- Email: [email protected]
Reduced hepatic insulin clearance in rats with dietary-induced obesity
Reduced Hepatic Insulin Clearance in Rats With Dietary-Induced Obesity Gunnar Strijmblad
and Per Bjijrntorp
insulin uptake in the in situ perfused liver from rats that were moderately obese after overfeeding was diminished in comparison with controls. The obese rats had higher levels of portal free fatty acids (FFA) and liver triglyceride contents but not of insulin concentration in the portal vein. There were strong negative correlations between hepatic triglyceride The perfusions were performed with lower FFA concentrations than those contents and insulin clearance (r - 0.8-0.9). found in vivo in the portal vein. It is suggested that the inhibited insulin uptake in the obese rats was due to exposure of these livers in vivo to elevated FFA concentrations, and that this inhibition remained during the experiment and was associated with the triglyceride contents of the livers. It is also suggested that this mechanism was responsible for the moderate peripheral hyperinsulinemia seen in these rats. A mechanism of regulation of insulin uptake in the liver via FFA and liver triglyceride might be of importance in several conditions with hyperinsulinemia and known elevation of portal FFA. and liver tri&ceride contents. B 1986 by Grune & Stratton, Inc.
H
YPERINSULINEMIA is an early and prevalent metabolic abnormality in obesity. Although hypersecretion of insulin has been considered the main cause to hyperinsulinemia, other factors have recently been brought to attention, such as modifications of the hepatic uptake of secreted insulin.’ A major portion of the insulin produced is taken up in the liver before it reaches peripheral circulation, and obviously changes in this uptake would therefore be of major consequences for peripheral concentrations of insulin. Recent studies in human obesity indicate that changes in hepatic uptake of insulin might indeed contribute to peripheral hyperinsulinemia. Based on measurements of connecting (C-) peptide concentrations in relation to insulin concentrations in plasma it has been suggested that peripheral hyperinsulinemia in conditions with low insulin secretion might be caused by diminished hepatic uptake of insulin.**’ Conclusions based on measurements of peripheral concentrations of C-peptide and insulin must be considered as tentative because they rest on a number of assumptions.4 Previous direct studies in the in situ perfused liver have, however, also indicated that obesity is followed by a decreased hepatic uptake of insulin.5+‘7 The studies by Karakash et al5 were performed on mice with a genetic obesity syndrome (ob/ob mice) with severe hyperinsulinemia. The decreased uptake of insulin in the liver was dependent on the degree of hyperinsulinemia and was considered secondary to the elevated secretion of insulin. Recent findings of statistical associations between an increase in lipolytically sensitive intraabdominal adipose tissue in abdominal obesity and hyperinsulinemia6 led to the hypothesis that free fatty acids (FFA) produced by this adipose tissue would interfer with hepatic uptake of insulin. Experiments in the in situ perfused rat liver seemed to support such a hypothesis. In these studies it was found that insulin uptake in the liver was inversely proportional to the FFA concentration in the perfusate in the physiologic range of portal FFA concentrations. Interestingly, very high FFA concentrations not only inhibited the uptake of insulin in the same perfusate, but also caused a long-term inhibition of insulin uptake.7 These results suggested that an accumulated product of FFA in the liver might be involved in the inhibition of insulin uptake. Such a product might be liver triglyceride. Metabolism,
Vol 35,
No 4 (April),
1986:
pp 323-327
In the present work the hepatic uptake of insulin was tested in rats rendered moderately obese by overfeeding. This was done with two purposes: first, to describe further the hepatic insulin uptake in this type of obesity, and second, to elucidate further possible mechanisms regulating the process of hepatic insulin uptake.
MATERIALS
AND
METHODS
Male Spraque-Dawley rats weighing 215 to 240 g were kept at constant humidity and temperature (21 OC to 23 “C) and with 12 hours light. They had free access to water and commercial rat pellets (EWOS, Sijdertllje, Sweden) containing by weight 5% fat, 55% carbohydrates, and 22.5% protein with sufficient minerals and vitamins. Part of the rats were overfed by being supplied with sweet cookies (Maryland cookies) and sweetened-condensed milk (Borden) for 25 days. Liver Perfusion
System
After an overnight fast, the rats were anaestetized with pentobarbital (50 mg/kg body weight intraperitoneally). The liver was kept in situ throughout the perfusion.’ A cannula was inserted into the portal vein for perfusion and the perfusate was collected through another cannula in the inferior vena cava. The perfusion system was kept at an environmental temperature of 38 “C and the rat carcass, containing the in situ perfused liver, was heated with a lamp. The perfusion medium was pumped at a constant rate (2.0 mL/min/g of wet liver) from a 40 mL reservoir to the lung in a recycling system. The lung consisted of 15 feet silastic tubing of I .47 mm internal diameter and 0.25 mm wall thickness (Dow Corning; Midland, Mich) coiled in a jar,’ into which a gas mixture of 95% O2 and 5% CO2 was flowing at a rate of 1 l/min. The perfusion medium consisted of Krebs-Ringer-bicarbonate buffer containing 3% bovine albumin (Sigma; St Louis) and human erythrocytes with the hemotocrite adjusted to 20%. Human blood
From the Clinical Metabolic Laboratory, Department of Medicine I, Sahlgren’s Hospital, University of Giiteborg, Giiteborg. Sweden. Address reprint requests to Dr Per Bjarntorp, Clinical Metabolic Laboratory, Department of Medicine I. Sahlgren’s Hospital. University of Giiteborg. Giiteborg, Sweden. o 1986 by Grune & Stratton, Inc. 0026-0495/86/3504Jl007$03.00/0 323
324
STRiiMBLAD AND BJijRNTORP
Table 1. Weights
of Body and Adipose Tissues
Table 2. Insulin and FFA Concentrations in Portal and Caval
in Overfed Rats and Control Rats Overfed In = 9) Body weight (g)
307 + 7
C0ntr0k (n - 6) 267 + 2
Veins in Overfed and Control Rats Overfed
P < 0.05
Insulin, v cava (pU/mL)
Epididymal fat pad (g)
3.1 * 0.2
1.9 * 0.3
< 0.05
Insulin, v porta (pU/mL)
Retroperitoneal fat pad (g)
2.9 k 0.3
1.5 + 0.3
< 0.05
FFA, v porta (pmol/L)
Mean f SEM.
Controls
9)
(n - 6)
22 k 2
16 + 3
(n =
P < 0.05
48 + 3
41 + 3
NS
550 * 15
364 -t 10
< 0.05
Mean + SEM.
(outdated donor blood) was washed three times with an equal volume of 0.9% NaCl and finally with Krebs Ringer bicarbonate buffer, and pH adjusted to 7.4 before perfusion. Glucose was added to a final concentration of 5.5 mmol/L, and insulin (crystalline, glucagon-free porcine insulin; Eli Lilly, Indianapolis) to a final concentration of 150 f 5 uU/mL (mean t SEM) perfusion medium. There were no differences in the perfusate concentrations of FFA between the two groups before (248 + 40 and 254 + 30, umol/L, obese and control rats, respectively, means f SEM) or after (243 f 33 and 251 t 10) perfusion, and the FFA concentrations did not change during the period of perfusion. Repeated checks of the function of this preparation have been done before,‘,“’ showing a minimal leakage of enzymes and potassium, no swelling, and a physiologic oxygen uptake. Nonfunctioning preparations were identified by elevated perfusion pressures (~15 cm H,O) and/or discoloration of the perfused liver. Measurements were started after an initial period of 15 minutes perfusion to achieve a constant oxygen uptake.” After perfusion, the livers were weighed and a part frozen for later analyses of protein’* and triglyceride” contents after homogenization in 0.25 mol/L sucrose. Retroperitoneal and epididymal fat pads were dissected out and weighed. Glucose was determined enzymatically at the beginning of perfusion (GLOX, KABI; Stockholm, Sweden), FFA at the beginning and end of perfusion according to Dole,” and insulin before and every fifth minute during 30 minutes perfusion by a solid phase radioimmunoassay (Phadebas, Pharmacia; Uppsala, Sweden). Insulin and FFA were also determined in blood samples from v. porta before perfusion. Curves for log insulin concentrations at different times of perfusion were constructed and evaluated for linearity by a computer. Curves with correlation coefficients below 0.95 were rejected if occurring. The hepatic clearance of insulin (k) was calculated from the formula of Mortimore et al’“: (2.3 x V/At) log (Ca/Cf) where k = mL perfusion medium cleared of insulin per minute, Ca and Cf
represent concentrations of insulin at the beginning and at the end of the time interval At, respectively, and V = volume of the medium.
The statistical method used was Student’s t-test. RESULTS
Overfeeding the rats was followed by a body weight increase of about 15% and heavier adipose tissue weights indicating a moderate obesity (Table 1). Table 2 shows the results of determinations of insulin and FFA concentrations in the v. cava and v. porta of the rats before perfusion cannulas were inserted into these veins. The insulin concentrations were higher in the portal vein than in the v. cava. Although insulin concentrations were significantly higher in the v. cava of the overfed animals, there was no significant difference in portal insulin concentrations. FFA concentrations were, however, higher in portal vein of the overfed than in the control rats. The livers of the overfed rats were heavier and contained more triglyceride per whole liver or per mg protein (Fig 1). Figure 2 shows the average relative insulin concentrations during 30 minutes perfusion of livers from obese and control rats. At each concentration the livers of the control rats had extracted significantly more insulin than the livers from obese rats. This is also apparent in Fig 3 where it is seen that insulin clearance is lower in the obese than in the control rats, particularly clear when expressed per weight unit of liver, where the clearance is decreased by about 30% in the livers from the obese animals. When insulin clearance was analyzed in relation to the triglyceride contents of the liver (Fig 4), strong correlations were found between these factors whether total liver trigly-
TRIGLYCERIDE
101
.z
loo-
e
Fig 1. Weight and triglyceride contents of livers from overfed rats (striped columns) and controls (open columns). Means f SEM.
325
REDUCED HEPATIC INSULIN CLEARANCE
100
1.5
0.3
90
1.0
70.
60. OBESE
‘t
5 B .
+ ’
0.2 -i .E
-i .c E
z
CONTROLS 0.5
0.1
10 i
> 5
10
15
20
25
30
Fig 3. Insulin clearance in livers from overfed (striped columns) and control (open columns) rats. Means + SEM.
MIN.
Fig 2. Insulin concentrations in the perfusate of livers from obese and control animals. Values given ES per cent of the 5 minutes perfusion value. Absolute insulin concentrations at the beginning of perfusion: 150 f 5 ulJ/mL, Means t SEM. Comparison between groups significant (P < 0.05) at all time points except 5 minutes.
ceride or liver triglyceride per unit protein were examined, or whether clearance was expressed per whole liver or per unit weight of the liver. DISCUSSION
The results of this study show that in rats rendered moderately obese by overfeeding with fat and carbohydrate, insulin uptake by the in situ perfused liver is diminished. Clearly, if this difference is valid for in vivo conditions it would contribute to the hyperinsulinemia of obesity, and increased insulin concentrations were also found in peripheral plasma in the obese rats examined in the present work. The insulin concentrations in v. porta were, however, not different suggesting that the moderate hyperinsulinemia seen in the obese rats was mainly due to the lower hepatic uptake of insulin. It is of obvious interest to try to understand why livers from diet-induced moderately obese rats take up less insulin than livers from control rats. First, exposure of the liver to high insulin concentrations are known to inhibit further uptake of insulin in the liver.’ The portal insulins of the rats examined were not very high in the fasting condition, and they were, as a matter of fact, not significantly elevated in the obese rats although numerically they were about 20% higher. It seems likely that the obese rats had a more marked
postprandial portal hyperinsulinemia than that seen in the fasting condition. It can therefore not be excluded that the liver was exposed to higher insulin concentrations in vivo in the obese rats, and that this might cause a diminished insulin clearance by the liver under these conditions. The lack of difference in portal insulin concentrations during fasting conditions makes it fairly unlikely, however, that this would be the explanation during the fasting situation, during which the perfusion experiments were performed. It has recently been shown that the concentration of FFA in the perfusate is important for the regulation of the uptake of insulin in the in situ perfused rat liver.’ The FFA concentrations in the perfusate in the present work were similar in both groups of rats, not only before, but also after the perfusion period. Therefore, the livers were exposed to the same concentrations of FFA during the experiment. The obese rats had elevated portal FFA, presumably due to their enlarged adipose tissue mass. It should be noted, however, that the FFA concentration in the perfusate was lower than the portal concentrations of FFA in both groups. This fact suggests that the FFA in the perfusate did not cause the inhibition of liver uptake of insulin in the obese rats; this liver had been exposed to more than double the perfusion concentration in vivo. It seems more likely that the in vivo exposure of the liver to elevated portal FFA had caused a remaining inhibition of insulin uptake during the experiment, in analogy with the results of a previous report, where exposure of the liver to very high FFA concentrations in the perfusate was followed by a remaining inhibition of insulin uptake.’ This remaining inhibition might be associated with liver triglycerides, because excess FFA taken up by the liver are at least partly transferred to liver triglycerides.15 The strong negative correlations between insulin
326
STR6MBLAD
AND BJihNTORP
0
0.20
2.
1,,,,,,,,I/;, 0
0
0
FjJ5 -_.::::
0
0.1
.
.
”
.
46
50
G
130
50
lcm
TRIGLYCERIDE (mg)
TRIGLVCERICE (mg)
TRIGLVCERIOE (rJg. mg pr0t-l)
TRIGLVCERIDE (Ng. mg prot?
Correlations
between
liver triglyceride
Y
X
Triglyceride
.
.
100
Fig 4.
. .
contents
and insulin clearance.
Equation
Clearance
130
r
Ptb
Y = -0.0072x
+ 200
-0.791
< 0.05
Y = -0.0013x
+ 0.260
-0.862
< 0.05
Y = -0.0096x
+ 2.00
-0.843
< 0.05
Y = -0.0015x
+ 0.250
-0.836
< 0.05
(mL x min)
(mg/liver) -
Clearance (mL x g-’
Triglyceride (ug x mg prot-‘) -
x min-‘)
Clearance (mL x min-‘1 Clearance (mL x 9-l
x min-‘)
uptake and the triglyceride contents of the livers in this work suggest that this is indeed the correct alternative. Hyperinsulinemia in human obesity is found particularly strongly associated with localization of excess adipose tissue in the abdominal regions.6 These regions are also sensitive to lipolytic stimulation, and would therefore cause an excessive
exposure of the liver to high FFA.16 When occurring chronitally this might then perhaps cause elevated hepatic triglyceride contents followed by diminished hepatic uptake of insulin, and consequently peripheral hyperinsulinemia. This important possibility should be further tested.
REFERENCES
1. Faber OK, Christensen K, Kehlet H, et al: Decreased insulin removal contributes to hypcrinsulinemia in obesity. J Clin Endocrino1 Metab 53:618-621, 1981 2. Bonora E, Zavaroni I, Coscelli C, et al: Decreased hepatic insulin extraction in subjects with mild glucose intolerance. Metabolism 32:438-446, 1983 3. Krotkiewski M, Rebuff&&rive M, Helm G, et al: Effects of physical training on plasma insulin and connecting (c-) peptide concentrations in obesity and diabetes mellitus type II. (submitted for publication) 4. Faber OK, Hagen C, Binder C, et al: Kinetics of human
connecting peptide in normal and diabetic subjects. J Clin Invest 62:197-203,1978 5. Karakash C, Assimacopoulos-Jeannet F, Jeanrenaud B: An anomaly of insulin removal in perfused livers of obese-hyperglycemic (ob/ob) mice. J Clin Invest 16:173-177, 1976 6. Krotkiewski M, Bjijrntorp P, Sjijstriim L. et al: Impact of Obesity on Metabolism in Men and Women. J CJin Invest 72:11501162.1983 7. Strijmblad G, Wirth A, Svedberg J, et al: Reduced hepatic insulin clearance by elevated fatty acids in the portal vein. J Clin Invest. (submitted for publication)
REDUCED HEPATIC INSULIN
CLEARANCE
8. Assimacopoulos-Jeannet F, Exton JH, Jeanrenaud B: Control of gluconeogenesis and glycogenolysis in perfused livers of normal mice. Am J Physiol 2252-32, 1973 9. Hamilton RL, Berry MN, Williams MC, et al: A simple and inexpensive membrane lung for small organ perfusion. J Lipid Res 15:152-186, 1974 10. Wirth A, Holm G, Bjiirntorp P: Effect of physical training on insulin uptake by the perfused rat liver. Metabolism 31:457-462, 1982 11. Dole VP: A relation between non-esterified fatty acids in plasma and the metabolism of glucose. J Clin Invest 33:150-155, 1956 12. Lowry OH, Rosebrough NJ, Farr AL, et al: Protein measurement with the Folin phenol reagent. J Biol Chem 193:265-275, 195i
327
13. Carbon LA: Determination of serum glycerides. Acta Sot Med (Uppsala) 64:208-213,1959 14. Mortimore GE, Tietze F, Stettin D: Metabolism of insulin I”’ studied in isolated perfused rat liver and hind limb preparations. Diabetes 8:307-314, 1959 15. Carbon LA, Boberg J, Hiigstedt B: Some physiological and clinical implications of lipid mobilization from adipose tissue, in Cahill G, Jr, Renold A-E (eds): Handbook of Physiology. American Physiol Sot, 1965, pp 6235-6244 16. Bjijrntorp P: Adipose tissue in obesity, in Hirsch J, van Itallie T (eds): Proceedings in Obesity Research IV. New York, International Congress on Obesity, London, Libbey, 1984 (in press) 17. Weiland D, Mondon CE, Reaven GM: Prolongation of Insulin Removal by Perfused Liver from Spontaneously Obese Rats. Diabetologia 16:173-177, 1979
and Per Bjijrntorp
insulin uptake in the in situ perfused liver from rats that were moderately obese after overfeeding was diminished in comparison with controls. The obese rats had higher levels of portal free fatty acids (FFA) and liver triglyceride contents but not of insulin concentration in the portal vein. There were strong negative correlations between hepatic triglyceride The perfusions were performed with lower FFA concentrations than those contents and insulin clearance (r - 0.8-0.9). found in vivo in the portal vein. It is suggested that the inhibited insulin uptake in the obese rats was due to exposure of these livers in vivo to elevated FFA concentrations, and that this inhibition remained during the experiment and was associated with the triglyceride contents of the livers. It is also suggested that this mechanism was responsible for the moderate peripheral hyperinsulinemia seen in these rats. A mechanism of regulation of insulin uptake in the liver via FFA and liver triglyceride might be of importance in several conditions with hyperinsulinemia and known elevation of portal FFA. and liver tri&ceride contents. B 1986 by Grune & Stratton, Inc.
H
YPERINSULINEMIA is an early and prevalent metabolic abnormality in obesity. Although hypersecretion of insulin has been considered the main cause to hyperinsulinemia, other factors have recently been brought to attention, such as modifications of the hepatic uptake of secreted insulin.’ A major portion of the insulin produced is taken up in the liver before it reaches peripheral circulation, and obviously changes in this uptake would therefore be of major consequences for peripheral concentrations of insulin. Recent studies in human obesity indicate that changes in hepatic uptake of insulin might indeed contribute to peripheral hyperinsulinemia. Based on measurements of connecting (C-) peptide concentrations in relation to insulin concentrations in plasma it has been suggested that peripheral hyperinsulinemia in conditions with low insulin secretion might be caused by diminished hepatic uptake of insulin.**’ Conclusions based on measurements of peripheral concentrations of C-peptide and insulin must be considered as tentative because they rest on a number of assumptions.4 Previous direct studies in the in situ perfused liver have, however, also indicated that obesity is followed by a decreased hepatic uptake of insulin.5+‘7 The studies by Karakash et al5 were performed on mice with a genetic obesity syndrome (ob/ob mice) with severe hyperinsulinemia. The decreased uptake of insulin in the liver was dependent on the degree of hyperinsulinemia and was considered secondary to the elevated secretion of insulin. Recent findings of statistical associations between an increase in lipolytically sensitive intraabdominal adipose tissue in abdominal obesity and hyperinsulinemia6 led to the hypothesis that free fatty acids (FFA) produced by this adipose tissue would interfer with hepatic uptake of insulin. Experiments in the in situ perfused rat liver seemed to support such a hypothesis. In these studies it was found that insulin uptake in the liver was inversely proportional to the FFA concentration in the perfusate in the physiologic range of portal FFA concentrations. Interestingly, very high FFA concentrations not only inhibited the uptake of insulin in the same perfusate, but also caused a long-term inhibition of insulin uptake.7 These results suggested that an accumulated product of FFA in the liver might be involved in the inhibition of insulin uptake. Such a product might be liver triglyceride. Metabolism,
Vol 35,
No 4 (April),
1986:
pp 323-327
In the present work the hepatic uptake of insulin was tested in rats rendered moderately obese by overfeeding. This was done with two purposes: first, to describe further the hepatic insulin uptake in this type of obesity, and second, to elucidate further possible mechanisms regulating the process of hepatic insulin uptake.
MATERIALS
AND
METHODS
Male Spraque-Dawley rats weighing 215 to 240 g were kept at constant humidity and temperature (21 OC to 23 “C) and with 12 hours light. They had free access to water and commercial rat pellets (EWOS, Sijdertllje, Sweden) containing by weight 5% fat, 55% carbohydrates, and 22.5% protein with sufficient minerals and vitamins. Part of the rats were overfed by being supplied with sweet cookies (Maryland cookies) and sweetened-condensed milk (Borden) for 25 days. Liver Perfusion
System
After an overnight fast, the rats were anaestetized with pentobarbital (50 mg/kg body weight intraperitoneally). The liver was kept in situ throughout the perfusion.’ A cannula was inserted into the portal vein for perfusion and the perfusate was collected through another cannula in the inferior vena cava. The perfusion system was kept at an environmental temperature of 38 “C and the rat carcass, containing the in situ perfused liver, was heated with a lamp. The perfusion medium was pumped at a constant rate (2.0 mL/min/g of wet liver) from a 40 mL reservoir to the lung in a recycling system. The lung consisted of 15 feet silastic tubing of I .47 mm internal diameter and 0.25 mm wall thickness (Dow Corning; Midland, Mich) coiled in a jar,’ into which a gas mixture of 95% O2 and 5% CO2 was flowing at a rate of 1 l/min. The perfusion medium consisted of Krebs-Ringer-bicarbonate buffer containing 3% bovine albumin (Sigma; St Louis) and human erythrocytes with the hemotocrite adjusted to 20%. Human blood
From the Clinical Metabolic Laboratory, Department of Medicine I, Sahlgren’s Hospital, University of Giiteborg, Giiteborg. Sweden. Address reprint requests to Dr Per Bjarntorp, Clinical Metabolic Laboratory, Department of Medicine I. Sahlgren’s Hospital. University of Giiteborg. Giiteborg, Sweden. o 1986 by Grune & Stratton, Inc. 0026-0495/86/3504Jl007$03.00/0 323
324
STRiiMBLAD AND BJijRNTORP
Table 1. Weights
of Body and Adipose Tissues
Table 2. Insulin and FFA Concentrations in Portal and Caval
in Overfed Rats and Control Rats Overfed In = 9) Body weight (g)
307 + 7
C0ntr0k (n - 6) 267 + 2
Veins in Overfed and Control Rats Overfed
P < 0.05
Insulin, v cava (pU/mL)
Epididymal fat pad (g)
3.1 * 0.2
1.9 * 0.3
< 0.05
Insulin, v porta (pU/mL)
Retroperitoneal fat pad (g)
2.9 k 0.3
1.5 + 0.3
< 0.05
FFA, v porta (pmol/L)
Mean f SEM.
Controls
9)
(n - 6)
22 k 2
16 + 3
(n =
P < 0.05
48 + 3
41 + 3
NS
550 * 15
364 -t 10
< 0.05
Mean + SEM.
(outdated donor blood) was washed three times with an equal volume of 0.9% NaCl and finally with Krebs Ringer bicarbonate buffer, and pH adjusted to 7.4 before perfusion. Glucose was added to a final concentration of 5.5 mmol/L, and insulin (crystalline, glucagon-free porcine insulin; Eli Lilly, Indianapolis) to a final concentration of 150 f 5 uU/mL (mean t SEM) perfusion medium. There were no differences in the perfusate concentrations of FFA between the two groups before (248 + 40 and 254 + 30, umol/L, obese and control rats, respectively, means f SEM) or after (243 f 33 and 251 t 10) perfusion, and the FFA concentrations did not change during the period of perfusion. Repeated checks of the function of this preparation have been done before,‘,“’ showing a minimal leakage of enzymes and potassium, no swelling, and a physiologic oxygen uptake. Nonfunctioning preparations were identified by elevated perfusion pressures (~15 cm H,O) and/or discoloration of the perfused liver. Measurements were started after an initial period of 15 minutes perfusion to achieve a constant oxygen uptake.” After perfusion, the livers were weighed and a part frozen for later analyses of protein’* and triglyceride” contents after homogenization in 0.25 mol/L sucrose. Retroperitoneal and epididymal fat pads were dissected out and weighed. Glucose was determined enzymatically at the beginning of perfusion (GLOX, KABI; Stockholm, Sweden), FFA at the beginning and end of perfusion according to Dole,” and insulin before and every fifth minute during 30 minutes perfusion by a solid phase radioimmunoassay (Phadebas, Pharmacia; Uppsala, Sweden). Insulin and FFA were also determined in blood samples from v. porta before perfusion. Curves for log insulin concentrations at different times of perfusion were constructed and evaluated for linearity by a computer. Curves with correlation coefficients below 0.95 were rejected if occurring. The hepatic clearance of insulin (k) was calculated from the formula of Mortimore et al’“: (2.3 x V/At) log (Ca/Cf) where k = mL perfusion medium cleared of insulin per minute, Ca and Cf
represent concentrations of insulin at the beginning and at the end of the time interval At, respectively, and V = volume of the medium.
The statistical method used was Student’s t-test. RESULTS
Overfeeding the rats was followed by a body weight increase of about 15% and heavier adipose tissue weights indicating a moderate obesity (Table 1). Table 2 shows the results of determinations of insulin and FFA concentrations in the v. cava and v. porta of the rats before perfusion cannulas were inserted into these veins. The insulin concentrations were higher in the portal vein than in the v. cava. Although insulin concentrations were significantly higher in the v. cava of the overfed animals, there was no significant difference in portal insulin concentrations. FFA concentrations were, however, higher in portal vein of the overfed than in the control rats. The livers of the overfed rats were heavier and contained more triglyceride per whole liver or per mg protein (Fig 1). Figure 2 shows the average relative insulin concentrations during 30 minutes perfusion of livers from obese and control rats. At each concentration the livers of the control rats had extracted significantly more insulin than the livers from obese rats. This is also apparent in Fig 3 where it is seen that insulin clearance is lower in the obese than in the control rats, particularly clear when expressed per weight unit of liver, where the clearance is decreased by about 30% in the livers from the obese animals. When insulin clearance was analyzed in relation to the triglyceride contents of the liver (Fig 4), strong correlations were found between these factors whether total liver trigly-
TRIGLYCERIDE
101
.z
loo-
e
Fig 1. Weight and triglyceride contents of livers from overfed rats (striped columns) and controls (open columns). Means f SEM.
325
REDUCED HEPATIC INSULIN CLEARANCE
100
1.5
0.3
90
1.0
70.
60. OBESE
‘t
5 B .
+ ’
0.2 -i .E
-i .c E
z
CONTROLS 0.5
0.1
10 i
> 5
10
15
20
25
30
Fig 3. Insulin clearance in livers from overfed (striped columns) and control (open columns) rats. Means + SEM.
MIN.
Fig 2. Insulin concentrations in the perfusate of livers from obese and control animals. Values given ES per cent of the 5 minutes perfusion value. Absolute insulin concentrations at the beginning of perfusion: 150 f 5 ulJ/mL, Means t SEM. Comparison between groups significant (P < 0.05) at all time points except 5 minutes.
ceride or liver triglyceride per unit protein were examined, or whether clearance was expressed per whole liver or per unit weight of the liver. DISCUSSION
The results of this study show that in rats rendered moderately obese by overfeeding with fat and carbohydrate, insulin uptake by the in situ perfused liver is diminished. Clearly, if this difference is valid for in vivo conditions it would contribute to the hyperinsulinemia of obesity, and increased insulin concentrations were also found in peripheral plasma in the obese rats examined in the present work. The insulin concentrations in v. porta were, however, not different suggesting that the moderate hyperinsulinemia seen in the obese rats was mainly due to the lower hepatic uptake of insulin. It is of obvious interest to try to understand why livers from diet-induced moderately obese rats take up less insulin than livers from control rats. First, exposure of the liver to high insulin concentrations are known to inhibit further uptake of insulin in the liver.’ The portal insulins of the rats examined were not very high in the fasting condition, and they were, as a matter of fact, not significantly elevated in the obese rats although numerically they were about 20% higher. It seems likely that the obese rats had a more marked
postprandial portal hyperinsulinemia than that seen in the fasting condition. It can therefore not be excluded that the liver was exposed to higher insulin concentrations in vivo in the obese rats, and that this might cause a diminished insulin clearance by the liver under these conditions. The lack of difference in portal insulin concentrations during fasting conditions makes it fairly unlikely, however, that this would be the explanation during the fasting situation, during which the perfusion experiments were performed. It has recently been shown that the concentration of FFA in the perfusate is important for the regulation of the uptake of insulin in the in situ perfused rat liver.’ The FFA concentrations in the perfusate in the present work were similar in both groups of rats, not only before, but also after the perfusion period. Therefore, the livers were exposed to the same concentrations of FFA during the experiment. The obese rats had elevated portal FFA, presumably due to their enlarged adipose tissue mass. It should be noted, however, that the FFA concentration in the perfusate was lower than the portal concentrations of FFA in both groups. This fact suggests that the FFA in the perfusate did not cause the inhibition of liver uptake of insulin in the obese rats; this liver had been exposed to more than double the perfusion concentration in vivo. It seems more likely that the in vivo exposure of the liver to elevated portal FFA had caused a remaining inhibition of insulin uptake during the experiment, in analogy with the results of a previous report, where exposure of the liver to very high FFA concentrations in the perfusate was followed by a remaining inhibition of insulin uptake.’ This remaining inhibition might be associated with liver triglycerides, because excess FFA taken up by the liver are at least partly transferred to liver triglycerides.15 The strong negative correlations between insulin
326
STR6MBLAD
AND BJihNTORP
0
0.20
2.
1,,,,,,,,I/;, 0
0
0
FjJ5 -_.::::
0
0.1
.
.
”
.
46
50
G
130
50
lcm
TRIGLYCERIDE (mg)
TRIGLVCERICE (mg)
TRIGLVCERIOE (rJg. mg pr0t-l)
TRIGLVCERIDE (Ng. mg prot?
Correlations
between
liver triglyceride
Y
X
Triglyceride
.
.
100
Fig 4.
. .
contents
and insulin clearance.
Equation
Clearance
130
r
Ptb
Y = -0.0072x
+ 200
-0.791
< 0.05
Y = -0.0013x
+ 0.260
-0.862
< 0.05
Y = -0.0096x
+ 2.00
-0.843
< 0.05
Y = -0.0015x
+ 0.250
-0.836
< 0.05
(mL x min)
(mg/liver) -
Clearance (mL x g-’
Triglyceride (ug x mg prot-‘) -
x min-‘)
Clearance (mL x min-‘1 Clearance (mL x 9-l
x min-‘)
uptake and the triglyceride contents of the livers in this work suggest that this is indeed the correct alternative. Hyperinsulinemia in human obesity is found particularly strongly associated with localization of excess adipose tissue in the abdominal regions.6 These regions are also sensitive to lipolytic stimulation, and would therefore cause an excessive
exposure of the liver to high FFA.16 When occurring chronitally this might then perhaps cause elevated hepatic triglyceride contents followed by diminished hepatic uptake of insulin, and consequently peripheral hyperinsulinemia. This important possibility should be further tested.
REFERENCES
1. Faber OK, Christensen K, Kehlet H, et al: Decreased insulin removal contributes to hypcrinsulinemia in obesity. J Clin Endocrino1 Metab 53:618-621, 1981 2. Bonora E, Zavaroni I, Coscelli C, et al: Decreased hepatic insulin extraction in subjects with mild glucose intolerance. Metabolism 32:438-446, 1983 3. Krotkiewski M, Rebuff&&rive M, Helm G, et al: Effects of physical training on plasma insulin and connecting (c-) peptide concentrations in obesity and diabetes mellitus type II. (submitted for publication) 4. Faber OK, Hagen C, Binder C, et al: Kinetics of human
connecting peptide in normal and diabetic subjects. J Clin Invest 62:197-203,1978 5. Karakash C, Assimacopoulos-Jeannet F, Jeanrenaud B: An anomaly of insulin removal in perfused livers of obese-hyperglycemic (ob/ob) mice. J Clin Invest 16:173-177, 1976 6. Krotkiewski M, Bjijrntorp P, Sjijstriim L. et al: Impact of Obesity on Metabolism in Men and Women. J CJin Invest 72:11501162.1983 7. Strijmblad G, Wirth A, Svedberg J, et al: Reduced hepatic insulin clearance by elevated fatty acids in the portal vein. J Clin Invest. (submitted for publication)
REDUCED HEPATIC INSULIN
CLEARANCE
8. Assimacopoulos-Jeannet F, Exton JH, Jeanrenaud B: Control of gluconeogenesis and glycogenolysis in perfused livers of normal mice. Am J Physiol 2252-32, 1973 9. Hamilton RL, Berry MN, Williams MC, et al: A simple and inexpensive membrane lung for small organ perfusion. J Lipid Res 15:152-186, 1974 10. Wirth A, Holm G, Bjiirntorp P: Effect of physical training on insulin uptake by the perfused rat liver. Metabolism 31:457-462, 1982 11. Dole VP: A relation between non-esterified fatty acids in plasma and the metabolism of glucose. J Clin Invest 33:150-155, 1956 12. Lowry OH, Rosebrough NJ, Farr AL, et al: Protein measurement with the Folin phenol reagent. J Biol Chem 193:265-275, 195i
327
13. Carbon LA: Determination of serum glycerides. Acta Sot Med (Uppsala) 64:208-213,1959 14. Mortimore GE, Tietze F, Stettin D: Metabolism of insulin I”’ studied in isolated perfused rat liver and hind limb preparations. Diabetes 8:307-314, 1959 15. Carbon LA, Boberg J, Hiigstedt B: Some physiological and clinical implications of lipid mobilization from adipose tissue, in Cahill G, Jr, Renold A-E (eds): Handbook of Physiology. American Physiol Sot, 1965, pp 6235-6244 16. Bjijrntorp P: Adipose tissue in obesity, in Hirsch J, van Itallie T (eds): Proceedings in Obesity Research IV. New York, International Congress on Obesity, London, Libbey, 1984 (in press) 17. Weiland D, Mondon CE, Reaven GM: Prolongation of Insulin Removal by Perfused Liver from Spontaneously Obese Rats. Diabetologia 16:173-177, 1979