The time course of sodium currents (INa) in squid giant axon was analysed using viscous non-electrolyte solutions on both sides of the axolemma. It slowed reversibly as the non-electrolyte concentration increased. The activation, deactivation (closing) and inactivation processes were slowed in a similar manner. The gating current of the sodium channel was also slowed to the same extent as the activation time constant. 2. The voltage dependence observed in a time constant vs. voltage relationship and a chord conductance vs. voltage relationship (activation curve), did not change significantly. 3. The gating kinetics have a similar temperature dependence in non-electrolyte solutions, showing that the basic gating mechanism did not change in these solutions and only a slight increase in the activation free energy was one of the main causes of slowing. 4. Eight non-electrolytes, formamide, ethylene glycol, glycerol, erythritol, glucose, sorbitol, sucrose and polyethylene glycol (mean molecular weight 600) were used. The amount of slowing was correlated with the gram concentration (g l-1) of non-electrolytes, but not with molar concentration (M) and solution osmolarity (osmol l-1). 5. The percentage changes of the time constant were expressed as a function of the relative change in solution viscosity, eta/eta0. The proportionality constants alpha in the relationship alpha (eta/eta0), and gamma in the relationship 100 (eta/eta0)gamma, obtained using different non-electrolytes, were close to 100% and 1, respectively. The simplest model to explain the results assumes that a slowing of a global conformational change is a consequence of sequential viscosity-dependent movements of local structures (viscosity model). 6. Values of alpha and gamma deviated frequently from those in an ideal case, i.e. 100% for alpha and 1 for gamma, and they scattered, having a tendency to decrease as a function of molecular weight. 7. The slowing was also expressed as an exponential function of the solution osmolarity. A predicted solute-inaccessible volume Va ranged (in nm3 per molecule) between 0.09 and 1.45. The value of Va increased as a logarithmic function of the molecular weight of the non-electrolyte. 8. This solute-inaccessible volume should be distributed in all hydrophilic parts of the sodium channel protein, but is not located in the channel conducting pore itself. The slowing of gating could be explained by a model in which a rate-limiting step is a hydration process that occurs after local small structural changes have exposed new, unhydrated faces (transient hydrated-states model). 9. Considering the opposite dependencies of parameters alpha (or gamma) and beta on the molecular weight, sodium channel gating is likely to reflect a combination of these two models, which are coupled in microscopic segment movements. We emphasize with this combination of models that fluctuating hydrophilic structures play an important role in determining time constants in the gating process.
