Modulation of calcium signals by intracellular pH in isolated rat pancreatic acinar cells

Abstract

1. We have investigated the interactions between intracellular pH (pH_i) and the intracellular free calcium concentration ([Ca²⁺]_i) in isolated rat pancreatic acinar cells. The fluorescent dyes fura-2 and BCECF were used to measure [Ca²⁺]_i and pH_i, respectively.

2. Sodium acetate and ammonium chloride (NH₄Cl) were used to acidify and alkalinize pH_i, respectively. Cytosolic acidification had no effect on [Ca²⁺]_i in resting pancreatic acinar cells, whereas cytosolic alkalinization released Ca²⁺ from intracellular stores.

3. Cytosolic acidification using either acetate or a CO₂-HCO₃⁻-buffered medium enhanced Ca²⁺ signals evoked by acetylcholine (ACh) and cholecystokinin (CCK). In contrast, both NH₄Cl and trimethylamine (TMA) inhibited Ca²⁺ signals during stimulation with either ACh or CCK. This inhibitory effect was also observed in the absence of extracellular Ca²⁺, and was therefore not due to changes in Ca²⁺ entry.

4. Calcium oscillations evoked by physiological concentrations of CCK were enhanced by cytosolic acidification and inhibited by cytosolic alkalinization.

5. In order to determine the effects of pH_i upon Ca²⁺ handling by intracellular Ca²⁺ stores, intraorganellar [Ca²⁺] was monitored using the low affinity Ca²⁺ indicator mag-fura-2 in permeabilized cells. Addition of NH₄Cl, which is expected to alkalinize intraorganellar pH, did not alter intraorganellar [Ca²⁺] in permeabilized cells, suggesting that changing intraorganellar pH does not release Ca²⁺ from intracellular stores. Addition of NH₄Cl or acetate also did not affect the rate of Ca²⁺ release induced by inositol 1,4,5-trisphosphate (InsP₃).

6. Modification of extraorganellar (‘cytosolic’) pH did not affect the rate of ATP-dependent Ca²⁺ uptake into stores, but did modify the rate of Ca²⁺ release evoked by submaximal concentrations of InsP₃. The rate of Ca²⁺ release was increased at more alkaline extraorganellar pHs. These results would suggest that manipulation of intraorganellar pH does not affect Ca²⁺ handling by the intracellular stores. In contrast, extraorganellar (‘cytosolic’) pH does affect InsP₃-induced Ca²⁺ release from the stores.

7. In conclusion, changes in intracellular pH in pancreatic acinar cells can profoundly alter cytosolic [Ca²⁺]. This may shed light on earlier observations whereby cell-permeant weak acids and bases can modulate fluid secretion in epithelia.

Alteration of pH_i using a wide range of weak acids, including HCO₃⁻, acetate, pyruvate and butyrate, has been shown to stimulate fluid and electrolyte transport in a number of epithelia (Ullrich, Radtke & Rumrich, 1971; Petersen, Wood, Schulze & Heintze, 1981; Case, Conigrave, Favaloro, Novak, Thompson & Young, 1982). In exocrine acinar cells, weak bases have been found to activate Ca²⁺-dependent K⁺ and Cl⁻ channels (Hayashi et al. 1992), and to modify the rate of Cl⁻ fluxes (Seagrave, Barker, Curry & Martinez, 1992). The mechanism(s) underlying these effects are unknown. However, one attractive possibility is the involvement of changes in cytosolic Ca²⁺ ([Ca²⁺]_i). Changes in [Ca²⁺]_i in acinar cells following receptor stimulation result in activation of luminal and basolateral membrane ion channels, and subsequent fluid secretion (Petersen, 1992). A large number of studies on a variety of cell types, including acinar cells, have illustrated interactions between changes in pH_i and Ca²⁺ signals. For example, cytosolic acidification increases [Ca²⁺]_i in collecting duct cells (Slotki, Schwartz & Alexander, 1993) and pancreatic acinar cells (Tsunoda, 1990). Furthermore, cytosolic alkalinization was also found to increase [Ca²⁺]_i in lymphocytes (Grinstein & Goetz, 1985), cultured smooth muscle cells (Siskind, McCoy, Chobanian & Schwartz, 1989), lacrimal acinar cells (Yodozawa, Speake & Elliott, 1997) and endothelial cells (Danthuluri, Kim & Brock, 1990).

In the present study we have investigated the relationship between pH_i and Ca²⁺ signalling in detail in isolated pancreatic acinar cells. We report that in resting cells, acetate has no effect on [Ca²⁺]_i, whereas NH₄Cl releases Ca²⁺ from intracellular stores. In contrast, cytosolic acidification during sustained or oscillatory Ca²⁺ signalling increases [Ca²⁺]_i, whereas cytosolic alkalinization inhibits Ca²⁺ signals. These effects appear to occur at the level of the intracellular Ca²⁺ store.

METHODS

Pancreatic acinar cell isolation

Small clusters of acinar cells were prepared from the pancreas of Sprague-Dawley rats, which were killed by fluothane inhalation. Isolated pancreata were digested with collagenase (680 U (g tissue)⁻¹) in physiological medium (see below for composition) supplemented with 0.12 mg ml⁻¹ soybean trypsin inhibitor (Sigma), 1% (w/v) bovine serum albumin (BSA), and an amino acid mixture (containing minimum essential amino acids; 2% v/v). Following 30 min incubation at 37°C, the tissue suspension was washed several times with enzyme-free medium and dissociated into a fine suspension of acini by repeatedly passing the suspension through a 5 ml plastic pipette tip. The resulting cell suspension was layered onto medium supplemented with 4% BSA, and centrifuged at 1000 g for 2 min. After discarding the supernatant and repeating the centrifugation step once more, cells were resuspended in 0.1% BSA medium. Cells were either used immediately or stored on ice until use.

Loading of fluorescent dyes

Portions (1 ml) of the cell suspension were incubated with one of the following dyes: 2 μM of the acetoxymethyl ester form of fura-2 (fura-2 AM) for 30 min at room temperature, 1 μM of the acetoxymethyl ester form of 2′,7′-bis(carboxyethyl)-5-(and-6)-carboxyfluorescein (BCECF AM) for 30 min at room temperature, or 5 μM mag-fura-2 AM for 45 min at 37°C. Loaded cells were then washed twice with 0.1% BSA containing-medium, and resuspended in physiological medium.

Fluorescence measurements

Dye fluorescence was measured using a system based on a Nikon Diaphot inverted microscope and was described previously (Berrie & Elliott, 1994). Dye-loaded acinar cells were allowed to adhere to a coverslip forming the base of a perfusion chamber and observed with a × 40 oil immersion lens (numerical aperture 1.3). Individual cells, or small clusters of cells were isolated from the remaining cells in view using an adjustable diaphragm. Cells were excited at wavelengths of 350 and 380 nm (emission wavelength, 500 ± 20 nm) for fura-2 and mag-fura-2, and 488 and 436 nm (emission wavelength, 526 ± 12.5 nm) for BCECF, using a filter wheel spinning at 40 Hz (Cairn Research Ltd, Faversham, Kent, UK).

Permeabilization of acinar cells

Mag-fura-2-loaded acinar cells were allowed to settle onto the glass coverslip and were then superfused with Ca²⁺-free ‘cytosolic’ medium containing (mm): 135 KCl, 1.2 KH₂PO₄, 1.39 MgCl₂, 0.5 EGTA, 0.5 HEDTA, 0.5 nitriloacetic acid (NTA) and 20 Hepes (pH 7.1 with KOH), supplemented with Streptolysin-O (SLO; 0.4 u ml⁻¹) to permeabilize the cells, as described in detail in van de Put & Elliott (1996). In those experiments where the effects of modifying extraorganellar pH on inositol 1,4,5-trisphosphate (InsP₃)-induced Ca²⁺ release were examined, the EGTA, HEDTA, and NTA Ca²⁺ buffers were replaced by 1.5 mm BAPTA, since the ability of BAPTA to bind Ca²⁺ is only very weakly dependent on pH (Tsien, 1980). The permeabilization process was followed on-line by monitoring the 350 nm fluorescence signal (near the isosbestic point for mag-fura-2, 360 nm). The 350 nm signal dropped significantly as the dye was lost from the cytosol (van de Put & Elliott, 1996). Acinar cells were subsequently perfused with Ca²⁺-uptake ‘cytosolic’ medium with the same composition as the Ca²⁺-free ‘cytosolic’ medium but with the addition of 0.175 mm CaCl₂. This gave free [Mg²⁺] and [Ca²⁺] of 0.9 mm and 0.2 μm, respectively (calculated according to Schoenmakers, Visser, Flik & Theuvenet, 1992). Calcium uptake into intracellular stores was initiated by superfusing the permeabilized cells with Ca²⁺-uptake ‘cytosolic’ medium containing 1 mm ATP. After a Ca²⁺ store loading period of 10 min permeabilized cells were stimulated with InsP₃.

Calibration of ratio values

In experiments measuring [Ca²⁺]_i, the uncalibrated 350:380 nm ratio signal is shown as an index of [Ca²⁺]_i, since absolute estimates of [Ca²⁺]_i were not routinely derived from the values of the 350:380 nm fluorescence ratio. However, a two-point calibration of the fura-2 signal was carried out on a number of pancreatic acinar cells (Grynkiewicz, Poenie & Tsien, 1985). Individual cells were initially treated with 1 μm thapsigargin and 1 μm ionomycin in Ca²⁺-free medium (containing 2 mm EGTA) to obtain fluorescence parameters for fura-2 under Ca²⁺-free conditions (R_min). The cells were subsequently superfused with thapsigargin and ionomycin in Ca²⁺-supplemented medium to obtain fluorescence parameters for Ca²⁺-saturated fura-2 (R_max). The values obtained were: R_max, 5.74 ± 0.58; R_min, 0.92 ± 0.03; and the ratio S_F,380/S_B,380, 6.34 ± 0.41 (all values n = 4); where S_F,380/S_B,380 refers to the fluorescence signals measured at 380 nm for Ca²⁺-free and Ca²⁺-saturated dye. The 350:380 nm ratio signal varied between 1.0 and 1.5 in resting cells, corresponding to an estimated [Ca²⁺]_i of between 15 and 120 nM (the K_d of fura-2 at 22°C was taken as 135 nM; Grynkiewicz et al. 1985). Submaximal concentrations of ACh (1–2 μm) typically raised the 350:380 fluorescence ratio to around 3.0, which corresponds to an estimated [Ca²⁺]_i of around 0.65 μm, while supramaximal ACh concentrations (10 μm) raised the 350:380 fluorescence ratio to around 4.5, corresponding to a [Ca²⁺]_i of around 2.5 μm.

Intracellular pH was calibrated from the 488:436 nm ratio signal using the high extracellular K⁺/nigericin technique (Thomas, Buchsbaum, Zimniak & Racker, 1979), as previously described (Yodozawa et al. 1997).

Effects of changes in pH_i on fura-2 and mag-fura-2

Since many of the experiments in the present paper involve treatments which simultaneously change both pH_i and [Ca²⁺]_i, it is necessary to consider whether changes in pH_i alter the behaviour of the Ca²⁺-sensing dyes. For mag-fura-2, a detailed in vitro study by Lattanzio & Bartschat (1991) found that the dye's K_d for Ca²⁺ was essentially unaltered by changes in pH over the range 5.5–7.5. For fura-2, the effects of pH on the dye's K_d have been extensively characterized both in solution and inside intact cells in a series of papers by Martínez-Zaguilán and co-workers (Martínez-Zaguilán, Martínez, Lattanzio & Gillies, 1991; Martínez-Zaguilán, Gurulé & Lynch, 1996a; Martínez-Zaguilán, Parnami & Lynch, 1996b). These authors showed that the K_d of fura-2 rises as pH becomes more acidic, with the effect being particularly marked at pH values below around 6.8 (see also Lattanzio, 1990). The values of pH_i recorded in the present study ranged from 8.0 to 6.4. However, in most experiments pH_i was between 7.8 and 7.0 (mean pH_i in unstimulated cells was around 7.3, while acetate acidified cells to a mean pH_i close to 7.0 and NH₄Cl alkalinized cells to a mean pH_i around 7.8; see Results). The data of Martínez-Zaguilán et al. (1995) show that the fura-2 K_d will vary only slightly over this pH range. Calculation from their data gives a change in the fura-2 K_d from 124 nM at pH 7.8 to 139 nM at pH 7.3 and 161 nM at pH 7.0.

Changes in the fura-2 K_d will lead to ‘apparent’ changes in [Ca²⁺]_i of equal magnitude. However, under most conditions in the present study these effects are small compared with the actual changes in 350:380 ratio/[Ca²⁺]_i recorded, and thus the actions of pH_i on the K_d of fura-2 do not materially alter any of the conclusions reached. To give a quantitative example, application of NH₄Cl caused a mean increase in the fura-2 ratio of about 0.4 ratio units in resting cells (see Results). Using the calibration parameters given above, and neglecting any effects of pH_i this corresponds to an increase in calculated [Ca²⁺]_i of at least 100%. The change in pH_i that accompanied this increase in [Ca²⁺]_i, an alkalosis of around 0.5 pH units, will cause the K_d of fura-2 to decrease by 10%, which will in turn cause an ‘apparent’ increase in [Ca²⁺]_i of around 10%. The contaminating effects of pH_i on the fura-2 K_d would thus account for 10% or less of the observed change in [Ca²⁺]_i in this situation. Overall, when cells are alkalinized with NH₄Cl, any increase in [Ca²⁺]_i measured using fura-2 will be overestimated by a maximum of around 10%, while any decrease in [Ca²⁺]_i will be underestimated by a similar amount. Conversely, when acetate is used to acidify the cell, any increase in [Ca²⁺]_i will be underestimated by a maximum of around 15%.

The experimental situation producing the largest change in pH_i in this study is wash-out of NH₄Cl, which typically reduced pH_i to values near 6.5–6.6. At a pH_i of 6.5 the fura-2 K_d is expected to be around 250 nM, almost doubled from 139 nM at pH_i 7.3 (Martínez-Zaguilán et al. 1996a). Therefore values of [Ca²⁺]_i recorded during the NH₄Cl wash-out phase (see Figs 2B, 7, 8 and 10) will almost certainly be substantially underestimated due to the accompanying reduction in fura-2 Ca²⁺ sensitivity. This effect can be expected to be most marked for wash-out following longer (6–7 min) exposures to NH₄Cl (Figs 2B and 10), where acid loading will be greater than following short (2–4 min) exposures (Figs 7 and 8).

Figure 2. The effect of acetate and NH₄Cl on [Ca²⁺]_i in resting pancreatic acinar cells.

Traces show records of the 350:380 nm fura-2 ratio signal as an index of [Ca²⁺]_i. A, the effect of 20 mm acetate; B, the effect of 20 mm NH₄Cl. A and B from two different cells.

Figure 7. Effect of NH₄Cl-induced cytosolic alkalinization on [Ca²⁺]_i during agonist stimulation.

These traces show the effect of NH₄Cl on ACh-evoked (A) and CCK-evoked (B) Ca²⁺ signals. The traces are from two different cells (both traces, n = 5).

Figure 8. Effect of NH₄Cl on Ca²⁺ entry.

A, the effect of NH₄Cl on [Ca²⁺]_i during Ca²⁺ entry induced by microsomal Ca²⁺-ATPase inhibition using thapsigargin (TG, 1 μm) and cyclopiazonic acid (CPA, 10 μm). B, NH₄Cl is applied during ACh stimulation in the absence of extracellular Ca²⁺. Traces shown are from two different cells.

Figure 10. Effect of cytosolic alkalinization on Ca²⁺ oscillations.

NH₄Cl was applied during a series of Ca²⁺ oscillations induced by 10 pm CCK.

Materials

The standard experimental medium contained (mm): 104 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 2 CaCl₂, 15 glucose and 25 Hepes. For Ca²⁺-free medium, Ca²⁺ was omitted and the medium also contained 0.5 mm EGTA. Media containing acetate, NH₄Cl or trimethylamine were made by equimolar replacement of 20 mm NaCl with sodium acetate, NH₄Cl, or trimethylamine. Experimental media were equilibrated with 100% O₂ prior to use, and pH was set to 7.4 with NaOH. The CO₂-HCO₃⁻-buffered medium was made by replacement of 25 mm NaHepes with 25 mm NaHCO₃, and was gassed continuously with 95% O₂-5% CO₂ to give a pH of 7.4. All experiments were carried out at room temperature (22–24°C).

Chemicals were obtained from the following suppliers: ACh, InsP₃, thapsigargin, cyclopiazonic acid and nigericin from Sigma; CCK-8 from Bachem UK Ltd; fluorescent dyes from Molecular Probes, Europe; Streptolysin-O from Difco. All other chemicals were of analytical grade.

Statistics

Results are given as means ±s.e.m. Tests for differences between two means were made using Student's t test.

RESULTS

Effects of acetate/ammonium chloride on intracellular pH

In order to establish the size of the pH_i changes evoked in experiments using acetate and NH₄Cl, pancreatic acinar cells were loaded with the H⁺-sensitive dye BCECF. The resting pH_i of pancreatic acinar cells in Hepes-buffered medium was found to be 7.33 ± 0.06 (n = 22), in good agreement with a previous study with BCECF-loaded pancreatic acinar cells (7.28; Muallem & Loessberg, 1990a). Addition of the weak acid acetate (20 mm) to the extracellular medium (extracellular pH constant at 7.4) decreased pH_i by 0.30 ± 0.02 pH units (n = 11) as shown in Fig. 1A. Intracellular pH recovery during acetate exposure was extremely slow, with a rate of only 0.002 ± 0.001 pH units min⁻¹ (n = 11). Application of a CO₂-HCO₃⁻-buffered medium acidified pH_i to a similar extent as 20 mm acetate (not shown; decrease in pH_i, 0.33 ± 0.04 pH units, n = 9).

Figure 1. The effect of acetate and NH₄Cl on intracellular pH.

Representative records of BCECF fluorescence calibrated as pH_i from two different pancreatic acinar cells. A, the effect of 20 mm acetate (extracellular pH, 7.4); B, the effect of 20 mm NH₄Cl.

Figure 1B shows that application of 20 mm NH₄Cl to pancreatic acinar cells resulted in a large alkaline shift in pH_i of 0.47 ± 0.04 pH units (n = 11). During exposure to NH₄Cl, there was some recovery towards resting pH_i (rate of reacidification, 0.03 ± 0.003 pH units min⁻¹, n = 11), probably resulting from the entry of NH₄⁺ (Roos & Boron, 1981). On removal of NH₄Cl, pH_i rapidly acidified as NH₄⁺ dissociated, releasing protons into the cytoplasm. Intracellular pH then slowly recovered towards resting pH_i at a rate of 0.016 ± 0.007 pH units min⁻¹ (n = 11). The slow pH_i recovery would suggest that the Na⁺-H⁺ exchanger is relatively inactive in unstimulated pancreatic acinar cells under these conditions. A different weak base, trimethylamine (20 mm), also alkalinized pH_i, although the pH_i change was less than with NH₄Cl (0.27 ± 0.02 pH units, n = 9; not shown).

Effects of intracellular pH change on [Ca²⁺]_i in resting pancreatic acinar cells

Figure 2A illustrates that acidifying the cytosol with acetate had no effect on cytosolic [Ca²⁺] in resting pancreatic acinar cells (n = 5). Cytosolic acidification with a CO₂-HCO₃⁻-buffered medium also had no effect on [Ca²⁺]_i (n = 4; not shown). In contrast, alkalinizing the cytoplasm with NH₄Cl transiently increased [Ca²⁺]_i (Fig. 2B). The mean change in ratio with 20 mm NH₄Cl was 0.41 ± 0.05 ratio units (n = 11). Similar changes in [Ca²⁺]_i were also noted with 20 mm trimethylamine (not shown; mean increase in ratio, 0.41 ± 0.07 pH units, n = 8). These observations are in contrast to an earlier study using pancreatic acinar cells, where addition of the weak acid propionate (20 mm) resulted in a small increase in [Ca²⁺]_i, while NH₄Cl failed to increase [Ca²⁺]_i (Tsunoda, 1990). Washout of NH₄Cl caused a slight but noticeable undershoot of the 350:380 nm ratio in approximately 50% of the cells studied (see e.g. Fig. 2B). However, this may represent an effect of the acidic pH_i on the fura-2 K_d rather than a real undershoot in [Ca²⁺]_i (see Methods for a discussion of this issue).

The ability of NH₄Cl to increase [Ca²⁺]_i in resting pancreatic acinar cells was examined in greater detail. In lacrimal acinar cells, NH₄Cl was suggested to release Ca²⁺ from the same internal store as agonists (Yodozawa et al. 1997). This possibility was explored in pancreatic acinar cells using a combination of NH₄Cl and ACh. Figure 3A illustrates that, following supramaximal stimulation with 10 μm ACh in a Ca²⁺-free medium, application of 20 mm NH₄Cl failed to release any additional intracellular Ca²⁺. In contrast, application of NH₄Cl prior to ACh stimulation resulted in clear intracellular Ca²⁺ release (Fig. 3B). No significant difference was observed between the magnitude of the ACh-induced increase in [Ca²⁺]_i in the presence and absence of NH₄Cl (change in ratio with ACh in the presence of NH₄Cl: 1.81 ± 0.11 ratio units, n = 6, compared with 1.79 ± 0.1 ratio units, n = 7, in the absence of NH₄Cl). These results suggest that ACh and NH₄Cl release Ca²⁺ from the same intracellular stores. The protocol adopted in Fig. 3B also demonstrates the ability of NH₄Cl to increase cytosolic Ca²⁺ in the absence of any extracellular Ca²⁺, implying that cytosolic alkalinization increases [Ca²⁺]_i via Ca²⁺ release from internal stores.

Figure 3. The effect of NH₄Cl on [Ca²⁺]_i under Ca²⁺-free conditions.

The effect of NH₄Cl following (A) and prior to (B) store depletion with 10 μm ACh. Traces shown are from two different cells and are each representative of five experiments.

Calcium release evoked by acetate during agonist stimulation

In pancreatic acinar cells, the secretagogues ACh and CCK induce both Ca²⁺ release from intracellular stores, via the generation of InsP₃, and subsequent Ca²⁺ entry from the external medium (Williams & Yule, 1993). In Fig. 4A and B, application of these agonists, at submaximal concentrations (ACh, 0.5 μm; CCK, 100–500 pm), resulted in a rapid increase in cytosolic [Ca²⁺]. When acetate was applied during the falling phase of the Ca²⁺ signal, there was a second transient increase in [Ca²⁺]_i, as previously observed in lacrimal acinar cells stimulated with ACh (Yodozawa, Grant & Elliott, 1994). Cytosolic acidification therefore enhanced agonist-evoked Ca²⁺ signals during stimulation with both ACh and CCK. The increase in [Ca²⁺]_i evoked by acetate varied considerably between different cells (compare Fig. 4A and B).

Figure 4. The effect of ACh/CCK and acetate on [Ca²⁺]_i.

These traces show the effect of acetate-induced cytosolic acidification during the falling phase of theACh-evoked (A) and CCK-evoked (B) Ca²⁺ signal. Traces shown are from two different cells and are each representative of five experiments.

Acetate could produce this increase in [Ca²⁺]_i by promoting Ca²⁺ entry from the external medium during agonist stimulation. This possibility was tested by repeating the protocol in Fig. 4A in the absence of extracellular Ca²⁺. In Fig. 5A, acetate application again resulted in a transient increase in [Ca²⁺]_i during ACh stimulation, arguing against any involvement of Ca²⁺ entry (n = 8). The transient increase in [Ca²⁺]_i evoked by acetate often resembled that of the initial phase of agonist stimulation (see Figs 5A and 4B), suggesting that the source of Ca²⁺ may be the intracellular store(s). The concentration effects of acetate were therefore investigated with respect to the degree of fullness of these stores. In Fig. 5B, the stores were maximally depleted using supramaximal ACh concentrations (10 μm), after which no second increase in [Ca²⁺]_i was observed upon acetate application (n = 11). The increase in [Ca²⁺]_i evoked by acetate during submaximal (2 μm) ACh stimulation (0.739 ± 0.15, n = 8; ratio change with ACh was 2.13 ± 0.31) was significantly larger (P < 0.0001) than during supramaximal (10 μm) ACh stimulation (0.05 ± 0.02, n = 11; ratio change with ACh was 3.39 ± 0.26). This would suggest that in agonist-stimulated pancreatic acinar cells the concentration effects of acetate are dependent upon the ‘fullness’ of the intracellular stores. The variable size of the acetate-evoked increase in [Ca²⁺]_i (compare Figs 4A, 4B, and 5A) thus probably reflects varying degrees of prior Ca²⁺ store emptying by agonists in different individual cells.

Figure 5. Effect of acetate on [Ca²⁺]_i during ACh stimulation.

These traces show the effect of acetate addition during submaximal (2 μm; A) and supramaximal (10 μm; B) ACh stimulation under Ca²⁺-free conditions. Traces shown are from two different cells.

It was important to ascertain if the acetate-evoked increase in [Ca²⁺]_i in agonist-stimulated cells resulted from a specific effect of acetate rather than an effect of intracellular acidification. This question was addressed using a CO₂-HCO₃⁻-buffered medium to acidify the cytosol (Roos & Boron, 1981). The Hepes-buffered medium was exchanged for a CO₂-HCO₃⁻-buffered medium during the plateau phase of agonist stimulation with 0.5 μm ACh. Figure 6 shows a representative experiment, illustrating that acidifying pH_i with CO₂ also increases cytosolic Ca²⁺ during agonist stimulation (n = 5). This therefore argues against any specific effects of acetate during the process of enhanced agonist-induced Ca²⁺ release.

Figure 6. Effect of cytosolic acidification on [Ca²⁺]_i during ACh stimulation.

Cytosolic acidification was achieved with a CO₂-HCO₃⁻-buffered medium (n = 5).

Calcium signals are inhibited by NH₄Cl during agonist stimulation

When NH₄Cl was applied to cells stimulated with submaximal concentrations of ACh or CCK (1 μm and 500 pm respectively), there was a strikingly rapid decrease in [Ca²⁺]_i back to the prestimulation level (Fig. 7A and B). Removal of NH₄Cl, increased [Ca²⁺]_i back to, or close to, the peak level. Indeed, this [Ca²⁺]_i‘overshoot’ may be substantially underestimated due to the increase in the K_d of fura-2 under these conditions. A comparable degree of inhibition of Ca²⁺ signals was observed upon application of 20 mm trimethylamine during ACh stimulation (not shown; n = 6).

The second, sustained phase of Ca²⁺ signalling during agonist stimulation of non-excitable cells is generally attributed to Ca²⁺ entry from the external medium (Putney, 1986). The inhibitory effect exerted by NH₄Cl during agonist stimulation in Fig. 7A and B might therefore reflect inhibition of Ca²⁺ entry from the external medium. This possibility was examined under conditions where Ca²⁺ entry could be studied in isolation, achieved using the microsomal Ca²⁺-ATPase pump inhibitors thapsigargin and cyclopiazonic acid (Mason, Garcia-Rodriguez & Grinstein, 1991). These inhibitors release Ca²⁺ from intracellular stores in the absence of any changes in inositol polyphosphate levels (Mason et al. 1991). As the stores become progressively depleted, Ca²⁺ entry is stimulated, i.e. capacitative or depletion-activated Ca²⁺ entry occurs (Putney, 1986). Surprisingly, 1 μm thapsigargin alone was unable to release Ca²⁺ from intracellular stores in our hands. Intracellular stores were effectively depleted only when using a combination of 1 μm thapsigargin and 10 μm cyclopiazonic acid, resulting in the observed increase in [Ca²⁺]_i in Fig. 8A. During the period of sustained raised [Ca²⁺]_i, the cytosol was alkalinized with NH₄Cl. An inhibitory effect of NH₄Cl alkalinization on Ca²⁺ entry would be expected to decrease [Ca²⁺]_i. However, this manoeuvre had no effect on [Ca²⁺]_i (n = 5). Therefore the decrease in [Ca²⁺]_i observed during agonist stimulation in Fig. 7A and B is most likely not due to impaired Ca²⁺ entry.

The experiment illustrated in Fig. 7A was also repeated in the absence of extracellular Ca²⁺. Figure 8B shows that cytosolic alkalinization with NH₄Cl still rapidly decreased [Ca²⁺]_i during stimulation with 1 μm ACh (n = 6). This observation reinforces the suggestion that inhibition of Ca²⁺ signals by NH₄Cl is not due to impaired Ca²⁺ entry.

A smaller return of ACh-induced response was observed upon removal of NH₄Cl in Ca²⁺-free medium (Fig. 7A) than in Ca²⁺-containing medium (Fig. 8B) even though the changes in pH_i should be very similar. This may reflect a lack of Ca²⁺ entry into the cell in the absence of extracellular Ca²⁺. Alternatively, this observation may be explained by greater store depletion under these conditions, since the cytosolic Ca²⁺ stores cannot be replenished in the absence of extracellular Ca²⁺.

Calcium oscillations and intracellular pH modification

In a number of cells, submaximal concentrations of CCK (10–150 pm) were used to induce Ca²⁺ oscillations. The exact CCK concentration required to elicit oscillations varied, probably reflecting both cell-to-cell variability and slight day-to-day variation in cell responsiveness. Figure 9A shows a typical experiment, where addition of 20 pm CCK elicited a series of small transient oscillations, or Ca²⁺ spikes, arising from baseline [Ca²⁺]_i. When acetate was applied during stimulation with CCK, there was a large increase in the amplitude of the Ca²⁺ spikes (n = 5). This enhancement of amplitude persisted for the period of cytosolic acidification associated with acetate application. Upon removal of acetate, but in the continued presence of CCK, Ca²⁺ signals were inhibited for approximately 8 min in this particular cell.

Figure 9. Effect of cytosolic acidification on Ca²⁺ oscillations.

The effect of 20 mm acetate (A) and CO₂-HCO₃⁻-buffered medium (B) on oscillations evoked by 20 pm and 100 pm CCK, respectively. Traces shown are from two different cells.

Similar observations were made using a CO₂-HCO₃⁻-buffered medium to acidify cytosolic pH during a series of Ca²⁺ oscillations. Figure 9B shows an experiment where application of CO₂ increased cytosolic Ca²⁺ above the amplitude of the spikes induced by 100 pm CCK alone. The elevated [Ca²⁺]_i persisted for the duration of CO₂ addition (n = 5). In three out of five cells, Ca²⁺ oscillations were inhibited for several minutes following removal of CO₂ from the extracellular medium. In the remaining two cells, spike amplitude was reduced upon return to a Hepes-buffered medium.

Experiments were also performed using NH₄Cl to alkalinize pH_i. Figure 10 shows an experiment in which regular large transient oscillations in [Ca²⁺]_i were produced using 10 pm CCK. Addition of NH₄Cl to the extracellular medium abolished the oscillations, and [Ca²⁺]_i transiently increased above the basal level (n = 5). Upon removal of NH₄Cl, the oscillations in [Ca²⁺]_i resumed. The first spike following removal of NH₄Cl took noticeably longer to return to baseline levels of [Ca²⁺]_i. The degree to which the basal [Ca²⁺] increased during NH₄Cl exposure, and the enhancement of Ca²⁺ spikes upon NH₄Cl removal, varied between the five individual cells tested, but the general pattern seen in Fig. 10 was observed in all five cells.

Calcium handling by the intracellular stores

The above data suggest that the application of acetate, and the subsequent cytosolic acidification, increases [Ca²⁺]_i in agonist-stimulated pancreatic acinar cells by an action at the level of the intracellular Ca²⁺ store. The intracellular store(s) involved are presumably those sensitive to InsP₃, as the acetate effect was only observed in the presence of an InsP₃-generating secretagogue. The situation with NH₄Cl is more complex, in that cytosolic alkalinization increased [Ca²⁺]_i in resting cells, whereas during agonist stimulation Ca²⁺ signals were inhibited. However, the increase in cytosolic [Ca²⁺] generated by cytosolic alkalinization in resting cells is clearly due to Ca²⁺ release from intracellular stores, since removal of extracellular Ca²⁺ did not inhibit the [Ca²⁺]_i response to NH₄Cl. Modification of pH_i could alter Ca²⁺ handling by the intracellular stores during agonist stimulation by acting on either the InsP₃-receptor Ca²⁺ release channel or the microsomal Ca²⁺-ATPase pump. In order to assess whether pH_i alters these processes we used the low-affinity Ca²⁺ indicator mag-fura-2 to monitor intraorganellar [Ca²⁺] in permeabilized cells. We examined the affects of altering both intraorganellar pH (using acetate and NH₄Cl) and extraorganellar pH (using medium of various pHs) upon ATP-driven Ca²⁺ uptake and InsP₃-induced Ca²⁺ release.

Intraorganellar pH and InsP₃-induced calcium release

The resting mag-fura-2 ratio of the permeabilized cells was 0.94 ± 0.02 (n = 52). Addition of ATP (1 mm) to the superfusate resulted in an increase in the 350:380 nm ratio (Fig. 11), indicating an increase in organellar [Ca²⁺] as a result of ATP-dependent Ca²⁺ uptake. The actual change in ratio with the addition of ATP was an increase of 0.74 ± 0.04 ratio units (n = 26). A steady-state level of Ca²⁺ uptake was attained within 4–5 min. We first attempted to examine the effects of changing intraorganellar pH on Ca²⁺ handling by the Ca²⁺ stores. The Ca²⁺ stores loaded by mag-fura-2 in permeabilized pancreatic acinar cells are located in the endoplasmic reticulum (ER; van de Put & Elliott, 1996, 1997). Although there are no direct measurements of ER lumen pH in intact cells, the pH within all vesicular subcompartments is thought to be acidic with respect to the cytosol (see e.g. Anderson & Pathak, 1985, for references), and recent direct measurements of pH inside the Golgi apparatus, the organelle most closely related to the ER, show a pH of 6.2–6.5 (Seksek, Biwersi & Verkman, 1995; Kim, Lingwood, Williams, Furuya, Manolson & Grinstein, 1996). The pH gradients between acidic organellar subcompartments and the cytosol can be disrupted by addition of permeant weak bases (e.g. methylamine, NH₄Cl, chloroquine), which accumulate in the acidic compartment and dissipate the pH gradient (see e.g. Niederau, Van Dyke, Scharschmidt & Grendell, 1986; De Lisle & Williams, 1987; Kim et al. 1996). However, when permeabilized cells were perfused with NH₄Cl (Fig. 11) or acetate (both 20 mm, medium pH constant at 7.1), there was no change in the ratio signal and hence in intraorganellar Ca²⁺ content (acetate, n = 6; NH₄Cl, n = 6). Figure 11 also shows that addition of 0.5 μm InsP₃ caused a large drop in the mag-fura-2 fluorescence ratio, demonstrating that the Ca²⁺ stores in the permeabilized cells were capable of releasing Ca²⁺ in response to stimuli.

Figure 11. Effect of ATP, NH₄Cl and IP₃ on the Ca²⁺ content of intracellular Ca²⁺ stores.

The trace shows the 350:380 nm mag-fura-2 fluorescence ratio in a cluster of Streptolysin-O permeabilized pancreatic acinar cells loaded with mag-fura-2.

We went on to examine the rate of InsP₃-induced Ca²⁺ release in combination with changing intraorganellar pH using acetate and NH₄Cl. A final concentration of 0.5 μm InsP₃ was added together with either 20 mm acetate or 20 mm NH₄Cl to alter intraorganellar pH. The initial rates of InsP₃-induced Ca²⁺ release were compared in the presence and absence of acetate/NH₄Cl. The rate of Ca²⁺ release obtained with 0.5 μm InsP₃ alone (0.46 ± 0.09 ratio units min⁻¹, n = 11) was not significantly different from the rate obtained in the presence of acetate (0.33 ± 0.08 ratio units min⁻¹, n = 10; P = 0.33 by an un-paired t test) or NH₄Cl (0.34 ratio units min⁻¹± 0.04, n = 11; P = 0.27 by unpaired t test). These results suggest that modifying intraorganellar pH does not affect either the steady-state intraorganellar [Ca²⁺] or the rate of InsP₃-induced Ca²⁺ release.

Extraorganellar pH and InsP₃-induced calcium release

Three different medium pH values were selected to examine the effects of extraorganellar (cytosolic) pH on ATP-dependent Ca²⁺ uptake and InsP₃-induced Ca²⁺ release - pH 6.7, 7.1 (control) and 7.6. Medium pHs of 6.7 and 7.6 were selected to reflect the changes in pH_i observed in the earlier experiments on applying acetate and NH₄Cl respectively to pancreatic acinar cells. The permeabilized acinar cells were continuously superfused with experimental medium at a given pH following permeabilization and during application of ATP (1 mm) and InsP₃ (0.5 μm).

There was no significant difference in the extent of ATP-dependent Ca²⁺ uptake by the intracellular stores (as judged by the steady-state change in ratio) at the different pHs. Pooling the data from all three pHs, the actual change in ratio with the addition of ATP was an increase of 0.76 ± 0.04 ratio units (n = 24). As described above (see Fig. 11), a steady-state level of Ca²⁺ uptake was achieved within 4 min, and subsequent application of InsP₃ rapidly released Ca²⁺ from the intracellular stores.

The rates of Ca²⁺ uptake and release at the different medium pHs are given in Table 1. The rate of Ca²⁺ uptake varied markedly between different cell clusters at any given value of pH, and mean Ca²⁺ uptake rates at each of the medium pHs were not significantly different from one another. In contrast, rates of InsP₃-induced Ca²⁺ release were fairly consistent between different cell clusters. Table 1 gives averaged data indicating that the rate of InsP₃-induced Ca²⁺ release increased with increasing extraorganellar pH. The rate of Ca²⁺ release at pH 6.7 was significantly slower than at pH 7.1 (P < 0.05, unpaired t test) and pH 7.6 (P < 0.02, unpaired t test). There was no significant difference between the rate of Ca²⁺ release at pH 7.1 and pH 7.6. In contrast to the above results, when permeabilized cells were treated with a fivefold higher concentration of InsP₃ (2.5 μm, a near-maximal concentration; see van de Put & Elliott, 1996), altering the pH had no effect on the rate of Ca²⁺ release (data not shown). This suggests that changing extraorganellar pH alters the sensitivity of the InsP₃ receptor to InsP₃ in pancreatic acinar cells.

Table 1.

Results of the study comparing the rate of Ca²⁺ uptake and release induced by 1 mm ATP and 0.5 μm InsP₃, respectively, in single permeabilized pancreatic acinar cells

Medium pH	Rate of Ca2+ uptake (t1/2) (s)	Initial rate of Ca2+ release (ratio units min−1)
6.7	246 ± 92 (7)	0.28 ± 0.08 (8)*
7.1	403 ± 167 (6)	1.55 ± 0.56 (6)
7.6	575 ± 127 (9)	2.93 ± 1.08 (6)

Assessments of uptake and release rates were carried out at different medium pHs - 6.7, 7.1 (control) and 7.6. The rate of Ca²⁺ uptake was analysed by fitting a mono-exponential time course by non-linear regression to the data. The rate of Ca²⁺ release was not well fitted by a mono-exponential time course and is therefore given as the initial rate of change of the 350:380 ratio/Δtime (minutes).

^*Significantly different from rate at pH 7.1 and 7.6. Values of n given in parentheses.

DISCUSSION

The data described above illustrate that there is a relationship between pH_i and [Ca²⁺]_i in pancreatic acinar cells both under conditions of sustained Ca²⁺ signalling and during Ca²⁺ oscillations. In the present study, acetate application evoked a cytosolic acidification but had no effect on [Ca²⁺]_i in unstimulated pancreatic acinar cells. In contrast, acidification of pH_i during agonist stimulation resulted in a second increase in [Ca²⁺]_i, as previously observed in lacrimal acinar cells (Yodozawa et al. 1994). In unstimulated cells, the weak base NH₄Cl transiently increased cytosolic [Ca²⁺] by mobilizing Ca²⁺ from internal stores. However, cytosolic alkalinization with NH₄Cl/trimethylamine dramatically inhibited Ca²⁺ signalling in the presence of secretory agonists. In addition, CCK-induced Ca²⁺ oscillations were enhanced by cytosolic acidification, and inhibited by cytosolic alkalinization. Therefore the effects of changing pH_i upon [Ca²⁺]_i in pancreatic acinar cells are essentially opposite depending upon whether the cells are resting or simultaneously stimulated with secretory agonists.

Ca²⁺ release in unstimulated cells

Application of NH₄Cl to resting pancreatic acinar cells induced a transient increase in [Ca²⁺]_i. Similar observations have been made in a number of other cell types (Siskind et al. 1989; Danthuluri et al. 1990), including our recent study in lacrimal acinar cells (Yodozawa et al. 1997). The resulting increase in [Ca²⁺]_i in resting cells was much smaller than that evoked by ACh stimulation and also occurred in the absence of extracellular Ca²⁺, indicating that intracellular stores act as the source of Ca²⁺ (Fig. 3B). As in lacrimal acinar cells, NH₄Cl appears to release Ca²⁺ from the same intracellular stores released by ACh, since application of NH₄Cl had no effect on [Ca²⁺]_i when the agonist-sensitive intracellular Ca²⁺ stores had been depleted with maximal concentrations of ACh under Ca²⁺-free conditions (Fig. 3A).

The Ca²⁺-releasing effect of alkalinization in resting cells might be explained by an increase in InsP₃ concentration above basal levels. In lacrimal acinar cells, NH₄Cl increased total inositol phosphate production, suggesting that NH₄Cl may have released Ca²⁺ via an increase in InsP₃ concentration (Yodozawa et al. 1997). However, intracellular alkalinization can also release Ca²⁺ from agonist-sensitive stores without altering basal InsP₃ levels, e.g. in bovine aortic endothelial cells (Danthuluri et al. 1990). Our experiments in permeabilized cells clearly show that cytosolic alkalinization would be expected to enhance the rate of Ca²⁺ release for a given intracellular InsP₃ concentration.

Ca²⁺ entry

Calcium entry has previously been reported to be sensitive to both pH_o and pH_i (Muallem, Pandol & Beeker, 1989). In pancreatic acinar cells, Tsunoda, Stuenkel & Williams (1990) found that intracellular alkalinization or acidification during Ca²⁺ influx, respectively, inhibited or enhanced Ca²⁺ influx. In the present study cytosolic acidification during agonist stimulation induced a secondary transient increase in [Ca²⁺]_i in pancreatic acinar cells. If acidification/H⁺ were enhancing Ca²⁺ entry, one might expect the resulting increase in [Ca²⁺]_i to be of a more sustained (rather than transient) nature, given that H⁺ extrusion during acetate exposure is relatively slow (Fig. 1A and B).

In both pancreatic (this study) and lacrimal acinar cells (Yodozawa et al. 1997), NH₄Cl application during prolonged agonist stimulation produced a rapid reduction in [Ca²⁺]_i. Upon removal of NH₄Cl, the Ca²⁺ signals showed an ‘overshoot’ and returned to the initial peak level (Fig. 7A and B), probably as a result of the acid loading associated with NH₄Cl removal (Fig. 1B). The ‘overshoot’ in [Ca²⁺]_i may indeed be substantially larger than indicated by the traces shown, since the K_d of fura-2 increases during large acidifications, which has the effect of making the dye less sensitive to [Ca²⁺]_i. The [Ca²⁺]_i response to agonists during the plateau phase of sustained signalling is derived from increased Ca²⁺ influx (Marty & Tan, 1989; Kwan, Takemura, Obie, Thastrup & Putney, 1990). Since these observations were made in the presence of extracellular Ca²⁺, they could relate to a reduction in Ca²⁺ entry during agonist stimulation as a result of NH₄Cl-induced cytosolic alkalinization. To test whether cytosolic alkalinization inhibited Ca²⁺ entry, microsomal Ca²⁺-ATPase inhibitors were employed to study the Ca²⁺ entry process in isolation following store depletion. Using a combination of cyclopiazonic acid and thapsigargin to deplete Ca²⁺ stores and promote Ca²⁺ entry, alkalinization with NH₄Cl had no effect on [Ca²⁺]_i (Fig. 8A). In addition, cytosolic alkalinization with NH₄Cl rapidly decreased [Ca²⁺]_i during agonist stimulation in the absence of extracellular Ca²⁺ (Fig. 8B) also arguing against any inhibitory effect of NH₄Cl on Ca²⁺ entry in pancreatic acinar cells.

Since agonists and antagonists of the muscarinic receptor are quarternary NH₄⁺ ions, it is conceivable that NH₄⁺ might interfere with muscarinic receptor function, thereby explaining the rapid decrease in [Ca²⁺]_i during ACh stimulation in pancreatic acinar cells. However, this would not explain why NH₄Cl also inhibited the [Ca²⁺]_i response to CCK (Fig. 7B).

Acetate enhances calcium signalling

The transient nature of the acetate-evoked increase in [Ca²⁺]_i during agonist stimulation, together with the observation that this increase in [Ca²⁺]_i does not persist following store depletion, suggests that acetate acts to release Ca²⁺ from intracellular stores. In addition, the intracellular store(s) involved are presumably those sensitive to InsP₃, since the acetate effect is only observed in the presence of ACh or CCK.

There are several possible mechanisms by which acetate might enhance Ca²⁺ release from the intracellular stores and thus account for the observed rise in [Ca²⁺]_i. Acetate/H⁺ might reduce Ca²⁺ re-uptake via inhibition of the microsomal Ca²⁺-ATPase. Inhibition of this type may produce a net increase in Ca²⁺ release from the intracellular stores. The data in Table 1 do not show any significant differences in the rate of Ca²⁺ uptake into stores in permeabilized pancreatic acinar cells at the various medium pHs. However, cell-to-cell variation in the rate of Ca²⁺ uptake was quite large, and this may have obscured the effect of pH.

Inositol 1,4,5-trisphosphate-mediated Ca²⁺ release from intracellular stores has been reported to be sensitive to H⁺ in a number of studies. Firstly, the binding of InsP₃ to the InsP₃ receptor increases at more alkaline pH in both brain and smooth muscle (Worley, Baraban, Supattapone, Wilson & Snyder, 1987; Joseph, Rice & Williamson, 1989; Varney, Rivera, Lopez Bernal & Watson, 1990). In addition, permeabilized human platelets released 30, 50 and 75% of stored Ca²⁺ when suspended in medium at pH 6.9, 7.1 and 7.4, respectively, showing that the affinity of the InsP₃ receptor for InsP₃ increases at more alkaline pHs (Brass & Joseph, 1985). In permeabilized smooth muscle cells, the sensitivity of the Ca²⁺ release mechanism to InsP₃ and the rate of InsP₃-induced Ca²⁺ release were both increased at more alkaline pHs in the presence of Ca²⁺, but pH had little effect in the absence of Ca²⁺ (Tsukioka, Iino & Endo, 1994). The fact that effects of pH_i were only observed in the presence of Ca²⁺ in the latter study suggests some degree of cross-talk between InsP₃, Ca²⁺ and H⁺ binding sites.

The available literature thus consistently shows an enhancement of the effects of InsP₃ at more alkaline pH_i. However, from the results on intact cells presented in this study, one might expect the converse to be true, i.e. the sensitivity and/or rate of InsP₃-induced Ca²⁺ release to be reduced at more alkaline pHs. This is suggested by the fact that the weak base NH₄Cl inhibits Ca²⁺ signalling during agonist stimulation, while the weak acid acetate enhances agonist-evoked Ca²⁺ release. This theory was examined directly by measuring intraorganellar [Ca²⁺] using mag-fura-2 in permeabilized pancreatic acinar cells. Addition of acetate/NH₄Cl at a constant medium pH did not affect the rate of InsP₃-evoked Ca²⁺ release. Since NH₄Cl is known to alkalinize vesicular compartments, this suggests that intraorganellar pH is unlikely to have a dramatic effect on Ca²⁺ release. Consistent with this conclusion, the collapse of the zymogen granule pH gradient in pancreatic acinar cells induced by a number of lysosomotropic agents, including methylamine, did not affect CCK- or carbachol-stimulated amylase secretion (De Lisle & Williams, 1987).

We went on to examine the effects of extraorganellar ‘cytosolic’ pH on InsP₃-evoked Ca²⁺ release. In single-cell intraorganellar [Ca²⁺] measurements, a concentration of 0.5 μm InsP₃ was used to release Ca²⁺ from the stores. Increasing extraorganellar pH between 6.7 and 7.6 enhanced the rate of InsP₃-induced Ca²⁺ release, probably by altering the sensitivity of the InsP₃ receptor to InsP₃. This is consistent with the work of Tsukioka et al. (1994). Calcium is a known co-agonist of InsP₃-induced Ca²⁺ release from intracellular stores (Finch, Turner & Goldin, 1991), and the coupling domain of the InsP₃ receptor contains putative binding sites for Ca²⁺ (Bezprozvanny & Ehrlich, 1995). It is possible that competition between Ca²⁺ and H⁺ could occur at the Ca²⁺ binding site(s). For example, an acidic pH_i may result in decreased Ca²⁺ binding to the InsP₃ receptor and hence a reduction in the open probability of the InsP₃ receptor-linked channel. This may explain some of the earlier observations by Tsukioka et al. (1994), as well as our own results in permeabilized cells. However, these data clearly cannot explain the enhanced Ca²⁺ release from intracellular stores observed with acetate during agonist stimulation, or the rapid inhibition of Ca²⁺ signals by cytosolic alkalinization.

The question remains, then, of how acidification enhances Ca²⁺ release from the intracellular stores. An increase in InsP₃ formation via an action of pH_i at some point in the phosphatidylinositol pathway is a plausible option. Unfortunately, we were unable to investigate this possibility because of recurring technical problems with measuring InsP₃ levels. In rat IMCD cells, NH₄Cl removal and nigericin-induced cytosolic acidification significantly raised basal InsP₃ levels under resting conditions (Slotki et al. 1993), while alkalinization raised the basal rate of inositol phosphate turnover in lacrimal acinar cells (Yodozawa et al. 1997).

Fluid secretion

The results of this study may be relevant to observations in many epithelia where a number of weak acids can substitute for HCO₃⁻ in fluid and electrolyte transport (Ullrich et al. 1971; Petersen et al. 1981; Case et al. 1982). For instance, in the rabbit mandibular salivary gland, acetate perfusion increased ACh-stimulated fluid secretion by more than 100% (Novak & Young, 1989). An increase in [Ca²⁺]_i, such as that observed with acetate during ACh application, would be expected to lead to an increase in fluid secretion by causing activation of Ca²⁺-dependent Cl⁻ and K⁺ channels. Interestingly, the acetate effect observed in Novak & Young's study was abolished in mandibular glands stimulated with supramaximal concentrations of ACh. In the present study, the Ca²⁺ releasing effect of acetate was also reduced with 10 μm ACh. An observation which argues against a mechanism based on changes in [Ca²⁺]_i is that salivary flow rate in the mandibular gland remained elevated above the control value for several hours during perfusion with acetate (Novak & Young, 1989). However, in pancreatic acinar cells (this study) and lacrimal acinar cells (Yodozawa et al. 1994), the effects of acetate on [Ca²⁺]_i were short lived. It therefore seems unlikely that enhancement of agonist-evoked intracellular Ca²⁺ release by acetate could fully explain the prolonged elevation in secretion in rabbit mandibular glands.

In perfused rabbit mandibular glands, the weak base NH₄Cl had dramatic inhibitory effects upon ACh-evoked fluid secretion (Lau, Donohue & Case, 1996). Even at concentrations as low as 0.5 mm, NH₄Cl totally inhibited fluid secretion. This observation is reminiscent of the inhibitory effects of NH₄Cl upon Ca²⁺ signalling in this study and in lacrimal acinar cells (Yodozawa et al. 1997), although the concentrations of NH₄Cl employed were quite different. Cytosolic alkalinization has been reported to influence electrolyte transport in a number of epithelia (see Lau, Elliott, Brown & Case, 1990; Yodozawa et al. 1997, for references).

In summary, many earlier observations indicate that cell permeant weak acids and bases, including acetate and NH₄Cl, significantly modulate acinar cell function. These effects have not generally been attributed to altered cytosolic Ca²⁺ handling. However, the present study clearly shows that alterations in pH_i induced by weak acids and bases can alter agonist-evoked [Ca²⁺]_i signals. Changes in pH_i have been reported to occur during agonist stimulation in many acinar cells, with the typical pattern of a transient intracellular acidosis followed by a sustained alkalosis (Muallem & Loessberg, 1990b; Seo, Larcombe-McDouall, Case & Steward, 1995; Steward, Poronnik & Cook, 1996). The data presented in this study suggest that these cytosolic pH changes occurring in acinar cells during stimulation are likely to modify Ca²⁺ signalling.