Hepatic Effects of Chlorpropamide Inhibition of Glucagon-stimulated Gluconeogenesis in Perfused Livers of Fasted Rats STANLEY A. BLUMENTHAL AND KYRA RIEGLE WHITMER
MATERIALS AND METHODS
SUMMARY In perfused livers of rats fasted for 24 h, glucagon (5 x 1O~10 M) significantly elevated tissue and perfusate levels of cyclic AMP and caused a twofold increase in glucose formation from lactate. Chlorpropamide (0.8 x 10~3 M) consistently blocked these effects. Measurements of metabolic intermediates suggest that chlorpropamide may inhibit gluconeogenesis by antagonizing the action of glucagon on the phosphoenolpyruvate cycle. In the experiments described, chlorpropamide did not lower hepatic ATP concentration or energy charge, and exerted its effects at perfusate concentrations comparable to serum concentrations reported in patients on maintenance doses of the drug. DIABETES 28:646-650, July 1979.
I
t is probable that sulfonylurea drugs reduce plasma glucose levels by a combination of pancreatic and extrapancreatic effects.1"4 For several years, the authors have employed the perfused rat liver preparation of Exton et al.5 in studies on the effects of chlorpropamide, a representative sulfonylurea drug, on hepatic carbohydrate metabolism. These studies have developed evidence that chlorpropamide not only augments the action of insulin, but also suppresses glucagon-stimulated gluconeogenesis when the liver is exposed to low insulin concentrations, i.e. after a 24-h fast. In addition, the drug causes a parallel suppression of cyclic AMP production.6-7 This communication describes efforts to identify the site, or sites, at which chlorpropamide inhibits glucagonmediated gluconeogenesis, and to determine whether this inhibition is contingent on a lowering of hepatic ATP concentration. From the Department of Medicine, Upstate Medical Center, State University of New York, and the Veterans Administration Hospital, Syracuse, New York. Accepted for publication 21 March 1979.
646
Livers of 120-160 g male Sprague-Dawley rats which had been fasted for 24 h were perfused for 125 min with oxygenated recirculating Krebs bicarbonate buffer. The buffer contained 3% bovine serum albumin and sufficient outdated human red blood cells to bring the hematocrit to 20%. The perfusion technique used in these studies has been described previously.6 The experimental protocol is summarized in Table 1. Perfusate volume approximated 80 cm3. Livers were frozen in liquid nitrogen at 125 min for assay of gluconeogenic intermediates, adenine nucleotides, and cyclic AMP. Samples of perfusate were taken at 65 and 125 min for measurement of glucose and cyclic AMP. In preparation for analysis of intracellular metabolites, samples of liver were extracted with cold perchloric acid and neutralized. Cyclic AMP was determined by the method of Gilman using protein kinase purified from beef skeletal muscle.8'9 Gluconeogenic intermediates and adenine nucleotides were measured by standard enzymatic methods and absorption spectrophotometry.10 Glucose was measured with glucose oxidase using the Beckman 670800 glucose analyzer (Beckman Instruments, Fullerton, California). RESULTS
Perfusion of livers with 13.4 mM lactate was associated with a mean glucose production rate of 17.1 /xmol/g wet wt liver/h. Addition of glucagon (5 x 10~10 M) caused a 2.4-fold increase in glucose production above control levels (Table 2). These production rates are essentially identical to those reported by Exton and Park in fasted livers perfused for 1 h before lactate and hormone addition.12 When chlorpropamide was added to the perfusate with glucagon, glucose production was depressed significantly toward control levels. The blockade of glucagon action was both more consistent and more complete at the higher perfusate concentration of the drug (0.8 x 10~3 M) than at the lower concentration (0.4 x 10~3 M). When given alone, chlorpropamide (0.8 x 10~3 M) did not alter significantly control levels of glucose output (Table 2).
DIABETES, VOL 28, JULY 1979
STANLEY A. BLUMENTHAL AND KYRA RIEGLE WHITMER
mean concentrations of gluconeogenic intermediates after 125 min of perfusion. These concentrations are expressed as percentages of the corresponding control values, and the means and standard errors of these values are also ChlorLacGlurecorded. It is apparent from Figure 1 that glucagon cagon^ propamide§ Experimental group tatef significantly reduced pyruvate levels and increased the — — + I. Control levels of malate (which was measured instead of oxalaceII. Glucagon-treated + + tate since we did not have access to a fluorimeter), phosIII. Chlorpropamide-treated + + phoenolpyruvate, 2-phosphoglycerate, .and 3-phosphoIV. Treated with glucagon glycerate. The addition of chlorpropamide to glucagon + • + chlorpropamide + + prevented the fajl in pyruvate and lessened the incre* Livers were perfused with recirculating oxygenated medium ments in malate, phosphoenolpyruvate, and 3-phosphoalone for 60 min. All additions to the perfusate were made between glycerate. Chlorpropamide alone produced a pattern of 60-125 min. Perfusions were terminated at 125 min. 1120 mg of sodium L(+)-lactate were added to each perfusate intermediates which closely resembled that in control between 60 and 65 min. Maximum initial medium concentration, livers between pyruvate and 3-phosphoglycerate and which 13.4 mM. did not differ significantly from the control pattern be% In groups II and IV, 0.15 fig glucagon was given by bolus tween phosphoglyceraldehyde and glucose 6-phosphate.
TABLE 1 Protocol for the addition of chlorpropamide, glucagon, and sodium lactate to the perfusion medium*
at 60 min followed immediately by constant infusion of 0.15 fig over the next 65 min. Maximum initial medium concentration, 5 x 10-'°M. § In groups III and IV, 20 mg of chlorpropamide were given by bolus at 60 min plus 20 mg by constant infusion over the next 65 min. Maximum initial medium concentration, 0.8 x 10~3 M.
DISCUSSION
It should be noted that the livers used in these experiments were derived from rats fasted for 24 h and were, therefore, exposed to low concentrations of insulin. In this regard, earlier studies had demonstrated that there was no measStimulation of glucose production by glucagon was ac- urable insulin in the perfusate after the first 60 min of companied by a significant increment in hepatic concen- perfusion. These preliminary studies also established that tration of cyclic AMP and in the accumulation of cyclic the concentration of liver glycogen was extremely low AMP in the medium (Table 3). When chlorpropamide was (
MATERIALS AND METHODS
SUMMARY In perfused livers of rats fasted for 24 h, glucagon (5 x 1O~10 M) significantly elevated tissue and perfusate levels of cyclic AMP and caused a twofold increase in glucose formation from lactate. Chlorpropamide (0.8 x 10~3 M) consistently blocked these effects. Measurements of metabolic intermediates suggest that chlorpropamide may inhibit gluconeogenesis by antagonizing the action of glucagon on the phosphoenolpyruvate cycle. In the experiments described, chlorpropamide did not lower hepatic ATP concentration or energy charge, and exerted its effects at perfusate concentrations comparable to serum concentrations reported in patients on maintenance doses of the drug. DIABETES 28:646-650, July 1979.
I
t is probable that sulfonylurea drugs reduce plasma glucose levels by a combination of pancreatic and extrapancreatic effects.1"4 For several years, the authors have employed the perfused rat liver preparation of Exton et al.5 in studies on the effects of chlorpropamide, a representative sulfonylurea drug, on hepatic carbohydrate metabolism. These studies have developed evidence that chlorpropamide not only augments the action of insulin, but also suppresses glucagon-stimulated gluconeogenesis when the liver is exposed to low insulin concentrations, i.e. after a 24-h fast. In addition, the drug causes a parallel suppression of cyclic AMP production.6-7 This communication describes efforts to identify the site, or sites, at which chlorpropamide inhibits glucagonmediated gluconeogenesis, and to determine whether this inhibition is contingent on a lowering of hepatic ATP concentration. From the Department of Medicine, Upstate Medical Center, State University of New York, and the Veterans Administration Hospital, Syracuse, New York. Accepted for publication 21 March 1979.
646
Livers of 120-160 g male Sprague-Dawley rats which had been fasted for 24 h were perfused for 125 min with oxygenated recirculating Krebs bicarbonate buffer. The buffer contained 3% bovine serum albumin and sufficient outdated human red blood cells to bring the hematocrit to 20%. The perfusion technique used in these studies has been described previously.6 The experimental protocol is summarized in Table 1. Perfusate volume approximated 80 cm3. Livers were frozen in liquid nitrogen at 125 min for assay of gluconeogenic intermediates, adenine nucleotides, and cyclic AMP. Samples of perfusate were taken at 65 and 125 min for measurement of glucose and cyclic AMP. In preparation for analysis of intracellular metabolites, samples of liver were extracted with cold perchloric acid and neutralized. Cyclic AMP was determined by the method of Gilman using protein kinase purified from beef skeletal muscle.8'9 Gluconeogenic intermediates and adenine nucleotides were measured by standard enzymatic methods and absorption spectrophotometry.10 Glucose was measured with glucose oxidase using the Beckman 670800 glucose analyzer (Beckman Instruments, Fullerton, California). RESULTS
Perfusion of livers with 13.4 mM lactate was associated with a mean glucose production rate of 17.1 /xmol/g wet wt liver/h. Addition of glucagon (5 x 10~10 M) caused a 2.4-fold increase in glucose production above control levels (Table 2). These production rates are essentially identical to those reported by Exton and Park in fasted livers perfused for 1 h before lactate and hormone addition.12 When chlorpropamide was added to the perfusate with glucagon, glucose production was depressed significantly toward control levels. The blockade of glucagon action was both more consistent and more complete at the higher perfusate concentration of the drug (0.8 x 10~3 M) than at the lower concentration (0.4 x 10~3 M). When given alone, chlorpropamide (0.8 x 10~3 M) did not alter significantly control levels of glucose output (Table 2).
DIABETES, VOL 28, JULY 1979
STANLEY A. BLUMENTHAL AND KYRA RIEGLE WHITMER
mean concentrations of gluconeogenic intermediates after 125 min of perfusion. These concentrations are expressed as percentages of the corresponding control values, and the means and standard errors of these values are also ChlorLacGlurecorded. It is apparent from Figure 1 that glucagon cagon^ propamide§ Experimental group tatef significantly reduced pyruvate levels and increased the — — + I. Control levels of malate (which was measured instead of oxalaceII. Glucagon-treated + + tate since we did not have access to a fluorimeter), phosIII. Chlorpropamide-treated + + phoenolpyruvate, 2-phosphoglycerate, .and 3-phosphoIV. Treated with glucagon glycerate. The addition of chlorpropamide to glucagon + • + chlorpropamide + + prevented the fajl in pyruvate and lessened the incre* Livers were perfused with recirculating oxygenated medium ments in malate, phosphoenolpyruvate, and 3-phosphoalone for 60 min. All additions to the perfusate were made between glycerate. Chlorpropamide alone produced a pattern of 60-125 min. Perfusions were terminated at 125 min. 1120 mg of sodium L(+)-lactate were added to each perfusate intermediates which closely resembled that in control between 60 and 65 min. Maximum initial medium concentration, livers between pyruvate and 3-phosphoglycerate and which 13.4 mM. did not differ significantly from the control pattern be% In groups II and IV, 0.15 fig glucagon was given by bolus tween phosphoglyceraldehyde and glucose 6-phosphate.
TABLE 1 Protocol for the addition of chlorpropamide, glucagon, and sodium lactate to the perfusion medium*
at 60 min followed immediately by constant infusion of 0.15 fig over the next 65 min. Maximum initial medium concentration, 5 x 10-'°M. § In groups III and IV, 20 mg of chlorpropamide were given by bolus at 60 min plus 20 mg by constant infusion over the next 65 min. Maximum initial medium concentration, 0.8 x 10~3 M.
DISCUSSION
It should be noted that the livers used in these experiments were derived from rats fasted for 24 h and were, therefore, exposed to low concentrations of insulin. In this regard, earlier studies had demonstrated that there was no measStimulation of glucose production by glucagon was ac- urable insulin in the perfusate after the first 60 min of companied by a significant increment in hepatic concen- perfusion. These preliminary studies also established that tration of cyclic AMP and in the accumulation of cyclic the concentration of liver glycogen was extremely low AMP in the medium (Table 3). When chlorpropamide was (
