Thromboxane A(2) (TXA(2)) has been proposed as a mediator of perioperative myocardial ischemia, vasoconstriction, and thrombosis. As these adverse events are minimized with epidural anesthesia, rather than general anesthesia, we hypothesized that local anesthetics would inhibit TXA(2)-receptor signaling. We used fluorometric determination of intracellular [Ca(2+)] in human K562 cells and 2-electrode voltage clamp measurements in Xenopus laevis oocytes expressing TXA(2) receptors. After 10-min incubation, lidocaine (IC(50): 1.02 +/- 0.2 x 10(-3) M), ropivacaine (IC(50): ropivacaine 6.3 +/- 0.9 x 10(-5) M), or bupivacaine (IC(50): 1.42 +/- 0.08 x 10(-7) M) inhibited TXA(2)-induced [Ca(2+)](i) in K562 cells. These data were confirmed in Xenopus oocytes recombinantly expressing TXA(2) receptors, with IC(50)s of bupivacaine 1.2 +/- 0.2 x 10(-5) M, R(+) ropivacaine 4.9 +/- 1.7 x 10(-4) M, S(-) ropivacaine 5.3 +/- 0.9 x 10(-5) M, and lidocaine 6.4 +/- 2.8 x 10(-4) M. Intracellular pathways activated by IP(3) and GTPgammaS were not significantly affected by the local anesthetics tested. QX314, a positively charged lidocaine analog, inhibited only if injected intracellularly (IC(50): 5.3 +/- 1.7 x 10(-4) M), indicating one local anesthetic target is most likely inside the cell. Benzocaine (largely uncharged) inhibited with an IC(50) of 8.7 +/- 1.8 x 10(-4) M. This suggests that some of the beneficial effects of regional anesthesia techniques might be due to direct interaction of local anesthetics with the functioning of membrane proteins.