Rabbit renal cortices were fractionated by collagenase dispersion and glomeruli, microvessels and tubuli purified on a discontinuous sucrose gradient. Binding experiments with (-)[125I]N6-(4-hydroxyphenylisopropyl)-adenosine ([125I]HPIA) provided evidence for the presence of A1-adenosine receptors in the glomerular and microvascular fraction. With glomeruli, saturation isotherms for specific [125I]HPIA binding were mono-phasic with a KD of 1.3 nmol/l and a Bmax of 7.7 fmol/mg protein. In kinetic experiments, an association rate constant of 4.9 X 10(5) (mol/l-1 s-1 and a dissociation rate constant of 4.3 X 10(-4) s-1 were obtained, yielding a KD of 0.9 nmol/l. Adenosine analogs displaced [125I]HPIA binding with a rank order of potency typical of A1-adenosine receptors; furthermore, binding was inhibited by methylxanthines and modulated by GTP. Saturation experiments with the microvessels revealed a KD of 1.9 nmol/l and a Bmax of 13.4 fmol/mg protein. However, no inhibition of glomerular and microvascular adenylate cyclase activity could be demonstrated, but instead both 5'-N-ethylcarboxamido-adenosine (NECA) and N6-(R-phenylisopropyl)-adenosine (R-PIA) stimulated enzyme activity, with EC50 values of 0.14 mumol/l and 1.5 mumol/l, respectively. The concentration-response curve for NECA was shifted to the right (factor 9) by 10 mumol/l 8-phenyltheophylline. On the other hand, computer simulation of biphasic curves (adenylate cyclase inhibition in the presence of activation via a stimulatory receptor) indicates that the failure to observe an A1-adenosine receptor-mediated inhibition of adenylate cyclase activity in the presence of stimulatory adenosine receptors may be attributable to methodological constraints.(ABSTRACT TRUNCATED AT 250 WORDS)