Kinetic Control of Multiple Forms of Ca²⁺ Spikes by Inositol Trisphosphate in Pancreatic Acinar Cells

Abstract

The mechanisms of agonist-induced Ca²⁺ spikes have been investigated using a caged inositol 1,4,5-trisphosphate (IP₃) and a low-affinity Ca²⁺ indicator, BTC, in pancreatic acinar cells. Rapid photolysis of caged IP₃ was able to reproduce acetylcholine (ACh)-induced three forms of Ca²⁺ spikes: local Ca²⁺ spikes and submicromolar (<1 μM) and micromolar (1–15 μM) global Ca²⁺ spikes (Ca²⁺ waves). These observations indicate that subcellular gradients of IP₃ sensitivity underlie all forms of ACh-induced Ca²⁺ spikes, and that the amplitude and extent of Ca²⁺ spikes are determined by the concentration of IP₃. IP₃-induced local Ca²⁺ spikes exhibited similar time courses to those generated by ACh, supporting a role for Ca²⁺-induced Ca²⁺ release in local Ca²⁺ spikes. In contrast, IP₃- induced global Ca²⁺ spikes were consistently faster than those evoked with ACh at all concentrations of IP₃ and ACh, suggesting that production of IP₃ via phospholipase C was slow and limited the spread of the Ca²⁺ spikes. Indeed, gradual photolysis of caged IP₃ reproduced ACh-induced slow Ca²⁺ spikes. Thus, local and global Ca²⁺ spikes involve distinct mechanisms, and the kinetics of global Ca²⁺ spikes depends on that of IP₃ production particularly in those cells such as acinar cells where heterogeneity in IP₃ sensitivity plays critical role.

Agonist receptors induce the release of Ca²⁺ from intracellular stores and thereby generate Ca²⁺ spikes, waves, or oscillations that play important roles in many cellular functions (Berridge 1993; Petersen et al. 1994; Clapham 1995). It is thought that positive feedback effects of Ca²⁺ on Ca²⁺-release channels, including both inositol 1,4,5-trisphosphate (IP₃)¹ (Iino 1989; Bezprozvanny et al. 1991) and ryanodine receptors (Endo et al. 1970), result in Ca²⁺-induced Ca²⁺ release (CICR) and contribute to the generation of such Ca²⁺ responses. Indeed, local Ca²⁺ release events induced by IP₃, such as puffs (Callamaras et al. 1998) and local Ca²⁺ spikes (Kasai et al. 1993; Thorn et al. 1993), are likely attributable to CICR mechanisms at IP₃ receptors, because they can be induced at constant concentrations of IP₃ (Wakui et al. 1989). However, it has remained unclear whether the generation of global Ca²⁺ spikes is also explained by CICR mechanisms (Bootman et al. 1997).

Pancreatic acinar cells represent an ideal system for investigating the mechanisms of agonist-induced generation of Ca²⁺ spikes. First, agonist-induced increases in the cytosolic concentration of Ca²⁺ ([Ca²⁺]_i) in these cells are mostly attributable to the generation of IP₃ from phosphatidylinositol 4,5-bisphosphate in a reaction catalyzed by phospholipase C (PLC) (Petersen 1992). Second, Ca²⁺ release channels are heterogeneously distributed along the polarized intracellular structures (Kasai et al. 1993), resulting in a fixed pattern of Ca²⁺ spike spread. The spikes are always initiated at the trigger zone, the apical pole of the secretory granule–containing region of the cell (Kasai and Augustine 1990; Nathanson et al. 1992; Toescu et al. 1992). Thus, the functioning of distinct Ca²⁺-release channels can be directly visualized. And third, agonists induce multiple forms of Ca²⁺ spikes in a dose-dependent manner; they can be local or global (Kasai et al. 1993; Thorn et al. 1993). Increases in [Ca²⁺]_i remain restricted to a discrete area or expand to entire cells in the local and global Ca²⁺ spikes, respectively. The global Ca²⁺ spikes further manifest at submicromolar or micromolar concentrations of Ca²⁺ (Ito et al. 1997). The existence of multiple forms of Ca²⁺ spikes in the acinar cells enables us to investigate their mechanisms in the same experimental conditions.

We have now characterized the Ca²⁺ spikes induced by spatially uniform and rapid increases in [IP₃]_i, generated by photolysis of caged IP₃, and compared them with the Ca²⁺ spikes induced by a natural stimulus, acetylcholine (ACh). If CICR mechanisms play a dominant role in ACh-induced Ca²⁺ spikes, then the time course of such spikes should resemble that of those induced by IP₃. We found that this was indeed the case for local Ca²⁺ spikes, but not for global Ca²⁺ spikes. Ca²⁺ imaging was performed with a low-affinity Ca²⁺ indicator, benzothiazole coumarin (BTC), that minimizes the effects of changes in intrinsic Ca²⁺ buffering in the cells and allowed us to quantify large increases in [Ca²⁺]_i without the problem of dye saturation (Ito et al. 1997; Kasai and Takahashi 1999).

Materials and Methods

Preparation of Acinar Cells

Acinar cells were dissociated from the pancreas of 5–7-wk-old mice by enzymatic treatment as described (Ito et al. 1997). For electrophysiological recording, the cells were dispersed in a small chamber in a solution (Sol A) containing 140 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM Hepes-NaOH (pH 7.4), and 10 mM glucose. ACh (Wako) was dissolved in Sol A and applied to cells through a glass pipette. Ca²⁺ indicators, fluo-3 or BTC (Molecular Probes), were dissolved in a solution (basic internal solution) containing 120 mM cesium glutamate, 5 mM CsCl, 50 mM Hepes-CsOH (pH 7.2), 1 mM ATP, 0.2 mM GTP, and 2 mM MgCl₂, and were then loaded into cells at a concentration of 200 μM by the patch clamp method. Caged IP₃ [d-myo-inositol 1,4,5-trisphosphate, P⁴⁽⁵⁾-1-(2-nitrophenyl)-ethyl ester; Calbiochem-Novabiochem] or caged GPIP₂ [1-(α-glycerophosphoryl)-d-myo-inositol 4,5-bisphosphate, P⁴⁽⁵⁾-1-(2-nitrophenyl)-ethyl ester; Calbiochem-Novabiochem] was also added to the basic internal solution. Osmolarities of the external and internal solutions were estimated to be ∼310 mOsM after addition of all chemicals (Semi-Micro Osmometer; Knauer). All experiments were performed under yellow light illumination (FL40S-Y-F; National) at room temperature (22–25°C).

Ca²+ Imaging

Confocal Ca²⁺ imaging was performed as described (Kasai et al. 1993), with the exception that fluo-3 was used as the Ca²⁺ indicator. Fluorescence from patch-clamped acinar cells was detected with a confocal laser scanning microscope (MRC-600; Bio-Rad) attached to an inverted microscope (IMT-2; Olympus) with an objective lens (DApo 40× UV/340 oil; Olympus). Fluo-3 was excited with an argon laser at 488 nm, and [Ca²⁺]_i was calculated from the ratio of fluorescence values during stimulation (F) to that obtained before stimulation (F ₀) according to the equation

1

where K and [Ca²⁺]₀ were assumed to be 0.39 and 0.1 μM, respectively. Values of F _max/F _min were estimated in vivo by assuming that the maximal [Ca²⁺]_i achieved in the presence of ACh (10 μM) was 10 μM (see Fig. 6A and Fig. B). The mean value of F _max/F _min thus obtained was 6.5 and was used to calibrate local Ca²⁺ spikes induced with a low concentration of IP₃ (see Fig. 1 A).

Figure 6.

Line-scan analysis of global Ca²⁺ spikes induced by ACh or caged IP₃. (A, C, and E) Increases in [Ca²⁺]_i recorded with the line-scan mode of the confocal microscope and with fluo-3 as the Ca²⁺ indicator. The trigger zone and basal area of the acinar cells are denoted by T and B, respectively. The cells were stimulated either with 10 μM ACh (A), or by rapid (C) or slow (E) photolysis (indicated by tracings above images) of 100 μM caged IP₃. (B, D, and F) Time courses of [Ca²⁺]_i in the trigger zones and basal areas for the cells shown in A, C, and E, respectively.

Figure 1.

Local Ca²⁺ spikes induced by photolysis of caged IP₃ in pancreatic acinar cells. Local increases in [Ca²⁺]_i were induced by photolysis of 5 μM caged IP₃. (A) Ca²⁺ images obtained with a confocal microscope and the high-affinity Ca²⁺ indicator fluo-3. (B) Time courses of [Ca²⁺]_i within the rectangles shown in A. (C) Ca²⁺ images obtained with a cooled CCD camera and the low-affinity Ca²⁺ indicator BTC. (D) Time courses of [Ca²⁺]_i within the rectangles shown in C. Dashed white lines in the black and white photographs shown in A and C indicate the secretory granule region of the cell. Vertical white bars in B and D indicate the times when the images shown in A and C were obtained; the arrow indicates the time of photolysis of caged IP₃ induced by ultraviolet (UV) irradiation.

Ca²⁺ imaging with a cooled CCD camera was performed as described (Ito et al. 1997). In brief, a recording chamber was placed on an inverted microscope (IX; Olympus) and observed through an objective lens (DApo 40× UV/340 oil). The [Ca²⁺]_i was measured with the Ca²⁺ indicator BTC. Monochromatic beams with wavelengths of 430 or 480 nm were isolated from light emitted by a xenon lamp with the use of a polychromator (T.I.L.L. Photonics), and were fed into one port of a light guide (IX-RFA caged; Olympus). The light was reflected by a dichroic mirror (DM500) placed beneath the objective lens, and fluorescent light emitted from the cells was captured with a cooled CCD camera system (T.I.L.L. Photonics) fixed at the side port of the microscope. The duration of image acquisition was 0.12 s, and the pairs of images were acquired every 0.24 s. [Ca²⁺]_i was estimated from BTC fluorescence as described (Ito et al. 1997). Calibration constants for BTC were R _max = 2.0 and K _Bβ = 112. To obtain Ca²⁺ images from BTC fluorescence, we first estimated the distribution of R _min in individual cells by averaging several frames of the resting distribution of R. This procedure was used to compensate for small heterogeneity in R _min within a cell, and to reduce noise levels, particularly at [Ca²⁺]_i values of <1 μM. The mean value of R _min (m[R _min]) was ∼0.55. Distributions of ΔR were then calculated by subtracting the distribution of R _min from that of R. From ΔR, [Ca²⁺]_i was estimated as K _Bβ·ΔR/(R _max – m[R _min] –ΔR).

The [Ca²⁺]_i in Ca²⁺ images was represented by pseudocolor coding, where 0.1, 0.3, 1, 3, and 10 μM were expressed as blue, sky blue, green, yellow, and red, respectively (Fig. 1, Fig. 2, Fig. 3, and Fig. 6).

Figure 2.

Global Ca²⁺ spikes or Ca²⁺ waves induced by photolysis of caged IP₃. (A and B) Homogeneous increase in [Ca²⁺]_i induced by photolysis of 100 μM caged IP₃. (C–H) Ca²⁺ waves induced by photolysis of 100 μM (C and D), 50 μM (E and F), or 10 μM (G and H) caged IP₃. Ca²⁺ images were obtained with a cooled CCD camera and BTC. Data are presented as described in the legend to Fig. 1.

Figure 3.

ACh-induced global Ca²⁺ spikes. Ca²⁺ waves induced by exposure of acinar cells to ACh at concentrations of 10 μM (A and B), 1 μM (C and D), or 50 nM (E and F) were imaged with a cooled CCD camera and BTC. The horizontal bars in (B), (D), and (F) indicate the duration of exposure to ACh.

Photolysis of Caged IP₃

We used a mercury lamp (IX-RFC or IMT-2-RFC; Olympus) as an actinic light source for photolysis of caged IP₃. Light from the mercury lamp was filtered through a 360-nm band-pass filter and fed into the second port of the light guide (IX-RFA caged or IMT-2-RFC caged; Olympus). Incorporation of a dichroic mirror (DM400) allowed the light guide to accommodate two light sources, one for photolysis of IP₃ and the other for excitation of the Ca²⁺ indicator. Illumination from the actinic light was gated through an electric shutter (Copal). We estimated that irradiation for 125 ms was necessary and sufficient for full activation of caged IP₃. For this calibration experiment, the irradiation was restricted to a recorded cell and not applied to a patch pipette to facilitate recovery of [IP₃]_i through the pipette, and photolysis was intermittently applied to the same cells. We found that Ca²⁺ responses depended on the duration of the irradiation, and reached the maximal response at 125 ms. In most experiments, we therefore set the duration of the opening of the shutter at 125 ms to achieve complete photolysis of caged IP₃, and the irradiation was applied to whole objective field including the tip of patch pipette to maintain [IP₃]_i constant as long as possible. In some experiments, a neutral density filter (10, 20, or 50%) was used to reduce the light intensity, in which case the concentration of photolyzed IP₃ was obtained by multiplying the concentration of caged IP₃ introduced into the cells by the relative light intensity. Only those data obtained from the first photolysis were used to avoid complications of preceding Ca²⁺ spikes.

Results

IP₃-Induced Local Ca²⁺ Spikes

We first investigated whether homogeneous and constant increases in [IP₃]_i could produce local Ca²⁺ spikes in the secretory granule area of pancreatic acinar cells similar to those induced by ACh. Photolysis of caged IP₃ was induced 2–5 min after the establishment of whole-cell perfusion, at which time the concentration of IP₃ in the cell should be equilibrated with that in the patch pipette. We monitored [Ca²⁺]_i with a confocal microscope and a high-affinity Ca²⁺ indicator dye, fluo-3. Local increases in [Ca²⁺]_i confined to small spots within the secretory granule area were detected immediately after photolysis of 5 μM caged IP₃ (Fig. 1 A). The spatial pattern of the IP₃-induced local Ca²⁺ spikes was similar to those induced by ACh (n = 7, data not shown; Kasai et al. 1993). The result is in accord with previous studies in which IP₃ was microinjected into the cells (Kasai et al. 1993; Thorn et al. 1993).

The time course of local Ca²⁺ spikes induced by photolysis of caged IP₃ (Fig. 1A and Fig. B) also was similar to that of ACh-induced local Ca²⁺ spikes (Kasai et al. 1993). We believe that IP₃-induced Ca²⁺ spikes per se do not cause IP₃ production, because, in the absence of receptor stimulation, increases in [Ca²⁺]_i alone could not give rise to the Ca²⁺ gradients characteristics of IP₃-induced Ca²⁺ spikes (Toescu et al. 1992; Maruyama et al. 1993). Thus, we believe that [IP₃]_i stays constant during IP₃-induced Ca²⁺ spikes, and that local Ca²⁺ spikes were mediated by CICR mechanisms as reported (Wakui et al. 1989; Thorn et al. 1996). The increases in [Ca²⁺]_i were always transient in the experiments described in this study. The transient nature of the responses is likely attributable to desensitization of IP₃ receptors, given that photolyzed caged IP₃ was continuously perfused from the patch pipette and that a metabolically stable analogue of caged IP₃, caged GPIP₂, also induced transient increases in [Ca²⁺]_i (n = 5, data not shown). Concentrations of caged IP₃ of <1 μM did not trigger detectable increases in [Ca²⁺]_i.

The local Ca²⁺ spikes also could be detected with the use of the low-affinity Ca²⁺ indicator BTC and a cooled CCD (charge-coupled device) camera (Fig. 1C and Fig. D). A focal and transient increase in [Ca²⁺]_i of ∼0.5 μM was detected in the trigger zone in response to photolysis of caged IP₃ (n = 5). The increases in [Ca²⁺]_i were confirmed by the appearance of Ca²⁺-dependent Cl⁻ currents (data not shown). The detection of local Ca²⁺ spikes with BTC allowed us to make a direct comparison with their properties with those of global Ca²⁺ spikes recorded with BTC.

IP₃-Induced Global Ca²⁺ Spikes

We next examined the effects of rapid photolysis of larger concentrations of IP₃ (10–100 μM). Ratiometric Ca²⁺ imaging with BTC was used for reliable estimation of amplitudes and time courses of changes in [Ca²⁺]_i persisting for >20 s. Because of substantial cell-to-cell variability in the responses, these experiments were performed with a large number of cells (n = 41). Photolysis of 100 μM caged IP₃ often resulted in large increases in [Ca²⁺]_i throughout the cell that were apparent within 0.24 s (Fig. 2A and Fig. B), the earliest time at which an image was collected by the CCD camera. The Ca²⁺ indicator (BTC) was not saturated with Ca²⁺ at these concentrations (Fig. 2 A), and it can therefore be concluded that the increases in [Ca²⁺]_i were relatively homogeneous and exceeded 10 μM throughout the cell. Thus, the capacity for Ca²⁺ release appeared to be distributed homogeneously throughout the cell. The abundance of IP₃ receptors in the basal area was also supported by the previous observation that IP₃ injection could directly trigger Ca²⁺ release in the basal area (Fig. 6 C of Kasai et al. 1993).

Photolysis of caged IP₃ at concentrations between 10 and 100 μM induced Ca²⁺ spikes that were initiated at the trigger zone (Fig. 2C, Fig. E, and Fig. G) as in the case with ACh-induced Ca²⁺ spikes. In fact, Ca²⁺ concentrations immediately (0.24 s) after photolysis of caged IP₃ were always larger in the trigger zone than in the basal area (Fig. 2D, Fig. F, and Fig. H). Furthermore, the initial Ca²⁺ concentrations in the trigger zone (initial [Ca²⁺]_t) and the basal area (initial [Ca²⁺]_b) depended on [IP₃]_i with median effective concentrations of 5 and 50 μM, respectively (Fig. 4A and Fig. B). These data suggest that IP₃ receptors in the basal area were ∼10 times less sensitive to IP₃ than those in the trigger zone. Gradual increases in [Ca²⁺]_i were detected throughout the cells after photolysis of caged IP₃, suggesting positive feedback effect of Ca²⁺ on Ca²⁺ release channels.

Figure 4.

Dose dependence of amplitudes of global Ca²⁺ spikes induced by caged IP₃ or ACh. (A and B) Semilog plots of [Ca²⁺]_i immediately (0.24 s) after photolysis of caged IP₃ in the trigger zone ([Ca²⁺]_t) (A) and in the basal area ([Ca²⁺]_b) (B), respectively. The smooth curves were drawn assuming median effective concentrations of 5 and 50 μM, respectively. (C and D) Semilog plots of peak [Ca²⁺]_i in the trigger zone ([Ca²⁺]_t) versus [IP₃]_i (C) or ACh concentration (D). The smooth curves were drawn assuming median effective concentrations of 10 and 0.5 μM for IP₃ and ACh, respectively. (E and F) Correlations between peak [Ca²⁺]_i in the trigger zone ([Ca²⁺]_t) and that in the basal area ([Ca²⁺]_b) for global Ca²⁺ spikes induced by IP₃ or ACh, respectively. The regression coefficients are 0.57 and 0.63 for IP₃ and ACh, respectively.

The peak amplitudes of the IP₃-induced Ca²⁺ spikes also depended on [IP₃]_i (see Fig. 4 C), as those of ACh-induced Ca²⁺ spikes did on the concentration of ACh (see Fig. 4 D). The amplitudes of Ca²⁺ spikes ranged from micromolar, with concentrations of >10 μM in the trigger zone (Fig. 2 C and 3 A), to intermediate (∼5 μM; Fig. 2 E and 3 C), to submicromolar (<1 μM) (Fig. 2 G and 3 E). The amplitudes of the smallest global Ca²⁺ spikes generated by IP₃ or ACh were <1 μM in most regions of the cell (Fig. 2 G and 3 E). The peak amplitudes of ACh-induced increases in [Ca²⁺]_i in the trigger zone were always larger than those in the basal area (Fig. 3 and Fig. 4 F). This Ca²⁺ gradient was not due to the gradient of [IP₃]_i, because similar Ca²⁺ gradients were induced by homogeneous increases in [IP₃]_i induced by caged IP₃ (Fig. 2 and Fig. 4 E). Thus, IP₃ receptors in the basal area was less sensitive to IP₃ than those in trigger zone even at the peak of Ca²⁺ spikes in the respective areas.

Time Courses of Global Ca²⁺ Spikes

Marked differences were evident in the time courses of the global Ca²⁺ spikes induced by caged IP₃ and of those induced by ACh (Fig. 5). First, the time-to-peak for Ca²⁺ spikes at the trigger zone induced by caged IP₃ was <1 s in most experiments, and was independent of [IP₃]_i (Fig. 5 A). In contrast, the time-to-peak for ACh-induced global Ca²⁺ spikes was >1 s in most experiments, and decreased as the concentration of ACh increased (Fig. 5 B). These data indicate that [IP₃]_i increases gradually during ACh stimulation, and that the rate of this increase is dependent on ACh concentration.

Figure 5.

Dose dependence of time courses of global Ca²⁺ spikes induced by caged IP₃ or ACh. (A and B) Semilog plots of the time-to-peak of Ca²⁺ spikes in the trigger zone versus concentration of IP₃ (A) or ACh (B). (C and D) Semilog plots of spike spread time versus concentration of IP₃ (C) or ACh (D). (E and F) Semilog plots of spike delay time versus concentration of IP₃ (E) or ACh (F). Correlation coefficients are 0.066 (P > 0.1), 0.45 (P < 0.001), 0.29 (P > 0.1), 0.63 (P < 0.001), 0.003 (P > 0.1), and 0.61 (P < 0.001) for A through F, respectively.

Second, the spread of Ca²⁺ spikes induced by caged IP₃ was faster than that of those induced by ACh. To quantify the rate of spread of Ca²⁺ spikes (Ca²⁺ waves), we defined the spike spread time as the difference between the times at which the half-maximal [Ca²⁺]_i was achieved in the trigger zone and in the basal area. The spread time for spikes induced by caged IP₃ was <0.7 s in most experiments, and was independent of [IP₃]_i (P > 0.1; Fig. 5 C). In contrast, the spread time for ACh-induced Ca²⁺ spikes was >0.7 s in most experiments, and it decreased as the concentration of ACh increased (Fig. 5 D).

Finally, the onset of Ca²⁺ spikes in the basal area was always delayed relative to that of Ca²⁺ spikes in the trigger zone for cells stimulated with ACh (Fig. 3), whereas little delay was observed for Ca²⁺ spikes induced by caged IP₃ (Fig. 2). We quantified the delay in the onset of Ca²⁺ spikes in the basal area by measuring the difference between the times at which [Ca²⁺]_i reached 0.5 μM in the trigger zone and in the basal area. The spike delay ranged between 0 and 0.24 s for IP₃-induced Ca²⁺ spikes (Fig. 5 E) and between 0.48 and 4 s for ACh-induced Ca²⁺ spikes (Fig. 5 F). Precise measurements of delay and spike spread times were not possible at high IP₃ concentrations with our cooled CCD camera operating at an acquisition interval of 0.24 s.

Line-Scan Analysis of Global Ca²⁺ Spikes

Therefore, we applied the line-scan mode of confocal laser scanning microscopy to analyze, in more detail, the speed of Ca²⁺ spikes (Ca²⁺ waves) induced by large concentrations of ACh (10 μM) or IP₃ (100 μM). We chose fluo-3 as the Ca²⁺ indicator for these experiments, because, unlike BTC, it was not excited by the ultraviolet light used for the activation of caged IP₃ and therefore permitted visualization of Ca²⁺ spikes during photolysis. The Ca²⁺ spikes induced by 10 μM ACh traversed the acinar cells with the spike spread time of 0.9 ± 1 s (mean ± SD, n = 7) and the spike delay of 0.9 ± 0.9 s (Fig. 6A and Fig. B; Kasai et al. 1993), whereas those induced by 100 μM IP₃ exhibited the mean spread time of 0.1 ± 0.3 s (n = 4) and the delay of 0.1 ± 0.3 s (Fig. 6C and Fig. D). These results were consistent with those obtained by two-dimensional imaging with BTC (Fig. 5). Thus, spread of ACh-induced Ca²⁺ spikes were consistently slower than those induced by rapid photolysis of caged IP₃ at all concentrations of IP₃ and ACh examined.

We postulated that the slow spread of ACh-induced Ca²⁺ spikes is due to slow generation of IP₃ and to sequential activation of Ca²⁺-release channels with heterogeneous sensitivities for IP₃. To test this hypothesis, we reduced the rate of photolysis of caged IP₃ by decreasing the intensity of the actinic light source to 10% of its original value, so that the increase in [IP₃]_i occurred over a period of 1 s. As predicted from our hypothesis, the spike spread time of the resulting Ca²⁺ spikes was increased to 0.7 ± 0.3 s (n = 5; Fig. 6E and Fig. F). More importantly, the spike delay was also prolonged to 0.8 ± 0.3 s, similar to the spike delay for ACh-induced Ca²⁺ spikes (Fig. 6A and Fig. B). Thus, an artificial slow increase in [IP₃]_i was required to reproduce the time course of ACh-induced global Ca²⁺ spikes.

Discussion

We have demonstrated that spatially homogeneous increases in [IP₃]_i can induce Ca²⁺ spikes in acinar cells that share most features of those induced by ACh, consistent with the role of IP₃ as the Ca²⁺-mobilizing messenger for this neurotransmitter. Our data have also confirmed that subcellular gradients of IP₃ sensitivities are important for the generation of all forms of Ca²⁺ spikes in these cells, and that IP₃ is a long-range messenger and act as a global signal in those cells with diameters less than 20 μM (Allbritton et al. 1992; Kasai and Petersen 1994). Moreover, we have shown that the temporal profile of [IP₃]_i affects the kinetics of global Ca²⁺ spikes.

Control of Global Ca²⁺ Spikes by IP₃ Production

The time courses of global Ca²⁺ spikes induced by instantaneous increases in [IP₃]_i were faster than those of ACh-induced Ca²⁺ spikes at all concentrations of IP₃ and ACh examined. This observation indicates that ACh-induced activation of PLC results in a gradual increase in [IP₃]_i, and that the kinetics of [IP₃]_i is a key determinant of the time course of global Ca²⁺ spikes. Thus, we propose a mechanism for the generation of Ca²⁺ spikes in which the time course of their spread reflects that of [IP₃]_i, and in which their extent and amplitude are determined by the maximal [IP₃]_i (Fig. 7). The control of Ca²⁺ spikes by IP₃ production can explain simply the key properties of agonist-induced Ca²⁺ spikes in exocrine gland cells. First, the spread of Ca²⁺ spikes is relatively slow (5–15 μm/s; Kasai and Augustine 1990; Jaffe 1991; Toescu et al. 1992). Second, their extent and speed depend on agonist type and concentration (Fig. 5; Nathanson et al. 1992; Kasai et al. 1993; Thorn et al. 1993; Sjoedin et al. 1997; Pfeiffer et al. 1998). And finally, their amplitude varies over a large concentration range (0.5 to >10 μM) depending on the agonist concentration (Fig. 4). Thus, global Ca²⁺ spikes in acinar cells predominantly reflect global increases in [IP₃]_i, which are predicted to reach a maximum 1 to 8 s after the application of ACh (Fig. 5 B).

Figure 7.

Control of Ca²⁺ spikes by IP₃ production in pancreatic acinar cells. Increases in [IP₃]_i can trigger spread of Ca²⁺ spikes (Ca²⁺ waves) by sequential activation of IP₃ receptors, if there is a spatial gradient of IP₃ sensitivity across the cell and if the rate of IP₃ production is slower than that of the CICR-mediated Ca²⁺ spikes. For simplicity, if we assume that the CICR mechanism occurred instantaneously, the distribution of [Ca²⁺]_i is expressed as a function of the IP₃ sensitivity at a specific place, K _IP3(x), and of [IP₃]_i at a specific time, [IP₃]_i (t). [Ca²⁺]_i is determined by a balance between CICR-amplified Ca²⁺ release and Ca²⁺ removal mechanisms. When [IP₃]_i is small, local Ca²⁺ spikes result; when [IP₃]_i is large, Ca²⁺ spikes spread further, and generate global Ca²⁺ spikes in a dose-dependent manner. More rapid increases in [IP₃]_i result in a faster spread of Ca²⁺ spikes. Thus, kinetics of IP₃ production play a key role in the amplitude, extent, and time course of Ca²⁺ spikes. The same mechanism is operative in the generation of Ca²⁺ spikes in those cells with heterogenous IP₃ receptors, even though there is no spatial gradient in IP₃ receptors.

Our data also support role of CICR mechanisms of Ca²⁺ release channels in global Ca²⁺ spikes, because gradual increases in [Ca²⁺]_i were induced in response to rapid photolysis of caged IP₃ (Fig. 2, C–H). However, these increases in [Ca²⁺]_i were too fast (Fig. 5A, Fig. C, and Fig. E) to account for ACh-induced global Ca²⁺ spikes (Fig. 5B, Fig. D, and Fig. F). Thus, it is conceivable that the CICR mechanism locally generates Ca²⁺ spikes, and that the increases in [IP₃]_i control the spread of such Ca²⁺ spikes. Since gradual increases in [IP₃]_i determine the kinetics of global Ca²⁺ spikes, it is likely that the positive feedback effect of Ca²⁺ on PLC plays a role in the generation of global Ca²⁺ spikes and oscillation in acinar cells as suggested in other preparations (Meyer and Stryer 1988; Harootunian et al. 1991; Hirose et al. 1999). In contrast, local Ca²⁺ spikes appear to be mediated solely by CICR mechanisms, because they occurred at constant level of [IP₃]_i (Fig. 1; Wakui et al. 1989; Thorn et al. 1996).

Given that the production of IP₃ by PLC is not instantaneous in any cell type, the resulting time-dependent increase in [IP₃]_i may be crucial to Ca²⁺ spikes in general. Moreover, long-range control of Ca²⁺ spike spread (Fig. 7) can be applied to cells in which gradients of IP₃ sensitivity exist (Inagaki et al. 1991; Fay et al. 1995; Lefevre et al. 1995; Robb-Gaspers and Thomas 1995; Missiaen et al. 1996; Simpson et al. 1997; Callamaras et al. 1998; Yamamoto-Hino et al. 1998). Thus, mechanisms of Ca²⁺ spiking generally involve (a) PLC dependent long-range control (Fig. 7), (b) local CICR mechanisms (Berridge 1993; Petersen et al. 1994; Clapham 1995), and (c) heterogeneity in the Ca²⁺ release channels.

Types of IP₃ Receptors

The control of global Ca²⁺ spikes by [IP₃]_i in pancreatic acinar cells is consistent with the previous observation that agonists and IP₃ each mobilize Ca²⁺ in a dose-dependent manner (Muallem et al. 1989; Petersen et al. 1991a,Petersen et al. 1991b). Our data further demonstrate that such dose-dependent control involves heterogeneity in the Ca²⁺-release processes distributed in various subcellular regions and results in a wide range of [Ca²⁺]_i (0.1 to >10 μM). The graded nature of Ca²⁺ spikes (Fig. 4C and Fig. D) may reflect the balance between Ca²⁺ release and clearance in vivo (van de Put et al. 1994).

It has reported that all three types of IP₃ receptors were expressed in acinar cells (Lee et al. 1997). The presence of type-1 IP₃ receptors may account for the initiation of Ca²⁺ spikes and oscillations in the trigger zone (Hagar et al. 1998; Miyakawa et al. 1999). The preferential localization of type-3 IP₃ receptors in the trigger zone (Nathanson et al. 1994) is possibly responsible for the large increases in [Ca²⁺]_i in this region, given the small inhibitory effect of Ca²⁺ on these receptors (Hagar et al. 1998). It is therefore suggested that the type-3 IP₃ receptor plays a specific role in cellular processes such as exocytosis that require high [Ca²⁺]_i (Ito et al. 1997; Kasai and Takahashi 1999). The Ca²⁺ release in the trigger zone exhibited a similar sensitivity (Fig. 4 A) to the type-3 IP₃ receptors in vivo (Miyakawa et al. 1999). The reasons for 10 times lower IP₃ sensitivity in the basal area (Fig. 4 B) remain to be clarified.

Acinar cells may differ from oocytes and smooth muscle cells in that the latter cell types express predominantly one type of IP₃ receptor, and Ca²⁺ spikes in these cells occur in an all-or-nothing manner (Lechleiter and Clapham 1992; Parker and Ivorra 1993; Iino et al. 1993). Thus, the distributions of distinct IP₃ receptors appear critical for Ca²⁺-dependent cellular functions.