This paper is concerned with pancreatic substances that can antagonize insulin. Previous work (Marks & Young, 1939; Thorogood & Zinmmermann, 1945; Mirsky, Futterman, Wachman & Perisutti, 1951) has already suggested that insulin antagonists exist in the pancreas. Evidence has been produced for the identity of these antagonists with glucagon, but there is other work suggesting that the pancreas may produce substances that can antagonize insulin in such tissue as muscle (Young, 1962; Antoniades, Renold, Dagenais & Steinke, 1960) when glucagon is without effect (Randle, 1958). Antoniades et al. (1960) suggested that insulin might be combined with a basic protein in pancreas. The studies reported below are an attempt to isolate and characterize at least one of the bovine pancreatic substances that is capable of antagonizing insulin in a biological system such as rat diaphragm. MATERIALS AND METHODS Materials. DEAE-cellulose and CM-cellulose were obtained from Whatman (W. R. Balston Ltd.). Before use DEAE-cellulose was treated with 2N-NH3 and 2M-NaCl; CM-cellulose was treated with 2N-acetic acid and 2M-NaCl; both were washed well with distilled water. The bovine insulin (22-2 i.u./mg.) was obtained from Boots Pure Drug Co. Ltd., Nottingham. The proteolytic enzymes trypsin and oc-chymotrypsin were recrystallized preparations, obtained from L. Light and Co. Ltd., Colnbrook, Bucks. The pancreatic elastase was prepared by the method of Hall & Czerkawski (1959). Lilley & Valiance-Owen (1961) have shown that dialysis tubing contains substances capable of antagonizing insulin, and that these substances can be removed by boiling the tubing in two changes of distilled water. It was deemed wise to take the same precaution with the 1 in. tubing (Visking) used in the present work. Unless otherwise stated all the reagents used were of AnalaR quality. Electrophoresis. The purity of the fractions was frequently checked by use of electrophoresis at high voltage. on paper and at low voltage on cellulose acetate (Oxoid Division of Oxo Ltd., London, E.C. 4) as supporting medium. The electrophoresis was usually performed in a 75 mM-barbiturate (veronal) buffer solution, pH 8-6, which,

Crude pancreatic extracts described as trypsin (Dudley, 1923; Shonle & Waldo, 1923; Witzemann & Livshis, 1923) were found to interfere with the biological activity of insulin. Epstein & Rosenthal (1924) showed that, whereas a mixture of pancreatic extracts and insulin was biologically inactive when injected into animals immediately after mixing, little or no proteolytic destruction of the insulin by the pancreatic protein could have occurred. Later workers, using purified enzymes, have shown that trypsin acts slowly on insulin, breaking off a heptapeptide and alanine, and that the residuum possesses little or no biological activity (Nicol, 1960; Carpenter & Baum, 1962). Young & Carpenter (1961) found that the biological inactivation of insulin is complete when as little as 20 % of the insulin has been split by trypsin, and