Effect of HCO(3)(-) on TPA- and IBMX-induced anion conductances in Necturus gallbladder epithelial cells.

Effects of HCO(3)(-) on protein kinase C (PKC)- and protein kinase A (PKA)-induced anion conductances were investigated in Necturus gallbladder epithelial cells. In HCO(3)(-)-free media, activation of PKC via 12-O-tetradecanoylphorbol 13-acetate (TPA) depolarized apical membrane potential (V(a)) and decreased fractional apical voltage ratio (F(R)). These effects were blocked by mucosal 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB), a Cl(-) channel blocker. In HCO(3)(-) media, TPA induced significantly greater changes in V(a) and F(R). These effects were blocked only when NPPB was present in both mucosal and basolateral compartments. The data suggest that TPA activates NPPB-sensitive apical Cl(-) conductance (g(Cl)(a)) in the absence of HCO(3)(-); in its presence, TPA stimulated both NPPB-sensitive g(Cl)(a) and basolateral Cl(-) conductance (g(Cl)(b)). Activation of PKA via 3-isobutyl-1-methylxanthine (IBMX) also decreased V(a) and F(R); however, these changes were not affected by external HCO(3)(-). We conclude that HCO(3)(-) modulates the effects of PKC on g(Cl)(b). In HCO(3)(-) medium, TPA and IBMX also induced an initial transient hyperpolarization and increase in intracellular pH. Because these changes were independent of mucosal Na(+) and Cl(-), it is suggested that TPA and IBMX induce a transient increase in apical HCO(3)(-) conductance.

Left atrial spontaneous echo contrast is highly associated with previous stroke in patients with atrial fibrillation or mitral stenosis.

BACKGROUND AND PURPOSE: Spontaneous echo contrast is a dynamic smokelike signal that is detected by transesophageal echocardiography in patients with stasis of blood in the left atrium. We designed this study to determine if spontaneous echo contrast is associated with an increased risk of previous stroke or peripheral embolism. METHODS: Forty-two patients with spontaneous echo contrast were identified (34 had atrial fibrillation or mitral stenosis; 8 had neither). Control subjects comprised 40 patients randomly selected from patients with atrial fibrillation or mitral stenosis who did not have spontaneous echo contrast at transesophageal echocardiography. The frequency of vascular risk factors, echocardiographic features, and stroke or peripheral embolism within 1 year of echocardiography were compared in the two groups. RESULTS: The frequency of traditional risk factors for stroke were the same in both groups, yet 9 of 42 patients with spontaneous contrast had stroke or peripheral embolism compared with only 1 of 40 control subjects (P < .02; relative risk, 10.6; 95% confidence interval, 1.3 to 88.4). In patients with nonvalvular atrial fibrillation, 6 of 12 patients with spontaneous contrast had a stroke or peripheral embolism compared with 1 of 28 patients without spontaneous contrast (P < .001; relative risk, 27.0; 95% confidence interval, 2.7 to 267.8). CONCLUSIONS: Spontaneous echo contrast is highly associated with previous stroke or peripheral embolism in patients with atrial fibrillation or mitral stenosis. Transesophageal echocardiography may enable stratification of cardioembolic risk in patients with nonvalvular atrial fibrillation.

NaCl transport by Madin Darby canine kidney cyst epithelial cells.

The mechanism of NaCl transport across the epithelium of intact MDCK cysts grown in a collagen gel matrix was investigated. Double-barreled microelectrodes were used to measure basolateral membrane PD (Vbl), transepithelial PD (Vt), and intracellular (Cli) and intralumenal (Clcy) Cl- activities in cysts under different conditions. In a control Ringer's solution (RS), Cli (60 +/- 1 mM) and Clcy (107 +/- 2 mM) exceeded the values corresponding to electrochemical equilibrium across the basolateral membrane and epithelium, respectively. Cli was reduced by superfusing the cysts with a low Cl- RS (Cli, 20 +/- 3 mM), a low Na+ RS (Cli, 40 +/- 4 mM), or by adding amiloride to the control RS (Cli, 46 +/- 1 mM). Cli was unaffected by removal of either K+ or HCO3- from the RS or by adding furosemide or SITS to the control RS. Vbl in the control RS was -50 +/- 2 mV and was affected only by removal from the RS of K+ (Vbl, -31 +/- 3 mV) or HCO3- (Vbl, -29 +/- 4 mV) or by the addition of SITS to the control RS (Vbl, -59 +/- 5 mV). Vt in control RS was -2 +/- 0.2 mV (lumen negative), and was increased by reducing bath Na+ (Vt, -37 +/- 2 mV) but not by reducing bath Cl-. These data indicate that Cl- is secreted in a basolateral to apical direction by the cyst epithelium. Basolateral Cl- transport probably occurs mainly by an electroneutral Cl-/HCO3- exchanger. Transepithelial Na+ transport seems to occur via a paracellular route which appears to be cation selective. These experiments also support the existence, in the basolateral membrane, of a Na+/K+ ATPase, a Na+/H+ exchanger, and possibly a Na+/HCO3-/CO3(2-) transporter.

A possible relationship between KCl symport and basolateral K(+)-conductance in Necturus gallbladder epithelial cells.

1. Apical membrane potential (Va), transepithelial potential (VT), fractional apical voltage ratio (FVa = delta Va/delta VT), tissue resistance (RT), and intracellular Cl- (aiCl) and K+ (aiK) activities were measured in isolated gallbladders maintained between oxygenated bicarbonate-free physiological media (23 degrees C, pH 7.2 or 8.2) in a divided chamber. The basolateral membrane potential (Vb) was calculated from the measured values of Va and VT. 2. Cl- removal from the serosal medium (which should accelerate coupled basolateral KCl exit) significantly depolarized Vb, decreased aiCl, decreased FVa, increased RT, and attenuated the depolarization of Vb (delta Vb) induced by high K+ added to the serosal side. These changes are consistent with a decrease in the K(+)-conductance of the basolateral membrane (gbK). 3. Addition of furosemide (an inhibitor of KCl cotransport) to the serosal medium induced significant increases in Vb, FVa, and high K(+)-induced delta Vb, indicating an increase in gbK. 4. In the presence of serosal furosemide, Cl- removal from the serosal medium did not significantly alter Vb, aiCl or delta Vb from their corresponding values when serosal Cl- was present. 5. Serosal furosemide had no significant effect on aiK and aiCl measured with double-barreled ion-selective microelectrodes. 6. These results suggest the possibility of a reciprocal relationship between gbK and the rate of basolateral KCl cotransport. This may contribute to the maintenance of aiK in gallbladder epithelial cells.

Calibration of ion-selective microelectrodes.

Measurements of intracellular and extracellular ion activities with ion-selective microelectrodes generally involve calibration of the electrodes in solutions of known composition and fitting of the calibration data with a theoretical expression. The Nicolsky equation is frequently used to describe the variation of electrode potential with primary ion activity in the presence of a constant amount of interference. In this report, we review the estimation of primary ion activities in calibration solutions and discuss the practical use of the Nicolsky equation. We describe a specialized computer program, developed in this laboratory, for routine input and editing of calibration data, fitting of data with the Nicolsky equation, and calculation of experimental ion activities from the fitted curve. Earlier versions of this program have proven helpful in several investigations in this and in other laboratories.

Liquid junction potentials calculated from numerical solutions of the Nernst-Planck and Poisson equations.

We present numerical solutions for the one-dimensional Nernst-Planck and Poisson system of equations for steady-state electrodiffusion. Commonly used approximate solutions to these equations invoke assumptions of local electroneutrality (Planck approximation) or constant electric field (Goldman approximation). Calculations were performed to test the ranges over which these approximate theories are valid. For a dilutional junction of a 1:1 electrolyte, separated from adjoining perfectly stirred solutions by sharp boundaries, the Planck approximation is valid for values of kappa dL greater than 10, where 1/kappa d is the Debye length of the more dilute solution. The Goldman approximation is valid for kappa cL less than 0.1 where 1/kappa c is the Debye length of the more concentrated solution. These results suggest that the modeling of electrodiffusive flows in and near membrane ion channels may require numerical solutions of this set of equations rather than the use of either limiting case.

Patch clamp recording from anterior pituitary cells identified by reverse hemolytic plaque assay.

The study of hormone secretion by anterior pituitary cells is complicated by the presence of multiple cell types. For unambiguous interpretation of data it is necessary to identify the cells from which measurements are made. We have described a reliable experimental approach involving the identification of cultured cells of a particular type with a reverse hemolytic plaque assay. The electrical characteristics of individual identified cells can then be studied using patch clamp recording. This electrophysiological approach is well suited to the study of complex systems in cultured cells. Although this combined approach requires some expertise in a variety of techniques, it is workable and should yield valuable information regarding the role of ion channels in the cellular control of hormone secretion by the anterior pituitary.

Gonadotropin-releasing hormone induces oscillatory membrane currents in rat gonadotropes.

Electrophysiological studies were performed to characterize membrane currents of rat gonadotropes under basal conditions and after exposure to secretagogues. Gonadotropes were identified in primary cultures of rat anterior pituitaries by a reverse hemolytic plaque assay. Giga-seal patch clamp recording with the cell-attached configuration was used to monitor membrane currents in these cells. Spontaneous spikes in basal current were seen. These were blocked by methoxyverapamil and probably reflect Ca2+-dependent action potentials. Brief GnRH stimulation induced slow oscillatory changes in membrane current that evolved into a series of large amplitude inward pulses after about 8 min. Treatment with TRH had no effect, and depolarization with K+ led to delayed inward currents without any oscillatory behavior. Under conditions of Ca2+ channel blockade, GnRH stimulation did not induce pulses of inward current, but did lead to oscillatory activation of a small conductance ion channel apparently selective for K+. Taken together these results suggest that GnRH induces oscillations in intracellular Ca2+ and that these oscillations are controlled by biochemical processes.

Double-barreled K+-selective microelectrodes based on dibenzo-18-crown-6.

Liquid ion-exchanger microelectrodes based on Corning code 477317 K+ exchanger are known to be much more sensitive to quaternary ammonium ions than to K+. In the presence of such cations, the capability of measuring K+ activities with Corning microelectrodes may be seriously impaired. We have developed a neutral carrier K+-selective microelectrode based on the crown ether dibenzo-18-crown-6. The crown ether cocktail contained (wt/wt) 2.3% dibenzo-18-crown-6, 0.8% Na-tetraphenylborate, 30.1% 2-nitrophenylocylether, and 66.8% O-nitrotoluene. Double-barreled crown ether and Corning microelectrodes were calibrated in KCl solutions with or without choline, acetylcholine, tetramethylammonium, imidazole, Na+, tris(hydroxymethyl)aminomethane (Tris), and N-methyl-D-glucamine. Both kinds of microelectrodes showed similar K+ over Na+, Tris, and N-methyl-D-glucamine selectivities. However, crown ether microelectrodes had immensely greater selectivities of K+ over quaternary ammonium ions and imidazole than Corning microelectrodes. Selectivity factors, defined as log K(ij)K, of crown ether microelectrodes with respect to K+ for tetramethylammonium, choline, acetylcholine, and imidazole were -1.92 +/- 0.13, -2.97 +/- 0.03, -1.75 +/- 0.15, and -1.30 +/- 0.20, respectively. Intracellular K+ activities measured in the same Necturus gallbladders with both kinds of microelectrodes did not differ significantly.

Rheogenic transport of basic and acidic amino acids across the brush border of Necturus small intestine.

In studies with isolated Necturus intestine, glutamate (Glu-) and Na+ each enhanced the mucosal influx of the other. Measurement of apical membrane potential, Va, with microelectrodes revealed a rapid depolarization with addition of 10 mM mucosal Glu-. This depolarization was Na+ dependent. Upon complete removal of Cl- from the bathing medium Va hyperpolarized and the Glu- -induced depolarization increased significantly. However, removal of Cl- did not alter the total Glu- influx. These data suggest that external Cl- attenuates the rheogenicity of Na+/Glu- cotransport in the apical membrane of the absorptive cells. We have presented a model consistent with these observations in which Cl- competes with one -COO- group of Glu- for its binding site on the carrier. The two complexes which may form, carrier/Glu-/2Na+ or carrier/Glu-/2Na+/Cl-, allow for either electrogenic or electroneutral transport of Glu-, depending on the ratio [Glu-]/[Cl-] in the extracellular fluid. In other experiments, addition of mucosal L-lysine (Lys+) induced a rapid depolarization of Va. In the presence of Na+, the depolarization appeared to be saturable with respect to Lys+ concentration. In Na+-free media, however, the depolarization increased with Lys+ concentration up to a maximum at 10 mM and then decreased to near zero at 30 mM. These data are consistent with a model for Lys+ entry in which an anionic site of the carrier can bind either Na+ or the epsilon-NH3+ group of Lys+. In this model transport of either complex, carrier-/Lys+ or carrier-/Lys+/Na+ (and return of the carrier to the extracellular surface) is rheogenic. However, at higher Lys+ concentrations, the epsilon-NH3+ group of a second Lys+ molecule may bind to the carrier forming a complex, carrier-/2Lys+, which is not transported.

A microcomputer interface for a digital audio processor-based data recording system.

An inexpensive interface is described that performs direct transfer of digitized data from the digital audio processor and video cassette recorder based data acquisition system designed by Bezanilla (1985, Biophys. J., 47:437-441) to an IBM PC/XT microcomputer. The FORTRAN callable software that drives this interface is capable of controlling the video cassette recorder and starting data collection immediately after recognition of a segment of previously collected data. This permits piecewise analysis of long intervals of data that would otherwise exceed the memory capability of the microcomputer.

Measurement of intracellular chloride activity in mouse liver slices with microelectrodes.

Steady-state membrane potential (Vm) and intracellular Cl- activity (aCli) were measured with double-barreled Cl(-)-selective microelectrodes in mouse liver slices. In bathing solutions (33.8 degrees C) containing pyruvate, glutamate, fumarate, and glucose, Vm and aCli were -27.6 +/- 1.0 mV and 32.6 +/- 1.5 mM, respectively. This apparent value of aCli exceeded the level required for passive distribution of this ion (aCleq = 26.4 +/- 1.3 mM) by 6.2 +/- 1.0 mM. This difference was essentially unchanged in experiments where (i) Na+ was replaced by choline, (ii) HCO3- was removed, and (iii) Cl- was replaced by gluconate. These data argue against the presence of Na+- or HCO3(-)-coupled Cl- transport mechanisms in the plasma membrane of mouse liver cells. This implies that aCli is in fact at equilibrium and interference with the response of Cl(-)-selective microelectrodes by intracellular anions is responsible for the apparent difference between aCli and aCleq. We found that Cl(-)-selective microelectrodes containing Corning 477315 ligand are sensitive to taurocholate, a representative bile salt. Their selectivity to taurocholate is about 60-times their selectivity towards Cl-. This suggests that interference of bile acids at concentrations normally present in hepatocytes with determinations of aCli can account for the apparent difference aCli-aCleq.

Cl(-)-selective microelectrodes: sensitivity to anionic Cl- transport inhibitors.

Cl(-)-selective microelectrodes, containing Corning code 477315 or 477913 liquid ion exchangers, are often used to measure extra- and intracellular Cl- activities in the presence of Cl- transport inhibitors such as furosemide, bumetanide, and the stilbene sulfonic acid derivative 4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid (SITS). Because these inhibitors are anions in the physiological pH range and have relatively high lipid solubilities, they would be expected to interfere with the response to Cl- of these microelectrodes. Preliminary reports have confirmed this expectation. We examined the effect of furosemide, bumetanide, and SITS on the Cl(-)-selective barrels of double-barreled microelectrodes containing Corning code 477315 liquid anion exchanger and suitable for impaling small cells (e.g., epithelial cells). The results showed that at pH 8.2 in pure solutions of furosemide and bumetanide, these microelectrodes gave linear responses to the logarithm of furosemide or bumetanide concentrations ranging from 1 X 10(-2) to 1 X 10(-4) M. In the physiological pH range both these inhibitors (in concentrations of 0.1 mM) interfered significantly with the response of the microelectrodes to Cl- (in concentrations ranging from 100 to 1 mM). Calculated electrode sensitivities, relative to Cl-, were approximately 150 for both these compounds. Microelectrodes of this type appeared to be approximately 1,000 times as sensitive toward SITS as they were toward Cl-.

Electrophysiology of L-lysine entry across the brush-border membrane of Necturus intestine.

Microelectrode measurements of apical membrane potentials (Va) in absorptive cells of isolated Necturus intestine showed that, in the presence or absence of external Na+, 10 mM lysine added to the mucosal medium caused rapid depolarization followed by slower repolarization of Va. In Na+-free media the effects of 10 mM lysine on Va were abolished by 10 mM leucine which alone had no effect on Va under these conditions. This indicates that uncoupled electrodiffusion of lysine plays little or no role in lysine entry across the brush-border membrane. When external Na+ was greater than 10 mM the maximum depolarization of Va (delta Va') induced by [Lys] ranging from 5 to 30 mM was a simple saturable function of [Lys]. In Na+-free media, the relationship between delta Va' and [Lys] was biphasic. At first, delta Va' increased with increasing [Lys] reaching a maximum at 10 mM lysine. When [Lys] was further increased, delta Va' declined progressively to reach zero or near zero values. A single transport pathway model is proposed to account for rheogenic lysine entry across the brush-border membrane in the presence and absence of Na+. This postulates an amino acid transporter in the membrane with two binding sites. One is an amino acid site specific for the alpha-amino-alpha-carboxyl group. The other is a Na+ site. Neutral amino acids (e.g. leucine) compete with lysine for the amino acid site. The Na+ site has some affinity for the epsilon-amino group of lysine. When external Na+ is high the Na+ site is essentially 'saturated' with Na+ and formation of a mobile complex between an amino acid and the transporter depends in a saturable fashion on amino acid concentration. In Na+-free media or in media containing low [Na+]; at low external [Lys] the epsilon-amino group of a lysine molecule (simultaneously attached to the amino acid site) interacts with the Na+ site to form a mobile complex, as external [Lys] is increased, attachment of different lysine molecules to each site of an increasing number of transporters to form nontransported or poorly transported complexes results in substrate inhibition of the rheogenic lysine transport process.

Cyclic AMP-induced changes in membrane conductance of Necturus gallbladder epithelial cells.

Enhanced cellular cAMP levels have been shown to increase apical membrane Cl- and HCO3- conductances in epithelia. We found that the phosphodiesterase inhibitor 3-isobutyl-1-methyl-xanthine (IBMX) increases cAMP levels in Necturus gallbladder. We used conventional open-tip and double-barreled Cl- -selective microelectrodes to study the effects of IBMX on membrane conductances and intracellular Cl- activities in gallbladders mounted in a divided chamber and bathed with Ringer's solutions at 23 degrees C and pH 7.4. In HCO3- -free media, 0.1 mM IBMX added to the mucosal medium depolarized the apical membrane potential Va, decreased the fractional resistance FR, and significantly reduced intracellular Cl- activity (aCli). Under control conditions, aCli was above the value corresponding to passive distribution across the apical cell membrane. In media containing 25 mM HCO3-, IBMX caused a small transient hyperpolarization of Va followed by a depolarization not significantly different from that observed in HCO3- -free Ringer's. Removal of mucosal Cl-, Na+ or Ca2+ did not affect the IBMX-induced depolarization in Va. The basolateral membrane of Necturus gallbladder is highly K+ permeable. Increasing serosal K+ from 2.5 to 80 mM, depolarized Va. Mucosal IBMX significantly reduced this depolarization. Addition of 10 mM Ba2+, a K+ channel blocker, to the serosal medium depolarized Va and, essentially, blocked the depolarization induced by IBMX. These results indicate that mucosal IBMX increases apical HCO3- conductance and decreases basolateral K+ conductance in gallbladder epithelial cells via a cAMP-dependent mechanism. The latter effect, not previously reported in epithelial tissues, appears to be the major determinant of the IBMX-induced depolarization of Va.

Electronic device for microelectrode recordings in epithelial cells.

A device is described that permits continuous measurement of electrophysiological parameters in epithelial tissues in the open-circuit mode. Transepithelial potential (VT) and microelectrode (either conventional or ion-selective) potential (VM) are directly measured. Application of transepithelial current pulses allows continuous monitoring of transepithelial resistance (RT) and the ratio between the changes in VM and VT induced by these pulses. Measurement of this ratio, which under some circumstances reflects the apical fractional resistance of the cellular pathway, is important in assessing membrane damage during microelectrode impalement and/or as an index that the microelectrode tip is inside a cell. This is particularly useful when the change in VM during impalement is small. Application of 0.5-nA current pulses through open-tip microelectrodes allows continuous recording of the microelectrode resistance (RM). In epithelia where the individual cells are electrically coupled this permits acceptable impalements (RM remains nearly constant) to be distinguished from those affected by tip potential artifacts due to plugging of the microelectrode tip (RM increases after penetration of the cell membrane). The device provides compensation for the IR voltage drop in the solution between the potential measuring salt bridges and the epithelial surfaces. The microelectrode electrometer has an input impedance greater than 10(15) and is provided with stray capacitance neutralization.

KCl cotransport: a mechanism for basolateral chloride exit in Necturus gallbladder.

K+ and Cl--selective double-barreled microelectrodes were used to study the effect of changes in external K+ concentration on intracellular Cl- activity (aiCl) in epithelial cells of Necturus gallbladder. Decreasing the K+ concentration simultaneously in both bathing solutions produced a decrease in aiCl. Steady-state values of aiCl were related to the values of the chemical potential gradient for K+ (delta microK) across either the apical or the basolateral cell membrane. A similar dependence between aiCl and delta microK appeared when the K+ concentration was changed in the serosal solution only. This indicates that aiCl depends on delta microK across the basolateral membrane. aiCl was virtually independent of the membrane potential. This supports the idea that both the mucosal and the basolateral membranes of Necturus gallbladder cells have very low passive permeabilities to Cl-. These results indicate that the exit of Cl- from Necturus gallbladder cells is driven by delta microK across the basolateral membrane, and suggest that KCl electroneutral coupled mechanism in this membrane plays an important role in transcellular Cl- transport.

Intracellular chloride activity and apical membrane chloride conductance in Necturus gallbladder.

Open-tip and Cl--selective microelectrodes were used to study the effect of external pH on apical membrane potential (Va) and intracellular chloride activity (aiCl) in epithelial cells of Necturus gallbladder. Increasing the pH from 7.2 to 8.2 in the mucosal, the serosal, or in both bathing solutions simultaneously, hyperpolarized Va (control value -60 +/- 5 mV) by about -6, -10 and -17 mV, respectively, but did not significantly change the transepithelial potential (VT = 0.3 +/- 0.5 mV). Identical hyperpolarizations were recorded with Cl--selective microelectrodes, even 40 min after changing external pH. Thus, aiCl (12 +/- 2 mM) remained essentially constant. The ratio fVa between the deflections in Va and VT produced by transepithelial current pulses, which is an approximate measure of the fractional resistance of the apical membrane, decreased when mucosal pH was increased, and increased when serosal pH was raised. The changes in Va and fVa are due, in part at least, to the known pH dependence of cell membrane K+ conductance (PK) in this tissue. The constancy of aiCl, despite significant increases in Va, indicates that cell membrane Cl- conductance (PCl) is virtually zero or decreases, with increased external pH, in a way that compensates for the increased driving force for Cl- exit. Experiments in which 90 mM gluconate or 90 mM methylsulfate were substituted for an equivalent amount of luminal Cl- did not provide any evidence for a significant contribution of Cl- ions, per se, to the emf or conductance of the apical membrane. They suggested, rather, a dependence of apical membrane cation permeability on luminal Cl- concentration. Since basolateral membrane PCl is known to be very low, the insensitivity of aiCl to Va is the consequence of a negligible electrodiffusive Cl- permeability at both cell membranes. Thus, overall, transcellular Cl- transport in Necturus gallbladder is, in large measure, effected by electroneutral processes.

Intracellular sodium activity and transcellular sodium transport in gallbladder.

As stated in the introduction, the purpose of this report has been to illustrate how the measurement of steady-state intracellular ionic activities, a technique that has proved to be of great importance in studying the energetics of transmembrane ionic transfer processes, can, with appropriate assumptions, be used to obtain information concerning the kinetics of these processes. Specifically, our analysis has focussed on transcellular Na+ transport in Necturus gallbladder and has shown that, given the steady-state values of Em, a1Na, and apical Na+ conductance for a particular set of conditions, it is possible to obtain estimates of net baso-lateral Na+ efflux (Na+ pump rate), net (and/or unidirectional) diffusive apical Na+ influx, and net coupled NaCl influx. It should be emphasized that the analysis outlined above is a preliminary essay in this direction. We present it here in the hope that, wit appropriate refinements, it may prove useful in unraveling the mechanisms by which drugs, hormones and other specific agents affect membrane function in epithelial and other systems.