Recently, the predominant molecular forms of cholecystokinin (CCK) in extracts of tissues from the central nervous system and small intestines of adult man and hog have been demonstrated to correspond to the COOH-terminal octa- and tetra-peptide of CCK. CCK is one of the established gastro-intestinal hormones capable of stimulating pancreatic exocrine secretion. In addition, pure natural porcine CCK, synthetic COOH-terminal octa-peptide of CCK and synthetic caerulein which contains a COOH-terminal penta-peptide identical to CCK have been shown to stimulate insulin and glucagon secretion in vivo and in vitro. Thus, it seems reasonable to expect some effect of these COOH-terminal fragments of CCK on pancreatic exocrine and endocrine secretion. The present study was, therefore, undertaken to compare the various COOH-terminal fragments of CCK with those effects on pancreatic exocrine and endocrine secretion in the rat.Pancreases from male Wistar rats, fed ad libitum, were isolated according to the technique of Kanno. Perfusate consisted of a Krebs Ringer bicarbonate buffer containing 4.6% Dextran T-70 and 0.25% bovine serum albumin. The effects of various COOH-terminal peptides of CCK (synthetized by Prof. Noboru Yanaihara, Shizuoka) at the doses ranging from 10-11 to 10-8M were studied in the presence of 50mg/dl or 150mg/dl glucose. Insulin (IRI) and glucagon (IRG) levels in the portal effluent were determined by radioimmunoassay using polyethyleneglycol and charcoaled dextran with antiserum 30K, respectively. Amylase activities in the pancreatic juice were assayed by a chromogenic method with blue-dyed starch polymer.In the presence of 50mg/dl glucose, IRI and IRG responses to the COOH-terminal octa-peptide of CCK (CCK-8) occurred at high doses, more than 10-9M. On the other hand, IRI and IRG releasing doses of CCK-8 resulted in lower rates of release of fluid and amylase when compared with those stimulated by the maximally effective doses of CCK-8.In the presence of 150mg/dl glucose, CCK-8 at a submaximally effective dose of 10-10M for the exocrine pancreas potentiated the IRI response to glucose stimulation. These results were completely consistent with our previous findings obtained with natural porcine CCK and caerulein. The COOH-terminal deca-peptide (CCK-10) and tetradeca-peptide (CCK-14) at a dose of 10-10M also potentiated the effluent IRI response to glucose stimulation. However, the IRI responses to CCK-8 and CCK-14 were biphasic, but the IRI response to CCK-10 was monophasic and transient. Moreover, the maximal IRI response just after the perfusion of these peptides was greater in CCK-8 than in CCK-10 or CCK-14. On the other hand, the effluent IRI responses to the COOH-terminal tetra-peptide of CCK (CCK-4) or deamidated CCK-8 (CCK-8-OH) at a dose of 10-10M was not significantly increased as compared to that obtained with glucose stimulation alone. In addition, the combined perfusion with 10-10M CCK-8 and 10-8M CCK-8-OH resulted in a lower IRI response than that obtained with the perfusion of CCK-8.From these observations, it may be concluded that CCK-8 at a submaximally effective dose for the exocrine pancreas has a glucose-dependent insulinotropic action and that the amino acids from the position 5 to 8 of the COOH-terminus and the NH2 -residue of the COOH-terminal phenylalanine are important for the insulinotropic action of CCK.