Pflügers Arch – Eur J Physiol (2000) 439:504–512 Digital Object Identifier (DOI) 10.1007/s004249900194

O R I G I N A L A RT I C L E

Danny Jans · Jeannine Simaels · Dana Cucu Wolfgang Zeiske · Willy Van Driessche

Effects of extracellular Mg2+ on transepithelial capacitance and Na+ transport in A6 cells under different osmotic conditions Received: 18 June 1999 / Received after revision: 4 October 1999 / Accepted: 18 October 1999 / Published online: 14 January 2000 © Springer-Verlag 2000

Abstract The electrophysiological characteristics of monolayers of cultured renal epithelial A6 cells were studied under short-circuit conditions. Replacing basolateral isosmotic (260 mOsm/kg H2O) media by hyposmotic (140 mOsm/kg H2O) solutions transiently increased the transepithelial capacitance (CT) by 57.3±2.3% after 16 min. The transepithelial Na+ current (INa) increased concomitantly from 4.2±0.7 to 26.1± 2.6 µA/cm2 with a time course that was noticeably slower, reaching its maximum after 60 min of hypotonicity. The transepithelial conductance (GT) increased synchronously with INa. Analysis of blocker-induced noise in INa, using the amiloride analogue 6-chloro-3,5-diaminopyrazine-2-carboxamide (CDPC), showed that the hypotonic shock increased Na+ channel density (NT) at the apical border. The presence of 10 mM Mg2+ on both sides of the epithelium suppressed the hypotonicity-induced CT increase to 14.3±0.5%, whereas the INa increase was even larger than without Mg2+. Both effects of Mg2+ were located at an extracellular, basolateral site, because apical administration was without effect, whereas the acute basolateral addition of Mg2+ at the moment of the hypotonic shock was sufficient. Interaction between Mg2+ and Ca2+ influenced the behaviour of CT. At constant osmolality (200 mOsm/kg H2O) 10 mM Mg2+ increased INa, leaving CT unaffected, whereas 10 mM Ca2+ stimulated both INa and CT. In the presence of 1 mM Mg2+, however, the Ca2+-induced CT increase was abolished. The failure of CT to increase during stimulation of INa by Mg2+ suggests that the divalent cation activates pre-existing channels in the apical membrane. Noise analysis showed that the natriferic effects of Mg2+ were also mediated by an increase in NT. The moderate D. Jans · J. Simaels · W. Zeiske · W. Van Driessche (✉) Laboratory of Physiology, K. U. Leuven, Campus Gasthuisberg O/N, B-3000 Leuven, Belgium e-mail: [email protected] Tel.: +32-16-345731, Fax: +32-16-345991 D. Cucu University of Bucharest, Faculty of Biology, Biophysical Lab, Bucharest 76208, Rumania

initial increase in CT in the presence of Mg2+ under hypotonic conditions, occurring in parallel with increases in GT and INa, reflects most likely Na+ channel insertion induced by the hypotonic treatment. However, the large, transient, Mg2+-sensitive increase in CT, not correlated with increases in GT and INa, seems to be unrelated to Na+ channel recruitment. Key words Mg2+ · Renal epithelia · Exocytosis · ENaC · Hypotonicity · Noise analysis

Introduction When cultured on permeable filter supports, A6 cells develop into polarized epithelia with a high transepithelial resistance. These monolayers show transport characteristics similar to mammalian cortical collecting duct cells [20]. Na+ is transported via uptake at the apical border through amiloride-blockable Na+ channels (ENaC-type) and extrusion at the basolateral side by the Na+-K+-adenosine triphosphatase (Na+-K+-ATPase). Transepithelial Na+ transport is regulated mainly at the apical border. The apical membrane sodium permeability can be raised by increasing the total number of active channels (NT), by an increase of open probability for the channel (Po) or by the vesicular insertion of new channels that would augment NT and the apical membrane surface area. The latter is reflected in an increase of the transepithelial capacitance (CT) that may reveal exocytotic processes [10]. In addition, we analysed the transepithelial conductance (GT) and the noise in the transepithelial Na+ current (INa) induced by the Na+ channel blocker CDPC to assess possible changes in Po and NT. Mg2+ is the most abundant divalent cation in the intracellular fluid. Intracellular Mg2+ is essential for ATPase activity and therefore a prerequisite for maintaining several forms of active ion transport. Extracellular Mg2+ blocks Ca2+ influx in the giant squid axon [2] and Mg2+ ions also block Ca2+ influx through the ligand-gated Nmethyl-D-aspartate (NMDA) channel by entering the

505

pore from either the extracellular or the cytoplasmic side of the membrane in a voltage-dependent manner [17]. In human tracheal secretory gland cells, extracellular Mg2+ inhibits exocytosis [22] by decreasing the rise in intracellular [Ca2+] ([Ca2+]i), which is a trigger for this process. Most interestingly, in the isolated urinary bladder of the toad Bufo marinus, another tight epithelial structure, a clear stimulation of Na+ transport has been reported after apical addition of 3 mM Mg2+ [1]. Several studies in natural and model tight epithelia have suggested that the activation of Na+ transport occurs through the insertion of Na+ channels in the apical membrane [8, 14, 18, 25, 26]. The activation of INa through channel insertion most likely results in an increase in membrane area, and thus in CT, accompanied by a parallel increase in GT. Such a relationship between INa and CT has been demonstrated during hormonal activation of Na+ transport [13, 27]. In the present study we explored further the relationship between CT and INa in A6 cells by the application of extracellular Mg2+ under different osmotic conditions. In the absence of extracellular Mg2+, the activation of Na+ transport elicited by decreasing basolateral solution osmolality evoked a biphasic increase in CT and INa. Initially, the time courses proceeded synchronously, whereas during the subsequent period an opposite behaviour was observed. The presence of extracellular Mg2+ at the basolateral, but not apical, side at the moment of the hypotonic shock significantly diminished the second transient part of the CT increase but further activated the hypotonic stimulation of INa. In addition, interaction between extracellular Mg2+ and Ca2+ was observed. At constant osmolality, 10 mM Ca2+ evoked an increase in CT that could be abolished by the presence of 1 mM Mg2+, while both divalent cations activated INa additively. It appears that the activation of INa by Mg2+ occurs through the opening of dormant Na+ channels in the apical membrane of A6 cells, while osmotic changes insert other vesicular material into the apical membrane, apart from new Na+ conducting channels.

Materials and methods Cell culture All experiments were carried out on monolayers of A6 cells cultured on permeable filter supports (pore size 0.2 µm) (Anopore, Nunc Intermed, Roskilde, Denmark). A6 cells used for these studies were obtained from Dr. J.P. Johnson (University of Pittsburgh, Pittsburgh, Pa., USA). The cells were seeded at a density of 105/cm2 and cultured at 28 °C in a humidified incubator in 1% CO2. The growth medium was a mixture (1:1) of Leibovitz’s L-15 and Ham’s F-12 media, supplemented with 10% fetal bovine serum (Sigma, St. Louis, Mo., USA), 3.8 mM L-glutamine, 87 IU penicillin and 87 µg/ml streptomycin. The growth medium was renewed twice weekly. For the experiments in this study we used cell passages 104–112 that had been cultured for 8–22 days. The epithelial monolayers were mounted with minimal edge damage in Ussing-type chambers for electrophysiological measurements. These chambers are suited for continuous perfusion of both compartments and rapid exchange of the solutions. All experiments

were carried out under short-circuit conditions by continuously clamping transepithelial voltage to zero with a low-noise or highspeed voltage-clamp for noise analysis or capacitance measurements, respectively. Solutions Hyposmotic solutions (140 mOsm/kg H2O) contained (in mM) 70 Na+, 2.5 K+, 2.5 HCO–3, 1 Ca2+ and 72 Cl– (pH 8.0). Isosmotic solutions (260 mOsm/kg H2O) were prepared by adding 65 mM NaCl. Solutions with an osmolality of 200 mOsm/kg H2O contained (in mM) 102 Na+, 2.5 K+, 2.5 HCO–3, 1 Ca2+ and 104 Cl– (pH 8.0). For the experiments with Mg2+, 0.5 and 2 mM MgCl2 was simply added to the solutions. For the experiments with 10 mM MgCl2 and 10 mM CaCl2, 15 mM NaCl was substituted. Hypotonic shock was initiated by decreasing the basolateral osmolality, in which case the apical perfusate was made hyposmotic at least 30 min in advance. We could use this procedure because the apical membrane of A6 epithelia is water impermeable, meaning that the cells do not swell as a consequence of this manoeuvre [12]. The benefits of using hyposmotic apical perfusates are, firstly, that an osmotic gradient across the tissue directed from the apical to basolateral side is avoided when lowering the osmolality of the basolateral bath. Such a gradient opens the paracellular pathway, thus increasing the shunt conductance. On the other hand, imposing the opposite osmotic gradient (basolateral to apical) during phases where the basolateral bath is isosmotic does not affect the shunt conductance. Second, with this procedure [Na+] in the apical bath can be kept constant when the osmolality is reduced, thus avoiding the current changes expected from bilateral reduction of [Na+]. Amiloride (10–4 M, Sigma) was used to determine the amiloride-sensitive component of the short-circuit current (Isc) that corresponds to INa. 6-Chloro-3,5-diaminopyrazine-2-carboxamide (CDPC) was obtained from Aldrich (Aldrich Chemical, Milwaukee, Wis., USA).

CT and GT measurements For these experiments we used a high-speed voltage clamp in which the phase shift between the voltage and current signal was minimal up to 100 kHz. A previous report [23] describes the equipment in detail and discusses extensively the theoretical background for the measurement of CT. Briefly, the hardware for CT measurements was based on two digital signal processing (DSP) boards (Model 310B, Dalanco Spry, Rochester, N.Y., USA) equipped with two high-speed (300 kHz) analogue-to-digital converters (14-bit) and two digital-to-analogue converters (12-bit). One DSP board was used to record GT and Isc. GT was measured by imposing a 1-Hz sine wave voltage of 5 mV to the tissue. Using the second DSP board, CT was measured with five sine waves that were subsequently imposed to the command voltage input of the clamp. In this study we used 2, 2.7, 4.1, 5.4 and 8.2 kHz. The figures in this paper show records at 4.1 kHz. Phase shifts and amplitude ratios between the voltage and current signals were calculated using regression analysis. With these data we calculated the parameters of the equivalent circuit of the epithelium represented by a simple RC network and consisting of a series resistance, CT and its equivalent parallel resistance. The graphical interface (Labview, National Instruments) enabled the real-time display of GT, Isc and CT. Model calculations based on a lumped, two-membrane model demonstrated that, in the high-frequency range, CT equals the equivalent capacitance of the series arrangement of the apical (Cap) and basolateral capacitance (Cbl) [23]: 1 = 1 + 1 CT Cap Cbl

(1)

As Cbl is about 12 times larger than Cap, changes in CT will reflect mainly alterations at the apical membrane, i.e. the result of endo-

506 and exocytotic processes at that border. In the tight epithelium that A6 cells represent, GT is determined largely by the apical membrane conductance that, under non-stimulated conditions, reflects mainly the permeability of the apical Na+ channels. Noise analysis A low-noise voltage-clamp was used to short-circuit the tissues in these experiments. Noise analysis was implemented using the weak Na+ channel blocker CDPC. We used a pulse protocol similar to that described by Blazer-Yost et al. [4]. The apical surface was exposed alternately to 10 or 40 µM CDPC for periods of 5 min. The mean percentage current reductions at 10 and 40 µM CDPC were 7.9±0.7 and 19.3±1.6% (n=6), respectively. Current noise at both blocker concentrations was amplified, digitized and Fourier-transformed to yield power density spectra. During each 5-min period we recorded noise spectra as the mean of 50 sweeps of 2 s duration, resulting in a fundamental frequency of 0.5 Hz. The amiloride-insensitive current (Iami) was measured by blocking the channels at the apical side with 50 µM amiloride. The blockersensitive, macroscopic current (IBNa)was calculated as: IBNa=Isc–Iami. The random interactions of CDPC with the Na+ channel induced a Lorentzian component in the power density spectra. The two Lorentzian parameters, the low-frequency plateau (So) and the corner frequency (fc), were determined by non-linear curve fitting of the spectra. Changes in INa influenced So, but did not affect fc. The latter varied linearly, however, with the blocker concentration. The on- (kon) and off- (koff) rates were calculated from: 2πfc=konB+koff

(2)

Fig. 1 Effects of lowering basolateral osmolality on transepithelial capacitance (CT), short-circuit current (Isc) and transepithelial conductance (GT) in cultured monolayers of renal epithelial A6 cells in the absence or presence of extracellular Mg2+. In this and all subsequent experiments, the monolayers were perfused with NaCl-Ringer’s solutions on the apical and basolateral sides. This figure compares the responses of CT, Isc and GT to a hypotonic shock in the absence (solid line) and presence (dashed line) of 10 mM MgCl2 on both sides of the monolayer, during the entire experiment. The apical perfusate was hyposmotic (140 mOsm/kg H2O). During the period indicated the basolateral osmolality was reduced from 260 to 140 mOsm/kg H2O. During hypotonic treatment, the basolateral [Na+] equalled the apical (70 mM). The traces are mean values from six tissues. For clarity, SEMs have been omitted

Na+

where B is the blocker concentration. Using the currents and the Lorentzian parameters (fc and So) obtained from noise analysis, we estimated single-channel current (iNa) and the density of open, amiloride-sensitive Na+ channels in the apical membrane (No). Single-channel currents in the presence of 10 µM CDPC (i10Na) were used as single-channel currents in the absence of blocker, since they do not differ significantly [4], according to: S10 (2π f 10 )2 iNa = i 10 = o 10 c Na 4 INa kon B

(3)

where S10o, f10c and I10Na are the values of So, fc and INa in the presence of 10 µM CDPC. No for functional channels at 10 µM CDPC (N10o) is given by: 10 / i10 No10 = INa Na

No in the absence of CDPC is calculated as follows: No = No10 (1 + Po B ) KB

(4)

(5)

where KB is the equilibrium coefficient for the effect of the blocker on open channels (≅koff/kon) in µmol/l. Po with respect to spontaneous channel fluctuations was calculated using KB from the fractional inhibition of the blocker-sensitive Na+ transport, I40/10Na, elicited by increasing the CDPC concentration from 10 to 40 µM: 40 / 10  1 – INa  (6) Po =  K 40 / 10  40 INa – 10  B The total number of channels (NT) is then calculated as: NT=No/Po

(7)

Statistical analysis Comparisons between groups of data were made using Student’s ttest for paired samples. P<0.05 was regarded as significant. Results are given as means±SEM with n being the number of tissues investigated.

Results Effect of Mg2+ during a hypotonic shock Tissues were initially exposed to isosmotic (260 mOsm/kg H2O) and hyposmotic (140 mOsm/kg H2O) solutions on the basolateral and apical side, respectively. As described in Materials and methods, these conditions evoke neither volume nor transport changes in A6 epithelia. Imposing hypotonic conditions by reducing the basolateral osmolality from 260 to 140 mOsm/kg H2O results in cell swelling and volume recovery [12]. Figure 1 shows the impact of this treatment on Isc, CT, and GT and demonstrates the effects of 10 mM Mg2+ added to both sides of the tissue. Immediately after exposure to hypotonicity, CT decreased quickly but transiently, reaching a minimum after 25 s and recovering within 2–3 min. This phenomenon, caused by closure and opening of the lateral interspaces (LIS) when the cells swell, has been described extensively previously [23]. In the absence of Mg2+, after recovery from this transient closure of the LIS, CT displayed a biphasic transient increase from 0.76±0.02 to a maximum of 1.20±0.02 µF/cm2 (n=6), 16 min after initiation of the perturbation. The time courses of the responses of Isc and GT to the hypotonic treatment were noticeably slower and proceeded synchronously. Both parameters increased biphasically and reached a first plateau at the time at which CT had attained its maximum. During the phase in which CT declined, GT and Isc increased steadily, reaching a plateau 60 min after initia-

507 Mg2+

Table 1 Effect of different extracellular concentrations of ([Mg2+]e) on the response of transepithelial capacitance (CT), short-circuit current (Isc) and transepithelial conductance (GT) during hypotonic treatment of monolayers of cultured renal epithelial A6 cells. Mg2+ was added to both sides of the tissue at least 30 min before imposing the hypotonic shock. Under isotonic (ISO) conditions, epithelia were exposed to 260 and 140 mOsm/kg H2O [Mg2+]e (mM)

0 0.5 2 10

CT (µF/cm2)

NaCl solutions on the basolateral and apical sides, respectively. Under hypotonic (HYPO) conditions the osmolality of the basolateral bath was lowered to 140 mOsm/kg H2O by NaCl removal. CT was recorded at the time of maximal increase; on average 16 min after beginning the osmotic perturbation. Isc was recorded at the end of the period when Na+ transport had reached a plateau value. Means±SEM for n tissues

Isc (µA/cm2)

GT (mS/cm2)

n

ISO

HYPO

ISO

HYPO

ISO

HYPO

0.76±0.02 0.73±0.01 0.77±0.01 0.74±0.01

1.20±0.02 0.79±0.01 0.92±0.01 0.85±0.01

4.2±0.7 1.4±0.1 1.7±0.1 5.4±0.6

26.1±2.7 21.5±2.8 26.3±1.6 30.7±1.6

0.25±0.06 0.16±0.04 0.12±0.01 0.30±0.08

0.39±0.05 0.27±0.04 0.33±0.02 0.50±0.03

6 6 6 6

tion of the hypotonic treatment. Isc increased from 4.2±0.7 in isosmotic conditions to 26.1±2.7 µA/cm2 at the end of the hypotonic period. The response of CT to hypotonicity was markedly altered in the presence of Mg2+. With 10 mM Mg2+ in both bathing solutions, CT increased modestly from 0.77±0.01 to 0.92±0.01 µF/cm2 (n=6), whereas Isc increased from 5.4±0.6 to 30.7± 1.6 µA/cm2. Importantly, during the hypotonic shock the initial increase of CT, GT and Isc occurred synchronously, both in the absence and in the presence of Mg2+. During the subsequent phase, the time courses of CT on one hand and GT and Isc on the other, developed asynchronously. The presence of Mg2+ abolished mainly the second increase of CT, while the plateau phase of GT and Isc disappeared due to additional activation of each. The impact of Mg2+ on the responses of CT and Isc to hypotonicity was also manifest when 2 mM Mg2+ was added to the solutions on both sides and even concentrations of Mg2+ as low as 0.5 mM decreased the CT increase to comparable levels. A summary of the results with different concentrations of extracellular Mg2+ is given in Table 1. These results show that relative small concentrations of Mg2+ in the extracellular fluid markedly impair the increase in CT and, by inference, membrane traffic evoked by hypotonicity. iNa, Po and NT following hypotonic shock Stimulation of transepithelial Na+ transport requires the elevation of Na+ uptake at the apical border. This can occur through an increase of the permeability of the apical membrane for Na+, by an increased driving force for apical Na+ uptake or both. Augmentation of the driving force should be reflected in an increase of iNa. Apical membrane permeability could increase by either the insertion of Na+ channels from an intracellular pool, by the activation of existing dormant channels in the membrane, or by increasing Po for already activated channels. The small increase of CT in the presence of Mg2+ under hypotonic conditions, described in the previous section, could be related to channel insertion by exocytosis. We applied noise analysis of the Na+ current to determine iNa, Po and NT. In this series of paired experiments, we compared the response of two sets of six tissues in con-

Fig. 2A,B Effects of lowering basolateral osmolality on the transepithelial Na+ current (INa), single-channel Na+ current (iNa), total channel density (NT) and channel open probability (Po) in the absence or presence of extracellular Mg2+. Tissues were incubated initially in isosmotic and hyposmotic NaCl solutions on the basolateral and apical side, respectively. After an equilibration period of at least 60 min the osmolality of the basolateral bath was lowered to 140 mOsm/kg H2O to induce a hypotonic shock. Noise analysis was performed by recording 6-chloro-3,5-diaminopyrazine-2-carboxamide (CDPC)-induced noise by exposing the tissue alternately to 10 and 40 µM CDPC for 5-min periods. A Pulse protocol of CDPC-induced noise. Amiloride (Ami, 50 µM) was applied apically at the end of the hypotonic shock to measure INa. B Results of noise analysis represented in bar diagrams. The data labelled iso were collected in a 30-min period preceding the hypotonic treatment of the cells; those labelled hypo were collected during the last 30 min of the hypotonic period. Data are compared from experiments in which no Mg2+ was present in the solutions (open bars) or where 10 mM Mg2+ was present bilaterally (hatched bars). *P<0.05 for comparisons between 0 and 10 mM Mg2+; #P<0.05 for comparisons between isosmotic and hyposmotic solutions, Student’s t-test for paired samples

508

trol solutions (0 mM Mg2+) with experiments in which 10 mM Mg2+ was present bilaterally. Figure 2A illustrates the protocol in which we exposed the apical surface alternately to 10 and 40 µM CDPC. Figure 2B gives an overview of the results obtained. Hypotonicity increased INa under both conditions, an effect that was more pronounced in the presence of 10 mM Mg2+. Furthermore, iNa decreased significantly in both control and Mg2+-treated tissues (P<0.05), indicating that Mg2+ and hypotonicity decreased the driving force for apical Na+ uptake. Po was activated significantly in the presence of Mg2+ (P<0.05) and reduced significantly by the hypotonic shock (P<0.05). The reduction of both iNa and Po by the hypotonic shock should result in a decrease of INa. The observed increase of INa must therefore be attributed to a marked increase in NT, an effect enhanced even further by Mg2+. It should also be noted that, in this series of paired experiments, INa recorded in the presence of Mg2+, under both isosmotic and hypotonic conditions, was significantly larger than in control. Contrary to hypotonicity, the increase in INa by Mg2+ is correlated with increases in both Po and NT. Under isotonic conditions, therefore, Mg2+ appears to activate dormant channels already present in the membrane and also to shift already activated channels towards an open state, whereas the hypotonic shock may involve exocytotic processes. Comparison of apical vs. basolateral addition of Mg2+ Experiments were performed to resolve the side dependence of the inhibitory effect of Mg2+ on the CT increase elicited by hypotonicity. Solutions containing 2 mM Mg2+ from the beginning of the experiment (chronic administration) were applied to either the apical or basolateral sides of the epithelium, while the opposite side remained Mg2+ free. Apical administration of the divalent cation did not noticeably affect the time course of the hypotonic-induced change of CT, which increased from 0.78±0.02 to 1.32±0.02 µF/cm2 (n=6), as shown in Fig. 3A. The plateau reached by Isc at the end of the hypotonic pulse was comparable to that in Fig. 1 in the absence of Mg2+ (Isc increased from 2.4±0.3 to 24.6±0.7 µA/cm2, n=6). On the other hand, the application of 2 mM Mg2+ to the basolateral side of the epithelium was clearly sufficient to evoke effects similar to the bilateral application. In this case CT increased modestly from 0.78±0.01 to 0.94±0.01 µF/cm2 (n=6), while Isc rose from 2.6±0.3 to 24.7±2.4 µA/cm2 (n=6). It is noteworthy that basolateral Mg2+ also abolished the biphasic pattern of the Isc increase as observed in Fig. 1. Mean steady-state Isc, GT and CT values are summarized in Table 2. Acute addition or removal of Mg2+ To this point, our data demonstrate that the effect of Mg2+ is located at the basolateral border. To establish whether Mg2+ exerts its effect extra- or intracellularly,

Fig. 3A,B Effect of Mg2+ on changes of CT, Isc and GT induced by a hypotonic shock. A Comparison of apical vs. basolateral presence of Mg2+. Mg2+(2 mM) was present in either apical (solid line) or basolateral (dashed line) bathing solution during the entire experiment (chronic administration). Experiments were performed as described in Fig. 1. Traces were obtained by averaging the responses from six sets of tissues. B Comparison of acute addition vs. acute removal of basolateral Mg2+ at the moment of the hypotonic shock. In one set of six tissues (dashed line), epithelia were incubated in Mg2+ free solutions and 2 mM Mg2+ was only present in the basolateral perfusate during the hypotonic treatment. In the other set (solid line), A6 monolayers were perfused at the basolateral side with a solution containing 2 mM Mg2+ until the hypotonic shock was applied. At that moment, Mg2+ was removed and added again when basolateral osmolality was restored. During the entire experiment the apical perfusate was hyposmotic and Mg2+ free

we added or removed Mg2+ acutely from the basolateral bath when the osmolality was reduced. The rationale for this type of experiment is that, if Mg2+ acts intracellularly, acute addition might fail to exert its effect, whereas during acute removal the effects should be preserved. Opposite effects would be expected for an extracellular site of action for Mg2+. Figure 3B clearly demonstrates that the effects of acute addition on CT and Isc were comparable to those recorded during chronic administration: the hypotonicity-induced increase in CT was largely depressed by the acute addition of Mg2+, whereas Isc was augmented to comparable levels. On the other hand, in experiments where Mg2+ was acutely removed we obtained a response in CT that was not noticeably different from the experiments in Fig. 1 in the absence of Mg2+. Mean values are displayed in Table 2. Moreover, the biphasic time course of Isc was preserved during acute removal and abolished during acute addition.

509 Mg2+

Table 2 Effect of 2 mM on the response of CT, Isc and GT during hypotonic treatment of A6 epithelia. Under ISO conditions, epithelia were exposed to 260 and 140 mOsm/kg H2O NaCl-Ringer’s solutions at the basolateral and apical sides, respectively. During HYPO conditions the osmolality of the basolateral bath was Type

CTRL Chronic Chronic Chronic Acute addition Acute removal

Side

Bilateral Basolateral Apical Basolateral Basolateral

CT (µF/cm2)

lowered to 140 mOsm/kg H2O by NaCl removal. CT was recorded at the time of maximal increase; on average 16 min after beginning the osmotic perturbation. Isc and GT were recorded at the end of the period when Na+ transport had reached a plateau value. Means±SEM for n tissues

Isc (µA/cm2)

GT (mS/cm2)

n

ISO

HYPO

ISO

HYPO

ISO

HYPO

0.76±0.02 0.77±0.01 0.78±0.01 0.78±0.02 0.78±0.01 0.77±0.03

1.20±0.02 0.92±0.01 0.94±0.01 1.32±0.02 0.91±0.02 1.21±0.04

4.2±0.7 1.7±0.1 2.6±0.3 2.4±0.3 2.5±0.1 2.5±0.7

26.1±2.7 27.9±0.9 24.7±2.4 24.6±0.7 26.7±0.4 19.6±2.1

0.25±0.06 0.12±0.01 0.14±0.03 0.11±0.01 0.10±0.01 0.15±0.02

0.39±0.05 0.36±0.02 0.33±0.03 0.29±0.02 0.33±0.01 0.29±0.02

6 6 6 6 6 6

of 1 mM MgCl2 during the stimulation by 10 mM CaCl2 abolished the increase of CT completely, while the increase of Isc was unaffected. These results support the idea that Mg2+ interferes with the role of Ca2+ in the exocytotic response, probably by inhibiting Ca2+ entry [7, 15, 17]. The further increase in Isc elicited by the addition of 1 mM Mg2+ to the Isc increase evoked by 10 mM Ca2+ was significant (P<0.05), indicating an additional effect on Na+ transport for both divalent cations. Effect of Mg2+on iNa, Po and NT at constant osmolality

Fig. 4 Effect of Mg2+ and Ca2+ on CT, GT and Isc at constant osmolality. Initially, epithelia were incubated bilaterally with solutions at 200 mOsm/kg H2O. After 120-min equilibration, 15 mM NaCl was replaced by 10 mM Mg2+ (dotted line), 10 mM Ca2+ (solid line), or 10 mM Ca2++1 mM Mg2+ (dashed line) on both sides of the epithelium. Responses of CT, Isc and GT are displayed. Each curve represents the mean of six tissues

Effect of Mg2+ and Ca2+ on CT, GT and Isc at constant osmolality It is generally accepted that Ca2+ plays a key role in the regulation of membrane traffic. Since Mg2+ reportedly inhibits Ca2+ entry [17], it is conceivable that the effects of Mg2+ on CT are related to effects on Ca2+ entry. To clarify possible interactions between the two divalent cations and to evaluate the stimulation of INa by Mg2+, we performed experiments at constant osmolality. In this case, we pre-stimulated Na+ transport by incubating the tissues in solutions at 200 mOsm/kg H2O, and added the divalent cations subsequently. To keep the osmolality constant at 200 mOsm/kg H2O, 15 mM NaCl was replaced by 10 mM XCl2, where X represents Ca2+ or Mg2+. Figure 4 demonstrates that 10 mM MgCl2 significantly increased Isc from 7.3±1.4 to 15.8±2.9 µA/cm2 while CT remained unaltered. CaCl2 (10 mM) increased CT from 0.84±0.02 to 0.99±0.03 µF/cm2 and Isc from 5.4±0.6 to 10.2±2.3 µA/cm2. Surprisingly, the presence

We used noise analysis to specify the stimulatory effects on INa of the addition of Mg2+ at constant osmolality. To exclude possible interference with processes activated by lowering basolateral osmolality, the stimulatory effects of Mg2+ were investigated at different osmotic values, i.e. 260 and 200 mOsm/kg H2O. We replaced 15 mM NaCl by 10 mM MgCl2 to keep solution osmolality constant and applied a slight hypotonic shock towards 200 mOsm/kg H2O by the removal of 33 mM NaCl. iNa, NT and Po were determined from the CDPC-induced fluctuation in INa. In the first set of six tissues, epithelial monolayers were initially bathed in solutions of 260 mOsm/kg H2O (➀ in Fig. 5) and 15 mM NaCl were replaced by 10 mM MgCl2 (②). Subsequently, solution osmolality was reduced to 200 mOsm/kg H2O (➂). Mean Isc recorded in the presence of 10 µM CDPC in the apical bath is depicted in Fig. 5A. Mean INa, iNa, NT and Po are displayed in Fig. 5B. Mg2+ (10 mM) tripled INa (P<0.05) by a threefold increase in NT. Mg2+ significantly stimulated Po (P<0.05), while iNa decreased. The osmotic shock increased INa via a rise in NT, although both iNa and Po decreased significantly (P<0.05). In the second series of experiments (Fig. 6), epithelia at 260 mOsm/kg H2O (➀ in Fig. 6) were submitted first to an osmotic shock towards 200 mOsm/kg H2O (②) after which 15 mM NaCl was replaced by 10 mM MgCl2 (➂). The increase in INa caused by both hypotonic treatment and Mg2+ could be attributed to an increase of NT, while both iNa and Po fell. Both sets of data suggest that extracellular Mg2+ stimulates Na+ transport independent-

510

Fig. 5A,B Effect of Mg2+ at 260 mOsm/kg H2O on iNa, Po and NT followed by a reduction of the osmolality to 200 mOsm/kg H2O. Cells were initially bathed in 260 mOsm/kg H2O solutions on both sides (➀). Thereafter, 15 mM NaCl was replaced by 10 mM Mg2+ bilaterally (②) and then the osmolality of the solutions on both sides of the epithelium was decreased to 200 mOsm/kg H2O by the removal of 33 mM NaCl (➂). A Time course of the change of Isc with 10 µM CDPC in the apical bath (see Materials and methods). The curve is the mean from six tissues. B Bar diagrams of INa, iNa, Po and NT. The hatched bars indicate the presence of Mg2+ in the solutions. *P<0.05 for comparisons between 0 and 10 mM Mg2+

Fig. 6A,B Effect of Mg2+ at 200 mOsm/kg H2O on iNa, Po and NT. Initially A6 epithelia were incubated in 260 mOsm/kg H2O solutions (➀). The osmolality was reduced bilaterally to 200 mOsm/kg H2O by removing 33 mM NaCl (②). Subsequently, 15 mM NaCl was replaced by 10 mM Mg2+ bilaterally (➂). A Time course of the change of Isc recorded in the presence of 10 µM CDPC in the apical bath (see Materials and methods). The curve is the mean from six tissues. B Bar diagrams showing INa, iNa, Po and NT. The hatched bars indicate the presence of Mg2+. *P<0.05 for comparisons between 0 and 10 mM Mg2+

ly of the osmotic state of the cells and via a pathway that is unaffected by osmotic changes.

depends on the membrane lipid composition and on the frequency because of dielectric dispersion phenomena taking place in the organization of the lipids in the membrane [9]. The latter phenomenon will certainly appear in the megahertz range, but is most likely negligible in the audio-frequency range used in this study. Assuming that the lipid composition of the plasma membrane remains unaltered in the protocols we employed, we can therefore assume safely that CT is proportional to membrane area. The results at constant osmolality demonstrate that Mg2+ increased the number of transport sites markedly without increasing membrane area. The failure of CT to increase during Na+ current activation by Mg2+ can be explained in two different ways. The first possibility is that the active channel molecules are not newly inserted by an exocytotic process but rather are already present in the membrane. The second is that insertion by exocytosis takes place together with an enhanced endocytotic process, thereby keeping membrane area approximately constant. The experiments at constant osmolality (Fig. 4)

Discussion In this paper we report data related to the effects of extracellular Mg2+ on CT and on the stimulation of Na+ transport under different osmotic conditions. Extracellular Mg2+ inhibited most of the capacitance increase evoked during a hypotonic challenge, without inhibiting Na+ transport. On the contrary, with Mg2+ in the bathing media the currents were even larger. Measuring the time course of the electrical capacitance allows real-time monitoring of the membrane traffic occurring in the plasma membrane. This approach is based on the generally accepted idea that an increase in the electrical capacitance is the consequence of exocytotic insertion of additional membrane material, such that CT is directly proportional to membrane area [24]. CT is also proportional to the dielectric constant of the membrane (ε). The latter

511

favour the first hypothesis, since they clearly demonstrate interaction between Ca2+ and Mg2+ on the exocytotic pathway. Indeed, the addition of 1 mM Mg2+ to the basolateral bath is sufficient to counteract the increase in CT evoked by 10 mM Ca2+. This CT increase is probably unrelated to insertion of Na+ conducting channels since Mg2+ addition did not attenuate the INa increase. On the contrary, INa increased significantly (P<0.05), although the CT increase was abolished. Even more striking is the discrepancy between channel density and membrane area when challenging the cells with hypotonic solutions. Hypotonicity increased INa conspicuously by elevating NT. The decrease in iNa during hypotonic treatment reflects the decrease in driving force across the apical barrier due to the depolarization caused by the increased Na+ permeability [16] as a consequence of the enormous elevation of NT. The steep initial CT increase occurred simultaneously with the initial changes in GT and Isc. This supports the assumption that vesicular fusion, evoked by the hypotonic shock, adds apical Na+ channels, a process apparently independent of the presence of extracellular Mg2+. The slower and transient increase of CT developed disparately from GT and Isc. This component is clearly Mg2+ sensitive and probably not related to Na+ channel insertion at all. Our data suggest that the Mg2+ effect is located at the basolateral membrane. Indeed, acute addition of Mg2+ at the basolateral membrane at the moment at which the hypotonic shock was induced resulted in effects similar to those of chronic administration. One can argue that the hypotonic shock, by opening a non-specific channel [19], allows a sudden Mg2+ influx, especially when the extracellular [Mg2+] is 10 mM. The finding that extracellular Mg2+ inhibits 86Rb-efflux through this non-specific channel during a hypotonic shock supports this idea [11]. However, it remains doubtful whether this channel is involved in the observed effects in the present study. Moreover, acute removal of extracellular Mg2+ at the moment of the hypotonic shock abolishes the observed effects of Mg2+. The addition of Mg2+ to the basolateral bath stimulates Na+ transport not only during the hypotonic shock, but also at constant osmolality. Our data suggest that this effect of Mg2+ is due to activation of dormant transport sites resulting in an increase in the number of conducting Na+ channels (i.e. NT). Furthermore, we found that Na+ transport could be activated by an increase of the extracellular [Ca2+]. The fact that both divalent cations exert similar and additive stimuli is reminiscent of a Ca2+-sensing mechanism. A so-called Ca2+-sensing receptor has been described in the basolateral membranes of renal cells [3, 21] and this may act also as a Mg2+ sensor [6]. In summary, our data are consistent with the idea that Mg2+ exerts extracellular effects at the basolateral side of A6 epithelia. First, the inhibitory effect of Mg2+ on the increase of CT either evoked by hypotonicity or by extracellular Ca2+ reveals a competitive inhibition of Mg2+, probably on the process of Ca2+ entry. This Ca2+ entry step is, however, not related to the activation of the transepithelial Na+ transport. Second, the increase of Isc

in the presence of Mg2+ is independent of solution osmolality and appears to be mimicked by an increase in extracellular Ca2+, although to a lesser extent. On the contrary, the stimulatory effect of both divalent cations on INa was additive. This behaviour suggests a receptor like that cloned originally by Brown et al. [5]. Whether such a mechanism is involved in the activation process of Na+ transport is not yet known. Third, the combination of the methods used is very important. By using noise analysis only, one might assume that the number of Na+ channels increases markedly upon the addition of Mg2+ or by hypotonic solutions. By measuring CT and GT it can be established that hypotonicity activates both CT and GT, while Mg2+ increases only GT and concomitantly INa. Acknowledgements We thank Dr. W-M. Weber for his critical comments and helpful suggestions on this paper. We are also grateful to Ms Els Larivière for performing some of the capacitance measurements. This project was supported by research grants from the “Fonds voor wetenschappelijk onderzoek – Vlaanderen” (G.0179.99), the Interuniversity Poles of Attraction Program – Belgian State, Prime Minister’s Office – Federal Office for Scientific, Technical and Cultural Affairs IUAP P4/23 and the bilateral program BIL96/23 of the Flemish government, providing the support of DC.

References 1. Aguilera AJ, Kirk KL, DiBona GF (1978) Effect of magnesium on sodium transport in toad urinary bladder. Am J Physiol 234:F192–F198 2. Baker PF, Hodgkin AL, Ridgway EB (1970) Two phases of calcium entry during the action potential in giant axons of Loligo. J Physiol (Lond) 208:80P-82P 3. Bapty BW, Dai LJ, Ritchie G, Jirik F, Canaff L, Hendy GN, Quamme GA (1998) Extracellular Mg2+- and Ca2+-sensing in mouse distal convoluted tubule cells. Kidney Int 53:583–592 4. Blazer-Yost BL, Liu X, Helman SI (1998) Hormonal regulation of ENaCs: insulin and aldosterone. Am J Physiol 274:C1373–C1379 5. Brown EM, Gamba G, Riccardi D, Lombardi M, Butters R, Kifor O, Sun A, Hediger MA, Lytton J, Hebert SC (1993) Cloning and characterization of an extracellular Ca2+-sensing receptor from bovine parathyroid. Nature 366:575–580 6. Brown EM, Vassilev PM, Hebert SC (1995) Calcium ions as extracellular messengers. Cell 83:679–862 7. Cessna SG, Chandra S, Low PS (1998) Hypo-osmotic shock of tobacco cells stimulates Ca2+ fluxes deriving first from external and then internal Ca2+ stores. J Biol Chem 273: 27286–27291 8. Coupaye-Gerard B, Kim HJ, Singh A, Blazer-Yost BL (1994) Differential effects of brefeldin A on hormonally regulated Na+ transport in a model renal epithelial cell line. Biochim Biophys Acta 1190:449–456 9. Daniel VV (ed) (1967) Dielectric relaxation. Academic Press, London 10. De Smet P, Erlij D, Van Driessche W (1997) Insulin effects on ouabain binding in A6 renal cells. Pflügers Arch 434:11–18 11. De Smet P, Li J, Van Driessche W (1998) Hypotonicity activates a lanthanide-sensitive pathway for K+ release in A6 epithelia. Am J Physiol 275:C189–C199 12. De Smet P, Simaels J, Van Driessche W (1995) Regulatory volume decrease in a renal distal tubular cell line (A6). II. Effect of Na+ transport rate. Pflügers Arch 430:945–953 13. Erlij D, De Smet P, Van Driessche W (1994) Effect of insulin on area and Na+ channel density of apical membrane of cultured toad kidney cells. J Physiol (Lond) 481:533–542

512 14. Fisher RS, Grillo FG, Sariban-Sohraby S (1996) Brefeldin A inhibition of apical Na+ channels in epithelia. Am J Physiol 270:C138–C147 15. Fleckenstein-Grün G (1996) Calcium antagonism in vascular smooth muscle cells. Pflügers Arch 432(Suppl 3):R53–R60 16. Granitzer M, Nagel W, Crabbé J (1991) Voltage dependent membrane conductances in cultured renal distal cells. Biochim Biophys Acta 1069:87–93 17. Kupper J, Ascher P, Neyton J (1996) Probing the pore region of recombinant N-methyl-D-aspartate channels using external and internal magnesium block. Proc Natl Acad Sci USA 93:8648–8653 18. Lacoste I, Brochiero E, Ehrenfeld J (1993) Control of Na+ and H+ transports by exocytosis/endocytosis phenomena in a tight epithelium. J Membr Biol 134:197–212 19. Li J, De Smet P, Jans D, Simaels J, Van Driessche W (1998) Swelling-activated cation-selective channels in A6 epithelia are permeable to large cations. Am J Physiol 275:C358–C366 20. Ling BN, Eaton DC (1989) Effects of luminal Na+ on single Na+ channels in A6 cells, a regulatory role for protein kinase C. Am J Physiol 256:F1094–F1103 21. Riccardi D, Lee WS, Lee K, Segre GV, Brown EM, Hebert SC (1996) Localization of the extracellular Ca2+-sensing receptor and PTH/PTHrP receptor in rat kidney. Am J Physiol 271:F951–F956

22. Sebille S, Pereira M, Millot JM, Jacquot J, Delabroise AM, Arnaud M, Manfait M (1998) Extracellular Mg2+ inhibits both histamine-stimulated Ca2+-signaling and exocytosis in human tracheal secretory gland cells. Biochem Biophys Res Commun 246:111–116 23. Van Driessche W, De Vos R, Jans D, Simaels J, De Smet P, Raskin G (1999) Transepithelial capacitance decrease reveals closure of lateral interspace in A6 epithelia. Pflügers Arch 437:680–690 24. Van Driessche W, Erlij D (1991) Cyclic AMP increases electrical capacitance of apical membrane of toad urinary bladder. Arch Int Physiol Biochim Biophys 99:409–411 25. Verrey F, Groscurth P, Bolliger U (1995) Cytoskeletal disruption in A6 kidney cells: impact on endo/exocytosis and NaCl transport regulation by antidiuretic hormone. J Membr Biol 145:193–204 26. Weng K, Wade JB (1994) Effect of brefeldin A on ADH-induced transport responses of toad bladder. Am J Physiol 266:C1069–C1076 27. Wills NK, Purcell RK, Clausen C, Millinoff LP (1993) Effects of aldosterone on the impedance properties of cultured renal amphibian epithelia. J Membr Biol 133:17–27