Regulation of Phosphoenolpyruvate Carboxykinase (GTP) Synthesis in Rat Liver Cells RAPID INDUCTION OF SPECIFIC mRNA BY GLUCAGON OR CYCLIC AMP AND PERMISSIVE EFFECT OF

Isolated rat liver cells maintained in suspension culture for 4 to 6 h synthesize the gluconeogenic cytosolic enzyme phosphoenolpyruvate carboxykinase at a rate approximately &fold lower than the in vivo hepatic rate. Glucagon rapidly re-induces phosphoenolpyruvate carboxykinase synthesis in such cells. The rate of enzyme synthesis doubles in 40 min and plateaus at a level 6- to 13-fold higher than in control cells 120 min after glucagon addition at maximal concentration. Consistent with the presumed role of cyclic AMP as a mediator of enzyme induction, the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, added simulta- neously with glucagon, shifts the hormone dose-re-sponse curve 2 log units to the left. Moreover, cyclic AMP supplied exogenously to the cells mimics the inductive effect of glucagon. Total cellular RNA isolated from hepatocytes induced by glucagon contains an increased level of mRNA coding for phosphoenolpyruvate carboxykinase, as determined by translational as- say. The kinetics and extent of the rise in mRNA level are adequate to explain the stimulation of enzyme syn- thesis. Although glucagon on its own induces a build-up of phosphoenolpyruvate carboxykinase mRNA and a commensurate stimulation of enzyme synthesis, the glucagon induction is very markedly P-enolpyruvate carboxykinase served as an estimate of the concentration of functional mRNA coding for the enzyme in total cellular RNA. The incorporation of leucine into total polypeptides and into P-enolpyruvate car- boxykinase was dependent on the RNA input within a range of concentrations comprised between 0 and 10 pg of total cellular RNA in a reaction volume of 80 pl. Routinely, the assay was performed using 6 or 9 pg of RNA. Radioactivity Measurements-Protein precipitates were dissolved in NCS Tissue Solubiliier and the tritium radioactivity measured by liquid scintillation spectrometry with an efficiency of about 50%. Cylindrical SDS-polyacrylamide gels containing immunoprecipitated P-enolpyruvate carboxykinase were fractionated in 2-mm portions. The radioactivity in the gel fragments was measured after elution in the liquid scintillation mixture described in Ref. 29. Ninety-five % of the radioactivity was released from the gel pieces and counted at an efficiency of about 50%. Prior to fractionation, the gels were scanned spectrophometrically at 280 nm in order to localize the position of P-enolpyruvate carboxykinase. The amount of r3H]leucine incorporated into P-enolpyruvate carboxykinase was computed by adding the radioactivity of the fractions corresponding to the enzyme band and subtracting a background estimated from the radioactivity of the neighboring fractions.


RAPID INDUCTION OF SPECIFIC mRNA BY GLUCAGON OR CYCLIC AMP AND PERMISSIVE EFFECT OF DEXAMETHASONE*
(Received for publication, January 28, 1982, and in revised form, June 2, 1982) Agusti Salavert and Patrick B. IynedjianS Isolated rat liver cells maintained in suspension culture for 4 to 6 h synthesize the gluconeogenic cytosolic enzyme phosphoenolpyruvate carboxykinase at a rate approximately &fold lower than the in vivo hepatic rate. Glucagon rapidly re-induces phosphoenolpyruvate carboxykinase synthesis in such cells. The rate of enzyme synthesis doubles in 40 min and plateaus at a level 6-to 13-fold higher than in control cells 120 min after glucagon addition at maximal concentration. Consistent with the presumed role of cyclic AMP as a mediator of enzyme induction, the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, added simultaneously with glucagon, shifts the hormone dose-response curve 2 log units to the left. Moreover, cyclic AMP supplied exogenously to the cells mimics the inductive effect of glucagon. Total cellular RNA isolated from hepatocytes induced by glucagon contains an increased level of mRNA coding for phosphoenolpyruvate carboxykinase, as determined by translational assay. The kinetics and extent of the rise in mRNA level are adequate to explain the stimulation of enzyme synthesis. Although glucagon on its own induces a buildup of phosphoenolpyruvate carboxykinase mRNA and a commensurate stimulation of enzyme synthesis, the glucagon induction is very markedly amplified when the cells are first preincubated with dexamethasone. The glucocorticoid by itself, however, does not have any substantial effect on the level of phosphoenolpyruvate carboxykinase mRNA or on the rate of enzyme synthesis. Its role can therefore be characterized as permissive.
Glucagon plays a key role in the adaptive response of hepatic metabolism during the transition from the fed to the fasting states. The actions of glucagon in the liver include a stimulation of glycogenolysis, gluconeogenesis, and fatty acid oxidation, and conversely an inhibition of glycogen synthesis, glycolysis, and fatty acid synthesis (1,2). These metabolic changes are acutely caused by alterations of enzyme activities due to phosphorylation, as exemplified by the regulation of phosphorylase, glycogen synthase, and pyruvate kinase (3,4). A second type of regulatory process, operating over a longer time range, consists of hormonal effects on the synthesis, and * This research was supported by Grant 3.632-0.80 from the Swiss National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom correspondence should be addressed at, Department of Pharmacology, University of Lausanne, 21 Rue du Bugnon, CH-1011 Lausanne, Switzerland. eventually on the content, of various enzymes (5-7). A classical example is the glucagon-induced increase in cytosolic P-enolpyruvate carboxykinase (phosphoenolpyruvate carboxykinase (GTP) EC 4.1.1.32) (a), the rate-limiting enzyme of gluconeogenesis, as a consequence of a selective stimulation of the enzyme synthesis (9). The accumulation of P-enolpyruvate carboxykinase is considered as an important factor to sustain active gluconeogenesis during fasting.
Our interests lie in the sequence of events which lead to an alteration of the synthesis rate of specific hepatic proteins, in particular P-enolpyruvate carboxykinase, following the interaction of glucagon with its membrane receptor. As in the case of enzyme regulation by phosphorylation, this sequence is currently thought to involve the initial activation of adenylate cyclase, resulting in a rise of intracellular cyclic AMP. In earlier studies, it was shown that the injection of glucagon or dibutyryl cyclic AMP to the rat results in the rapid accumulation in the liver of functional mRNA coding for P-enolpyruvate carboxykinase (10)(11)(12)(13). Similarly, the cyclic AMP-dependent induction of hepatic tyrosine aminotransferase synthesis was reported to result from an increase in translatable enzyme mRNA (14, 15). These data strongly suggest a role of cyclic AMP at the transcriptional or post-transcriptional levels, as opposed to the translational level, in the regulation of gene expression in the mammalian liver. However, the molecular processes underlying this regulation remain to be elucidated.
Previous studies on the regulation of hepatic P-enolpyruvate carboxykinase synthesis by glucagon or cyclic AMP were carried out either in vivo in the whole rat or in vitro in hepatoma derived cell culture systems. Both experimental approaches have obvious limitations. First, the synthesis of Penolpyruvate carboxykinase in the rat liver is affected by a variety of hormones, including epinephrine, the glucocorticoids, insuIin, and possibly still other factors, in addition to glucagon (16). The interaction between these multiple factors makes it difficult to define the mechanism of action of individual inducers in uiuo. Furthermore, experiments involving precursor incorporation techniques or the use of inhibitors of macromolecule synthesis are often technically difficult or unsatisfactory in the whole animal. On the other hand, hepatoma cells cultured in vitro may display defective or abnormal mechanisms of response to hormones. For these reasons, we have undertaken a study of the regulation of P-enolpyruvate carboxykinase synthesis in normal rat liver cells maintained in suspension culture. The present report describes the rapid, large, and consistent induction of enzyme synthesis elicited in this system by glucagon or cyclic AMP, as a consequence of the accumulation of functional P-enolpyruvate carboxykinase mRNA. The system is also used to investigate the permissive mRNA in Hepatocytes 13405 role of the glucocorticoids in the cyclic AMP-mediated induction of the enzyme.

EXPERIMENTAL PROCEDURES
Materials-Collagenase from Clostridium histolyticum (Type IV, batch 100F-6810) was purchased from Sigma. Bovine serum albumin (Fraction V powder from Sigma) was extensively dialyzed against 5 m~ Na phosphate buffer, pH 7.4, and lyophilized before its addition to the incubation media. Crystallized highly purified porcine glucagon was a generous gift of Dr. Jorgen Schlichtkrull of the Novo Research Institute. It was dissolved in 0.003 N HCl, 2% (w/v) glycerol, and 0.1% (v/v) phenol and stored as a stock solution at 2 "C. Dexamethasone was bought from Sigma. It was dissolved in ethanol at a concentration of 14.5 mM and diluted with water to a concentration of 0.44 mM immediately before addition to incubated cells. 3-isobutyl-I-methylxanthine (from Sigma) was dissolved in 0.1 N NaOH at a concentration of 18.1 mM and neutralized with 1 M HCl immediately before addition to the incubation medium. All other biochemicals were supplied by Sigma, Boehringer Mannheim, or Bio-Rad. Penicillin and streptomycin were supplied by Grand Island Biological Co. ~- [3,4,5-3H]leucine (specific activity, 110-140 Ci/mmol) was purchased from New England Nuclear or Amersham International Ltd. NCS Tissue Solubilizer was from Amersham.
Animals-Male Wistar rats, bought from Kleintierfarm Madoerin Inc. (Fuellinsdorf, Switzerland) and weighing between 180 and 260 g at the time of the experiments, were used. Twenty-four h before the experiments, food was removed from the cages. After 22 h of fasting, the animals were fed 7 g of glucose/kg of body weight by gavage and offered chow pellets. They were killed for hepatocyte isolation 2 h after gavage. The usefulness of the fasting-refeeding cycle in bringing about a uniformly low rate of hepatic P-enolpyruvate carboxykinase synthesis has been shown previously (10). For experiments with adrenalectomized rats, the removal of the adrenal gland was performed under ether anesthesia 3 to 5 days before hepatocyte isolation.
After surgery, the rats received I% (w/v) NaCl to drink instead of water.
Liver Cell Isolation-Surgery and liver perfusion were carried out basically as described by Seglen (17). The calcium-free buffer, collagenase buffer, suspension and washing buffers were similar to those of Seglen, except that all were supplemented with 5 m~ glucose, 100 units/ml of penicillin, IoOpg/ml of streptomycin, and 2 mM glutamine. Nineteen amino acids were also added to provide final concentrations mimicking plasma levels in the rat (18). Moreover, bovine serum albumin was present at 1% (w/v) in suspension and washing buffers. All the solutions were thoroughly gassed by bubbling with 0 2 before use. Care was taken to avoid bacterial contamination throughout the perfusion, isolation, and incubation procedures. The liver perfusion was performed in two steps, first using the calcium-free buffer and subsequently a recirculating collagenase buffer containing 0.045% (w/ v) collagenase. The digestion time was strictly limited to 10 minutes. After perfusion, the liver was transferred to a petri dish containing 75 ml of suspension buffer and the cells were dispersed using a dog comb. The cells were filtered once through a 250-pm hole nylon mesh and centrifuged at 200 rpm (i.e. 7.2 X g,J for 2.5 min in a bench top centrifuge. The cells were washed three times in 80 ml of washing buffer, centrifugation between washes being performed as above. A stock suspension of cells was made up by dispersing the final cell pellet in an incubation medium containing: 137 m~ NaCI, 5.36 mM KCl, 1.22 m~ CaCIZ, 0.64 mM MgC12, 1.10 mM KH2P04, 0.70 mM NaZSO,, 25 mM NaHC03, 30 m~ 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid, pH 7.4. The medium was supplemented before use with 10 m~ glucose as well as with glutamine, other amino acids, and antibiotics at the concentrations indicated above. Dialyzed bovine serum albumin was also added at 2% (w/v). The medium was saturated with oz/COz (95:5) before use. The cells were counted in a Neubauer haemocytometer. In twenty-five experiments, the cell yield per liver was 489 f 37 X 10' cells (mean f S.E.). The stock suspension of cells was made up to contain approximately 20 x 10' cells/d.
Incubation of Ceiis-In the initial series of experiments, in which only the relative rate of P-enolpyruvate carboxykinase synthesis was measured, samples of about 8 X 10' cells (i.e. 0.4 ml of stock suspension) were distributed into individual incubation flasks (25-d polyethylene vials for scintillation counting) containing 4.0 ml of supplemented incubation medium. In subsequent experiments, in which both P-enolpyruvate carboxykinase mRNA level and relative rate of enzyme synthesis were simultaneously determined, 60 x IO' cells were incubated in individual 250-ml Erlenmeyer flasks containing 60 ml of incubation medium. In all experiments, incubation was carried out at 37.5 "C under a gas phase consisting of OZ/COZ (95:5) in a metabolic incubator with reciprocating shaking at about 1 0 0 cycles/min. Hormones and other additions to the cells were made from concentrated stock solutions and control cells received the vehicle only. The viability of the cells after isolation and at various times of incubation was examined by the trypan blue exclusion test at a final dye concentration of 0.25% (w/v).
Pulse-labeling of Cells-In the experiments involving the measurement of the relative rate of P-enolpyruvate carboxykinase synthesis only, the cells were pulse-labeled with [3H]leucine by Method A, as follows. At chosen times, 5 X IO6 cells were transferred from the incubation flasks to 12-ml polystyrene tubes, spun at 1 , O O O rpm for 3 s in the bench top centrifuge and washed once at room temperature in 2 ml of leucine-free incubation medium. The washed cell pellet was suspended for protein labeling in 1 ml of leucine-free incubation medium containing 30 pCi/ml of [3H]leucine. Incubation in this medium was continued for 20 min at 37.5 "C, followed by two washes in 5 ml of washing buffer without albumin. The final cell pellet was suspended in 1 ml of a homogenization buffer containing 10 mM Trisl HC1, pH 7.4, 200 m~ sucrose, 60 mM NaCl, and 1 mM EDTA. Cell homogenization was accomplished by three cycles of freezing-thawing, using liquid Nz. In the experiments involving the determination of enzyme synthesis rate and enzyme mRNA level at the same time, samples of 200 X lo3 cells were pulse-labeled using Method B. The cells were dispensed into 12-ml tubes containing 10 ml of leucine-free incubation medium and spun at 300 rpm for 2 min at room temperature. The supernatant was discarded and 0.2 ml of leucine-free incubation medium containing 60 pCi of r3H]leucine was added. The cells were incubated for 20 min in this medium, washed once in 5 ml of washing buffer without albumin and homogenized as described above in 0.2 ml of homogenization buffer. The cell extracts were stored at -80 "C until further analysis.
Determination of Amino Acid Incorporation into Total Cellular and Soluble Proteins and into P-enolpyruuate Carboxykinase-The incorporation of C3H]leucine into total cell protein was measured by trichloroacetic acid precipitation of a sample of homogenate, as described by Ballard et al. (19). The homogenate was then centrifuged at 111,OOO X g , , for 35 min at 2 "C. The particle-free supernatant was used for the assay of P-enolpyruvate carboxykinase activity (20-22), as well as for the determination of [3H]leucine incorporation into soluble proteins by trichloroacetic acid precipitation and into Penolpyruvate carboxykinase by direct immunoprecipitation. The latter procedure was performed using 0.4 ml of high speed supernatant after pulse-labeling by Method A or 0.15 ml of high speed supernatant after pulse-labeling by Method B. In the fvst case, the total amount of P-enolpyruvate carboxykinase in the antigen-antibody reaction was comprised between 60 and 100 milliunits, consisting of the endogenous enzyme present in the cell extract and, when required, of exogenous unlabeled P-enolpyruvate carboxykinase added in the form of rat liver cytosol. In the second case, the amount of endogenous Penolpyruvate carboxykinase was negligible and 60 milliunits of carrier enzyme were added. A specific antibody against cytosolic P-enolpyruvate carboxykinase (23) was added in 20% excess (in terms of precipitable milliunits) over the enzyme amount, so as to precipitate P-enolpyruvate carboxykinase quantitatively. The antigen-antibody reaction mixture, containing 0.6% (w/v) Triton X-405, was incubated at 37 "C for 30 min and 2 "C for 3 h. At the completion of the reaction, the mixture was dispensed on top of 0.5 ml of 1 M sucrose dissolved in 10 m~ Na phosphate buffer, pH 7.4, 150 mM NaCl, 5 m~ leucine, and 0.6% (w/v) Triton X-405. The immunoprecipitate was pelleted through the sucrose cushion by centrifugation at 1,750 X g,,, and 2 "C for 30 min. The pellet was washed three times by resuspension in the detergent-containing solution. The washed immunoprecipitate was submitted to electrophoresis in cylindrical SDS'-polyacrylamide gels as described previously (IO).
Slab gel electrophoresis analysis of pulse-labeled soluble proteins was performed according to the method of Laemmli (24). For these experiments, the cells were pulse-labeled by Method B. The slab gels were processed for fluorography as described (23).
The protein concentration in total homogenate and high speed supernatant was determined by the Lowry method (25). The total protein content of liver cells was 966 2 37 pg/lO' cells and the soluble Protein content 413 f 15 pg/lO' cells (means f S.E., 18 determinations).
Isolation of Total Cellular RNA-Fifty-five X IO' cells were used The abbreviation used is: SDS, sodium dodecyl sulfate.

P-enolpyruvate Carboxykinase mRNA in Hepatocytes
for each RNA extraction. A t appropriate times of incubation, the cells were pelleted by centrifugation at 300 rpm for 2 min in the bench top centrifuge and washed once in 50 ml of ice-cold, albumin-free washing of total cellular RNA, or alternatively was quickly frozen in liquid NS buffer. The washed cell pellet was used immediately for the extraction and stored at -80 "C until RNA isolation. The RNA was isolated by the guanidinium thiocyanate procedure, exactly as described by Chirgwin et al. (26), using initially 7 ml of the guanidinium thiocyanate solution for the homogenization of 55 X IO6 cells. The isolated RNA was precipitated twice from 0.2 M K acetate, pH 5.2, with two volumes of ethanol and finally dissolved in boiled deionized water at a concentration of 5 pg/pl. The RNA samples were stored at -80 "C. The yield of total RNA from 55 X IO6 cells was 995 f 25 pg (mean f S.E., 30 preparations).

Messenger RNA Translation Assay in the Wheat Germ Cell-free Protein Synthesis System-Total
cellular RNA was translated in a wheat germ cell-free protein synthesis system, using the incubation conditions described in detail previously (27,28). After the incubation, a high speed supernatant of the reaction mixture was obtained by centrifugation at 111,OOO X g-and 2 "C for 35 min. A sample of high speed supernatant was precipitated with trichloroacetic acid for the determination of [3H]leucine incorporation into total released polypeptides, which served as an estimate of the overall template activity of the RNA. The remainder of the high speed supernatant was used for the direct immunoprecipitation of P-enolpyruvate carboxykinase which was synthesized by the wheat germ extract at the direction of liver cell RNA. The immunoprecipitation was accomplished as described above with respect to the liver cell extracts. The amount of r3H]leucine incorporated into immunoprecipitable P-enolpyruvate carboxykinase served as an estimate of the concentration of functional mRNA coding for the enzyme in total cellular RNA. The incorporation of leucine into total polypeptides and into P-enolpyruvate carboxykinase was dependent on the RNA input within a range of concentrations comprised between 0 and 10 pg of total cellular RNA in a reaction volume of 80 pl. Routinely, the assay was performed using 6 or 9 pg of RNA.
Radioactivity Measurements-Protein precipitates were dissolved in NCS Tissue Solubiliier and the tritium radioactivity measured by liquid scintillation spectrometry with an efficiency of about 50%. Cylindrical SDS-polyacrylamide gels containing immunoprecipitated P-enolpyruvate carboxykinase were fractionated in 2-mm portions. The radioactivity in the gel fragments was measured after elution in the liquid scintillation mixture described in Ref. 29. Ninety-five % of the radioactivity was released from the gel pieces and counted at an efficiency of about 50%. Prior to fractionation, the gels were scanned spectrophometrically at 280 nm in order to localize the position of Penolpyruvate carboxykinase. The amount of r3H]leucine incorporated into P-enolpyruvate carboxykinase was computed by adding the radioactivity of the fractions corresponding to the enzyme band and subtracting a background estimated from the radioactivity of the neighboring fractions.

Viability and Protein Synthesis Capacity of Incubated
Liver Cells-Using the two-step collagenase perfusion method of Seglen (17) for the isolation of liver cells, we routinely obtained 500 X lo6 hepatocytes from the Liver of adult male rats weighing about 230 g. Immediately after isolation, 97 k 0.15% of the cells (mean f S.E., 10 experiments) were viable, as estimated by the trypan blue exclusion test. The apparent viability of the cells incubated in suspension remained excellent for several hours, 94.5 & 0.48% of the cells still excluding the dye after 5 h and 90.1 2 0.66% of the cells after 7 h of incubation. As others (30, 31), we noticed a tendency of suspended cells to aggregate during incubation. In general, this tendency became manifest around 5 h of incubation and resulted in the formation of small clusters (less than 15 cells), so that accurate cell counting was still possible at 7 h. Howevel, larger clumps preventing cell enumeration at the end of the incubation formed in about 16% of the experiments without obvious reason.
Consistent with the high viability figures, the capacity of the cells for protein synthesis was well preserved throughout the incubation. Cells were pulse-labeled with ['Hlleucine for 20 min periods starting at 15 min, 4.5 and 6.5 h of incubation, in order to determine the amino acid incorporation into total cellular proteins and into the soluble fraction of these proteins. As may be seen in the second and third columns of Table I, the cells maintained a nearly steady level of leucine incorporation during incubation. In subsequent series of experiments we sometimes noted small (less than 30%) but significant decreases in the total protein synthetic activity of the cells during the fist 4.5 h of incubation. In all cases, however, the rate of amino acid incorporation remained stable from 4.5 h until the end of the experiments.
In the same experiments, we also measured the incorporation of [3H]leucine into P-enolpyruvate carboxykinase by immunoprecipitation of the enzyme (Table I, fourth column).
From these data, the relative rate of P-enolpyruvate carboxykinase synthesis, i.e. the ratio of the radioactivity in the enzyme over the radioactivity in soluble protein, was computed. The data, expressed in ppm, are given in the last column of Table I. In contrast to the stability of total protein synthesis, the relative rate of P-enolpyruvate carboxykinase synthesis decreased 7-fold in 4.5 h of incubation and a further 2-fold during the next 2 h. This sharp decline in enzyme synthesis does not reflect a gross alteration in the pattern of protein synthesis in the hepatocyte, as shown by electrophoretic analysis of the total soluble proteins synthesized at various times of incubation ( Fig. 1, tracks 1-3). Within the limits of resolution of conventional SDS-polyacrylamide gel electrophoresis, there was no obvious qualitative change in protein synthesis during the course of a 6.5-h incubation. These data agree with earlier reports demonstrating the ability of liver cells in suspension culture to synthesize several plasma proteins at relatively constant rates for 12 h or more (30, 32).
Glucagon Effect on P-enolpyruvate Carboxykinase Synthesis-We reasoned that the extremely low rate of P-enolpyruvate carboxykinase synthesis in liver cells preincubated in hormone-free medium for a few hours would constitute an advantageous feature for enzyme induction studies. Glucagon was added at a concentration of M to cells preincubated for 4.5 h and the incubation was continued in the presence of the hormone for a further 2 h, after which the cells were pulselabeled with ['Hlleucine. Extracts of cells incubated with or without glucagon were reacted with the anti-P-enolpyruvate carboxykinase antibody and the immunoprecipitates were analyzed on SDS-polyacrylamide gels, as shown in Fig. 2.

Total protein synthesis and relative rate of P-enolpyruvate carboxykinase synthesis in isolated liver cells in suspension culture
Liver cells from fasted-refed rats were incubated as described under "Experimental Procedures." At the times indicated, cell samples were pulse-labeled for 20 min in the presence of [3H]leucine. Cell extracts were prepared for the determination of amino acid incorporation into total proteins. A high speed supernatant was obtained from these extracts for the measurement of amino acid incorporation into soluble proteins and into P-enolpyruvate carboxykinase. The initial labeling period was started 15 min after placing the cells in the incubator. Data are the means f S.E. for 9 different cells preparations. PEPCK, P-enolpvruvate carboxykinase. The hormonal effect was selective, as demonstrated by gel electrophoresis of the total complement of soluble proteins made by control or glucagon-treated cells (Fig. 1, tracks 3 and  4). There is no apparent difference due to glucagon in the synthesis rate of the proteins separated by this method. In such gels, P-enolpyruvate carboxykinase migrates with the -"  front of the prominent 70,000 molecular weight polypeptide, tentatively identified as pro-albumin, and is therefore masked by the latter.

45K-
The time-course of the glucagon-dependent induction of Penolpyruvate carboxykinase in preincubated liver cells is presented in Fig. 3. Glucagon was present a t a concentration of M and cells were pulse-labeled a t 40-min intervals after its addition. From a starting value of 428 ppm, the relative rate of enzyme synthesis rose approximately 7-fold to 2,723 ppm in 120 min and plateaued a t this level. A clear-cut increase in P-enolpyruvate carboxykinase synthesis was noticeable as early as 40 min following hormone addition. On the contrary, the rate of enzyme synthesis in control cells incubated without glucagon slowly declined to barely detectable levels by the end of the incubation. In both treated and control cells, the rate of amino acid incorporation into total protein remained unchanged throughout the experiment.
Dose-Response Relationship of Glucagon Effect-The next experiments were performed in order to investigate the concentration-dependence of the glucagon induction. As in previous experiments, the cells were preincubated without hormone for 4.5 h. Glucagon was then added to provide concentrations ranging from 4 X 10"' to 4 X M and the effect on P-enolpyruvate carboxykinase synthesis was measured 2 h later. It should be mentioned that glucagon is degraded by liver cells (33) and that the hormone concentration might well have decreased during the 2-h incubation. The data in Fig. 4 show that a 2-fold increase in the relative rate of enzyme synthesis occurred at a nominal glucagon concentration of 1.6 X lo-' M and that maximal induction took place with concentrations comprised between and 4 x 10" M. Total protein synthesis was not affected by glucagon at any concentration.
In order to assess the presumed role of cyclic AMP as a mediator of the inductive effect of glucagon, the cells were also challenged with the hormone in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine. As may be seen in Fig. 4, the latter compound strongly potentiated the effect of submaximal concentrations of glucagon.
Cyclic AMP Effect on P-enolpyruvate Carboxykinase Synthesis-In preliminary experiments dealing with the direct effect of cyclic AMP added externally to liver cells, we found that concentrations of 0.3 m~ cyclic AMP and 0.5 mM 3isobutyl-1-methylxanthine would elicit a maximal induction of P-enolpyruvate carboxykinase synthesis in preincubated liver cells. Cyclic AMP at 1 m~ or higher, on the other hand, caused a consistent inhibition of total protein synthesis, accompanied by a decrease in the relative rate of P-enolpyruvate carboxykinase compared to the maximally induced rate (data not shown). Fig. 5 shows the kinetics of induction occurring when maximal concentrations of cyclic AMP and 3-isobutyl-1-methylxanthine were provided to liver cells 5.5 h after isolation. From a value of 332 ppm at the time of inducer addition, the relative rate of enzyme synthesis reached 3,980 ppm in 2 h. In these experiments, the time course of action of the cyclic nucleotide was somewhat delayed relative to the glucagon effect (compare Figs. 5 and 3). The short lag might have been due to a diffusion barrier to the penetration of the nucleotide in the liver cell. We determined that the maximal increase in intracellular cyclic AMP following the addition of glucagon occurred in 4 to 5 min (experiments not shown). It is conceivable that the build-up of an effective intracellular cyclic AMP level would take more time after the external addition of the nucleotide. After 2 h, however, the cells exposed to cyclic AMP had a rate of enzyme synthesis identical to cells incubated in the presence of 10" M glucagon (Fig. 5).
Glucocorticoid-Glucagon Interaction in the Regulation of Hepatic P-enolpyruvate Carboxykinase-The glucocorticoids play a complex role in the regulation of P-enolpyruvate carboxykinase synthesis in the whole animal. Whereas they induce the enzyme in the kidney (34) as a consequence of an increase in specific P-enolpyruvate carboxykinase mRNA (23,28,29), they are ineffective to stimulate the synthesis of hepatic P-enolpyruvate carboxykinase in the intact rat, possibly because of the inhibitory effect of insulin released after glucocorticoid administration (35). Taking advantage of our hormone-responsive in vitro system, we attempted to clarify the role of the glucocorticoids in the regulation of hepatic Penolpyruvate carboxykinase. We f i s t asked whether the presence of a glucocorticoid in the incubation medium would prevent the fall of P-enolpyruvate carboxykinase synthesis normally seen in hepatocytes maintained in vitro. Freshly isolated liver cells were incubated with 1O"j M dexamethasone or, as before, in hormone-free medium. After 4.5 h, they were pulse-labeled in order to measure enzyme synthesis. From an initial relative rate of 2362 436 ppm, P-enolpyruvate carboxykinase synthesis dropped to 450 f 183 ppm (mean ? S.E., 5 experiments) in the presence of dexamethasone, not significantly different from 224 k 101 ppm in control cells. Thus, the glucocorticoid was clearly unable to sustain P-enolpyruvate carboxykinase synthesis at the level prevailing at the start of the incubation.
The second question was to know whether cells preincubated with dexamethasone would exhibit a stronger inductive response to glucagon. That this was indeed the case is shown in Table 11. Glucagon caused a 13-fold stimulation of P-enolpyruvate carboxykinase in control cells and a 43-fold stimulation in cells incubated with dexamethasone. The amplification of the response to glucagon due to dexamethasone was, therefore, approximately 3-fold (43 + 13). The relative rate of enzyme synthesis in cells induced by glucagon in the presence of dexamethasone is comparable to the maximal rate seen in

I1
Effect of dexamethasone on glucagon-dependent induction of P-enolpyruvate carboxykinase Liver cells from intact, fasted, glucose-refed (cf. Table 111) rats were preincubated for 4.5 h in hormone-free medium or in the presence of dexamethasone M). Glucagon M) or vehicle were then added and the incubation continued for an experimental period of 2 h. Samples of cells were pulse-labeled with [3H]leucine after the preincubation period (0 time) and at the end of the experimental period (2 h). The cells were processed for the determination of radioactive leucine incorporation into total soluble proteins and into P-enolpyruvate carboxykinase as described under "Experimental Procedures." The data are the means f S.E. of 5 experiments. PEPCK, P-enolpyruvate carboxykinase.
vivo in the liver of fasted rats, which is between 20,000 and 30,000 ppm (36).
A further question which arose then was whether glucagon by itself was at all able to stimulate the synthesis of Penolpyruvate carboxykinase in the complete absence of glucocorticoids. It was possible that the effect of glucagon seen after preincubation in hormone-free medium was promoted by residual glucocorticoid complexed to its receptor inside the hepatocyte. To elucidate this point, we performed experiments using liver cells isolated from rats which had been adrenalectomized 3 to 5 days prior to the experiments. These cells appeared to be more fragde than normal ones, only 81.2 k 3.1% of the cells (mean k S.E., 6 experiments) excluding trypan blue after 7 h of incubation instead of the usual 90%. The rate of amino acid incorporation into protein, however, was the same as in normal cells and, as in the latter, it was not affected by any of the hormonal treatments. Glucagon added to these cells following a preincubation in hormone-free medium caused a 2.8-fold stimulation of P-enolpyruvate carboxykinase synthesis in 2 h (Fig. 6 ) . Although more modest than in normal cells, the induction was nevertheless obvious in all the experiments. When the cells were primed with dexamethasone, glucagon produced an %fold increase in the relative rate of P-enolpyruvate carboxykinase synthesis, corresponding to a 6-fold enhancement over the effect of glucagon acting alone (18 + 2.8). As in normal cells, the glucocorticoid by itself was devoid of any marked effect on P-enolpyruvate carboxykinase synthesis.

Effects of Glucagon and Dexamethasone on P-enolpyruvate
Carboxykinase mRNA-Several studies in the whole rat have shown that an injection of dibutyryl cyclic AMP results in the build-up of P-enolpyruvate carboxykinase mRNA in the liver, supporting the view that the cyclic AMP-mediated induction of the enzyme is achieved by a transcriptional or post-transcriptional mechanism (10-13). Earlier in vitro experiments using Reuber H-35 hepatoma cells as a model system suggested, however, that cyclic AMP might also stimulate the synthesis of P-enolpyruvate carboxykinase by a translational mechanism because RNA synthesis inhibitors appeared not to prevent enzyme induction (37,38). To explain

TABLE I11
Effect of dexamethasone on glucagon-dependent induction of P-enolpyruvate carboxykinase mRNA Liver cells from intact, fasted-glucose refed rats were preincubated for 5 h either in hormone-free medium or in the presence of dexamethasone M). Glucagon (lo-' M ) or vehicle.were then added and the incubation was continued for 2 h. At the end of this period, total cellular RNA was isolated as described under "Experimental Procedures." Translation assay of the RNA was performed in the wheat germ cell-free protein synthesis system using 6 pg of RNA. The high speed supernatant of the translation mixture was used to measure the [3H]leucine incorporation into total released polypeptides and into P-enolpyruvate carboxykinase. The total incorporation values were corrected by subtracting the background incorporation due to the translation of mRNA endogenous to the wheat germ extract. The data are for a suuernatant volume of 130 ~1 .
Values are means f S.D. of three experirrmts. PEPCK, P-enolpyrkate carboxykinase. the synergism between dexamethasone and glucagon observed in the present experiments, we reasoned that dexamethasone might induce the accumulation of P-enolpyruvate carboxykinase mRNA, but that this mRNA could not be translated efficiently in the cell in the absence of a stimulus capable of raising the intracellular concentration of cyclic AMP. This hypothesis was tested by measuring the level of P-enolpyruvate carboxykinase mRNA in cells preincubated for 5 h with or without dexamethasone M) and subsequently challenged with glucagon ( M) for a period of 2 h. Total cellular RNA was isolated from such cells and assayed in a wheat germ translation system for its capacity to direct the incorporation of amino acids into total polypeptides and specifically into P-enolpyruvate carboxykinase. The level of functional Penolpyruvate carboxykinase mRNA was expressed as the ratio of mRNA-directed leucine incorporation into enzyme over mRNA-directed leucine incorporation into total products (Table 111). Three important results emerge from these experiments. First, glucagon caused a 10-fold increase in the level of P-enolpyruvate carboxykinase mRNA in liver cells preincubated in hormone-devoid medium. Second, the level of enzyme mRNA was not different in cells incubated with or without dexamethasone. Third, P-enolpyruvate carboxykinase mRNA was induced 30-fold by glucagon in the presence of dexamethasone, as compared to 10-fold in its absence. Thus, the glucocorticoid amplied the induction of the enzyme mRNA by a factor of 3. Since the glucagon-dependent effects on the level of P-enolpyruvate carboxykinase mRNA were quantitatively similar to the hormonally induced alterations in enzyme synthesis reported in Table 11, we concluded that enzyme induction was not achieved by a translational mechanism, neither in the absence of dexamethasone nor after preincubation with the glucocorticoid.
The next experiments were performed to determine whether changes in the rate of enzyme synthesis could be accounted for by equivalent changes in mRNA level throughout the induction time-course. The kinetics of mRNA accumulation was established in cells induced by glucagon or cyclic AMP after preincubation in the presence of dexamethasone.
At all time points, the ongoing rate of P-enolpyruvate car- boxykinase synthesis in the cells was also measured. The induction of P-enolpyruvate carboxykinase mRNA and the stimulation of enzyme synthesis by glucagon are shown in Fig.

, A and B.
Corresponding data for the effect of cyclic AMP plus 3-isobutyl-1-methylxanthine are displayed in C and D.
With both glucagon and cyclic AMP and over the entire experimental period, the increase in functional mRNA level was adequate to explain the coincident stimulation of enzyme synthesis, in agreement with a transcriptional or post-transcriptional mechanism of enzyme induction. In these experiments, the action of exogenous cyclic AMP was as rapid as that of glucagon, whereas it appeared slightly slower when the cells were not preincubated with dexamethasone (compare Figs. 7 and 5). A possible explanation is that incubation in the presence of dexamethasone somehow results in a better penetration of the cyclic nucleotide through the liver cell membrane.

DISCUSSION
Isolated liver cells in suspension culture for 7 to 8 h provide an excellent experimental system for studying acute hormonal effects on the synthesis of hepatic P-enolpyruvate carboxykinase. An attractive feature of the system is the very low rate of enzyme synthesis in cells preincubated in hormone-devoid medium for a few hours. In the adult rat, the rate of hepatic P-enolpyruvate carboxykinase synthesis is lowest during refeeding after a fasting period. Under these circumstances, the values reported are in the range of 2000 to 4000 ppm (10,36), similar to those measured in liver cells immediately after cell isolation (Table I). Our data demonstrate that the relative rate of enzyme synthesis is maintained at this level only as long as the liver is exposed to the internal milieu of the whole animal. Removal from this milieu results in a further decline of P-enolpyruvate carboxykinase synthesis. When superimposed on the extremely low basal rate of enzyme synthesis prevailing after a few hours of preincubation, hormonal effects are magnified relative to those seen in vivo in the liver of the whole animal.
Glucagon causes a rapid, substantial, and consistent stimulation of the synthesis of P-enolpyruvate carboxykinase in isolated, preincubated liver cells. In several series of experiments, the rate of enzyme synthesis increased between 6-and 13-fold in 2 h following the addition of glucagon at

M.
Effective hormone concentration ranged from IO-' to 4 x M. A similar dose-response relationship was reported by Gurr and Potter (39) for the glucagon-dependent induction of tyrosine aminotransferase in hepatocyte suspensions or monolayers. Another response to glucagon, the induction of an amino acid transport system, however, was shown to occur at lower hormone concentrations, the maximal effect being observed around lo-' M glucagon (40,41). The nutritional status of the donor animals, the collagenase used for cell isolation, and the incubation conditions may be responsible for such differences. It is interesting to note that, in spite of the large effect of glucagon on P-enolpyruvate carboxykinase synthesis, we did not observe increases in enzyme activity in induced cells. This apparent discrepancy is not surprising in view of: (i) the relatively long half-life of hepatic P-enolpyruvate carboxykinase (about 6 h, Ref. 36), (ii) the large pool of preformed enzyme in cells freshly isolated from fasted-refed rats, and (iii) the very low ongoing rate of enzyme synthesis in these cells. Under these circumstances, even considerable increases of enzyme synthesis would not be expected to result in any substantial changes in the assayable level of P-enolpyruvate carboxykinase over short periods of time. The same argument probably also explains why Oliver et al. (42) did not observe acute effects of glucagon on P-enolpyruvate carboxykinase activity in freshly plated hepatocyte monolayers, while they and other authors (43) reported a rapid rise in enzyme level in older cultures with a much reduced intracellular enzyme pool.
Two lines of evidence support the contention that cyclic AMP is the mediator of the glucagon induction. First, the inhibitor of phosphodiesterase, 3-isobutyl-l-methylxanthine, strongly potentiated the hormonal effect. The half-maximum concentration of glucagon was shifted leftward from approximately 2.5 X lo-@ to 4 X M in the presence of the phosphodiesterase inhibitor. Second, cyclic AMP itself was able to induce the synthesis of P-enolpyruvate carboxykinase directly. At maximal concentrations, cyclic AMP and glucagon were equieffective inducers.
Glucagon, acting via cyclic AMP, induces the synthesis of P-enolpyruvate carboxykinase in the liver cell as the result of an increase in enzyme mRNA level. In the present experiments, we noted that the stimulation of enzyme synthesis caused by glucagon or directly by cyclic AMP was always accompanied by a commensurate increase in the level of functional P-enolpyruvate carboxykinase mRNA, as measured by translation assay. Previous in vivo studies from several laboratories, using cDNA hybridization as well as translation assays, have shown that the mRNA coding for Penolpyruvate carboxykinase accumulates in the rat liver following an injection of dibutyryl cyclic AMP, concomitantly with the induction of enzyme synthesis (10)(11)(12)(13)44,45). Taken together, these data discount the possibility of a translational type of induction mechanism, which would call for effects on P-enolpyruvate carboxykinase synthesis not paralleled by equivalent changes in specific mRNA level.
An important objective of this work was to clarify the role of the glucocorticoids in the regulation of hepatic P-enolpyruvate carboxykinase synthesis. In the intact rat, the enzyme is not induced after an injection of hydrocortisone or triamcinolone. There is, however, a slight stimulation of the already enhanced rate of P-enolpyruvate carboxykinase synthesis in the diabetic animal following glucocorticoid administration (35). These observations suggested that the glucocorticoid effect on P-enolpyruvate carboxykinase was normally offset by a rise of the plasma insulin level (35). The present data show that the glucocorticoids by themselves are not effective inducers of P-enolpyruvate carboxykinase synthesis in the rat liver cell. The level of mRNA encoding the enzyme does not increase in hepatocytes incubated for several hours in the presence of dexamethasone. While ineffective when acting alone, the glucocorticoids nevertheless play an important regulatory role for the expression of P-enolpyruvate carboxykinase mRNA. This role becomes manifest after addition of glucagon or cyclic AMP. Under the concerted actions of a glucocorticoid and of glucagon or its second messenger, the level of enzyme mRNA and the relative rate of enzyme synthesis increase 30-to 50-fold. The rate of enzyme synthesis observed in vitro is then at least equal to the fully induced rate measured in vivo in the liver of fasting rats (36). In the absence of dexamethasone, on the contrary, glucagon causes a more limited accumulation of mRNA and the rate of Penolpyruvate carboxykinase synthesis remains far lower than the induced rate in the liver of the starved animal. Recently, Kletzien et al. (46) reported data on the regulation of Penolpyruvate carboxykinase activity in hepatocytes maintained in monolayer culture which are entirely consistent with the hormonal effects on enzyme mRNA level and rate of synthesis described here. Their data and ours demonstrate the permissive role of the glucocorticoids in the regulation of hepatic cytosolic P-enolpyruvate carboxykinase.
The molecular basis for the glucocorticoid-cyclic AMP interaction in the regulation of P-enolpyruvate carboxykinase mRNA in the liver remains to be elucidated. The hormonally induced accumulation of functional mRNA may result from a stimulation of the transcription of the structural gene, a stimulation of the processing of the primary transcript into mature mRNA, a stabilization of the mRNA against degradation, or from a combination of several of these mechanisms. The glucocorticoids are thought to regulate the transcription of specific genes, probably following the interaction of the hormone-receptor complex with DNA sequences in the vicinity of the concerned genes (47,48). Cyclic AMP is also involved in the regulation of gene expression in eucaryotic cells by influencing the synthesis of specific mRNAs (49,50). It will be of considerable interest to see whether the two co-inducers of hepatic P-enoipyruvate carboxykinase mRNA act cooperatively to enhance the rate of transcription of the enzyme gene, or whether each one acts at a different step of the gene expression pathway.