The use of physical methods in determining gramicidin channel structure and function.

The various details provided by physical methods have been integrated into a coherent picture of the cation transport process. The free energy profile describes simultaneously the thermodynamic and kinetic factors that govern conductance. Localization of ion binding sites near each end of the channel by NMR and X-ray crystallographic techniques demonstrates that there is an energy barrier separating the two ends. There is no a priori reason to suspect any barrier to ion entry into the channel, and recent molecular modeling computations confirm this intuition. Therefore, the channel is best described as a two-site, one-barrier channel. However, the movement of ions across the central barrier over a distance of 1.9 nm is really a multistep process best described as Brownian motion (even the short steps from one pair of carbonyls to the next is more diffusive than activated for ions as large as Na+), the entry step is probably best considered as a compound step requiring correction for interfacial polarization and diffusion limitation, the binding sites in the channel ought not be considered to be in equilibrium with the bath, and quasi-knock off behavior (binding of a second ion at the entry facilitates release of a first ion from the exit) is probably the rule at physiological permeant ion concentrations and higher. Progress will probably focus on channel kinetics. The channel backbone probably undergoes conformational changes as the ions pass which, according to solid state NMR results, may last long enough to provide the channel with a sort of memory and which may give rise to excess single channel noise. These conformational changes are further reflected in shifts in side chain positions which, in turn, may be partly responsible for changes in the single-channel lifetime, which appears to be exquisitely sensitive to side chain-lipid interactions. Reasonable explanations for the impermeance of divalent cations, anions, and medium-sized organic cations such as guanidinium have proven elusive at first, but it now appears that divalent cations bind water too tightly, anions bind water in an orientation that produces unfavorable contacts between water oxygens and peptide carbonyl oxygens at the channel entry, and iminium ions bind strongly to the flexible peptide carbonyls at the channel entry.