Previous studies demonstrated that the proximal tubule (PT) responds to isolated increases in basolateral ([CO(2)](BL)) or "bath" CO(2) concentration by increasing the HCO(3)(-) reabsorption rate (J(HCO(3))). Blockade of the rabbit apical AT(1) receptor or knockout of the mouse AT(1A) receptor eliminates these effects, demonstrating a requirement for luminal ANG II that the PT itself synthesizes. In the present study, we examined the effects of the ACE inhibitor lisinopril on J(HCO(3)) in isolated perfused rabbit PTs (S2 segment), using out-of-equilibrium solutions to make isolated changes in [CO(2)](BL) at a fixed baseline HCO(3)(-) concentration of 22 mM and fixed baseline pH of 7.4. Adding 60 or 240 nM lisinopril (in vitro K(i): 0.5 or 1.2 nM) to the lumen had no effect. These results are not consistent with the hypothesis that the PT secretes either angiotensinogen or ANG I. However, adding 60 nM basolateral lisinopril significantly decreased J(HCO(3)) at a [CO(2)](BL) of 20%. Moreover, 240 nM basolateral lisinopril decreased baseline (i.e., at 5% CO(2)) J(HCO(3)) by one-half and completely eliminated the response to altering [CO(2)](BL) from 0 to 20%, but left intact the stimulatory effect of 10(-11) M basolateral ANG II. At extremely high concentrations (i.e., 100 microM), luminal lisinopril replicated the effects of 240 nM basolateral lisinopril. Our data are consistent with the hypothesis that lisinopril readily crosses the basolateral (but not apical) membrane to block ACE in a vesicular compartment. We conclude that the isolated PT predominantly secretes preformed ANG II, rather than angiotensinogen or ANG I.