It has been suggested that the survival of animals deprived of water is in direct proportion to their ability to concentrate their urine, and that the high concentration of their urine is achieved at the expense of an increased secretion of the antidiuretic hormone. The problem investigated here was to see whether the decrease of the urinary volume observed in laboratory rats deprived of water was the sole factor in the mechanism of water preservation and if so, whether the reduction of urinary secretion was the result of a corresponding increase in secretion of the antidiuretic hormone. METHODSWhite adult male rats of body weight ranging from 250 to 325 g were fed on a standard diet, 100 g of which contained: protein 14 g, fat 4 g, and soluble carbohydrate 49 g; calorific yield 400 call 100 g. From its composition, it could be calculated that its metabolic water was 52 ml./100 g food: its amount of moisture (i.e. 'preformed water') was 13-0 ml./100 g. The water content of food and of faeces was estimated by drying in an oven at 1040 C for 48 hr. The rats were kept in metabolism cages, at a temperature of 210 C; the degree of humidity was not recorded. Urine separators were fixed to the cages and the faeces were collected separately in a vessel with narrow neck so as to diminish water evaporation. Whenever possible fresh faeces were collected for the estimation of their water content. For the estimation of its antidiuretic activity urine was collected in a solution of 3 % (v/v) acetic acid (final concentration, 0-2-0-3 %). Pituitary glands were removed, dried and extracted as specified in the British Pharmacopoeia (1953); the amount of acetic acid used in their extraction did not exceed 0-2 ml./gland of a 0-25 % acetic solution (Bentley & Dicker, 1955). Urine samples were analysed for their Na and K content, using an EEL flame photometer; occasional Cl estimations were performed.The antidiuretic activity of urine or of dissected neurohypophyseal glands was estimated by intravenous injection in rats under ethanol anaesthesia, with the water load kept constant (Dicker, 1953). The vasopressor activity of the neurohypophysis was estimated on rats anaesthetized with urethane and injected with dibenamine (Dekansky, 1952) and the oxytocic activity on a rat's isolated uterus (Holton, 1948). Each assay of the antidiuretic, vasopressor or oxytocic activity consisted of four doses, two of the standard and two of the unknown, the ratio, high to low dose, being the same for standard and unknown solutions. Results were expressed in terms of the antidiuretic or vasopressor activity of a solution of vasopressin, or in terms of oxytocic activity of a solution of oxytocin.

Friedman, Hinke & Friedman (1956) have shown that the injection of hypertonic sodium chloride solution into the common carotid arteries of nonanaesthetized hydrated rats produced a smaller antidiuretic effect in senescent than in younger animals. They attribute the decrease of the antidiuretic effect in old rats to failure of the neurohypophysis to react to the stimulation of hypertonic saline solutions.The accidental finding, during a series of routine assays, that intravenous injections of known amounts of vasopressin into old rats under ethanol anaesthesia did not produce the expected antidiuretic effect, led us to reinvestigate the problem with a view to determining whether the decreased antidiuretic response reported by Friedman and his associates was really due to a decreased responsiveness of the neurohypophysis to osmotic stimuli or rather to a decreased sensitivity of the kidneys to the antidiuretic hormone. METHODSOld white rats of about 350 days and an average body weight of 370 g (range 325-415 g) were compared with control adult animals, about 100 days old with an average weight of 250 g (range 210-270 g). The old rats had been reared either on a diet containing 13-7 % of protein (diet D41) or on one containing 20 % protein (diet D 86). Control rats had been fed on diet D 41 only. The role of the protein content of the diet on renal function in rats has been reported previously (Dicker, 1949).For pressor assays, a modification of Dekanski's method (1952) was used. The blood pressure was recorded from the femoral artery. A carotid artery, as well as a vein, was cannulated, so that it was possible not only to assess the effect of an intracarotid hypertonic sodium chloride solution on the blood pressure, but also to estimate it in terms of intravenously injected vasopressin. A (2 + 2) dose assay of up to four groups was used. The results were calculated by standard statistical methods (B.P. 1953); the errors fell well within the limits stated previously (Dicker & Nunn, 1957).Antidiuretic effects following intracarotid injections of various solutions in non-anaesthetized hydrated rats were estimated as follows: the common carotid artery and the bladder were cannulated under light ether anaesthesia; 1 hr after the operation, when the animal had entirely recovered from anaesthesia, 5 ml. water/100 g was given by stomach tube. When the water diuresis was well established, vasopressin (Pitressin, Parke, Davis and Co.) or hypertonic sodium chloride solution was injected into the carotid artery. The antidiuretic effect was estimated

The fate of the antidiuretic activity of vasopressin has been investigated in experiments in vivo (Burn & Singh Grewal, 1951;Heller, 1952Heller, , 1953Ginsburg & Heller, 1953;Dicker, 1954;Dicker & Greenbaum, 1954) and in experiments in vitro (Heller & Urban, 1935;Birnie, 1953;Dicker & Greenbaum, 1956). The antidiuretic activity of vasopressin appears to be inactivated by both liver and kidneys though the manner in which the inactivation takes place is still not clear. As for the metabolism of the pressor activity of vasopressin, little is known about its fate in the body beyond the fact that the pressor activity of post-pituitary gland extracts appears to be less stable than the antidiuretic activity (Heller, 1939).Since vasopressin has been synthesized there can be no doubt that both activities belong to the same octapeptide molecule. According to van Dyke (1955) the pharmacological activities of the vasopressin are abolished by cleavage of the disulphide bond. As, however, the two activities of vasopressin are so different, there is a possibility that the metabolism and the rate of excretion of the pressor activity are different from those of the antidiuretic activity. The present investigation is a contribution to this problem. METHODSMale albino rats of approximately 250 g were used. They had been bred and reared in the Department's animal house and fed on a standard diet containing: protein 14 g, fat 4 g, and soluble carbohydrate 49 g/100 g.The pressor activity was assayed on rats injected with urethane and dibenamine (Dekanski, 1952). The errors of the method fell within the limits stated previously (Dicker & Nunn, 1957). Nephrectomy and evisceration were performed on rats under ethanol anaesthesia, induced by the administration of 5 ml./100 g of a 12% (v/v) ethanol solution, and maintained, if necessary, by intravenous injections of 1 ml. of a 10% ethanol solution. The 'evisceration' consisted in the removal of the whole of the gastro-intestinal tract with the pancreas and spleen: in what follows the word is used with this special meaning. The liver, however, remained in 8itu, but its whole blood supply, except that from the vena cava, was stopped by ligation of the coeliac and the superior mesenteric arteries and their branches, and of the portal vein and its branches.

No abstract