Orthograde and retrograde microperfusion experiments were conducted in Sprague-Dawley rats to evaluate the participation of vasoconstrictive eicosanoids as mediators of tubuloglomerular feedback (TGF) signals. Retrograde perfusion with 160 microM arachidonic acid (AA) added to a hypotonic solution enhanced the stop-flow pressure (SFP) feedback responses compared with those obtained with the control hypotonic solution (delta SFP, 1.6 +/- 0.4 vs. 10.1 +/- 0.7 mmHg with AA). Blockade of thromboxane A2 (TxA2) with the receptor blocker EP 092 or the synthesis inhibitor UK 38485 did not alter the magnitude of the SFP feedback responses obtained with an isotonic solution. Similarly, nordihydroguaiaretic acid, a lipoxygenase inhibitor, did not alter maximal SFP feedback responses. Although indomethacin (5 mM) did induce attenuated SFP feedback responses (delta SFP, 9.5 +/- 0.7 vs. 0.5 +/- 0.4 mmHg with indomethacin), normal feedback responses were restored within 15-90 s after cessation of indomethacin perfusion. Additionally, SFP feedback responses were not inhibited with 5 mM piroxicam, a different cyclooxygenase inhibitor. These data fail to support a role for either TxA2 or lipoxygenase end products as mediators of TGF signals. The rapid restoration of feedback responses after indomethacin exposure and the lack of blockade with piroxicam suggest that transmission of feedback signals is not dependent on cyclooxygenase products.