Mechanisms of Receptor-mediated Ca 2 + Signaling in Rat Hepatocytes "

The Ca2+ signal observed in individual fura-2-loaded hepatocytes stimulated with the al-adrenergic agonist phenylephrine consisted of a variable latency period, a rapid biphasic increase in the cytosolic free Ca2+, followed by a period of maintained elevated cytosolic Ca2+ (plateau phase) that depended on the continued presence of both agonist and external Ca". Microinjection of guanosine-5'-0-(3-thiophosphate) elicited a Ca2+ transient with the same basic features. The Ca2+ transient resulting from microinjecting inositol 1,4,5trisphosphate (Ins-1,4,5-P3) occurred with essentially no latency period and consisted of a rapid spike that decayed back to preinjection levels within 15 s. Microinjection of inositol 1,4,5-trisphosphorothioate (thio-Ips), a nonmetabolizable analog of Ins-1,4,5-P3, elicited a Ca2+ transient that was initially identical to that observed with Ins-1,4,5-P3, except that the cytosolic Ca2+ remained elevated. The maintained thio-Ipsinduced Ca2+ increase was dependent on the presence of external Ca2+, suggesting an activation of Ca2+ influx. Reintroduction of external Ca2+ in the presence of 5 p~ phenylephrine to Ca2+-depleted cells resulted in a 2-fold greater rate of rise in the cytosolic Ca2+ compared to the rate observed upon Ca2+ addition to cells Ca2+-depleted by preatement with thapsigargin. The rate of Ca2+ rise upon Ca2+ addition to cells microinjected with thio-IP3 was similar to that observed with phenylephrine. Coinjection of the cells with thioIPS plus heparin reduced the rate of Ca2+ rise upon Ca2+ addition to that observed in thapsigargin-treated cells. These data indicate that the mechanism responsible for receptor-mediated stimulation of Ca2+ entry into hepatocytes involves not only capacitative Ca2+ entry but also an additional component mediated directly by Ins1 ,4,5-P3.


Mechanisms of Receptor-mediated Ca2+ Signaling in Rat Hepatocytes"
(Received for publication, March 4, 1991) Carl A. HansenS, Lijun Yang, and John R. Williamson$ From the DeDartment of Biochemistry and Biophysics, University of Pennsylvania School of Medicine, Philadelphia; Pennsyluania 19104 " " The Ca2+ signal observed in individual fura-2-loaded hepatocytes stimulated with the al-adrenergic agonist phenylephrine consisted of a variable latency period, a rapid biphasic increase in the cytosolic free Ca2+, followed by a period of maintained elevated cytosolic Ca2+ (plateau phase) that depended on the continued presence of both agonist and external Ca". Microinjection of guanosine-5'-0-(3-thiophosphate) elicited a Ca2+ transient with the same basic features. The Ca2+ transient resulting from microinjecting inositol 1,4,4, occurred with essentially no latency period and consisted of a rapid spike that decayed back to preinjection levels within 15 s. Microinjection of inositol 1,4,5-trisphosphorothioate (thio-Ips), a nonmetabolizable analog of Ins-1,4,5-P3, elicited a Ca2+ transient that was initially identical to that observed with Ins-1,4,5-P3, except that the cytosolic Ca2+ remained elevated. The maintained thio-Ipsinduced Ca2+ increase was dependent on the presence of external Ca2+, suggesting an activation of Ca2+ influx. Reintroduction of external Ca2+ in the presence of 5 p~ phenylephrine to Ca2+-depleted cells resulted in a 2-fold greater rate of rise in the cytosolic Ca2+ compared to the rate observed upon Ca2+ addition to cells Ca2+-depleted by preatement with thapsigargin. The rate of Ca2+ rise upon Ca2+ addition to cells microinjected with thio-IP3 was similar to that observed with phenylephrine. Coinjection of the cells with thio-IPS plus heparin reduced the rate of Ca2+ rise upon Ca2+ addition to that observed in thapsigargin-treated cells. These data indicate that the mechanism responsible for receptor-mediated stimulation of Ca2+ entry into hepatocytes involves not only capacitative Ca2+ entry but also an additional component mediated directly by Ins-1 ,4,5-P3.
The concentration of cytosolic free Ca2+ regulates many intracellular events. These include the control of metabolism, secretion, and muscle contraction, as well as other functions mediated by hormones, neurotransmitters, and growth factors. Hormone-mediated Ca2+ signaling is initiated by an agonist-induced activation of phospholipase C, which hydrolyzes phosphatidylinositol 4,5-bisphosphate into two intracel-*This study was supported by an American Heart Association Grant-in-aid and National Institutes of Health Grant DK 15120. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertlsement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
To whom requests for reprints should be addressed. lular messengers, Ins-1,4,5-P3' and diacylglycerol (1,2). The ensuing rise in cytosolic free Ca2+ results from an Ins-1,4,5-P3-mediated release of Ca2+ from internal stores and from an activation of Ca2+ influx (2,3). The mechanism of agoniststimulated extracellular Ca2+ entry has not yet been resolved, but in many cell types it is distinct from voltage-operated Ca2+ channels (4). Current experimental data suggests three possible mechanisms: 1) receptor-activated Ca2+ channels, in which receptor activation and channel opening are intimately interconnected (5,6), perhaps via a G protein (5,7), 2) second messenger operated Ca2+ channels, which are gated either by 4,3,4,12), or Ca2+ itself (13), and 3) capacitative calcium entry, in which activation of Ca2+ entry is linked to emptying of the internal Ins-1,4,5-P3-sensitive Ca2+ store (14,15). It is likely that the various mechanisms operate to different extents in different cell types. In this study, we have characterized the Ca" transient elicited by phenylephrine in hepatocytes at the single cell level and subsequently attempted to determine the essential intracellular signals required to generate the various aspects of the Ca2+ transient. This was accomplished by activating the signal transduction process distal to receptor activation by microinjecting specific compounds that are currently thought to mediate aspects of this process.

EXPERIMENTAL PROCEDURES
Hepatocyte Preparation-Hepatocytes were isolated from livers of fed, male Sprague-Dawley rats weighing 180-220 g by collagenase digestion as described previously (16). Cells were resuspended at 2 X lo5 cells/ml in Leibowitz-15 media supplemented with 20 m M Hepes and 5.5 mM glucose. Cells were attached to poiy-D-Iysine coated coverslips by incubating 3 ml of cell suspension for 1 h at 37 "C, at which point the media was changed and unattached cells removed. Cells were used from 1.5 to 6 h post-plating.
Measurement of Cytosolic Free Ca""Ce1ls attached to coverslips were transferred to a Leiden cell chamber and incubated in 3 ml of modified Hank's buffer consisting of 20 mM Hepes, 137 m M NaC1, 5.4 m M KC1, 4.2 mM NaHCO:<, 0.44 m M KH,PO,, 0.33 mM NaHP04, 1.3 mM CaC12, 1.0 mM MgCI2, and 5.5 mM glucose at pH 7.4, unless otherwise noted in the text. The chamber was maintained a t 30 "C on the microscope stage with a Medical Systems TC-102 temperature controller. Approximately 20 cells/coverslip were loaded with fura-2 free acid via pressure microinjection. A fine tip micropipette (R, = 50-60 MO) was filled with a solution of 10 m M fura-2 free acid in 125 mM KC1 and impaled into the hepatocyte. The content of the micropipette was introduced into the cell by applying a 20 psi pressure pulse for 200 ms using a General Valve Picospritzer. This injection protocol delivered approximately 50-150 femtoliters, determined using radioisotope, which represents approximately 1-3% of the total cell volume.
The fluorescence of a dye-loaded cell was imaged using a 1 0 0~ Nikon Fluor objective and a Nikon Diaphot epifluorescence microscope illuminated with alternating (100 Hz) 340 and 380 nm (5-nm half-band width) excitation light. Fluorescence emission was collected at 520 nm (50-nm half-band width) by a photomultiplier tube. Output from the fluorometer was digitized, ratioed, and analyzed by an IBM AT clone based analysis system (Indec Systems). Calibration of the fluorescence signal followed the method of Grynkiewicz et al. (17) and was performed in situ by addition of 10 ~L M of ionomycin (R,nax), followed by addition of 5 mM EGTA, pH 8.0 (Rmin). Values obtained were 15, 0.4, and 8.6 for R,.,, Rmi,, and @, respectively. A Kd of 224 nM was used (17).
In certain experiments, the fura-2-loaded cells were reimpaled with another fine tip micropipette containing either Ins-1,4,5-P3, thio-IP3', Ins-1,3,4,5-P4, or thio-IP, in 125 mM KC1 and the contents introduced into the cell as discussed above. Agonists were delivered to the cell surface via a perfusion pipette (10 pm inner diameter) located approximately 30 pm from the cell by applying a 1-psi pressure pulse for various durations.
Materials-Ins-1,4,5-Pa and ionomycin were obtained from Calbiochem. Thio-IP:, and thio-IP, were from Du Pont-New England Nuclear. Fura-2 was from Molecular Probes. Ins-1,3,4,5-P4 was synthesized as described by Cerdan et ul.(18). Collagenase was obtained from Worthington Chemicals, Leibowitz L-15 culture media and phenylephrine were from Sigma, and GTP-yS was from Boehringer Mannhein. All other chemicals were reagent grade or better. Thin wall borosilicate glass with filament (1.0 mm outer diameter X 0.75 mm inner diameter) was obtained from A-M Systems.

RESULTS
The Hormone-induced Ca2+ Transient- Fig. 1 illustrates characteristic responses of individual fura-2-loaded hepatocytes exposed to increasing concentrations of phenylephrine to activate the al-adrenergic receptors in the plasma membrane. A threshold concentration of hormone had to be exceeded in order to elicit any change of the cytosolic free Ca'+, which varied between individual cells (compare cell A uersus B, Fig. 1). In the cell of Fig. L4, 0.2 p~ phenylephrine produced no effect, while 0.4 p~, after a long delay, produced a submaximal increase of peak Ca'+ followed by a brief plateau phase. As the concentration of phenylephrine applied to the cell was raised further, the Ca2+ peak height continued to increase, the latency period before the initial rapid Ca2+ rise continued to shorten, but the level of the sustained Ca2+ increase during the plateau phase remained relatively unchanged. In other cells, exemplified by the cell in Fig. lB, the peak height of the Ca2+ spike became maximal at a lower phenylephrine concentration than that required to produce the shortest latency period. The latency period is thought to represent the time required to generate sufficient Ins-1,4,5-P3 to bind to and activate the Ins-1,4,5-P, receptor (2,19). The pattern obtained from cell B suggests that the slowest steps for agonist-induced production of Ins-1,4,5-P3 may occur prior to phospholipase C activation, at the level of the receptor-(= protein interaction. At all phenylephrine concentrations, the cytosolic free Ca2+ returned to the prestimulated resting level of about 125 nM Ca2+ within 30 s after terminating agonist delivery from the perfusion pipette. This type of hormone response is similar to that observed with angiotensin I1 stimulation of single N1E-115 neuroblastoma cells (20) and many other cell types after agonist stimulation. A small portion of the hepatocytes responded upon phenylephrine stimulation with repetitive spikes in the cytosolic free Ca2+ concentration, as has been observed by others (21-23).
Internal Ca2+ mobilization appears to predominate initially over the phase of Ca2+ entry, as illustrated by the data in Fig.  2, which show results obtained by varying the exposure time of the cell to 2.5 p~ phenylephrine from 2 to 60 s. The peak height of the Ca2+ transient was not appreciably affected by the exposure time, but the secondary Ca2+ plateau phase could not be readily distinguished unless the hormone was delivered for a time longer than 15 s. In addition, the initial phase of the Ca'+ transient was not affected by a decrease in the extracellular Ca'+ concentration to about 1 p M just prior to hormone addition, while the plateau phase was abolished by removing extracellular Ca2+ (data not shown). This reflects an increased rate of Ca'+ entry into the cells consequent to receptor stimulation. In hepatocytes, the plateau phase of the Ca2+ transient is not affected by voltage-dependent Ca2+ channel antagonists or low concentrations of inhibitory cations (19). These data suggest, therefore, that while the two processes of intracellular Ca2+ mobilization and Ca2+ entry are receptor-mediated, they have different kinetic, temporal, and sensitivity characteristics.
Post activate phospholipase C via a GTP-binding protein (1,2). The non-hydrolyzable analog of GTP, GTPyS, has been used extensively to investigate the presence and role of G proteins in receptor coupling since its binding to the GTP site of the a-subunit results in a long term activation of the G-protein (24). Fig. 3 shows the results from three separate cells in which GTPyS was rapidly introduced into a fura-2-loaded hepatocyte through pressure microinjection. The resulting Ca2+ transients were associated with a variable latency period, an abrupt increase of Ca2+ to a peak value, followed by a sustained period of elevated Ca2+. In one of the cells illustrated, the intracellular free Ca2+ exhibited a damped oscillatory behavior. These data illustrate that the basic features of the hormone-induced Ca2+ transient (latency of the response, intracellular Ca'+ release, and stimulated Ca'+ entry) were reproduced by a direct activation of a G protein(s) and represent activation of post-receptor events.
In contrast to the sustained increase in the cytosolic free Ca'+ induced by microinjection of GTPyS, Fig. 4

(trace A )
shows that microinjection of Ins-1,4,5-P3 caused a rapid increase of Ca2+, which then decayed back to the resting Ca'+ level within 10-15 s. The height of the peak Ca'+ reached by pressure microinjection of a maximal concentration of Ins-1,4,5-PC1 was similar to the peak height of the initial Ca'+ spike observed with a maximal amount of phenylephrine. The increase in the cytosolic free Ca2+ occurred in less than 400 ms, the minimum time resolution of the instrumentation. It has been shown in stop-flow experiments using permeablized hepatocytes that release of internal Ca'+ stores was initiated within 20 ms of presenting Ins-1,4,5-Py to its receptor (25). The almost immediate increase of Ca2+ observed with Ins-1,4,5-P3 injections showed that the latency period observed with hormone addition or GTPyS microinjection was caused by a combination of the time required for activation of the Gprotein and for the GTP-bound a-subunit to activate one of the isoforms of phospholipase C. These data support the argument that the phospholipase C activity has to be increased to a point where the rate of production of Ins-1,4,5-P:, exceeds its rate of metabolism, thereby allowing a critical concentration to be produced for binding to the Ins-1,4,5-Ps receptor with opening of Ca'+ channels in intracellular vesicles (19,26).
Mechanism of Hormone-stimulated Ca2+ Entry-During hormone-stimulated Ca'+ signaling, Ins-1,4,5-P, is being continuously produced and, in hepatocytes with hormones such as vasopressin and phenylephrine, its concentration is maintained elevated for as long as the receptor remains occupied (29). As shown in Fig. 5, increased levels of Ins-1,4,5-P3 were maximal within 5 s after addition of 10 PM phenylephrine to a suspension of ["Hlmyo-inositol-labeled hepatocytes. Following the initial rise, however, Ins-1,4,5-P3 levels decreased to a new steady-state level within 30 s that was 30% of the peak level. It is during this period of an agonist-generated increase Downloaded from in the steady-state level of cellular Ins-1,4,5-P, that the plateau phase of the Ca2+ transient occurs. The question arises whether the continued presence of elevated Ins-1,4,5-P3 levels in hepatocytes plays a role other than maintaining a depletion of the Ins-1,4,5-P3-sensitive Ca2+ stores. Metabolism of Ins-1,4,5-P3 can be circumvented by using the non-hydrolyzable analog of Ins-1,4,5-P3, thio-IPS. In hepatocytes, this compound is only slightly less potent than Ins-1,4,5-P3 as a Ca2+ mobilizing agent, and it is not a substrate for either the 5phosphomonoesterase or the Ins-1,4,5-P3 3-kinase (30). Fig.  4, truce B, shows results obtained by microinjecting thio-IP, into hepatocytes incubated in 1.3 mM Ca2+-containing medium. Introduction of thio-IP, caused a similar rapid increase in the cytosolic free Ca2+ as observed with Ins-1,4,5-P3, reaching peak levels equivalent to maximal hormone stimulation. Unlike addition of Ins-1,4,5-P3, however, the intracellular Ca2+ level remained elevated after addition of thio-IP,. Fig. 6 shows that upon removal of extracellular Ca2+ after microinjection of thio-IP, into a cell bathed in normal Ca2+ medium, the thio-IP3-induced rise of the cytosolic free Ca2+ concentration returned toward resting values. Reintroduction of 1.3 mM external Ca2+ by turning off the perfusion pipette containing 3 mM EGTA in Ca2+-free modified Hank's media resulted in a subsequent increase of the cytosolic Ca2+ concentration as Ca2+ entered the cell. These data suggest, therefore, that thio-IPB was stimulating Ca2+ entry into the cell.
The data in Fig. 7 provide additional support for a thio-IP3-stimulated Ca2+ influx. Representative results from two separate hepatocytes are shown in which the cells were first exposed for 10 min to nominally Ca2+-free medium (approximately 1 PM) and then exposed to 25 PM phenylephrine for 90 s in order to deplete the Ins-1,4,5-P3-sensitive Ca2+ stores. Separate experiments showed that a subsequent stimulation of these Ca2+-depleted cells with phenylephrine did not elicit an increase in the cytosolic free Ca2+ concentration. After the cytosolic free Ca2+ had returned to resting levels, the cell was microinjected with thio-IP3. This caused an immediate small increase of the cytosolic free Ca2+, which presumably reflects the residual amount of Ca2+ in the Ins-1,4,5-P3-sensitive Ca2+ stores in this experiment. Fifteen s after the injection of thio-IP,, the cell was exposed to 1.3 mM Ca2+. The cytosolic free Ca2+ increased rapidly and leveled off at about 600 nM (cf. Fig. 6). In contrast, microinjection of thio-IP, produced no effect by itself on the cytosolic free Ca2+, and there was no increase in the cytosolic free Ca2+ upon readdition of Ca2+ to  FIG. 8. Enhancement of the rate of increase in the cytosolic free Ca2+ upon readdition of external Ca2+ to Ca2+-depleted hepatocytes by phenylephrine. Cells were depleted of internal Ca'+ either by the procedure described in Fig. 7 (circles, squares) or by incubation with 500 nM thapsigargin in nominally Ca'+-free media for 10 min (triangles). External Ca2+ was then reintroduced by perfusing the cells with modified Hank's buffer containing either 1. the medium. The result of the thio-IP, study on Ca2+ entry was similar to that obtained in the controls, which consisted of microinjecting 125 mM KC1 followed by Ca2+ addition to the medium (data not shown). Hence, the stimulated Ca2+ entry observed with thio-IP, appears to be specific to this Ins-1,4,5-P3 analog.
In contrast to the lack of enhanced Ca2+ entry in Ca2+depleted cells, readdition of Ca2+ in the presence of 5 PM phenylephrine produced a stimulation of Ca2+ influx. Fig. 8 shows mean results of a series of experiments in which intracellular Ca2+ stores were first depleted by perfusion of cells with phenylephrine or thapsigargin under Ca2+-free conditions followed by readdition of Ca2+ at the arrow ( t = 15 s). When Ca2+ was reintroduced in the presence of 5 p M phenylephrine (Fig. 8, squures), the rate at which cytosolic free Ca2+ increased was stimulated more than 6-fold compared to control cells in which Ca2+ was added in the absence of phenylephrine (Fig. 8, circles). These data indicate that events associated with receptor activation were necessary for stimulated Ca2+ entry. One hypothesis for receptor-stimulated Ca2+ entry, the capacitative model, proposes that Ca2+ entry is enhanced when the Ins-1,4,5-P3-sensitive Ca2+ store is emptied. Since the thio-IPs injection studies demonstrated that there was a small amount of residual Ca2+ in the Ins-1,4,5-P3sensitive Ca2+ stores in some experiments, it was possible that this Ca2+ was adequate to diminish activation of Ca2+ entry. T o deplete the residual Ca2+ in the Ins-1,4,5-P3-sensitive Ca2+ store, the cells were placed in nominally Ca2+-free medium containing 500 nM thapsigargin. Thapsigargin is an inhibitor of the Ca2+-ATPase that pumps Ca2+ into the Ins-1,4,5-P3sensitive Ca2+ stores and causes them to gradually lose their stored Ca2+ (31). Under these conditions the cytosolic Ca2+ initially increased as Ca2+ was released from the Ca2+ stores and then fell to below the normal resting value with establishement of a new basal level within 7 min (data not shown). Addition of phenylephrine after thapsigargin to cells bathed in Ca2+-free medium failed to elicit an increase in cytosolic free Ca2+, indicating that the thapsigargin treatment completely emptied the agonist-sensitive Ca2+ store. Upon readdition of external Ca2+ to thapsigargin-treated cells, (Fig. 8,  triangles), the rate at which the cytosolic Ca2+ increased was greater than that observed with control cells (Fig. 8, circles), but less than that observed after Ca2+ addition in the presence of phenylephrine.
The statistical significance of the changes in cytosolic free Ca2+ are more clearly seen from the data in Table I, which shows the mean rate of change of the cytosolic Ca2+ over 10s intervals following Ca2+ addition to Ca2+-depleted cells exposed to different conditions. In the thapsigargin-treated cells (column 11) the rate of increase of cytosolic free Ca2+ was maximal about 60 s after CaZ+ addition, while with phenephrine treatment (column 111) the rate was maximal after about 40 s, and thereafter declined to a rate comparable to that observed after thapsigargin treatment. Thus, the primary effect of phenylephrine relative to thapsigargin was to enhance the rate of Ca2+ entry over the time interval from 25 to 55 s after Ca2+ addition. The tailing off of the phenylephrine-induced rate of Ca2+ entry may indicate negative feed-back effects resulting from increases of the cytosolic free Ca'+. Nevertheless, these data indicate that a portion of the receptor-stimulated Ca2+ entry in hepatocytes resulted from a capacitative-like mechanism, but they also suggest that there was an additional factor, produced upon receptor occupancy, which further stimulated the rate of Ca2+ entry.
A likely candidate for this factor is the Ins-1,4,5-P3 generated upon receptor activation. As shown in Fig. 9 (circles), addition of external Ca2+ to cells injected with thio-IP, resulted in a high rate of Ca2+ entry, similar to that observed with phenylephrine. Table I, column IV, shows that the rate of rise of the cytosolic Ca2+ in thio-IP3-treated cells was significantly increased relative to all other conditions by 5 s after Caz+ addition and reached a maximal rate after 15 s. If the only mechanism underlying activation of Ca2+ entry was capacitative Ca2+ entry, there should have been no difference between the kinetics of thio-IP3-injected and thapsigarginpretreated cells, since the thapsigargin pretreatment should have fully primed the capacitative Ca'+ entry mechanism. The greater maximum rate of increase of the cytosolic Ca2+ and the shorter period required to obtain the maximum rate indicated that thio-IP, had preactivated the Ca'+ entry mechanism to a greater extent than thapsigargin pretreatment. The only difference between thio-IP:3-injected and phenylephrine-treated cells was that an additional 10 s was required to obtain the maximum rate of Ca2+ increase with phenylephine. This is not unexpected since it requires approximately 10 s following phenylephrine stimulation for peak In~-1,4,5-P:~ accumulation (see Fig. 5).
Heparin has been shown to be a competitive inhibitor of Ins-1,4,5-P3 binding to its receptor (32). When heparin was coinjected with thio-IP:3 into hepatocytes incubated in the presence of extracellular Ca2+, a substantial diminution in the initial rapid thio-IP3-induced Ca'+ transient was observed, indicating that under these conditions, heparin substantially blocked internal mobilization of Ca2+ (data not shown). Reintroduction of external Ca2+ to Ca'+-depleted hepatocytes injected with the same proportion of thio-IP, to heparin decreased the rate of increase of the cytosolic free Ca2+ to approximately half that observed with thio-IP3 alone (Fig. 9). Interestingly, the maximum rate of Ca2+ increase in the presence of heparin was the same as that observed after thapsigargin pretreatment of the cells (Table I, columns I1 and V). Assuming that the thio-IP3 is acting identically to hormone-generated Ins-1,4,5-P3, these data clearly suggest that Ins-1,4,5-P3 has a direct role in the stimulation of Ca'+ entry into hepatocytes. The fact that heparin coinjection with

Effect ofphenylephrine and thio-IP, on Ca'+ entry into Ca"-depleted hepatocytes
Cells were Ca*+-depleted as described in the legends to Figs. 7 and 8. The rate of increase in cytosolic Ca'+ over 10 s intervals after addition of extracellular Ca2+ was calculated for each cell. The number of hepatocytes in each group were 17, 27,8, 12, and 8 for columns I-V, respectively. FIG. 9. Inhibition by heparin of the thio-IP3 -enhanced rate of increase in the cytosolic free Ca2+ following readdition of external Ca2+ to Ca2+-depleted hepatocytes. Cells were depleted of internal Ca" by the procedure described in Fig. 7. Subsequently the cells were microinjected with either thio-IP3 (circles) or thio-IP3 plus heparin (squares). External Ca2+ was then reintroduced by perfusing the cells with modified Hank's buffer containing either 1.3 mM Ca2+ at t = 15 s (arrow) for 75 s from a perfusion pipette. The average rate of increase in the cytosolic free Ca2+ was, for circles, 254 f 50 nmol/min, n = 12, and for sqmres, 138 & 43 nmol/min, n = 8. The micropipette contained either 100 p M thio-IPs in 125 mM KC1 (circles) or 100 pM thio-IPs plus 10 mg/ml heparin in 125 mM KC1 (squares).
t.hio-IP3 did not completely inhibit Ca2+ entry is consistent with part of the total Ca'+ influx being through a capacitative mechanism, but may also reflect an incomplete inhibition of the putative Ins-1,4,5-P3-gated mechanism by heparin.

DISCUSSION
The data presented in this paper demonstrate that the basic features of the Ca2+ transient elicited by al-adrenergic stimulation in hepatocytes can be related to the generation of Ins-1,4,5-P3. The latency period, which is the aspect of the Ca2+ transient most affected by changes in agonist concentration, results from the time required for ligand-bound receptor to activate its G-protein, which in turn activates a subtype of phospholipase C to the point where the amount of Ins-1,4,5-P3 rises to a level that can discharge the Ca2+ from internal pools. This discharge of the internal Ca'+ stores appears as a rapid spike of increased cytosolic free Ca2+ that immediately decays to prestimulation levels if the agonist is present for less than 15 s. Since direct microinjection of Ins-1,4,5-P3 produces a similar rapid increase in Ca2+ which then returns to basal levels, the rise of cytosolic Ca'+ during the initial activation phase would appear to simply follow the hormoneinduced production of Ins-1,4,5-P3. If the presence of agonist is maintained, the initial Ca2+ increase decays to a new plateau level that is significantly higher than the initial resting level of Ca'+. This plateau phase is relatively independent of agonist concentration but is dependent on the presence of extracellular Ca'+ and returns to the resting level when agonist is removed. Injection of either GTP-yS, which presumably sustains an enhanced formation of Ins-1,4,5-P3 by maintaining the G-protein/phospholipase C interaction in an activated state, or thio-IP3, also produce a plateau phase. These data indicate a role for Ins-1,4,5-P3 in maintaining elevated levels of cytosolic free Ca'+, the important point being that activation of the hormone-sensitive Ca2+ entry mechanism via the al-adrenergic receptor requires the continued formation of Ins-1,4,5-P3. Ins-1,4,5-P3 enhances Ca'+ entry not only as a consequence of it mobilizing Ca2+ from internal stores, thereby triggering a capacitative type mechanism, but also by an apparent specific activation of a Ca2+ channel in the plasma membrane. This conclusion is based on the assumption that thio-IP3 mimics the action of Ins-1,4,5-P3. One mechanism for the action of Ins-1,4,5-P3 on Ca2+ entry would be through an Ins-1,4,5-P3-gated Ca2+ channel located in the plasma membrane. There have been several reports that suggest the presence of these channels in several cell types. These include electrophysiological data obtained in patch-clamped T cells (8) and the observation that Ins-1,4,5-P3 can induce Ca2+ release from purified plasma membrane vesicles from liver (33) and platelets (34,35). These data were further supported by the observation that antipeptide antibodies to amino acid sequences of the cerebellum Ins-1,4,5-P3 receptor reacted with protein in the plasma membranes of Purkinje cells (36). However, it now appears that the basis for the immunoreactivity results from a close apposition of the internal Ins-1,4,5-P3 receptors to the plasma membrane (37-39). Similarly, purified rat liver plasma membranes positively react to polyclonal antibodies raised against the COOH terminus sequence of the Ins-1,4,5-P3 receptor: but again, it is not yet clear that this does not represent contamination of the plasma membrane preparation with Ins-1,4,5-P3 receptor from microsomes (see Ref. 33). Hence, it is clearly premature to conclude that hepatocyte plasma membranes contain an integral Ins-1,4,5-P3 receptor/Ca2+ channel.
If hepatocytes possess both plasma membrane and internal membrane Ins-1,4,5-P3 receptors, there must be differences in their regulatory properties, since there is a distinct temporal difference between mobilization of internal stores and activation of extracellular Ca2+ entry. At present there is insufficient data to conclude that distinct subsets of Ins-1,4,5-P3 receptor exist in hepatocytes. However, the properties of the Ins-1,4,5-P3 receptor can be quite different between different cell types. For example, in brain (32, 40) and uterus membranes (41), Ins-1,4,5-P3 binding to its receptor is inhibited by Ca2+, while in adrenal cortex membranes (41), Ca2+ has no effect on Ins-1,4,5-P3 binding and in liver membranes (42) Caz+ enhances Ins-1,4,5-P3 binding. Both high and low affinity Ins-1,4,5-P3-binding sites have been characterized in liver (43) and adrenal cortex membranes (44). In liver, it appears that it is the hormone-induced rise in Ca2+ that converts the Ins-1,4,5-P3 receptor from the low to the high affinity state (43). The basis for the differential effects of Ca2+ is not yet known but may be related to the presence of the regulatory Ca2+-binding protein calmedin (45) or to the phosphorylation state of the receptor (46, 47). Hence, the temporal difference between activation of internal Ins-1,4,5-P3 receptors and plasma membrane Ins-1,4,5-P3 receptors could result from their association with different regulatory proteins, which results in a requirement for different ancillary factors to induce activation of the Caz+ channel.
Molecular cloning studies of the Ins-1,4,5-P3 receptor have indicated that it is related to the other major intracellular Ca'+ release channel, the ryanodine receptor (48, 49). The ryanodine receptor possesses a large NH2-terminal domain (50), forming the foot structure that interacts with the dihydropyridine receptor (the L-type Ca2+ channel) of the plasma membrane in skeletal muscle excitation-contraction coupling. Since the Ins-1,4,5-P3 receptor also possesses a large cytoplasmic NH2-terminal domain, it has been suggested (51) that it might interact with plasma membrane proteins involved in regulating Ca2+ entry in an analogous fashion. Information relating the Ca2+ content of the internal Ca2+ store could then be sensed by the luminal domain of the Ins-1,4,5-P3 receptor and then transmitted to the plasma membrane to signal Ca2+ entry. This hypothesis is supported by the observation that a significant portion of the 1,4,5-P3 receptors are located in the subplasma membrane space (33,38,39). This postulated protein interaction not only provides a mechanism for capacitative Ca2+ entry, but also provides a mechanism for the putative interaction of cytosolic regulatory molecules, like Ins-1,4,5-P3, to coordinate internal and extracellular Ca'+ mobilization. Whereas internal Ca2+ release would require only Ins-1,4,5-P3, maximal stimulation of extracellular Ca2+ entry would require both the release of internal Ca2+ and the presence of Ins-1,4,5-P3, as observed in this paper. Depleted internal Ca2+ stores in the absence of Ins-1,4,5-P3 would not necessarily be an effective signal for activating Ca'+ entry. That is, the conformational changes in the large cytoplasmic domain induced by Ins-1,4,5-P3 binding would not only cause opening of the intrinsic Ca'+ channel of the Ins-1,4,5-P3 receptor, but also modify the interaction of this domain with associated plasma membrane proteins, thus promoting Ca'+ entry. Hence, this scheme would not require actual integral plasma membrane Ins-1,4,5-P3 receptors for Ins-1,4,5-P3 to contribute to the gating of Ca2+ entry.
The observation that maximal stimulation of Ca2+ entry in hepatocytes requires both the release of internal Ca2+ stores and the presence of Ins-1,4,5-P3 is supported by the recent study of Kass et al. (52). They observed that hormone-stimulated Ca'+ entry in hepatocytes had a significantly longer latency period than did internal release and that simply emptying the Ins-1,4,5-P3-sensitive Ca'+ pool, by treating the cells with 2,5-di(tert-butyl)-l,4-benzohydroquinone (53), did not lead to a significant stimulation of Ca2+ entry. Only upon addition of hormone to Ca'+-depleted cells was Ca2+ entry activated. Their study did not identify the hormone-generated factors, but from our study it appears that Ins-1,4,5-P3 is one of the factors.
Despite intensive research by many laboratories to elucidate the mechanism that underlies hormone-stimulated Ca2+ entry, no simple consensus has emerged. This is not surprising, since studies on single cells have demonstrated that the Ca2+ signal generated by activation of different receptor systems is quite diverse, being capable of differentially activating internal mobilization relative to extracellular Ca'+ entry, and coordinating these responses to generate Ca2+ waves, repetitive Ca'+ spikes, and gradients of Ca2+ within the cytosolic compartment (see Ref. 54). In endothelial (55) and parotid cells (15), the Ca'+ status of the internal stores appears to be a major factor, whereas in hepatocytes (52, this study), hormone generated Ins-1,4,5-P3 is essential, and in lacrimal cells (12), the presence of both Ins-1,4,5-P3 and Ins-1,3,4,5-P4 appear to be necessary to activate Ca2+ entry. Given the diversity of receptor systems coupled to Ca'+ signaling, the variety of responses that can be elicited from the same cell and the probable complex network of cross-talk that exists between these systems, it is likely that additional external factors will be found that regulate and coordinate internal Ca'+ mobilization and extracellular Ca2+ entry.