Abstract. Bilayer membranes, formed from various phospholipids, were studied to assess the influence of the charge of the polar head groups on the membrane conductance mediated by neutral "carriers" of cations and anions. The surface charge of an amphoteric lipid, phosphatidyl ethanolamine, was altered by varying the pH, and the surface charge of several lipids was screened by increasing the ionic strength of the solution with impermeant monovalent and divalent electrolytes. The-surface charge should be a key parameter in defining the membrane conductance for a variety of permeation mechanisms; conductance measurements in the presence of carriers may be used to estimate the potential difference, due to surface charge, between the interior of the bilayer and the bulk aqueous phase. The large changes in conductance observed upon varying the surface charge density and the ionic strength agree with those predicted by the Gouy-Chapman theory for an aqueous diffuse double layer. Explicit expressions for the dependence of the membrane conductance on the concentrations of the carrier, the permeant ion, the surface charge density, and the ionic strength are presented.With the development of the artificial phospholipid bilayer membrane,1 2 the relationship between phospholipid composition and permeability properties of the membrane has become accessible to experimental study. The significance of the fluidity of the membrane for solute permeation has been shown.3 Moreover, surface potential measurements on phospholipid monolayers and studies of the electrophoretic mobility of phospholipid dispersions have demonstrated that the presence of a charged polar head group produces a substantial potential at the lipid-solution interface. Such a potential should influence the concentration of ions at the interface, and hence, the permeability properties of the membrane. Studies on phospholipid vesicles4 5 and bilayers6 7 have indeed shown that the anion to cation permselectivity is influenced by the charge of the membrane. Furthermore, the conductance of a bilayer due to the iodide ion or certain weak acids is affected by the charge on the polar head group of the phospholipid.8 9 The present study examines theoretically and experimentally how certain molecules which enhance the membrane conductance, probably by acting as carriers10-12 of particular cations and anions, may be used as "probes" to distinguish between effects of charge and of other variables, such as the fluidity of the bilayer interior and the lipid solubility of the complex.

A B S T R A C T The electrical properties of "inward" rectifying egg cell membranes of the starfish Mediastera aequalis have been studied in the presence of K+-TI + mixtures. When the ratio of the external concentrations of these ions is changed while their sum is kept constant, both the conductance and the zero-current membrane potential go through a minimum, showing clear discrepancies from theoretical results based on conventional electrodiffusion models (e.g., Goldman's equation). By contrast, when the ratio of the two concentrations is fixed and their sum varied, the potential follows an ideal Nernst slope, consistent with Goldman's equation. The membrane conductance which, according to previous studies on similar membranes, is to be viewed as a function of the displacement of the membrane potential from its resting value AV, shows marked differences between the cases in which K + or T1 ÷ are the predominant ions: when K ÷ is the predominant permeant ion in solution, the addition of small amounts of T1 + inhibits the current, while corresponding blocking effects of K + on the current are not observed when TI + is the predominant permeant ion. Also, the time course of the conductance during voltage clamp is different in the two cases, being much faster in T1 + than in K ÷ solution for comparable values of AV. Most of the above features are accounted for by a model in which it is assumed that the ionic channels have external binding sites for cations and that their permeability properties depend on the species of the cation bound (K + or TI ÷ in the present experiments).

This paper, the last in a series of three, characterizes the electrical properties of phospholipid bilayer membranes exposed to aqueous solutions containing nonactin, monactin, dinactin, and trinactin and Li(+), Na(+), K(+), Rb(+), Cs(+), and NH 4 (+) ions. Not only are both the membrane resistance at zero current and the membrane potential at zero current found to depend on the aqueous concentrations of antibiotic and ions in the manner expected from the theory of the first paper, but also these measurements are demonstrated to be related to each other in the manner required by this theory for "neutral carriers". To verify that these antibiotics indeed are free to move as carriers of cations, cholesterol was added to the lipid to increase the "viscosity" of the interior of the membrane. Cholesterol decreased by several orders of magnitude the ability of the macrotetralide antibiotics to lower the membrane resistance; nevertheless, the permeability ratios and conductance ratios remained exactly the same as in cholesterolfree membranes. These findings are expected for the "carrier" mechanism postulated in the first paper and serve to verify it. Lastly, the observed effects of nonactin, monactin, dinactin, and trinactin on bilayers are compared with those predicted in the preceding paper from the salt-extraction equilibrium constants measured there; and a close agreement is found. These results show that the theory of the first paper satisfactorily predicts the effects of the macrotetralide actin antibiotics on the electrical properties of phospholipid bilayer membranes, using only the thermodynamic constants measured in the second paper. It therefore seems reasonable to conclude that these antibiotics produce their characteristic effects on membranes by solubilizing cations therein as mobile positively charged complexes.

The single-channel recording technique was employed to investigate the mechanism conferring ATP sensitivity to a metabolite-sensitive K channel in insulin-secreting cells. ATP stimulated channel activity in the 0-10/zM range, but depressed it at higher concentrations. In inside-out patches, addition of the cAMPdependent protein kinase inhibitor (PKI) reduced channel activity, suggesting that the stimulatory effect of ATP occurs via cAMP-dependent protein kinase-mediated phosphorylation. Raising ATP between 10 and 500/~M in the presence of exogenous PKI progressively reduced the channel activity; it is proposed that this inactivation results from a reduction in ldnase activity owing to an ATP-dependent binding of PKI or a protein with similar inhibitory properties to the kinase. A model describing the effects of ATP was developed, incorporating these two separate roles for the nucleotide. Assuming that the efficacy of ATP in controlling the channel activity depends upon the relative concentrations of inhibitor and catalytic subunit associated with the membrane, our model predicts that the channel sensitivity to ATP will vary when the ratio of these two modulators is altered. Based upon this, it is shown that the apparent discrepancy existing between the sensitivity of the channel to low ATP concentrations in the excised patch and the elevated intracellular level of ATP may be explained by postulating a change in the inhibitor/kinase ratio from 1:1 to 3:2 owing to the loss of protein kinase after patch excision. At a low concentration of ATP (10-20 #M), a nonhydrolyzable ATP analogue, AMP-PNP, enhanced the channel activity when present below 10 #M, whereas the analogue blocked the channel activity at higher concentrations. It is postulated that AMP-PNP inhibits the formation of the kinase-inhibitor complex in the former case, and prevents phosphate transfer in the latter. A similar mechanism would explain the interaction between ATP and ADP which is characterized by enhanced activity at low ADP concentrations and blocking at higher concentrations.

Some effects of the external pH on Ca channels were studied in a hybridoma cell line (mAb-7B), by using the whole-cell configuration of the patch-clamp technique. As the pH was lowered, both the activation and the inactivation curves shifted toward less negative membrane potentials, suggesting a pH-induced decrease of an external negative surface potential, sensed by the mechanism of gating. The potential for half-activation, V1/2, and that for half-inactivation, Vh, were related by a straight line with a slope of one. The inward current varied exponentially with V1/2, as would be expected if the field inside the channel and the Ca2 concentration at the entrance were sensitive to the surface potential. However, the reversal potential and the outward current were unaltered by changes in the pH. Under the hypothesis that the channel senses the surface potential, all these results, as well as the nernstian behavior of the reversal potential with respect to Ca2+, observed in previous studies, are accounted for by a three-barrier, two-ion model for a channel, provided it is assumed that the potential in the channel drops almost entirely across the barrier adjacent to the external solution.Voltage-gated Ca channels are found in hybridoma cell lines constructed by fusion of mouse myeloma cell line S194 and mouse splenic B lymphocytes (1). Voltage-clamp type analyses revealed that the current through the channel is carried in the inward direction by external Ca ions and in the outward direction by internal alkali cations. No other types of voltage-gated ion channels were found in these cells (2).In the present experiments, the pH of the external solution was changed keeping all other external and internal ion concentrations constant. The gating kinetics of the membrane current shifted in the positive direction along the voltage axis when the external pH was reduced. This indicates that the surface potential is altered by the external H+ concentration and that the potential changes are sensed by the gating mechanism. The aim of this work is to investigate whether or not a similar surface potential is also sensed by the channel itself. In studies of Ca channels some investigators have simply assumed that surface potential changes alter the field in the channel as well as in the surrounding lipid area, whereas others have asserted that only gating is altered, not the channel conductance. In most cases the assumptions had no clear experimental support. The same question has also been raised for the Na channel, leading to different conclusions depending on the type of preparation (3, 4). The conclusion that the Na channel in the node of Ranvier does not sense the surface potential (4) has been deduced from the behavior of the outward current, which was found to be insensitive to the ionic strength and the external pH. Difficulties in isolating the outward current have been the major obstacle to performing equally significant experiments with Ca channels.However, in the present preparation we can record clear outward ...