We hypothesized that tumor necrosis factor-alpha (TNF-alpha) mimics endotoxin in attenuating endothelium-dependent vasodilation and smooth muscle constriction of rat renal arteries, and that tyrosine kinase is involved. Isolated rat renal arteries (n =6 per group), pretreated for 2 h by genistein (4',5,7-trihydroxyisoflavone, 10 microg/mL, a tyrosine kinase inhibitor) or vehicle, were exposed for 2 h to recombinant human (rh) TNF-alpha (100 ng/mL) or vehicle. rhTNF-alpha attenuated (P < 0.05) the constriction response to depolarizing 125 mM KCl (952.6+/-125.3 mg/mm vs. 1191.4+/-136.8 mg/mm in rhTNF-alpha-exposed and control segments, respectively), but did not affect the constriction response to norepinephrine (NE, 0.01-10 microM). Genistein did not affect the constriction response to KCl. The concentration-response relation to NE in genistein-pretreated control segments showed (P < 0.05) a rightward shift, while the maximum constriction was not affected. Genistein did not prevent a reduction (P < 0.05) by rhTNF-alpha in the maximum response to NE (721.7+/-42.4 mg/mm vs. 999.8+/-84.4 mg/mm in controls). The endothelium-dependent relaxation induced by (acetyl choline) ACh (0.001-1.0 microM) was attenuated (P < 0.05) by rhTNF-alpha (39.4%+/-6.7% and 77.4%+/-10.0% in rhTNF-alpha-exposed and control segments, respectively). The reduction (P < 0.05) in maximum ACh-induced relaxation after exposure to rhTNF-alpha was not affected by genistein (44.6%+/-3.4% and 70.8% x 2.2% in genistein-pretreated rhTNF-alpha-exposed and control segments, respectively). Hence, the attenuated endothelium-dependent relaxation and smooth muscle constriction of rat renal arteries following short-term rhTNF-alpha exposure, mimicking the effect of endotoxin, does not involve the activity of tyrosine kinase. The latter may be involved in pharmacomechanical coupling, by increasing Ca2+ sensitivity, but less in the electromechanical coupling of smooth muscle constriction.