Search

The permeability of K channels to various cations is studied in myelinated nerve. Ionic currents under voltage clamp are measured in Ringer solution containing tetrodotoxin and a high concentration of the test ion. Reversal potentials for current in K channels are determined and used with the Goldman-Hodgkin-Katz equation to calculate relative permeabilities. The ratios P(Tl):P(K):P(Rb):P(NHNH4) are 2.3:1.00:0.92:0.13. No other ions are found to be measurably permeant including Li(+), Na(+), Cs(+), methylamine, guanidine, hydrazine, or hydroxylamine. The ratio P(Na)/P(K) is less than 0.01. Potassium conductance is depressed at pH values below 5.0. Leakage conductance is higher in K, Rb, Cs, NH(4), and Tl Ringer than in Na Ringer, but the selectivity sequence probably is not the same as for K channels. The hypothesis is offered that the narrowest part of the K channel is a circle of oxygen atoms about 3 A in diameter with low electrostatic field strength.

The relative permeability of sodium channels to eight metal cations is studied in myelinated nerve fibers. Ionic currents under voltage-clamp conditions are measured in Na-free solutions containing the test ion. Measured reversal potentials and the Goldman equation are used to calculate the permeability sequence: Na(+) approximately Li(+) > Tl(+) > K(+). The ratio P(K)/P(Na) is 1/12. The permeabilities to Rb(+), Cs(+), Ca(++), and Mg(++) are too small to measure. The permeability ratios agree with observations on the squid giant axon and show that the reversal potential E(Na) differs significantly from the Nernst potential for Na(+) in normal axons. Opening and closing rates for sodium channels are relatively insensitive to the ionic composition of the bathing medium, implying that gating is a structural property of the channel rather than a result of the movement or accumulation of particular ions around the channel. A previously proposed pore model of the channel accommodates the permeant metal cations in a partly hydrated form. The observed sequence of permeabilities follows the order expected for binding to a high field strength anion in Eisenman's theory of ion exchange equilibria.

The sodium channel of the excitability mechanism in nerve membranes is about as permeable to hydroxylamine and hydrazine cations as it is to sodium ions. It is impermeable to methylamine cations. This selectivity is explained by supposing that an oxygen group in the channel must receive a hydrogen bond from the permeating cation at the same time as the cation lies against another negatively charged oxygen acid. If these conditions are not satisfied the cation cannot permeate. Sodium ions can satisfy this hydrogen-bonding requirement if they have a water of hydration. The H(2)O.Na complex also has almost the same dimensions as the hydroxylamine and hydrazine cations. This hydrated ion is probably part of the critical complex between sodium ions and the selectivity mechanism of the sodium channel.

The effect of tetraethylammonium ion (TEA) on the voltage clamp currents of nodes of Ranvier of frog myelinated nerve fibers is studied. The delayed K currents can be totally abolished by TEA without affecting the transient Na currents or the leakage current in any way. Both inward and outward currents disappear. In low TEA concentrations small K currents remain with normal time constants. The dose-response relationship suggests the formation of a complex between TEA and a receptor with a dissociation constant of 0.4 mM. Other symmetrical quaternary ammonium ions have very little effect. There is no competition between TEA and agents that affect the Na currents such as Xylocaine, tetrodotoxin, or Ca ions. The pharmacological data demonstrate that the Na, K, and leakage permeabilities are chemically independent, probably because their mechanisms occupy different sites on the nodal membrane. The data are gathered and analyzed by digital computer.

Quaternary ammonium ions were applied to the inside of single myelinated nerve fibers by diffusion from a cut end. The resulting block of potassium channels in the node of Ranvier was studied under voltage-clamp conditions. The results agree in almost all respects with similar studies by Armstrong of squid giant axons. With tetraethylammonium ion (TEA), pentyltriethylammonium ion (C(5)), or nonyltriethylammonium ion (C(9)) inside the node, potassium current during a depolarization begins to rise at the normal rate, reaches a peak, and then falls again. This unusual inactivation is more complete with C(9) than with TEA. Larger depolarizations give more block. Thus the block of potassium channels grows with time and voltage during a depolarization. The block reverses with repolarization, but for C(9) full reversal takes seconds at -75 mv. The reversal is faster in 120 mM KCl Ringer's and slower during a hyperpolarization to -125 mv. All of these effects contrast with the time and voltage-independent block of potassium, channels seen with external quaternary ammonium ions on the node of Ranvier. External TEA, C(5), and C(9) block without inactivation. The external quaternary ammonium ion receptor appears to be distinct from the inner one. Apparently the inner quaternary ammonium ion receptor can be reached only when the activation gate for potassium channels is open. We suggest that the inner receptor lies within the channel and that the channel is a pore with its activation gate near the axoplasmic end.

The relative permeability of sodium channels to 21 organic cations was studied in myelinated nerve fibers. Ionic currents under voltage-clamp conditions were measured in sodium-free solutions containing the test cation. The measured reversal potential and the Goldman equation were used to calculate relative permeabilities. The permeability sequence was: sodium approximately hydroxylamine > hydrazine > ammonium approximately formamidine approximately guanidine approximately hydroxyguanidine > aminoguanididine >> methylamine. The cations of the following compounds were not measurably permeant: N-methylhydroxylamine, methylhydrazine, methylamine, methylguanidine, acetamidine, dimethylamine, tetramethylammonium, tetraethylammonium, ethanolamine, choline, tris(hydroxymethyl)amino methane, imidazole, biguanide, and triaminoguanidine. Thus methyl and methylene groups render cations impermeant. The results can be explained on geometrical grounds by assuming that the sodium channel is an oxygen-lined pore about 3 A by 5 A in cross-section. One pair of oxygens is assumed to be an ionized carboxylic acid. Methyl and amino groups are wider than the 3 A width of the channel. Nevertheless, cations containing amino groups can slide through the channel by making hydrogen bonds to the oxygens. However, methyl groups, being unable to form hydrogen bonds, are too wide to pass through.

Voltage clamp measurements on myelinated nerve fibers show that tetrodotoxin, saxitoxin, and DDT specifically affect the sodium channels of the membrane. Tetrodotoxin and saxitoxin render the sodium channels impermeable to Na ions and to Li ions and probably prevent the opening of individual sodium channels when one toxin molecule binds to a channel. The apparent dissociation constant of the inhibitory complex is about 1 nM for the cationic forms of both toxins. The zwitter ionic forms are much less potent. On the other hand, DDT causes a fraction of the sodium channels that open during a depolarization to remain open for a longer time than is normal. The effect cannot be described as a specific change in sodium inactivation or as a specific change in sodium activation, for both processes continue to govern the opening of the sodium channels and neither process is able to close the channels. The effects of DDT are very similar to those of veratrine.

The voltage dependence of the voltage clamp responses of myelinated nerve fibers depends on the concentration of divalent cations and of hydrogen ions in the bathing medium. In general, increases of the [Ca], [Ni], or [H] increase the depolarization needed to elicit a given response of the nerve. An e-fold increase of the [Ca] produces the following shifts of the voltage dependence of the parameters in the Hodgkin-Huxley model: m(infinity), 8.7 mv; h(infinity), 6.5 mv; tau(n), 0.0 mv. The same increase of the [H], if done below pH 5.5, produces the following shifts: m(infinity), 13.5 mv; h(infinity), 13.5 mv; tau(n), 13.5 mv; and if done above pH 5.5: m(infinity), 1.3 mv; h(infinity), 1.3 mv; tau(n), 4.0 mv. The voltage shifts are proportional to the logarithm of the concentration of the divalent ions and of the hydrogen ion. The observed voltage shifts are interpreted as evidence for negative fixed charges near the sodium and potassium channels. The charged groups are assumed to comprise several types, of varying affinity for divalent and hydrogen ions. The charges near the sodium channels differ from those near the potassium channels. As the pH is lowered below pH 6, the maximum sodium conductance decreases quickly and reversibly in a manner that suggests that the protonation of an acidic group with a pK(a) of 5.2 blocks individual sodium channels.