Imaging of Intracellular Calcium Stores in Individual Permeabilized Pancreatic Acinar Cells STORES IN PERMEABILIZED PANCREATIC ACINAR CELLS*

Several lines of evidence suggest that the existence of a heterogeneous population of inositol 1,4,5-trisphos-phate (Ins(1,4,5)P 3 )-sensitive Ca 2 (cid:49) stores underlies the polarized agonist-induced rise in cytosolic Ca 2 (cid:49) concentration ([Ca 2 (cid:49) ] i ) in pancreatic acinar cells (Kasai, H., Li, Y. X., and Miyashita, Y. (1993) Cell 74, 669–677; Thorn, P., Lawrie, A. M., Smith, P. M., Gallacher, D. V., and Pe-tersen, O. H. (1993) Cell 74, 661–668). To investigate whether the apical pole of acinar cells contains Ca 2 (cid:49) stores which are relatively more sensitive to Ins(1,4,5)P 3 than those in basolateral areas, we studied Ca 2 (cid:49) han-dling by Ca 2 (cid:49) stores in individual streptolysin O (SLO) permeabilized cells using the low affinity Ca 2 (cid:49) indicator Magfura-2 and an in situ imaging technique. The uptake of Ca 2 (cid:49) by intracellular Ca 2 (cid:49) stores was ATP-dependent. A steady-state level was reached within 10 min, and the free Ca 2 (cid:49) concentration inside loaded Ca 2 (cid:49) stores was estimated to be 70 (cid:109) M . Ins(1,4,5)P 3 induced Ca a electron analysis. a of The of

In many non-excitable cell types, agonist stimulation results in repetitive oscillations of the free cytosolic Ca 2ϩ concentration ([Ca 2ϩ ] i ) 1 arising largely from Ca 2ϩ release from intracel-lular stores (1). In the polarized exocrine acinar cell of exocrine glands, these agonist-induced intracellular Ca 2ϩ signals are not spatially homogeneous. Thus acetylcholine-or cholecystokinin-octapeptide-induced [Ca 2ϩ ] i rises are initiated in the luminal pole of the cytosol with the Ca 2ϩ wave subsequently spreading into the basolateral areas of the cell (2)(3)(4). In some cases little or no [Ca 2ϩ ] i increase at all is observed in basolateral regions (5,6). This spatial pattern of [Ca 2ϩ ] i signaling has been suggested to be important for both unidirectional fluid secretion (2,7) and exocytosis (8).
In permeabilized pancreatic acinar cells, as in many other permeabilized cell systems, the intracellular messenger D-myoinositol 1,4,5-trisphosphate (Ins(1,4,5)P 3 ) mobilizes Ca 2ϩ from non-mitochondrial intracellular Ca 2ϩ stores (1,9). Recent evidence in both permeabilized (10 -12) and intact cells (5,6) favors the existence of a heterogeneous population of Ins(1,4,5)P 3 -sensitive Ca 2ϩ stores. In a recent study (12) it was suggested that the Ca 2ϩ imaging data obtained by several laboratories (see above) might be explained by the existence of stores more sensitive to Ins(1,4,5)P 3 localized in the apical pole with stores less sensitive to Ins(1,4,5)P 3 being located in basolateral areas.
To address this question we have now imaged the Ca 2ϩ concentration within Ca 2ϩ stores in pancreatic acinar cells using the in situ imaging technique originally described by Hofer and Machen (13). The low affinity Ca 2ϩ indicator Magfura-2 was loaded into intracellular stores and the properties of these stores were studied in streptolysin O-permeabilized cells. ATP-driven Ca 2ϩ uptake and Ins(1,4,5)P 3 -dependent Ca 2ϩ release could be clearly demonstrated within individual permeabilized pancreatic acinar cells. However, we were unable to detect any subcellular regional differences in Ins(1,4,5)P 3 -sensitivity. This may indicate that cytosolic modulators of Ins(1,4,5)P 3 -operated Ca 2ϩ channels and/or the involvement of ryanodine receptors, underlie the polarized nature of [Ca 2ϩ ] i signaling patterns in acinar cells.

EXPERIMENTAL PROCEDURES
Pancreatic Acinar Cells-Acinar cells were prepared from the pancreas of one 200-g male Sprague-Dawley rat. The cell isolation procedure used was the same as described previously for rabbit pancreas (12). After isolation, acinar cells were resuspended in 6 ml of a HEPES/ Tris-buffered (pH 7.4) physiological medium which contained: 133 mM NaCl, 4.2 mM KCl, 1.0 mM CaCl 2 , 1.0 mM MgCl 2 , 5.8 mM glucose, 0.2 mg/ml soybean trypsin inhibitor, an amino acid mixture according to Eagle (14), 1% (w/v) bovine serum albumin, and 10 mM HEPES. The pH of the medium was set at 7.4 with Tris. Cells were either used immediately or stored in 1-ml portions on ice until use.
Loading of Pancreatic Acinar Cells with Magfura-2-Pancreatic acinar cells were resuspended in the physiological medium described above with the addition of 5 M Magfura-2-AM. After a 30-min incubation at 37°C, cells were washed twice with physiological medium containing 0.1% bovine serum albumin. Cells were allowed to settle on a poly-L-lysine-coated glass coverslip which formed the bottom of a perfusion chamber.
Imaging of Magfura-2-loaded Pancreatic Acinar Cells-The imaging system used was based on an inverted epifluorescence Nikon Diaphot microscope and a ϫ 40 oil immersion lens (numerical aperture: 1.3). A field containing 5-15 cells was selected and the dye-loaded cells were excited alternately with light at 340 and 380 nm using a filterwheel (Lambda 10, Sutter Instruments; 340-and 380-nm band-pass filters were from Ealing Electro-Optics) and a dichroic mirror (400-nm dichroic mirror, 420-nm barrier filter). The emitted fluorescence was captured and digitized at 12-bit resolution by a slow scan CCD camera (Digital Pixel Ltd., Brighton, United Kingdom). An IBM-compatible personal computer and an imaging software package (Kinetic Imaging Ltd., Liverpool, UK) was used to drive the filterwheel and camera and store acquired images. The size of the silicon sensor of the camera and the ϫ 40 objective allowed images of a field 90 ϫ 135 m to be captured. A 3 ϫ 3 binning was applied to the individual pixels on the image sensor to give a spatial resolution of 0.67 m/pixel. Since acinar cells have virtually no detectable autofluorescence (results not shown) an empty area of a coverslip was used to determine background levels. All images were background corrected. The ratio of the 340 nm and 380 nm excitations of paired, background-subtracted images were calculated offline. All experiments were performed at room temperature.
Since a considerable amount of the total accumulated Magfura-2 was present in the cytosolic compartment, the permeabilization process could be followed on-line (as loss of cytosolic dye) by using the imaging system. At the start of the permeabilization procedure, loaded cells were excited at the isosbestic wavelength for Magfura-2, i.e. 360 nm. Permeabilization was achieved within 10 min and, as a consequence, a significant drop in fluorescence was observed as cytosolic Magfura-2 was lost into the incubation medium (results not shown). Perfusion of the permeabilized cells was subsequently continued with the Ca 2ϩ uptake medium devoid of SLO, as described above.
Ca 2ϩ Uptake and Release Experiments-Permeabilized cells were continuously perfused throughout the experiments. Ca 2ϩ uptake by intracellular Ca 2ϩ stores was initiated by superfusing cells with a medium containing 1 mM ATP and a free Ca 2ϩ concentration of 0.2 M; the free Mg 2ϩ concentration remained 0.9 mM (free divalent cation concentrations were calculated again according to Schoenmakers et al. (15)). Mitochondrial Ca 2ϩ uptake inhibitors were not included in the medium, since mitochondrial Ca 2ϩ uptake has previously been shown not to occur at this ambient free Ca 2ϩ concentration (16). After a loading period of 15 min, permeabilized cells were stimulated with Ins(1,4,5)P 3 , thapsigargin, or cyclic ADP-ribose at various concentrations as described in the text and in the captions of the figures. The Ca 2ϩ concentration in the stores was monitored by determining the ratio of the emitted fluorescence at 340 and 380 nm excitation. Exposure times for individual images were 300 ms, and the interval times between ratio images is indicated in the legend of the figure.
Calibration of Ratio Values-The Ca 2ϩ ionophore 4-Br-A23187 (2 M) was used to equilibrate extra-and intra-compartment Ca 2ϩ and thus to impose determined free Ca 2ϩ concentrations inside the Magfura-2 containing compartments. The medium used was a 137 mM KCl solution buffered with 20 mM HEPES/KOH to pH 7.1. Media were prepared with various free Ca 2ϩ concentrations (as indicated) in the absence of Mg 2ϩ (since Mg 2ϩ gradients are also dissipated by Ca 2ϩ ionophores and Mg 2ϩ is able to alter the fluorescent properties of Magfura-2). Free Ca 2ϩ concentrations were calculated taking account of the ionic strength of the solution. The activity coefficient for Ca 2ϩ was calculated by using the Guggenheim approximation to the Debye-Hueckel Limiting Law (17,18).
Electron Microscopy on Intact and Permeabilized Pancreatic Acinar Cells-Intact or SLO-permeabilized acinar cells (permeabilization procedure, see above) were diluted 1:1 into fixation buffer (0.1 M sodium cacodylite, 5% glutaraldehyde). Preparations were washed twice, postfixed with 1% osmium tetroxide in 0.1 M sodium cacodylate buffer (pH 7.4), dehydrated, and embedded in resin. A Reichert Ultracut E microtome was used to cut 60-nm sections. Sections were stained using uranyl acetate and lead citrate, and a Philips transmission electron microscope (200) was used to visualize the samples. The magnifications used are indicated in the figures.
Analysis of Data-The mean ratio values mentioned under "Results" were calculated by determining ratio values for each of the individual cells in a field. From these values, an average value for a particular experiment was determined. These values were in turn averaged between preparations to give the values presented in the text. The value of n given thus refers to the number of cell preparations on which a given experiment was performed, although the total number of individual cells analyzed was between 19 and 180 for different experimental procedures.

Effect of Streptolysin O Treatment on Pancreatic Acinar
Cells-The morphology of pancreatic acinar cells was altered after permeabilization. Acinar cells had a more swollen appearance in the light microscope, and the cytosol became paler, reflecting loss of proteins. Permeabilized cells coupled to each other in doublets or triplets clearly retained their polarized morphology, whereas the polarity in individual cell was, just as in intact cells, largely lost (see e.g. Fig. 5B for an example). To study the effect of SLO treatment on the cellular structure in more detail, electron microscopy was performed on both intact and permeabilized cells. Fig. 1A shows a representative triplet of intact acinar cells. Cells retained a polarized morphology, since the large densely stained secretory granules were restricted to the apical pole. The endoplasmic reticulum retained its typical tubular arrangement and was present in all parts of the cell, although it was more abundant in basolateral regions. A representative example of permeabilized pancreatic acinar cells is shown in Fig. 1B. The cells had a much brighter appearance, reflecting loss of the majority of the cytosolic content after permeabilization. The picture shows two coupled cells which were, similar to intact cells, polarized. The endoplasmic reticulum was rearranged but appeared to be continuous. The changes in endoplasmic reticulum morphology may be due to the swelling of the cells after permeabilization.
Ca 2ϩ Uptake by Permeabilized Pancreatic Acinar Cells-After permeabilization, cells were exposed to an "intracellular" medium with an ambient free Ca 2ϩ concentration of 0.2 M Ca 2ϩ and 1 mM ATP. This resulted in an increase in the Magfura ratio, presumably reflecting an increase in the free Ca 2ϩ concentration in intracellular Ca 2ϩ stores (Fig. 2). To rule out the possibility that changes in the Magfura ratio were due to changes in intra-organelle [Mg 2ϩ ], rather than [Ca 2ϩ ], the experimental protocol was adapted by perfusing cells initially with Ca 2ϩ uptake medium devoid of Ca 2ϩ but containing ATP and Mg 2ϩ . Fig. 2 shows that the ratio remained unaltered and could only be increased by including Ca 2ϩ in the medium. The combined addition of Ca 2ϩ and ATP increased the Magfura ratio from 0.53 (S.E. ϭ 0.04; n ϭ 14 cell preparations and 137 individual cells analyzed) to 1.44 (S.E. ϭ 0.04; n ϭ 16 cell preparations and 178 individual cells analyzed) within 10 min. This increased level remained constant for a considerable period thereafter (at least 10 min).
To demonstrate that the Magfura signal was not saturated in loaded Ca 2ϩ stores under these conditions, the Ca 2ϩ ionophore 4-Br-A23187 (2 M) was included and the ambient free Ca 2ϩ concentration in the medium was increased to 1 mM (for details of the medium used, see "Experimental Procedures"). This treatment resulted in a further increase in the ratio to 3.43 (S.E. 0.02; n ϭ 3 cell preparations and 19 individual cells analyzed; results not shown, but see also Fig. 3). This indicates that the free Ca 2ϩ concentration of steady-state loaded intracellular Ca 2ϩ stores of pancreatic acinar cells is not in the millimolar range. A tentative calibration was made by perfus-ing permeabilized cells with Ca 2ϩ ionophore and a medium containing different ambient free Ca 2ϩ concentrations. Fig. 3 shows the result of this calibration. The average Magfura ratio in stores loaded at steady-state was 1.44 (see above), and therefore the free Ca 2ϩ concentration in loaded intracellular Ca 2ϩ stores is estimated to be 70 M. The minimum ratio, obtained with permeabilized cells in a Ca 2ϩ free medium, including ionophore, was 0.37 (S.E. ϭ 0.02; n ϭ 4 cell preparations and 37 individual cells analyzed). The average ratio in unloaded stores without ionophore in the same set of experiments was 0.47 (S.E. ϭ 0.03; n ϭ 4 cell preparations and 37 cells analyzed). This shows that, before initiation of Ca 2ϩ uptake, the Ca 2ϩ  Effects of Ins(1,4,5)P 3 on Loaded Intracellular Ca 2ϩ Stores-Perfusion of permeabilized pancreatic acinar cells with an Ins(1,4,5)P 3 containing medium resulted in a release of Ca 2ϩ , indicated by a fall in the Magfura ratio. Fig. 4, A and B, show the effect of 0.3 and 1.0 M Ins(1,4,5)P 3 , respectively, on loaded stores in permeabilized cells. Ins(1,4,5)P 3 induced a release of Ca 2ϩ in both cases. The rate of release, and the new reduced intravesicular [Ca 2ϩ ] achieved, both varied with the concentration of Ins(1,4,5)P 3 applied. The intravesicular Ca 2ϩ levels could be further reduced by increasing the Ins(1,4,5)P 3 concentration, which also indicated that the release process was not desensitized despite the 12 min presence of submaximal doses of Ins(1,4,5)P 3 . At 10 M Ins(1,4,5)P 3 , a maximal Ca 2ϩ releasing effect was obtained and the ratio was reduced to 0.62 (S.E. ϭ 0.02, n ϭ 5 cell preparations and 52 individual cells analyzed). This level was close to the ratio in unloaded stores, indicating that the Ca 2ϩ stores were virtually emptied by a maximally effective dose of Ins(1,4,5)P 3 (10 M was shown to be a maximally effective dose, since 30 M did not induce any further Ca 2ϩ release). The half-maximal effect of Ins(1,4,5)P 3 occurred at 0.3 M, a value slightly lower than, but in a range similar to that reported by others using permeabilized pancreatic acinar cell preparations (e.g. Refs. 9,10,19,and 20).
In permeabilized gastric epithelial cells, mitochondrial uptake of Ca 2ϩ contributed substantially to the ATP-and Ca 2ϩdependent increase in the signal from compartmentalized Magfura-2 (21). Thus a considerable part of the ATP-dependent Ca 2ϩ pool in this cell type was Ins(1,4,5)P 3 -insensitive but sensitive to mitochondrial Ca 2ϩ uptake inhibitors. In pancreatic acinar cells, however, a maximal dose of Ins(1,4,5)P 3 released virtually all the Ca 2ϩ that had been taken up in an ATP-dependent manner, indicating that mitochondria were not active under the conditions used (see above). In addition, when Ca 2ϩ uptake was performed in the presence of the mitochondrial inhibitors antimycin and oligomycin, no change in uptake characteristics was observed (results not shown). These observations confirm previous studies with radiotracer techniques on permeabilized pancreatic acinar cells, in which it was shown that mitochondrial Ca 2ϩ uptake was inactive at an ambient free Ca 2ϩ concentration identical to that used in the present study (16).
Effects of Ins(1,4,5)P 3 on Cellular and Subcellular Level-We went on to study the kinetics and sensitivity of Ins(1,4,5)P 3induced Ca 2ϩ release in individual cells in more detail. The "steps" in store Ca 2ϩ content on increasing Ins(1,4,5)P 3 resemble the so-called "quantal" Ca 2ϩ release previously observed in these and other cells (10,12,22). The effect of Ins(1,4,5)P 3 was reversible, since reperfusion with control medium resulted in a reuptake of Ca 2ϩ (this reversibility was tested in four experiments and was observed in all cases). Fig. 5A also shows that the effects of Ins(1,4,5)P 3 were observed in all seven cells analyzed from this field. Ins(1,4,5)P 3 evoked a simultaneous and equal response in all cells, indicating that there were no intercellular differences in Ins(1,4,5)P 3 sensitivity.
The action of Ins(1,4,5)P 3 was studied in more detail by analyzing the ratio images of the experiment. Fig. 5, B and C show a brightfield and a fluorescence image respectively of the selected field of cells. The brightfield image shows again that permeabilized cells organized in triplets clearly maintained their polarized morphology. The first ratio image (Fig. 5D, image 1) shows Ca 2ϩ stores in permeabilized cells loaded to a steady-state level. The ratio intensity throughout the different regions of the cells was not homogeneous, although virtually all regions had a ratio value of about 1 or higher (i.e. the free [Ca 2ϩ ] was 40 M or higher). Stimulation with a low dose of Ins(1,4,5)P 3 (0.3 M) resulted in a simultaneous decrease of Ca 2ϩ levels in all regions of the permeabilized cells (Fig. 5D,  images 2 and 3). Elevation of the dose to 1.0 M (Fig. 5D, image  4) and then to 10.0 M (Fig. 5D, image 5) resulted again in a simultaneous reaction in all subcellular regions in all permeabilized acinar cells. Subsequent removal of Ins(1,4,5)P 3 resulted in Ca 2ϩ reuptake in all regions of the permeabilized cells (Fig. 5D, image 6). (The regional analysis, as presented in Fig.  5D, was performed in four additional experiments; in two experiments sequential Ins(1,4,5)P 3 additions were made as presented in Fig. 5, and in the other two experiments, 0.3 M Ins(1,4,5)P 3 added to loaded stores of permeabilized cells. All experiments analyzed gave similar results to those presented in Fig. 5). To further demonstrate the uniform Ins(1,4,5)P 3 sensitivity, the kinetics of Ins(1,4,5)P 3 -induced Ca 2ϩ release were compared between selected areas of interest in apical and basolateral regions in the same field of cells. Fig. 6A shows the selected areas and Fig. 6B shows the averaged and normalized kinetics of Ca 2ϩ release induced by Ins(1,4,5)P 3 in apical and basolateral regions. Again, the results demonstrate that both regions were equally sensitive to Ins(1,4,5)P 3 . Taken together, the results demonstrate that Ca 2ϩ stores in SLO-permeabilized acinar cells display neither regional nor intercellular differences in their sensitivity toward Ins(1,4,5)P 3 .

Effect of Thapsigargin on Ins(1,4,5)P 3 -induced Ca 2ϩ
Release-To study the effect of Ca 2ϩ pump activity on Ins(1,4,5)P 3 -induced Ca 2ϩ release, thapsigargin was used to completely block all Ca 2ϩ pumping into the intracellular Ca 2ϩ stores (12,16). Addition of thapsigargin (1 M) resulted in a slow but sustained efflux of Ca 2ϩ from the stores with kinetics similar to that observed in permeabilized acinar cell suspensions (Fig. 7). Addition of 0.1 M Ins(1,4,5)P 3 resulted in an increased efflux rate, indicating that Ins(1,4,5)P 3 -operated Ca 2ϩ channels were indeed activated at this low dose. After 5 min the Ins(1,4,5)P 3 concentration was increased to 0.3 M; the efflux rate increased again and remained elevated. Under these conditions suboptimal doses of Ins(1,4,5)P 3 were indeed more effective in depleting stores compared with the situation where Ca 2ϩ pumps remained active. The Ins(1,4,5)P 3 -sensitive Ca 2ϩ stores in the acinar cells were virtually depleted after prolonged treatment with 0.3 M Ins(1,4,5)P 3 , since a further increase of the Ins(1,4,5)P 3 concentration to 1.0 M evoked only a minor further release of Ca 2ϩ . DISCUSSION The major aim of the present study was to characterize the spatial organization of intracellular Ca 2ϩ stores in pancreatic acinar cells. To address this question we imaged Ca 2ϩ stores in individual permeabilized pancreatic acinar cells using a Ca 2ϩsensitive dye compartmentalized in organelles. Our main finding is that Ins(1,4,5)P 3 -sensitive Ca 2ϩ stores are located throughout the acinar cell cytoplasm and that no regional differences in Ins(1,4,5)P 3 sensitivity exist, at least in the absence of cytosolic modulatory factors.
During the cell permeabilization process, cytosolic factors are lost via the pores created by SLO in the plasma membrane (23). One of our concerns was that SLO treatment affected intracellular structures. Both conventional and electron microscopy demonstrated that SLO permeabilized pancreatic acinar cells had a more swollen appearance. However, the cell architecture remained polarized, since the localization of zymogen granules remained restricted to the apical pole of the cells. Other studies on SLO-permeabilized pancreatic acinar cells have also shown that they retain their polarity and remain functionally active, both in terms of agonist-or Ins(1,4,5)P 3stimulated Ca 2ϩ release from intracellular stores and in terms of agonist-or Ins(1,4,5)P 3 -or Ca 2ϩ -stimulated enzyme secretion (24,25). Electron microscopy revealed that the endoplasmic reticulum was less strictly arranged compared with intact cells, an effect that was most likely caused by the swelling. We therefore cannot rule out the possibility that some rearrangement of the endoplasmic reticulum Ca 2ϩ stores may have occurred. However, our hypothesis is that the relative position of the components of the Ca 2ϩ stores is most likely not altered given that the permeabilized cells clearly retained their polarized morphology.
The characteristics of store loading, and the ratio values in unloaded and loaded stores of permeabilized pancreatic acinar cells, were similar to those observed in permeabilized gastric epithelial cells by Hofer and Machen (13). Our estimate of the free intra-organellar Ca 2ϩ concentration in steady-state loaded intracellular Ca 2ϩ stores as 70 M is also in broad agreement with the value of 127 M found in gastric epithelial cells. It might be argued that Mg 2ϩ interferes to some extent with the Ca 2ϩ signals reported by Magfura-2. However, this seems unlikely, since (i) Mg 2ϩ is not transported in an ATP-dependent manner and (ii) the resting Magfura ratio in unloaded stores was very low, i.e. nearly equivalent to the minimum ratio for the dye in the total absence of all divalent cations. In addition, if Mg 2ϩ was present inside stores its free concentration would need to exceed a value of 150 M to give a significant contribution to the Magfura-2 signal, since the apparent affinity of the dye for Mg 2ϩ is very low, i.e. 1.5 mM (26). The Magfura-2 signal in organelles was clearly not saturated with Ca 2ϩ under normal conditions, since the ratio was increased markedly by exposure of Ca 2ϩ stores to the Ca 2ϩ ionophore 4-Br-A23187 in the presence of 1 mM ambient Ca 2ϩ .
Recent developments have allowed measurements of Ca 2ϩ levels inside the endoplasmic reticulum by targeting the Ca 2ϩsensitive bioluminescent protein aequorin to this organelle (27)(28)(29). In the first work of this type, Kendall et al. (27,28) have reported that in COS-7 cells free Ca 2ϩ inside the endoplasmic reticulum was around 1-5 M, approximately 5-20 times the free cytosolic Ca 2ϩ concentration. Very recently, however, it has been reported by Montero et al. (29) that Ca 2ϩ concentrations inside the endoplasmic reticulum of HeLa cells exceeded 100 M. By using the Ca 2ϩ surrogate Sr 2ϩ , these workers concluded that even millimolar free concentrations of divalent cations could occur within the endoplasmic reticulum and they argued from this that Ca 2ϩ levels might reach similar values. Taken at face value, however, the widely differing estimates reported using the aequorin technique suggest that a ubiquitous conclusion about Ca 2ϩ levels inside Ca 2ϩ stores cannot be reached. Although, as discussed above, we cannot exclude the possibility that properties of the endoplasmic reticulum are altered during permeabilization, we suggest that our estimate of a free Ca 2ϩ concentration within the endoplasmic reticulum of 70 M might well be applicable to loaded Ca 2ϩ stores in intact pancreatic acinar cells.
The second messenger Ins(1,4,5)P 3 released Ca 2ϩ from intracellular Ca 2ϩ stores in permeabilized pancreatic acinar cells. The characteristics of this release were similar to the fluxes observed in suspensions of permeabilized acinar cells using the radioactive tracer 45 Ca 2ϩ (10,12). Ca 2ϩ release was of a quantal nature, apparently due to the compensatory action of the organelle Ca 2ϩ pump during suboptimal stimulation. Thus, suboptimal concentrations of Ins(1,4,5)P 3 were much more efficient in releasing Ca 2ϩ in the absence of Ca 2ϩ pumping activity. An interesting observation was that all permeabilized cells showed similar sensitivities to Ins(1,4,5)P 3 . This observation rules out the often raised possibility that intercellular differences in Ins(1,4,5)P 3 sensitivity determine the quantal nature of Ins(1,4,5)P 3 -induced Ca 2ϩ release.
Compartmentalized dye techniques similar to those applied here have been employed in a number of cell types, including hepatocytes, gastric epithelial cells, AR4 -2J pancreatoma cells, and smooth muscle cells (13, 21, 30 -32). In both hepatocytes (31) and DDT 1 MF-2 smooth muscle cells (32) the intracellular Ca 2ϩ stores function as a single homogeneous pool, and electron microscopy has shown that the endoplasmic reticulum, which presumably acts as the intracellular Ca 2ϩ storage compartment, is a continuous compartment. However, the properties of Ins(1,4,5)P 3 -induced Ca 2ϩ release from intracellular Ca 2ϩ stores clearly differed between the two cell types. In permeabilized hepatocytes attached to coverslips, the Ins(1,4,5)P 3 -induced Ca 2ϩ release was non-quantal, in that, while the kinetics of Ca 2ϩ release depended on the concentration of Ins(1,4,5)P 3 used, all doses of Ins(1,4,5)P 3 eventually induced total depletion of the Ca 2ϩ stores. In smooth muscle cells, in contrast, Ca 2ϩ release induced by Ins(1,4,5)P 3 was of a quantal nature, similar to what we have observed for pancreatic acinar cells. Agonist stimulation of intact acinar cells initiates a rise of cytosolic Ca 2ϩ in the apical pole of the cell, with the increase in [Ca 2ϩ ] i subsequently spreading into basolateral areas of the cell (2)(3)(4)(5)(6)(7). It is notable, however, that some reports indicate that [Ca 2ϩ ] i rises uniformly throughout acinar cells upon agonist stimulation (33,34). In recent studies on intact acinar cells, employing the combination of imaging techniques and patch-clamp recording, evidence was obtained for a heterogeneous distribution of Ca 2ϩ stores (5,6). By infusing acinar cells with a low dose of Ins(1,4,5)P 3 , or its non-metabolizable analogue inositol 1,4,5-trisphosphorothioate, it was shown that Ca 2ϩ spikes could be generated exclusively in the apical pole. These results therefore strongly support the idea that a heterogeneous population and distribution of Ins(1,4,5)P 3 -sensitive Ca 2ϩ pools do exist in individual pancreatic acinar cells. Biochemical evidence suggested that this store heterogeneity might be explained by differences in numbers of Ins(1,4,5)P 3operated Ca 2ϩ channels and/or by differences in sensitivity to Ins(1,4,5)P 3 (12). In particular, it was suggested that, during suboptimal stimulation, the most sensitive stores were completely depleted, whereas less sensitive stores remained partially filled due to a compensatory pumping mechanism. This model (12) could explain a number of observations in intact acinar cells in which the apical pole Ca 2ϩ stores display higher apparent Ins(1,4,5)P 3 and Ca 2ϩ sensitivity (5,6). In the present study we have examined directly whether Ca 2ϩ stores were heterogeneously distributed in individual permeabilized cells. However, no subcellular differences in Ins(1,4,5)P 3 sensitivity could be detected. One possible interpretation of this result is that subcellular regional differences in Ins(1,4,5)P 3 sensitivity depend critically on some aspects of cellular or cytoskeletal or endoplasmic reticulum architecture, which is disrupted by permeabilization. However, as discussed above, we feel it is unlikely that cell permeabilization results in a major redistribution of Ca 2ϩ stores. If our results can be extrapolated to the intact acinar cell, an alternative explanation of why Ca 2ϩ starts to rise in the apical region on agonist stimulation may be the selective presence of an additional Ca 2ϩ release mechanisms in this area of the cell (5,35). Several studies suggest the existence of cyclic ADP-ribose-induced Ca 2ϩ release which might be mediated by ryanodine receptors (35,36). The experimental evidence in those studies has been interpreted to suggest that the combined activation of Ins(1,4,5)P 3 -sensitive and cyclic ADP-ribose-sensitive mechanisms is required to explain polarized Ca 2ϩ spike generation. However, 5 M cyclic ADPribose failed to change the Magfura signal in permeabilized pancreatic acinar cells, whereas in the same cells Ins(1,4,5)P 3 induced a normal response. 2 It is possible that cytosolic factors, which are lost during permeabilization and the subsequent extensive perfusion, may be required for the cyclic ADP-ribose response.
Interestingly, other lines of evidence argue against our finding that no regional differences in Ins(1,4,5)P 3 sensitivity exist. Several isoforms of the Ins(1,4,5)P 3 receptor are known to be expressed in pancreatic acinar cells (37,38). Therefore, multiple isoforms are likely to be translated into functionally operating receptors in pancreatic acinar cells, possibly with nonhomogeneous distributions within the cell. So far, immunocytochemistry with antibodies directed against Ins(1,4,5)P 3 receptors has shown that these receptors are present in the apical pole of pancreatic and airway gland acinar cells (24,39). In pancreatic acinar cells only type 3 Ins(1,4,5)P 3 receptors were detected, with no evidence being found for the presence of type 1 Ins(1,4,5)P 3 receptors (24). This observation is surprisingly for two reasons: (i) Ins(1,4,5)P 3 sensitivity in basolateral areas has been demonstrated many times in intact cells (5,6) and in permeabilized cells (this study) and (ii) more than half of the Ins(1,4,5)P 3 receptor mRNA expressed in the whole pancreas was mRNA of type 1 receptors (37,38).
The present study is, however, consistent with Ins(1,4,5)P 3 binding studies in a pancreatic microsomal fraction, which revealed the presence of a single class of binding sites (20). In addition, the dose-response curve for Ins(1,4,5)P 3 -induced Ca 2ϩ release gave no indications of the presence of multiple binding sites (e.g. Refs. 19 and 20). Multiple Ins(1,4,5)P 3 binding sites and a broad dose-response relationship for Ins(1,4,5)P 3 -induced Ca 2ϩ release would be expected if the intrinsic properties of Ins(1,4,5)P 3 receptor subtypes included different Ins(1,4,5)P 3 sensitivities. If the spatial pattern of Ca 2ϩ signaling observed in pancreatic acinar cells is not a result of the intrinsic properties of Ins(1,4,5)P 3 receptors, it must involve additional cytosolic factors regulating the opening of these receptor-operated ion channels. Multiple kinases and cytosolic factors like Ca 2ϩ are known to be involved in the complex regulation of Ins(1,4,5)P 3 receptors (1,40,41). In peripherial tissues, cytosolic Ca 2ϩ levels and/or the phosphorylation status of Ins(1,4,5)P 3 receptors have been shown to play an important role in controlling Ca 2ϩ release mechanisms (42,43).
In conclusion, imaging of Ca 2ϩ within intracellular stores revealed that Ins(1,4,5)P 3 -sensitive Ca 2ϩ stores are found in all regions of the polarized pancreatic acinar cell. Furthermore, the stores did not display a heterogeneous sensitivity toward Ins(1,4,5)P 3 . The polarized Ca 2ϩ signaling observed in intact acinar cells is therefore likely to be controlled by additional cytosolic factors and/or ryanodine receptors possibly present in the apical pole of acinar cells.