Effects of selective insulin or glucagon deficiency on glucose turnover H. L. A. LICKLEY,
G. G. ROSS, AND
M. VRANIC
Departments of Physiology and Surgery, and Women’s College Hospital, University of Toronto, Toronto, Ontario LICKLEY, H. L. A., G. G. Ross, AND M. VRANIC. Effects of selective insulin or glucagon deficiency on glucose turnover. Am. J. Physiol. 236(3): E255-E262, 1979 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 5(3): E255-E262, 1979.-To study the importance of glucagon and insulin in diabetes, somatostatin (ST) was infused, alone or with insulin or glucagon, in 11 conscious dogs. Plasma immunoreactive insulin (IRI) and glucagon (IRG) levels fell 65 -+ 4% and 33 t 3%, respectively, with somatostatin infusion. Glucose production (RB) assessed by [3-3H]glucose, [2-3H]glucose, or [l-14C]glucose decreased transiently. This is in contrast to the rise in R, seen after insulin withdrawal in depancreatized dogs, which have normal levels of IRG. Thus, suppression of IRG with somatostatin prevented an increase in R, in spite of suppression of IRI. When near basal IRG levels were provided during ST infusion in normal dogs, R, increased, indicating that glucagon contributes to the acute development of diabetes. When basal IRI levels were provided with ST, suppression of R, was maintained, suggesting that the transience of the metabolic effects of ST-induced glucagon suppression requires concomitant insulin suppression. A comparison of glucose turnover measured using different tracers showed that ST-related hormonal changes did not alter the rate of futile cycling in the liver. ST induced a rise in plasma free fatty acid (FFA) levels, attributed solely to insulin deficiency, as glucagon suppression did not significantly alter FFA concentrations when normal insulin levels were maintained.
diabetes mellitus; somatostatin; depancreatized production; free fatty acids; [3-3H]glucose; [ l-‘4C]glucose; futile cycle; Cori cycle
dogs; glucose [2-3H]glucose;
IT HAS BEEN SHOWN in man (10) and in dogs (2,3,8) that when both plasma immunoreactive insulin (IRI) and glucagon (IRG) levels are decreased concomitantly by the infusion of somatostatin (ST), initially there is no resultant hyperglycemia, but rather a decrease in plasma glucose levels in spite of the acute insulin deficiency. This is in contrast to the hyperglycemia noted in insulindeprived depancreatized dogs (33). Depancreatized dogs have been shown to have normal basal levels of glucagon (4, 32), and, if insulin-deprived, they exhibit a consistent and marked increase in the glucagon moiety (34). Other laboratories have also reported glucagon immunoreactivity in dogs after pancreatectomy (20, 21). The main source of the extrapancreatic glucagon in these depancreatized dogs is the fundus of the stomach (22,34). This gastrointestinal glucagon cannot be distinguished immunochemically from pancreatic glucagon (22, 29, 32). It 0363-6100/79/0000-OOOOooo$Ol.25
Copyright
0 1979 the American
Physiological
is possible that glucagon is essential for the acute development of diabetes and that suppression of glucagon prevents the rapid rise in glucose production that would be expected during somatostatin-induced insulin deficiency. In the present study somatostatin was given alone or together with an intravenous infusion of glucagon or with an intraportal infusion of insulin to compare the acute effects of a combined deficiency of insulin and glucagon on glucose turnover with the effects of a selective de% ciency of either insulin or glucagon. The rates of infusion of glucagon and insulin have been chosen in order to approach normal basal plasma levels of these hormones. The changes in glucose turnover during the administration of somatostatin, with or without glucagon or insulin replacement, were studied during a primed constant tracer infusion using either a simultaneous infusion of [ l-‘4C]glucose together with [2-3H]glucose or [3-3H]glucase alone as the tracer material. These results have been published in abstract form previously (17, 27). Insulin is known to inhibit lipolysis, but the role of glucagon in lipolysis is not as clearly delineated. In phar macological doses glucagon can stimulate lipolysis in diabetic man (18), but physiological doses did not increase free fatty acid (FFA) levels in depancreatized, insulin-infused dogs (6). Thus, changes in FFA levels were measured during infusion of somatostatin to determine the effects of a combined deficiency of insulin and glucagon on lipolysis. FFA levels were also monitored during infusion of somatostatin together with basal insulin replacement to study the effects of selective glucagon deficiency on lipolysis. MATERIALS
AND
METHODS
Preparation of experimental animals. Experimental studies were carried out on 11 healthy male mongrel dogs weighing 12-25 k. The animals received a high-protein diet containing 200 g of dog chow (Wayne Tailwagger, Allied Mills) and 400 g of beef chunks (Dr. Ballard’s) daily. In the first group of five animals, three polyvinyl cannulas were inserted under a short general anesthetic 2 days prior to each experiment (pentobarbital sodium, Abbott Laboratories). The cannula for infusion of labeled glucose was inserted in the saphenous vein and advanced until the tip was in the inferior vena cava just below the entrance of the hepatic vein. The cannulas for infusion Society
E255
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
E256 of somatostatin and glucagon were inserted into the cephalic veins and the sampling cannula was introduced into the superior vena cava via the jugular vein. In the second group of six animals laparotomy was performed (1 wk prior to each experiment). Under general anesthetic (pentobarbital sodium induction followed by inhalation anesthesia using Fluothane and nitrous oxide) a polyvinyl cannula was inserted into the portal vein through a tributary of the splenic vein for the infusion of insulin. In addition, a cannula was inserted into the cephalic vein for the administration of somatostatin, and a sampling cannula was introduced into the inferior vena cava via a saphenous vein. Experimental procedures. All experiments were conducted in conscious, trained, unsedated animals after an overnight fast of 18 h. I. In 5 dogs, the experimental study consist.ed of 1) a 50-min period of tracer equilibration, 2) a 40-min control period, 3) a 50-min infusion of cyclic somatostatin (Ayerst Laboratories), 4) a 50-min recovery period, and 5) a 50min infusion of somatostatin (0.83 ,ug/kg per min) together with glucagon (Eli Lilly and Co., Toronto, Ontario). Somatostatin was dissolved in a solution of normal saline containing 0.2% bovine serum albumin adjusted to pH 6.0-7.0 immediately prior to each experimental study. Somatostatin was infused at a rate of 0.83 r_Lg/kg per min. Glucagon was infused into a peripheral vein at a rate of between 0.13 pg/kg-h and 0.43 pg/kg-h in order to provide concentrations of IRG within the range normally seen in the portal venous blood. II. In four additional dogs the experimental study consisted of 1) an initial 50-min period of tracer equilibration, 2) a 40-min control period, 3) a 75-min infusion of somatostatin given at a rate of 0.5 pg/kg per min, 4) a 50min recovery period, and 5) a 75-min infusion of somatostatin (0.5 ,ug/kg per min) together with porcine insulin (Connaught Laboratories, Toronto, Ontario). Insulin was infused at a rate of 200 pU/kg per min. This insulin infusion rate has been shown to provide normal basal insulin levels (6). In two additional dogs a similar protocol was followed, but the somatostatin and somatostatinplus-insulin infusion periods were reversed. In was confirmed that the results obtained were not dependent on the order in which somatostatin and somatostatin plus insulin were given. Tracer methods and calculations. In part I, a double tracer method was employed. A solution of [ 1-‘4C]glucose (New England Nuclear, Boston, MA) and [2-3H]glucose (Amersham/Searle, Arlington Heights, IL) were mixed immediately prior to each experiment to ensure a constant ratio of the two tracer solutions so that there were 2.0 ,&i/ml of [1-14C]glucose and 6.25 ,&i/ml of [2-3H]glucose. A simultaneous infusion of [ l-14C]glucose and [2-3H]glucose was given to determine if changes in glucagon and/or insulin concentrations could affect the rate of loss of tritium into futile cycles (15). A priming dose equivalent to the amount of tracer infused in 105 min was given at time 0 and the infusion was delivered at 0.152 ml/min. In part II, [3-3H]glucose (New England Nuclear) was used as tracer (5.0 &i/ml). The tritium label at this position of the glucose molecule is lost to body water and
LICKLEY,
ROSS,
AND
VRANIC
is not reincorporated into glucose (1, 15). Thus, the use of [3-“HIglucose as a tracer eliminates the underestimation of glucose production due to recycling-as seen when [l-‘4C]glucose is employed as the tracer substance, and it also eliminates the overestimation of glucose production due to futile cycling-as seen when [2-3H]glucose is used as the tracer substance (1, 21). A priming dose equivalent to the amount infused in 90 min was given at time 0 and the infusion was delivered at 0.152 ml/min. The rates of production (rate of appearance, R,) and utilization (rate of disappearance, Rd) of endogenous glucose were determined by the method of primed tracer infusion (4, 7, 23). It has been demonstrated in using this method with dogs that the kinetics of exogenously infused glucose or inulin (a metabolically inert polysaccharide) can be predicted with accuracy under both steadystate and rapidly changing nonsteady-state conditions (23, 24). Because the specific activity of plasma glucose did not reach a plateau until approximately 50 min, only data from the last 40 min (t = 50-90 min) of the initial control period were used for the calculation of base-line turnover values. The sliding-fit technique was applied to glucose concentration and specific activities to calculate R, and Rd (4). Laboratory methods. Blood for glucose determination was placed in heparinized tubes containing sodium fluoride. The plasma was separated by centrifugation and the plasma glucose was determined by the glucose oxidase method with the use of the Beckman glucose analyser. An aliquot of plasma was then deproteinized with a mixture of equal volumes of 5% zinc sulphate and 0.3 N barium hydroxide. I. When [l-14C]glucose and [2-3H]glucose were given together as tracer, the supernatant was passed through an ion exchange resin (Ag. 2-X8, Bio-Rad Laboratories, Richmond, CA) to remove labeled metabolites of glucose. An aliquot of the supernatant was then evaporated to dryness in order to eliminate any tritium present as 3HZ0. The residue was redissolved in water and “H and 14C were counted simultaneously in Aquasol (New England Nuclear) according to the method of Hetenyi and Reynolds (12), with the use of three independent channels on a Nuclear Chicago Mark I liquid scintillation spectrometer. A second sample of the column eluate was used to determine the extent of recycling of the 14Clabel via the tricarbon intermediates (Cori cycle) (25). In all experiments aliquots of the infused tracer were treated in the same manner as the plasma samples, and 3H and 14C were counted. II. When [3-3H]glucose was used as tracer, an aliquot of the supernatant was evaporated and the residue dissolved in water and counted in Aquasol by liquid scintillation spectrometry. The supernatant was not passed through an ion exchange resin because all tritium from position 3 is lost to body water (15) and is therefore not present in metabolites of glucose. We have compared radioactivity of the supernatant passed through an ion exchange resin (Ag. 2-X8), with that not passed through resin and found no difference. An aliquot of plasma was extracted in Dole’s solution for determination of free fatty acid (FFA) levels using a radiochemical method (13).
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
INSULIN
AND
GLUCAGON
DEFICIENCY
AND
GLUCOSE
E257
TURNOVER
For the determination of plasma immunoreactive glucagon, 4.0-ml samples of blood were transferred to chilled tubes containing 0.2 M EDTA (24 mg/ml) and 0.2 ml (2,000 KIU) Trasylol (FBA Pharmaceuticals, New York City). Plasma glucagon was determined using 30 K antiserum, which is considered to be reactive with the Cterminal portion of the glucagon molecule and which crossreacts only very weakly with gut glucagonlike immunoreactivity (9). 1251-labeledglucagon and purified glucagon for standards were obtained from Novo Research Institute (Denmark). Serum insulin was assayed in triplicate by the method of Hales and Randle (ll), with the use of the Amersham/Searle kit. The described method was modified by double-diluting the antibodies so that the assay would be sensitive to insulin concentrations as low as 4 jJJ/ml. Statistical analysis was done using the paired t test.
250
200 c2 aI p150 u
ST+ G
u" $100
ST 50
l-
r
2150
20
3 2
10
z -
0
L, u I
I
I
I
I
.85+18pCJ/d 078+26pCJ/d
T
250
ST+G
ST
160
ST+G
Minutes 1. Effects of infusion of somatostatin (0.8 ,ug/kg per min, W) or somatostatin together with glucagon (0.13-0.43 pg/kg per h, (O----- 0) on (A) serum immunoreactive insulin (IRI) concentrations ($J/ml), (B) percentage change in serum immunoreactive glucagon (IRG) concentrations, and (C) plasma glucose concentrations (mg/lOO ml) in 5 normal dogs. Vertical bars represent SEM. Basal values for IRG are indicated.
mg/kg-min
02.d
0.46 mg/kg-min
50 I 0
2
0.49
6 o‘=
I. The effects of the infusion of somatostatin alone or tsgether with glucagon on circulating insulin, glucagon, and plasma glucose levels are shown in Fig. 1. Immedi-
a.24 2
& 6 100
RESULTS
FIG.
3.31*0.53 mg / kg-min o 2.982 0.52 mg/ kg-min l
I 10
I 20
I 30
1 40
Minutes 2. Effects of infusion of somatostatin (0.8 ,ug/kg per min, O---O) or somatostatin together with glucagon (0.13-0.43 ,ug/kg per h, O----O) on (A) percentage changes in rate of appearance (RB) of glucose and (B) percentage changes in rate of disappearance (Rd) of glucose in 5 normal dogs. Vertical bars represent SEM. Basal values for R, and Rd are indicated. FIG.
ately after the start of the somatostatin infusion there was a 64% decrease in serum IRI and a 28% decrease in plasma IRG concentrations. Both hormones remained suppressed throughout the infusion period (P < 0.025). With the infusion of somatostatin alone, plasma glucose concentrations fell slightly (P < 0.01 at 30 min). When glucagon was infused concurrently with somatostatin, in order to compensate for the decrease in plasma IRG levels induced by the somatostatin infusion, serum IRI was suppressed to levels comparable to those seen with the infusion of somatostatin alone. Glucagon was infused into a peripheral vein, and it was assumed that the IRG levels in the portal system were equivalent to those measured in the peripheral circulation. Thus, although plasma glucagon levels rose with the infusion of glucagon in all but one animal, plasma IRG levels during the glucagon infusion period remained within the reported basal range for IRG in portal blood (28). In contrast to the changes seen during the infusion of somatostatin alone, the concomitant infusion of somatostatin and glucagon resulted in an increase in plasma glucose concentration from 100 t 2 mg/lOO mg to 145 t 10 mg/lOO mg. The effects of the infusion of somatostatin alone or together with glucagon on glucose R, and glucose Rd is shown in Fig. 2. The infusion of somatostatin resulted in a small but significant decrease in the rate of production of glucose (P < 0.05 at 20 min). Rd fell slightly, reaching a 17 t 5% drop by the end of the infusion (P < 0.05). With the infusion of somatostatin and glucagon there was a marked increase in glucose production, which was significant at each time point; and there was a slight increase in Rd, which was significant at the 40-min time point. The changes in Rd parallel the changes in plasma glucose concentration during both infusion periods.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
E258
LICKLEY,
The use of [Z-“HIglucose as a tracer resulted in consistently higher values for R, as compared to those obtained using [ l-14C]glucose corrected for recycling (Table 1). A comparison of the ratio of R, calculated using [23H]glucose to the ratio of R, calculated using [l-‘“Clglucose for all time points in four of the animals indicates that the mean percentage difference in R, is approximately 22%. The consistency of turnover results obtained using both of these isotopes is also illustrated in Fig. 3, where it is shown that the specific activities of the [2-3H]glucose (stippled area) and [ 1-14C]glucose (solid line) follow coincident paths throughout the experimental period except at the end of the glucagon infusion period when an acceleration of futile cycling was indicated. The specific activity of [2-3H]glucose measured in one animal was used as a representative example and all others were normalized to yield identical values at t = 90 mm. II. The effects of the infusion of somatostatin alone or together with basal insulin replacement on serum IRI, TABLE
I Dog No.
1. Ratios I Initial cp”,:;;“b
r
of rates of glucose production Somatostatin
Second Control Period
95- 105 min
r 1.17 1.05 1.63 1.23
Somatostatin
+ Glucagon
195205 min
215225 min
205215 min
225235 min
1.11 1.07 1.26 1.25 1.02 1.04 1.04 1.14 1.19 1.80 1.09 1.33 1.11 1.35 1.88 1.35 1.28 1.13 1.19 1.41 0.17 0.06 0.07 0.16
The rates of glucose production were calculated by using either specific activities of [2-3H]glucose (Ra3H) or of [ l-‘4C]glucose corrected for recycling (R,14C). The data are expressed as the ratio of the two rates (R,“H: R,14C) prior to, during, and after infusion of somatostatin and during infusion of somatostatin and glucagon in 4 normal dogs. Time intervals given are minutes from start of the primed tracer
ROSS,
AND
VRANIC
plasma IRG, plasma glucose, and FFA levels are indicated in Fig. 4. It can be seen that the infusion of somatostatin again induced a marked decrease in serum IRI and plasma IRG and there was also a decrease in plasma glucose concentration. However, with the intraportal administration of insulin at 200 pU/kg per min, together with somatostatin, basal insulin levels were maintained. There was a 46 t 4% decrease (P < 0.001) in serum IRG within 15 min of the start of the somatostatin and insulin infusion that was similar to the suppression of IRG seen with the infusion of somatostatin alone. Glucagon suppression persisted throughout the 70-min infusion period. There was a similar initial decrease in plasma glucose levels (P c 0.05 by 30 min) seen with the administration of somatostatin, whether or not basal insulin replacement was provided. However, the decrease in plasma glucose concentration reached a plateau by the 30-min time point with the infusion of somatostatin alone, whereas it continued to decrease when insulin was also given. With the infusion of somatostatin, circulating levels of FFA rose significantly (P < 0.05) to a maximum of 188 t 32% above basal levels and remained elevated until the end of the infusion period. With insulin replacement, however, the apparent somatostatin-induced rise in FFA was not statistically significant. The effects on Ra and Rd of the infusion of somatostatin alone or together with basal insulin replacement are shown in Fig. 5. Initially there was an equivalent decrease in R, with the infusion of somatostatin or somatostatin together with insulin. The percentage decrease in R, was 28 t 5% (P < 0.005) at the 15- and 30.min time points with the infusion of somatostatin alone; however this effect on Ra did not persist --- and - by 45 min the decrease -in R, was no longer significant. In contrast, the 27 t 6% sunnression (P < 0.01) in R, seen with the administration of *somatostatin and insulin was maintained throughout the experimental period. There was a similar decrease in Rd with the administration of somatostatin or somatostatin together with basal insulin replacement until the 30-
FIG. 3. Specific activities (means t SE) of [2-“HIglucose (stz&Zed area) and [ l-14C]glucose (solid line) prior to and during infusions of somatostatin and of somatostatin and glucagon in 4 normal dogs. Specific activity of [2-3H]glucose in dog 2 was used as a representative example and all others were normalized to yield identical values at t = 90 min.
Minutes
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
INSULIN
AND
GLUCAGON
OL 0 L1L -
100
80
w
40
.-
ON
I
100
AND
GLUCOSE
I
I
1
1
150228
A I
L
I
1
I
I
pCJ/Id
r I
-
ST
u r
e
ST+INS
60 50 225
I-
I
I
1
f-
I
I
T
ST
tb E f o\Q
E259
TURNOVER
0 180~62pg/ml
n
l
c
DEFICIENCY
parable decrease in IRI (26, 33). In these latter animals insulin deficiency resulted in a marked increase in glucose production and plasma glucose concentrations, whereas utilization of glucose was decreased. Such depancreatized animals have normal levels of circulating glucagon (4, 34). Thus, the comparison between depancreatized dogs and somatostatin-infused normal dogs indicates that the presence of glucagon is essential for the diabetogenic effect of acute insulin deficiency. In the present study three different tracers were used: [ l-‘4C]glucose and [2-3H]glucose were administered simultaneously in one group of experiments; and [3-3H]glucose was used in the remaining experiments. The rate of production of glucose calculated using [2-3H]glucose as tracer was consistently approximately 20% higher than that with [l-‘4C]glucose. This difference has been attributed to futile cycling (14). In diabetes mellitus the rate of futile cycling is threefold higher than normal (30). Although in the present studies the absolute turnover values differed depending on the tracer employed, with the infusion of somatostatin the ratio of R, (as measured by [2-3H]glucose) to R, (measured using [l-14C]glucose as tracer) (Ra3H: R,14C) remained within a narrow range (Table 1). This indicates that somatostatin or somatostatin-induced decreases in insulin and glucagon did not alter the rate of futile cycling. There was a comparable maximum percentage decrease in R, with [l-14C]glucose (26 t 3%), [2-3H]glucose (26 t 3%), and [3-3H]glucose (28 t 6%). Thus, the changes in R, occurring in response to the infusion of somatostatin were not dependent on the tracer employed. In addition, it was shown that Ra3H: R,l*C did not increase after the cessation of the somatostatin infusion, suggesting that there was no substantial deposition of glycogen in the liver during the somatostatin infusion period; for, if glucose is incorporated into
150
ST+ INS
o 4.142 0.75 mg/kg-min
125 l
3.96 + 0.82 mg/kg-
min
100 I
0
I 15
1
1
I
J
30
45
60
75
(s
Minutes FIG. 4. Effects of infusion of somatostatin (0.5 pg/kg per min, O-----O) or somatostatin together with basal insulin replacement (200 pU/kg per min, 0 -----0) on (A) serum IRI concentrations (pU/ml), (B) percentage changes in IRG, (C) plasma glucose concentrations (mg/lOO), and (D) percentage changes in free fatty acid (FFA) levels. Vertical bars represent SEM. Basal values for IRG and FFA are indicated.
60
I
1
1
I
100 -r
Pu
ST *:
80
P u
min time point; subsequently Rd diverged, paralleling the changes in plasma glucose concentration.
ST+ INS . 4.072
6
\o .
DISCUSSION
Bihormonal deficiency of insulin and glucagon induced by somatostatin. The infusion of somatostatin in this study resulted in suppression of insulin and glucagon release, which was associated with decreased R, and plasma glucose, as has been reported by several investigators (2, 3, 27). This is in contrast to the changes that we have noted in pancreatectomized dogs with a com-
*
1
O\o
60
0.82 mg/
k;
min
Lo 4.12 ,+ o.68mg/kg-min I
I
I
I
0
15
30
45
I 60
Minutes
5. Effects of infusion of somatostatin (0.5 g/kg per min, M) and somatostatin together with basal insulin replacement (200 on (A), percentage changes in rate of appearU/kg per min, 0 -----0) ance of glucose (R*), and (B), percentage changes in rate of disappearance of glucose (Rd), in 5 normal dogs. Vertical bars represent SEM. Basal values for R, and Rd are indicated. FIG.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
E260
LICKLEY,
q
Pancreatectomy
(lnsulin~)
Somatostatin
Insulin Glucagon
Somatostatin
lal Glucagon
+
&
i
(InsulinJ)
! I
Glucose Production
:I
T
i
Ii
_ __
_
FIG. 6. Effects of insulin deficiency induced either by pancreatectomy (Cl) or by concurrent infusion of somastatin and glucagon (a), as compared to effects of a combined deficiency of insulin and glucagon induced by infusion of somatostatin alone (m). Maximal percentage changes (means t SE) from control period are shown for glucose concentration and rate of production of glucose (R,). Pancreatectomy data were obtained from results of previous experiments (33). Pancreatectomized dogs were initially maintained on basal portal infusion of insulin (227-270 U/kg per min).
glycogen during the somatostatin infusion, 14C but not 3H would be incorporated. Thus during the postsomatostatin period some of the [l-‘4C]glucose from liver glycogen would reappear in plasma due to glycogenolysis, thus increasing the Ra3H: R,14C ratio. SeZectiue ins&in deficiency. The infusion of somatostatin together with glucagon replacement resulted in acute insulin deficiency and an accompanying rise in R, and plasma glucose. These results were comparable to those seen previously in depancreatized dogs (33), again confirming the importance of glucagon for the hyperglycemic effect of acute insulin deficiency and also indicating that the effects on glucose production of pancreatic and gastrointestinal glucagon are comparable (26). The effects of somatostatin, alone or together with glucagon, are summarized in Fig. 6. The effects of acute insulin deficiency induced by somatostatin in normal dogs are compared to the effects of sudden withdrawal of insulin in five depancreatized dogs. When exogenous insulin was withdrawn from the depancreatized dogs for 60 min,
ROSS,
AND
VRANIC
serum IRI decreased by 60%. However, extrapancreatic glucagon was within the normal range both before and after the cessation of the insulin infusion (34). The insulin deficiency resulted in significant increments in plasma glucose production and concentration. These effects were abolished in the normal dogs when glucagon was also suppressed by means of a somatostatin infusion. There was no statistically significant difference in the increase in glucose concentration and R, between the depancreatized dogs and the normal dogs that were concurrently infused with glucagon and somatostatin. The results observed with the infusion of somatostatin and glucagon also suggest that the liver is more sensitive to increases in plasma glucagon when plasma insulin is decreased, as also noted by others (2, 3); for, when depancreatized dogs were infused with basal insulin, it required 6 times the amount of glucagon that was given in the present studies to achieve a comparable increase in Ra (5). This increased responsiveness of the liver to glucagon, seen with insulin deficiency, is in contrast to the results of studies in which an increase in circulating insulin did not play an important role in regulating liver sensitivity to glucagon in the resting dog (5). At basal glucagon concentrations similar changes in Ra were seen in depancreatized dogs and in the normal dogs that received somatostatin and glucagon. Thus it appears that the effects of somatostatin are due to suppression of insulin release rather than to a direct effect of somatostatin. This has also been clearly demonstrated by Cherrington et al. (3), who gave basal insulin and glucagon replacement to dogs during somatostatin infusions and found that Ra and plasma glucose were restored to norma1 values. An indication that the diabetic state also increases the rate of futile cycling in the liver (30) was indicated only at the end of the selective insulin deficiency period. Sete&ive ghcagon deficiency. When intraportal basal insulin replacement was provided during a somatostatin infusion there was a comparable decrease in R, to that seen with the infusion of somatostatin alone. This effect on R, is maintained throughout the infusion of somatostatin and insulin, whereas the effect of somatostatin alone in Ra appears to be transient; for in 8 of the 11 animals studied Ra returned toward preinfusion values within 45 min of the start of the somatostatin infusion. Thus suppression of glucagon release may be capable of preventing the rise in Ra seen with insulin lack, but with effect of glucabasal insulin replacement the restraining gon d.eficiency on R, is prolonged, suggesting that i nsulin sensitizes the liver to the effects of glucagon deficiency. The transient natu re of the effects of somatostatin glucose production has also been noted by others (2, 31). The administration of basal insulin replacement together with somatostatin in man (19) permitted a prolonged maintenance of suppression of glucose production, as in our studies in dogs. The administration of somatostatin resulted in a significant rise in FFA concentrations. The slight increase in FFA levels occurring in response to the administration of somatosta tin and insulin was not significant. Thus insulin replacement has prevented the rise in FFA seen
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
INSULIN
AND
GLUCAGON
DEFICIENCY
AND
GLUCOSE
E261
TURNOVER
with a somatostatin infusion. Selective glucagon deficiency cannot be considered to be antilipolytic under the conditions of the present study, and this finding is consistent with the observation that glucagon administration had no lipolytic effects in the dog (6). However, glucagon has a direct stimulatory effect on ketogenesis in the liver when insulin is deficient, which does not occur as a result of either increased hepatic FFA uptake or increased lipolysis (16). In conclusion, suppression of glucagon release prevented the rise in R, and plasma glucose that occurred with acute insulin deficiency in depancreatized dogs. When basal glucagon levels were provided during insulin deficiency, there was a rise in R, and plasma glucose levels, suggesting that glucagon plays an important part in the acute development of diabetes. When basal insulin levels were provided during glucagon deficiency, t,here
was a prolonged suppression of R, that was not observed during bihormonal insulin and glucagon deficiency, suggesting that insulin sensitizes tlhe liver to the effects of glucagon deficiency. The increased FFA levels observed during somatostatin infusion are attributed solely to insulin deficiency because selective glucagon deficiency did not affect lipolysis. We are indebted t.o J. Wilkins and N. Kovacevic for their excellent assistance. We thank Dr. R. Deghengyi (Ayerst, Research Laboratories, Montreal, Quebec) for the somatostatin and Connaught Laboratories, Toromo, Ontario for t.he insulin used in the studies. This investigation was supported by the Medical Research Council of Canada, the Canadian Diabetes Association, and the Dorothy Frances Graham Research Fund, Women’s College Hospital. G. Ross was a graduate student in the Department, of Physiology. Received
19 Dec. 1977; accepted
in final
form
17 Oct. 1978.
REFERENCES 1. ALTSZULER,
N., A. BARKAI, C. BJERKNES, B. GOTTLIEB, AND R. Glucose turnover values in the dog obtained with various species of labeled glucose. .4m. J. Physiol. 229: 1662-1667, 1975. ALTSZULER, N., B. GOTTLIEB, ,~ND T. HAMPSHIRE. Interaction of somatostatin, glucagon and insulin on hepatic glucose output in the normal dog. Diabetes 25: 116-121, 1976. CHERRINGTON, ,4. D., J. L. CHIASSON, J. E. LILJENQUIST, A. S. JENNINGS, V. KELLER, AND W. W. LACY. The role of insulin and glucagon in the regulation of basal glucose production in the postabsorptive dog. J. Clin. Invest. 58: 1407-1415, 1976. CHERRINGTON, A. D., R. KAWAMORI, S. PEK, AND M. VRANIC. Arginine infusion in dogs. Model for the roles of insulin and glucagon in regulating turnover and free fatty acid levels. Diabetes 23: 805-815, 1974. CHERRINGTON, A. D., AND M. VRANIC. Effect of interaction between insulin and glucagon on glucose turnover and FFA concentrations in normal and depancreatized dogs. kfetabolism 23: 729-774, 1974. CHERRINGTON, A. D., M. VRANIC, I-). FONO, AIL’D N. KOVACEVIC. Effect of glucagon on glucose t,urnover and plasma free fatty acids on depancreatized dogs maint,ained on matched insulin infusion. Can. J. PhysioZ. Pharmacol. 50: 946-954, 1972. DEBODO, R. C., R. STEELE, N. ALTSZULER, A. DUNN, AND J. BISHOP. On the hormonal regulation of carbohydrate metabolism. Studies with C’” glucose. Recent Prog Horm. Res. 19: 445-488, 1963. DOBBS, R., H, SAKURAI, H. SASAKI, G. FALOONA. I. VALVERDE, D. BAETENS, L. ORCI, AND R. UNGEH. Glucagon: role in the hyperglycemia of diabetes mellitus. Science 187: 544-547, 1975. FALOONA, G. R., AND R. H. UNGER. Glucagon. In: Methods of Hormone Radioimmunoassa.y, edited by B. M. Jaffe and H. R. Behrman. New York: Academic, 1974, p. 317-390. GERICH, J. E., M. LORENZI, V. SCHNEIDER, J. KARAN, J. RIVIER, R. GUILLEMIN, AND P. FORSHAM. Effects of somat,ostatin on plasma glucose and glucagon levels in human diabetes mellitus. N. En&. J. Med. 291: 544-547, 1974. HALES, C. N., AND I? J. RANDLE. Immunoassay of insulin with insulin-antibody precipitat,e. Biochem. J. 88: 137-146, 1963. HETENYI, G. JR., AND J. REYNOI~DS. A simple procedure for oounting doubly labeled quenched samples by liquid scintillation spectrometry. Int. J. AppZ. Radia. Isot. 18: 332-334, 1967. Ho, R. J. Radiochemical assay of long-chain fatty acids using ““Ni as tracer. AnpaL. Biochem. 36: 105-113, 1970. ISSEKUTZ, B. JR., M. ALLEN, AND I. BORKWO. Estimation of glucose turnover in the dog with glucose-2-T and glucose-U’“C. Ant. J. Physiol. 222: 710-712, 1972. KATZ, J., AND A. DUNN. Glucose-2-T as a tracer for glucose met#abolism. Biochemistry 6: 1-5, 1967. KELLER, U., J. L. CHIASSON, J. E. LII,JENQUIST, A. D. CHERRINGTON, A. S. JENNINGS, AND 0. B. CROFFORD. The roles of insulin, STEELE.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11. 12.
13. 14.
15. 16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
glucagon and free fatty acids in the regulat,ion of ketogenesis in dogs. Diabetes 26: 1040-1051, 1977. LICKLEY, L., AND M. VRANIC. Effects of selective glucagon deficiency (Abstract). CZin. Res. 24: 66A, 1976. LILJENQUIST, J. E., J. D. BOMBOY, S. B. LEWIS, B. C. SINCLAIR-SMITH, P. W. FELTS, W. W. LACY, 0. B. CROFFORD, AND G. W. LIDDLE. Effects of glucagon on lipolysis and ketogenesis in normal and diabetic men. J. CZin. Inuest. 53: 190-197, 1974. LILJENQUIST, J. E., G. L. MUELLER, A. D, CHERRINGTON, U, KELI,ER, J. L. CHIASSON, J. M. PERRY, W. W. LACY, AND D. RABINOWITZ. Evidence for an important, role of glucagon in the regulation of hepatic glucose production in normal man. 9. CZin. Inuest. 59: 369-374, 1977. MASHITER, K., P. G. HARDING, M. CHOU, G. D. MASHITER, J, STOUT, D. DIAMOND, AND J. B, FIELD. Persistent pancreatic glucagon but not insulin response to arginine in pancreatectomized dogs. Endocrinology 96: 678-693, 1975. MATSUYAMA, T., AT\;D P. P. FOA. Plasma glucose, insulin, pancreatic and enteroglucagon levels in normal and depancreatized dogs. Proc. Sot. Exp. BioZ. 147: 97402, 1974. MORITA, S., K. 1301, C. YIP, AND M. VRANIC. Measurement and partial characterization of immunoreactive glucagon in gastrointestinal t,issues of dogs. Diabetes 25: 1018-1025. 1976. RADZIUK, J., K. NORWICH, AND M. VRANIC. The measurement of glucose turnover in non-steady state and its experimental validation. Am. J. PhysioZ. 234: E84-E93, 1978 or Am. J. Physiol.: EndocrinoZ. Metab. Gastroirztest. Physiol. 3: E84-E93, 1978. RADZIUK, J., K. NORWICH, AND M. VRANIC. Measurement and validation of non-steady turnover rates with application to the insulin and glucose systems Federation Proc. 33: 1855-1864, 1974. REXCHARDT, G. A. JR., N. F. MOURY, N. J. HOCHELLA, A. L. PATTERSON, AND S. WEINHOUSE. Quantitative estimation of Cori cycle in the human. J. BioZ. Chem. 238: 495-501, 1963. Ross, G., L. LICKLEY, AND M. VRANIC. Role of extrapancreatic glucagon in regulation of glucose turnover in depancreatized dogs. Am. J. Physiol. 234: E213-E219 or Am. J. PhysioZ.: EndocrinoZ. Metab. Gastrointest. Physiol. 3: E213-E219, 1978. Ross, G., AND M. VRANIC. Effects of somatostatin induced decreases in immunoreactive insulin (IRI) and glucagon (IRG) on glucose turnover (Abstract). Diabetes 24, Suppl. 2: 408, 1975. SAKURAI, H., R. DOBBS, AND R. H. UNGER. Somatostatin-induced changes in insulin and glucagon secretion in normal and diabetic dogs. J. CZin. Invest. 54: 1395-1402, 1974. SASAKI, H., B. RUBALCAVA, D. BAETENS, E. BLAZQUEZ, C. B. SRIKANT, L. ORCI, AND R. H. UNGER. Identification of glucagon in the gastrointestinal tract. J. Clin. Invest. 56: 135-175, 1975. SHAW, W., T. ISSEKUTZ, AND B. ISSEKUTZ, JR. Gluconeogenesis from glycerol at rest and during exercise in normal, diabetic and methyl prednisolone-treated dogs. MetaboZism 25: 329-339, 1976. SHERWIN, R. S., R. HENDLER, R. DEFRONZO, J. WAHREN, AND P.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.
E262 Glucose homeostasis during prolonged suppression of glucagon and insulin secretion by somatostatin, Proc. NatZ. Acad. Sci. USA 74: 348-352, 1977. 32. VRANIC, M., R. ENGERMAN, K. DOI, S. MORITA, AND C. YIP. Extrapancreatic glucagon in the dog. MetaboZism 25, Suppl. 1: 1469-1473, 1976. 33, VRANIC, M., R. KAWAMORI, S. PEK, N. KOVACEVIC, AND G. A, FELIG.
LICKLEY,
ROSS,
AND
VRANIC
The essentiality of insulin and the role of glucagon in regulating glucose utilization and production during strenuous exercise in dogs. J. Chin. Invest. 57: 245-255, 1976, 34. VRANIC, M., S. PEK, AND R, KAWAMORI. Increased “glucagon immunoreactivity” in plasma of totally depancreatized dogs. Diabetes 23: 905-912, 1974. WRENSHALL.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 21, 2019.