Evidence for the Role of Phosphorylase Kinase, Protein Kinase C, and Other Ca2+-sensitive Protein Kinases in the Response of Hepatocytes to Angiotensin 11 and Vasopressin*

Angiotensin 11, catecholamines, and vasopressin can stimulate the phosphorylation of 10 hepatic cytosolic proteins via a Ca"+-linked, cyclic AMP-independent mechanism, To explore the role of known Caz+-sensi-tive protein kinases in this response, [32PlP043--la- beled hepatocytes were stimulated with various agonists, the cytoplasmic proteins were separated on two- dimensional gels, and the resulting autoradiographs were computer analyzed. The role of phosphorylase kinase was examined using hepatocytes from gsdjgsd rats which are deficient in this enzyme. The phosphorylation state of phosphorylase was not increased by glucagon, angiotensin 11, or vasopressin in hepatocytes from the gsdlgsd animals. The phosphorylation state of all other substrates was changed by glucagon or the Ca2+-linked hormones to the same extent in gsdlgsd hepatocytes as in normal Wistar controls, suggesting that phosphorylase kinase plays a restricted role in the hormone response, The role of the Ca2+- and phospho-lipid-sensitive protein kinase (protein kinase C) was examined by stimulating hepatocytes with phorbol es- ters which are thought to activate protein kinase C by substituting for diacylglycerol. Phorbol esters in- creased the phosphorylation state of 3 bovine esters 100% dimethyl sulfoxide, angiotensin I1 vaso- pressin 154 NaC1. the cells 100-fold concentrated stocks. Incubations by lysing the cells to a slight modification (17) of the of Janski and This lyses the cells with and the cytoplasmic proteins from the rest of the cell in 6-10 s minimizing the of proteolysis. The cytoplasmic then for two-dimensional gel electrophoresis as assay of Ca2' -fluxes or rylase a in a regular phosphate Krebs-Ringer L-lactate and pyruvate were reduced and respectively, in these experiments.


Recipient of United States Public Health Service Postdoctoral
Fellowship  11, vasopressin, and al-adrenergic agonists stimulate different membrane events including phosphatidylinositol turnover (4-9) and Ca" fluxes (1-3, 10,11). Although these two types of hormones generate very different biochemical messengers, their net effects on carbohydrate metabolism are similar, with both stimuli leading to increased glycogenolysis (1-3) and gluconeogenesis (12-14). Moreover, both types of stimuli appear to control the activity of the important regulatory enzymes of carbohydrate metabolism by changing their phosphorylation state (15-17). These observations suggest that hepatocytes contain protein kinases responsive to a variety of intracellular messengers but that the substrate specificity of these enzymes overlaps to a significant extent (17, 18).
Evidence from a number of experimental approaches suggests that the effect of glucagon on the phosphorylation state of hepatic enzymes is mediated by the cyclic AMP-dependent protein kinase (16,(19)(20)(21). In contrast, the nature and role of the protein kinases that respond to a Caz+-linked hormone such as vasopressin are unknown. Although a number of possibilities exist, the relevant kinases have not been identified. Based on the literature, the candidates must include the Ca2+-and phospholipid-dependent protein kinase (protein kinase C) (22,231 and Ca2+-calmodulin-requiring kinases such as phosphorylase kinase (24) and glycogen synthase kinase (25,261. Other enzymes, such as Ca2+-regulated phosphatases or unknown Ca2+-sensitive kinases, may also participate. The purpose of the present study was to define the role of some of these enzymes in the response of hepatocytes to Caz+linked hormones by studying protein phosphorylation in the intact cell. The role of phosphorylase kinase was examined using hepatocytes from gsdlgsd animals that are deficient in this enzyme (27). The role of protein kinase C was explored by stimulating cells with phorbol esters which have been shown to activate this kinase in intact platelets (28,29). The role of other Ca2+-sensitive phosphorylation events was examined by treating cells with the Ca2+ ionophore A23187. The results indicate that the full response to a Ca2+-linked hormone requires the participation of phosphorylase kinase, pro-and stimulation times were chosen to give maximal responses (16). Glucagon was dissolved in 0.01 N NaOH containing 1 mg/ml of crystalline bovine serum albumin, phorbol esters and A23187 were dissolved in 100% dimethyl sulfoxide, and angiotensin I1 and vasopressin in 154 mM NaC1. All agents were added to the cells as 100fold concentrated stocks. Incubations were terminated by lysing the cells according to a slight modification (17) of the method of Janski and Cornell (31). This procedure lyses the cells with digitonin and separates the cytoplasmic proteins from the rest of the cell in 6-10 s minimizing the chances of proteolysis. The cytoplasmic proteins were then prepared for two-dimensional gel electrophoresis as described (17). When cells were used for the assay of Ca2' -fluxes or phosphorylase activity, they were prepared as described above but resuspended a t a concentration of 10-20 mg of protein/ml in a regular phosphate Krebs-Ringer bicarbonate buffer. The concentrations of L-lactate and pyruvate were reduced to 16 mM and 4 mM, respectively, in these experiments.
Two-dimensional Gel Electrophoresis, Autoradiography, and Computer Analysis-Cytoplasmic proteins were resolved by two-dimensional gel electrophoresis exactly as described (17). Gels were stained with Coomassie brilliant blue and dried on filter paper. Autoradiography was performed with Kodak X-Omat K (XK-1) film for 10-14 days. This double emulsion film proved superior to others (e.g. Du-Pont Chronex 4) because it combines fine grain, low optical background (about 0.2 A at 550 nm) and a moderately fast response to '"P. The use of slightly more [3zP]P02-in the incubation mixtures and Kodak XK-1 film produced a higher quality autoradiograph for computer analysis. The films were developed in a Kodak automatic processor and scanned on an Optronics C-4000 high speed densitometer at a raster size of 200 pz. The data (about 450,000 optical density readings per film) was stored on a 9-track tape. The optical density of the spots on the film was integrated by a substantially revised version of the program described by Garrison and Johnson (32).
Assays-Phosphorylase was assayed as [14C]glucose incorporated into glycogen by the filter paper method of Gilboe et al. (33) modified to use the assay mix recommended by Stalman and Hers (34). Samples were prepared for this assay by freezing the cells in a final concentration of 150 mM NaF, 10 mM Mes,' pH 6.1, as described (16). Protein was assayed by the method of Lowry et al. (35) using crystalline bovine albumin as a standard. The M , and PI of the proteins in the two-dimensional gel system was estimated using standard curves as outlined previously (17).
Measurement of Ca2+ Fluxes in Intact Cells-Calcium fluxes were measured using "Ca2+ as a tracer by an adaptation of the procedures described by Blackmore et al. (10) and Barritt et al. (11). Freshly prepared hepatocytes were washed 3 times in 10 volumes of Ca2+-free Krebs-Ringer bicarbonate buffer containing 1 mM EGTA. Cells were resuspended at a concentration of 0.2-1 X lo7 cells/ml and preincubated for 30-60 min with 1 pCi/ml of 'SCaC1z, 16 mM L-lactate, and 4 mM pyruvate in a shaking water bath at 37 "C in stoppered tubes which had been gassed with 95% 0 2 , 5% COz. After preincubation, cells were centrifuged at 50 X g and resuspended in Krebs-Ringer bicarbonate buffer containing 2.5 mM Ca2+. Two hundred p1 of the cell suspension were added to 5-ml plastic tubes containing vehicle or varying concentrations of hormone. Cells were incubated at 37 "C for 6-10 min under 95% 02, 5% COz and the incubations were terminated by the addition of 5 ml of an ice-cold wash solution containing 154 mM NaC1, 2.3 mM Lac&, 1% bovine serum albumin (Fraction V). The medium Ca2+ was separated from that in the cells by one of two methods. In some experiments, the cells were removed from the medium by centrifugation (13,000 X g for 1 min in an Eppendorf minifuge) through a layer of bromododecane ( p = 1.04) much as described by Barritt et al. (11). The 45CaY' in the medium and in the cell pellet was counted in a liquid scintillation counter with an efficiency of about 35%. In other experiments, the contents of the tubes were immediately poured over a Whatman GF/A filter which had been presoaked in the wash solution containing NaCI, LaCl3, and albumin. Tubes were rinsed with an additional 5 ml of wash solution and the rinse was filtered. 4sCa on the filters was measured in a liquid scintillation counter with an efficiency of 30-35%. Each hormone concentration was paired with a separate control.
The two methods of separating the cells from the medium were carefully compared using different hormones and found to give identical results.* The filter method was used in most experiments because samples could be processed much more rapidly. Regardless of the method used to separate cells from medium, the Ca2+ flux results are expressed as the per cent of 45Ca efflux caused by a given agonist as compared to the matched control.
Suppliers-The reagents used in this study were obtained from the following sources. Calculations and Expression of Results-Two-dimensional gel data presented under "Results" were chosen as representative from 4-8 experiments. Averaged data are presented as the mean k S.E. Differences between groups of data were evaluated for significance with the Student's t-test.

Expansion of the Autoradiograph Numbering System
The phosphorylated proteins in the cytosolic fraction of the hepatocytes were resolved on two-dimensional polyacrylamide gels and displayed on autoradiographs. In a previous report, the 37 darkest spots on the authoradiograph were assigned numbers and 6 of these proteins were identified as important regulatory enzymes in intermediary metabolism (see Fig. 4 and Table I of Ref. 17). Because of the technical improvements described under "Materials and Methods," two more spots were visualized on the autoradiographs. Moreover, the phosphorylation of each of these proteins is increased by Ca2+-linked hormones such as vasopressin or angiotensin but not by glucagon (see Table I, below). In order to keep the original numbering system intact, these new spots were designated "a" and "b." The alphabetical designation of these and future spots will allow expansion of the spot list without creating confusion. The positions of spots a, b, and the 21 other spots analyzed in the present experiments are indicated in Fig. 2. The molecular weights and isoelectric points of the phosphorylated forms of a and b are: a, M, = 70,000, PI = 5.8; b, M, = 56,000, PI = 5.7.

Role of Phosphorylase Kinase
Hepatocytes Deficient in Phosphorylase Kinase-Previous work has shown that phosphorylase kinase in the gsdlgsd rat is inactive in a number of tissues, including the liver (27). As a result, neither glucagon nor the Ca2+-linked hormones are able to stimulate phosphorylase activity in hepatocytes from these animals (27,36). T o ensure that the gsdlgsd animals used in this study were of the proper phenotype, phosphorylase activity was measured before and after treatment with glucagon or vasopressin. Basal phosphorylase activity in the gsdlgsd hepatocyte was 0.14 pmol of glucose/mg of protein/ 15 min, about 50% lower than basal activity in the Wistar rat. Treatment of the cells with M glucagon or 24 nM vasopressin for 2 min yielded an activity of 0.16 pmol of glucose/mg of protein/ 15 min, an 18% increase in activity.  (Table IV acid: EGTA. ethvlene glvcol bis(0-aminoethyl ether) (-N,N,N',N'-and (1-3, 16)). These results agree with those of other investetraacetic acid; p i , isoelectric point; PMA, 4~-phorbol-12-myristate-

Ca*+-sensitive Protein Kinases in Hepatocytes 3285
phosphorylase also focus as multiple isoelectric species. The protein (spot 4) is identified by the arrows pointing upward toward one major isoelectric form in each panel. One of the benchmark proteins, spot 10, whose phosphorylation state does not change with hormone treatment is also shown. These proteins are used to verify that the 32P specific activities, the amount of protein loaded on the gel, and the autoradiographic exposures are constant within an experiment (17).
Although approximately equal amounts of protein are visible in the stained gels from Wistar or gsdlgsd hepatocytes, there is a large difference in the amount of phosphate incorporated into phosphorylase in the two types of cells. Even in the basal state, the phosphorylase in the Wistar hepatocyte contains measurable amounts of phosphate and this is markedly increased by treatment of the cells with glucagon. In contrast, the phosphorylase molecule in the gsdlgsd hepatocyte contains very little phosphate in the basal state and its content increases only slightly (an average of 1.8-fold-see Table I, below) following treatment of the cells with glucagon. Moreover, neither angiotensin I1 nor vasopressin is able to stimulate the phosphorylation of phosphorylase in the gsd/ gsd hepatocyte (see Table I, below). These results document that phosphorylase kinase does not respond to hormones in the gsdlgsd hepatocyte. Moreover, the observation that phosphorylase in the liver of gsdlgsd animals contains very little phosphate in both the basal and stimulated states provides a molecular explanation for the lack of activity observed in the standard phosphorylase assay (above text and Refs. 27 and 36).
Role of Phosphorylase Kinase in the Response to Hormones-Previous work has demonstrated that stimulation of hepatocytes with glucagon or a hormone such as vasopressin stimulates the phosphorylation of separate but overlapping sets of substrates (17,18). In order to examine the role of phosphorylase kinase in the overall phosphorylation response, [:"P]PO~--labeled hepatocytes from gsdlgsd animals were stimulated with glucagon, angiotensin 11, or vasopressin and the cytoplasmic proteins were separated on two-dimensional polyacrylamide gels. The pattern of phosphorylated proteins observed on the autoradiographs did not differ from those of normal Wistar hepatocytes (data not shown). The integrated density information on the autoradiographs is presented in Table I along with matched controls performed with hepatocytes from Wistar rats.3 Four major points are evident from the data. First, in agreement with previous results, the quantitative effects of angiotensin I1 and vasopressin were very similar, therefore the data were combined and presented in one column (17). Second, as noted in Fig. 1, none of the hormones significantly increased the phosphorylation of phosphorylase (spot 4) in hepatocytes prepared from gsdlgsd animals. Third, the phosphorylation state of three substrates unique to the Ca2+-linked hormones (a, b, and 29) is still increased in hepatocytes from the gsdlgsd rat. Finally, excluding phosphorylase, there is no statistical difference between the effects of glucagon or the Ca"-linked hormones on the phosphorylation of their respective substrates in the hepatocytes from the two types of animals.
These results suggest that phosphorylase kinase plays a restricted role in the phosphorylation of hepatic proteins in response to glucagon or either of the Ca*+-linked hormones.
The most correct control animal would be the normal New Zealand strain (NZR/Mh) from which the gsdlgsd rats were derived. However, since these animals were not available and our entire data base has been obtained with Wistar rats, Wistar animals were used as controls.
tigators (27,36) and confirm that the animals used were of the proper phenotype. Molecular Basis of the Lesion-A molecular explanation for the inability of hormones to activate phosphorylase in the gsdlgsd rat is shown in Fig. 1. Hepatocytes from Wistar or gsdlgsd animals were labeled to equal specific activities and treated with M glucagon and equal amounts of cytoplasmic proteins loaded on two-dimensional gels. The position of phosphorylase in the gel system has been determined (17) and the figure shows the region of the gel containing the molecule. Phosphorylase is identified by the downward pointing arrow in the left-hand panels of the figure representing the stained proteins. The molecule has a M, of 93,000 and focuses as a string of 5 spots over the pH range of 6.5-6.7. The reason for the multiple isoelectric points is unknown; however, crystalline muscle phosphorylase also has multiple isoelectric forms. The multiple isoelectric forms do not appear to be due to multiple phosphorylation of phosphorylase from either tissue (37). The right-hand side of Fig. 1 presents the corresponding autoradiographs. The phosphorylated forms of  Tables I and 11). About 35 pg of protein was loaded on the gels. B, identical with A except that the hepatocytes were prepared from gsdlgsd rats. About 35 pg of protein was loaded on the gels.

A comparison of the effect of glucagon and the Ca2+-linked hormones on the phosphorylation state of cytoplasmic proteins from Wistar and gsdlgsd rats
Cells were labeled with [3'P]P04, stimulated with 100 nM glucagon for 4 min, 24 nM vasopressin or 100 nM angiotensin for 3 min, and cytoplasmic proteins were resolved on two-dimensional gels. Autoradiographs were prepared and the density information from 21 phosphoproteins was selected for computer analysis. The position of the proteins is shown in Fig. 2. The effect of hormones on the phosphorylation state (optical density) of each spot is presented below as -fold over control. Molecular weights and spot identities were determined as described under "Materials and Methods." The data for angiotensin I1 and vasopressin were combined into one column because they were not statistically different. Except for phosphorylase (spot 4) in the gsdlgsd animals ( p > 0.05), the effects of hormones on all proteins in A is significant ( p < 0.05).  Blanks in this column indicate that the identity of the protein is unknown.

___
Phosphorylase is clearly a substrate for the kinase but apparently none of the other phosphoproteins resolved in the gel system are affected by this enzyme. Therefore, other Caz+sensitive protein kinases must be involved in the hormone response. C-In 1982, Castagna et al. (28) demonstrated that phorbol esters, such as PMA could directly activate protein kinase C purified from brain. The kinetic effects of phorbol esters on the enzyme appeared to be identical with those of a series of diacylglycerol compounds (38) and it was proposed that phorbol esters activated the kinase by replacing the requirement for diolein (28). In addition, recent work has shown that a putative phorbol ester receptor co-purifies with protein kinase C through two column steps (39). Most importantly, using platelets as an experimental system, it has been demonstrated by a number of criteria that phorbol esters cells activate protein kinase C in the intact cell (28,29).

Role of Protein Kinase C Phorbol Esters and Protein Kinase
In light of these observations, phorbol esters were used to examine the role of protein kinase C in the response of hepatocytes to the Ca"-linked hormones. Hepatocytes were labeled with [3zP]POi-, stimulated with vasopressin or PMA and the cytoplasmic proteins were resolved on two-dimensional gels. Fig. 2 presents autoradiographs made from such an experiment. As noted in Table I,

Dose Response and Time Course of the PMA Effects-Dose
response and time course experiments were performed to establish the conditions required for maximal effects of phor-bo1 esters. Initially, 1 pg/ml of PMA was added to the hepatocytes and cytoplasmic proteins were prepared after 0.5, 1, 2, 3, and 10 min of stimulation. Two-dimensional autoradiographs were computer-analyzed and the time course of the phosphorylation of spots a, b, and 29 was measured. Four benchmark proteins were also analyzed and did not change.
The time course of the phosphorylation of all three proteins was very similar. An observable increase (about 40% of maximal) was apparent after 30 s of stimulation and maximal effects occurred between 2.0 and 3.0 min of treatment. This time course is very similar to that seen in platelets (28). The phosphorylation of all three proteins was maintained through- out the 10-min experimental period (n = 3). Based on these results, 2-3 min was used as a standard period of stimulation in all subsequent experiments. Dose-response curves were performed in a manner analogous to the time course experiments using a 2-min stimulation period and PMA concentrations ranging between 50 and 2000 ng/ml. The phosphorylation of all three proteins, a, b, and 29, showed a similar sensitivity to PMA. The half-maximal stimulation occurred between 100 and 200 ng/ml with maximal effects observed in the range of 500 and 1000 ng/ml. No inhibitory effects of higher doses (2000 ng/ml) were observed ( n = 3). Rased on these data, PMA was used at a dose of 1000 ng/ml in subsequent experiments. The sensitivity of hepatocytes to PMA appears to be slightly less than that observed in platelets (28,40) and other systems (41,42). However, phorbol esters are known to be lipophilic (41) and the hepatocytes are used at high cell concentrations in these experiments. Thus, the lack of sensitivity may be due to the large number of cells in the incubations.

Role of Kinases Sensitive to Ca'+ Influx
Effects of Ca" Ionophore-Previous results using onedimensional gels have demonstrated that A23187 could stimulate the phosphorylation of a few proteins in hepatocytes (15). Moreover, known calmodulin-requiring protein kinases such as phosphorylase kinase (24) and myosin light chain kinase (43)(44)(45) are activated by treatments which elevate cytosolic Ca'' levels (46)(47)(48)(49)(50). Therefore, in an attempt to investigate the role of putative calmodulin-requiring protein kinases in the hepatic response to vasopressin, hepatocytes were stimulated with the Car+ ionophore A23187. Care was taken to keep the time of exposure short (90-180 s) to minimize toxic effects of this compound on mitochondrial function.
T h e lower right section of Fig. 2 shows that treatment of the cells with 30 V M A23187 for 3 min changes the phosphorylation of 7 cytoplasmic proteins. These proteins are identified by arrowheads. Note that these 7 proteins comprise the rest of the 10 proteins affected by treating cells with vaso-

Ca2+-sensitiue Protein Kinases in Hepatocytes
pressin. Indeed, if cells are treated with a combination of A23187 and PMA, they respond exactly as if they had been treated with vasopressin (autoradiograph not shown, but see Table I1 below).
Dose Response and Time Course of A23187 Effects-Doseresponse and time course experiments were performed to ensure that maximal effects of A23187 on protein phosphorylation were observed. These experiments were performed in a manner analogous to those described above for phorbol esters and the autoradiographs were computer-analyzed. As shown in Table I1 below, the effect of A23187 on three phosphoproteins (spots 13, 17, and 23) is small. Therefore, only the effects of A23187 on spots 4, 34, 35, and 36 were analyzed along with 4 benchmark proteins. The time course of the effect of 10 p~ A23187 on the phosphorylation of the four proteins was similar and was about 30% of maximal a t 1 min, near maximal a t 2-3 min, and appeared to decline after 5-8 min (n = 3). No effects were observed on the benchmark proteins. Consequently, A23187 dose-response curves were performed with a 2-3 min stimulation. One p~ ionophore caused about a 30% response and maximal responses were observed following doses of 10-30 WM. The response of three proteins (spots 4, 35, and 36) was similar with spot 34 lagging somewhat behind the others (n = 3). In light of these results, all subsequent experiments were performed with a dose of 20-30 p M A23187 for 2-3 min.
Quantitation of Phosphorylation Changes following Treatment with A23187 and PMA-The data of Fig. 2 suggest that stimulation of hepatocytes with the combination of A23187 and PMA should mimic the response of the cell to vasopressin. To test this hypothesis, [32P]POi--labeled cells were stimu-lated with maximal doses of vasopressin, A23187, PMA, or A23187 + PMA and the autoradiographs were subjected to computer analysis. Table I1 presents the quantitative effects of vasopressin on the phosphorylation of the 10 proteins identified in the upper right quadrant of Fig. 2. As anticipated, PMA mimics the effects of vasopressin on only 3 proteins (spots a, b, and 29) while A23187 mimics the effects on the other 7 (spots 4, 13, 17, 23, 34, 35, and 36). While A23187 may have caused a small increase in the phosphorylation of spot 29, the effects were of borderline significance = 0.05).
The far right column shows that the combination of ionophore and phorbol ester is able to quantitatively reproduce the effects of vasopressin on protein phosphorylation. This result suggests that stimulation of hepatocytes with a hormone such as angiotensin I1 or vasopressin activates at least two distinct events within the cell. Apparently, phorbol esters and calcium ionophores can be used as probes to activate the separate segments of the hormone response.
Effects of Other Phorbol Esters-The effects of phorbol esters other than PMA have been examined in a wide variety of assay systems with very consistent results. PMA is usually the most potent compound with 4@-phorbol-didecanoate, 4Pphorbol-dibenzoate, and 40-phorbol-dibutyrate being about 10-20-fold less potent. 4a-Phorbol-didecanoate and 4P-phorbol-13-monoacetate are usually inactive (41). The expected structure activity pattern was observed when these compounds were tested for their ability to activate purified protein kinase C (28). In order to support the hypothesis that phorbol esters were activating protein kinase C in hepatocytes, the effects of the six analogues on the phosphorylation state of spots a, b, and 29 were examined. Table 111 demonstrates that   TABLE I1 Quantitative effects of vasopressin, PMA, and A23187 on the phosphorylation state of 10 cytoplasmic proteins Hepatocytes were labeled with ["P]POj-, stimulated with 24 nM vasopressin for 3 min, 1 pg/ml of PMA for 2 min, 30 @M A23187 for 3 min, or both PMA and ionophore for 3 min and the cytoplasmic proteins were resolved on two-dimensional gels. Autoradiographs were prepared and computer-analyzed as described under "Materials and Methods." The density information is presented below as -fold over control. All other details are as described in the legend to Table I

TABLE I11
A structure-activity study of the effects of various phorbol esters on protein phosphorylation Hepatocytes were labeled with [3'P]P0,"-and stimulated with the indicated concentration of phorbol ester for 2 min and the cytoplasmic proteins were resolved on two-dimensional gels. Autoradiographs were prepared and computer-analyzed as described under "Materials and Methods." The density information is presented below as -fold over control. Five benchmark proteins were also analyzed and the results were omitted because they were analogous to those in Tables  I and 11. Active and Inactive refer to the ability of the phorbol ester to activate purified protein kinase C (28).

TABLE IV
The effect of vasopressin, A23187, and various phorbol esters on the activity ofphosphorylase in hepatocytes Cells were incubated with 40 mM glucose for 20 min to suppress basal activity, stimulated with the indicated concentration of agonist for 2 min, and prepared for the phosphorylase assay as described under "Materials and Methods." Active and Inactive refer to the ability of the phorbol ester to activate purified protein kinase C (28).

Agonist added
Concentration Phospho~lase activity ' None of the phorbol esters tested increased phosphorylase activity ( p > 0.1).
PMA, 4P-phorbol-didecanoate, 4P-phorbol-dibutyrate, and 4/3-phorbol-dibenzoate were about equally effective in stimulating the phosphorylation of the three proteins. However, 10-20-fold more of the latter three compounds was required. 4ol-Phorbol-didecanoate and 4~-Phorbol-13-monoacetate were inactive a t doses up to 10 pg/ml. These results are very similar to the structure-activity relationships observed for the activation of purified protein kinase C and support the hypothesis that proteins a, b, and 29 are substrates for this enzyme in the intact cell.

Effect of Phorbol Esters on Ca2+ Fluxes
Effect on Phosphorylase-A number of studies have shown that the phosphorylation state and activity of phosphorylase in hepatocytes closely follows the level of free Ca2+ in the cytosol (46,47). Table 11 demonstrates that phorbol esters do not increase the phosphorylation state of phosphorylase (spot 4) in hepatocytes, suggesting that these compounds do not raise Ca2+ levels in liver cells. Since there are inconclusive reports on the effect of phorbol esters on Ca2+ levels in other types of cells (41), the effects of phorbol esters on Ca2+ fluxes were tested in hepatocytes. These experiments were performed indirectly, by assaying their effects on phosphorylase, and directly, by monitoring Ca2+ fluxes in intact cells. Table  IV presents the effects of a series of phorbol esters on the activity of phosphorylase in hepatocytes. The controls in the top portion of the table show that vasopressin and A23187 provide the expected stimulation of phosphorylase activity. However, none of the phorbol esters increased phosphorylase activity. These results confirm the prediction made from the data in Table 11.
Effect on Ca2+ Fluxes-When hepatocytes are completely equilibrated with 45Ca2+ and stimulated with hormones such as angiotensin I1 or vasopressin, a Ca2+ efflux is observed over the following 6-15 min (10, 11).  Tables 11 and IV and, taken together, the data provide strong evidence that phorbol esters do not elicit Caz+ fluxes in hepatocytes.

DISCUSSION
The ability of angiotensin 11, vasopressin, and cyl-adrenergic agonists to stimulate hepatic carbohydrate metabolism through a Ca2+-requiring, cyclic AMP-independent mechanism is well recognized (1-3, 51). In addition, previous work has shown that important regulatory enzymes in the glycogenolytic and gluconeogenic pathways can be controlled by these hormones via Ca2+-dependent protein phosphorylation reactions (15)(16)(17). The object of this study was to explore the role of various Ca2+-sensitive protein kinases in the response of hepatocytes to hormones such as angiotensin I1 or vasopressin. The data support the conclusion that these hormones activate a minimum of three Ca2+-sensitive protein kinases: phosphorylase kinase, protein kinase C and at least one other enzyme whose identity is unknown. Moreover, since the substrates for some of these enzymes in the gel patterns have been identified (17), important conclusions can be drawn regarding the substrates phosphorylated by the respective kinases in the intact cell. This information helps to define the role of the particular protein kinase in the overall metabolic response of the cell.
Phosphorylase kinase is functionally inactive in the liver of the gsdlgsd rat (27,36). Therefore, hepatocytes prepared from these animals provide an excellent experimental system to examine the role of this kinase in the response of the cell to glucagon or the Ca2+-linked hormones. The data presented in Fig. 1 and Table I clearly show that phosphorylase (spot 4) is not phosphorylated in response to hormones in hepatocytes from the gsdlgsd animal. However, the potentially more interesting result is that phosphorylase is the only substrate observed for phosphorylase kinase in these experiments. Quantitation of the effects of glucagon, angiotensin, or vasopressin on 15 other cytoplasmic phosphoproteins in hepatocytes from Wistar or gsdlgsd animals provided no measurable differences between the response of these two cell types to hormones (Table I). These results suggest that phosphorylase kinase has a very restricted role in the biochemical response of the hepatocyte to a hormone such as glucagon or vasopressin. The data support and extend the observations of Clark et al. (52) showing that catecholamines still inactivate pyruvate kinase in hepatocytes from gsdlgsd animals. Moreover, they provide confirmation, in the intact cell, of the data obtained by Chrisman et al. (24) with purified hepatic phosphorylase kinase showing that the enzyme has a very restricted set of substrates in uitro. On the other hand, it must be cautioned that the data in this report cannot be used to argue that phosphorylase is the only substrate in the cell for phosphorylase kinase. Only cytoplasmic proteins were examined in these experiments and even within this set of proteins, certain important enzymes such as glycogen synthase and acetyl-coA carboxylase do not enter the focusing dimension of the gel system (17). How many other potential substrates exist in other fractions of the cell or outside the M , and PI limits used in this work is unknown. However, since in uitro work with purified skeletal muscle phosphorylase kinase has shown that it can phosphorylate a number of substrates including glycogen synthase (53-55) and troponin (56), studies are planned to answer these important questions.
Protein kinase C is a Ca2+-requiring kinase that has a wide distribution in mammalian tissues (57,58). While the complete role of this kinase in the hormonal response of cells is unknown, it is noteworthy that this protein kinase can bind to membranes (39,42) and is markedly stimulated by the first product of phosphatidylinositol breakdown, diacylglycerol (23, 38). Therefore, this enzyme is thought to participate in signal transduction for hormones that elicit the phosphatidylinositol response (59, 60). Since phorbol esters have been demonstrated to activate this protein kinase in the intact cell (28,291, these compounds were used to explore the role of the enzyme in the response of hepatocytes to hormones such as angiotensin 11 or vasopressin. The data obtained yield five important conclusions. First, stimulation of the cells with a phorbol ester elicits only a part of the vasopressin response (3 of 10 proteins, substrates a, b, and 29). Second, all of these protein substrates are unique for Ca2+-linked hormones; none of them are phosphorylated in response to glucagon (Table  I). Third, spots a, b, and 29 are relatively low in concentration in the cell and are not visible in gels stained with Coomassie blue (they can be seen if the gels are stained with silver, data not shown). Fourth, as is obvious from Table I, these proteins do not represent any of the regulatory enzymes so far identified in the gel pattern. Finally, stimulation of the cells with Ca2+ ionophores does not cause an increase in the phosphorylation state of these proteins (Table 11).
Taken together, these results suggest that protein kinase C is responsible for a portion of the phosphorylation response observed following stimulation of the cells with a Ca2+-linked hormone. However, the present data do not identify any role for this enzyme in the regulation of glycogen metabolism. Indeed, no effects of phorbol esters were observed on phosphorylase activity or on Ca2+ fluxes in the cell (Table IV and Fig. 3). Although a yet undiscovered role in gluconeogensis cannot be eliminated, no effect of phorbol esters was observed on phosphoproteins known to regulate the gluconeogenic pathway (Tables I and 11). These observations suggest that protein kinase C does not participate directly in the regulation of carbohydrate metabolism and therefore must be involved in some other aspect of hepatic function.
Another intriguing result is that Ca2+ influx does not increase the phosphorylation of the putative protein kinase C substrates, suggesting that some other messenger activates this kinase. An obvious candidate would be a diacylglycerol released during the breakdown of phosphatidylinositol induced by angiotensin I1 or vasopressin. In this regard, it is encouraging that synthetically prepared 1-oleoyl-2-acetylglycerol, a diacylglycerol analogue that can cross the cell membrane (61), exactly mimics the effects of phorbol esters on the phosphorylation of spots a, b, and 29 in hepatocytes! This observation clearly supports the idea that angiotensin and vasopressin can activate protein kinase C via changes in diacylglycerol levels. Finally, it must be realized that while the results suggesting that hormones or phorbol esters activate protein kinase C in the intact cell are convincing, it has not been proven that this is the only effect of phorbol esters in cells. Indeed, a host of effects of these compounds have been described (41). Thus, additional proof that protein kinase C is involved in the phosphorylation of substrates a, b, and 29 in liver will have to await in uitro studies with purified kinases and substrates.
The majority of the cytosolic proteins (spots 13, 17, 23, 34, 35, 36) phosphorylated in response to angiotensin 11 and vasopressin are substrates for protein kinases other than phosphorylase kinase or protein kinase C. An important discovery is that the quantitative effects of hormones on these substrates can be mimicked by addition of the Ca2+ ionophore J. C . Garrison

Ca2+-sensitive Protein
A23187 to hepatocytes. Clearly, A23187 causes a rise in the cytosolic Ca2+ concentration (10 and text). It is significant that one of the enzymes activated by this event is phosphorylase kinase, since this enzyme is known to be stimulated by Ca2+ and calmodulin (24). A similar situation exists in platelets (61, 62) and smooth muscle (48)(49)(50)631 where A23187 or Ca2+-linked hormones increase the phosphorylation of myosin light chain via another enzyme known to require Ca2+ and calmodulin for full activity, myosin light chain kinase (43)(44)(45). Based on the above and consideration of the literature (64,65), it is tempting to argue that the hepatic proteins phosphorylated in response to A23187 are substrates for Ca2+calmodulin-regulated kinases or phosphatases. While there is no direct experimental proof of this hypothesis, a number of known calmodulin-requiring enzymes may explain the observed results.
One known Ca2+-calmodulin-requiring protein kinase in liver is glycogen synthase kinase (25, 26). When studied in vitro with purified substrates, this enzyme can phosphorylate glycogen synthase, smooth muscle myosin light chain, casein, and phosvitin. Whether pyruvate kinase (spot 13), phenylalanine hydroxylase (spot 17), and other cytosolic proteins are substrates for this enzyme has not been reported. Another consideration is that, to date, this enzyme has only been studied in rabbit liver (25,26) and information about the rat liver enzyme is lacking. Therefore, conclusions about the role of this protein kinase in the response of rat hepatocytes to angiotensin or vasopressin must await further studies. Another interesting discovery is that calcineurin appears to be a Ca2+-calmodulin-stimulated phosphatase (66). This finding is intriguing because hormones and A23187 decrease the phosphorylation of one protein, spot 36 (Tables I and 11). If a similar protein phosphatase exists in liver, this enzyme may participate in the response to hormones along with the protein kinases discussed above. While the nature of all the enzymes mediating the response of the cell to a Ca2+ influx is not known, it is clear at least that some of the substrates affected by these enzymes are those involved in regulating carbohydrate metabolism. For example, phosphorylase (spot 4) and pyruvate kinase (spot 13), proteins important in controlling glycogenolysis and gluconeogenesis, are both substrates for the Ca2+-calmodulin-requiring enzymes. This finding is consistent with the observation that the effects of angiotensin I1 or vasopressin on the activity of these enzymes require Ca2+ ion in the bathing medium (1-3, 16). Further work with purified enzymes will be necessary to fully understand the nature of all the Ca2+-sensitive enzymes involved in this segment of the hormone response.
Perhaps the most interesting implication of the data obtained in this study is that binding of a hormone such as vasopressin to a cell membrane appears to elicit responses within the cell via two different messengers, diacylglycerol and Ca2+ ion. Each of these signals can activate distinct protein kinases which phosphorylate only a portion of the substrates affected by the hormone itself. Apparently, phorbol esters can activate one segment of this response by mimicking the effect of diacylglycerol on protein kinase C. One potentially important conclusion from this line of reasoning is that activation of protein kinase C does not appear to initiate Ca2+ fluxes in the hepatocyte (Fig. 2, 3, and Table IV). This observation may help explain the apparent lack of correlation between phosphatidylinositol metabolism and phosphorylase activation observed in some studies (67)(68)(69). A23187 appears to activate the other portion of the hormone response by activating at least one Ca*+-calmodulin-requiring enzyme (phosphorylase kinase) and perhaps other calmodulin-regu-lated enzymes acting on their respective substrates. Recent evidence suggests that these biochemical mechanisms may exist in a number of cell types and are capable of regulating a variety of intracellular events. Extensive data have been accumulated in platelets where thrombin appears to stimulate protein kinase C via diacylglycerol and myosin light chain kinase via Ca2+ ion (61,62). In addition, thyrotropin-releasing hormone appears to regulate protein phosphorylation and prolactin release in GH, cells through very similar mechanisms (70, 71). Clearly, the cellular responses to hormones such as angiotensin 11, thrombin, thyrotropin-releasing hormone, or vasopressin are complex and new insights into the role of protein phosphorylation in cell function will be gained by further study of these systems.

J C Garrison, D E Johnsen and C P Campanile
vasopressin. and Ca2+-sensitive protein kinases in the response of hepatocytes to angiotensin II Evidence for the role of phosphorylase kinase, protein kinase C, and other