Tolbutamide stimulates exocytosis of glucagon by inhibition of a mitochondrial-like ATP-sensitive K⁺ (K_ATP) conductance in rat pancreatic A-cells

Abstract

1. Capacitance measurements were used to examine the effects of the sulphonylurea tolbutamide on Ca²⁺-dependent exocytosis in isolated glucagon-secreting rat pancreatic A-cells.

2. When applied extracellularly, tolbutamide stimulated depolarization-evoked exocytosis 4.2-fold without affecting the whole-cell Ca²⁺ current. The concentration dependence of the stimulatory action was determined by intracellular application through the recording pipette. Tolbutamide produced a concentration-dependent increase in cell capacitance. Half-maximal stimulation was observed at 33 μm and the maximum stimulation corresponded to a 3.4-fold enhancement of exocytosis.

3. The stimulatory action of tolbutamide was dependent on protein kinase C activity. The action of tolbutamide was mimicked by the general K⁺ channel blockers TEA (10 mm) and quinine (10 μm). A similar stimulation was elicited by 5-hydroxydecanoate (5-HD; 10 μm), an inhibitor of mitochondrial ATP-sensitive K⁺ (K_ATP) channels.

4. Tolbutamide-stimulated, but not TEA-induced, exocytosis was antagonized by the K⁺ channel openers diazoxide, pinacidil and cromakalim.

5. Dissipating the transgranular K⁺ gradient with nigericin and valinomycin inhibited tolbutamide- and Ca²⁺-evoked exocytosis. Furthermore, tolbutamide- and Ca²⁺-induced exocytosis were abolished by the H⁺ ionophore FCCP or by arresting the vacuolar (V-type) H⁺-ATPase with bafilomycin A₁ or DCCD. Finally, ammonium chloride stimulated exocytosis to a similar extent to that obtained with tolbutamide.

6. We propose that during granular maturation, a granular V-type H⁺-ATPase pumps H⁺ into the secretory granule leading to the generation of a pH gradient across the granular membrane and the development of a positive voltage inside the granules. The pumping of H⁺ is facilitated by the concomitant exit of K⁺ through granular K⁺ channels with pharmacological properties similar to those of mitochondrial K_ATP channels. Release of granules that have been primed is then facilitated by the addition of K⁺ channel blockers. The resulting increase in membrane potential promotes exocytosis by unknown mechanisms, possibly involving granular alkalinization.

In a recent report (Bokvist et al. 1999), we demonstrated the presence of sulphonylurea receptors and ATP-sensitive K⁺ (K_ATP) channels in rat pancreatic A-cells. Inhibition of K_ATP channel activity may account for the previously reported stimulatory effects of the sulphonylureas on glucagon secretion (Grodsky et al. 1977; Efendic et al. 1979). Closure of the K_ATP channels leads to membrane depolarization and opening of voltage-dependent N- and L-type Ca²⁺ channels, which culminates in the initiation of Ca²⁺-dependent exocytosis (Gromada et al. 1997). Interestingly, high-affinity sulphonylurea-binding sites are not only found in the plasma membrane of rat A-cells. As in the insulin-secreting B-cells, they also associate with glucagon-containing granules (Carpentier et al. 1986). The role of the granular sulphonylurea-binding sites is not known. However, it is tempting to speculate that they, by analogy to what appears to be the case in insulin-secreting mouse pancreatic B-cells (Eliasson et al. 1996; Barg et al. 1999; Smith et al. 1999), participate in the regulation of glucagon secretion by interaction with the exocytotic machinery. In this paper we have investigated this aspect and provide circumstantial evidence for the participation of a granular mitochondrial-like K_ATP channel in the control of exocytosis in the glucagon-secreting pancreatic A-cells.

METHODS

Preparation of single rat A-cells and pituitary somatotrophs

Male Lewis rats (250–300 g; Møllegaard, Lille Skensved, Denmark) were anaesthetized with pentobarbital (100 mg kg⁻¹i.p.) and killed by decapitation. The use of animals was approved by the local ethical committee for animal studies. The pancreatic duct was ligated distally and injected with an ice-chilled solution of 600 U ml⁻¹ collagenase, 5 μg ml⁻¹ DNase and 5.6 mm glucose in Hepes-buffered saline solution (HBSS). The pancreas was removed and incubated for 2 × 4.5 min in a shaking water bath at 37°C (200 strokes min⁻¹; amplitude, 5 cm). The suspension was passed through a 14 gauge i.v. catheter, centrifuged and resuspended in the bottom layer of a discontinuous Ficoll gradient (13, 19.5, 20.5 and 24.5 %). After centrifugation for 15 min at 800 g at room temperature, the islets were recovered from the interfaces and washed in HBSS, and finally hand-picked under a stereomicroscope. The islets were stored at 4°C overnight in RPMI 1640 tissue culture medium (Gibco BRL, Life Technologies Ltd, Paisley, UK) with 10 % fetal calf serum. The islets were dispersed into single cells using dispase and pancreatic A-cells were separated by fluorescence-activated cell sorting as described elsewhere (Josefsen et al. 1996). Based on the hormone contents and the glucose sensitivity of electrical activity, we estimate that the preparation contained > 80 % A-cells and < 3 % B-cells (Josefsen et al. 1996; Gromada et al. 1997). The cell suspension was plated on 35 mm diameter Petri dishes and incubated in a humidified atmosphere for up to 3 days in RPMI 1640 medium supplemented with 10 % (v/v) heat-inactivated fetal calf serum, 100 IU ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin.

Pituitary somatotrophs from albino male Sprague-Dawley rats (250 g; Møllegaard) were prepared according to Raun et al. (1998). The animals were killed by decapitation. The cell suspension contained ∼70 % pituitary cells. After isolation the cells were plated onto plastic Petri dishes and maintained for up to 2 days in Dulbecco's modified Eagle's medium supplemented with 25 mm Hepes, 4 mm glutamine, 0.075 % NaCO₃, 0.1 % non-essential amino acids, 2.5 % fetal calf serum, 3 % horse serum, 10 % fresh rat serum, 1 nm T₃ and 40 μg l⁻¹ dexamethasone (pH 7.30). All tissue culture media were obtained from Life Technologies.

Electrophysiology

Pipettes were pulled from borosilicate glass, coated with Sylgard near their tips and fire polished. When filled with pipette solutions, the electrodes had a resistance of 3–4 MΩ. Exocytosis was measured as increases in cell capacitance (Neher & Marty, 1982) using, except for Fig. 1, an EPC-9 patch-clamp amplifier (HEKA Elektronik, Lamprecht/Pfalz, Germany) and Pulse software (version 8.30; HEKA Elektronik). The interval between two successive points was 0.2 s and the measurements of cell capacitance were initiated < 5 s after establishment of the whole-cell configuration. In Fig. 1, changes in cell capacitance were elicited by voltage-clamp depolarizations to 0 mV from a holding potential of −70 mV using an EPC-7 patch-clamp amplifier (List Elektronik) and in-house software written in AxoBasic (Axon Instruments) as detailed elsewhere (Ämmäläet al. 1993). All experiments, except those in Fig. 1 in which the perforated-patch whole-cell configuration was used, were performed using the standard whole-cell recording mode. The volume of the recording chamber was 0.4 ml and the solution entering the bath (1.5–2 ml min⁻¹) was maintained at +33°C.

Figure 1. Tolbutamide stimulation of depolarization-induced exocytosis in rat A-cells.

A, Ca²⁺ current (I_Ca, middle) and membrane capacitance (C_m, bottom) increases evoked by 500 ms depolarizations from −70 to 0 mV (V_m, membrane potential; top) before and 2 min after the addition of 0.1 mm tolbutamide. B, time-dependent changes in exocytotic responses before (−2 and 0 min) and 2, 4 and 6 min after addition of 0.1 mm tolbutamide (•). No changes in the exocytotic capacity were observed over time in a parallel series of control experiments conducted in the absence of tolbutamide (○). C, histogram summarizing the integrated Ca²⁺ current (Q_Ca) in the absence (−) and presence (+) of tolbutamide (Tolb) before (0 min) and 2 min after addition of tolbutamide. The data are means ±s.e.m. of 5 experiments. *P < 0.01 vs. Control.

Solutions

The extracellular medium consisted of (mm): 138 NaCl, 5.6 KCl, 2.6 CaCl₂, 1.2 MgCl₂, 5 Hepes (pH 7.4 with NaOH) and 5 D-glucose. The extracellular solution used for the measurement of cell capacitance evoked by voltage-clamp depolarization contained (mm): 118 NaCl, 20 TEA-Cl, 5.6 KCl, 1.2 MgCl₂, 2.6 CaCl₂, 5 Hepes (pH 7.40 with NaOH) and 5 glucose. TEA-Cl was included in the medium to block the outward delayed rectifying K⁺ current, which otherwise obscures the smaller Ca²⁺ current (Rorsman & Trube, 1986). The pipette solution used for the infusion experiments consisted of (mm): 125 potassium glutamate, 10 KCl, 10 NaCl, 1 MgCl₂, 5 or 9 CaCl₂, 3 Mg-ATP, 10 EGTA and 5 Hepes (pH 7.15 with KOH). In the experiments where internal K⁺ was replaced with Cs⁺ (Fig. 5B), the pipette solution consisted of (mm): 125 caesium glutamate, 10 CsCl, 10 NaCl, 1 MgCl₂, 5 CaCl₂, 3 Mg-ATP, 10 EGTA and 5 Hepes (pH 7.15 with CsOH). The free Ca²⁺ concentration of the resulting buffers was estimated as 0.22 or 2 μm using the binding constants of Martell & Smith (1971). For perforated-patch whole-cell experiments (Fig. 1), the pipette solution was composed of (mm): 76 Cs₂SO₄, 10 KCl, 10 NaCl, 1 MgCl₂ and 5 Hepes (pH 7.35 with CsOH). Electrical contact 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 series conductance was stable and > 35–40 nS. Pinacidil, glibenclamide, glipizide, diazoxide and tolbutamide were obtained from RBI (Natick, MA, USA). Rp-cAMPS (R_P-cyclic 3′,5′-hydrogen phosphorothioate adenosine triethylammonium salt) was from Biolog (Hamburg, Germany). Bafilomycin A₁ was obtained from Calbiochem (La Jolla, CA, USA). All other chemicals were purchased from Sigma. In the experiments utilizing the PKA and PKC inhibitors Rp-cAMPS and staurosporine, the cells were first pre-incubated for > 20 min with 0.1 mm Rp-cAMPS or 0.1 μm staurosporine. The protein kinase inhibitors were then included at the same concentrations in the pipette solution dialysing the cell interior during the electrophysiological measurements.

Figure 5. TEA and quinine stimulate exocytosis in rat A-cells.

A, increases in cell capacitance elicited by 0.22 μm free Ca²⁺ in the absence and presence of 10 mm TEA. B, mean rates of linear capacitance increase (ΔC_m/Δt) measured in the absence (Control) and presence of TEA (10 mm), with TEA + tolbutamide (both 0.1 mm), quinine (10 μm), TEA + diazoxide (10 and 0.1 mm), TEA + pinacidil (10 and 0.1 mm) and TEA + cromakalim (10 and 0.1 mm), and following equimolar replacement of K⁺ in the pipette solution with Cs⁺ alone or in combination with tolbutamide. Data are means ±s.e.m. of 5–8 individual cells under each experimental condition. *P < 0.01.

Data analysis

In the infusion experiments, the exocytotic rate was estimated by using a linear component of the capacitance increase lasting at least 30 s during the first 60 s following the establishment of the whole-cell configuration, ignoring any rapid changes occurring during the initial ∼10 s. Statistical significances were evaluated using Student's t test for unpaired data (Figs 1, 3, 5B and 8E and F), Dunnett's test for multiple comparisons to a single control (Figs 2, 4, 7 and 8A–D) and Duncan's test when comparing multiple means within a series of experiments (Figs 5 and 6). Data are quoted as mean values ±s.e.m. of the indicated number of experiments (= cells, n).

Figure 3. Tolbutamide stimulates exocytosis in a PKC-dependent fashion.

Changes in cell capacitance were elicited by intracellular dialysis of single rat A-cells with 0.22 μm free Ca²⁺ in the absence or presence of 0.1 mm tolbutamide. A and C, cells were treated with 0.1 μm staurosporine (A) or 0.1 mm Rp-cAMPS (C). B and D, the histogram summarizes mean linear rates of capacitance increase (ΔC_m/Δt) occurring in the absence and presence of tolbutamide in cells treated with staurosporine (B; Stauro) or Rp-cAMPS (D). Data are given as mean values ±s.e.m. of 4–5 experiments. *P < 0.01vs. control.

Figure 8. Tolbutamide- and Ca²⁺-induced exocytosis depend on the transgranular H⁺ gradient.

A, changes in cell capacitance elicited by 0.22 μm free Ca²⁺ in the absence or presence of 0.1 mm tolbutamide alone or in the simultaneous presence of tolbutamide and 0.1 mm FCCP. The histogram depicts the mean linear rates of capacitance increase (ΔC_m/Δt) in the absence and presence of tolbutamide or in the simultaneous presence of tolbutamide and FCCP. B, effects of FCCP on high free [Ca²⁺]_i (2 μm)-induced exocytosis. The histogram depicts the mean linear rate of capacitance increase in the presence or absence of FCCP. C, as in A except that FCCP was replaced with 5 μm bafilomycin A₁ (Baf A₁). D, effect of bafilomycin A₁ on high free [Ca²⁺]_i-induced exocytosis. E, oligomycin (1 μg ml⁻¹; Oligom) does not prevent the ability of tolbutamide to stimulate exocytosis in the presence of 0.22 μm free Ca²⁺. F, effects of NH₄⁺ (0.5 mm) on exocytosis elicited by 0.22 μm free Ca²⁺. Data are mean values ±s.e.m. of 3–11 individual cells. *P < 0.01.

Figure 2. Intracellular application of tolbutamide stimulates exocytosis in rat A-cells.

A, increases in cell capacitance observed during the first 2 min after establishment of the standard whole-cell configuration elicited by intracellular infusion with a Ca²⁺-EGTA buffer with a free Ca²⁺ concentration of 0.22 μm in the absence (Control) or presence of 0.1 mm tolbutamide in the pipette solution. Throughout the recording, the cell was clamped at −70 mV in order to avoid activation of the voltage-dependent Ca²⁺ channels that would otherwise interfere with the capacitance measurement. B, dose-response relationship for tolbutamide-induced stimulation of exocytosis. The rate of capacitance increase (ΔC_m/Δt) was measured during the linear component of capacitance increase as indicated by the dotted line in A. The line is the best fit of the mean data to the Hill equation. C, histogram depicting mean rates of increase in cell capacitance in the absence and presence of 0.1 mm tolbutamide, 100 nm glibenclamide or 100 nm glipizide. Data are mean values ±s.e.m. of 4–11 different experiments. *P < 0.01vs. Control.

Figure 4. K⁺ channel openers inhibit tolbutamide-induced exocytosis.

A, changes in cell capacitance when A-cells were dialysed with 0.22 μm free Ca²⁺ in the presence of 0.1 mm tolbutamide alone or in combination with diazoxide. B, mean rates of linear capacitance increase (ΔC_m/Δt) in the presence of tolbutamide alone (Tolb) and when diazoxide (dia), pinacidil (pin) or cromakalim (crom) were included in the pipette solution in the continued presence of the sulphonylurea. C, changes in cell capacitance when A-cells were infused with a maximally stimulatory free Ca²⁺ concentration (2 μm) in the absence and presence of diazoxide. D, mean rates of capacitance increase in the absence and presence of diazoxide, pinacidil or cromakalim. In all experiments the concentration of the K⁺ channel openers was 0.1 mm. The data are means ±s.e.m. of 3–11 different experiments. *P < 0.01vs. Tolb (B) or Control (D).

Figure 7. The K⁺ ionophore nigericin abolishes tolbutamide- and Ca²⁺-induced exocytosis.

A, increases in cell capacitance elicited by 0.22 μm free Ca²⁺ in the presence of tolbutamide (0.1 mm) alone or in combination with nigericin (10 μm). B, histogram depicting mean linear rates of capacitance increase (ΔC_m/Δt) in the presence of tolbutamide alone or in combination with nigericin or valinomycin (0.2 mm). C and D, as in A and B, except that exocytosis was stimulated by 2 μm free Ca²⁺ in the absence of tolbutamide. Data are means ±s.e.m. of 5 individual cells under each experimental condition. *P < 0.01vs. Tolb (B) or Control (D).

Figure 6. 5-Hydroxydecanoate stimulates exocytosis in rat A-cells.

A, increases in cell capacitance elicited by 0.22 μm free Ca²⁺ in the absence (Control) and presence of 10 μm 5-hydroxydecanoate (5-HD). B, mean rates of linear capacitance increase (ΔC_m/Δt) measured in the absence and presence of 5-HD (10 μm) and in the simultaneous presence of 5-HD (10 μm) and tolbutamide (0.1 mm), diazoxide (0.1 mm) or TEA (10 mm). Data are means ±s.e.m. of 4–10 individual cells under each experimental condition. *P < 0.01.

RESULTS

Tolbutamide potentiates depolarization-induced exocytosis in rat A-cells

Figure 1A illustrates whole-cell Ca²⁺ currents and the associated changes in cell capacitance elicited by 500 ms depolarizations from −70 to 0 mV in an intact rat A-cell using the perforated-patch configuration. Under basal conditions, the integrated Ca²⁺ current amounted to 6.9 pC and a capacitance increase of 31 fF was evoked. The latter value corresponds to the discharge of 16 granules using a conversion factor of 2 fF per granule (Gromada et al. 1997). Two minutes after inclusion of 0.1 mm tolbutamide in the bathing solution, the same membrane depolarization produced an integrated Ca²⁺ current of 6.7 pC and a capacitance increase of 81 fF (161 % stimulation). The effects of tolbutamide are summarized in Fig. 1B, which illustrates the amplitude of the exocytotic responses against the time elapsed after the addition of tolbutamide. It can be seen that tolbutamide produced a 234 ± 16 % (P < 0.01; n = 5) stimulation of exocytosis within 2 min. For comparison, the exocytotic responses in cells not stimulated by tolbutamide are shown. It can be seen that there were no spontaneous changes of the exocytotic capacity that could contribute to the observed stimulation by tolbutamide. The stimulatory action of tolbutamide on exocytosis was not associated with a change of the integrated Ca²⁺ current (Fig. 1C).

Tolbutamide stimulates exocytosis evoked by intracellular infusion of Ca²⁺

It seems likely that the effects of tolbutamide on exocytosis are exerted following its uptake into the A-cell. It has previously been demonstrated in pancreatic B-cells that tolbutamide remains a potent stimulator of exocytosis when applied intracellularly through the patch electrode during standard whole-cell recordings (Barg et al. 1999). Subsequent experiments were therefore performed using the latter configuration. This has the additional advantage that the diffusion barrier represented by the plasma membrane is removed. Following establishment of the whole-cell configuration and infusion of a Ca²⁺ buffer with a free Ca²⁺ concentration of 0.22 μm, exocytosis was initiated and was observed as a gradual capacitance increase (Fig. 2A, Control). In general, the cell capacitance reached a new steady-state level within 2 min. It is clear that inclusion of 0.1 mm tolbutamide in the pipette solution exerted a strong stimulatory action on the increase in cell capacitance (Fig. 2A, Tolbutamide). On average, tolbutamide evoked a 3-fold stimulation of both the rate and total amplitude of the capacitance increase. The effect of tolbutamide on exocytosis was dependent on dose (Fig. 2B). No stimulation of exocytosis was observed with 1 μm tolbutamide. At higher concentrations, tolbutamide accelerated the increase in cell capacitance by 39–339 %. Approximating the data points to the Hill equation yielded values of the association constant (K_a) and co-operativity factor of 33 μm and 6.5, respectively. Maximal stimulation of exocytosis was seen at concentrations of tolbutamide ≥ 100 μm (Fig. 2B).

Figure 2C shows that the stimulatory action of tolbutamide on exocytosis was mimicked by the second-generation sulphonylureas glibenclamide and glipizide. When applied at a concentration of 100 nm, exocytosis was increased 3-fold by both compounds (P < 0.01; n = 4 for glibenclamide and n = 5 for glipizide). Accordingly, these compounds were ≤ 1000-fold more potent than the prototype sulphonylurea tolbutamide.

Tolbutamide stimulates PKC-dependent exocytosis

In pancreatic B-cells tolbutamide stimulates exocytosis by a protein kinase C (PKC)-dependent mechanism (Eliasson et al. 1996). This also applies to pancreatic A-cells as suggested by the failure of tolbutamide to be stimulatory following inhibition of PKC by staurosporine (0.1 μm for 20 min; Fig. 3A and B). By contrast, inhibition of PKA by Rp-cAMPS (0.1 mm) did not affect the ability of tolbutamide to stimulate an increase in cell capacitance (Fig. 3C). Under these experimental conditions, tolbutamide enhanced the exocytotic response 3-fold (P < 0.01; n = 5; Fig. 3D).

Diazoxide inhibits tolbutamide-induced exocytosis

Figure 4A shows that inclusion of the K⁺ channel opener diazoxide (0.1 mm) in the pipette solution antagonized tolbutamide-induced exocytosis. Diazoxide completely inhibited the tolbutamide-evoked increase in cell capacitance and the rate of capacitance increase in the presence of tolbutamide and diazoxide was not different from that observed under control conditions (absence of drugs). Other K_ATP channel openers such as pinacidil and cromakalim (both 0.1 mm) also showed an inhibitory action and reduced exocytosis to a similar extent (Fig. 4B). This contrasts with what appears to be the case in mouse B-cells (S. Barg, L. Eliasson & E. Renström, personal communication). The K⁺ channel openers (0.1 mm) also reduced exocytosis evoked by infusion of high [Ca²⁺] buffer (2 μm free Ca²⁺; Fig. 4C and D).

K⁺ channel blockers mimic the stimulatory action of tolbutamide on exocytosis

In the following experiments, we investigated possible mechanisms by which tolbutamide stimulates exocytosis. We first investigated whether other K⁺ channel blockers share the effects of the sulphonylureas. Figure 5A illustrates the finding that inclusion of the broad-spectrum K⁺ channel blocker TEA (10 mm) in the pipette solution elicited a stimulation of exocytosis similar to that observed with tolbutamide. On average, TEA accelerated the exocytotic response 2.4-fold (P < 0.01; n = 6; Fig. 5B). Tolbutamide was unable to produce a further stimulation when applied in the presence of 10 mm TEA (Fig. 5B) suggesting that the actions of tolbutamide and TEA converge on the same cellular mechanism, possibly a granular K⁺ conductance. Moreover, quinine (another widely used K⁺ channel blocker) likewise stimulated exocytosis 2.8-fold (P < 0.01; n = 5) when applied at a concentration of 10 μm. We also tested the effects of the K⁺ channel openers diazoxide, pinacidil and cromakalim on TEA-stimulated exocytosis; all compounds were ineffective (Fig. 5B). Finally, we investigated the consequences of substituting Cs⁺ for K⁺ in the pipette solution. As shown in Fig. 5B, this had little effect on either basal or tolbutamide-stimulated secretion.

Tolbutamide stimulates exocytosis by activation of a mitochondrial-like K_ATP channel

Figure 6A shows that inclusion of 5-hydroxydecanoate (5-HD; 10 μm), an inhibitor of mitochondrial K_ATP channels (Hu et al. 1999), in the pipette solution stimulated exocytosis to an extent comparable to that observed with tolbutamide. A similar increase in the rate of exocytosis was observed at a 10 times higher concentration of 5-HD (data not shown). Furthermore, no additive effect on the secretion rate was obtained when either tolbutamide (0.1 mm) or TEA (10 mm; Fig. 6B) was added in the simultaneous presence of 5-HD. This finding is simplest to explain if 5-HD blocks the same K⁺ conductance as tolbutamide or TEA. In keeping with this idea, diazoxide (0.1 mm) was capable of reversing the stimulatory action of 5-HD on exocytosis (Fig. 6B). Consistent with results reported by others (Jagger et al. 1993), these effects of 5-HD on exocytosis were not paralleled by an inhibitory action of plasma membrane K_ATP channel activity (not shown).

Dissipation of a granular K⁺ gradient inhibits tolbutamide-evoked exocytosis

The above data indicate that tolbutamide stimulates exocytosis by closure of a granular K⁺ conductance. The finding that replacement of cytoplasmic K⁺ with Cs⁺ had no effect on exocytosis further argues that these channels mediate the efflux of K⁺ from the interior of the granules to the cytoplasm. A role for a granular K⁺ conductance in the stimulatory effect of tolbutamide on exocytosis is also suggested by the finding that inclusion of the K⁺ ionophore nigericin (10 μm) in the pipette solution abolished tolbutamide-induced exocytosis (Fig. 7A and B). The same effect was obtained with 0.2 mm valinomycin (Fig. 7B). Interestingly, the K⁺ channel ionophores also reduced exocytosis evoked by infusion of high [Ca²⁺] buffer (2 μm free Ca²⁺; Fig. 7C and D).

It has recently been reported that DIDS-sensitive Cl⁻ channels participate in the stimulatory action of tolbutamide on Ca²⁺-induced exocytosis in mouse B-cells (Barg et al. 1999). We therefore explored whether such channels also participate in tolbutamide-evoked exocytosis in the A-cell. However, inclusion of 0.2 mm DIDS in the pipette solution dialysing the cell did not affect the ability of tolbutamide to stimulate exocytosis (not shown). The same negative finding was made with 0.1 mm NPPB, another Cl⁻ channel blocker.

Tolbutamide-induced exocytosis depends on a transgranular H⁺ gradient

The above data suggest that the K⁺ gradient across the granular membrane might be important for tolbutamide-induced exocytosis. Secretory granules characteristically have an acidic pH. In the next series of experiments we investigated the importance of an acidic intragranular pH for exocytosis. Inclusion of the H⁺ ionophore FCCP (0.1 mm) completely inhibited tolbutamide-evoked secretion (Fig. 8A). FCCP also reduced control secretion by > 50 %. This is consistent with the observation that FCCP (0.1 mm) also reduced secretion induced by high [Ca²⁺]_i (Fig. 8B).

Insulin-containing secretory granules possess a vacuolar-type H⁺-ATPase that maintains the granule interior at or below pH 6 (Hutton, 1982; Orci et al. 1986). To investigate the effect of the granular H⁺ pump on tolbutamide-evoked secretion, cells were incubated in the presence of the V-type H⁺-ATPase inhibitor bafilomycin A₁ (5 μm). Bafilomycin A₁ not only abolished the ability of tolbutamide to stimulate exocytosis (Fig. 8C) but also reduced exocytosis under control conditions as well as exocytosis stimulated by high free [Ca²⁺]_i (2 μm; Fig. 8D). Similar effects were observed with 0.1 mm DCCD, another V-type H⁺-ATPase inhibitor (Bowman et al. 1988; Yoshimori et al. 1991; data not shown). We can exclude the possibility that FCCP acts by interfering with mitochondrial ATP generation. Although mitochondrial ATP production undoubtedly is compromised by FCCP, this is unlikely to have any major effect on the cytoplasmic concentration of ATP under these experimental conditions, since this is determined by the supply of exogenous ATP through the recording electrode. Indeed, as shown in Fig. 8E, application of the mitochondrial inhibitor oligomycin (1 μg ml⁻¹) was without effect on tolbutamide-stimulated exocytosis. Similar results were obtained when cells were pre-treated for 15 min with 1 μg ml⁻¹ oligomycin. We point out that cytoplasmic Ca²⁺ is effectively buffered by the presence of a high intracellular EGTA concentration so the effects of the mitochondrial inhibitors cannot simply be attributed to release of mitochondrial Ca²⁺.

The above data suggest that granular pH is important for controlling exocytosis. Previous experiments have demonstrated that addition of weak bases neutralizes granular pH and results in swelling of the granules due to the existence of an active accumulation mechanism for H⁺ across the granular membrane (Pace & Sachs, 1982). Interestingly, addition of the weak base ammonium (NH₄⁺, 0.5 mm) to the pipette solution produced a strong, but transient increase in the rate of exocytosis (Fig. 8F).

Effects of sulphonylureas on exocytosis in pituitary somatotrophs

To test whether the action of the sulphonylureas on exocytosis is confined to islet cells we investigated the effect of tolbutamide on exocytosis in rat pituitary somatotrophs. In a series of five experiments, 0.1 mm tolbutamide in the pipette solution increased the exocytotic responses 1.4-fold (P < 0.01) from 7.8 ± 0.9 fF s⁻¹ under control conditions to 18.8 ± 1.7 fF s⁻¹ in the presence of the sulphonylurea (Table 1). Glibenclamide was as effective as tolbutamide in stimulating exocytosis (Table 1). The stimulatory action of tolbutamide could be fully antagonized by 0.1 mm diazoxide (P < 0.01; n = 5), pinacidil (P < 0.01; n = 4) and cromakalim (P < 0.01; n = 4). In accordance with the data obtained in A-cells, TEA (10 mm) and 5-HD (10 μm) stimulated the rate of exocytosis 1.6-fold (P < 0.01 for both conditions). No effect was observed by including DIDS (0.2 mm) in the pipette solution (Table 1).

Table 1. Tolbutamide stimulates exocytosis in rat pituitary cells.

Condition	ΔCm/Δt(fF s−1)	n
Control	7.8 ± 0.9	5
Tolbutamide	18.8 ± 1.7*	5
Glibenclamide	16.6 ± 2.3*	5
5-Hydroxydecanoate	20.5 ± 2.1*	4
TEA	21.6 ± 2.2*	5
Tolbutamide + DIDS	21.2 ± 1.0	5
Tolbutamide + diazoxide	7.2 ± 1.2†	5
Tolbutamide + pinacidil	6.6 ± 1.5†	4
Tolbutamide + cromakalim	8.4 ± 2.1†	4

Increases in cell capacitance were elicited by intracellular infusion of a pipette solution with a free Ca⁺² concentration of 0.22 μm. Throughout the recording, the cell was clamped at −70 mV in order to avoid activation of the voltage-dependent Ca²⁺ channels, which would otherwise interfere with the measurements. The rates of capacitance increase (ΔC_m/Δt) were measured over the first 60 s after establishment of the wholecell configuration (excluding the initial 10 s). The following concentrations were used: tolbutamide, diazoxide, pinacidil and cromakalim, 0.1 nm; glibenclamide, 100 nm TEA, 10 nm; 5 hydroxydecanoate, 10 μm and DIDS, 0.2 nm.

^*P < 0.01vs. Control

^†P < 0.01vs. Tolbutamide.

DISCUSSION

Using capacitance measurements of exocytosis, we demonstrate here that tolbutamide produces a 3-fold potentiation of Ca²⁺-dependent exocytosis in rat pancreatic A-cells. This effect is mimicked by second-generation sulphonylureas and is observable at therapeutic concentrations. Similar observations were made in rat somatotrophs (Table 1) suggesting that the processes we describe could have implications for exocytosis in general. Our data are at first glance similar to those previously documented in mouse pancreatic B-cells (Eliasson et al. 1996; Barg et al. 1999; Smith et al. 1999). However, the observations reported here make it likely that different mechanisms are involved, the most notable being the involvement of K⁺ channels rather than Cl⁻ channels.

Role for PKC in tolbutamide-induced exocytosis

We demonstrate that tolbutamide potentiates exocytosis by a mechanism dependent on PKC activity and that it is exerted at a level distal to a rise in [Ca²⁺]_i. This is consistent with the results previously obtained in mouse B-cells (Eliasson et al. 1996) as well as in MIN6 insulinoma cells (Tian et al. 1998). Recently it was demonstrated that tolbutamide releases mitochondrial Ca²⁺ and based on this observation it was argued that the resultant elevation of [Ca²⁺]_i leads to stimulation of PKC (Smith et al. 1999). However, it is difficult to see how this mechanism could possibly be operational under the conditions used in most of our experiments because Ca²⁺ was strongly buffered using 10 mm EGTA. Any Ca²⁺ released from intracellular stores would consequently be expected to be chelated before any activation of PKC had taken place.

Tolbutamide stimulates exocytosis by interference with a granular K⁺ conductance

Several features of the effects of the sulphonylureas on exocytosis are suggestive of an involvement of K_ATP channels with pharmacological features similar to those documented for the mitochondrial variety of this channel. For example, the sensitivity of exocytosis to pinacidil and 5-HD conforms to the properties expected for a mitochondrial-like K_ATP channel. Interestingly, mitochondrial K_ATP channels require PKC activity for maintenance of their activity (Sato et al. 1998). It is tempting to speculate that the PKC dependence of the stimulatory effect of tolbutamide on exocytosis is attributable to this mechanism. Moreover, the recent observation that pinacidil and cromakalim also reversed the enhancement of secretion by glibenclamide in PC12 cells (Taylor et al. 1999) suggests that similar channels also participate in the control of exocytosis in other cells exhibiting regulated secretion.

It seems reasonable to conclude that tolbutamide-stimulated exocytosis is associated with the closure of K⁺ channels, which mediate the exit of K⁺ from the granule interior to the cytoplasm. This proposal rests on the finding that replacement of cytoplasmic K⁺ with Cs⁺ does not interfere with the stimulatory action of tolbutamide. The effects of the sulphonylureas and the K⁺ channel openers may well be mediated by a granular sulphonylurea receptor, possibly related to that which is part of the plasma membrane K_ATP channels, but it seems unlikely that this is the mechanism by which TEA and quinine act. The effects of these K⁺ channel blockers, which are chemically unrelated to the sulphonylureas, are more probably mediated by direct binding to the K⁺ channel itself.

How does tolbutamide enhance exocytosis in A-cells?

As in other neuroendocrine cells, the secretory granules of the rat A-cell belong to at least two functionally distinct pools: the readily releasable and the reserve pools. In the A-cell, the readily releasable pool has been estimated to contain ∼50 granules (Gromada et al. 1997). This value corresponds to a capacitance increase of ≤ 100 fF. Since this value is much less than the change normally observed in the Ca²⁺-infusion experiments (500–1000 fF), we conclude that a significant part of the observed responses is attributable to the mobilization of new granules from the reserve pool.

Figure 9 summarizes a model that attempts to incorporate both the acidic intragranular pH and the involvement of tolbutamide-sensitive mitochondrial-like K_ATP channels in the granular membrane. Under basal conditions, the activity of the V-type H⁺-ATPase leads to granular uptake of protons, acidification and the generation of a positive membrane potential. The concomitant efflux of K⁺ through granular K_ATP channels ensures that the membrane potential does not become so positive as to constitute an insurmountable barrier for further H⁺ pumping. It should be emphasized that in a small structure like the granule (diameter, 200 nm; volume, ∼4 al), a change in pH of 2 units (i.e. between pH 7 and pH 5) corresponds to the net uptake of only 20 protons assuming no buffering. Although this assumption is clearly unrealistic, this estimate nevertheless illustrates that the net transport of only a few protons is required to account for the observed pH gradients. The exit of K⁺ is favoured by the positive interior of the granules. The initial K⁺ content of the granules we attribute to some sort (or a combination) of exchange process(es), e.g. a Na⁺–K⁺ symport or a H⁺–K⁺ antiport. However, we point out that because of the positive voltage inside the granules, the concentration of K⁺ inside the granules need not be particularly high. For example, assuming that the voltage difference between the cytoplasm and the granule interior is ∼50 mV (Hutton, 1989), and a cytosolic K⁺ concentration of 135 mm, it can be estimated from the Nernst equation that K⁺ will flow out of the granule provided the concentration is ≥ 20 mm.

Figure 9. Model summarizing the results.

See text for details. PCO, potassium channel opener; SUR, sulphonylurea receptor; Ψ, membrane potential.

This model is consistent with the fact that mature granules have an acidic interior. It seems possible that an acidic granule interior may be required not only for the final processing of the hormone but also to make the granules release competent. In this context it may be significant that vacuole acidification is required for the pairing of the SNARE proteins in yeast (Ungermann et al. 1999). If this also applies to the SNARE proteins involved in the fusion between the granules and plasma membrane, then it is easy to see the connection between an acidic interior of the granules and their preparation for release.

How do the effects of the sulphonylureas fit into this scenario? We hypothesize that addition of the sulphonylureas leads to the closure of the granular K⁺ channels. These channels are pharmacologically similar to the mitochondrial K_ATP channels in being sensitive to 5-HD and pinacidil but are different from the K_ATP channels in the plasma membrane. The reduction of K⁺ efflux makes the granular membrane potential more positive (compare effects on B-cell membrane potential by K_ATP channel closure). The increase in membrane potential prevents H⁺ pumping by the H⁺-ATPase. Consequently, one would expect the granule to alkalify in the presence of tolbutamide. Precisely how alkalinization promotes exocytosis remains unclear but it is certainly of interest that transient Ca²⁺-dependent pH jumps (∼1 pH unit alkalinization) have been reported to precede exocytosis in mucosal mast cells (Williams et al. 1999) and that infusion of low concentrations of NH₄⁺ stimulates exocytosis to the same extent and with a similar time course to that observed with tolbutamide (Fig. 8F). The concept predicts that addition of tolbutamide should prevent priming of new granules as it counteracts intragranular acidification. This may be the reason why exocytosis in the A-cell often plateaus within 60 s after establishment of the whole-cell configuration (see Fig. 2A). This we attribute to an interrupted supply of new granules for release. Clearly it is now important to determine how the granular pH varies during the different stages of the release processes and how K⁺ channel modulators affect it. We speculate that the granular K_ATP channels are only active when situated in the secretory granules and that they inactivate when the granular membrane is inserted into the plasma membrane, accounting for the lack of effect of these K_ATP channel blockers in the whole-cell recordings of K_ATP currents (Bokvist et al. 1999). Finally, the observation that the inhibitory actions of the K⁺ channel openers and ionophores were not restricted to tolbutamide-stimulated exocytosis and that they also suppressed exocytosis evoked by high [Ca²⁺]_i suggests that the tolbutamide-sensitive mechanism we outline here is not only important for pharmacological regulation of glucagon release but also forms part of the mechanism utilized physiologically for Ca²⁺-dependent exocytosis.