The effect of hypoglycemia induced by the injection of glucagon-free insulin on adrenal medullary secretion was studied in 4 totally pancreatectomized dogs and in 3 sham-pancreatectomized controls. There was a clear-cut statistically significant increase in the output of catecholamines in the pancreatectomized animals, although the mean increase was less than it was in the shamoperated dogs. It is concluded that glucagon from the pancreas is not an obligatory mediator of the adrenal medullary response to hypoglycemia in the dog. (Endocrinology 74: 656, 1964) V EULER and his associates (1) and Sarcione et al. (2) have suggested that the increase in adrenal medullary secretion produced by hypoglycemia may not be due to the depression in blood sugar per se. Because hypoglycemia is a stimulus for the release of glucagon (3) and glucagon stimulates the adrenal medulla by a direct action on medullary cells (2, 4-6), Sarcione et al. postulated that the adrenal medullary response to hypoglycemia was caused by the increase in circulating glucagon. To test this hypothesis, we determined the effect of pancreatectomy on the adrenal medullary response to hypoglycemia induced in dogs by the injection of glucagon-free insulin. Materials and Methods Seven 10.2-18.2 kg male mongrel dogs which had been fasted for 48 hr were studied. Each was anesthetized with pentobarbital, and the pancreas was exposed through a right rectus incision. In 4 dogs the pancreas was removed, whereas in the remaining 3 it was left in place (shampancreatectomy controls). The wounds were then sutured, and in each dog the right lumboadrenal vein was cannulated through a flank incision by the method of Hume and Nelson (7). Cannulas were also placed in one femoral artery for collecting blood and monitoring blood pressure and in one femoral vein for administering blood and insulin. Arterial and adrenal venous blood collections were begun 1 hr or more after cannulation of the lumboadrenal vein and from 3 to 4^ hr after pancreatectomy or sham-pancreatectomy. Collections were made 15 and 5 min before and 15, Received October 4, 1963. Supported by USPHS Grants AM-06704-01, A-4426 and H-6285. 1 USPHS Medical Student Summer Research Fellow. 30, 60, 90, 120, 150 and 180 min after rapid intravenous administration of 3 units/kg body weight of glucagon-free insulin. A volume of bank blood from normal dogs equal to the amount of blood collected was reinfused after each collection. Arterial blood pressure was continuously recorded by means of a Statham strain gauge and a Grass Model 5 polygraph. Arterial blood was collected in tubes containing sodium fluoride, and its glucose content was measured by the glucose oxidase method (8). The timed adrenal venous blood specimens were collected in iced heparinized tubes and centrifuged promptly; the plasma was then frozen. The norepinephrine and epinephrine content of the plasma samples was subsequently determined by the method of Goldfien et al. (9). Output of these hormones was calculated by multiplying their concentrations by the adrenal venous plasma flow per min. Results and Discussion The mean values for blood sugar, epinephrine and norepinephrine output, and blood pressure in the control and pancreatectomized dogs are shown in Fig. 1 and 2. Blood pressure levels were similar in the 2 groups. Epinephrine secretion increased 14-fold or more in each of the sham-pancreatectomized dogs and 11-fold or more in each of the pancreatectomized dogs. The increased catecholamine secretion in the control animals was due mostly to an increase in epinephrine output, similar to the increase seen in normal dogs following injection of smaller doses of commercial insulin (10, 11). There was also a significant increase in catecholamine output (p <0.05) in the pancreatectomized dogs. Therefore, glucagon from the pancreas cannot be a necessary mediator of the adrenal medullary response in the dog. However, 2 Crystalline glucagon-free insulin supplied by Eli Lilly & Co., Lot No. 499667.