Originals Basic In Vitro Effects of Glycosylated Insulin and Glucagon in Perfused Liver of the Rat T. Ikeda, H. Ochi, I. Ohtani, K. Fujiyama, T. Hoshino, Y. Tanaka, T. Takeuchi and H. Mashiba The First Department of Internal Medicine, Tottori University School of Medicine, Yonago, Japan
Using perfused liver of the rat, the hepatic uptake of glycosylated insulin (GI) and glucagon (GG) and its effects on hepatic glucose output were investigated. Insulin and glucagon were glycosylated in ambient high glucose concentration, and GI80 or GG80 (insulin or glucagon incubated with 0.08 % glucose), GI350 or GG350 (incubated with 0.35 % glucose), and GI1000 or GG1000 (incubated with 1 % glucose) were prepared. The liver was perfused with the medium containing 1000 uU/ml insulin and 200 pg/ml glucagon or 200 uU/ml insulin and l000 pg/ml glucagon. The fractional uptake of insulin or glucagon by perfused liver was not significantly altered by the glycosylation. In the liver perfused with 1000 uU/ml insulin and 200 pg/ml glucagon, glucose output was not changed by the glycosylation of the hormones, while in the liver perfused with 200 uU /ml insulin and l000 pg/ml glucagon, GI1000 reduced its biological activity, as reflected by insulin-mediated decrease in glucose output. These results suggest that in the liver insulin incubated with markedly high concentration of glucose reduces its biological activity at a physiological concentration in the presence of high concentration of glucagon Key words Glycosylated Insulin Glycosylated Glucagon - Perfused Rat Liver - Hepatic Glucose Output
Introduction Nonenzymatic glycosylation can involve many structural and plasma proteins, including peptide hormones (Day, Thorpe and Baynes 1979; McFarland, Catalano, Day, Thorpe and Baynes 1979; Miller, Gravallese and Bunn 1980; Rosenberg, Modrak and Hassing 1979). It has been demonstrated that insulin can also be glycosylated (Anzenbacher and Kalous 1975; Dolhofer and Wieland 1979). The glycosylated insulin showed a reduced biological activity in isolated rat adipocytes (Dolhofer and Wieland 1979) and in man (Lapolla, Tessari, Poli, Valerio, Duner, Iori, Fedele and Crepaldi 1988), suggesting an impaired action of insulin in diabetes mellitus. When insulin is glycosylated in vivo, other peptides, such as glucagon, may also be glycosylated. Glucagon is known to
Horm. metab. Res. 24 (1992) 555-557 © Georg Thieme Verlag Stuttgart • New York
antagonize the insulin effect on the glucose metabolism in the liver, and the molar ratio of insulin to glucagon has an important role in hepatic glucose metabolism. However, no precise data exist regarding the hepatic action of the glycosylated insulin and glucagon. The aim of our study was to elucidate the effect of glycosylated insulin and glucagon on hepatic glucose output. Materials and Methods Animals Male Wistar albino rats weighing approximately 150 g were used in the present study. Materials Dextran T-70 was purchased from Green Cross Co., Osaka, Japan. Bovine serum albumin (Fraction V) and lactic acid were obtained from Sigma Chemical (St. Lotus, MO). DL-a-alanine was purchased from Wako Chemical (Osaka, Japan). Glycosylation of insulin or glucagon Porcine insulin or porcine glucagon (Novo Industry) was glycosylated in ambient high glucose concentration according to Dolhofer and Wieland (1979). In brief, 1 ml of short-acting insulin or glucagon was independently mixed with 3 ml 0.002 N HC1. This solution was further diluted with 32 ml phosphate-buffered saline (pH 7.4), and 10%, 3.5% or 0.8% glucose (4ml) or 0.9% NaCl (4ml) was added. Thus the final glucose concentration was 1 %, 0.35 %, 0.08 % or 0 %. The pH was again adjusted to 7.4. The solution was incubated at 37 °C for 3 days. The amount of glycosylated peptides in the sample was quantified according to Kennedy, Mehl, Riley Merimee (1981) and expressed as nanomoles of 5-hydroxymethylfurfural (HMF) per milligram of protein. Perfusion of the liver The liver of fed rats was perfused as previously described (Ikeda, Mokuda, Tominaga and Mashiba 1989; Ikeda, Yoshida, Honda, Ito, Murakami, Mokuda, Tominaga and Mashiba 1987). After anesthesia with intraperitoneal pentobarbital sodium (40 mg/kg), the abdomen of the rat was opened through a midline incision. The intestines were then placed to the animal's left. The thin strands of connective tissue between the right lobe of the liver and the vena cava were cut, and a loose ligature was placed around the inferior vena cava above the right renal vein, the superior mesenteric and celiac arteries and the portal vein. The portal vein was then cannulated and the perfusion pump was started. The ligatures around the portal vein and the arteries were tied. The thorax was then opened and an outflow cannula was inserted through the right atrium into the thoracic vena cava. Finally, the
Received: 14 Oct. 1991
Accepted: 30 March 1992
Downloaded by: National University of Singapore. Copyrighted material.
Summary
Horm. metab. Res. 24 (1992)
T. Ikeda, H. Ochi, I. Ohtani, K. Fujiyama, T. Hoshino, Y. Tanaka et al.
ligature around the abdominal inferior vena cava was tied, thus closing the circuit. The liver was perfused without recirculation with a synthetic medium at a flow rate of 3.5 ml/min/g liver weight in situ. Perfusion
methods
The perfusate was gassed with 95 % O2 and 5 % CO2, and pH was maintained at 7.4 during perfusion. The liver was perfused with the synthetic medium that consisted of a Krebs-Ringer bicarbonate buffer containing 0.5 % BSA, 4.6 % Dextran T-70, 2 mM alanine, 2mM lactic acid and 5.5 mM glucose. Glucose concentration in the effluent from control liver was almost the same from 15 to 80 min of perfusion time. Therefore, after perfusion with the medium not containing insulin and glucagon for 50 min, the experimental liver was perfused with the medium containing 1000 uU/ml insulin and 200pg/ ml glucagon or 200uU/ml insulin and 1000 pg/ml glucagon for following 30 min. The glucose output for 30 min (from 50 to 80 min of perfusion time) was compared with that for previous 30 min (from 20 to 50 min of perfusion time). The effluent was collected every 5 min and stored at -20 °C until the time of assay. To prevent the glucose metabolism by erythrocytes and the influence on insulin assay by hemolysis, erythrocyte-free medium was used for the present perfusion study. Mean weight of six livers was 7.5 g. Therefore, perfusion flow rate was maintained at 26.5 ml/min.
Effect of glycosylation on insulin and glucagon assay Anti-insulin and anti-glucagon antibodies still bind to glycosylated hormones and binding remains quantitative. Thus, glycosylation of hormones did not impair the radioimmunoassay for insulin and glucagon. Therefore, the concentration of glycosylated hormones was measured by radioimmunoassay. Fractional uptake of glycosylated insulin and glycosylated glucagon by perfused liver As shown in Table 1, there were no significant differences in the fractional uptake of glycosylated insulin (GI) or glycosylated glucagon (GG) at the same concentration, respectively. Table 1 Fractional extraction of insulin and glucagon, and changes in glucose output by perfused rat liver. The values are means ±SD. Each number is 6. insulin (200 u/ml) glucagon (1000 pg/ml)
Calculations Fractional extraction of insulin (glucagon) by the liver was calculated by the formula: [insulin (glucagon) infused for 30 min (from 50 to 80 min of perfusion time) - effluent insulin (glucagon) for 30 min (from 50 to 80 min of perfusion time)] x 100 %/insulin (glucagon) infused for 30 min. The combined effect of insulin and glucagon was calculated by the formula: glucose output from the liver perfused with insulin and glucagon (from 50 min to 80 min of perfusion time) - glucose output from the liver pefused without insulin and glucagon (from 20 min to 50 min of perfusion time). Measurements The oxygen consumption in the liver was calculated from the difference between influent and effluent oxygen concentrations. Glucose concentration was measured by glucose oxidase method. Insulin was measured using the modification of double-antibody method of Morgan and Lazarow (1983). Glucagon was also measured by radioimmunoassay (Nishino, Kodaira, Shin, Imagawa, Shima, Kumahara, Yanaihara and Yanaihara 1981). Statistical
analysis
The values are expressed as means + SD. Statistical comparisons were performed by analysis of variance and two-tailed Student's nonpaired t-test. Results Oxygen consumption in perfused
liver
The oxygen consumption was 1.6-2.0 (J,mol 02/min/g in the liver. In vitro glycosylation
of hormones
The amount of insulin and glucagon glycosylated in vitro at 0..08%, 0.35%, or 1 % glucose was 1.7-2.5, 6.6-9.2, or 9.8-13.9nmol 5HMF, respectively.
hepatic extraction
(%)
changes in glucose output (|xmol/30 min)
Gl0 GG0
38.6±7.1 18.5 ±3.7
Gl80 GG80
insulin (1000 uU/ml) glucagon (200 pg/ml) hepatic extraction
(%)
changes in glucose output (umol/30 min)
267+ 89
22.3 ±3.9 20.1 ±4.1
-175 + 62
38.4 ±7.0 19.1+4.0
310±101
21.0±4.1 19.6±4.0
-168±71
GI350 GG350
36.9 ±7.3 19.5±4.1
309±110
21.7±4.2 22.1 ±3.8
-192 + 91
Gh1000 GG1000
35.7 ±7.2 20.0 ±3.9
467 ±128*
21.6 ±3.7 21.1 ±4.1
-183±79
Glo or GGo: non-glycosylated insulin or glucagon. GIso or GGso: insulin or glucagon incubated with 80 mg/dl glucose. GI350 or GG350: insulin or glucagon incubated with 350 mg/dl glucose. Gh1000 or GG1000: insulin or glucagon incubated with 1000 mg/dl glucose. *p < 0.05, significantly different from the liver perfused with Glo and GGo.
Insulin and glucagon-induced changes in glucose output The infusion of 200(iU/ml insulin and l000 pg /ml glucagon increased glucose output from the liver, and the glucose output from the liver perfused with GI1000and GG1000 was 75 % higher than that perfused with Gl0 and GG0, while the infusion of l000 uU/ml insulin and 200 pg/ml glucagon decreased hepatic glucose output. There were no significant differences in the decrease in glucose output in the group receiving insulin 1000 uU/ml and glucagon 200pg/ml. Discussion In the presence of 200 (iU/ml insulin and 1000 pg/ml glucagon, the increase in glucose output was significantly greater in the liver perfused with GI1000 and GG1000- This suggests that insulin action may be reduced, while glucagon
Downloaded by: National University of Singapore. Copyrighted material.
556
Glycosylation of Insulin
There have been no reports regarding the hepatic extraction of glycosylated glucagon. In the present study, hepatic extraction of glucagon was not changed by the glycosylation, suggesting that hepatic extraction of glucagon may not be impaired by glycosylation. It is unclear whether the present phenomenon has a physiological significance, because the half-life of insulin is very short (Sherwin, Kramer, Tobin, Insel, Liljenquist, Berman and Andres 197'4) and the ambient glucose concentration in diabetic patients is lower than the one decreased effect observed in our experiments. Komjati, Bratusch-Marrain and Waldhausl (1986) have reported that inhibition of hepatic glucose production by an equal total dose of pulsatile insulin infusion was more pronounced than during continuous insulin infusion in perfused liver. Further studies should be necessary. In summary, we conclude that insulin incubated with markedly high concentration of glucose (1000 mg /dl) reduces its biological activity in the liver in the presence of high concentration of glucagon.
557
References Anzenbacher, P., V. Kalous: Binding of d-glucose to insulin. Biochim. Biophys. Acta 386: 603-607 (1975) Caro, J. P, J. M. Amatruda: Evidence for modulation of insulin action and degradation independently of insulin binding. Amer. J. Physiol. 240: E325-E332(1981) Day, J. P, S. R. Thorpe, J. W. Baynes: Nonenzymatically glycosylated albumin. J. Biol. Chem. 254: 595-597 (1979) Dolhofer, R., O. H. Wieland: Preparation and biological properties of glycosylated insulin. FEBS Lett. 100: 133-136 (1979) Honey, R. N., S. Price: The determinants of insulin extraction in the isolated perfused liver. Horm. Metab. Res. 11: 111-117 (1979) Ikeda, T, O. Mokuda, M. Tominaga, H. Mashiba: Glucose intolerance in thyrotoxic rats: role of insulin, glucagon, and epinephrine. Amer. J. Physiol. 255: E843-E849 (1989) Ikeda, T., T. Yoshida, M. Honda, Y Ito, I. Murakami, O. Mokuda, M. Tominaga, H. Mashiba: Effect of intestinal factors on extraction of insulin in perfused rat liver. Amer. J. Physiol. 253: E603-E607 (1987) Kennedy, L., T. D. Mehl, W. J. Riley, T. J. Merimee: Non-enzymatically glycosylated serum protein in diabetes mellitus. An index of shortterm glycemia. Diabetologia 21: 94-98 (1981) Komjati, M., P. Bratusch-Marrain, W. Waldhausl: Superior efficacy of pulsatile versus continuous hormone exposure on hepatic glucose production in vitro. Endocrinology 118: 312-319 (1986) Lapolla, A., P. Tessari, T. Poll, A. Valerio, E. Duner, E. Iori, D. Fedele, G. Crepaldi: Reduced in vivo biological activity of in vitro glycosylated insulin. Diabetes 37: 787-794 (1988) McFarland, K. K, E. W. Catalano, J. F. Day S. R. Thorpe, J. W. Baynes: Nonenzymatic glucosylation of serum proteins in diabetes mellitus. Diabetes 28: 1011-1014 (1979) Miller, J. A., E. Gravallese, H. E Bunn: Nonenzymatic glycosylation of erythrocyte membrane proteins: relevance to diabetes. J. Clin. Invest. 65: 896-901 (1980) Misbin, R. N., T. J. Merimee, J. M. Lowenstein: Insulin removal by isolated perfused rat liver. Am. J. Physiol. 230: 171-177 (1976) Morgan, C. R., A. Lazarow: Immunoassay of insulin: two antibody system. Plasma insulin of normal, subdiabetic, and diabetic rats. Diabetes 12: 115-126(1963) Nishino, I, T. Kodaira, S. Shin, K Imagawa, K. Shima, Y Kumahara, C. Yanaihara, N. Yanaihara: Glucagon radioimmunoassay with use of antiserum to glucagon C-terminal fragment. Clin. Chem. 27: 1690-1697(1981) Rabkin, R., G. M. Reaven, C. E. Mondon: Insulin metabolism by liver, muscle, and kidneys from spontaneously diabetic rats. Amer. J. Physiol. 250: E530-E537 (1986) Rosenberg, H, J. B. Modrak, J. N. Hassing: Glycosylated collagen. Biochem. Biophys. Res. Commun. 91: 498-501 (1979) Shanker, T. P., S. Drake, S. S. Solomon: Insulin resistance and delayed clearance of peptide hormones in cirrhotic rat liver. Amer. X Physiol. 252: E772-E777 (1987) Sherwin, R. S., K. J. Kramer, J. D. Tobin, P. A. Insel, J. E. Liljenquist, M. Berman, R. Andres: A model for kinetics of insulin in man. J. Clin. Invest. 53: 1481-1492 (1974) Requests for reprints should be addressed to: Tadasu Ikeda, M. D. The First Department of Internal Medicine Tottori University School of Medicine Nishi-machi 36-1 Yonago 683 Japan
Downloaded by: National University of Singapore. Copyrighted material.
action increases. In our preliminary perfusion study, the stimulating effect of 200 or l000 pg/ml GG1000 on hepatic glucose output was not significantly different from that of 200 or 1000 pg/ml GGo, respectively. GI1000 may reduce its biological activity at a physiological concentration (200uU/ml). The infusion of l000 uU/ml insulin and 200pg/ml glucagon decreased glucose output similarly in all groups, suggesting that glycosylation of insulin does not reduce its biological activity at a supra-physiological concentration (1000 ) These results were in agreement with the report of Dolhofer and Wieland (1979) who have observed a reduced biological activity of glycosylated insulin (incubated with 4 % glucose) in vitro. To date, the effect of insulin incubated with low concentration of glucose has not been examined. The GI80 or GI350, insulin incubated with 0.08 % or 0.35 % glucose, hardly reduced its biological activity in the present study. When incubated with a markedly high concentration of glucose, insulin may reduce its biological activity. An increase in the perfusate insulin concentration, which is known to reduce insulin uptake in perfused liver {Honey and Price 1979; Misbin, Merimee and Lowenstein 1976), resulted in a lowering of fractional insulin extraction in the present study, suggesting that the present liver is available for evaluating the insulin uptake. Although the fractional uptake of insulin could be a summation of receptor binding, insulin internalization, and non-specific binding, glycosylated insulin reduces its biological activity probably at a postreceptor level because hepatic extraction of insulin was not changed by glycosylation. However, several investigators have suggested that insulin action can be modulated independently of its hepatic extraction (Cam and Amatruda 1981; Rabkin, Reaven and Mondon 1986; Shanker, Drake and Solomon 1987). Glycosylated insulin may reduce its biological activity through other mechanisms.
Horm. metab. Res. 24 (1992)