM) on exocytosis. In the presence of calphostin C, exocytosis evoked by 500 ms depolarizations to 0 mV amounted to 39 ± 14 and 45 ± 9 fF in the absence and presence of PMA, respectively (not shown).

Figure 2. Stimulatory action of ACh on exocytosis does not involve protein kinase C.

A, whole-cell Ca²⁺ currents (I_Ca) and exocytosis (ΔC_m) evoked by membrane depolarizations (500 ms; V_m) before and 2 min after addition of ACh (250 μM) in the presence of 2 μM forskolin and 1.5 μM of the PKC inhibitor calphostin C. B, histograms summarizing effects on changes of cell capacitance (ΔC_m) and peak Ca²⁺ current (I_ca) evoked by ACh. C, as in A but the cells had been pre-treated with PMA (100 nM for > 20 h) to downregulate PKC. D, as in B but experiments conducted in the presence of PMA as described in C. In A and C, the dotted lines indicate the zero-current level and the prestimulatory capacitance level, respectively. Data are mean values ±s.e.m. of 5 experiments. *P < 0.05.

ACh likewise remained stimulatory when PKC was inhibited by staurosporine (100 nM). In the presence of this PKC inhibitor, the exocytotic responses amounted to 32 ± 7 fF in the control solution and increased by 91 ± 29 fF (P < 0.05; n = 5) 2 min after adding ACh (250 μM; data not shown). Further support for the idea that ACh exerts its stimulatory action independently of PKC activation comes from the observation that ACh remained stimulatory when applied to B-cells in which PKC had been downregulated by exposure to PMA (100 nM) for > 20 h (Fig. 2C -D). Long-term exposure of B-cells to a high concentration of PMA is well established to effectively downregulate PKC activity in mouse B-cells (Arkhammar et al. 1989). These data argue that although activation of PKC certainly is a long-term consequence of adding ACh, this is not required for the acute stimulatory effects of the neurotransmitter on exocytosis which are the focus of this study.

Acetylcholine stimulation of depolarization-evoked exocytosis involves mobilization of intracellular Ca²⁺ stores

In addition to activating PKC, exposure to ACh also leads to generation of InsP₃, which mobilizes Ca²⁺ from intracellular stores. We next investigated to what extent the stimulation of exocytosis may be attributable to this effect. Figure 3A shows the Ca²⁺ currents and exocytotic responses in a B-cell which had been pre-treated with thapsigargin (0.5 μM for > 20 min), an inhibitor of the Ca²⁺-ATPase in the endoplasmic reticulum. Depletion of intracellular Ca²⁺ stores by thapsigargin did not interfere with the ability of a depolarization to elicit exocytosis. However, thapsigargin removed the ACh-induced stimulation of exocytosis without affecting the whole-cell Ca²⁺ current (Fig. 3A-C). This argues that although intracellular Ca²⁺ stores do not normally contribute to depolarization-evoked secretion, they are involved in the ACh-induced enhancement of Ca²⁺-induced secretion. The inability of ACh to be effective in the presence of thapsigargin cannot be attributed to this Ca²⁺-ATPase inhibitor depleting the RRP of granules. Accordingly, application of PMA (100 nM) to activate PKC increased the exocytotic responses by 118 ± 23 % (P < 0.05; n = 7) also in the presence of thapsigargin (Fig. 3B). Collectively, the observations made in the presence of thapsigargin argue that the stimulatory action of ACh depends on mobilization of Ca²⁺ from intracellular stores rather than activation of PKC.

From Figs 1–;3, it is also clear that the time course of endocytosis (seen as a capacitance decrease), in agreement with what we have previously reported (Eliasson et al. 1996), varies significantly from one cell to another. Whereas endocytosis in some cells is rapid and complete within 10 s (i.e. control in Fig. 1A), the capacitance increase is sustained in other cells and requires ≥ 1 min for complete recovery (Fig. 2).

Depolarization-induced elevations in [Ca²⁺]_i shape the exocytotic response

Studies on chromaffin cells have indicated that Ca²⁺ plays a dual role in the secretory process: it both triggers exocytosis of the secretory granules from the RRP and promotes their recruitment for subsequent release (von Rüden & Neher, 1993; Smith et al. 1998). The latter effect operates at [Ca²⁺]_i lower than those required to initiate exocytosis. In Fig. 4 we investigated whether a small elevation of [Ca²⁺]_i is sufficient to accelerate granule mobilization and thus increases the size of the RRP. This possibility was studied by measurements of cell capacitance before, during and after a 2 min membrane depolarization from -70 mV to -50 mV (Fig. 4A) under both standard (Fig. 4B -D) and perforated-patch whole-cell conditions (Fig. 4F). A 2 min depolarization to -50 mV was used for these experiments to maintain the standard interval between two consecutive depolarizations in order to ascertain that any changes of the exocytotic responses were not simply due to more extensive refilling of the RRP. In a series of five experiments, the membrane depolarization to -50 mV increased [Ca²⁺]_i from a basal of 209 ± 14 nM to 292 ± 26 nM (P < 0.02), which decayed to 225 ± 24 nM (P < 0.01) upon returning to the holding potential of -70 mV (Fig. 4E). This slight elevation of [Ca²⁺]_i, whilst not being sufficient to evoke much secretion by itself, transiently increased the exocytotic capacity of the B-cells and the amplitude of the capacitance increases elicited by voltage-clamp depolarizations rose by 210 ± 48 % (P < 0.05; n = 5) over that seen prior to the 2 min depolarization to -50 mV (standard whole-cell, Fig. 4B) and by 240 ± 98 % (P < 0.05; n = 5) using the perforated-patch configuration (Fig. 4F). As shown in Fig. 4E, these increases in cell capacitance were not due to an increased amplitude of the depolarization-induced [Ca²⁺]_i transients.

To elucidate the mechanisms by which Ca²⁺ promotes the transient overfilling of the RRP, the pipette solution was supplemented with 100 μM of the calmodulin binding domain (290-309) of calmodulin-dependent kinase II (CaM kinase II). This synthetic peptide is a potent calmodulin antagonist and a specific inhibitor of CaM kinase II (Payne et al. 1988). As evident from Fig. 4C and consistent with an earlier report (Ämmäläet al. 1993), this peptide reduced the exocytotic response by 46 ± 16 % when included in the pipette solution; from a control value of 45 ± 10 fF to 29 ± 8 fF (P < 0.05; n = 5). The effect was not associated with a reduction in the peak Ca²⁺ current which amounted to 54 ± 11 pA (n = 5) under control conditions and 49 ± 12 pA (n = 5) in the presence of the calmodulin binding domain (290-309). Interestingly, the transient stimulation of exocytosis following the 2 min membrane depolarization to -50 mV was almost abolished (39 ± 11 fF; n = 5). By contrast, inclusion of a scrambled peptide (see Methods) did not reduce the stimulatory effect of subthreshold depolarization on exocytosis (Fig. 4D).

We have also explored the involvement of CaM kinase II by the use of KN-62. However, as described previously (Ämmäläet al. 1993), this compound inhibits the Ca²⁺ current and we have therefore investigated its effects on exocytosis by intracellular dialysis of the B-cell with a Ca²⁺-EGTA solution with a free Ca²⁺ concentration of 2 μM (cf. Renström et al. 1997). Under control conditions the rate of capacitance increase measured over the initial 60 s after establishment of the standard whole-cell configuration was 27 ± 2 fF s⁻¹ (n = 5). This decreased to 6 ± 2 fF s⁻¹ (n = 6; P < 0.001) in cells treated with KN-62 (10 μM for 15 min and 10 μM added to the pipette solution). The calmodulin binding domain (290-309) affected exocytosis evoked by Ca²⁺ infusion similarly (not shown).

Acetylcholine stimulates exocytosis in hyperpolarized B-cells

ACh promotes InsP₃-production and mobilizes intracellular Ca²⁺ even in hyperpolarized B-cells. We have previously demonstrated that InsP₃-induced increases in [Ca²⁺]_i elicit a short-lived stimulation of exocytosis (Ämmäläet al. 1993). Figure 5 shows simultaneous measurements of [Ca²⁺]_i and changes in cell capacitance (C_m) in a single voltage-clamped B-cell before, during and after addition of ACh. Under control conditions, a 500 ms membrane depolarization from a holding potential of -70 mV to 0 mV evoked a transient increase in [Ca²⁺]_i from basal 80 nM to a peak value of 0.95 μM, which was associated with a 24 fF increase in cell capacitance. Sustained exposure to ACh (250 μM, 3 min) transiently elevated [Ca²⁺]_i and reached a peak concentration as high as 0.7 μM. The transitory nature of the ACh-induced rise in [Ca²⁺]_i is likely to reflect depletion of the intracellular Ca²⁺ pools. Close inspection revealed that the rising phase of the [Ca²⁺]_i increase consisted of at least three distinct spikes which occurred at a frequency of ∼0.2 Hz. The ACh-evoked increase in [Ca²⁺]_i was associated with a substantial increase in cell capacitance even when the membrane potential was maintained at -70 mV thus precluding the contribution of influx of Ca²⁺ through the voltage-dependent Ca²⁺ channels. As indicated by the vertical dotted lines, each [Ca²⁺]_i spike was associated with increases in cell capacitance. The increases in cell capacitance did not correlate with any concomitant changes of the measured membrane conductance (not shown). In a series of 20 cells, the total increase in cell capacitance elicited by addition of ACh amounted to 468 ± 52 fF. The increase in exocytosis was followed by a gradual return towards the pre-stimulatory capacitance level, presumably reflecting the retrieval of secreted membrane by endocytosis (cf. Eliasson et al. 1996).

The fact that ACh when applied to a voltage-clamped B-cell evoked a transient increase in [Ca²⁺]_i with no apparent sustained component (cf. Miura et al. 1997) argues that the refilling of the intracellular pools occurs via voltage-gated Ca²⁺ channels. ACh acts by opening a membrane conductance leading to membrane depolarization which in turn results in activation of the voltage-gated Ca²⁺ channels but the channel need not be Ca²⁺ selective itself. It is of interest that a membrane conductance fulfilling these criteria was recently described in B-cells (Roe et al. 1998).

Despite the massive exocytosis evoked by addition of ACh, secretion elicited by a subsequent 500 ms voltage-clamp depolarization was nevertheless strongly potentiated by the agonist, from a basal value of 32 ± 2 fF to 101 ± 12 fF (P < 0.01; n = 6) in the presence of ACh. This indicates that ACh, in addition to promoting exocytosis, must also accelerate the refilling of the RRP and that the latter action predominates.

In the cells where it was possible to simultaneously measure exocytosis and [Ca²⁺]_i, the rate of capacitance increase observed in the presence of ACh (145 ± 24 fF s⁻¹, n = 6) was not much smaller than that evoked by the voltage-clamp depolarization (202 ± 10 fF s⁻¹ during the 500 ms pulse; n = 6). The corresponding increases in [Ca²⁺]_i were 0.93 ± 0.04 and 0.68 ± 0.03 μM (n = 6) for the voltage-clamp depolarization and the ACh-evoked [Ca²⁺]_i transient, respectively.

Application of ACh also affected the time course of the recovery of the depolarization-induced increases in [Ca²⁺]_i (Fig. 5, inset). The action was manifested as a marked retardation of the falling phase and the time constant (τ) increased from 0.8 ± 0.1 s to 4.5 ± 0.6 s (P < 0.01, n = 9) in the presence of ACh. This effect was not associated with changes of the amplitude of the [Ca²⁺]_i transient, the holding current or the magnitude of the voltage-gated Ca²⁺ current (I_Ca). The effects of ACh on [Ca²⁺]_i handling were mimicked by inclusion of 50 μM InsP₃ in the pipette solution during whole-cell experiments and pre-treatment of the cells with thapsigargin; τ rose from a basal value of 0.8 ± 0.1 s to 4.9 ± 0.5 s in the presence of InsP₃ (P < 0.001; n = 8) and from 0.7 ± 0.2 to 4.1 ± 0.6 s (n = 9) in the absence and presence of the Ca²⁺-ATPase inhibitor, respectively (P < 0.005; data not shown).

Acetylcholine enhances exocytosis evoked by repetitive membrane depolarizations

Since a small elevation of resting [Ca²⁺]_i is sufficient to produce a marked potentiation of exocytosis (Fig. 4) it is easy to envisage how ACh-induced elevations of [Ca²⁺]_i or the slower recovery of the [Ca²⁺]_i transients in the presence of the neurotransmitter can be expected to have pronounced effects on the secretory process by accelerating granule mobilization with resultant overfilling of the RRP. This should be detectable as an increased size of the amount of exocytosis that can maximally be elicited during intense and repetitive stimulation. Figure 6A shows trains of 500 ms depolarizations applied at 1 Hz in the absence and presence of ACh until a depolarization failed to evoke a further capacitance increase. Under control conditions (Fig. 6A, left), the maximum increase in cell capacitance was limited to 67 ± 21 fF (n = 5), in fair agreement with that previously reported elsewhere (Renström et al. 1997). The exhaustion of the exocytotic capacity is likely to reflect depletion of the RRP rather than inactivation of the Ca²⁺ current with resultant suppression of Ca²⁺-induced secretion. This is suggested by the observation that the integrated Ca²⁺ current measured at the end of the train, when secretion had ceased, was only reduced by 27 ± 14 % with respect to the first depolarization. We have demonstrated elsewhere that a reduction of this magnitude only interferes marginally with the exocytotic capacity (Renström et al. 1996). Two minutes following ACh stimulation (250 μM), the train of depolarizations produced a maximal increase in cell capacitance of 231 ± 37 fF which was only associated with a 15 ± 8 % inhibition of the integrated Ca²⁺ current between the first and last membrane depolarization (Fig. 6A, right).

Figure 6. ACh increases the size of the readily releasable pool of granules.

A, trains of 500 ms voltage-clamp depolarizations from -70 to 0 mV were applied at a frequency of 1 Hz (V_m) using the perforated-patch whole-cell configuration. The trains of depolarizations were applied in the absence and presence of 250 μM ACh and went on until a depolarization failed to evoke a further increase in cell capacitance (lower trace). The middle trace shows the associated Ca²⁺ current (I_Ca). The interval between the trains of depolarizations was 3 min in order to allow for refilling of the readily releasable pool of granules and ACh included in the bath solution when two consecutive trains applied under control conditions evoked similar (± 10 %) changes in cell capacitance. The Ca²⁺ currents selected for display were the ones evoked by the first depolarization of the train and the first that failed to elicit a further increase in cell capacitance and were taken as indicated by the vertical dotted lines. B, increase in cell capacitance (ΣΔC_m) displayed against the Ca²⁺ entry (ΣQ_Ca) under control conditions (^) and in the presence of ACh (•). The dotted lines were derived by a least-squares fit to the observed values. Also indicated in B is the Ca²⁺ entry at which exocytosis is half-maximal in the presence or absence of ACh. Data represent mean values ±s.e.m. of 5 experiments. The difference between control and ACh-stimulated exocytosis was significant (P < 0.01) for all points.

An alternative approach to derive the size of the RRP is to compare the amplitude of the exocytotic responses evoked by the two first depolarizations of the train as first described by Gillis et al. (1996). If there is a limited pool of granules in the cell and the release probability of the granules is unchanged, then the second depolarization will evoke a smaller increase in cell capacitance simply because there are fewer granules for Ca²⁺ to act on. The size of the RRP and the increase in cell capacitance evoked by the first and second depolarizations (ΔC_m,1 and ΔC_m,2, respectively) is given by:

(1)

where S represents the sum of ΔC_m,1 and ΔC_m,2 and R the ratio ΔC_m,2/ΔC_m,1 (see Gillis et al. 1996 for derivation). The accuracy of eqn (1) depends greatly on the value of R with the precision decreasing as R increases towards 1. In a series of five experiments, the value of R averaged 0.55 ± 0.09 ascertaining the validity of this protocol. Using the approach by Gillis et al. (1996) the size of the RRP increased from 50 ± 8 fF (n = 5) under control conditions to 256 ± 68 fF (n = 5; P < 0.025 vs. control) after addition of ACh. Both values are in fair agreement with those obtained by the alternative protocol described above. We acknowledge that the method of Gillis et al. (1996) might result in a slight overestimation of the RRP under our experimental conditions since some mobilization is likely to occur during 1.5 s between the onset of the first pulse and the end of the second depolarization. However, given the time course of mobilization (see Fig. 7), showing little replenishment during the first 3 s, the error is likely to be small.

The relationship between the cumulative Ca²⁺ entry (ΣQ_Ca) and the sum of the step increases in cell capacitance (ΣΔC_m, to compensate for the influence of endocytosis) for the first six voltage-clamp depolarizations before and after addition of ACh stimulation is summarized in Fig. 6B. Application of ACh was associated with a small reduction of the integrated Ca²⁺ current. The stimulation of exocytosis can therefore not be explained in terms of stimulated Ca²⁺-dependent secretion. The slope of the relationships between ΣQ_Ca and ΣΔC_m provides an estimate of the apparent efficacy of Ca²⁺ as an initiator of exocytosis. It increased from 3.3 ± 1.1 to 12.5 ± 2.5 fF pC⁻¹ in the absence and presence of ACh (P < 0.05; n = 5), respectively. We considered the possibility that exposure to ACh may influence the Ca²⁺ dependence of the secretory machinery. From the relationship in Fig. 6B, it is possible to estimate that half-maximal exocytosis is observed at ∼6 pC both in the presence and absence of ACh. This crude analysis suggests that sensitization of the exocytotic apparatus to Ca²⁺ is unlikely to contribute to the stimulation of exocytosis. Rather the fourfold stimulation of exocytosis compares favourably with the change of the pool size. We conclude that the enhancement of exocytosis is the result of Ca²⁺ acting on a larger pool of granules but that the release probability of the individual granule is unaffected.

Acetylcholine accelerates refilling of the RRP

Our data suggest that ACh increases the size of the RRP by acceleration of granule mobilization. We next attempted to estimate the kinetics of granule mobilization (Fig. 7) by measuring the rate of recovery from complete depletion of the RRP. This was done by first applying a 200 ms depolarization to determine the exocytotic capacity. Next (45 s after the first depolarization) a train consisting of 10-14 depolarizations (500 ms and 1 Hz) was applied to deplete the RRP. The rate of recovery of the RRP was then monitored by application of 200 ms depolarizations at various intervals (3-80 s) after the train. The protocol was first applied under control conditions (Fig. 7A) and subsequently (using the same cell) 2 min after addition of 250 μM ACh (Fig. 7B). In a series of six experiments, the train of voltage-clamp depolarizations increased cell capacitance from 76 ± 17 fF under control conditions to 256 ± 17 fF (P < 0.05) in the presence of ACh, in close agreement with the values obtained in Fig. 6.

The rates of recovery in the absence and presence of ACh are summarized in Fig. 7C. Under control conditions, no exocytosis was observed 3 and 5 s after application of the train. The exocytotic responses subsequently increased towards the control amplitude (elicited by the 200 ms depolarization applied before the train) with complete recovery observed at ≥ 60 s. A similar relationship was observed in the presence of ACh but in this case, a small secretory response was already observable 5 s after the train. From the relationship between the amplitude of the exocytotic responses and the interval between the train and the pulse, it was possible to estimate τ for the recovery of the RRP. It decreased from a control value of 31 ± 5 s to 19 ± 1 s (P < 0.05; n = 6) in the presence of ACh (Fig. 7C). It was not feasible to apply the train prior to each short depolarization. As a result, there is a risk that we underestimate the rate of recovery as each depolarization will produce some depletion of the RRP. However, the fact that the rate of recovery is well described by a single exponential for all data points (including the short intervals) argues that partial depletion of the RRP by the 200 ms depolarizations does not interfere with the measurements. The acceleration of recovery of the RRP by ACh was associated with a small reduction in the peak Ca²⁺ current (Fig. 7D). It is worthy of note that, irrespective of whether ACh was present or not, the amplitude of the Ca²⁺ current was almost constant throughout the experiment and the observed increase in exocytotic capacity can accordingly not be explained as the recovery of the Ca²⁺ current from Ca²⁺-induced inactivation of the Ca²⁺ channels.

Acetylcholine stimulates exocytosis via activation of CaM kinase II

We next investigated the molecular mechanisms by which ACh potentiates Ca²⁺-induced exocytosis. Given that ACh increases [Ca²⁺]_i (Fig. 5) and that a small increase in resting [Ca²⁺]_i promotes exocytosis in a CaM kinase II-dependent fashion (Fig. 4), we explored whether activation of this kinase also underlies the stimulatory action of ACh on secretion. Calmodulin binding domain (290-309) was selected to investigate the involvement of CaM kinase II. Figure 8A shows the Ca²⁺ currents and associated changes in cell capacitance following 500 ms voltage-clamp depolarizations from -70 to 0 mV before and after ACh (250 μM) stimulation using the standard whole-cell configuration. It is clear that ACh remained stimulatory during these experimental conditions and the amplitude of the depolarization-evoked capacitance change increased from a basal 54 ± 18 fF to 122 ± 31 fF (n = 9) 2 min following addition of the neurotransmitter. This stimulation is close to that observed when the recordings were made using the perforated-patch configuration (Fig. 1). Again the stimulatory effect cannot be accounted for by increased Ca²⁺ entry and, if anything, the peak Ca²⁺ current fell from 46 ± 11 pA under control conditions to 41 ± 13 pA in the presence of ACh (n = 9). Inclusion of 100 μM calmodulin binding domain (290-309) in the pipette solution reduced the exocytotic response under control conditions by 52 % (P < 0.05; n = 5) and it amounted to 21 ± 9 fF. More importantly, the stimulatory effect of ACh on exocytosis was abolished by this manoeuvre (Fig. 8C -D). These effects could not be attributed to interference of Ca²⁺ entry and the peak current amplitude was unaffected by both ACh and the CaM kinase II inhibitor (compare Fig. 8B and D). By contrast, ACh remained stimulatory in the presence of 100 μM of the scrambled peptide (Fig. 8E -F).

DISCUSSION

Exocytosis in the pancreatic B-cell is triggered by a rise in [Ca²⁺]_i. This Ca²⁺-dependent regulation of insulin secretion includes not only the rapid fusion of the granules with the plasma membrane but also pre-fusion events such as the mobilization of the granules from the reserve pool, their priming and docking at the release sites. Using capacitance measurements of exocytosis we demonstrate here that ACh produces a prompt 2.3-fold potentiation of Ca²⁺-dependent exocytosis. Below we consider the underlying cellular mechanisms.

Release of intracellular Ca²⁺ elicits a transient stimulation of exocytosis

Insulin secretion stimulated by nutrient secretagogues that cause membrane depolarization, such as glucose, results from an elevation of [Ca²⁺]_i just beneath the membrane due to Ca²⁺ entry through voltage-gated L-type Ca²⁺ channels (Theler et al. 1992; Bokvist et al. 1995). These Ca²⁺ domains develop upon opening of the Ca²⁺ channels and collapse almost immediately (within 300 ms) upon their closure. As a result, exocytosis normally proceeds only during the periods the Ca²⁺ channels are active (Ämmäläet al. 1993; Bokvist et al. 1995). Similar localized Ca²⁺ transients have been observed in rat B-cells stimulated by the muscarinic agonist carbachol (Theler et al. 1992). However, as previously documented in chromaffin cells (von Rüden & Neher, 1993) and now demonstrated also in pancreatic B-cells, moderate but long-lasting increases in [Ca²⁺]_i (Fig. 4) accelerate the rate of granule mobilization and thus promote the replenishment of the release-competent pool of secretory granules. The significance of the latter processes is illustrated by the observations that only a short-lived stimulation of secretion can be elicited under experimental conditions when mobilization is not operational (Renström et al. 1997; Eliasson et al. 1997).

Addition of ACh is well established to elevate [Ca²⁺]_i by InsP₃-dependent release from intracellular Ca²⁺ stores and produces increases in [Ca²⁺]_i with highly variable patterns (Prentki et al. 1988). This variability is also reflected in the time course of depolarization-independent secretion and the exocytotic responses (Fig. 5). During these ACh-induced steps, the measured rate of exocytosis (145 fF s⁻¹) approaches that attained during a voltage-clamp depolarization (202 fF s⁻¹) although the [Ca²⁺]_i transient was 25 % lower. The fact that intracellular Ca²⁺ release is capable of eliciting exocytosis at a high rate suggests that the Ca²⁺ release sites are situated in the immediate vicinity of the secretory granules. Indeed, preliminary ultrastructural studies using electron microscopy indicate that endoplasmic reticulum-like structures juxtapose to the secretory granules near the plasma membrane (T. Kanno, S. Göpel, P. Rorsman & L. Eliasson, unpublished observations). A similar situation exists in pituitary gonadotropes where localized Ca²⁺ increases due to the release from internal stores play a prominent role in the control of exocytosis. In these cells, hormone-induced formation of InsP₃ appears to release Ca²⁺ selectively from stores just beneath the plasma membrane and thus raise [Ca²⁺]_i to exocytotic levels at the release sites (Tse et al. 1997). In this context it may be pertinent that InsP₃-dependent Ca²⁺ release mechanisms have been described in zymogen granules from pancreatic acinar cells (Gerasiminko et al. 1996).

Stimulation of exocytosis by ACh does not require activation of PKC

Muscarinic receptor activation in the B-cell leads not only to formation of InsP₃ but also to production of diacylglycerol with resultant long-lasting activation of PKC (Weng et al. 1993). This prompted the proposal that activation of PKC and sensitization of the secretory machinery to Ca²⁺ underlie the prolonged effects of cholinergic agonists on insulin secretion (Weng et al. 1993). This concept would indeed be consistent with the observation that downregulation of PKC activity with phorbol esters abolished the ability of carbachol to potentiate sustained insulin secretion. However, the same study revealed that the initial secretory response to carbachol was little affected by this procedure (Persaud et al. 1991). This is in agreement with our present observation that ACh-induced stimulation of exocytosis was influenced neither by the PKC inhibitors staurosporine and calphostin C, nor by downregulation of PKC activity. Activation of PKC by PMA promptly (< 2 min) stimulates exocytosis in intact B-cells (Ämmäläet al. 1994 and Fig. 3A) and the effect is poorly reversible. The rapid reversibility of the acute exocytotic response following removal of ACh from the bath, which distinguishes this effect from the sustained PKC-dependent action of ACh (Weng et al. 1993), is easier to reconcile with the idea that changes in [Ca²⁺]_i, rather than activation of PKC, are important for the initial ACh-induced component of insulin secretion.

ACh affects the duration of [Ca²⁺]_i transients

In addition to mobilizing Ca²⁺ from intracellular stores, ACh also extends the duration of the depolarization-evoked [Ca²⁺]_i transients. Given that the duration, but not the amplitude, of the [Ca²⁺]_i transients was affected and that the effects were similar to those obtained upon inhibition of Ca²⁺ uptake mechanisms, it seems plausible that interference with the [Ca²⁺]_i handling mechanisms are involved. We propose that ACh, when applied at the high concentration used in this study, produces a constant elevation of InsP₃ which serves to effectively short-circuit the intracellular Ca²⁺ stores: sequestered Ca²⁺ is instantly released due to the permanent activation of the InsP₃-receptor channels (Hagar et al. 1998). Indeed, inclusion of InsP₃ exerted virtually identical effects on depolarization-induced [Ca²⁺]_i transients. Another possibility is that voltage-gated Ca²⁺ entry stimulates Ca²⁺ release from the InsP₃-sensitive Ca²⁺ stores which in turn contributes to the slowed decay of the increases in [Ca²⁺]_i elicited by the voltage-clamp depolarizations. However, this alternative seems less likely given that the amplitude of the [Ca²⁺]_i transients is the same in the presence and absence of ACh/InsP₃. The different time constants obtained for the recovery of the RRP and the decay of the depolarization-induced increases in [Ca²⁺]_i (19 s vs. 5 s in the presence of ACh) suggest that once the process of refilling has been activated, it proceeds even at basal [Ca²⁺]_i (cf. Smith et al. 1998).

ACh-induced stimulation of granule mobilization increases size of the RRP

The slow disappearance of Ca²⁺ from the cytoplasm in response to membrane depolarization in the presence of ACh may well be physiologically significant since a small increase (100 nM) in ‘resting’[Ca²⁺]_i strongly potentiates Ca²⁺-induced secretion. We therefore propose that the effects of ACh on Ca²⁺ handling contribute significantly to the enhancement of secretion observed during cholinergic stimulation.

The stimulation of exocytosis observed following (ACh-induced) elevation of [Ca²⁺]_i results from stimulation of granule mobilization and ‘overfilling’ of the RRP so that more granules are available for release once [Ca²⁺]_i rises to exocytotic levels. The size of the RRP (estimated from the maximum increase in cell capacitance that can be elicited by a train of depolarizations) increased 3.5-fold, from ∼35-40 granules (67-76 fF) under basal conditions to 120-130 granules (231-256 fF) in the presence of ACh. The [Ca²⁺]_i dependence of the refilling has been estimated to be half-maximal at Ca²⁺ concentrations around 0.2-0.3 μM in mouse B-cells (Renström et al. 1997) and chromaffin cells (Smith et al. 1998). This value is obviously very close to that attained during a 2 min depolarization to -50 mV, which failed to evoke exocytosis itself but markedly enhanced secretion elicited by a subsequent 500 ms depolarization to 0 mV.

Our data suggest that ACh increased not only the size of the RRP but also the rate of granule replenishment. The experimental protocol we employed involves complete emptying of the RRP. This allows us to exclude the possibility that the acceleration of mobilization results from suppression of depriming, i.e. the removal of granules from the RRP, rather than an increased supply rate. Our data are reminiscent of those recently reported in chromaffin cells (Smith et al. 1998) where [Ca²⁺]_i increased the size of the RRP by acceleration of mobilization, the maximum pool size being attained at approximately 0.6 μM [Ca²⁺]_i. In this study, activation of PKC was found to produce a further increase in the RRP by increasing the supply rate whilst not much influencing the time course of recovery. Our data indicate that ACh increases the RRP in pancreatic B-cells by at least two mechanisms. First, it decreases the time constant for the recovery of the RRP from 31 to 19 s. Secondly, the capacity of the refilling machinery is increased so that the actual number of granules that pass from the reserve pool into the RRP is increased. The number of granules residing in the RRP at any time after complete depletion is given by:

(2)

where RRP is the pool size at time t, RRP_max the size of the RRP when it is maximally filled and τ the measured time constant for the RRP recovery. The rate of supply (RRP’) of granules from the reserve pool into the RRP can be derived from eqn (2) yielding the formula:

(3)

Insertion of the observed values for RRP_max and τ (76 fF and 31 s and 256 fF and 19 s in the absence and presence of ACh, respectively) indicates that the maximum supply rate (measured at time zero) increased 12-fold from a basal value of ∼2 fF s⁻¹ to ∼25 fF s⁻¹ in the presence of the agonist. The latter values correspond to rates of mobilization of 1 and > 10 granules per second.

CaM kinase II controls Ca²⁺- and ACh-induced exocytosis in the B-cell

Calmodulin is a major Ca²⁺ binding protein and plays an important role in insulin release (Sugden et al. 1979). By binding to calmodulin, Ca²⁺ activates Ca²⁺-calmodulin-dependent protein (CaM) kinases. One such CaM kinase is the multifunctional enzyme CaM kinase II, the presence of which has been documented in the pancreatic B-cell (Hughes et al. 1993). This enzyme is activated when B-cells are exposed to glucose (Wenham et al. 1994) or muscarinic agonists (Babb et al. 1996). Previous evidence for the involvement of CaM kinase II in regulating exocytosis in mouse B-cells was provided by the observation that calmodulin binding domain (290-309) and KN-62 reduced exocytosis evoked by voltage-clamp depolarizations (Ämmäläet al. 1993). In contrast, KN-62 failed to affect Ca²⁺-induced insulin secretion from permeabilized HIT cells and it was argued that CaM kinase II does not regulate exocytosis in this insulinoma cell line (Li et al. 1992). In the present study, we consistently observed that the calmodulin binding domain (290-309) not only abolished the effects of ACh and elevated [Ca²⁺]_i on exocytosis, it also reduced Ca²⁺-evoked exocytosis under control conditions by 40-50 % (Figs 4 and 8). This suggests that activation of CaM kinase II is required both for the acceleration of granule mobilization under conditions of stimulated exocytosis and for the maintenance of the secretory capacity under control conditions. However, the fact that some exocytosis remains observable in the presence of the CaM kinase II inhibitor argues that this enzyme does not participate in the late Ca²⁺-dependent fusion events. This role is likely to be fulfilled by other Ca²⁺ binding proteins such as synaptotagmin (Lang et al. 1997).

Previous data have led to the proposal that first phase insulin secretion reflects the release of granules belonging to the RRP (Eliasson et al. 1997). In this scenario, second phase insulin secretion is envisaged to result from the ATP-, Ca²⁺- and time-dependent mobilization of granules from the reserve pool. The finding that disruption of actin filaments primarily inhibits second phase insulin secretion (Li et al. 1994), with little effect on the first phase, would indeed be consistent with such an idea and suggests that interactions between the cytoskeleton and the secretory granules are required for the replenishment of the RRP. Circumstantial evidence for a role for actin-myosin interactions in the mobilization of secretory vesicles in the pancreatic B-cell is provided by the presence of myosin on the insulin-containing granules (Howell, 1984). CaM kinase II also associates with the secretory granules in the B-cells (Möhlig et al. 1997). In insulin-secreting cells, substrates for CaM kinase II include (1) a subunit of tubulin (Colca et al. 1983); (2) microtubuli-associated protein 2 (Krueger et al. 1997) and (3) myosin light chain (Niki et al. 1993; Li et al. 1994). Clearly, the presence of these CaM kinase II substrates raises the possibility that this kinase regulates a number of processes controlling the interactions between the cytoskeleton and the secretory granules. Another potential for CaM kinase II-dependent phosphorylation is the synapsin I-like protein that has recently been identified in MIN6 insulinoma cells (Matsumoto et al. 1995). This possibility is suggested by analogy to the situation in neurons where synapsins regulate the interactions between synaptic vesicles and the cytoskeleton (Llinás et al. 1991). Phosphorylation of synapsin by CaM kinase II leads to the dissociation of the synaptic vesicles from the cytoskeleton thus facilitating their translocation to and fusion with the plasma membrane. Our data suggest that CaM kinase II plays an important role in ACh-stimulated exocytosis in the B-cell but they are likely to be of significance for the control of exocytosis in general. In fact, any condition leading to a small but sustained elevation of [Ca²⁺]_i will mimic the action of ACh on granule mobilization and refilling of the readily releasable pool. Such conditions include exposure of the B-cell to glucagon-like peptide 1 and other hormones increasing cytoplasmic cyclic AMP-levels (Gromada et al. 1998). It remains to be established to what extent granule mobilization involves a chemical modification of the secretory granules or their physical translocation within the cell. Electron microscopy on pancreatic B-cells has so far given little support for the existence of distinct pools of secretory granules. However, it has recently become possible to visualize the movements of secretory granules within intact insulinoma cells which contain fewer secretory granules than normal pancreatic B-cells. These studies have demonstrated that ACh-evoked Ca²⁺ mobilization from internal stores accelerated granule movements (Hisatomi et al. 1996). These movements were organized and directed towards a certain part of the cell and were dependent on activation of Ca²⁺-calmodulin-dependent phosphorylation of myosin light chains (Niwa et al. 1998). This indicates, that physical translocation of secretory granules does take place in the intact B-cell. However, the quantitative importance of such granular movements remains to be established. In chromaffin cells evanescent-wave fluorescence microscopy indicates that most of the functional replenishment of the RRP is attributable to recruitment of granules already situated in the vicinity of the release sites (Steyer et al. 1997).