The biopsy materials of the sural nerve and the quadriceps femoris muscle obtained from 37 diabetic patients were studied by light and electron microscopy. The ages at the onset were below 30 years in 13 cases.
Sural nerve studies
1) Degeneration or disintegration of the myelin sheaths was observed in all cases of adult onset and 83% of cases of juvenile onset diabetes.
2) Cell inclusions of lipid were noted in 13 of 18 adult cases and 3 of 5 juvenile cases in the cytoplasm of Schwann cells of the myelinated nerve fibers.
3) Degeneration of the myelin sheaths was seen more frequently in adult cases than in juvenile cases. Therefore such changes might be related with aging process to some extent.
4) In one case of juvenile type, degeneration of the axon without accompanying changes of the myelin sheaths was observed. This finding suggested that the changes of the axon preceded that of the myelin sheaths in some occasions while most of the diabetic neuropathy started from Schwann cell changes with subsequent myelin sheath degeneration.
Muscle studies
1) Light microscopic observations of the muscle revealed mixture of both neurogenic and myogenic changes.
2) Light microscopically, focal degeneration seemed to be one of the features. Electron microscopically intact muscle fibers and completely degenerated ones were seen neighboringly as frequent observations.
3) Abnormally increased lipopigment between the myofibrils was another frequent finding.
4) Thickning of the basement membranes of the capillaries was remarkable and related with the diabetic changes of the ocular fundus.
5) Accumulations of the glycogen granules were often observed between myofibrils.
6) Crystalline inclusion bodies within the mitochondria, which had been reported in some types of myopathies for the past several years, were observed in one case and this was the first observation as far as diabetes mellitus was concerned.

View full abstract

A mixture of 10 μCi/kg of ^14C-U-maltose and 1 g/kg of 10% maltose solution was injected intravenously to normal rabbits and the radioactivities in blood, urine and several organs were determined.
The blood maltose level decreased from 563 mg/100 ml at the time of 5 minutes to 17 mg/100 ml 4 hours after maltose injection. Blood glucose slightly increased from the fasting level of 150 mg/100 ml to 238 mg/100 ml 40 minutes after the administration of maltose. ^14C-maltose rapidly disappeared from the blood stream, reaching the base-line within 4 hours. In contrast, circulating ^14C-glucose rose gradually after the maltose injection and exceeded ^14C-maltose at the time of 90 minutes.
In urine, 24% of total radioactivities was excreted as ^14C-maltose, 6% as ^14C-glucose and 12% as other ^14C-compounds.
The changes in radioactivities in each organ were divided into two types. In the organs such as the kidneys, the liver, the small intestine, the brain or the muscle, ^14C-glucose levels exceeded the level of ^14C-maltose throughout the experiment for 12 hours. In the other organs such as the heart, the lungs, the lymph nodes, the large intestine or the pancreas, ^14C-maltose at the time of 20 minutes was higher than ^14C-glucose but thereafter ^14C-glucose exceeded ^14C-maltose which rapidly decreased after maltose injection. Approximately 10% of total radioactivities were observed in the skeletal muscle at the time of 60 minutes and the radioactivities were measured at the organs in order of the kidneys, the large intestine, the liver, the small intestine, the lungs and the heart.
^14C -maltose was incorporated into glycogen in the heart and the muscle as much as ^14C-glucose was done. In the liver, however, the incorporation of ^14C-maltose into glycogen was minimum. In the brain, ^14C-glucose was detected whereas ^14C-maltose was not observed. In the gall bladder, 0.2% of total radioactivities was excreted for 6 hours.
These experiments, indicate that 70% of ^14C-maltose injected to normal rabbits was metabolized in various organs in different patterns.

View full abstract

The qualitative and quantitative determinations on the intestinal absorption of insulin in rabbits have been performed immunologically and biologically, and following results were obtained;
1. In experiments with the isolated perfused intestine in which artificial plasma was circulated through its vascular system, a considerable amount of insulin disappeared from the intestinal lumen (10.8-30.9% of the added insulin) and was recovered from the venous effluent (5.9-15.9%) during the two hour perfusion. In these experiments, the amount of insulin absorbed was directly correlated to the dose of insulin added. By the acrylamide-gel electrophoresis of the venous effluent, there was clearly demonstrated a component with the relative mobility similar to crystalline insulin.
2. Administration of insulin into the stomach and duodenum at a dose of 100 U/kg did not cause a significant increase in plasma insulin levels. When insulin at doses over 40 U/kg was infused to the upper jejunum via an indwelling catheter, there was a significant increase in plasma insulin levels, followed by a significant fall in blood glucose. A considerable increase in plasma insulin was observed following administration of insulin at doses over 20 U/kg to the Thiry-Vella loop of the jejunum, in which no pancreatic secretion entered. With a dose of 100 U/kg, the mean plasma insulin rose promptly to the peak of 241 μU/ml followed by a severe hypoglycemia.
3. The fraction of insulin absorbed from the intestine in vivo was estimated by measuring the increase in plasma insulin in the portal vein and the portal vein blood flow. The amount of insulin absorbed was found to be small, corresponding to the amount less than 2% from the jejunum and less than 4% from the ileum during the time of three hours.
From these results, it is concluded that insulin can be absorbed from the intestine of mammals in a biologically and immunologically active form, but the fraction of insulin absorbed is relatively small. The relatively small absorption of insulin might be attributable either to inactivation of a portion by intestinal proteolytic enzymes or to inherent physical difficulties to diffusion imposed by protein character.

View full abstract

A 52-year-old female was admitted because of unconsciousness. She suffered from peritonitis at the age of 30 years. She had been treated for hypertention since 38 years of age. Her blood glucose at admission was 33 mg/100 ml and she responded to the intravenous administration of glucose. The fasting levels of blood glucose ranged 52 to 80 mg/100 ml. Oral glucose loading revealed a diabetic glucose tolerance and hypoglycemic coma was not provoked by tolbutamide injection, although blood glucose fell 63 to 42 mg/100 ml at the time of 120 minutes. Total immunoreactive insulin (IRI) of untreated plasma was found more than 4, 000 μU/ml. Under the diagnosis of insulinoma, she was undergone laparotomy. On operation, adenoma of the pancreas was not revealed and the partial resection of the head of the pancreas was done.
Insulin content of the pancreas was 3.8 U/kg and the histological examination showed an increase in numbers of the Langerhans' islets.
Her plasma was proved to bind '311-insulin by the method of gel filtration with Sephadex G 50, paper chromatography or dextran-coated charcoal. The binding protein was identified to be gamma globulin upon paper electrophoresis and was absorbed by anti-human gamma globulin rabbit sera. Endogenous insulin as well as 131 I-insulin bound to patient's plasma was dissociated by acidification of plasma. Protein fraction, obtained by acidification and gel filtration of patient's plasma, reacted with pork, beef or human insulin. With insulin from bonito, however, her plasma did not show any immunological reaction. The maximal binding capacity of her plasma to human insulin was determined approximately as 5, 600 μU/ml. A positive skin anaphylaxis reaction was observed in a guinea pig immunized with patient's plasma but any precipitation reaction was not demonstrated by Ouchterlony method. It is concluded that the insulin binding protein is antibody to insulin and hat insulin antibody must be produced against endogenous insulin, since she had never received any insulin administration.
Clinical features, hypoglycemia, decreased glucose tolerance, increased level of IRI and hyperplasia of the pancreatic islets, were interpreted by the presence of insulin antibody in plasma. The patient findings were consistent with a case of insulin autoimmune syndrome coined by Hirata.

View full abstract

An autopsied case of thirty-two-year old man with Kimmelstiel-Wilson syndrome associated with pancreatolithiasis and sustained hyperinsulinism was reported with special refrences to the developing mechanism of hypoglycemic attacks, which were recurrently observed in this case.
Clinically, the diagnosis of diabetes mellitus associated with insuloma was strongly suggested, but the only histological evifindingdence in pancreas at autopsy was a marked decrease in the amount of normal structure which was replaced with large amount of fibrous connective tissue. We failed to verify the existence of insuloma in the pancreas in spite of its thorough examination. This, however, does not exclude the possibility of ectopic insuloma.

View full abstract

Serum insulin responses during 100 g oral glucose tolerance test were studied at 3-4 month intervals in subjects with normal glucose tolerance. Obese people had higher insulin levels than non-obese people both at fasting and after glucose load. Insulin responses were similar between the two tests when the body weight did not change, whether the subjects were obese or non-obese. Hyperresponse of insulin in obese people decreased significantly after weight losses of 5-10 kg. Significant elevation of insulin response was observed in patients with mild pulmonary tuberculosis who gained 3.5-7.0 kg body weight, but still remained within the normal weight range. The degree of obesity dose not seem to be the sole factor causing insulin hyperresponse in these patients.

View full abstract

Our previous study showed that maltose did not stimulate insulin secretion from the Perfused rat pancreas for 30 minutes. However, slow effect of maltose on insulin secretion is expected, when maltose cleaves glucose in the B-cells of the pancreas. Male Wistar rats (200-250 g) were fasted for 20 hours. Preparation of the isolated pancreas was performed according to the procedure of Grodsky with a small modification. Krebs-Ringer bicarbonate buffer containing dextran (4.5 %) was used as perfusate.
In this experimental system, biphasic increase in insulin secretion was obtained by glucose (300 mg/100 ml), i. e. rapid secretion of insulin at 3 minutes and slow and sluggish increase thereafter. On the contrary an increase in insulin secretion appeared at 30 minutes after maltose (300 mg/100 ml) perfusion and continued until 90 minutes. This result shows that maltose stimulates insulin secretion without early rapid phase as seen in glucose perfusion.

View full abstract