Insulin induced hypoglycemia and tolbutamide administration accompanied by much smaller falls in blood glucose were associated with increased pancreatic vein IRG concentrations in laparotomized dogs under barbiturate anesthesia. Hyperglycemia may have diminished pancreatic IRG release. Changes in IRG levels in peripheral and jejunal veins during hypoglycemia were unremarkable. In dogs, pancreatectomy was associated with no change in IRG release into a jejunal vein. Very small doses of glucagon administered intraportally during hyperglycemia were effective in promoting IRI release. It is concluded that a fall in blood glucose from the basal state is a stimulus, and hyperglycemia is an inhibitor, of pancreatic IRG release. It is possible that glucagon may act in the basal state to deliver repeated small pulses of glucose and insulin into the circulation.

Isolated perfused intestine of rat was used to demonstrate the glucose-stimulated release of glucagon-like immunoreactivity (GLI) into serosal secretions. The released GLI was characterised using immunoaffinity chromatography on columns of immobilised antibodies specific for the N (residues 1 to 18) and for the C (residues 19-29) terminal portions of glucagon followed by gel-filtration. The immunoreactivity was present in a variety of molecular species. These include a large GLI which has a molecular weight about 12000 and binds to antibodies specific for the N-terminal portion of glucagon and two polypeptide fractions with molecular weight closer to that of glucagon. While one fraction of the small GLI boun both to antibodies specific for the C-terminal and N-terminal portions of glucagon the other bound only to the former antibodies. The relevance of these findings to the origins of circulating GLI and the possible precursor relationship between large and other forms of GLI is discussed.

Antibodies to secretin have been reported using crude secretin conjugated to albumin (Young, Lazarus, Chisholm and At• kinson 1967). The present study reports the production of antibodies against microgram quantities of synthetic secretin and natural secretin in the unconjugated form.

Plasma insulin levels were measured in 8 patients of whom 5 were surgically proven to have, and 3 had strongly suggestive evidence of, a functioning islet cell tumour. The fasting plasma insulin level was elevated in only 3 patients, despite repeated sampling. The administration of tolbutamide produced an excessive insulin response in 5 of the 8 patients, glucagon produced an excessive insulin response in 4 of 6 patients tested, 1-leucine produced a rise in plasma insulin in 1 of 2 subjects tested and the only subject studied after 50g. oral glucose had a high insulin response. The insulin response was highest after tolbutamide in all but one patient and this would appear to be the single most valuable test. However false negatives may occur and therefore the whole range of tests may have to be performed to establish the diagnosis. Two patients were studied during surgery and in one the insulin content of the tumour was assayed after extraction with acid alcohol. Four patients were studied again after treatment by surgery (2) and diazoxide (2) when normal insulin levels were found.

No abstract