We used the absorbance spectrum of the pH-sensitive dye dimethylcarboxyfluorescein to monitor intracellular pH (pHi) in the isolated perfused S3 segment of the rabbit proximal tubule, and examined the effect on pHi of switching from a HEPES to a CO2/HCO3- buffer in the lumen and/or the bath (i.e., basolateral solution). Solutions were titrated to pH 7.40 at 37 degrees C. With 10 mM acetate present bilaterally (lumen and bath), this causing steady-state pHi to be rather high (approximately 7.45), bilaterally switching the buffer from 32 mM HEPES to 5% CO2/25 mM HCO3- caused a sustained fall in pHi of approximately 0.26. However, with acetate absent bilaterally, this causing steady-state pHi to be substantially lower (approximately 6.9), bilaterally switching to CO2/HCO3- caused a transient pHi fall (due to the influx of CO2), followed by a sustained rise to a level approximately 0.18 higher than the initial one. The remainder of the experiments was devoted to examining this alkalinization in the absence of acetate. Switching to CO2/HCO3- only in the lumen caused a sustained pHi fall of approximately 0.15, whereas switching to CO2/HCO3- only in the bath caused a transient fall followed by a sustained pHi increase to approximately 0.26 above the initial value. This basolateral CO2/HCO3(-)-induced alkalinization was not inhibited by 50 microM DIDS applied shortly after CO2/HCO3- washout, but was slowed approximately 73% by DIDS applied more than 30 min after CO2/HCO3- washout. The rate was unaffected by 100 microM bilateral acetazolamide, although this drug greatly reduced CO2-induced pHi transients. The alkalinization was not blocked by bilateral removal of Na+ per se, but was abolished at pHi values below approximately 6.5. The alkalinization was also unaffected by short-term bilateral removal of Cl- or SO4=. Basolateral CO2/HCO3- elicited the usual pHi increase even when all solutes were replaced, short or long-term (> 45 min), by N-methyl-D-glucammonium/glucuronate (NMDG+/Glr-). Luminal CO2/HCO3- did not elicit a pHi increase in NMDG+/Glr-. Although the sustained pHi increase elicited by basolateral CO2/HCO3- could be due to a basolateral HCO3- uptake mechanism, net reabsorption of HCO3- by the S3 segment, as well as our ACZ data, suggest instead that basolateral CO2/HCO3- elicits the sustained pHi increase either by inhibiting an acid-loading process or stimulating acid extrusion across the luminal membrane (e.g., via an H+ pump).