Profiles of permeation through Na-channels

Abstract

South American frogs (genus, Phyllobates) have already advanced the cause of Na-channel biophysics with their gift of the alkaloid, batrachotoxin (BTX). This toxin greatly prolongs the normally transient open state of the Na-channel, permitting its conductance mechanism to be rigorously explored in artificial lipid bilayers. Now, skeletal muscle from the Chilean toad, Caudiverbera caudiverbera, has become the gourmet choice of Nachannels for an in-depth analysis of conductance substates in the article by Naranjo and Latorre ( 1) of this issue. This paper serves to distill a substantial body of research on biophysical aspects of ion movement through this important class of voltage-gated channels. To fully appreciate the baroque nature of the arguments in this field, a sense of history required. For certain species ofsingle BTX-modified Na-channels, several groups have focused primarily on conductance behavior in the range of [Na+] less than 500 mM and optimistically interpreted the relationship between unitary conductance ('y) and symmetrical [Na+] as a MichaelisMenten function for a channel with a single binding site and a Km for Na+ of 8-40 mM (2-4). However, this lovely parallel between the behavior of an ion channel and that ofan uncomplicated enzyme blithely disregards the classic literature of macroscopic electrophysiology, which produced evidence for multi-site occupancy by alkali cations (5). Furthermore, there was the small unmentioned detail, that ionic strength was not constant in these measurements, leaving open the possibility of a large variation in surface potential arising from fixed charges associated with the channel protein. The multi-ion question could be cavalierly discarded by saying that the BTX-modified Na+ channel was different from the normal one, but the question of surface charge persisted. Green et al. (6) took aim on this latter issue by measuring conductance of canine brain Na+channels in salt solutions up to 3.5 M NaCl. Even at this high concentration ofNa+, the conductance still was not fully saturated, and the channel was clearly not Michaelis-Menten. By introducing surface charge near the mouth of the channel with the aid of Gouy-Chapman (GC) theory, Green et al. (6) were able to model their data on the basis ofa single-ion channel with substantial negative surface charge. This fixed charge was proposed to cause a large enhancement in the local Na+ concentration at low ionic strength which boosts conductance in this limit. As NaCl increased, negative surface potential screened and greatly diminished, which results in unmasking of the intrinsic low affinity of the channel for Na+, estimated to have a KD = 1.5 M. Although the GC theory ofplanar surface charge has a hallowed reputation in describing the surface electrostatics oflipid bilayers, it has a small quirk when it comes to trying to understand what goes on at the entrance to a channel's vestibule at low ionic strength. Because this theory based on the geometry of an infinite plane of smeared negative charge, the surface cation concentration approaches a non-zero value in the limit oflow ionic strength. When coupled to an ion channel, this theory predicts, that conductance saturates at some non-zero value, as [Na+] in the bulk solution reduced to zero. Whether this occurs or not can be very difficult to determine, for the range at which this effect kicks in, can be well below 10 mM NaCl. As Cai and Jordan (7) have recently pointed out, with respect to the question ofhow protein charge distributions affect the conductance of a channel, GC theory is not only too simple but also inappropriate. This a fine thing to say if one has access to the three-dimensional structure of a channel with the location of all the relevant surface charges, however for Na-channels, the reality otherwise. Nevertheless, by numerically solving the Poisson-Boltzmann equation, Cai and Jordan (7) showed that for an hourglass-shaped, single-site channel with negative surface charge very near the entrance to the pore, conductance does approach zero at low Na+ and approximates the kind of behavior found by Green et al. (6). This ameliorated the surface charge question, but short-shrifted the issue of multiple occupancy. How many ions can simultaneously bind in the pore? By glancing sideways into the fields of structurally related Cachannels and especially, K-channels, it not hard to tell that the ion occupancy number greater than one. Can a Na-channel be so different? Swinging the pendulum back to the other extreme, Ravindran et al. (8) reexamined Na+ conductance ofthe rat muscle channel over the wider range of 0.5 to 3,000 mM Na+. When the data