Akira Ohneda

Glucagon response to arginine after treatment of diabetes mellitus.

To investigate the aminogenic glucagon response in diabetes mellitus, arginine infusion tests were carried out on twenty-four diabetic patients before and after treatment. Eleven healthy men served as a control group. Plasma glucagon was measured by radioimmunoassay using an antiserum, G21, specific for pancreatic glucagon. Out of twenty-four patients, five were treated with diet alone, eight with sulfonylurea, and eleven with insulin. In all these diabetic groups, the glucose tolerance improved after treatment for diabetes mellitus, while the insulin response to the glucose did not show any remarkable change. The fasting levels of the plasma glucagon did not differ from that of the normal subjects both before and after treatment. Hyperresponsfveness of the plasma glucagon to arginine infusion was observed in all diabetic groups, in comparison with that of the normal controls. The exaggerated response of the plasma glucagon to arginine was lowered following appropriate treatment in each diabetic group. However, as far as the changes in glucagon area during the arginine test are concerned, the aminogenic hyperresponsiveness of the plasma glucagon was reduced prominently in the diabetic group treated with sulfonylurea. The relationship between the response of glucose and plasma insulin and between glucose and glucagon to arginine was investigated, and the importance of the changes in the insulin:glucagon ratio was emphasized. Moreover, the possibility that long-term administration of a sulfonylurea may reduce an exaggerated glucagon response to arginine was discussed.

Hypoglycemia due to apparent autoantibodies to insulin. Characterization of insulin-binding protein.

A fifty-two year old female, admitted when unconscious, had a blood glucose level of 33 mg./100 ml. Glucose injection restored consciousness, and oral glucose loading revealed a diabetic curve. Hypoglycemic coma was not produced by tolbutamide. Total immunoreactive insulin extracted from fasting plasma ranged from 824 to 1,400 μU. per milliliter. Laparotomy did not reveal adenoma of the pancreas. The specimen of the pancreas contained 3.8 U. of insulin per gram and histologically showed an increase in the numbers of islets. Her plasma was proved to bind 131-I-insulin by gel filtration, paper chromatography and dextran-coated charcoal methods. The binding protein was identified to be gamma globulin upon paper electrophoresis and was absorbed by anti–human globulin rabbit serum. The gamma globulin was characterized as IgG on immunoelectrophoresis with radioautography. Protein fractions from the patients plasma, freed from insulin by gel filtration after acidification, reacted with pork, beef or human insulin, and dose response curves were depicted. With bonito insulin, however, it did not show any immunologic reaction. A positive skin anaphylaxis reaction was demonstrated in a guinea pig immunized with the patients plasma, but no precipitation reaction was observed. Since the patient had never received insulin injections, the results suggest that the insulin-binding protein contains antibodies which are produced against endogenous insulin.

A Malignant Insulinoma Transformed into a Glucagonoma Syndrome

A 55-yr-old man was admitted to the Hospital of Tohoku University in 1970, when physical examination and laboratory findings revealed malignant insulinoma with metastases to the liver. Streptozotocin treatment, totaling 79.5 g for 1 yr, improved his clinical symptoms. After discharge he was well until 1977, when he complained of palpitation and dyspnea, and he was hospitalized in September 1977 because of anorexia and a loss of body weight. Laboratory findings revealed anemia, with a decrease both in serum albumin and iron. His fasting blood glucose was 74 mg/dl and plasma insulin (IRI) 25 μU/ml. His plasma IRI increased slightly after the stimuli of glucose, arginine, and glucagon. In contrast, his plasma glucagon (IRG) increased at fasting (3372 pg/ml) and became markedly elevated after arginine or glucose was administered. After his final admission, diazoxide improved his symptoms temporarily; streptozotocin could not be administered because of dyspnea and palpitation. On 24 January 1978, he fell suddenly into a coma and 2 days later he died. An autopsy revealed a large pancreatic tumor and metastases to the liver. Measurement of hormones from the tissues revealed a smaller amount of IRI in the pancreatic tumor and liver metastases compared with that in the uninvolved pancreas. In contrast, IRG contents had increased in the pancreatic tumor and hepatic metastases. In addition to the insulin cells, numerous cells with glucagon and somatostatin were observed in morphologic examinations. This indicated a mixed endocrine pancreatic tumor exhibiting typical symptoms of insulinoma initially and a glucagonoma syndrome later.

Hepatic steatosis and the elevated plasma insulin level in patients with endogenous hypertriglyceridemia

Among 31 nonobese or obese patients with endogenous hypertriglyceridemia, hepatic steatosis was found by histologic examination of the biopsied specimen in 17 patients, and it was severe in six patients, They had no history of excessive alcohol intake. Chemical analysis revealed that the lipid accumulated in the liver was triglyceride. The hypertriglyceridemic patients, with or without histologic steatosis, showed significantly increased responses of both plasma insulin and blood glucose to oral glucose load compared with control subjects. The responses were more exaggerated in the hypertriglyceridemic patients with steatosis than in the hypertriglyceridemic patients without steatosis. Analysis of correlations between five variables (liver triglyceride, plasma insulin, blood glucose, body weight index, and serum triglyceride) was done on 15 subjects whose liver triglyceride values were quantified, and highly significant correlations were found between liver triglyceride and plasma insulin, blood glucose, or body weight index. A step wise multiple regression analysis performed on the five variables with liver triglyceride as the dependent variable revealed that the plasma insulin level was the most closely related variable, and the blood glucose level the next. The prediction equation for liver triglyceride as a function of plasma insulin and blood glucose levels (r = 0.91, p greater than 0.001) accounted for 84 percent of the total variance of liver triglyceride. It was shown that the decay of intravenously injected insulin in plasma was not delayed in the hypertriglyceridemic patients with steatosis, while the insulin sensitivity examined after intravenous insulin injection significantly decreased in the hypertriglyceridemic patients with or without steatosis, thus suggesting that the hyperinsulinemia in the hypertriglyceridemic patients was due to an increased insulin secretion associated with the decrease in the insulin sensitivity. Therefore, the elevated plasma insulin and blood glucose levels--or the insulin insensitivity by itself--might be the essential abnormalities in patients with endogenous hypertriglyceridemia, which, in extreme cases, might lead to massive triglyceride accumulation in the liver.

Response of plasma glicentin to intraduodenal administration of glucose in piglets

Controversial results concerning the secretion of glicentin prompted us to investigate the response of circulating glicentin to intraduodenal administration of glucose in piglets. A 20% solution of glucose (2 g/kg) was administered into the duodenum of six piglets in a fully conscious state. As blood glucose rose, plasma insulin increased to a peak of 21 +/- 4 microU/ml. Plasma glucagon, determined by C-terminal-specific antiserum, was 70 +/- 30 pg/ml at fasting and slightly increased after the glucose load. Plasma immunoreactive glucagon measured by cross-reacting glucagon antiserum increased from the baseline of 1563 +/- 260 to a peak of 4738 +/- 415 pg/ml at 120 min. Plasma glicentin determined by antiserum R 64 was 463 +/- 81 pmol/l at baseline and reached a peak level of 1081 +/- 174 pmol/l at 90 min. The percent changes of plasma glucagon from the fasting level measured by cross-reacting antiserum and glicentin were 296 and 233%, respectively. There was a significant correlation between plasma glucagon measured by cross-reacting antiserum and glicentin (r = 0.817, P less than 0.001). Chromatography of plasma obtained during glucose load revealed the heterogeneity of glicentin. It can be concluded from the present study that glicentin is clearly secreted in response to intraluminal administration of glucose.

Insulinotropic action of human glicentin in dogs

Glicentin has been demonstrated to be released in response to the intraluminal administration of nutrients, but its biological action remains unknown. To clarify the effect of glicentin on the endocrine function of the pancreas, the present study was performed using an in vivo local circulation system of the canine pancreas. During infusion of 0.5% solution of glucose or arginine, 100 and 400 pmol glicentin and 400 pmol glucagon were administered into the pancreaticoduodenal artery (PA) within 10 minutes at 40-minute intervals successively. During glucose infusion, blood glucose in the femoral artery did not change following administration of 100 pmol glicentin, but slightly increased following 400 pmol glicentin. Plasma insulin (immunoreactive insulin [IRI]) in the pancreaticoduodenal vein (PV) increased significantly only following infusion of 400 pmol glicentin. Plasma glucagon (immunoreactive glucagon [IRG]), measured with a specific antiserum to the C-terminal portion of glucagon, did not change following administration of 100 pmol glicentin, but was slightly elevated following 400 pmol glicentin. Plasma total IRG, measured with a nonspecific antiserum, increased promptly after administration of 100 and 400 pmol glicentin. During arginine infusion, the response of plasma IRI to glicentin was markedly exaggerated both in dosages of 100 and 400 pmol. From the present study it was concluded that human glicentin clearly increases insulin release from the canine pancreas.

CHARACTERIZATION OF CIRCULATING IMMUNOREACTIVE GLUCAGON IN RESPONSE TO INTRADUODENAL ADMINISTRATION OF FAT IN DOGS

Experimental studies were carried out to characterize the total immunoreactive glucagon (IRG) of plasma in response to the intraduodenal administration of fat in dogs. When butter or corn oil, in a dose of 2 g per kg of body weight, was administered into the duodenum of conscious dogs, total IRG in the vena cava increased significantly from the fasting level. In the experiment using conscious dogs with triple catheters, the rise in total IRG of plasma after fat loading was observed to consist of an early increase of total IRG in the pancreatic vein and a later rise of total IRG in the mesenteric vein. The rise in total IRG of plasma after fat loading continued after the removal of the pancreas. When butter was administered into the distal ileum of conscious dogs, total IRG in the vena cava increased for 2 hr, whereas pancreatic glucagon remained unchanged. The direct infusion of chyle into the pancreatic artery failed to induce an increase in total IRG in the pancreatic vein. From this experiment it was concluded that the increase in circulating total IRG after fat loading derives from gut glucagon-like immunoreactivity and from pancreatic glucagon. This study also suggests that the secretion of pancreatic glucagon is mediated through enteric signals, probably pancreozymin.

Response of plasma glicentin to fat ingestion in piglets

In order to elucidate the response of plasma glicentin to fat ingestion, butter, glycerol or palmitate was administered into the duodenum of piglets in a fully conscious state and plasma glicentin and glucagon were determined. Butter instillation did not change blood glucose. Plasma triglyceride rose gradually 120 min after butter loading. Plasma insulin and glucagon measured by antiserum specific to the C-terminal slightly increased following butter administration and plasma total glucagon and glicentin increased gradually and significantly. The increments of total glucagon and glicentin were 179 and 158%, respectively. However, chromatography of porcine plasma obtained during fat loading revealed heterogeneity of glicentin-related peptides. Glycerol ingestion induced a slight rise of plasma total glucagon. Administration of palmitate revealed an increase in plasma total glucagon and glicentin. The present study clearly demonstrates the secretion of glicentin following fat ingestion, which might be caused by the hydrolysates of triglyceride, as suggested in previous dog experiments.

Interactions of Obesity and Glucose-stimulated Insulin Secretion in Familial Hypertriglyceridemia

Plasma lipids and lipoproteins, glucose tolerance, plasma insulin response to glucose load, and liver function were examined in 81 relatives of 12 index cases with primary endogenous hypertriglyceridemia, hyperinsulinemia, and hepatic steatosis, as well as in 90 nonrelatives, including the spouses, as controls. Insulin hypersecretion (with or without glucose intolerance), endogenous hypertriglyceridemia, and abnormal liver function suggesting hepatic steatosis were shown to exist in the relatives mostly in combined fashion. Correlation analysis and stepwise multiple regression analysis revealed that the combined disorder developed on the basis of obesity. The incidence of diabetes mellitus was significantly high in the relatives (14.8 per cent) as compared with the normal Japanese population (3.5 per cent). Although the vertical transmission of the combined disorder was noted in almost all pedigrees, the frequency distribution analysis of insulin response, glucose tolerance, and plasma triglyceride showed the histograms of these variables similarly skewed to the right as compared with those in the controls, with no apparent bimodality. In view of the hitherto suggested role of insulin in triglyceride metabolism, it is concluded that hyperinsulinemia coupled with obesity seems to be the basic trait of this form of familial hypertriglyceridemia and hepatic steatosis, though the mode of transmission remains to be elucidated.

Secondary Diabetic Retinopathy in Chronic Pancreatitis

We found abnormal glucose tolerance curves in 45 of 47 patients with chronic pancreatitis and observed secondary diabetic retinopathy in eight of 45 cases showing slight changes in the fundus. Abnormal glucose tolerance curves were somewhat related to the exocrine dysfunction of the pancreas. Slightly abnormal glucose tolerance curves were observed frequently in patients with chronic pancreatitis, and both insulin and glucagon responses were decreased. We could not explain the cause of the low frequency of secondary diabetic retinopathy in pancreatic diabetes from the results of insulin and glucagon response tests.