No abstract

The angiotensin converting enzyme inhibitor, captopril, given to rats in their drinking water (about 40 mg/day) for 6 days caused an increase in intake of hypertonic NaCl solution which began 1‐2 days after the captopril was started and reached a plateau after 4‐5 days. Twice‐daily subcutaneous injections of captopril (15 mg per injection) elicited a sodium appetite similar in pattern to that seen with oral administration. The rats remained in sodium and fluid balance during oral captopril treatment and the haematocrit did not alter. Captopril infused directly into the ventricles (12 micrograms/h), or captopril reaching the brain from the periphery across a leaky blood‐brain barrier, suppressed the sodium appetite which normally follows oral captopril. Continuous intravenous infusion of captopril at rates high enough to block angiotensin converting enzyme in the brain (25, 50 or 500 mg/day) did not cause sodium appetite. As soon as the rate was reduced to a low value (5 mg/day), NaCl intake increased. In conclusion, moderate levels of circulating captopril which do not cross the blood‐brain barrier in sufficient amounts to block cerebral angiotensin converting enzyme, result in an increase in circulating angiotensin I which stimulates sodium appetite when it is converted to angiotension II in the brain.

SUMMARY1. In experiments lasting 8 h, low (0O5 mg kg-') or medium (5 mg kg-') subcutaneous doses ofthe angiotensin-converting enzyme inhibitor captopril were mildly dipsogenic in sham-operated rats, much more so in rats subjected to bilateral ureteric ligation and not at all in bilaterally nephrectomized rats.2. Rats with ligated ureters drank enough water to gain weight during the experiments. All other groups lost weight.3. The enhanced responsiveness of rats with ligated ureters, despite fluid retention, shows that captopril-induced drinking was not secondary to increased renal fluid loss.4. Ureteric ligation alone which caused some increase in renin secretion was mildly dipsogenic compared with sham operation. Captopril caused further increases in plasma renin concentration and more drinking suggesting that the captopril response is renin-dependent.5. The failure of the nephrectomized rat to drink after captopril also shows that the response is renin-dependent.6. The highest dose (50 mg kg-') of captopril did not at first stimulate drinking, though water intake increased later. Slowness to drink was not the result of general depression of behaviour since drinking in response to subcutaneous hypertonic NaCl or intracranial angiotensin II was not inhibited by the highest dose.7. Slowness to drink after the highest dose was attributable to blockade of converting enzyme centrally as well as peripherally. This meant that the increased circulating angiotensin I resulting from peripheral blockade of converting enzyme was only slowly converted to angiotensin II in the brain.8. When cerebral conversion ofangiotensin I was prevented by a single intracranial injection of 25 Asg captopril, drinking in response to the lower doses of captopril was also inhibited in normal rats and in rats with ligated ureters. 9. The same intracranial dose of captopril also inhibited drinking in response to intracranial injections of renin or angiotensin I, but not angiotensin II. The time course of inhibition of renin-induced drinking was similar to that of inhibition of subcutaneous captopril-induced drinking.10. In conclusion, subcutaneous captopril causes increased water intake through activation of the renal renin-angiotensin system, an effect that is enhanced when the system has already been partly activated by ureteric ligation. Increased circulating * Thouron Scholar, University of Pennsylvania. R. M. ELFONT AND J. T. FITZSIMONSangiotensin I resulting from blockade of peripheral converting enzyme must be converted to angiotensin II in the brain in order to stimulate drinking. Drinking is not the consequence of increased fluid loss.

R224-R230, 1989.--[125I]iodocyanopindolol ([125I]ICYP) and [3H]rauwolscine were used to quantitate, respectively, the beta- and alpha 2-adrenergic receptors in freshly isolated bovine cerebral microvessels and in pericyte cultures derived from these microvessels. Morphological and immunocytochemical criteria distinguished the pericytes from endothelial cells. Competitive binding studies established the specificity of the radioligand binding. The maximal number of binding sites (Bmax) for [125I]ICYP in the pericytes constituted only 8% of that in the microvessels (3.5 +/- 1.3 vs. 44.4 +/- 6.6 fmol/mg protein). In contrast, the Bmax for [3H]rauwolscine in the pericytes was 50% of that in the microvessels (55.4 +/- 11.8 vs. 111.1 +/- 9.5 fmol/mg protein). The dissociation constants for both [125I]ICYP and [3H]rauwolscine were similar in the two preparations. No alpha 1-adrenergic receptors, as defined by the specific binding of [3H]prazosin, were identified either in the pericytes or microvessels. Overall, our results suggest that pericytes contribute minimally to the total beta-adrenoceptor number of cerebral microvessels, and thus the beta-adrenoceptors must be located predominantly on endothelial cells. However, the contribution of pericytes to the total alpha 2-adrenoceptor number of the microvessels may be substantial.

In small (0.5 mg/kg) subcutaneous doses, the angiotensin-converting enzyme inhibitor, captopril, greatly enhanced drinking in response to caval ligation in the rat. Drinking was not secondary to urinary water loss since the rats developed a substantial positive fluid balance. High (50 mg/kg) subcutaneous doses of captopril reduced drinking to a level below that following caval ligation alone. This effect could be mimicked by giving repeated intracerebroventricular injections of captopril (total amount 110 micrograms) to rats treated with the lower subcutaneous dose of captopril. With this combination, therefore, not only did the lower dose enhancement disappear, the basal caval ligation drinking response was also reduced with a total dose of captopril of less than 2% of the higher subcutaneous dose alone. These results show that, when conversion of angiotensin I to angiotensin II is prevented in the brain as well as systemically, drinking in response to caval ligation is reduced although not entirely prevented. The original report that such drinking is multifactorial, depending on angiotensin as well as nonangiotensin mechanisms, is confirmed.