Long Distance Communication between Muscarinic Receptors and Ca 2 (cid:1) Release Channels Revealed by Carbachol Uncaging in Cell-attached Patch Pipette*

We have investigated the characteristics of cytosolic Ca 2 (cid:1) signals induced by muscarinic receptor activation of pancreatic acinar cells that reside within intact pancreatic tissue. We show that these cells exhibit global Ca 2 (cid:1) waves and local apical Ca 2 (cid:1) spikes. This is the first evidence for local Ca 2 (cid:1) signaling in undissociated pancreatic tissue. The mechanism of formation of localized Ca 2 (cid:1) signals was examined using a novel approach in-volving photolysis of caged carbachol inside a patch pipette attached to the basal surface of an acinar unit. This local activation of basal muscarinic receptors elicited local cytosolic Ca 2 (cid:1) spikes in the apical pole more than 15 (cid:2) m away from the site of stimulation. In some experiments, local basal receptor activation elicited a Ca 2 (cid:1) wave that started in the apical pole and then spread toward the base. Currently, there are two com-peting hypotheses for preferential apical Ca 2 (cid:1) signaling. One invokes the need for structural proximity of the cholinergic receptors and the Ca 2 (cid:1) release channels in the apical pole, whereas the other postulates long distance communication between basal receptors and the channels. Our intrapipette uncaging experiments pro-vide definitive evidence for long distance communication between basal muscarinic receptors and apical Ca 2 (cid:1) Intrapipette Uncaging— Standard patch clamp technique was adapted using the EPC-8 amplifier and Pulse software (HEKA). We used pipettes having resistances of 2–4 megohms. Pipettes were back- filled with the extracellular solution containing, in addition, either 5 mg/ml caged fluorescein dextran-10K or 50 (cid:2) M -5 m M caged CCh. Seals of 5–10 giga-ohm resistance were obtained between the pipette and the basal plasma membrane of individual acinar cells in small clusters. Uncaging was achieved during the scanning procedure by exposure of pre-defined regions to UV laser light at 364 and 351 nm. The regions were selected to surround the pipette but not expose the cell.

In response to agonist stimulation, many cell types produce transient elevations in the cytosolic Ca 2ϩ concentration ([Ca 2ϩ ] i ), which remain localized to a specific subcellular domain (1). In isolated pancreatic acinar cells, a low concentration of acetylcholine (ACh) 1 can evoke repetitive local cytosolic Ca 2ϩ spikes, which are confined to the granule containing apical pole (2). These local Ca 2ϩ spikes can activate exocytosis through the apical plasma membrane (3) and open Ca 2ϩ -dependent Cl Ϫ channels present exclusively in the apical plasma membrane, thereby regulating acinar fluid secretion (4).
Agonist-receptor interaction causes activation of phospholipase C and the generation of inositol 1,4,5-trisphosphate (IP 3 ), which in turn opens Ca 2ϩ release channels in the endoplasmic reticulum membrane (5), thereby explaining the liberation of Ca 2ϩ from this intracellular store (6). Because the G-protein coupled receptors for cholecystokinin are localized in the basolateral plasma membrane as demonstrated in early autoradiographic studies on intact pancreas (7), the classical view of pancreatic acinar cell stimulation involves the binding of agonist to plasma membrane receptors located at the base. It is this end of the cell that is close to blood vessels and pancreatic nerve terminals from where physiologically released agonists approach the cell (8). If agonist-receptor interaction occurs at the base and Ca 2ϩ release occurs at the opposite (apical) pole, there is a need for a long distance Ca 2ϩ -releasing intracellular messenger. The established Ca 2ϩ -releasing messenger IP 3 (9,10) as well as the more recently discovered Ca 2ϩ -liberating agents cyclic ADP-ribose and nicotinic acid adenine dinucleotide phosphate each elicit repetitive cytosolic Ca 2ϩ spikes in the apical pole (2,11,12). Therefore, it has been proposed that messenger(s) generated at the basal membrane in response to agonist stimulation easily diffuse across the cell and release Ca 2ϩ primarily in the apical pole because of the high density of Ca 2ϩ release channels in this part of the cell (2,(13)(14)(15)(16)(17).
The view that the primary site of the neurotransmitter-or hormone-elicited intracellular Ca 2ϩ release is determined not by the site of the neurotransmitter or hormone-receptor interaction but by the localization of the Ca 2ϩ release channels (13) has been challenged in a recent study (18), which used immunofluorescence to localize G-protein coupled receptors in isolated pancreatic acinar cell clusters. This study shows localization of receptors close to the apical end of the cell in regions near to the tight junctions. The authors suggest that apical Ca 2ϩ signals are preferentially produced in that region as a result of nearby production of intracellular messengers. The apparent contradiction between this study (18) and the previously held view of agonist-evoked local Ca 2ϩ signal generation (13) lead us to test the ability of pancreatic acinar cells to produce apical Ca 2ϩ release in response to basal stimulation. Because all previous Ca 2ϩ signaling studies in the exocrine pancreas were carried out on isolated acinar cells or cell clusters, we first demonstrated that in intact undissociated pancreatic tissue, ACh stimulation preferentially produces [Ca 2ϩ ] i elevations in the apical region of the acinar cells. Using intrapipette uncaging of carbachol (CCh), we then show that activation of muscarinic receptors in a very small restricted area of the basal plasma membrane is able to produce localized Ca 2ϩ responses at the opposite end of the cell in the apical pole.

EXPERIMENTAL PROCEDURES
Cell Preparation, Solutions, Dye Loading, and Chemicals-All of the experiments were performed at room temperature (22-24°C). The pancreas was removed from mice killed by cervical dislocation. The isolated pancreas was immediately placed into extracellular solution, which contained (in mM): 140 NaCl, 4.7 KCl, 1.13 MgCl 2 , 10 HEPES, 10 glucose, 1 mM CaCl 2 , pH 7.2 (adjusted by NaOH). Freshly isolated mouse pancreatic clusters of acinar cells were prepared as described previously (2)  Confocal Imaging-Fluorescence images were obtained using either a Zeiss LSM 510 confocal microscope or a Leica SP2 confocal microscope with ϫ40 oil-immersion or ϫ63 water-immersion objective. During the experiments, intact pancreatic tissue was immobilized on a glass coverslip using either a needle through the tissue or a slice hold-down (Warner Instruments). Intact tissue or cells adhered to poly-L-lysinecoated slides were imaged on the stage of an inverted microscope (Axiovert 100M, Zeiss, or DMIRBE from Leica). Fluo-4 AM, Mitotracker Green FM, and caged fluorescein dextran-10K were excited by a 488 nm laser line, and emission was collected through a bandpass filter of 505-550 nm. Fura Red was excited at 488 nm, and emitted light was collected through a long pass 560-nm filter or bandpass 580 -680-nm filter. Excitation of Lysotracker Red was by a 543-nm laser line, and emission was collected through a long pass 570-nm filter. Hoechst 33342 was excited by UV light (364 nm), and emission was collected between 400 and 450 nm. Intrapipette Uncaging-Standard patch clamp technique was adapted using the EPC-8 amplifier and Pulse software (HEKA). We used pipettes having resistances of 2-4 megohms. Pipettes were backfilled with the extracellular solution containing, in addition, either 5 mg/ml caged fluorescein dextran-10K or 50 M-5 mM caged CCh. Seals of 5-10 giga-ohm resistance were obtained between the pipette and the basal plasma membrane of individual acinar cells in small clusters. Uncaging was achieved during the scanning procedure by exposure of pre-defined regions to UV laser light at 364 and 351 nm. The regions were selected to surround the pipette but not expose the cell.

Subcellular Localized Calcium Signaling in Intact Epithelial
Tissue-Confocal fluorescence microscopy was used to image the Ca 2ϩ -sensitive dye, Fluo-4, in the cytoplasm of cells that were stimulated by bath perfusion of physiological agonists. We used an undissociated pancreatic preparation to image large areas of intact pancreatic tissue (see Fig. 1A) and showed that application of ACh triggered repetitive [Ca 2ϩ ] i elevations in nearly all of the cells (n ϭ 25 experiments). An analysis of the fluorescence changes in individual cells revealed the spatiotemporal profile of these Ca 2ϩ oscillations (Fig. 1A). In some cells, lower levels of stimulation caused [Ca 2ϩ ] i oscillations that were restricted entirely to the secretory granule region at the apical end of the cell (Fig. 1A, images 1-3 (n ϭ 9)). Increasing the ACh concentration changed the profile of each transient from local to global. At higher doses, the [Ca 2ϩ ] i transient became a wave, which was initiated at the apical pole and spread toward the basal plasma membrane (Fig. 1A, images 4 -7 (n ϭ 22)). As the intensity of stimulation increased, the oscillations became superimposed on an elevated base line (Fig. 1A). These observations demonstrate that pancreatic acinar cells in their native environment can produce apically localized [Ca 2ϩ ] i elevations and global Ca 2ϩ waves initiated in the apical region.
Specific patterns of Ca 2ϩ signaling are translated into the responses of cellular organelles such as exocytosis of secretory granules (19), changes of mitochondrial metabolism (20,21), and nuclear gene expression (22). Also, the positioning of intracellular organelles can have profound implications for patterns of [Ca 2ϩ ] i signaling (23,24). We assessed the intracellular position of the major intracellular organelles in live cells within intact pancreatic tissue. The secretory (zymogen) granules were labeled using Lysotracker Red (based upon the affinity of the dye for acidic compartments), nuclei were stained using Hoechst 33342, and mitochondria were stained using Mitotracker Green FM. The majority of acidic compartments (mostly representing zymogen granules) were found near the apical (luminal) membrane (Fig. 1B). The acinar lumen was often visible as an unstained region between the cells. The nucleus or nuclei were always found outside the zymogen granule region toward the base (Fig. 1B). In isolated pancreatic acinar cells, several studies have shown that mitochondria have distinct and specific intracellular locations. The majority of the mitochondria are positioned in a perigranular belt, which limits the spread of apically localized Ca 2ϩ signals (23). This positioning was also apparent in the intact tissue, because the strongest mitochondrial staining was in the perigranular region (Fig. 1B). In addition, Mitotracker Green staining also revealed the perinuclear and subplasmalemmal mitochondria that have been identified in isolated cells (Fig. 1B) (24). The intracellular positioning of mitochondria is thought to have important functional interactions with nearby Ca 2ϩ signaling pathways (24 -26). Our observations in the intact pancreas highlight the relevance of the polarity of organelles and Ca 2ϩ signaling in cells within their native environment.
Long Distance Signaling Demonstrated by Intrapipette Uncaging of Carbachol-Acinar cells within the intact pancreas can produce local apical [Ca 2ϩ ] i spikes when ACh is applied to the basal side of the epithelium. This highlights the need for understanding the mechanism by which the local apical Ca 2ϩ signals are generated. Here we have studied the patterns of Ca 2ϩ signaling induced by exclusive stimulation of basally located receptors.
To specifically excite basal receptors, we have combined electrophysiological and photolytic (uncaging) techniques. A patch pipette containing caged agonist in standard extracellular solution was used to isolate a small area of the basal membrane The pipette forms a giga-ohm resistance seal with the membrane in cell-attached configuration. The pipette solution is identical to that surrounding the cells with the exception that it also contains a caged compound such as caged fluorescein dextran-10K as shown here. Ultraviolet laser light targeted at the pipette (shown as region bounded by dashed orange line throughout) causes uncaging of the caged compound exclusively inside the pipette. Graph shows the time courses of fluorescence from free fluorescein and the calcium-sensitive dye Fura red, which was loaded into the cytoplasm of the cells, from the regions shown in the transmitted light image. Short lasting uncaging (points shown by orange arrows) causes rapid increases in fluorescein fluorescence without triggering any cellular Ca 2ϩ release. Finally, ACh triggers a marked reduction in the Fura Red fluorescence intensity, which corresponds to a [Ca 2ϩ ] i rise. The lower panel shows images of fluorescein (green) and Fura Red (red) fluorescence before and after uncaging.
of a single acinar cell within a cluster of cells (3-10 cells). We used isolated cell clusters because the patching of cells within the intact pancreas was impossible because of the capsule surrounding the tissue. During continuous superfusion of the cell clusters, a high resistance (5-10 giga-ohms) seal was formed between the basal cell membrane and the glass pipette. Such a configuration could then be used to photolyse caged carbachol by UV laser light exposure of pre-defined regions (shown as dashed orange lines throughout) surrounding the pipette but avoiding the cells.
To test this experimental configuration, we used caged fluorescein dextran-10K in the pipette and loaded the cytoplasm of the cells with the Ca 2ϩ -sensitive dye, Fura Red. This allowed us to characterize the intrapipette uncaging and test whether UV exposure itself would cause any [Ca 2ϩ ] i elevation (Fig. 2). As expected, each UV exposure caused an increase in the fluorescence of fluorescein inside the pipette as the compound was uncaged (Fig. 2, n ϭ 4). The increase in fluorescence peaked at the first frame obtained after uncaging and then declined, initially rapidly and then more slowly. This decline presumably reflects dilution of the product of uncaging in the solution in the rest of the pipette, which is much larger than the volume exposed to UV light. The uncaging itself did not cause any change in [Ca 2ϩ ] I , whereas subsequent application of ACh to the bath triggered a large rapidly evolving [Ca 2ϩ ] i transient (Fig. 2). These experiments showed that we can selectively expose a small region of basal plasma membrane to a photolytically activated compound. This technique is equally applicable to studies using site-specific stimulation of many other cell types including neurons, which produce polarized responses to stimulation. The rapid time course and range of UV exposure makes intrapipette uncaging faster, more sensitive, and better regulated than other local stimulation techniques (2).
We tested the ability of CCh to trigger changes in [Ca 2ϩ ] i measured by Fluo-4 in the cytoplasm of the cells when uncaged inside a cell-attached pipette. We found that intrapipette uncaging of carbachol at the basal pole was often able to trigger short lasting [Ca 2ϩ ] I elevations, which were restricted to the extreme apex of the cell (Fig. 3A, n ϭ 11 cells). The distance between the basal membrane under the pipette and the point of Ca 2ϩ release was always over 15 m and up to a distance of 25 m. Approximately 60% of the cells responded to uncaging by producing Ca 2ϩ signals, and each uncaging event usually only triggered a single Ca 2ϩ spike. Intriguingly, in the other experiments, it was not possible to trigger Ca 2ϩ release even with very strong and/or long lasting UV exposure. This inability of even strong uncagings to trigger Ca 2ϩ signals in 40% of the cells could be explained either by insufficient production of second messengers or by absence of receptors under the patch pipette because of receptor clustering in other regions of the basal membrane. The clustering of transmembrane receptors has been reported previously (27) and is proposed to underlie high sensitivity areas of membrane that are able to amplify low level extracellular stimuli.
We compared the size of the [Ca 2ϩ ] i rises triggered by intrapipette CCh uncaging to those caused by bath application of ACh (10 nM) (Fig. 3A). Short lasting uncaging events (UV exposure of 100 -500 ms) consistently generated smaller responses than ACh in the same cell (n ϭ 4). For these durations of uncaging, the [Ca 2ϩ ] i signals were localized to the apical part of the cells.
To be certain that only muscarinic receptors under the patch pipette would be responsible for generating the intracellular Ca 2ϩ release, we carried out experiments in which we made sure that even in the unlikely event that some of the uncaged CCh could escape the pipette, no activation of receptors outside the area of the plasma membrane covered by the patch pipette would occur. We blocked all muscarinic receptors outside the pipette by adding atropine (100 M) after the gigaseal formation. This atropine concentration always blocked [Ca 2ϩ ] i elevations triggered by ACh (1 M) applied to the bath (n ϭ 12). Under these conditions, short intrapipette uncaging was still able to trigger a localized [Ca 2ϩ ] i elevation at the extreme apical pole of the cell (Fig. 3B, n ϭ 6 cells). This confirms that the pipette does indeed form an impenetrable seal with the membrane (otherwise the extremely high atropine concentration in the bath would have prevented the uncaged CCh in the pipette from interacting with the muscarinic receptors under the patch pipette) and that only receptors under the pipette can be activated following uncaging.
Global Ca 2ϩ Waves Triggered by Basal Intrapipette CCh Uncaging-The time course and size of the [Ca 2ϩ ] i elevations described above correlate with a low level of ACh activation in studies on isolated cells (2). Therefore, we tested whether stronger activation of receptors in a small area of basal plasma membrane could generate larger global [Ca 2ϩ ] i rises. Intrapipette uncaging of longer duration (800 ms-1.5 s) or repetitive short uncagings were able to trigger global Ca 2ϩ waves (Fig. 4,  n ϭ 4). Such waves were always initiated in the apical region and propagated toward the basal plasma membrane, demonstrating that activation of a small area of basal plasma membrane is able to produce significant activation of intracellular Ca 2ϩ signaling mechanisms. In some of our intrapipette uncaging experiments (n ϭ 4), we could trigger a Ca 2ϩ wave that propagated from the stimulated cell to other cells in the same cluster. This suggests that the level of second messenger produced by intrapipette uncaging is sufficient to support the spreading of the [Ca 2ϩ ] I rise via gap junctions.
The long distance signaling demonstrated directly in this study must rely on the production of some diffusible intracellular messenger following receptor activation. IP 3 , cyclic ADPribose, and nicotinic acid adenine dinucleotide phosphate are all able to release Ca 2ϩ from intracellular stores in pancreatic acinar cells and are involved in agonist-evoked responses (11, 28 -30). Such small water soluble messengers are highly diffusible in the cytoplasm (13), and it seems likely that one or more of them transmits the Ca 2ϩ release signal from the activated basal receptor to the Ca 2ϩ channels in the apical part of the endoplasmic reticulum store (31).
In the majority of our experiments, it was possible to trigger large [Ca 2ϩ ] i transients despite the fact that only a small area of plasma membrane underwent stimulation. We estimate that the area of plasma membrane covered by the patch pipette is at most 2% of the total surface area (based upon the ⍀-shaped membrane inside the pipette tip and the total cell surface estimated from our confocal images). If the distribution of the muscarinic receptors were reasonably even, this would suggest that occupancy of a relatively small proportion of the receptors is sufficient to trigger not only local but also global and even intercellular [Ca 2ϩ ] i signals. The ACh-detecting system in the basal membrane must therefore be highly sensitive.