Mechanisms of subcellular cytosolic Ca2+ signaling evoked by stimulation of the vasopressin V1a receptor.

Receptor activation may result in distinct subcellular patterns of Ca2+ release. To define the subcellular distribution of Ca2+i signals induced by stimulation of the vasopressin V1a receptor, we expressed the cloned receptor in Xenopus oocytes. Oocytes were then loaded with fluo-3 and observed using confocal microscopy. Vasopressin induced a single concentric wave of increased Ca2+ that radiated inward from the plasma membrane. With submaximal stimulation, however, regions of the Ca2+ wave spontaneously reorganized into repetitive (oscillatory) waves. Focal stimulation of a small part of the plasma membrane resulted in a Ca2+ wave which began at the point of stimulation, radiated toward the center of the cell, then reorganized into multiple foci of repetitive, colliding waves and spirals of increased Ca2+i. The pattern of Ca2+ signaling induced by focal or global stimulation was not altered in Ca(2+)-free medium, although signals did not propagate as fast. Finally, subcellular Ca2+ signaling patterns induced by vasopressin were inhibited by caffeine, while neither vasopressin nor microinjection of inositol trisphosphate blocked caffeine-induced increases in cytosolic Ca2+. Thus, stimulation of the V1a receptor in this cell system induces a complex pattern of Ca2+ signaling which is influenced by (1) the magnitude of the stimulus, (2) the distribution of the surface receptors that are stimulated, and (3) mobilization of Ca2+ from the extracellular space as well as from two distinct endogenous Ca2+ pools. The manner in which a single type of receptor is activated may represent an important potential mechanism for subcellular Ca2+i signaling.

Receptor activation may result in distinct subcellular patterns of Ca2+ release. To define the subcellular distribution of Ca2+i signals induced by stimulation of the vasopressin VI, receptor, we expressed the cloned receptor in Xenopus oocytes. Oocytes were then loaded with fluo-3 and observed using confocal microscopy. Vasopressin induced a single concentric wave of increased Ca2+ that radiated inward from the plasma membrane. With submaximal stimulation, however, regions of the Ca2+ wave spontaneously reorganized into repetitive (oscillatory) waves. Focal stimulation of a small part of the plasma membrane resulted in a Ca2+ wave which began at the point of stimulation, radiated toward the center of the cell, then reorganized into multiple foci of repetitive, colliding waves and spirals of increased Ca2+i. The pattern of Ca2+ signaling induced by focal or global stimulation was not altered in Ca2+-free medium, although signals did not propagate as fast. Finally, subcellular Ca2+ signaling patterns induced by vasopressin were inhibited by caffeine, while neither vasopressin nor microinjection of inositol trisphosphate blocked caffeine-induced increases in cytosolic Ca2+. Thus, stimulation of the VI, receptor in this cell system induces a complex pattern of Caz+ signaling which is influenced by (1) the magnitude of the stimulus, (2) the distribution of the surface receptors that are stimulated, and (3) mobilization of Ca2+ from the extracellular space as well as from two distinct endogenous Ca2+ pools. The manner in which a single type of receptor is activated may represent an important potential mechanism for subcellular Ca2+i signaling.
The temporal pattern of a cytosolic Ca2+ (Ca2+;) signal is thought to be important for its effect as a second messenger (1)(2)(3)(4)(5), and this pattern is often influenced by the magnitude of agonist stimulation (1, 2, 5). The spatial pattern of a Ca2+i signal may also be important for cell regulation (6)(7)(8), and stimulation of different types of surface receptors on the same cell induces different subcellular patterns of Ca2+ signaling (9). Stimulation of the vasopressin V1 receptor leads to an inositol trisphosphate (IP3)'-mediated Ca2+i increase (3, 4, lo), but the subcellular pattern of Vl-induced increases in Ca2+i has only recently been explored (4, 11). The VI receptor is found in a number of tissues, including liver, brain, kidney, vascular smooth muscle, and platelets (IO), and the VI, subtype of the receptor has been cloned from rat liver (12). The subcellular distribution of vasopressin-induced Ca2+ signals has been difficult to investigate in individual hepatocytes because of the small size of these cells (20-25 pm in diameter) relative to the speed at which Ca2+; waves may traverse them (up to 100 pm/s) (11,13). To better define Ca2+; signals generated by stimulation of the V1, receptor, and to insure that the Ca2+i signals were specifically induced by that receptor, we used the cloned cDNA for the receptor (12) to express it in much larger cells, Xenopus oocytes (-1000 pm in diameter). Such cells are too thick for detailed subcellular resolution of fluorescence from Ca2+ dyes if epifluorescence imaging is used, so the oocytes were observed using confocal microscopy (9, 14). The specific aims of this work were (1) to characterize subcellular Ca2+ signaling patterns induced by stimulation of the V1. receptor and (2) to determine the mechanisms responsible for these signaling patterns.

MATERIALS AND METHODS
Arc-vasopressin, caffeine, HEPES, and EGTA were obtained from Sigma, fluo-3 and IP3 were obtained from Molecular Probes, and ionomycin was obtained from Calbiochem. Xenopus oocytes were harvested and incubated at 18 "C in a modified Barth's medium. After 24 h, the oocytes were injected with mRNA (15-20 ng) synthesized from cDNA for the VI. receptor previously cloned from rat liver (12).
After an additional 48-72 h, oocytes were microinjected with the Ca2+-sensitive fluorescent dye fluo-3 (15) (48 ng). Within 30-120 min, the cells were transferred to a chamber on the stage of a Zeiss Axiovert microscope, perifused a t room temperature with a HEPES-buffered solution (1.2 ml/min), and observed using a Bio-Rad MRC-600 confocal imaging system. For studies in Caz+-free medium, EGTA (1 mM) was added to the HEPES buffer and Caz+ was withheld. To focally stimulate oocytes, a WPI Pneumatic PicoPump was used t o deliver 125 nl of vasopressin (1 p M ) onto the surface of the oocyte over 1-2 s. To minimize the surface area of membrane exposed to vasopressin, the microinjection apparatus delivered the vasopressin in the direction opposite to the direction of flow in the chamber. Microinjections of IP3 or CaCl, were performed using this same apparatus, although injection volumes were 30-80 nl (approximate oocyte volume, 1.5 pl).  and 140 s after stimulation, respectively. Increasing fluorescence intensity corresponds to increasing cytosolic Ca", but Ca2+ concentrations are not quantified further, because fluo-3 cannot be ratio-imaged (15). Oocyte diameter in this focal plane is 1175 pm, and depth of focus is 20 pm. I , change in fluorescence over time a t locations I , 2, and 3 indicated in A. There is an abrupt, sustained increase in Ca2+ a t each location and the lag time between stimulation and response increases as a function of distance from the plasma membrane. resolution of 1.76 pm/pixel were obtained; these cells typically have a diameter of 1000-1500 pm. Images were recorded a t a rate of 1 s" using a Panasonic TQ3031F optical memory disc recorder and were subsequently analyzed using an Itex Series 151 image processor. In selected experiments, confocal line scanning microscopy (11,13) was performed instead of time-lapse confocal microscopy. In the line scanning mode of confocal microscopy, fluorescence measurements are restricted to a single line across the plane of focus. This approach induces negligible photobleaching of fluo-3 (13) and allows the subcellular distribution of Ca2+; signals over time to be displayed in a single image, provided that Ca2+ wave propagation occurs along the line which is scanned (11,13). Changes in fluorescence intensity were used to reflect changes in cytosolic Ca2+, but Ca" concentrations were not quantified further, because fluo-3 cannot be ratio-imaged (15). Although the distribution of fluo-3 within the cytosol appeared slightly inhomogenous in some oocytes, oocytes were studied at least 30 min after fluo-3 microinjection and redistribution of the dye was never seen during experiments. Results for all studies are expressed as mean ? S.E., and comparisons between groups were made using the two-sample t test. pm/s; p < 0.0001). In half of the 10 oocytes stimulated with 1 or 10 nM VP, regions of the Ca'+ wave spontaneously reorganized into rapid repetitive (oscillatory) waves (Fig. 2). Stimulation with 0.1 nM vasopressin ( n = 4) induced similar oscillatory waves, but which began circumferentially at or near the plasma membrane (Fig. 2). No response to vasopressin (1 p~) was observed in oocytes injected with fluo-3 but not mRNA (n = 5 ) . These observations indicate that the magnitude of a stimulus may affect the subcellular pattern of Ca2+ release. A similar phenomenon was previously reported in acetylcholine-stimulated Xenopus oocytes expressing the cloned muscarinic MS receptor (9). However, the current work extends those observations by demonstrating that the different Ca2+ signaling patterns seen with increasing concentrations of agonist represent a progression. Specifically, minimal concentrations of vasopressin (0.1 nM) induce repetitive sub-plasmalemma1 Ca2+ waves, while higher concentrations (1-10 nM) induce a single concentric wave that reorganizes into these repetitive waves, and maximal stimulation (1 p~) leads to a single, faster wave that exhibits no repetitive behavior ( Figs. 1 and 2). These Ca'+ signaling patterns appear to be the spatial correlate of the progression of temporal patterns that often occurs with increasing concentrations of agonist; low concentrations induce sustained Ca'+ oscillations, higher concentrations induce transient Ca2+ elevations that may become (dampened) oscillations as the original signal decays, and .maximal stimulation leads to a single sustained CaZ+ elevation, often of greater amplitude than is seen in oscillatory signals (2,3,16). Thus, these findings demonstrate a parallel between the spatial and temporal aspects of cytosolic Ca2+ signals, as has previously been suggested (16).
Perifusion of a small part of the plasma membrane with vasopressin (1 p~, n = 8 ) resulted in an intracellular Ca'+ wave which began nearest the point of stimulation, radiated across that region of the cell at 8.7 & 2.1 pm/s (no different than the Ca2+ wave speed induced by global stimulation with 1 pM vasopressin), then reorganized along the wavefront into multiple foci of smaller, repetitive, colliding waves and spirals of increased Ca2+i which persisted for over 10 min. The average speed of these repetitive waves was over 40% greater than that of the original wave (12.4 f 0.6 pm/s, p < 0.0005), and the frequency of these Ca'+ oscillations ranged from 6 to 10 min-'. This periodic behavior within the cytosol demonstrates that the intracellular milieu behaves as an excitable medium (14,19). In such systems, the radius of curvature of these waves is linearly related to wavefront velocity, and this relationship has been described in detail (14,19). The change in cytosolic Ca'+ over time, as measured at a single point within the oocyte, was also very different depending on which point was chosen. In fact, sustained oscillations, dampened oscillations, and sustained elevations of Ca" could all be observed simultaneously a t different locations within a single oocyte. These observations demonstrate that the distribution of surface receptors which are stimulated also may affect the subcellular pattern of Ca2+ release within a cell.
No new or additional Ca2+ signal was elicited by re-exposure of a small region of plasma membrane to vasopressin (oocytes were restimulated within 1-2 min; n = 3). In addition, oocytes in which a small region of membrane was stimulated, followed by stimulation of the entire membrane, responded to the latter by a single, concentric inward-directed Ca'+ wave that excluded the intracellular region responding to the initial stimulus ( n = 2). These findings demonstrate that a refractory period follows local release of Ca2+ stores, which also supports the previous observation that the interior of Xenopus oocytes behaves like a regenerative excitable medium (14). However, refractory periods of over 1 min, as observed here, are over an order of magnitude longer than predicted (14). Longer refractory periods appear to result from more intense stimulation of receptors and are associated with faster Ca'+ wave propagation, which may reflect greater depletion of local Ca2+ stores (20).
It has been postulated that complex subcellular patterns of Ca'+ signaling are mediated by linkage of receptors to G proteins (9, 14). T o determine whether activation of plasma membrane Ca'+ channels (which are thought to be receptoroperated for VI. receptors (21,22)) is also involved, we investigated signaling patterns in Ca2+-free medium (Fig. 3). The   Stimulation of the VI, receptor induces an increase in cytosolic Ca2+ which is initiated largely by Ca2+ release from IP3-sensitive stores (4, 12, 15, 16).
This results in propagation of single or repetitive Ca2+ waves across the cytosol, in both isolated rat hepatocytes expressing the native VI, receptor (4, 11, 13) and Xenopus oocytes expressing the cloned receptor. It is hypothesized that such Ca2+ waves propagate by an autocatalytic reaction-diffusion network (i.e. CICR) rather caffeine was added subsequently (arrow). Caffeine induces a gradual, sustained increase in Ca2+i, no different from the pattern seen in Ca2+-containing medium (see Fig. 4c). Fluorescence over time was determined by time-lapse confocal microscopy, and the fluorescence at a single intracellular location is shown. The oocyte was stimulated with vasopressin beginning at t = 0 s, and caffeine was added subsequently (arrow). Caffeine induces a gradual, sustained increase in Ca2+,, no different from the pattern seen in the absence of vasopressin (see Fig. 4c).  Fig. 4c). Tracing is representative of findings from three experiments in which caffeine was introduced 2-5 min after IP,. In contrast, reinjection of IP, within 2-6 min of the initial microinjection failed to elicit a detectable increase in fluorescence ( n = 3; data not shown). than by simple diffusion of the Ca2+ released from IP3-sensitive stores (3,16,18). The CICRpool is thought to be distinct from IP3-sensitive Ca2+ stores (3, 16), in part because Ca2+ oscillations can be induced independent of changes in IP3 (23, 24) but are dependent upon mobilization of CICR stores (25). We used caffeine to investigate the role of CICR in the generation of subcellular Ca2+ signals, since the open probability of CICR Ca2+ channels is increased in a concentrationdependent manner by caffeine (26, 27), while caffeine has little affinity for the IP3 receptor (28). Oocytes perifused with 20 mM caffeine displayed a gradual but sustained increase in cytosolic Ca2+ ( n = 7). The Ca2+ increase began at the plasma membrane and spread slowly inward (Fig. 4)). Caffeine also induced this subcellular Ca2+ pattern in oocytes in Ca2+-free medium (n = 6; Fig. 5), in oocytes pretreated with 1 WM vasopressin (n = 4; Fig. 6), and in oocytes in which a supramaximal concentration of IP3 (100 p~) had been microinjected ( n = 3; Fig. 7)). In contrast, in oocytes in which IP, had been microinjected, reinjection of IP3 failed to elicit a detectable increase in fluorescence ( n = 3; data not shown). Together, these findings suggest that oocytes contain endog- enous caffeine-sensitive Ca2+ stores that are not depleted by mobilization of the IP3-sensitive Ca2+ pool. An alternative explanation is that caffeine instead mobilizes the IP3-sensitive Ca2+ pool itself. Although this hypothesis cannot be excluded by the current work, it may be less likely because caffeine is thought not to interact with the IP3 receptor (27-30). In addition, both caffeine and ryanodine slow the speed of agonist-induced Ca2+ waves but neither agent abolishes the waves or alters their direction (31), which further supports the hypothesis that caffeine acts on the ryanodine-sensitive rather than the IP3-sensitive Ca2+ pool. Oocytes pretreated with caffeine displayed no subsequent Ca2+ increase in response to 1 nM vasopressin (n = 5 ) , and those oocytes stimulated instead with 1 PM vasopressin ( n = 8) displayed either no global Ca2+ increase ( n = 1 of 8 ) or a faint circumferential Ca2+ wave which propagated to the center of the cell ( n = 7 of 8). The fluorescence increase was quantified in those seven oocytes showing a global response to vasopressin by dividing maximal fluorescence by basal fluorescence at a specific point within each cell (32). This increase in fluorescence was onefifth the increase seen in pair-matched oocytes stimulated with vasopressin but not caffeine ( p < 0.01, Wilcoxon signed rank test). These findings suggest that vasopressin-induced Ca2+ signals depend on mobilization of Ca2+ from caffeinesensitive stores. This is consistent with the previous observation that caffeine inhibits the IP3-mediated increase in current across Ca2+-activated C1-channels (33). It was also previously observed that caffeine alone fails to increase current across Ca2+-activated C1-channels (33), so it was concluded that caffeine does not elevate Ca2+i. In the current work, however, Ca2+, signals in the oocyte were observed directly using confocal fluorescence microscopy. Since caffeine induces an increase in fluo-3 fluorescence even in Ca2+free medium, these observations suggest that caffeine indeed releases Ca2+ from endogenous stores, despite the fact that caffeine has been reported not to increase current across Ca2+activated C1-channels (33). Finally, microinjection of IP3 (10 or 100 pM) resulted in a single Ca2+ wave that radiated across the cell from the point of injection, even when IP3 was not injected near the plasma membrane (Fig. 8). In contrast, and in agreement with previous observations (19,33) Each measurement of fluorescence along that line is stacked horizontally on top of subsequent measurements. Fluorescence along the line was measured every 0.3 s, for a total of 2.6 min (top to bottom). Ionomycin induces a gradual increase in fluorescence which is limited to an annular region beneath the cell membrane. This is typical of the pattern seen in all ionomycin-stimulated oocytes. C, confocal image of the oocyte after -4 min of stimulation with ionomycin. Note the narrow annulus of increased fluorescence immediately beneath the cell membrane. The width of the annulus is -60 pm, less than 15% of the oocyte radius. D, ionomycin-induced change in fluorescence over time at two distinct points within the oocyte. A gradual, sustained increase in Ca2+i which begins after a brief latency period is seen just beneath the cell membrane (solid line), but no increase in Ca2+, is evident 113 pm further into the oocyte (dotted line). Ionomycin (1 p~) is added to the perifusate at t = 0 s and is present continuously thereafter.
phore ionomycin (1 p~) induced an increase in cytosolic Ca2+ that was limited to the subplasmalemma (n = 7; Fig. 9). Thus, a focal increase in cytosolic Ca2+, induced either by microinjection of CaC12 or by perifusion with ionomycin, failed to elicit a nondiminished Ca2+ wave which propagates across the oocyte. Together, these findings suggest that the subcellular pattern of vasopressin-induced Ca2+ signals depends on the sequential release of Ca2+, first from IPS-sensitive and then from caffeine-sensitive stores. Although caffeine-sensitive Ca2+ stores are thought to be responsible for CICR, the finding by us and others (19,33) that focal increases in cytosolic Ca2+ per se do not lead to CICR suggests that different or additional messengers are needed to trigger Ca2+ release from CICR stores (19, 33). Since cytosolic Ca2+ modulates Ca2+ release from IPS-sensitive stores (34, 35), it has been hypothesized that the source of CICR in Xenopus oocytes is the IPSsensitive rather than an lP3-insensitive Ca2+ pool (19). However, the observation that caffeine-sensitive Ca2+ stores can be mobilized even after maximal stimulation with vasopressin (1 pM) or IP3 (10 or 100 pM) suggests that caffeine-sensitive and IP3-sensitive Ca2+ stores are distinct.
Vasopressin-induced Ca2+ signals regulate a number of functions in the hepatocyte, ranging from glycogenolysis (36) to canalicular contraction (37, 38) to bile secretion (39). Intercellular coordination of Ca2+ signals occurs in hepatocytes (11,40) and other epithelia (41), and such coordination may be dependent in part upon the concentration of the stimulus (42). The subcellular distribution of Ca2+ signals within individual cells may also be important for the regula-tion of epithelial cell function (6), and this work suggests that stimulation of the hepatocyte VI. receptor may result in patterns of subcellular Ca2+i signaling which are more complex than previously realized. The current work also illustrates that the subcellular pattern of Ca2+ release induced by stimulation of the VI. receptor differs from the pattern induced by stimulation of either the muscarinic Mz or M3 receptor, which are the only other receptors for which such patterns have been reported (9). Specifically, the V,.-induced response always begins circumferentially along the cell membrane and radiates inward. In contrast, the Mz-induced response begins in the cell interior and is manifested by multiple foci of repetitive Ca2+ pulsations (9), while the M3-induced response at lower acetylcholine concentrations is similar to the M2 response, but at higher concentrations is a single Ca2+ wave that envelopes the oocyte (9). The precise relationship between such complex Ca2+ signals and Ca2+-mediated cell functions remains to be established, however.
In summary, stimulation of the vasopressin VI. receptor in this cell system induces a complex subcellular pattern of Ca2+i signaling which is influenced by the magnitude of the stimulus, the distribution of the surface receptors that are stimulated, and mobilization of endogenous as well as exogenous Ca2+. Furthermore, this Ca2+ signaling pattern appears to require sequential mobilization of endogenous Ca2+, first from IP3-sensitive and then from caffeine-sensitive Ca" stores. It has been suggested that the spatial organization of a Ca2+ signal is intimately related to its temporal organization (16); these findings support that hypothesis, since we observed that the temporal organization of a Ca2+ signal is dependent upon the intracellular location at which the signal is measured. Moreover, there appears to be a parallel relationship between the spatial and temporal components of Ca2+ signals. Modulation of the subcellular distribution of Ca2* may represent an important potential mechanism by which activation of a single type of receptor can regulate cell function.