Transcellular sodium transport in cultured cystic fibrosis human nasal epithelium NIELS J. WILLUMSEN AND RICHARD C. BOUCHER Division of Pulmonary Diseases,Department of Medicine, School of Medicine, University of North Carolina, Chapel Hill, North Carolina 27514

WILLUMSEN, NIELS J., AND RICHARD C. BOUCHER. Transcellular sodium transport in cultured cystic fibrosis human nasal epithelium. Am. J. Physiol. 261 (Cell Physiol. 30): C332-C341, 1991.-Cystic fibrosis (CF) airway epithelia exhibit raised transepithelial Na’ transport rates, as determined by opencircuit isotope fluxes and estimates of the amiloride-sensitive equivalent short-circuit current (I,,). To study the contribution of apical and basolateral membrane paths to raised Na’ transport in CF, CF nasal epithelial cultures were studied with double-barreled Na+-selective microelectrodes and the Ussing chamber technique. Intracellular Na’ activity (aya) was 24.1 t 1.5 mM (n = 36), a value similar to a:’ of normal nasal epithelial cells. Reduction of luminal [Na’] to 3 mM abolished leq and reduced a:‘. Amiloride (lo-” M) abolished Ieq but increased aFa from 20 * 2 to 36 k 7 mM (n = 10). Amiloride-induced increase in ara was not affected by serosal [Na’] reduction but was blocked by preexposure to reduced luminal [Na’]. Amphotericin B increased Ieq during amiloride exposure, indicating that amiloride did not inhibit Na+-K’-ATPase. Ouabain abolished Icy and slowly raised apa. Reduction of serosal [Na’] led to a decrease in aya that was blocked by bumetanide. It is concluded that 1) CF airway epithelia exhibit an increased apical membrane Na’ permeability, 2) aFa is regulated to a normal level in CF cells despite increased transcellular Na’ fluxes, 3) the abnormal increase in a,Nain response to amiloride is dependent on luminal Na+, 4) Na’ is transported across the basolateral membrane by a bumetanide-sensitive cotransport mechanism, and 5) ouabain inhibits the basolateral Na+-K’ATPase, causing slow dissipation of the chemical and electrical gradients across the cell membranes.

api .cal membrane; basolatera .l memhuman airway epithelium; brane; amiloride; bumetan .ide; amphotericin B; ouabain

FIBROSIS (CF) is associated with ion transport abnormalities at a variety of anatomic sites including airways (la), pancreas (13), sweat glands (14), digestive tract (1, 9), and lymphocytes (49). Like other affected epithelia, CF nasal epithelium is characterized by the abnormal regulation of the Cl conductance in the apical cell membrane (8, 12, 14, 16, 17, 21). However, epithelia from airway regions are also characterized by an increased basal rate of Na+ absorption. Increased Na+ absorption has been demonstrated in intact tissues by electrophysiological and isotopic techniques (3, 6). In epithelial cultures, the amiloride-sensitive equivalent short-circuit current (I,,) of CF cultures is two to three times larger than in normal preparations (2, 21, 22). Recently, evidence has been presented that the increased CYSTIC



$1.50 Copyright

amiloride-sensitive Ieq in CF is associated with an increased absolute conductive apical Na’ permeability (2, 18) Ih an accompanying paper (20), we have characterized the intracellular Na+ activity and Na’ transport pathways in the normal human nasal epithelium. The increased rate of Na+ absorption in CF nasal epithelia should be associated with altered properties of specific cellular Na+ transport pathways in each membrane. Furthermore, based on relationships reported in other Na+absorbing epithelia, e.g., frog skin (10, ll), it is possible that the magnitude of intracellular Na’ activity (a:“) might yield information whether the increased Na+ transport rate is primarily a consequence of an increased apical Na+ permeability, in which case aya is expected to be higher than normal, or an increased activity of the basolateral Na+-K+-adenosinetriphosphatase (ATPase), in which case aFa is expected to be lower than normal. Consequently, in the present studies, we.determined Na + transport rates and electrochemical dn .ving forces and calculated the apical membrane Na+ permeability to compare these parameters in CF and normal preparations (20). MATERIALS



Tissue specimens.Nasal specimens were obtained from 13 cystic fibrosis subjects (mean age 11.4 t 1.2 yr, range 7-19 yr, 9 males, 4 females) and were typically polyps removed for nasal obstruction. All procedures were approved by the University of North Carolina Committee for the Protection of the Rights of Human Subjects. Tissue cultures. The cell culture procedures have been described in detail elsewhere (23). In brief, cells were isolated from freshly excised specimens by protease (Sigma IV), concentrated, and plated on collagen membranes in the bottom of plastic tissue culture cups. The cells were fed for 5 days with serum-free F-12 medium containing the following additives (F-12/7X): insulin, epidermal growth factor, cholera toxin, transferrin, hydrocortisone, triiodothyronine, and endothelial cell growth substance. Thereafter, they were fed with F-12/ 7X media supplemented (1:1) with 3T3 fibroblast-conditioned media containing 1% fetal bovine serum. After the 5th day in culture, the transepithelial potential difference (PD) developed by the culture wa.s measured daily to detect the occurrence of confluency. Cell preparations were routinely studied within 1 day of the devel-

0 1991 the American



Downloaded from by ${individualUser.givenNames} ${individualUser.surname} ( on January 13, 2019.



opment of the maximal PD. Solutions and drugs. The composition of the Krebs bicarbonate Ringer solution (KBR) was (in mM) 140 Na+, 120 Cl-, 5.2 K+, 25 HCO;, 2.4 HPOZ-, 0.4 H,PO,, 1.1 Ca”, and 5 glucose. Two different Na+-free Ringer solutions were employed: choline Cl- Ringer solution and N-methyl-Dglucamine (NMDG) Ringer solution. These Ringer solutions were identical to KBR except that all but 3 mM Na’ was substituted by choline and NMDG, respectively. Amiloride ( 10m4 M, a gift from Merck Sharpe & Dohme, West Point, PA), bumetanide (10v4 M, a gift from Leo Pharmaceuticals, Copenhagen, Denmark), ouabain (lo-” M, Sigma, St. Louis, MO) and amphotericin B (10v5 M, Sigma) were added to the KBR from concentrated stock solutions. All solutions were warmed to 37°C and gassedwith a 95% OZ-5% CO, gas mixture to maintain pH = 7.4. Microelectrodes and electrophysiological measurements. The microelectrodes were fabricated as previously reported (22). In brief, the tip of the Na’-selective barrel was backfilled with a droplet of neutral Na+ carrier (Na’ cocktail, Fluka, cat. No. 71176, Ronkonkoma, NY) and subsequently filled with 3 M KCl. The reference barrel was filled with either 0.5 or 3 M KCl. The electrodes responded linearly to logarithmic changes in Na+ activity throughout the range 5-100 mM with a sensitivity (S) or slope of 52.8 k 0.5 mV (n = 94) per decalog increase in Na’ activity. No corrections for interfering cations (K’ and Ca’+) were performed because calibration curves and quantitative calculations indicated little effect on the a,Nadeterminations (20). The response time of the Na+-selective electrode was generally in the order of 5 s. The transepithelial and intracellular electrophysiological techniques employed were as described in previous papers (21, 22). In brief, the tissue culture cups were mounted in a modified Ussing chamber for measurements of the transepithelial PD ( VJ and resistance (R,). In the case of symmetrical Ringer solutions, Ieq was determined as Ies = V/R,. In the caseof Na+-free solution on one side of the preparation, V/R, yields the equivalent transepithelial current at Vt = 0 mV (It), which is the sum of the currents through the cellular and the paracellular (or shunt) path. All microelectrode impalements were performed perpendicular to the outer (luminal) surface of the cell preparation. The fractional apical membrane resistance (fRa) was calculated from fR a = Ra/(Ra + Rb) = AVa/AVt


where Ra and RI, are the apical and basolateral membrane resistances, respectively, and AVa and A& are the deflections in the apical membrane potential (Va) and in Vt induced by the current pulses. From Va and the signal from the Na+-selective electrode barrel ( VNa), a,Nawas calculated from the equation aNa= aNa o . l(pha - vall~ C

where ara is the Na’ concentration in KBR, AVNa is the change in VNa upon an impalement, and S is the sensitivity defined above. The criteria for acceptance of im-



palements were as reported in previous papers (21, 22). Transepithelial fluxes. For measurements of transepithelial unidirectional isotopic Na’ fluxes, CF cultures matched on the basis of Vt and R, (~25% difference) were mounted in Ussing chambers bathed by 6 ml KBR that was gassedwith a 95% Oa-95% CO, gas mixture and warmed to 37°C. “2Na+ (10 &i) was added to either the luminal or serosal bath, and solution samples were taken from the sink (1 ml) and source solutions (0.05 ml) every 10 min for 40 min before and after amiloride (10s4 M) addition to the luminal solution. The samples were counted in a gamma spectrometer (LKB, Bromma, Sweden), and data were analyzed by a standard computer program. Sign conventions and statistics. The apical membrane potential is referenced to the luminal (grounded) bath potential: Va = tic - $1. The transepithelial potential and basolateral membrane potential are both referenced to the serosal bath: Vt = $1 - &. and Vb = qC - tis. $Q,&, and tic represent the electrical potentials of the luminal and the serosal baths and the intracellular compartment, respectively. Ieq is negative when flowing from lumen to serosa (inward cation flux or outward anion flux). The changes in parameters in response to experimental maneuvers were analyzed by a two-tailed, paired Student’s t test. Significance levels are given in the tables. All values are given as means t SE. RESULTS

Baseline a: and electrophysiological parameters. An illustrative sequence of impalements of a CF human nasal epithelial culture with a relatively high transepithelial potential difference is shown in Fig. 1. During the 11 min traced, Vt spontaneously varied between -54 and -59 mV. Among the five depicted impalements, Va varied between 18 and 22 mV, fRa varied between 0.31 and 0.37, and a,Navaried between 20 and 23 mM. Table 1 summarizes mean electrophysiological parameters from 32 CF preparations. The mean apa was 24 t 2 mM and the distribution of the individual a,Navalues is depicted in Fig. 2. The electrochemical driving force for Na’ across the apical membrane was 51 mV, directed from lumen to cell. In contrast to normal preparations (19), Va was not correlated with Vb (r = 0.15, Fig. 3A), whereas it was linearly correlated with Vt (r = 0.86, Fig. 3B) with a slope of -0.95. Like normal preparations (20), ley was not correlated with ara (r = 0.15, Fig. 3C) whereas, in contrast to normals, Ieq was correlated with fRa (r = 0.68, Fig. 30). Effect of reduction of luminal [Na+]. We studied the effect of removal of luminal Na’ on the apa and electrophysiological properties in eight preparations. Figure 4 depicts the changes in Vt, Vay and VNainduced by reduction of luminal [Na’] to 3 mM in an illustrative experiment. Vt decreased and reversed (from -21 to +20 mV) and was accompanied by a 44-mV hyperpolarization of Va from -38 to -82 mV. VNa hyperpolarized by 50 mV, from -86 to -136 mV. The greater hyperpolarization of VNathan Va reflects a decrease in aFa from 14 to 11 mM within -3 min. The decrease in aya was associated with

Downloaded from by ${individualUser.givenNames} ${individualUser.surname} ( on January 13, 2019.








40 to 830 t 140 Q cm’), fRa increased (from 0.59 t 0.05 to 0.79 t 0.02), and Ieq was virtually abolished (from -93 t 12 to -3 t 1 pA/cm2). Routinely, aya increased (from 20.2 t 1.9 to 35.8 t 6.5 mM) in response to luminal amiloride exposure. The increase in aFa was significant at all times between 5 and 20 min after amiloride administration. Interactions between reduction of luminal [Na+] and amiloride. The discrepancy in the behavior of CF airway epithelial aFa in response to amiloride compared with luminal Na+-free solutions led us to study interactions between these two maneuvers that affect Na+ entry into the cell across the luminal membrane. We exposed five V,, hV) preparations to the sequence of amiloride followed by amiloride-Na+-free Ringer solution in the luminal bathing solution. Within 10 min after luminal exposure to amiloride-Na+-free Ringer solution, a 62% reduction in a,Nacompared with amiloride-only exposure was noted 0 I 2 3 4 5 6 7 8 9 IO II (from 26 t 9 to 10 t 3 mM, Fig. 8). The shift from TIME (mid amiloride-only superfusion to amiloride-Na+-free superFIG. 1. Recordings from a series of 5 subsequent impalements of a fusion induced no effect on the fRa and induced small cystic fibrosis (CF) human nasal epithelial culture with a doublebut significant changes in Va (7 t 2 mV hyperpolarizabarreled Na+-selective microelectrode. Top trace, transepithelial election; P < 0.05, n = 5) and Vb (4 t 1 mV depolarization; trical potential; middle trace, signal from reference electrode, which during a successful impalement is a measure of apical membrane P C 0.05, n = 5). potential ( V,); and bottom trace, signal from Na’-selective electrode, Amiloride exposure during luminal perfusion with 3 which during an impalement represents driving force for Na’ flow mM Na+ solutions led to no increase in ara but did lead across apical cell membrane. Vertical deflections are responses to to rapid increases in fRa and R, (from 0.60 to 0.76 and transepithelially pulsed currents. Due to relatively long response time from 615 to 771 0. cm’, respectively, Fig. 5). These effects of Na+-selective electrode, deflections on VN, trace have smaller amplitudes than deflections on V, trace. were associated with reversal of Vt (from -4 t 6 to 6 t 0 mV) and a hyperpolarization of the apical membrane changes in R, and fR, as expected for a conductive apical (from -40 t 3 to -54 t 4 mV). Na’ path: an increase in R, and fR,. Effect of amphotericin B. To examine if amiloride Figure 5 collects the changes in electrophysiological raised a,Naby any inhibitory effect on the basolateral properties and a,Nafrom eight experiments for a 15min Na’-K’-ATPase, we attempted to increase apa by noninterval after luminal Na’ reduction. In the lo- to 15 selective permeabilization of the apical membrane with min interval, czFawas reduced from 23 t 5 to 8 t 2 mM amphotericin B and test whether an increase in Ieq was and the apical membrane had hyperpolarized from -2 t induced. The effect of amphotericin B (lo-” M) on amil5 to -40 t 3 mV. Vt depolarized from -45 t 10 to -4 t oride-pretreated preparations was similar to the one ob6 mV. served in normal nasal epithelia (20). Figure 9 depicts a Effect of amiloride. We studied the effect of inhibition typical experiment. Amphotericin B led to a rapid inof the apical Na’ conductance with amiloride (10s4 M) crease in Vt (from -9 to -46 mV), a depolarization of on bioelectric parameters and ara in 10 CF preparations. the apical membrane (from -48 to -16 mV), and a A typical amiloride response is illustrated in Fig. 6. Vt decrease in R, (from 277 to 135 Qcm’) and fR, (from decreased from -23 to -1 mV, Va hyperpolarized from 0.79 to 0.25). Amphotericin B administration was asso-10 to -58 mV, and VNa hyperpolarized from -56 to ciated with a rapid increase in aFa from 16 to 45 mM, -94 mV. The fact that VNahad changed 7 mV less than which induced an increase in Ies from -33 to -344 PA/ Va at 1 min indicates that the a,Naincreased in response cm’. to amiloride. During the first minute after amiloride, Ieq Transepithelial Na+ fluxes. To determine the effect of almost vanished. amiloride on transepithelial Na+ transport, isotopic Figure 7 shows the effect of amiloride on the ara and unidirectional Na’ fluxes were measured under opencircuit conditions (Table 3). Under basal conditions, a bioelectric parameters for a period of 20 min after amilnet inward Na’ flux of 1.82 peg. cmS2.h-l was detected. oride exposure. The effect of amiloride on electrophysiDuring amiloride exposure, the net absorptive flux of ological parameters of CF cultures was similar to that Na’ was abolished. reported in previous papers (2, 6, 21): Va hyperpolarized Effect of removal of serosal Na+ before and after amil(from -8 t 5 to -50 t 3 mV), R, increased (from 450 t V+ hV)













1. Bioelectric properties and intracellular Na+ activity (a?) of cultured cystic fibrosis human nasal epithelia K, mV -33k3



are means





t SE; n = 32 preparations.

R,, &cm2 413t35 Abbreviations


Transcellular sodium transport in cultured cystic fibrosis human nasal epithelium.

Cystic fibrosis (CF) airway epithelia exhibit raised transepithelial Na+ transport rates, as determined by open-circuit isotope fluxes and estimates o...
2MB Sizes 0 Downloads 0 Views