The route of Ca2+ entry during reloading of the intracellular Ca2+ pool in pancreatic acini.

To trace the route of Ca2+ entry and the role of the cytosolic Ca2+ pool in reloading of the internal stores of pancreatic acinar cells, Mn2+ influx into Fura 2-loaded cells and the effect of 1,2-bis(2-aminophenoxyethane-N,N,N',N'-tetraacetic acid (BAPTA) on Ca2+ storage in intracellular stores and reloading were examined. Treatment of acini suspended in Ca2(+)-free medium with carbachol (cell stimulation) or carbachol and atropine (reloading period) resulted in 2-fold increase in the rate of Mn2+ influx. Increasing Ca2+ permeability of the plasma membrane by elevation of extracellular pH from 7.4 to 8.2 further increased the rate of Mn2+ influx observed during cell stimulation and the reloading period. Loading the acini with BAPTA by incubation with 50 microM of the acetomethoxy form of BAPTA (BAPTA/AM) was followed by a transient reduction in free cytosolic Ca2+ concentration ((Ca2+]i). To compensate for the increased Ca2+ buffering capacity in the cytosol the acini incorporated Ca2+ from the external medium. Although BAPTA prevented changes in free cytosolic Ca2+ concentration during carbachol and atropine treatment, it had no apparent effect on Ca2+ content of the internal stores or the ability of agonists to release Ca2+ from these stores. Loading the cytosol with BAPTA considerably reduced the rate of Ca2+ reloading. These observations are not compatible with direct communication between the medium and the inositol 1,4,5-trisphosphate releasable pool and provide direct evidence for Ca2+ entry into the cytosol prior to its uptake into the intracellular pool, both during cell stimulation and the Ca2+ reloading.

To trace the route of Ca2+ entry and the role of the cytosolic Ca2+ pool in reloading of the internal stores of pancreatic acinar cells, Mn2+ influx into Fura 2loaded cells and the effect of 1,2-bis (2-aminophenoxyethane-N,N,N',N'-tetraacetic acid ( Stimulation of non-excitable cells by Ca2+ mobilizing agonists results in an increase of [Ca*+]il due to activation of at least 2 passive pathways. The first involves release of Ca2+ from intracellular stores by In-1,4,5-P3 (1). This second messenger activates a Ca*+ channel (2, 3) either in a discrete component of the endoplasmic reticulum (4,5) or in a separate organelle, the calciosome (6) HEPES, 4-(2.hydroxyethyl).l-piperazineethanesulfonic acid.
second pathway is across the plasma membrane and therefore results in Ca*+ entry into the cells by a receptor operated Ca2+ channel or pathway (7)(8)(9)(10)(11)(12)(13)(14). In the case of pancreatic exocrine cells, the latter pathway has been shown to be essential for the sustained phase of secretion, the former for the initial phase of secretion (15,16). During stimulation, regulation of intracellular Ca*+ also involves active transport. It has been shown that stimulation also results in changes in the turnover rate of both the plasma membrane Ca2+ pump (17, 18) and the Ca*+ pump in the membrane of the intracellular Ca*+ store (19-21). Restoration of the resting state in such cells requires reloading of the intracellular Ca2+ store which has been shown universally to depend on Ca*+ uptake from the medium, thus requiring Ca2+ transit across the plasma membrane. Indeed it has been shown that stimulation enhances plasma membrane Ca2+ entry by 5-7-fold (7, 18) and that this increased flux is maintained as long as the internal stores are depleted of Ca'+, even long after cell stimulation has terminated (9, 13). Reuptake of Ca*+ into the intracellular pool also requires maintained activity of the pool Ca*+ pump which is increased during stimulation (19)(20)(21). The mechanism of regulation of Ca*+ reloading has remained controversial.
Reloading of the intracellular store occurs at a constant level of [Ca"]i, identical to the resting level (9,22-24). Since Ca*+ entry is required for reloading, this finding has been taken to mean that there was a direct entry of medium Ca*+ into the intracellular store, without transit through the cytoplasm, a "privileged" pathway for Ca2+ (22,25), or that Ca*+ entry occurs through a restricted region where the endoplasmic reticulum and the plasma membrane are closely opposed (26). This hypothesis would require direct contact between the Ca2+ stores membrane and the plasma membrane or a vesicle shuttling mechanism between the two membranes. However, if the pool Ca2+ pump remains activated during reloading and is dominant with respect to Ca*' entry, it is quite possible to conceive of reloading at resting or even below resting [Ca*+]*.
Using Mn2+ entry as an index of plasma membrane Ca2+ permeability and a cytosolic Ca2+ chelator such as BAPTA, it is possible to exclude a direct pathway for Ca*' reloading. During reloading of pancreatic acini, we have shown an increased rate of Mn*+ influx. Further, BAPTA decreases the rate but not the extent of reloading of the cell Ca'+ store. The pathway for reloading must therefore transit the cytosolic compartment of the cell, and regulation of [Ca'+], during reloading is due to the properties of the plasma membrane entry pathway and the pool Cap+ pump. The acini were then dissolved by heating at 60 "C in 1 ml of 1 M NaOH for 20 min, and 45Ca was counted using standard liquid scintillation counting.

Plasma Membrane
Permeability-The rate and extent of quenching of Fura 2 fluorescence by Mn*+ was used to determine the effect of agonists on plasma membrane permeability to Ca'+. Fig. 1 shows the entry of Mn2+ into pancreatic acini under different conditions. In the absence of any added agonist, there was a relatively slow quench of intracellular Fura 2. This shows that there is a basal plasma membrane Mr?' permeability even in the absence of agonists.
Following maximal carbachol stimulation in this Ca'+-free medium, there was a rapid increase in [Ca*+]& due to release of intracellular Ca*+. Subsequently, the cells reduced [Ca2+li to slightly below resting levels, due to Ca*+ export across the plasma membrane. When Mn2+ was added to these stimulated cells, there was a more rapid quenching of Fura 2 fluorescence. This can be interpreted as showing an increased rate of Ca2+ entry due to stimulation ( Fig. 1 b). A similar result was seen when atropine was used to terminate the stimulation. Thus, following stimulation by carbachol and inhibition by atropine (to initiate reloading), the addition of Mn*+ resulted in a larger rate of quench than in the absence of any treatment (Fig. lc). Table I   showing that the stimulated cells, whether treated subsequently with atropine or not, had a larger quantity of intracellular Mn*+. This is consistent with the enhanced rate of Mn2+ entry. Furthermore, this shows that the cells are not able to actively export Mn2+.
An increase of extracellular pH increases Ca2+ entry through agonist activated Ca2+ channels (31). The experiments in Fig. 2 were therefore carried out at pH 8.2. It can be seen that the rate of Mn2+ entry was enhanced in each of the experimental situations.
The resting rate of Mn'+ induced quench was increased by 1.44-fold. The entry, post carbachol or carbachol and atropine treatments, was enhanced by approximately 36fold. Furthermore, following stimulation, digitonin had no effect on Fura 2 fluorescence showing that the  Fig. 3, b and c, show that BAPTA interacts with Fura 2. When recorded at excitation wavelengths of 340 (Fig. 3b) and 385 nm (Fig. 3c), BAPTA increased fluorescence at both 340 and 385 nm, the increase at the latter wavelength being greater. It has been suggested that these data are due to sustained reduction of [Ca2+], by BAPTA (33). However, when cells were loaded with Fluo 3 in Ca*' containing medium, there was only a transient decrease in the Fluo 3 signal showing that over the time period of the experiment in Fig.  3, b and  fluorescence is less than that of Fura 2 (not shown), all further fluorescence experiments were done using Fura 2. In six different experiments it was found that incubating acini with 50 pM BAPTA/AM for 15 min in the absence of Ca*' or 20 min in the presence of Ca2+ in the incubation medium was sufficient to completely buffer changes in [Ca'+] upon stimulation with carbachol. At intermediate loading of bAPTA, the effect of cytosolic Ca*+ chelation on the [Ca*+]i signal can be explored. Fig. 3, b and c, shows that incubating the acini with 50 pM BAPTA/AM for 7 min significantly reduced the [Ca2+li increase induced by carbachol. The Fluo 3 data show that at the time of carbachol addition resting [Ca'+]i had been restored. The fall in [Ca*+]i induced by atropine returned to base line but with a slower time course than that observed without BAPTA (Fig. 3 a-c). Further, there was only a small increment in [Ca"], when CCK-OP was added (Fig. 3, b and c) in contrast to the equivalent release seen in the absence of BAPTA (Fig. la). This could be due to lower Ca2+ content of the stores or due to buffering of [Ca'+]i. To examine the various possibilities, studies were performed with 45Ca.
Effect of Ca*' Chelation on %a Response-It has been shown that an efficient means of labeling the intracellular pool with 45Ca2+ is to cycle the acini through an unloading phase with carbachol and then allow loading in the presence of atropine and %a*' (18, 34). The result of such labeling is illustrated in Fig. 4, where the bottom curve shows direct uptake in the absence of cycling and the upper curve shows the effect of cycling. Following 45Ca2f labeling by the cycling procedure, BAPTA loading, and suspending the acini in 45Ca2+-free medium, there is a gradual loss of tracer Ca*+. In control acini, when no BAPTA is present, a steady state of counts is reached at about 80% of initial value, with a tlh of about 1.5 min. The addition of CCK-OP now gave a rapid release of the counts, with about 10% of initial 45Ca remaining associated with the acini 3 min after the addition of CCK-OP (Fig. 5). In case of the BAPTA containing cells, about 35% of the counts were lost initially with a tlh of about 3 min. However, following CCK-OP there was a marked reduction in the rate of 45Ca2' loss as compared to the rate in the absence of BAPTA, although the final level was similar. This is consistent with buffering of [Ca2+li by BAPTA and consequent reduction of plasma membrane Ca2+ pump activity accounting for the lower rate of loss of Ca*+ in the BAPTA-treated cells (Fig. 5). To exclude further that the effect of BAPTA was due to a reduction in the size of the intracellular pool, and to provide evidence that BAPTA affected the rate, but not the magnitude of reloading, the rate of 45Ca uptake was studied. Acini were suspended in nominally Ca*+-free medium, stimulated with carbachol to deplete the internal Ca2+ pool, and loaded with BAPTA.
Subsequent to BAPTA loading, Ca*+ reloading was initiated by the addition of a mixture of atropine and Ca2+ labeled with 45Ca. BAPTA increased the extent of Ca2+ uptake into unstimulated acini but decreased the rate of Ca*+ reloading (Fig. 6A). To estimate the effect of BAPTA on reloading rate, 45Ca uptake into control acini was subtracted from 45Ca uptake into cycled acini (Fig. 6B). It can be seen that BAPTA decreased the reloading rate by approximately 3-fold although the internal stores completely reloaded with Ca2+ within 10 min of incubation at 37 "C. in nominally Ca*+-free solution A were stimulated with carbachol for 5 min at 37 "C. Then 2 mM CaC12 (+45Ca) and 20 PM atropine were added. After 7.5 min of incubation at 37 "C a portion of the acini was transferred to a tube containing BAPTA/ AM to a final concentration of 50 PM and the incubation continued for an additional 15 min. The "cycled" (0) and "cycled" BAPTAloaded acini (A) were then collected by centrifugation and resuspended in solution A containing 2 mM of unlabeled CaCl*. After 5 min of incubation a portion of acini from each group was stimulated with 10 nM CCK-OP (0, A). At the indicated times samples were removed to measure 45Ca2+ content in the acini. The initial 45Ca content of "cycled" acini was taken as 100% control and "5Ca content was calculated as % of control. The figure shows the mean + S. E. of three separate experiments.  Non-excitable cells respond to Ca2+ mediated agonists in general by releasing intracellular Ca2+ and by increasing Ca*+ entry across the plasma membrane; the proportionality between these two pathways varies between cells and between agonists. Substantial evidence indicate that Ca2+ release from intracellular stores is mediated by the production of In-1,4,5-Pz in the cytosol (1). In contrast, there appear to be a variety of mechanisms involved in agonist stimulated Ca2+ entry into non-excitable cells. For example, In-1,4,5-P3 has been reported to induce Ca*+ entry in lymphocytes (35) and mast cells (36). In lacrimal acinar cells (37) both In-1,4,5-P3 and In-1,3,4,5-P* appeared to be necessary for activation of Ca*+ entry. In hepatocytes, glucagon and epinephrine appear to induce two separable pathways of Ca*+ influx (38).
A property which must exist in all such cells, however, is a Ca*+ pathway for reloading of intracellular stores following discharge by In-1,4,5-P3. A major question that has arisen is the anatomical properties of the reloading pathway. One model has suggested direct communication between the medium and the In-1,4,5-P3 releasable pool (22,25,26). Another, more conventional model, view that uptake of Ca*+ occurs downhill across the plasma membrane into the cytosol and hence, by pump mediated uptake into the intracellular Ca*+ store of the endoplasmic reticulum or calciosome (12,39). A third alternative is a vesicle cycling model whereby Ca*+ is transported into the pool by means of endosomal cycling between plasma membrane and endoplasmic reticulum. This latter suggestion arose from the observation that In-1,3,4,5-P4 was able to augment Ca*+ influx only in cells where the internal stores were depleted by In-1,4,5-P3 (37). It was hypothesized that In-1,3,4,5-P4 facilitated fusion between organelles containing the In-1,4,5-P3 activated channel (perhaps calciosomes) and a portion of the endoplasmic reticulum attached to the plasma membrane (40).
In both the direct communication model and in the cycling model, the loading of the endoplasmic reticulum should depend directly on the Ca*+ content of the medium. However, the extent of reloading appears constant over wide variations in medium Ca2+ concentration (0.5-2 mM) (9). Inaddition, during the reloading period, Ca*+ influx into the pool is at least 12-fold faster than Ca*+ efflux from the pool (18, 34). Furthermore, the direct pathway would seem to remove the need for a Ca2+ pump, given the normal constancy of extracellular Ca2+. In both the direct and the cycling model, additional regulatory systems would have to be invoked, involving stimulation and inhibition of cycling or of direct communication. On the other hand, for the series model the maintained reloading at resting cytosolic Ca*+ level (9,22-24) necessitates a determination of accessibility of the cytosolic pool during reloading and the effect of buffering the cytosolic Ca2+ pool on the properties of the intracellular stores.
Recently, the effect of the tumor promoter thapsigargin and methacholine on Ca2+ entry into parotid acinar cells were compared (41). It was concluded that these compounds activate Ca2+ entry by depletion of the intracellular Ca*+ pool through different mechanisms (41). However, the role of the thapsigargin-activated Ca*+ entry pathway in reloading was not defined, and these experiments did not investigate the role of the cytosolic Ca*+ pool in reloading. Here, we have examined both Mn2+ and BAPTA effects as indices of the role of the cytoplasmic Ca2+ pool in the reloading of the intracellular Ca*+ store.
In the pancreatic acini and in other cell types Mn2+ appears to act as a congener of Ca*+ in passive Ca2+ pathways (14,32,42). In the particular case of the pancreas, Mn2+ entry is stimulated by agonists such as carbachol or CCK and similar to Ca*' (9), Mn*+ entry remains stimulated after atropine inhibition of carbachol stimulation. As for Ca2+ (31), Mn2+ entry is enhanced by elevation of medium pH. Inhibition of carbachol stimulation of In-1,4,5-P3 levels is achieved within l-2 min following atropine (43,44). Nevertheless, there is an increased rate of Mn2+ entry into the cytosolic pool which is maintained following atropine inhibition of the carbachol response. Recently, while these studies were in progress, similar observations were reported by studying the effect of agonist and antagonist on Mn2+ entry into umblical-vein endothelial cells (14). In these cells, Ca*+ and Mn*+ entry during stimulation and reloading were significantly higher than those observed with pancreatic acinar cells. Thus, despite the differences in the absolute change in plasma membrane permeability, in both cell types Mn2+ enters the cytosolic pool during the reloading phase. These data are consistent with reloading of the stores by Ca2+ entry across plasma membrane, mixing of this Ca*+ with the cytosolic Ca*+ pool and uptake into the ER system by the Ca2+ pump. Mn2+ would act as a Ca*+ cogener in terms of entry across plasma membrane and mixing with the Ca2' in the cytosol but is not incorporated into the store since the Ca2+ pump poorly transport Mn2+ (45). However, these data do not show directly the involvement of the cytosolic Ca2+ pool in reloading.
A second line of evidence suggesting that a series model is correct and that the cytosolic pool is used for reloading was obtained by using an intracytoplasmic Ca2+ buffer. BAPTA/ AM does not act as a buffer outside the cell. However, following cellular uptake and release of BAPTA, the data presented above show that BAPTA acts as a buffer for cytosolic Ca2+. The chelator did not change the steady state level of [Ca2+]i, as monitored by the fluorescence of Fluo 3. Nevertheless, the presence of BAPTA resulted in an attenuated, but normal time course for the initial carbachol-induced Ca2+ increase, but slowing of the atropine response and marked attenuation of the second release of intracellular Ca*+ by CCK. These data could be interpreted as an effect of cytosolic buffering on the reloading of the Ca2+ stores, consistent with a series model, not consistent with a communication or cycling model.
The presence of a changed cytosolic buffering capacity due to BAPTA makes quantitative comparisons with control conditions difficult when using another buffer, Fura 2, to monitor changes in [Ca2+]i. This difficulty was overcome using 45Ca2+ measurements. %!a*' fluxes show that BAPTA did not affect the amount of Ca*+ stored in the releasable pool (Fig. 4) nor did BAPTA affect the amount of Ca*+ released by CCK-OP (Fig. 5). If BAPTA had been included in the pool and buffered Ca2+ in the intracellular store, both quantities should have changed. Thus, BAPTA/AM does not appear to enter the site of the Ca*+ storage directly, or if it does, it is not hydrolyzed by esterases in the store, nor does a significant amount of free BAPTA accumulate in the site of Ca*+ storage. The presence of BAPTA in the cytosol of the pancreatic acinar cells however, slowed the rate of reloading of the intracellular Ca2+ store by about 3-fold. This is consistent with a buffering action of BAPTA on the Ca2+ being used for reloading the store. Since BAPTA is acting as a cytosolic buffer for Ca'+, it appears that these data also are consistent with the series model and exclude the direct or cycling models of the Ca2+ loading.
Since [Ca'+], is maintained at or just below the resting level during Ca2+ reloading (9,22-24), there must be a fine balance between the rate of Ca2+ entry into the cytoplasm and removal of Ca2+ from the cytoplasm by the pool Ca2+ pump. Furthermore, this reloading can occur long after termination of stimulation (9,13,22). It is, therefore, likely that the regulatory mechanisms are built into the pathway itself, rather than being dependent on residual effects of agonists. This suggests a link between the plasma membrane Ca2+ pathway responsible for reloading the intracellular Ca2+ store, and the store itself or the Ca*+ pump involved in store replenishment. The most likely candidate for such a regulator would be the level of Ca*+ itself.
In summary, Mn2+ entry into the cytoplasm is enhanced during reloading, and cytosolic Ca2+ chelation by BAPTA reduced the rate of Ca*+ reloading into intracellular Ca*+ stores, without affecting the size of the store. These data exclude models where there is direct communication between extracellular medium and an In-1,4,5-P3 releasable Ca2+ stores and favor a model whereby reloading is achieved by entry into and uptake from the cytoplasm of the cell.