The molecular mechanism and potential dependence of the Na+/glucose cotransporter

Activity of the Na+/glucose cotransporter endogenously expressed in LLC-PK1 cells was measured using whole cell recording techniques under three different sodium concentration conditions: 1) externally saturating, zero trans; 2) 40 mM external, zero trans; and 3) externally saturating, 50 mM trans. Activity of the transporter with increasing concentrations of sugar was measured for each set of conditions, from which the maximal current for saturating sugar, Im, was determined. The Im measured shows substantial potential dependence for each set of conditions. The absolute Im and the relative potential dependence of Im compared among the various solute conditions were used to identify which loci in the transport cycle are responsible for potential-dependent changes in function. The experimental data were compared with the predicted Im values calculated from an eight-state, sequential, reversible model of a transport reaction kinetic scheme. Predictions derived from assignment of rate limitation and/or potential dependence to each of the 16 transitions in the transport pathway were derived and compared with the measured data. Most putative models were dismissed because of lack of agreement with the measured data, indicating that several steps along the transport pathway are not rate limiting and/or not potential dependent. Only two models were found that can completely account for the measured data. In one case, translocation of the free carrier must be rate limiting, and both extracellular sodium-binding events as well as translocation of both free and fully loaded carrier forms must be potential-dependent transitions. In the second case, translocation of the free carrier and dissociation of the first sodium to be released intracellularly must be equivalently rate limiting. In this case only the two translocation events are required to be potential dependent. The two external sodium-binding events might still be potential dependent, but this is not required to fit the data. Previous reports suggest that the first model is correct; however, no direct experimental data compel us to dismiss the second option as a feasible model.