EFFECT OF EXPRESSING THE WATER CHANNEL AQUAPORIN-1 ON THE CO2 PERMEABILITY OF XENOPUS OOCYTES

It is generally accepted that gases such as CO2 cross cell membranes b y dissolving in the membrane lipid. No role for channels or pores in g as transport has ever been demonstrated. Here we ask whether expressio n of the water channel aquaporin-1 (AQP1) enhances the CO2 permeabilit y of Xenopus oocytes. We expressed AQP1 in Xenopus oocytes by injectin g AQP1 cRNA, and we assessed CO2 permeability by using microelectrodes to monitor the changes in intracellular pH (pH(i)) produced by adding 1.5% CO2/10 mM HCO3- to (or removing it from) the extracellular solut ion. Oocytes normally have an undetectably low level of carbonic anhyd rase (CA), which eliminates the CO2 hydration reaction as a rate-limit ing step. We found that expressing AQP1 (vs. injecting water) had no m easurable effect on the rate of CO2-induced pH(i) changes in such low- CA oocytes: adding CO2 caused pH(i) to fall at a mean initial rate of 11.3 x 10(-4) pH units/s in control oocytes and 13.3 x 10(-4) pH units /s in oocytes expressing AQP1. When we injected oocytes with water, an d a few days later with CA, the CO2-induced pH(i) changes in these wat er/CA oocytes were more than fourfold faster than in water-injected oo cytes (acidification rate, 53 x 10(-4) pH units/s). Ethoxzolamide (ETX ; 10 mu M), a membrane-permeant CA inhibitor, greatly slowed the pH(i) changes (16.5 x 10(-4) pH units/s). When we injected oocytes with AQP 1 cRNA and then CA, the CO2-induced pH(i) changes in these AQP1/CA ooc ytes were similar to 40% faster than in the water/CA oocytes (75 x 10( -4) pH units/s), and ETX reduced the rates substantially (14.7 x 10(-4 ) pH units/s). Thus, in the presence of CA, AQP1 expression significan tly increases the CO2 permeability of oocyte membranes. Possible expla nations include 1) AQP1 expression alters the lipid composition of the cell membrane, 2) AQP1 expression causes overexpression of a native g as channel, and/or 3) AQP1 acts as a channel through which CO2 can per meate. Even if AQP1 should mediate a CO2 flux, it would remain to be d etermined whether this CO2 movement is quantitatively important.