Cyclosporin therapy in vivo attenuates the response to vasodilators in the isolated perfused kidney of the rabbit

1. The endothelium releases a number of vasoactive compounds, including the vasodilator prostaglandins, prostacyclin (PGI2) and prostaglandin E2 (PGE2) and endothelium-derived relaxing factor (EDRF), which play an important role in the regulation of vascular tone in the microcirculation. Nephrotoxicity is the major complication of cyclosporin (CS) therapy and is related to an increase in intrarenal vascular tone. Endothelial cell generation of PGI2 is inhibited by CS although this cannot fully explain the changes in vascular tone observed. We have investigated the possibility that EDRF-dependent vasodilatation is also affected by CS therapy in vivo. 2. CS nephrotoxicity was induced in rabbits with CS (15 mg kg-1 per day s.c. for 20 days (n = 6]; 6 rabbits were given CS vehicle (Veh) and 9 animals were studied without any treatment. Creatinine clearance fell significantly during treatment in the CS-treated rabbits (11.78 +/- 1.5 ml min-1, mean +/- s.e. mean, to 7.79 +/- 1.2 after 20 days treatment) but did not change in the vehicle-treated animals. 3. The responses to the endothelium-dependent (acetylcholine (ACh] and endothelium-independent (nitroprusside (NP) and PGI2) vasodilators were assessed in indomethacin-treated isolated perfused kidneys (IPKs) from untreated, CS- and Veh-treated animals. Vascular tone was induced with a constant infusion of noradrenaline 150 nM and the perfusion rate adjusted to produce a perfusion pressure of 90 mmHg. Perfusate flow rate (22.3 +/- 4.6 vs 20.4 +/- 3.1 ml min-1) and glomerular filtration rate (2.04 +/- 0.37 vs 1.88 +/- 0.16 nl min-1) did not differ between IPKs from CS- and Veh-treated animals. 4. The vasodilator response to ACh was reduced in the kidneys from CS-treated animals compared with those from untreated and Veh-treated animals (mean reduction 35.3 + 2.3% compared with Veh) as were the responses to both NP (42.8 + 3.6%) and PGI2 (27.7 + 7.4%). 5. This suggests that CS nephrotoxicity is not mediated via an effect on endothelium-dependent responses and that it is more likely that CS has a direct effect on vascular smooth muscle.