CATION PERMEABILITY OF A CLONED RAT EPITHELIAL AMILORIDE-SENSITIVE NA+ CHANNEL

1. Conductance of heterotrimeric rat epithelial Na+ channels (alpha, b eta, gamma-rENaCs) for Li+ and Na+ in planar lipid bilayers was a non- linear function of ion concentration, with a maximum of 30.4 +/- 2.9 p S and 18.5 +/- 1.9 pS at 1 M Li+ and NA(+), respectively. 2. The alpha ,beta,gamma-rENaC conductance measured in symmetrical mixtures of Na+- Li+ (1 M) exhibited an anomalous mole fraction dependence, with a mini mum at 4:1 Li+ to Na+ molar ratio. 3. Permeability ratios P-K/P-Na and P-Li/P-Na of the channel calculated from the biionic reversal potenti als were dependent on ion concentration: P-K/P-Na was 0.11 +/- 0.01, a nd P-Li/P-Na was 1.6 +/- 0.3 at 50 mM; P-K/P-Na was 0.04 +/- 0.01 and P-Li/P-Na was 2.5 +/- 0.4 at 3 M, but differed from the ratios of sing le-channel. conductances in symmetrical Li+, Na+ or K+ solutions. The permeability sequence determined by either method was Li+ > Na+ > K+ m uch greater than Rb+ > Cs+. 4. Predictions of a model featuring two bi nding sites and three energy barriers (2S3B), and allowing double occu pancy, developed on the basis of single ion current-voltage relationsh ips, are in agreement with the observed conductance maximum in single ion experiments, conductance minimum in the mole fraction experiments, non-linearity of the current-voltage curves in biionic experiments, a nd the concentration dependence of permeability ratios. 5. Computer si mulations using the 2S3B model recreate the ion concentration dependen cies of single-channel conductance observed for the immunopurified bov ine renal amiloride-sensitive Naf channel, and short-circuit current i n frog skin, thus supporting the hypothesis that ENaCs form a core con duction unit of epithelial Na+ channels.