Evidence from studies of temperature-dependent changes of D-glucose, D-mannose and L-sorbose permeability that different states of activation of the human erythrocyte hexose transporter exist for good and bad substrates.

(1) The inhibition constant of L-sorbose flux from fresh human erythrocytes by D-glucose, Ki(sorbose) increases on cooling from 50 degrees C to 30 degrees C from 5.15 +/- 0.89 mM to 12.24 +/- 1.9 mM; the Ki(sorbose) of D-mannose increases similarly, indicating that the process is endothermic. (2) The activation energy Ea(sorbose) of net L-sorbose exit is 62.9 +/- 3.1 kJ/mol; in the co-presence of 5 mM D-glucose Ea(sorbose) is reduced to 41.7 +/- 1.6 kJ/mol (P < 0.005). (3) Cooling from 35 degrees C to 21 degrees C decreases the Ki(inf, cis) of auto-inhibition of D-glucose net exit from 5.2 +/- 0.3 mM to 1.36 +/- 0.06 mM; the Ki(inf, cis) of D-mannose falls from 10.9 +/- 1.65 mM to 5.7 +/- 0.3 mM. (4) The activation energy of D-glucose zero-trans net exit is 34.7 +/- 2.1 kJ/mol and that of D-mannose exit is 69.4 +/- 3.7 kJ/mol (P < 0.0025). (5) The exothermic and exergonic processes of auto-inhibition of D-glucose net exit are larger than those for D-mannose (P < 0.03). These data are consistent with D-glucose binding promoting an activated transporter state which following dissociation transiently remains; if an L-sorbose molecule binds within the relaxation time after D-glucose dissociation, it will have a higher mobility than otherwise. Cooling slows the relaxation time of the activated state hence raises the probability that L-sorbose will bind to the glucose-activated transporter. D-Glucose donates twice as much energy to the transporter as D-mannose, consequently produces more facilitation of flux. This view is inconsistent with the alternating carrier model of sugar transport in which net flux is considered to be rate-limited by return of the empty carrier, but is consistent with fixed two-site models.