Effects of Insulin-Like Growth Factor-I on Insulin and Glucagon Release from Isolated Perfused Rat Pancreas

Transcription

1 Effects of Insulin-Like Growth Factor-I on Insulin and Glucagon Release from Isolated Perfused Rat Pancreas KOICHI KAWAI, SEIJI SUZUKI, KAZUE TAKANO*, NAOMI HIZUKA*, YASUKO WATANABE AND KAMEJIRO YAMASHITA Department of Internal Medicine, Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Ibaraki 305, Japan *Department of Medicine, Institute of Clinical Endocrinology, Tokyo Women's Medical College, Tokyo 162, Japan Abstract The infusion of recombinant insulin-like growth factor I (IGF-I) causes a decrease in plasma glucose and insulin. In this study we examined whether or not IGF-I directly affects islet hormone release by means of a rat pancreas perfusion system. A superphysiologic concentration of IGF-I (2nM) elicited a slight but significant decrease in insulin release under the perfusate glucose concentration of 120mg/dl. A pharmacological concentration of IGF- I (200nM) significantly suppressed the increase in insulin release in response to an increase in the perfusate glucose concentration (from 4.5mM to 12.8 mm), but did not affect the decrease in insulin release in response to a decrease in the perfusate glucose concentration (from 6.9mM to 2.8mM). Glucagon release was not influenced by IGF-I in these experiments.. These results suggest that IGF-I potentially inhibits the insulin release from islet B-cells directly, but its pathophysiological significance may be slight considering its partial inhibition at superphysiologic concentrations and its stable plasma level. Insulin-like growth factor I (IGF-I; somatomedin C) mediates the somatotrophic actions of growth hormone and is expected to be a therapeutic tool for increasing the height of patients of short stature because large amounts of it have been produced by a recombinant DNA technique. Recombinant human IGF-I infused into hypophysectomized rats and insulin deficient Received July 3, 1990 Address for correspondence: KOICHI KAWAI, M. D. & Ph. D. Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Ibaraki 305, Japan diabetic rats stimulated their growth (Schoenle et al., 1982; Scheiwiller et al., 1986). However, it also exerts metabolic actions that closely resemble those of insulin; IGF-I caused a decrease in plasma glucose and insulin levels in animals and man (Zapf et al., 1986; Guler et al., 1989; Takano et al., 1990a). The results with the glucose clamp technique in rats suggest that the decrease in plasma insulin is derived from the direct effect of IGF-I on insulin secretion from islet B-cells (Jacob et al., 1989). Insulin release from cultured fetal islets is reduced by adding IGF-I to

2 KAWAI et al. Endocrinol. Japon. December 1990 the culture medium (Otonkoski et al., 1988). However, it is probable that IGF-I stimulates the glucose uptake in peripheral tissues and the reduced plasma glucose thereby suppresses the insulin release. We therefore studied the effects of IGF-I on islet hormone release from isolated perfuseed rat pancreases to evaluate the direct effects of IGF-I on islet hormone release. Materials and Methods Pancreas perfusion The pancreases were isolated from male Wistar rats, weighing about 250g, under pentobarbital anesthesia after an overnight fast. The isolated pancreas was perfused by the method of Grodsky et al. (1963) slightly modified. A Krebs-Ringer bicarbonate solution containing 4% dextran T-70 (Pharmacia Fine Chemicals, Uppsala, Sweden), 0.2% bovine serum albumin (BSA), 5mM each of pyruvate, fumarate and glutamate, glucose (its concentration is shown in the figure legends), and 3mM arginine was equilibrated with a 95% 02-5% CO2 mixture at 37 Ž. The flow rate of the perfusate was 2ml/ min. Recombinant human IGF-I donated by the Fujisawa Pharmaceutical Co., Ltd. (Osaka, Japan) was dissolved in 0.9% saline containing 0.2% BSA and glucose at the same concentration as that of the perfusate, and administered through a side-arm syringe at a rate of 0.1 ml/ min. The diluent was infused in control experiments. A glucose solution dissolved in 0.9% saline containing 0.2% BSA was administered through a side-arm syringe in the experiment to change the glucose concentration of the perfusate during the experiment (Figs. 2 and 3) at a rate of 0.1ml/min. The venous perfusate was collected at 1-min intervals and stored at -40 Ž until the assay. Assays Insulin (IRI) and glucagon (IRG) were assayed by RIA, according to the methods of Herbert et al. (1965) and Faloona and Unger (1974) with E-7 antibody (kindly donated by Dr. H. von Schenck), respectively. In the insulin assay, porcine insulin was used as a standard and 200 nm IGF-I did not crossreact in this insulin assay system at all. In our laboratory, the interassay variations were 6.1% and 10.9% for insulin and glucagon, respectively, and the intraassay variations were 7.3% and 6.5%, respectively. Statistics The data were analyzed by analysis of variance, and the Newman-Keuls procedure for multiple comparisons was performed (Fig. 1). The degree of decrease in total output of hormone release from its preceding basal level was calculated by measuring the area under the curves in Figs. 1 and 2, and compared by Student's t-test for nonpaired data (Table 1). All data herein are expressed as the mean }SEM. A P value of 0.05 or less was considered signi- Results Effects of IGF-I on insulin and glucagon release from isolated perfused rat pancreas under the constant perfusate conditions The infusion of IGF-I (2 nm and 200 nm) caused a significant decrease in insulin and glucagon release under the pefusate condition of 125mg/dl glucose plus 3nM arginine (Fig. 1). However, the infusion of diluent also suppressed insulin and glucagon release (Fig.1). When the decreases in the total outputs of insulin and glucagon from their preceding basal levels due to IGF-I infusion was compared with those due to diluent (Table 1-A), a significant decrease was obtained only in insulin release during the 2nM IGF-I infusion. Under the low glucose condition (50mg/dl glucose plus 3mM arginine), the basal IRI level was low (6.0 }0.6ƒÀU/ml,N=7) and a significant decrease from its preceding basal level was scarcely obtained during the infusion of 2nM IGF-I (data not shown). The basal glucagon level was very high in this perfusate condition (1907 }335 pg/ml, N=7), and it sharply decreased from its high starting level. Therefore, no significant

3 Vol.37, No.6 IGF-I AND PANCREATIC HORMONE RELEASE IGF-1 Fig.1. Effects of IGF- I on insulin and glucagon release from isolated perfused rat pancreas. Basal perfusate contained 125 mg/dl glucose and 3 mm arginine. *p< 0.05 vs. mean of preceding basal level. A: Fig. 1. Table1. Statistical calculations for Fig. 1 and Fig. 2

4 KAWAI et al. Endocrinol. Japon. December 1990 B: Fig. 2. change was obtained during the infusion of IGF-I (data not shown). Effects of IGF-I on insulin and glucagon release in response to change in glucose concentration in the perfusate In this study, the glucose concentration in the perfusate changed in the presence and absence of IGF-I infusion. The order of infusion of IGF-I and its diluent was performed alternately (Figs. 2 and 3, A and B) to exclude changes in insulin and glucagon release during the experiment irre- IGF-1 (200 nm) Fig.2. Changes in insulin and glucagon release from isolated perfused rat pancreas in response to increase in glucose concentration of perfusate in the presence or absence of IGF-I.

5 Vol.37, No.6 IGF-I AND PANCREATIC HORMONE RELEASE spective of the effects of IGF-I. When the glucose concentration of the perfusate was increased to 230mg/dl from its basal level of 80mg/dl, the insulin release increased in a biphasic manner and glucagon release decreased. The degree of these changes was the same when 200nM IGF-I was coinfused in the first place (Fig. 2-A). On the other hand, when the diluent was coinfused from the first, the degree of increase in insulin release was significantly less in the presence of 200nM IGF-I (Fig 2-B, Table 1-B). The decrease in the glucagon level did not change significantly. When the glucose concentration of the perfusate was decreased to 50mg/dl from its basal level of 125mg/di, insulin release was suppressed and glucagon release was enhanced. The degree of these changes was not influenced by coinfusion of 200nM IGF-I (Fig. 3). Fig.3. Changes in insulin and glucagon release in response to the decrease in the glucose concentration of perfusate in the presence or absence of IGF-1.

6 KAWAI et al. Discussion The present study with a rat pancreas perfusion system demonstrates that IGF-I slightly inhibits insulin release under the physiological hyperglycemic condition (Fig. 1) and insulin release in response to an increase in the glucose concentration of the perfusate (Fig.2). Significant inhibition of insulin release in response to the increase in the perfusate glucose concentration was obtained only when the diluent was coinfused from the first (Fig. 2-B). As an explanation of this result, we suppose that the insulin response to the second infusion of glucose is weaker than that to the first infusion in the absence of IGF-I throughout the experiment. Therefore, when the IGF-I was coinfused from the first (Fig. 2-A), a similar degree of insulin response must have been obtained due to slight inhibition of insulin release which responded to the Endocrinol. Japon. December 1990 ma level of IGF-I is almost constant all day long, and the serum IGF-I concentration These results are essentially consistent extracted with acid-ethanol is around 200 with the results reported by Leahy and ng/ml (Takano et al., 1990a). Less than Vandekerkhove (1990) during the preparation 5 percent of the total serum IGF-I seems of this report. They reported that 2ng/ml to be an active form freed from its binding of IGF-I (0.25nM) significantly suppresses protein (Hizuka et al. submitted for publication). Therefore, this inhibition should insulin release and 200ng/ml (25nM) of IGF-I decreased it to 35% of the basal not play a significant role in nutrient homeostasis. In its therapeutic use, its adminis- level. However, some of their conclusions seems to have resulted from an improperly tration might cause a significant suppression prepared experiment and insufficient control of insulin release concomitant with the decrease in plasma glucose as mentioned in studies. Therefore, we conclude that superphysiologic concentrations of IGF-I (the the introduction. The infusion of recombinant IGF-I in normal human subjects for serum IGF-I level freed from its binding protein) partially inhibit the augmented 6 days (resulting plasma IGF-I level, about insulin release brought on by high glucose 700 ng/ml) did not cause a change in blood conditions, probably through a direct action glucose or fasting insulin, although fasting on islet B-cells. A recent study has demonstrated the presence of high affinity receptor for IGF-I on rat pancreatic B cells (Van Schravendijk et al., 1987). The present study does not fully explain the results brought on by an in vivo study with the glucose clamp technique in rats, in which plasma insulin was suppressed to an undetectable level during the infusion of IGF-I and plasma IGF-I was ng/ml (Jacob et al., 1989). A similar discrepancy is observed in insulin-insulin feedback. In in vivo studies the exogenous insulin clearly inhibits endogenous insulin secretion and decreased plasma C-peptide (Sando et al., 1970; Liljenquist et al., 1978 ; DeFronzo et al., 1981). On the other hand, a pharmacological amount of insulin does not inhibit insulin release from the isolated perfused pancreas (Marincola et al., 1983). This discrepancy has been explained by an experiment showing that indirect neural control plays an important role in the insulin-insulin auto-feedback mechanism (Stagner et al., 1986). We think this explanation also explain the discrepancy between in vivo and in vitro studies concerning IGF-I's inhibition of insulin secretion. Under physiological conditions, the plas- plasma C-peptide was suppressed (Guler et al., 1989). In our study, the sc injection of IGF-I (0.1mg/kg) 30 min after breakfast for 7 days caused a suppression of the increase in plasma insulin after the meal without effecting plasma glucose in healthy volunteers (Takano et al., 1990b). Thus,

7 Vol. 37, No. 6 IGF-I AND PANCREATIC HORMONE RELEASE the direct suppression of insulin release by IGF-I seems to contribute to the defence against the hypoglycemic effect of IGF-I in its therapeutic use. Acknowledgements This report was supported by a research grant from the Foundation for Growth Science (Japan). We thank Miss Junko Tanaka for her expert technical assistance. References DeFronzo, R. A., C. Binder, J. Wahren, P. Felig, E. Ferrannini and O. K. Faber (1981). Sensitivity of insulin secretion to feedback inhibition by hyperinsulinemia. Acta Endocrinol.(Copenh) 98, Faloona, G. R. and R. H. Unger (1974). Glucagon. In: Methods of Hormone Radioimmunoassay (B. M. Jaffe and H. R. Behrman eds.), Academic Press, New York. pp Grodsky, G. M., A. A. Batt, L. L. Bennett, C. Vcella, N. B. McWilliams and D. F. Smith (1963). Effects of carbohydrates on secretion of insulin from isolated rat pancreas. Am. J. Physiol. 2:5, Guler, H.-P., C. Schmid, J. Zapf and E. R. Froesch (1989). Effects of recombinant insulinlike growth factor I on insulin secretion and renal function in normal human subjects. Proc. Natl. Acad. Sci. USA. 86, Herbert, V., K.-S. Lau, C. W. Gottlieb and S. J. Bleicher (1965). Coated charcoal immunoassay of insulin. J. Clin. Endocrinol. Metab. 25, Jacob, R., E. Barrett, G. Plewe, K. D. Fagin and R. S. Sherwin (1989). Acute effects of insulin-like growth factor I on glucose and amino acid metabolism in the awake fasted rat. J. Clin. Invest. 83, Leahy, J. L. and K. M. Vandekerkhove (1990). Insulin-like growh factor-i at physiological concentrations is a potent inhibitor of insulin secretion. Endocrinology 126, Liljenquist, J. E., D. L. Horwitz, A. S. Jennings, J. L. Chiasson, U. Keller and A. H. Rubenstein (1978). Inhibition of insulin secretion by exogenous insulin in normal man as demonstrated by C-peptide assay. Diabetes 27, Marincola, F., W. Frank, W. Clark, M. Douglas and R. Merrell (1983). The independence of insulin release and ambient insulin in vitro. Diabetes 32, Otonkoski, T., M. Knip, I. Wong and O. Simell (1988). Effects of growth hormone and insulin-like growth factor I on endocrine function of human fetal islet-like cell clusters during long-term tissue culture. Diabetes 37, Sando, H., Y. Kanazawa and T. Kuzuya (1970). Effect of bonito insulin on endogenous insulin secretion in dogs. Am. J. Physiol. 218, Scheiwiller, E., H. P. Guller, J. Merriweather, C. Scandella, W. Maerki, J. Zapf and E. R. Froesch (1986). Growth restoration of insulindeficient diabetic rats by recombinant human insulin-like growth factor I. Nature (Lond) 323, Schoenle, E., J. Zapf, R. E. Humbel and E. R. Froesch (1982). Insulin-like growth factor I stimulates growth in hypophysectomized rat. Nature (Lond) 296, Stagner, J., E. Samols, K. Polonsky and W. Pugh (1986). Lack of direct inhibition of insulin secretion by exogenous insulin in the canine pancreas. J. Clin. Invest. 78, Takano, K., N. Hizuka, K. Asakawa, I. Sukegawa, K. Shizume and H. Demura (1990a). Effects of sc administration of recmbinant insulin-like growth factor I (IGF-I) on normal human subjects. Endocrinol. Japon. 37, Takano, K., N. Hizuka and H. Sekino (1990b). Safety and pharmacokinetics study of recombinant human IGF-I (FK780) by seven days continuous administration in healthy volunteers: Phase I clinical study. Yakuri to Chiryo 18, (In Japanese) Van Schravendijk, C. F. H., A. Foriers, J. L. Van den Brande and D. G. Pipeleers (1987). Evidence for the presence of type I insulinlike growth factor receptors on rat pancreatic A and B cells. Endocrinology 121, Zapf, J., C. Hauri, M. Waldvogel and E. R.

8 KAWAI et al. Endocrinol. Japon. December 1990 Froesch (1986). Acute metabolic effects and half-lives of intravenously administered insulinlike growth factors I and II in normal and hypophysectomized rats. J. Clin. Invest. 72,