Acute Stimulation by Glucocorticoids of Gluconeogenesis from Lactate/ Pyruvate in Isolated Hepatocytes from Normal and Adrenalectomized Rats*

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY hy The American Society of Biological Chemists, Inc. VoI. 260, No. 23, Issue of October 15, pp ,1985 Printed in U.S.A. Acute Stimulation by Glucocorticoids of Gluconeogenesis from Lactate/ Pyruvate in Isolated Hepatocytes from Normal and Adrenalectomized Rats* Frank D. SistareS and Robert C. Haynes, Jr.4 From the Department of Pharmwology, Uniuersity of Virginia, Charlottesuilk, Virginia (Received for publication, January 17,1985) Dexamethasone stimulated gluconeogenesis from precursors and not to an inhibition of utilization. While much lactatelpyruvate in suspensions of hepatocytes isolated of the glucocorticoid effect might be attributable to mobilifrom both adrenalectomized and normal fasted rats. zation of glucogenic amino acids from peripheral tissues and This stimulation was observed in incubations with 1 a secondary rise in levels of insulin, actions of glucocorticoids mm pyruvate and at a lactatelpyruvate ratio of 25 but in vitro on liver to stimulate gluconeogenesis have been demnot at a ratio of At a lactatelpyruvate ratio of onstrated. Haynes (3, 4), Uete and Ashmore (5), and Eisen , the stimulation by dexamethasone was pro- stein et al. (6) have demonstrated direct stimulations of glugressively enhanced as the pyruvate concentration was coneogenesis following glucocorticoid addition to rat liver decreased to 0.25 mm. Concurrent administration of a maximally stimulating concentration of dexamethasone with angiotensin I1 or glucagon yielded an additive stimulation at all concentrations of the peptide hormones tested. No potentiating or permissive actions of acute glucocorticoid administration were observed using hepatocytes from either normal or adrenalectomized animals. The acute stimulation by dexamethasone was antagonized by prior addition of progesterone or cortexolone to the hepatocyte suspensions. Triamcinolone and corticosterone also stimulated gluconeogenesis. Concentrations of the active glucocorticoids needed to elicit half-maximal stimulations (K,J were approximately 100 nm for dexamethasone and triamcinolone and 400 nm for corticosterone. Deoxycorticosterone, 17a-methyltestosterone, and 5&d1hydrocortisol did not stimulate. Stimulation of gluconeogenesis by dexamethasone was seen following a lag averaging 9 min after the time of steroid addition. Preliminary evidence suggests that this effect was not dependent upon a stimulation of protein synthesis, but the observed stimulation and inhibition of control rates of gluconeogenesis by cycloheximide and cordycepin, respectively, demonstrate the difficulties of working with such inhibitors in attempting to answer this question. Glucocorticoids have long been known to regulate glucose homeostasis. Long et al, (1) were first to show that administration of adrenal cortex extracts to fasted adrenalectomized rats leads to increased blood glucose levels and liver glycogen deposition. Welt et al. (2) later demonstrated that this effect was clearly attributable to a stimulation of production from * This research was supported by Grant 5R01 AM from the National Institutes of Health, United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $Present address: Building 4, Room 134, National Institutes of Health, Bethesda, MD To whom correspondence should be addressed. slices. Comparing gluconeogenesis in perfused livers from control and adrenalectomized rats, Friedmann et al. (7), and later Exton et al. (8, 9) proposed that glucocorticoids play a permissive role in the hormonal regulation of gluconeogenesis. Livers from adrenalectomized rats, shown to be less responsive to glucagon and epinephrine, became fully responsive within 60 min of dexamethasone addition to the liver perfusions. Dexamethasone alone caused no significant stimulation in those studies. The defect in adrenalectomized livers was not due to an impairment in stimulation of CAMP levels or activation of CAMP-dependent protein kinase. Recently, Friedmann (10) has pointed out, however, that the nature of the defect following adrenalectomy is elusive and not easily reproducible. Eigler et al. (11) have shown that a physiologically relevant increment in normal cortisol serum levels, which alone does not alter rates of glucose production can potentiate the effects of physiological elevations of glucagon or epinephrine in the normal dog. Recently, Allan and Titheradge (12) reported that dexamethasone administered in vivo to fed rats and maintained in vitro in incubations of subsequently isolated hepatocytes stimulated gluconeogenesis and increased the sensitivity of gluconeogenesis to stimulation by glucagon and epinephrine. The objectives of the present investigation, then, were to reproduce and further study any acute direct, permissive, and potentiating actions following in vitro glucocorticoid addition to suspensions of isolated hepatocytes from 24-h fasted rats. EXPERIMENTAL PROCEDURES Methods-Hepatocytes were isolated essentially as described by Seglen (13) from male Wistar rats ( g) starved for h. Isolated hepatocytes were incubated at a concentration of 0.5 mg of cell protein/ml in a modified Krebs-Ringer bicarbonate buffer (40.7 mm NaCl, 4.7 mm KC1,2.5mMCaCl2, 1.2 mm MgSOI, 1.2 mm KH2P04, 100 mm TES, 22.8 mm NaHCOa, ph 7.4, at 37 c) with 95% 02, 5% Con as the gas phase. Substrates were lactate and pyruvate adjusted to provide specific concentrations during the period in which glucose formation was measured. At designated time points, aliquots of the cell suspensions were removed, extracted with cold 0.3 M perchloric acid, and neutralized The abbreviations used are: TES, N-tris[hydroxymethyl]methyl- 2-aminoethanesulfonic acid; ACTH, adrenocorticotropic hormone.

2 Acute Glucocorticoid Stimulation of Gluconeogenesis with KHC03. Extracellular lactate, pyruvate, and glucose were determined by enzyme-linked fluorometric procedures (14). Protein was determined as described by Lowry et al. (15) using crystallized bovine serumalbuminas a standard.rates ofglucoseproductionwere calculated using linear regression analysis of measurements from samples taken at 60, 80, 100, and 120 min after the start of each incubation. Lactate and pyruvate concentrations were averaged from measurements of samples taken at 80 and 100 min. At the time of liver perfusion, blood was collected from the inferior vena cava which was cut just prior to the initial flushing of the liver, andserumcorticosteroneconcentrationsweredeterminedfluorometrically according to the method of Guillemin et al. (16). Rats were adrenalectomized under pentobarbital anesthesia, using a dorsal approach, 5-7 days prior to hepatocyte isolation and maintained on 0.9% saline ad libitum. Materials-Hexokinaseandglucose-6-phosphatedehydrogenase were obtained from Boehringer Mannheim. Crystallized, lyophilized -201 bovine serum albumin was from Miles Laboratories. Dexamethasone, 8r - I1 - T E glucagon,corticosterone,triamcinolone,progesterone,cortexolone,.- 0 c deoxycorticosterone, 17a-methyltestosterone, 5P-dihydrocortisol, col- *A lagenase (type V, lot 33F-6817), and lactate dehydrogenase (from beef 4.E 6- heart) were from Sigma. [Asp, Ile5]Angiotensin I1 (human) octapep- q tide was from the United States Biochemical Corp.,a 0, RESULTS Effects of Adrenalectomy on Hormonal Stimulation of Glu- coneogenesis-for these studies, substrate concentrations Q) + were carefully manipulated. The extracellular pyruvate con- OL centration was maintained at 1 mm, and the lactate/pyruvate CDGA ratio was fixed at 0.6, 7, and 25. The effects of additions of NL ADX NL ADX NL AOX maximal concentrations of glucagon, angiotensin 11, or dexa methasone to suspensions of hepatocytes from normal and Ratio of Lactate: Pyruvate adrenalectomized animals were compared. FIG. 1. Effects of dexamethasone, glucagon, and angioten- Serum corticosterone concentrations measured 5-7 days sin I1 on glucose production rates in hepatocytes from normal and adrenalectomized rats. Preparation and incubation of hepafollowing adrenalectomy were approximately 5% of normal. tocytes from normal and adrenalectomized rats and measurement of (Adrenalectomized rats had a mean concentration of 1.5 & glucose, lactate, and pyruvate were as described under Experimental 0.1 pg/100 ml compared to normal values of 33 & 3 pg/loo Procedures. Hepatocytes from normal animals (NL, solid symbols, ml.) This low level of fluorescence confirms that adrenals and solid lines) or animals adrenalectomized for 5-7 days (ADX, open were completely removed, since the trace of residual fluores- symbols, and broken lines) were incubated at a concentration of 0.5 cence remaining after adrenalectomy is not believed to repmgof protein/ml at 37 C under 95% 02, 5% C02 with 2.1 mm pyruvate only; 6.75 mm lactate, 1.8 mm pyruvate; or 25 mm lactate, 0.65 mm resent adrenocortical steroid (17). Fig. 1 shows the depend- pyruvate, which yielded approximate lactate/pyruvate ratios of 0.6, ence of gluconeogenesis on the lactate/pyruvate ratio as de- 7, or 25,., respectively, all at a pyruvate concentration of approximately scribed in the preceding paper (18) and, furthermore, that 1 mm (average of values at 80 and 100 min). Dexamethasone was adrenalectomy did not diminish control rates of gluconeo- dissolved in methanol and evaporated to dryness on the bottom of a genesis from 1 mm pyruvate at any of the three lactatel flask prior to addition of buffer to yield a final concentration of 50 gm pyruvate ratios examined. The effects of maximally stimula- (D, 0). At 60 min, glucagon (100 nm) (G, 0, O), angiotensin I1 (10 nm) (A, A, A), or vehicle (C) was added. For the experiments tory concentrations of hormones on hepatocytes from normal using hepatocytes from adrenalectomized rats, each point and bar and adrenalectomized animals are also presented in Fig. 1. At represent the mean f. S.E. of three experiments. For experiments the lowest lactate/pyruvate ratio, 0.6 (where cells were incu- with normal hepatocytes, n = 5-7. bated starting with pyruvate only, and lactate levels rose to approximate this ratio between 80 and 100 min), glucagon administration inhibited glucose production by approximately 20% in hepatocytes from both normal and adrenalectomized rats. Angiotensin I1 and dexamethasone each stimulated glu- cose production by approximately 20% in both groups of hepatocytes. At the lactate/pyruvate ratio of 7, dexamethasone had only a marginal effect in either group. Mean stimulations by angiotensin I1 and glucagon were reduced, however, following adrenalectomy. At the lactate/pyruvate ratio of 25, the dexamethasone stimulation in hepatocytes from normal and adrenalectomized rats reached 30-40%. The mean stimulations due to angiotensin TI and glucagon were again dimin- ished 30-50% by adrenalectomy, consistent with the previous reports (7-9) that adrenalectomy results in reduced hormonal responsiveness. This apparent effect of chronic glucocorticoid withdrawal was originally pursued in these studies because of the reported ability of dexamethasone to reverse acutely the effects of adrenalectomy. The effect of combining dexamethasone with either glucagon or angiotensin TI was therefore s3 $ E 2-0-.; 4- examined in these same experiments. As shown in Fig.2, when dexamethasone was added to hepatocytes from adrenalectomized rats 60 min prior to angiotensin I1 or glucagon, the glucagon inhibition at the lactate/pyruvate ratio of 0.6 was not enhanced. The glucagon and angiotensin I1 stimulations seen at the lactate/pyruvate ratio of 25 were approximately additive with the dexamethasone stimulation seen at this lactate/pyruvate ratio. At the lactate/pyruvate ratio of 7, where little or no dexamethasone stimulation was noted, still no potentiation or permissive return to full responsiveness was found. Therefore, no acute permissive or potentiating effect of dexamethasone was observed at any lactate/pyruvate ratio tested with a pyruvate concentration of 1 mm. However, as was also reported in the preceding paper (18), a direct stimulation of gluconeogenesis by glucocorticoids occurs at very low and very high lactate/pyruvate ratios. At nearphysiological lactate/pyruvate ratios when the pyruvate concentration is 1 mm, little glucocorticoid stimulation is seen. Explored next were the possibilities that permissive or It

3 12756 Acute Glucocorticoid Stimulation of Gluconeogenesis Or I -201 Ratio of Lactate : Pyruvate FIG. 2. Effects of dexamethasone in combination with maximally stimulating concentrations of glucagon and angiotensin I1 in hepatocytes from adrenalectomized rats. Preparation and incubation of hepatocytes and measurement of glucose, lactate, and pyruvate were as described under Experimental Procedures and in the legend to Fig. 1. The pyruvate concentration was 1 mm. Dexamethasone (50 p ~ final, concentration) dissolved in methanol (solid lines and filled symbols) or methanol alone (broken lines and open symbols) was evaporated to dryness onto the bottom of flasks prior to addition of buffer. At 60 min, vehicle (H), glucagon (100 nm) (0, 0), or angiotensin I1 (10 nm) (A, A) was added to the dexamethasone and control flasks. Each point and bar represent the mean f S.E. of three experiments. potentiating actions of glucocorticoids might be revealed only at lower, near-physiological substrate concentrations or that glucocorticoids might only potentiate subsaturating concentrations of angiotensin I1 or glucagon. Fig. 3 compares the dose-response curves for angiotensin I1 and glucagon stimulation of gluconeogenesis from mm pyruvate at a lactate/pyruvate ratio of in hepatocytes from normal and adrenalectomized rats. Here, a 25-35% reduction in maximal response was noted following adrenalectomy, but there was no loss in sensitivity to either hormone. Glucagon s K,,, was approximately 2-3nM, and angiotensin 11 s Kact was approximately 100 PM in hepatocytes from both normal and adrenalectomized animals. Furthermore, as shown in Fig. 4, the concurrent presence of 50 p~ dexamethasone in these experiments with hepatocytes from adrenalectomized rats did not potentiate the action of any submaximal concentration of angiotensin I1 or glucagon. The stimulation due to dexamethasone was essentially additive to the stimulations seen at each concentration of angiotensin I1 or glucagon. While stimulation of gluconeogenesis by maximal doses of angiotensin I1 or glucagon might be diminished following adrenalectomy, no evidence for an acute potentiating or permissive role of dexamethasone was found using this hepatocyte suspension system. Since a sizable direct, acute dexamethasone effect was observed in hepatocytes from both normal and adrenalecto- mized rats at high lactate/pyruvate ratios and also at low substrate concentrations with physiological lactate/pyruvate ratios, these effects were explored further in hepatocytes from normal rats, while investigations into the effects of adrenalectomy were abandoned. Effect of Substrate Concentration on Acute Stimulation of Gluconeogenesis by Glucocorticoids-In the experiments of Fig. 5, the mean lactate/pyruvate ratio was maintained at and the lactate and pyruvate concentrations were adjusted so that the mean pyruvate concentration measured at 80 and 100 min ranged between 0.2 and 2.0mM. Dexamethasone Ot b/ I I I Log [Hormone] (molar) FIG. 3. Sensitivity of glucagon and angiotensin I1 stimulation of gluconeogenesis in hepatocytes from normal and adrenalectomized rats. Measurement of glucose production rates and the preparation and incubation of hepatocytes from normal rats (solid lines and solid symbols) and rats adrenalectomized for 5-7 days (broken lines and open symbols) were as described under Experimental Procedures. Hepatocytes were incubated at 0.5 mg of protein/ml at 37 C under 95% 02, 5% CO, with 3.2 mm lactate only as initial substrate. This resulted in pyruvate concentrations of mm and mean lactate/pyruvate ratios of during the period in which glucose production rates were determined. Vehicle, glucagon (0, 0), or angiotensin I1 (A, A) was added at 60 min to yield the final concentrations indicated. Glucose production rates are expressed as a percentage of stimulation over control rates. Each point represents the mean f S.E. of three to four experiments. produced a mean 40% stimulation of gluconeogenesis at the lowest pyruvate concentration tested, and the effect steadily decreased to a mean of 10% in this series of experiments at saturating pyruvate levels. Dexamethasone did not act by increasing the concentration of pyruvate and lowering the lactatelpyruvate ratio, since after 80 and 100 min of exposure to dexamethasone, measured pyruvate levels were not significantly higher than, nor were lactate/pyruvate ratios significantly lower than, controls (p > 0.4). Steroid Specificity of the Acute Stimulation of Gluconeo- genesis by Glucocorticoids-To assess further the physiological relevance of the acute stimulation of gluconeogenesis by glucocorticoids, the in vitro potency was compared to known circulating levels of adrenal steroids in the rat, and steroid the specificity of the effect was examined. Since, as noted, high lactate/pyruvate ratios and low pyruvate levels enhanced the effects of dexamethasone, in this series of experiments, cells were initially incubated with 3.6 mm lactate only, and 60 min were allowed for pyruvate levels to build before samples were taken for glucose measurement. The sensitivity of the gluconeogenic response to the glucocorticoids corticosterone, dexamethasone, and triamcinolone was tested in the experiments summarized in Fig. 6. Dexamethasone and triamcinolone appeared to be equipotent, both stimulating half-maximally at about 100 nm. Corticosterone was found to be less potent, exhibiting a K,,, of approximately 400 nm. Lactate and pyruvate concentrations, measured at 80 and 100 min following the start of the incubations and averaged to extrapolate to and represent the midpoint of the 60- min interval during which glucose production rates were determined, again indicated no significant differences in pyr-

4 Acute Glucocorticoid Stimulation of Gluconeogenesis Log [An 01 Glucagon] (molar) FIG. 4. Sensitivity of gluconeogenesis in hepatocytes from adrenalectomized rats to stimulation by glucagon and angiotensin I1 in the presence or absence of high concentrations of dexamethasone. Conditions were the same as those described in the legend to Fig. 3. Dexamethasone (to provide 50 pm concentration in the incubation medium) dissolved in methanol (solid lines and solid symbols) or methanol alone (broken lines and open symbols) was evaporated to dryness in flasks prior to addition of buffer. After 60 min, vehicle or the indicated concentrations of glucagon (0, 0) or angiotensin I1 (A, A) were added. Glucose production rates are expressed as a percentage stimulation over control rates. Each point represents the mean f S.E. of three to four experiments. i:il 60.= 30 rn z I I I Pyruvate (mmolar) FIG. 5. Effect of pyruvate concentration on dexamethasone stimulation of glucose production. Preparation and incubation of hepatocytes and measurement of lactate, pyruvate, and glucose production rates were as described under Experimental Procedures. Hepatocytes (0.5 mg of protein/ml) were incubated at 37 C under 95% 02, 5% COZ with varying substrate concentrations, all of which yielded a mean lactate/pyruvate ratio of as determined from the average of measurements of samples taken at 80 and 100 min. Dexamethasone (50 WM), where indicated (W), was present at the beginning of the incubation (dissolved in methanol and evaporated to dryness prior to addition of buffer). Glucose production rates are expressed as percentage stimulations over control rates. Each point represents the mean f S.E. of five experiments. uvate concentrations or lactate/pyruvate ratios between control incubations and those with glucocorticoids (20 PM) present ( p > 0.2). In the experiments of Table I, the experimental design was altered slightly. Nonglucocorticoids and antiglucocorticoids were dissolved in methanol and evaporated to dryness in each flask as described for the previous experiments. After allowing a 40-min incubation to build pyruvate concentrations, a plas- ~~ 2x10-102x10-9 2x10-* ~XIO-~ 2x10-6 2~10-5 2x10-4 [Glucocorticoid] (molar) FIG. 6. Dose-response relationship of acute glucocorticoid stimulation of gluconeogenesis. Preparation and incubation of hepatocytes and measurement of glucose production rates were as described under Experimental Procedures. Hepatocytes (0.5 mg protein/ml) were incubated at 37 C under 95% 9,5% COZ with 3.6 mm lactate only as initial substrate. The desired quantity of corticosterone (A), triamcinolone (O), or dexamethasone (W) was dissolved in methanol and evaporated to dryness in each flask before addition of buffer. Control flasks received the same volume of methanol evaporated to dryness. Glucose production rates are expressed as the percentage of stimulation over control production rates. Each point represents the mean * S.E. of four to five experiments. tic tab, onto which dexamethasone in methanol had been evaporated, was dropped into the desired flasks. After 20 additional min, glucose samples were taken and production was measured over the ensuing hour. The results indicate that the nonglucocorticoid steroids deoxycorticosterone and 5/3- dihydrocortisol and the antiglucocorticoids 17a-methyltestosterone, progesterone, and cortexolone did not stimulate gluconeogenesis. In fact, very high levels of 17a-methyltestosterone, deoxycorticosterone, and progesterone inhibited gluconeogenesis. Furthermore, 56-dihydrocortisol did not block the dexamethasone stimulation, while progesterone and cortexo- lone at concentrations that did not alter control rates of glucose production blocked the dexamethasone effect. As a single control experiment, cortexolone was shown not to inhibit glucagon s stimulation of gluconeogenesis, suggesting only that the blockby cortexolone was not a nonspecific inhibition of gluconeogenic stimulation. In the presence of inhibitory concentrations of the antiglucocorticoid progesterone, reduced levels of pyruvate and very high lactate/pyruvate ratios were noted. Progesterone (100 PM) yielded a pyruvate level of 0.03 f 0.01 mm and a

5 12758 Acute Glucocorticoid Stimulation of Gluconeogenesis TABLE I Steroid specificity of the acute glucocorticoid stimulation of gluconeogenesis Preparation and incubation of hepatocytes and measurement of glucosewere as described under Experimental Procedures. The desired quantities of deoxycorticosterone (DOC), 17a-methyltestosterone (17a-MT), progesterone, 5P-dihydrocortisol (5B-DHC), and cortexolone were dissolved in methanol and evaporated to dryness in the bottom of flasks prior to addition of buffer. Hepatocytes (0.5 mg of protein/ml) were incubated at 37 C under 95% 02, 5% C02 with 3.6 mm lactate only as initial substrate. After 40 min, a plastic tab, onto which a quantity of dexamethasone in methanol had been evaporated to dryness, was added to flasks as indicated to yield a final concentration of 50 p ~ Control. flasks received a plastic tab onto which the same volume of methanol was evaporated to dryness. In a single experiment, glucagon (500 nm) was added rather than a control tab. Glucose production rates are expressed as a percentage of stimulation over control rates. Each number represents the mean f S.E. of four to five experiments except for the single experiment with glucagon f cortexolone. Hormones present min min Stimulation over control % Dexamethasone, 50 p~ +28 f 3 DOC, 100 pm -79 f 3 DOC, 20 pm +5 f 3 17a-MT, 100 pm -62 f 8 17a-MT, 20 pm +6 f 4 Progesterone, 100 p~ -99 f 1 Progesterone, 20 PM +3 f 3 Progesterone, 20 /IM Dexamethasone, 50 p~ +16 t- 3 5@-DHC, 100 /LM +12 f 3 5@-DHC, 100 pm Dexamethasone, 50 p~ +34 k 6 Cortexolone, 100 PM +2 f 3 Cortexolone, 100 pm Dexamethasone, 50 pm +6 f 3 Glucagon, 500 nm +44 Cortexolone, 100 p~ Glucagon, 500 nm +41 lactate/pyruvate ratio of 131 k 46 (mean k S.E., n = 3). The inhibition of gluconeogenesis by this steroid, then, appeared to result from or cause an impairment of pyridine nucleotide oxidation and therefore impairment of pyruvate generation. The locus of this inhibitory effect was not further investigated. Inuestigation into the Mechanism of Stimulation of Glucose Production by Glucocorticoids-The experiments presented in Fig. 7 were designed to determine the time of onset of the glucocorticoid effect on glucose synthesis. Similar to the experiments of Table I, after allowing a 40-min incubation with 3.6 mm lactate present initially, a plastic tab, onto which dexamethasone in methanol had been evaporated, was dropped into the hepatocyte suspension, and glucose samples were taken at 20-min intervals from 40 to 120 min after the start of the incubation. A paired t test of five pairs of glucose concentrations at 40 min, prior to addition of dexamethasone, revealed no difference (p > 0.2), while at 60 min, only 20 min after dexamethasone addition, glucose concentrations were significantly greater than controls (p < 0.001). Furthermore, linear regression lines calculated from the data of five pairs of control and dexamethasone-treated hepatocyte preparations revealed an intersection at only 9 k 2 min after dexamethasone administration. This unexpectedly early onset suggests that a dexamethasone-induced stimulation of protein synthesis is an unlikely mechanism to explain the elevation of glucose production. Nevertheless, a series of experiments was designed to test this possibility since, for example, glucocorticoids can inhibit glu- cose transport by rat thymus cells within 15 min by a mechanism that can be prevented by actinomycin D (19). Control, C a &, 300- E L a, m = E m a, I I I I I I Time (minutes) FIG. 7. Time of onset of stimulation of gluconeogenesis by dexamethasone. Preparation and incubation of hepatocytes and measurement of glucose were as described under Experimental Procedures. Hepatocytes (0.5 mg of protein/ml) were incubated at 37 C under 95% 02, 5% C02 with 3.6 mm lactate only as initial substrate. After 40 min, a plastic tab, coated with a film of dexamethasone prepared by drying a methanolic solution of the steroid, was added to the flasks as indicated to yield a final concentration of 50 p~ (broken line). At 40 min, control flasks (solid line) received a plastic tab onto which methanol alone had been evaporated. Linear regression analyses using glucose values from 60,80,100, and 120 min were used to generate the straight lines drawn through the points. This single experiment is representative of five such replications. angiotensin 11-stimulated, and dexamethasone-stimulated glucose production rates were titrated over a wide range of concentrations of cordycepin, an inhibitor of mrna synthesis, and cycloheximide, an inhibitor of protein synthesis. In these experiments, cells were incubated with substrate and inhibitors for 40 min prior to dexamethasone addition and 60 min prior to angiotensin I1 addition. As shown in Fig. 8, no concentration of cordycepin inhibited dexamethasone-stimu- lated gluconeogenesis without also inhibiting control and angiotensin 11-stimulated rates. Concentrations of cycloheximide known to inhibit protein synthesis in other systems (20) further enhanced angiotensin 11- and dexamethasone-stimulated rates along with control rates. The additivity of the dexamethasone and cycloheximide stimulations suggests, however, that dexamethasone does not stimulate gluconeogenesis by promoting protein synthesis. While not shown in Fig. 8, the effect of 10 ~ L M cycloheximide on the stimulation of gluconeogenesis by glucagon was also tested in a single experiment. Glucagon alone yielded a 69% stimulation, while glucagon plus cycloheximide yielded a 63% stimulation. These results are consistent with the report (21) that cycloheximide increases CAMP levels in liver and adipose tissue. DISCUSSION While acute in vitro stimulations of gluconeogenesisby glucocorticoids had been reported in liver slices and isolated perfused livers from rats, no such stimulation had been re-

6 - ao- Acute Glucocorticoid Stimulation of Gluconeogenesis I I IO 100 Log [Inhibitor] FIG. 8. Effects of cycloheximide and cordycepin on stimulation of gluconeogenesis by dexamethasone and angiotensin 11. Preparation and incubation of hepatocytes and measurement of glucose production rates were as described under "Experimental Procedures." Hepatocytes (0.5 mg of protein/ml) were incubated at 37 "C under 95% 02, 5% COz with 3.2 mm lactate only as initial substrate. The desired concentration of cordycepin (log mg/ml, broken lines and open symbols) or cycloheximide (log pm, filled symbols and solid lines) was present at the beginning of the incubation. Control flasks received neither cordycepin nor cycloheximide. At 40 min, a plastic tab coated with a film of dexamethasone was added to flasks as indicated (0, B) to yield a final concentration of 50 p ~ Flasks. that received no hormone were given a plastic tab onto which methanol alone had been evaporated (0,O). At 60 min, angiotensin I1 (10 nm) was added to flasks as indicated (A, A). Glucose production rates are expressed as a percentage of change from control rates determined in the absence of any inhibitor or hormone. Each point represents the mean of two to threexperiments from different cell preparations. ported in freshly isolated hepatocytes from either adrenalectomized or normal animals. As noted in the preceding paper (18), stimulation of gluconeogenesis by dexamethasone was noted in suspensions of hepatocytes from fasted rats at a pyruvate concentration of 1 mm with lactate/pyruvate ratios of 17 and greater. In the present investigation, stimulation of gluconeogenesis by dexamethasone was again noted at a lactate/pyruvate ratio of 25 when the pyruvate concentration was 1 mm. In addition, we have now further shown that as the pyruvate concentration was reduced while keeping a relatively constant mean lactate/pyruvate ratio of10-13, the percentage stimulation of gluconeogenesis by dexamethasone increased. When hepatocytes were initially incubated with mM lactate only in the presence of dexamethasone, corticosterone, or triamcinolone, stimulation of gluconeogenesis was observed without a concurrent increase in the generation of pyruvate or decrease in the lactate/pyruvate ratio. Taken together, these observations argue, therefore, that the stimulation of gluconeogenesis by glucocorticoids is not simply due to enhanced removal of cytosolic NADH and does not strictly depend upon low concentrations of pyruvate, but rather depends upon low level(s) of some metabolic intermediate(s) whether induced by a low pyruvate concentration or very high cytosolic NADH/NAD+ ratios. Physiological concentrations of lactate and pyruvate are approximately 1 and 0.1 mm, respectively, in the blood. Even at the low cell concentrations used in this study, it would be difficult to hold these substrate levels very steady for any length of time. Furthermore, at low cell density and pyruvate concentrations of 0.1 mm, steady state levels of metabolites become difficult to measure. It is not surprising, therefore, that hormone actions are not usually examined at low, nearphysiological concentrations of lactate and pyruvate. The ability to reproduce successfully effects of hormones such as angiotensin I1 or glucagon at high levels of substrate reinforces any reluctance to work at low substrate concentrations. Very little effect of glucocorticoids can be seen at high substrate concentrations, however. When substrates are reduced, the percentage glucocorticoid stimulation becomes progressively larger. The investigations presented in the following paper (22) were designed to explore this observation in greater depth. With pyruvate only as starting substrate, a consistent dexamethasone stimulation was also observed. As noted previously (18), however, this stimulation correlated with an increase in lactate generation from the added pyruvate and could thus be explained on the basis of an enhanced generation of cytosolic NADH that would be expected to increase rates of glucose production. Observations that glucagon inhibits gluconeogenesis from pyruvate as the only substrate have been used to argue that pyruvate kinase is being inhibited. Conversely, the observation that stimulation of gluconeogenesis by angiotensin I1 and other Ca2+-mobilizing agents with pyruvate only as substrate has been used to argue that these hormones do not work by inhibiting pyruvate kinase. Thus, one might argue, since dexamethasone stimulates glucose production when pyruvate only is added, that dexamethasone, too, does not function by inhibiting pyruvate kinase. However, it is safer to conclude simply that, with pyruvate only as initial substrate, dexamethasone and angiotensin I1 are not stimulating gluconeogenesis by inhibiting pyruvate kinase. Under a different set of conditions, these hormones might inhibit pyruvate kinase flux in the cell. Since the dexamethasone stimulation is seen at high lactate/pyruvate ratios or low substrate concentrations but not at high substrate concentrations and intermediate lactate/pyruvate ratios, whereas angiotensin I1 and glucagon are much more versatile stimulators, it is reasonable to conclude that the mechanism whereby dexamethasone stimulates gluconeogenesis is different from that of angiotensin I1 or glucagon. Further support for this conclusion are the data of Fig. 4 showing that the effects of dexamethasone are additive to those of glucagon and angiotensin 11. If maximally effective concentrations of two agents are operating through identical mechanisms, their combined administration wouldbe less than additive. It is reassuring that the glucocorticoid effect is seen at lower, near-physiological substrate levels and at physiological lactate/pyruvate ratios rather than only under nonphysiological conditions. No evidence could be gathered favoring the view that the glucocorticoid stimulation of gluconeogenesis seen in this system resulted from glucocorticoid-altered gene expression in spite of the fact that altered gene expression is a widely accepted model that can explain almost all of the biological effects of glucocorticoids. Information that has been gathered regarding the regulation of tyrosine aminotransferase or glutamine synthetase by glucocorticoids in hepatoma cell lines and in freshly isolated rat hepatocytes exemplifies the characteristics of steroid action attributable to altered gene expression. In such studies, concentrations needed to elicit half-maximal elevations of either enzyme activities or mrna coding for the enzymes are typically nm for dexamethasone (23-26) and nm (23, 26) for corticosterone. Furthermore, when glucocorticoid binding to cytosolic extracts has been examined, binding affinities for dexamethasone and corticosterone were found that correlated well with their relative K,,, values (23,

7 12760 Acute Glucocorticoid Stimulation of Gluconeogenesis 26). A lag ranging from 30 to 120 min is characteristically observed between the time the hormone is added and enzyme induction is first noted. The effects of steroids can be inhibited by protein or mrna synthesis inhibitors. Glucocorticoid agonists such as dexamethasone, triamcinolone, and corticosterone will induce enzymes, dexamethasone and triamcinolone being slightly more potent than corticosterone. Nonglucocorticoids will not, or only partially, promote induction. Antiglucocorticoids will not promote induction, and prior administration of antiglucocorticoids will block the induction seen by glucocorticoids. Physiological blood levels of corticosterone in the rat vary according to sex, breed, time of day, and exposure to stress. In the male rat, levels have been reported as low as 250 nm (17) and as high as 1.8 p~ (27). Measurements from rats used in this study averaged 1.0 f 0.1 p ~ The. fraction of unbound corticosterone can be estimated to be 5-10%of total (28). Within this range of total corticosterone levels reported, then, free corticosterone concentrations are expected to range from approximately 10 to 200 nm. While, typically, enzyme inductions are approximately 2-10 times more sensitive to glucocorticoids than the acute effects on gluconeogenesis observed in these studies, the gluconeogenic potencies are near the physiological range. Of the steroids tested in this study, the gluconeogenic effect was shown to be specific to the known glucocorticoids, with dexamethasone and triamcinolone being slightly more potent than corticosterone. Furthermore, antiglucocorticoids known to inhibit glucocorticoid binding blocked glucocorticoid-stimulated glucose production, while the inactive steroid 5p-di- hydrocortisol did not. These characteristics are consistent with receptor-mediated action, but the sensitivities may be too low to ascribe the actions to binding of receptors believed responsible for mediating steroid activation of gene transcription. Furthermore, the very short time of onset and apparent insensitivity to cycloheximide inhibition suggest that this glucocorticoid effect does not result from a glucocorticoid receptor-mediated activation of gene transcription and subsequent stimulation of protein synthesis. The steroid specificity suggests, however, that the activation of gluconeogenesis is mediated by an allosteric modulation of a lower affinity glucocorticoid-specific binding site. It is teleological for a mammal, which responds so rapidly to acute hypoglycemia and acute stress with an acute rise in corticosterone, to have developed a mechanism whereby the effector organ responsible for glucose production can respond rapidly to that hormone. Other direct effects of glucocorticoids have been observed that do not fit the gene activation model. The most widely studied glucocorticoid effect not requiring gene activation is the postulated first step in the gene activation cascade: binding and allosteric activation of the glucocorticoid receptor. Rapid feedback inhibitions by glucocorticoids of the pituitary to decrease ACTH secretion (29) and the hypothalamus to decrease corticotropin-releasing factor secretion (30) are seen within minutes and appear to operate through a mechanism that is independent of a stimulation of protein synthesis. Ray et al. (31), some years ago, demonstrated increased glycogen deposition and increased blood glucose levels within 60 min of glucocorticoid administration to rats given doses of actinomycin D shown to inhibit protein synthesis. Thus, the slightly lower sensitivity of hepatocytes to corti- costerone, dexamethasone, and triamcinolone in these studies, the rapid onset of action, and the insensitivity to protein synthesis inhibition suggest that the glucocorticoid stimulation observed is due to direct allosteric mediation and may not be a result of the sequence of binding and activation of the knownglucocorticoid receptors. The sensitivity is not unrealistically low, however; and the steroid specificity characteristics are consistent with the possibility that this effect is physiologically meaningful. Furthermore, while an understanding of this stimulation of gluconeogenesis may be useful for elucidation of the apparent allosteric mechanism of glucocorticoid activation, this allosteric glucocorticoid activation may also prove to be a useful phenomenon for further understanding of the regulation of gluconeogenic flux. Acknowledgment-We gratefully acknowledge the helpful participation of Ralph A. Picking in carrying out these experiments and his extremely valuable contribution in preparing this manuscript. REFERENCES 1. Long, C. N. H., Katzin, B., and Fry, E. G. (1940) Endocrinology 26, Welt, I. D., Stetten, D., Jr., Ingle, D. J., and Morley, E. H. (1952) J. Bwl. Chem. 197, Haynes, R. C., Jr. (1962) Endocrinology 71, Haynes, R. C., Jr. (1964) Endocrinology 75, Uete, T., and Ashmore, J. (1963) J. Bwl. Chem. 238, Eisenstein, A. B., Berg, E., Goldenberg, D., and Jensen, B. (1964) Endocrinology 74, Friedmann, N., Exton, J. H., and Park, C.R. (1967) Biochem. Biophys. Res. Commun. 29, Exton, J. H., Friedmann, N., Wang, E. H.-A., Brineaux, J. P., Corbin, J. D., and Park, C. R. (1972) J. Biol. Chem. 247, Exton, J. H., Miller, T. B., Harper, S. C., Jr., and Park, C. R. (1976) Am. J. Physiol. 230, Friedmann, N. (1984) Physiol. Reu. 64, Eigler, N., Saccl, L., and Sherwin, R. S. (1979) J. Clin. Inuest. 63, Alan, E. H., and Titheradge, M. A. (1984) Biochem. J. 219, Seglen, P. 0. (1976) Methods Cell Biol. 13, Lowry, 0. H., and Passoneau, J. V. (1972) A Flexible System of Enzymatic Analysis, pp. 175, 200, 213, Academic Press, New York 15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, Guillemin, R., Clayton, G. W., Smith, J. D., and Lipscomb, H. S. (1959) J. Lab. Clin. Med. 53, Guillemin, R., Clayton, G. W., Smith, J. D., and Lipscomb, H. S. (1958) Endocrinology 63, Sistare, F. D., and Haynes, R. C., Jr. (1985) J. Biol. Chem. 260, Mosher, K.M., Young,D. A., and Munck, A. (1971) J. Biol. Chem. 246, Schreiber, G., and Schreiber, M. (1973) Subcell. Biochem. 2, Wititsuwannakul, D., and Kim, K.-H. (1977) Biochem. Biophys. Res. Commun. 76, Sistare, F. D., and Haynes, R. C., Jr. (1985) J. Biol. Chem. 260, Rousseau. G. G.. Barker. J. D.. and Tomkins., G. (1972).. J. Mol. Biol. 67,99-1i5 24. Crook. R. B.. Louie. M.. Deuel. T. F.. and Tomkins. G. M. (1978).. J. Biol. Chem. 253, Ernest, M. J., Chen, C.-L., andfeigelson, P. (1977) J. Biol. Chem. 252, Rousseau, G. G., and Schmit, J.-P. (1977) J. Steroid Biochem. 8, Glick, D., Von Redlich, D., and Levine, S. (1964) Endocrinology 74, Keller, N., Sendelbeck, L. R., Richardson, U. I., Moore, C., and Yates, F. E. (1966) Endocrinology 79, Portanova, R., and Sayers, G. (1974) Biochem. Biophys. Res. Commun. 56, Jones, M. T., and Hillhouse, E. W. (1976) J. Steroid Biochem. 7, Ray, P. D., Foster, D. O., and Lardy, H. A. (1964) J. Biol. Chem. 239,