Insulin Control of Hepatic Glucose Production1 Ltreiy D ~ a ImstI'f14fe s for Medical Research, betvisr'l G E I Z E ~Hospitab, CII Mo~trcrrC,Qlrehec H3T bE2
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Received May 24, 1974 Matsuura, N., Cheng, J. S. & Malant, N. (1975) Insulin Control of Hepatic Glucose Production. CUIT. 8. Bioclze~7z.53, 28-36 Insulin injected intravenously caused a rapid, marked decrease in hepatic glucose secretion in the rabbit. as determined by an isotope-dilution procedure. This was associated with a decrease in the concentrations of gluconeogenic intermediates from phospkoenolpyruvate to triose phosphates, inclusive, compatible with inhibition of gl~aconeogenesis art phosphoenolpyruvate carbowykinase. The concentration of glucose 6-phosphate was unaltered but that of hepatic glucose was reduced. The specific activities of the hewose phosphates, relative to that of liver glucose, were the same in control and insulin-treated animals. These observations can be explained by a decrease in the activity of glucose-6ghosphatase. It is concluded that this enzyme is a control point for hepatic glucose prsduction and is inhibited by insulin. In the rat. ins~alinproduced a rapid fall in blood sugar. The hepatic glucose output remained normal despite a fall in hepatic glucose 6-phosphate concentration during the initial period of insulin action. This suggests that glucose-6-phosphatarse activity was increased. Subsequently the concentration of glucose 6-phosphate returned to norlraal with no change in the rate of glucose ~roduction.The data suggest that in the rat, insulin produces a transient increase in gluco:,e-6-phosphatase activity. Matsuura, N., Cheng, J. S. & Kalarat, W. (1975) Insulin Control of Hepatic Gl~acose Production. Curr. B. Bfoci~enr. 53, 28-36 L'injection intraveineuse d'insuline .=ntraine une diminution rapide et anarqa~kede la stcrttion du glucose hkpatique chez le lapin, tel que dCmontr6 par dilution isotopiqne. Cette diminution est associte & une chute de la concentration des intermtdiaires de la ntogl~acog$n&se,depuis le phssphotnolpyruvate jusqaa'a~ax triose phosphates inclusivernent, cz qui est compatible avec une inhibition de la ntoglucog6n&se B l'ktape de la phosphodnolpyruvate carboxykinase. La concentration du glucose 6-phosphate demeure inchangCe alors que eelle du glucose htpatique est rkduite. Les activitts spCcifiqlaes des hexose phosphates par rapport B celle du glucose hkpatique sont les mErnes chez les anirnaux tCmoins et les animaux traitts B l'insuline. Ces observations peuvent s'expliquer par une diminution de l'activitk de la glucose-6-phosphatase. Cette enzyme exerce donc Lan contr8le sur la production du gluccse hkpatique et elle est inhibhe par I'insuline. Chez le rat, l'insuline dinain~aerapidement le glucose sangrain; la sortie du glucose hCpatique demeure normale en dtpit d'une baisse du gluzose 4-phosphate hkpatiqrne durant la pkriode initiale de I'action insuliniqbae. Ceci s~aggkreune augmentation de B'activit6 de la gl~acose-6-phosphatase.Par la suite, la concentration du glucose 6-phosphate reviemt B la normale sans changen11:nt dans la production du glucose. Ces donnkes siaggkrent que, chez le rat, l'insuline entraine une augmentation transitoire de lqactivit6 de [Tradiait par le journal] la glucose-6-phosphatase.
It has been shown by a variety of experimental techniques that insulin rapidly decreases the hepatic release of glucose. These techniques include, perfaasion of the isolated liver (1-31, measurement sf the glucose concentration gradient across an Eck fistula (4-6) or between hepatic artery and vein (71, and isotope dilution 'Supported by a grant from the Medical Research Ccsrancil of Canada. 'Postdoctcsral fellow of the Medical Research Council of Canada.
methods ( 8-1 1) . However, the mechanism by which insulin produces this effect on glucose secretion is not established. There is substantial evidence that diabetes increases the activity of the "key9' gluconeogcnic enzymes, pyruvate casboxylase, phosphoenolpyruvate carboxykinase, fructose diphssplaatase, rind glucose-6-pkosplaatass, and that insulin administration lowers these enzyme activities to normal (12) over a period of hours or days. An effect sf insulin on phosphoenolpyruvate produc-
M A T S U U R A ET AL.: HEPATIC GLUCOSE PRODUCTION
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
tion from pyruvate is also indicated by changes in hepatic concentration of gluconeogenic intermediates several hours after administration of the hormone and by the reverse changes after administration of anti-insulin serum (13 ) . The present report deals with an attempt to localize the hepatic site of action of insulin during the interval knmediately after its administration, when there is a marked decrease in glucose secretion.
Materials and Methods
Anin~uls Sprague-Dawley rats weighing 150-200 g and New Zealand rabbits weighing 2.5-3.0 kg were used. The goal of the experiments was to determine the effect of insulin on hepatic secretion of glucose, on the hepatic concentration of glaaconeogenic intermediates, and on their specific activities at a given time after administration of [14C]glucose. These measurements were possible in a single procedure in the rabbit. Rabbits were given acepromazine maleate intramuscularly and 1% lidocaine locally for anesthesia; polyethylene tubing (No. 160 Clay-Adams) was inserted into the internal jugular vein and passed to the junction of the venae cavae. Penicillin G (50 000 u/kg) was injected intramuscularly daily for 3 days. Cannulae were filled with heparin solution (70 u/ml) and washed with heparin-sdine solution once or twice a day, for 3 days prior to experimentation ( 14). After an overnight fast, [U-lJC]glucose (12.5 or 50 pCi per animal) was injected through the cannula; beginning 30 min later, blood samples were taken every 5 rnin into heparinized syringes and centrifuged immediately to separate the plasma. Immediately after the fifth sample, insulin 0.2 u/kg or 0,15 M NaCl was injected and two more blood samples were taken at 5-min intervals. The animal was then quickly anesthetized by the intravenous injection of pentobarbital, 50 mg/ kg, the abdomen was quickly opened, and the left lobe of the liver was immediately r e m ~ v e d after freeze clamping (15). The time lapse from injection of pentobarbital to freezing of samples was less than 30 s. Measurements were made of concentration and specific activity of glucose in the plasma and of the gluconeogenic intermediates in the liver. In the rat, separate procedures were used to measure the hepatic secretion of glucose and the concentration of the metabolic intermediates. Under pentobarbital anesthesia (40 mg/kg intraperitoneally) an indwelling catheter (polyethylene tubing, Clay-Adams No. PESO) was placed in the external jugular vein. For the secretion rate, ~14Cflqglucose (25 pCi) together with insulin, 0.1 U / 100 g, or an equivalent volume sf 0.15 M NaC1, was injected through the cannula; blood samples of 0.1 ml were taken from the cut tip of the tail at 4-min intervals for the next 14 min, and the specific activity of the plasma glucose was determined. These experiments were performed with two groups of animals. with and without pentobarbital anesthesia. For the concentrations of the metabolites, insulin or saline was injected
29
as above; 4, 6, 8, or 10 min later, the animal was anesthetized by the injection of 50 mg of sodium pentobarbital per kilogram through the cannula, the abdomen was quickly opened, and parts of both right and left lobes of the liver were removed by freeze clamping. Less than 20 s elapsed from injection of pentobarbital to freezing of liver samples. Rabbits were deprived of food for 16 h prior to experimentation; rats were without food for 4 h. Blood Sanlples Plasma was deproteinized (16) and an aliquot of the protein-free supernatant was assayed for glucose ( 17). The remaining supernatant was passed through a column (0.7 >( 6 cm) of Dowex 1 x 8 (acetate) to remove organic acids. The glucose was eluted with 5 ml of water, Aliquots were used for glucose assay and for measurement of glucose radioactivity by liquidscintillation counting. Prep~rcrtionof Liver Extracts Frozen liver was kept in liquid nitrogen until deproteinization. The tissue ( 1-2 g of rat, 3-5 g of rabbit) was weighed, then pulverized in a compression mortar (Thermovac Industries) that was pre-cooled in liquid nitrogen. The powdered liver was transferred to a glass homogenizer and homogenized at 0 "C with 3.5 volumes of 0.6 N perchloric acid. Protein was removed by centrifugation in the cold at 3000 g for 10 min and re-extracted with 2 volumes of 0.3 N perchloric acid. The combined supernatants were titrated to p H 6.5-7.0 by the addition of 30% KOH with vigorous stirring. After having stood 15 min at 0 "C,the precipitate of K61O1 was removed by centrifugation in the cold. The residue was extracted (18) and analyzed (19) for DNA. Measurement of Metabolite concentration.^ Enzymatic analyses were made for glc-1-P (20). glc-6-P and fru-6-P (21 ) fru-1,6-Py and triose-P (22), 3-P-glyceric acid, 2-P-glyceric acid and P-enolpyruvate (23), pyruvate (24), lactate (25), malate (26), and for ATP (27), ADP and AMP (28), and uridine diphssphate glucose ( U D P G ) (29) after the liver extract had been shaken with Florisil (60-100 mesh) to remove flavines (30). Measurements were made either spectrophotometrically or fluorimetrically. Glucose and glycogen in the liver extract were separated from the phosphorylated compounds on a column (0.4 x 5 c m ) of Dowex 1 x 8 (formate) 200-400 mesh. Two milliliters of liver extract were applied to the column and the neutral saccharides were eluted with 4 ml of water, lyophilized, then redissolved in 1 ml of water. Glycogen was precipitated with ethanol and removed by centrifugation. The supernatant was evaporated under a stream of air, redissolved in a known amount of water, and assayed for glucose (17), The glycogen was hydrolyzed with 1 N HCI for 2+ h at 100 "C, then neutralized with NaOH and assayed for glucose.
Measuremerrt of Specific Activities o f Intermediates These were isolated by a modification of the ionexchange chromatography procedure of Bartlett (3 1) followed by thin-layer and paper chromatography; the
CAN. J . BIOCHEM. VOL. 53, 1975 Liver exlraet
measurements seporotion fer meas&lrsment
of specific octirity I
I
0.1N formic
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
CB1014mi
00e and glycopen
loetie ocid
Dowex t I
I
Q.5N formic
1.W Bormic
I 7 0.3N ommonium 0.5N ammenivm
I
discord
I
olk. p'cse Dower 50 paper chromatog
GIs-6-C, elc
04k. p'cse paper shsamotes
FIG.1. Flow diagram for isolation sf gluconeogenic intermediates; details in text.
hexose phosphates were then hydrolyzed and the constitarent hexoses isolated by paper chromatography for measurement of specific activity (Fig. 1 ). Liver extract (4-6 ml) was applied to a Dowex 1 x 8 (formate) column (0.4 x 5 cm) at 4 "C. Glucose and glycogen in the neutral eluate were separated as described above. The glycogen was hydrolyzed and assayed for glucose and radioactivity. The glucose fraction was chromatographed (32) on prewashed Whatman No. 1 filter paper. The glucose standard was visualized with silver nitrate (33). The corresponding portion of the sample chromatogram was eluted with water; the eluate was applied to a column (0.5 x 5 cm) of Dowex 1x8 (acetate) to remove an inhibitor of the glucose oxidase reaction (34). Portions of the emaaesat were assayed for glucose and radioactivity. Control tubes contained the eluate from a gaper blank treated in an identical manner. Lactic acid, eluted with 0.1 N formic acid, was lyophilized, then redissolved in 100 pl of water. A 206~1portion was assayed enzymatically and a 50-p1 portion was purified by thin-layer chromatography (t.1.c.) on cellulose, with a solvent system of 96% ethanol water-25% NMIOH (100:12:16, v/v) (35). A lactic-acid standard was visualized with bromocresol purple. The corresponding area s f the sample was taken for measurement sf radioactivity. The 1.0 N formic-acid fraction containing hexose msnophosphntes (Fig. 1) was lyophilized, then treated with 1 N Pic1 for 6 min at 108 "C to hydrolyze glc-1P (36). (It was established that this treatment also hydrolyzed about 20% of the fru-6-P but none of the glc-6-P.) The mixture was immediately placed on a column (0.7 x 6 cm) of nowex 1 x 8 (formate) and washed with 5 ml of water. The emuent contained the free glucose and fructose. The undegraded glcd-P and fru-6-P were eluted with 15 ml of 2 M formic acid.
Both fractions were liyophilized. The glucose fraction was redissolved In 50 pB of water and chromatographed on paper as described above. Then the glucose was eluted, dried, and dissolved in 1.2 ml sf 0.05 M triethanolamine buffer (pH 7.6). One milliliter was assayed for glucose. A measured portion of the reaction mixture was then assayed for radioactivity in a liquidscintillation counter, with an external standard for quench correction. The glc-6-P and fru-6-P were redissolved in 1 0 0 ~ 1of HnO and applied to an ECTEOLA-cellulose t.1.c. plate (37 ) . A 20-cm length of pleated filter paper was clamped to the upper end of the plate to permit development for 18 h. The solvent was 0.025 M ammonium tetraborate (pH 10) 95% ethanol (2 :3, v/v) . The hexose monaphosphates were localized by radioautsgraphy, then extracted from the t.1.c. adsorbent with 3 x I mi of 1.5 N formic acid, and lyophi8ized. The separated phosphate esters were hydrolyzed with intestinal alkaline phosphatase. One milligram of enzyme in 1 ml of 0.05 Ad sodium-bicarbonate buffer, pH 9.3, containing 0.01 M magnesium acetate was added to each sample and incubated for 60 min at 37 "C. The whole incubation mixture was then passed through a column (0.5 x 7 cm) of Dowex 1x8 (acetate) layered on Dowex 50WX8 (H') to remove ions and protein. Fructose and glucose derived from fru-6-P and glc-6-P were eluted from the column with water, lyophilized, redissolved in 50 pl of water, and separated by paper chromatography as previously described. Glucose derived from glc-6-P was assayed and counted as for the glucose of glc-I-$. Fructose was similarly assayed enzymatically (38), then counted. UDPG eluted with 0.3 M amrmoniurn formate (Fig. 1) was dried, hydrolyzed with 0.5 N MCl at 100 "C for 2 h, and then passed through a Dowex 1x8 (formate) column (8.5 x 6 cm). The specific activity of glucose derived from UDPG was determined as for glc-1-P.
MAFSUURA EF AL.: HEPATIC GLUCOSE PRODUCTION
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Fru-1,6-PBeluted with 0.5 M ammonium formate was dried, then hydrolyzed with alkaline ghosghatase. Fructose was isolated by pager chromatography and the specific activity determined as described for fru6-P. Calculations of plasma-glucose flux rates were made according to Bunn et al. (8). Since the values obtained for successive measurement intervals of each phase of the experiment (pre- and post-insulin in the rabbit, post-saline and post-insulin in the rat) showed little variation. only average values are presented for each phase. Cilemicals Uniformly labelled [14C]glucose ( 192 mCiimmol ) was obtained from the New England Nuclear Corp. Low-zinc insulin (M-240 LZ),containing 0.05% glucagon, was kindly supplied by LiIly and Company, Indianapolis, Ind. Aldolase, glycerol-3-phosphate dehydrogenase, and triose phosphate isomerase were purchased from Boehringer Mannheim Company. Glucose oxidase reagent was purchased from Worthington Biochemical Corp. All other enzymes and nucleotides were obtained from the Sigma Chemical Company.
Results
( A ) Rabbit (i)Plasrna Glucose Flux In preliminary experiments it was demonstrated that injection of saline had no effect on the rate of fall of plasma glucose specific activity (Fig. 2 ) . Insulin caused an immediate decrease in the rate of fall of plasma glucose specific activity. There was a calculated decrease in glucose 3% of the control vdue and an influx to 23 increase in glucose efaux to 306 Z!Z 49% of control. These effects of insulin were subsequently confiimed by three experiments in which the csnstant infusion technique of isotope administration was used (10). The rate of glucose influx fell to 21-45% of the control value. (ii)Corzce~ztratiorzof In termed iates in Liver The absolute values for control and insulintreated animals are given in Table 1. Figures 3 and 4 show the values in the latter group expressed as a percent of control. It is clear that there was a decrease in the concentrations of glucose, and of the metabolites between malate and fru-1 ,6-P2. Nucleotide concentrations were measured to obtain an index of the redox state of the cell. The cytoplasmic free WAD/free NADH ratios (Table 1 ), calculated from the lactate/pyruvate ratio (301, were not significantly different in the two groups. There were no differences between the groups in the concentrations of the adenosine
*
I
Insulin Insulin
Saline
\ ?
'
t
.--o
insulin
Minutes
FIG. 2. Eflect of saline and insulin (0.2 u/kg intravenously) on plasma-glucose concentration and specific activity in the rabbit. Two preliminary experiments are shown.
phosphates, and the ATP/AMP ratios were comparable with those reported by others (39, 40). It was found that with an ATP/AMP ratio >3, the glc-6-P concentration was independent of this ratio. As the ratio fell below 3, the glc-6-P concentration rose abruptly; such results were considered to be the result of anoxia, and experiments with a ratio below 3 were discarded. (iii)Specific Activities of Hexose Phosphates The specific activities relative to that of liver glucose are shown in Fig. 5. In both groups of animals the values for liver glucose and plasma glucose were virtually identical. That of glc-6-P was much lower than that of hepatic glucose and there was an increase in the specific activity along the glycolytic and glycogenic pathways so that the values for fru-1,6-P2, glc-1-P, and UDPG were significantly higher ( y < 0.005) than that of glc-6-P. The only significant difference between control and insulin-treated animals was the higher activity of glc-1-P in the control group ( P < 0.02). ( B ) Rat (i) Plasrna Glucose Flux In contrast to the rabbit, the rat did not respond to insulin with a decrease in glucose flux
32
CAN. 3. BIOCMEM. VOL. 53, 1975
TABLE I . Csnce~~trations of gluconeogenic intermediates and related compounds in rabbit liver 10 min after injection of insulina
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Metabolite
Control
Insulin-treated
P
Glucose, plasma Qmg/lQOmI) Glucose, liver (pmoH/g) G%ucose6-phosphate (nmol/g) Fructose 6-phosphate (nmol/gj Fructose l,6-dipkosphate (mmaoH/g) Triose phosphate (nrnol/g) 3-Phospheglyceric acid (nmol/g) 2-Phesphoglyceric acid (nrnol/g) Phosphoenolpyruvate (nmol/'g) Pyruvate (nmsl/g) Lactate (nmsl/g) Malate (nmol/g) Glucose 1-phosphate (nmol! g) Uridiame dipbosphate glucose (nmoH,/g) Glycogen (pmol/g j ATP (ymol/g) ADP (trmollg) AMP (btrnoH/g) ATP/ADP ATP/AMP WAD/ NADH aSix animals per group.
FIG.3. Crossover plot showing the effect of insulin on gluconeogenic intermediates in rabbit liver. Insulin was given intravenously (0.2 u/kg) and liver removed 10 min later. Values from insulin-injected animals ( ( 2 )are expressed as a percentage of the mean control value (@); vertical bars represent S.E.M. Abbreviations: PEP, phospkoenolpyruvate; 2-PGA, 2-phospfnoglyceric acid; 3-PGA, 3-gkosphoglyceric acid; trisse-P, trliose phosphate; FDP, fre~etsse, I ,6-diphosphate; F - 6 2 , fructose 6-phosphate: G-6-P, glucose 6-phosphate; G, glucose.
from Iivcr to plasma; the rate of glucose release was constant over the 14-min interval following insulin injection. Since no differences were sb-
served between anesthetized and non-anesthetized rats, the results have been pooled for presentation in 'Table 2.
33
MATSUUWA ET AL.: HEPATIC GLUCOSE PRODUCTION
(ii) Concentration of Intermediates irl Liver Absolute values are shown in Table 3 and the time course sf changes in relative concentrations x
1
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
150
I
__----
_.---
I
'I----------\
I
I
/
T,/
e Control
lnrerlin
5
R;. 4. Crossover plot showing the effect of insulin on glycogenic intermediates in rabbit liver. Insulin was given intravenously (0.2 u/kg) and liver removed 10 min later. Values from insulin-injected animals (0) are expressed as a percentage of the mean control value ( 0 ) ;vertical bars represent S.E.M. Abbreviations: 6 ,glucose; G-6-P, glucose 6-phosphate; G1P,glucose I-phosphate; UDPG, uridine diphosphate glucose. TABLE 2. Effect of insulin on flux rates of plasma glucose in ratn
Control (10) Hnsulin (1 1 )
P
after insulin administration is shown in Fig. 6 . In the first few minutes after insulin there were decreases in the levels of glucose and glc-6-P which were greater than that of plasma glucose. These returned toward or above starting HeveIs even though the concentration of plasma glucose continued to fall rapidly. Glc-1-P and fru-6-P followed patterns almost parallel to that of glc-6-]Pa
Inflow (rng/min per 100 ml blood)
Outflow (naglrnin per BOO ml blood)
2.9250.3 3.138k0.52 N.S.
2.73 kQ.35 7.88f 0a61
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Received May 24, 1974 Matsuura, N., Cheng, J. S. & Malant, N. (1975) Insulin Control of Hepatic Glucose Production. CUIT. 8. Bioclze~7z.53, 28-36 Insulin injected intravenously caused a rapid, marked decrease in hepatic glucose secretion in the rabbit. as determined by an isotope-dilution procedure. This was associated with a decrease in the concentrations of gluconeogenic intermediates from phospkoenolpyruvate to triose phosphates, inclusive, compatible with inhibition of gl~aconeogenesis art phosphoenolpyruvate carbowykinase. The concentration of glucose 6-phosphate was unaltered but that of hepatic glucose was reduced. The specific activities of the hewose phosphates, relative to that of liver glucose, were the same in control and insulin-treated animals. These observations can be explained by a decrease in the activity of glucose-6ghosphatase. It is concluded that this enzyme is a control point for hepatic glucose prsduction and is inhibited by insulin. In the rat. ins~alinproduced a rapid fall in blood sugar. The hepatic glucose output remained normal despite a fall in hepatic glucose 6-phosphate concentration during the initial period of insulin action. This suggests that glucose-6-phosphatarse activity was increased. Subsequently the concentration of glucose 6-phosphate returned to norlraal with no change in the rate of glucose ~roduction.The data suggest that in the rat, insulin produces a transient increase in gluco:,e-6-phosphatase activity. Matsuura, N., Cheng, J. S. & Kalarat, W. (1975) Insulin Control of Hepatic Gl~acose Production. Curr. B. Bfoci~enr. 53, 28-36 L'injection intraveineuse d'insuline .=ntraine une diminution rapide et anarqa~kede la stcrttion du glucose hkpatique chez le lapin, tel que dCmontr6 par dilution isotopiqne. Cette diminution est associte & une chute de la concentration des intermtdiaires de la ntogl~acog$n&se,depuis le phssphotnolpyruvate jusqaa'a~ax triose phosphates inclusivernent, cz qui est compatible avec une inhibition de la ntoglucog6n&se B l'ktape de la phosphodnolpyruvate carboxykinase. La concentration du glucose 6-phosphate demeure inchangCe alors que eelle du glucose htpatique est rkduite. Les activitts spCcifiqlaes des hexose phosphates par rapport B celle du glucose hkpatique sont les mErnes chez les anirnaux tCmoins et les animaux traitts B l'insuline. Ces observations peuvent s'expliquer par une diminution de l'activitk de la glucose-6-phosphatase. Cette enzyme exerce donc Lan contr8le sur la production du gluccse hkpatique et elle est inhibhe par I'insuline. Chez le rat, l'insuline dinain~aerapidement le glucose sangrain; la sortie du glucose hCpatique demeure normale en dtpit d'une baisse du gluzose 4-phosphate hkpatiqrne durant la pkriode initiale de I'action insuliniqbae. Ceci s~aggkreune augmentation de B'activit6 de la gl~acose-6-phosphatase.Par la suite, la concentration du glucose 6-phosphate reviemt B la normale sans changen11:nt dans la production du glucose. Ces donnkes siaggkrent que, chez le rat, l'insuline entraine une augmentation transitoire de lqactivit6 de [Tradiait par le journal] la glucose-6-phosphatase.
It has been shown by a variety of experimental techniques that insulin rapidly decreases the hepatic release of glucose. These techniques include, perfaasion of the isolated liver (1-31, measurement sf the glucose concentration gradient across an Eck fistula (4-6) or between hepatic artery and vein (71, and isotope dilution 'Supported by a grant from the Medical Research Ccsrancil of Canada. 'Postdoctcsral fellow of the Medical Research Council of Canada.
methods ( 8-1 1) . However, the mechanism by which insulin produces this effect on glucose secretion is not established. There is substantial evidence that diabetes increases the activity of the "key9' gluconeogcnic enzymes, pyruvate casboxylase, phosphoenolpyruvate carboxykinase, fructose diphssplaatase, rind glucose-6-pkosplaatass, and that insulin administration lowers these enzyme activities to normal (12) over a period of hours or days. An effect sf insulin on phosphoenolpyruvate produc-
M A T S U U R A ET AL.: HEPATIC GLUCOSE PRODUCTION
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
tion from pyruvate is also indicated by changes in hepatic concentration of gluconeogenic intermediates several hours after administration of the hormone and by the reverse changes after administration of anti-insulin serum (13 ) . The present report deals with an attempt to localize the hepatic site of action of insulin during the interval knmediately after its administration, when there is a marked decrease in glucose secretion.
Materials and Methods
Anin~uls Sprague-Dawley rats weighing 150-200 g and New Zealand rabbits weighing 2.5-3.0 kg were used. The goal of the experiments was to determine the effect of insulin on hepatic secretion of glucose, on the hepatic concentration of glaaconeogenic intermediates, and on their specific activities at a given time after administration of [14C]glucose. These measurements were possible in a single procedure in the rabbit. Rabbits were given acepromazine maleate intramuscularly and 1% lidocaine locally for anesthesia; polyethylene tubing (No. 160 Clay-Adams) was inserted into the internal jugular vein and passed to the junction of the venae cavae. Penicillin G (50 000 u/kg) was injected intramuscularly daily for 3 days. Cannulae were filled with heparin solution (70 u/ml) and washed with heparin-sdine solution once or twice a day, for 3 days prior to experimentation ( 14). After an overnight fast, [U-lJC]glucose (12.5 or 50 pCi per animal) was injected through the cannula; beginning 30 min later, blood samples were taken every 5 rnin into heparinized syringes and centrifuged immediately to separate the plasma. Immediately after the fifth sample, insulin 0.2 u/kg or 0,15 M NaCl was injected and two more blood samples were taken at 5-min intervals. The animal was then quickly anesthetized by the intravenous injection of pentobarbital, 50 mg/ kg, the abdomen was quickly opened, and the left lobe of the liver was immediately r e m ~ v e d after freeze clamping (15). The time lapse from injection of pentobarbital to freezing of samples was less than 30 s. Measurements were made of concentration and specific activity of glucose in the plasma and of the gluconeogenic intermediates in the liver. In the rat, separate procedures were used to measure the hepatic secretion of glucose and the concentration of the metabolic intermediates. Under pentobarbital anesthesia (40 mg/kg intraperitoneally) an indwelling catheter (polyethylene tubing, Clay-Adams No. PESO) was placed in the external jugular vein. For the secretion rate, ~14Cflqglucose (25 pCi) together with insulin, 0.1 U / 100 g, or an equivalent volume sf 0.15 M NaC1, was injected through the cannula; blood samples of 0.1 ml were taken from the cut tip of the tail at 4-min intervals for the next 14 min, and the specific activity of the plasma glucose was determined. These experiments were performed with two groups of animals. with and without pentobarbital anesthesia. For the concentrations of the metabolites, insulin or saline was injected
29
as above; 4, 6, 8, or 10 min later, the animal was anesthetized by the injection of 50 mg of sodium pentobarbital per kilogram through the cannula, the abdomen was quickly opened, and parts of both right and left lobes of the liver were removed by freeze clamping. Less than 20 s elapsed from injection of pentobarbital to freezing of liver samples. Rabbits were deprived of food for 16 h prior to experimentation; rats were without food for 4 h. Blood Sanlples Plasma was deproteinized (16) and an aliquot of the protein-free supernatant was assayed for glucose ( 17). The remaining supernatant was passed through a column (0.7 >( 6 cm) of Dowex 1 x 8 (acetate) to remove organic acids. The glucose was eluted with 5 ml of water, Aliquots were used for glucose assay and for measurement of glucose radioactivity by liquidscintillation counting. Prep~rcrtionof Liver Extracts Frozen liver was kept in liquid nitrogen until deproteinization. The tissue ( 1-2 g of rat, 3-5 g of rabbit) was weighed, then pulverized in a compression mortar (Thermovac Industries) that was pre-cooled in liquid nitrogen. The powdered liver was transferred to a glass homogenizer and homogenized at 0 "C with 3.5 volumes of 0.6 N perchloric acid. Protein was removed by centrifugation in the cold at 3000 g for 10 min and re-extracted with 2 volumes of 0.3 N perchloric acid. The combined supernatants were titrated to p H 6.5-7.0 by the addition of 30% KOH with vigorous stirring. After having stood 15 min at 0 "C,the precipitate of K61O1 was removed by centrifugation in the cold. The residue was extracted (18) and analyzed (19) for DNA. Measurement of Metabolite concentration.^ Enzymatic analyses were made for glc-1-P (20). glc-6-P and fru-6-P (21 ) fru-1,6-Py and triose-P (22), 3-P-glyceric acid, 2-P-glyceric acid and P-enolpyruvate (23), pyruvate (24), lactate (25), malate (26), and for ATP (27), ADP and AMP (28), and uridine diphssphate glucose ( U D P G ) (29) after the liver extract had been shaken with Florisil (60-100 mesh) to remove flavines (30). Measurements were made either spectrophotometrically or fluorimetrically. Glucose and glycogen in the liver extract were separated from the phosphorylated compounds on a column (0.4 x 5 c m ) of Dowex 1 x 8 (formate) 200-400 mesh. Two milliliters of liver extract were applied to the column and the neutral saccharides were eluted with 4 ml of water, lyophilized, then redissolved in 1 ml of water. Glycogen was precipitated with ethanol and removed by centrifugation. The supernatant was evaporated under a stream of air, redissolved in a known amount of water, and assayed for glucose (17), The glycogen was hydrolyzed with 1 N HCI for 2+ h at 100 "C, then neutralized with NaOH and assayed for glucose.
Measuremerrt of Specific Activities o f Intermediates These were isolated by a modification of the ionexchange chromatography procedure of Bartlett (3 1) followed by thin-layer and paper chromatography; the
CAN. J . BIOCHEM. VOL. 53, 1975 Liver exlraet
measurements seporotion fer meas&lrsment
of specific octirity I
I
0.1N formic
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
CB1014mi
00e and glycopen
loetie ocid
Dowex t I
I
Q.5N formic
1.W Bormic
I 7 0.3N ommonium 0.5N ammenivm
I
discord
I
olk. p'cse Dower 50 paper chromatog
GIs-6-C, elc
04k. p'cse paper shsamotes
FIG.1. Flow diagram for isolation sf gluconeogenic intermediates; details in text.
hexose phosphates were then hydrolyzed and the constitarent hexoses isolated by paper chromatography for measurement of specific activity (Fig. 1 ). Liver extract (4-6 ml) was applied to a Dowex 1 x 8 (formate) column (0.4 x 5 cm) at 4 "C. Glucose and glycogen in the neutral eluate were separated as described above. The glycogen was hydrolyzed and assayed for glucose and radioactivity. The glucose fraction was chromatographed (32) on prewashed Whatman No. 1 filter paper. The glucose standard was visualized with silver nitrate (33). The corresponding portion of the sample chromatogram was eluted with water; the eluate was applied to a column (0.5 x 5 cm) of Dowex 1x8 (acetate) to remove an inhibitor of the glucose oxidase reaction (34). Portions of the emaaesat were assayed for glucose and radioactivity. Control tubes contained the eluate from a gaper blank treated in an identical manner. Lactic acid, eluted with 0.1 N formic acid, was lyophilized, then redissolved in 100 pl of water. A 206~1portion was assayed enzymatically and a 50-p1 portion was purified by thin-layer chromatography (t.1.c.) on cellulose, with a solvent system of 96% ethanol water-25% NMIOH (100:12:16, v/v) (35). A lactic-acid standard was visualized with bromocresol purple. The corresponding area s f the sample was taken for measurement sf radioactivity. The 1.0 N formic-acid fraction containing hexose msnophosphntes (Fig. 1) was lyophilized, then treated with 1 N Pic1 for 6 min at 108 "C to hydrolyze glc-1P (36). (It was established that this treatment also hydrolyzed about 20% of the fru-6-P but none of the glc-6-P.) The mixture was immediately placed on a column (0.7 x 6 cm) of nowex 1 x 8 (formate) and washed with 5 ml of water. The emuent contained the free glucose and fructose. The undegraded glcd-P and fru-6-P were eluted with 15 ml of 2 M formic acid.
Both fractions were liyophilized. The glucose fraction was redissolved In 50 pB of water and chromatographed on paper as described above. Then the glucose was eluted, dried, and dissolved in 1.2 ml sf 0.05 M triethanolamine buffer (pH 7.6). One milliliter was assayed for glucose. A measured portion of the reaction mixture was then assayed for radioactivity in a liquidscintillation counter, with an external standard for quench correction. The glc-6-P and fru-6-P were redissolved in 1 0 0 ~ 1of HnO and applied to an ECTEOLA-cellulose t.1.c. plate (37 ) . A 20-cm length of pleated filter paper was clamped to the upper end of the plate to permit development for 18 h. The solvent was 0.025 M ammonium tetraborate (pH 10) 95% ethanol (2 :3, v/v) . The hexose monaphosphates were localized by radioautsgraphy, then extracted from the t.1.c. adsorbent with 3 x I mi of 1.5 N formic acid, and lyophi8ized. The separated phosphate esters were hydrolyzed with intestinal alkaline phosphatase. One milligram of enzyme in 1 ml of 0.05 Ad sodium-bicarbonate buffer, pH 9.3, containing 0.01 M magnesium acetate was added to each sample and incubated for 60 min at 37 "C. The whole incubation mixture was then passed through a column (0.5 x 7 cm) of Dowex 1x8 (acetate) layered on Dowex 50WX8 (H') to remove ions and protein. Fructose and glucose derived from fru-6-P and glc-6-P were eluted from the column with water, lyophilized, redissolved in 50 pl of water, and separated by paper chromatography as previously described. Glucose derived from glc-6-P was assayed and counted as for the glucose of glc-I-$. Fructose was similarly assayed enzymatically (38), then counted. UDPG eluted with 0.3 M amrmoniurn formate (Fig. 1) was dried, hydrolyzed with 0.5 N MCl at 100 "C for 2 h, and then passed through a Dowex 1x8 (formate) column (8.5 x 6 cm). The specific activity of glucose derived from UDPG was determined as for glc-1-P.
MAFSUURA EF AL.: HEPATIC GLUCOSE PRODUCTION
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Fru-1,6-PBeluted with 0.5 M ammonium formate was dried, then hydrolyzed with alkaline ghosghatase. Fructose was isolated by pager chromatography and the specific activity determined as described for fru6-P. Calculations of plasma-glucose flux rates were made according to Bunn et al. (8). Since the values obtained for successive measurement intervals of each phase of the experiment (pre- and post-insulin in the rabbit, post-saline and post-insulin in the rat) showed little variation. only average values are presented for each phase. Cilemicals Uniformly labelled [14C]glucose ( 192 mCiimmol ) was obtained from the New England Nuclear Corp. Low-zinc insulin (M-240 LZ),containing 0.05% glucagon, was kindly supplied by LiIly and Company, Indianapolis, Ind. Aldolase, glycerol-3-phosphate dehydrogenase, and triose phosphate isomerase were purchased from Boehringer Mannheim Company. Glucose oxidase reagent was purchased from Worthington Biochemical Corp. All other enzymes and nucleotides were obtained from the Sigma Chemical Company.
Results
( A ) Rabbit (i)Plasrna Glucose Flux In preliminary experiments it was demonstrated that injection of saline had no effect on the rate of fall of plasma glucose specific activity (Fig. 2 ) . Insulin caused an immediate decrease in the rate of fall of plasma glucose specific activity. There was a calculated decrease in glucose 3% of the control vdue and an influx to 23 increase in glucose efaux to 306 Z!Z 49% of control. These effects of insulin were subsequently confiimed by three experiments in which the csnstant infusion technique of isotope administration was used (10). The rate of glucose influx fell to 21-45% of the control value. (ii)Corzce~ztratiorzof In termed iates in Liver The absolute values for control and insulintreated animals are given in Table 1. Figures 3 and 4 show the values in the latter group expressed as a percent of control. It is clear that there was a decrease in the concentrations of glucose, and of the metabolites between malate and fru-1 ,6-P2. Nucleotide concentrations were measured to obtain an index of the redox state of the cell. The cytoplasmic free WAD/free NADH ratios (Table 1 ), calculated from the lactate/pyruvate ratio (301, were not significantly different in the two groups. There were no differences between the groups in the concentrations of the adenosine
*
I
Insulin Insulin
Saline
\ ?
'
t
.--o
insulin
Minutes
FIG. 2. Eflect of saline and insulin (0.2 u/kg intravenously) on plasma-glucose concentration and specific activity in the rabbit. Two preliminary experiments are shown.
phosphates, and the ATP/AMP ratios were comparable with those reported by others (39, 40). It was found that with an ATP/AMP ratio >3, the glc-6-P concentration was independent of this ratio. As the ratio fell below 3, the glc-6-P concentration rose abruptly; such results were considered to be the result of anoxia, and experiments with a ratio below 3 were discarded. (iii)Specific Activities of Hexose Phosphates The specific activities relative to that of liver glucose are shown in Fig. 5. In both groups of animals the values for liver glucose and plasma glucose were virtually identical. That of glc-6-P was much lower than that of hepatic glucose and there was an increase in the specific activity along the glycolytic and glycogenic pathways so that the values for fru-1,6-P2, glc-1-P, and UDPG were significantly higher ( y < 0.005) than that of glc-6-P. The only significant difference between control and insulin-treated animals was the higher activity of glc-1-P in the control group ( P < 0.02). ( B ) Rat (i) Plasrna Glucose Flux In contrast to the rabbit, the rat did not respond to insulin with a decrease in glucose flux
32
CAN. 3. BIOCMEM. VOL. 53, 1975
TABLE I . Csnce~~trations of gluconeogenic intermediates and related compounds in rabbit liver 10 min after injection of insulina
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
Metabolite
Control
Insulin-treated
P
Glucose, plasma Qmg/lQOmI) Glucose, liver (pmoH/g) G%ucose6-phosphate (nmol/g) Fructose 6-phosphate (nmol/gj Fructose l,6-dipkosphate (mmaoH/g) Triose phosphate (nrnol/g) 3-Phospheglyceric acid (nmol/g) 2-Phesphoglyceric acid (nrnol/g) Phosphoenolpyruvate (nmol/'g) Pyruvate (nmsl/g) Lactate (nmsl/g) Malate (nmol/g) Glucose 1-phosphate (nmol! g) Uridiame dipbosphate glucose (nmoH,/g) Glycogen (pmol/g j ATP (ymol/g) ADP (trmollg) AMP (btrnoH/g) ATP/ADP ATP/AMP WAD/ NADH aSix animals per group.
FIG.3. Crossover plot showing the effect of insulin on gluconeogenic intermediates in rabbit liver. Insulin was given intravenously (0.2 u/kg) and liver removed 10 min later. Values from insulin-injected animals ( ( 2 )are expressed as a percentage of the mean control value (@); vertical bars represent S.E.M. Abbreviations: PEP, phospkoenolpyruvate; 2-PGA, 2-phospfnoglyceric acid; 3-PGA, 3-gkosphoglyceric acid; trisse-P, trliose phosphate; FDP, fre~etsse, I ,6-diphosphate; F - 6 2 , fructose 6-phosphate: G-6-P, glucose 6-phosphate; G, glucose.
from Iivcr to plasma; the rate of glucose release was constant over the 14-min interval following insulin injection. Since no differences were sb-
served between anesthetized and non-anesthetized rats, the results have been pooled for presentation in 'Table 2.
33
MATSUUWA ET AL.: HEPATIC GLUCOSE PRODUCTION
(ii) Concentration of Intermediates irl Liver Absolute values are shown in Table 3 and the time course sf changes in relative concentrations x
1
Can. J. Biochem. Downloaded from www.nrcresearchpress.com by McMaster University on 11/20/14 For personal use only.
150
I
__----
_.---
I
'I----------\
I
I
/
T,/
e Control
lnrerlin
5
R;. 4. Crossover plot showing the effect of insulin on glycogenic intermediates in rabbit liver. Insulin was given intravenously (0.2 u/kg) and liver removed 10 min later. Values from insulin-injected animals (0) are expressed as a percentage of the mean control value ( 0 ) ;vertical bars represent S.E.M. Abbreviations: 6 ,glucose; G-6-P, glucose 6-phosphate; G1P,glucose I-phosphate; UDPG, uridine diphosphate glucose. TABLE 2. Effect of insulin on flux rates of plasma glucose in ratn
Control (10) Hnsulin (1 1 )
P
after insulin administration is shown in Fig. 6 . In the first few minutes after insulin there were decreases in the levels of glucose and glc-6-P which were greater than that of plasma glucose. These returned toward or above starting HeveIs even though the concentration of plasma glucose continued to fall rapidly. Glc-1-P and fru-6-P followed patterns almost parallel to that of glc-6-]Pa
Inflow (rng/min per 100 ml blood)
Outflow (naglrnin per BOO ml blood)
2.9250.3 3.138k0.52 N.S.
2.73 kQ.35 7.88f 0a61
