INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE MAMMALIAN CORTICAL collecting duct (CCD) plays a major role in regulating renal Na⁺ absorption and as such is important in controlling total body Na⁺ homeostasis. The general model of Na⁺ reabsorption in CCD involves coordinated Na⁺ entry across the apical membrane through Na⁺ channels and Na⁺ exit across the basolateral membrane by Na-K-ATPase (39). Previous studies of isolated perfused CCD and model culture epithelia (such as A6 cells) have yielded invaluable information about the membrane properties and transport mechanisms involved in Na⁺ regulation in this segment. These studies indicate that many hormones [e.g., adrenal steroids, arginine vasopressin (AVP), phorbol esters, bradykinin, and endothelin] contribute to regulation of Na⁺ transport in CCD (5, 6, 39). Among these, aldosterone (Aldo) is the main stimulant of Na⁺ reabsorption in the mammalian collecting duct.

Administration of Aldo stimulates Na⁺ reabsorption by augmenting Na⁺ entry across the apical Na⁺ channels and by subsequently activating basolateral Na-K-ATPase (35, 39). Because of the importance of this process for Na⁺ balance, the cellular and molecular mechanisms of hormonal regulation of Na⁺ transport remain under intense study using a variety of experimental preparations. Cell culture models provide many advantages to address issues that are difficult to study in vivo. However, most cell culture studies of the effects of mineralocorticoids have used the A6 amphibian cell line rather than mammalian cells (16). M-1 cells, a mammalian cell line derived from microdissected CCD of a mouse transgenic for the early region of SV40 virus, maintain many characteristics of CCD and were used for this study. Although M-1 cells have been utilized since their development for a variety of studies, particularly addressing epithelial Na⁺ channels, the hormonal regulation of transepithelial transport in these cells has not been fully characterized. Therefore, the purpose of the present studies was to determine the acute and chronic hormonal regulation of Na⁺ transport in these mammalian cells.

The present study was designed to address several important issues. First, we characterized the transepithelial transport characteristics of these cells and the conditions under which M-1 cells in serum-free media can be used to study transport. Second, we characterized the differential effects of Aldo and dexamethasone (Dex) on transport. We were interested in comparing the effects of both hormones, the time course during which an effect on transport developed, and the possibility that their actions overlap. Third, we investigated the effects of other hormones that acutely modulate Na⁺ transport in other CCD preparations. These included epidermal growth factor (EGF), phorbol esters [phorbol myristate acetate (PMA)], and AVP. Finally, we investigated whether these mammalian cells contain an endogenous protease activity as has been recently reported in A6 cells.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell culture. The M-1 cell line was originally derived by Stoos et al. (47) from a mouse transgenic for the early region of simian virus 40 [strain Tg(SV40)Bri/7]. The M-1 cells used in this study were obtained from Drs. Geza Fejes-Toth, B. Stoos, and J. Garvin and belonged to the same strain as the ones originally described. Cells were used from passages 11-17 with no obvious changes in electrical properties among different passages. The cells were initially cultured in plastic flasks (Costar, Corning, NY) and grown to confluence in a humidified incubator at 37°C and 5% CO₂. The cultures were initially maintained in a defined medium consisting of equal amounts of Ham's F-12 and low-glucose DMEM (Sigma), supplemented with 2 mM L-glutamine, 50 U/ml penicillin, 50 mg/ml streptomycin, 5% fetal bovine serum, growth promoting factors (transferrin, insulin, sodium selenite, 6.25 mg/ml each), and 100 nM Dex. Media were changed every 2 days. After cells reached confluence, usually in ~7 days, they were passaged by trypsinization and plated onto semipermeable membranes for electrophysiological studies or on coverslips for intracellular Ca²⁺ ([Ca²⁺]_i) measurements. For transepithelial measurements, cells were seeded on semipermeable membranes, 24 mm in diameter (Transwells, Costar). The seeding density was 2.5 × 10⁵ cells/Transwell and the active surface area was 4.52 cm². On the third day after seeding, the culture medium was changed to the identical media described above except without serum. Cells were maintained in this media (Dex) until formation of confluent monolayers. Cells seeded on coverslips for optical measurements were treated in exactly the same way and were usually studied 5-8 days after seeding. Cell monolayers were confirmed to be confluent by development of adequate transepithelial resistance (R_te) and voltage (V_te), which were monitored daily. At this point, the media was changed again to contain Aldo (1 µM), or Dex (100 nM), or vehicle (ethanol). Electrical monitoring of V_te and R_te was maintained daily, and media changes were kept at 2-day intervals.

Electrophysiological transepithelial measurements. Measurements of V_te, R_te, and short circuit current (I_eq) were obtained by two methods. In the first method, the Transwells with the M-1 cells forming a monolayer were mounted in a Ussing-type chamber, especially designed to accept Transwells (EVC-4000; WPI, Sarasota, FL). A Transwell containing no cultured cells filled with media or mammalian Ringer solution was used to null junction potentials and to compensate for the resistance of the system. The confluent monolayer of M-1 cells separated the apical and basolateral compartments. The two compartments were filled with continuously circulating symmetrical solutions which contained 25 mM bicarbonate, and were continuously bubbled with 5% CO₂ (pH 7.4). The whole apparatus was jacketed by warm air and maintained at a constant temperature of 37°C. Two low-resistance Ag:AgCl electrodes made contact with the two chambers through Ringer-Agar bridges placed in the solution lines and were used to measure V_te. Two other Ag:AgCl electrodes were placed very close on either side of the tissue through 3 M KCl-agar bridges and were used to inject a current across the epithelial monolayer. The monolayer culture was maintained in an essentially open-circuit condition and V_te was continuously measured in this manner. During periodic intervals (every 5 min), V_te was clamped at 0 mV (short-circuit condition), and I_eq was measured. R_te was calculated from the ratio of V_te to I_eq.

In the second method, V_te was measured by means of a set of two Ag:AgCl electrodes (STX electrodes, WPI) using an ohm/volt meter (EVOM, WPI). The voltage readings were nulled to zero when V_te was measured across a Transwell with no cultured cells. Transepithelial measurements on confluent M-1 monolayers were performed while the Transwells sat undisturbed in the culture plates. Measurements were made under sterile conditions in a laminar flow hood and thus repeated measurements could be obtained. V_te was determined with the basolateral chamber as reference. R_te was obtained by injecting current across the epithelium for a period of <1 s. The I_eq was calculated as the ratio of V_te to R_te (corrected for the resistance of the fluid). The distance between the two electrodes was constant in all experiments. I_eq was normalized, by dividing I_eq by the surface area (4.52 cm²) of active membrane. Using this method, we were able to monitor V_te and R_te on a daily basis to verify formation of confluent monolayers without contaminating the cultures. Multiple measurements on the same Transwell are not possible using the Ussing chamber. To study the acute effects of hormones, measurements using EVOM were obtained from Transwells placed in a converted incubator maintained at 37°C and at 5% CO₂. After a period of equilibration (1-2 h) in the incubator, multiple and periodic measurements of V_te and R_te were obtained once every 3 or 5 min. Once baseline readings had stabilized, the experimental agent (see below) was added to the appropriate compartment(s) in three Transwells at a time, while vehicle was added to another set of three Transwells whose V_te and R_te closely matched those of the hormone-treated ones. In this manner, the effects could be compared in a paired fashion while eliminating wide variations in transport that would inevitably occur as the cultures grow in time.

Measurements of intracellular Ca²⁺. Changes in [Ca²⁺]_i were determined from fluorescence measurements of fura 2 trapped intracellularly. Cells grown on coverslips were washed twice with mammalian Ringer solution and then incubated at 37°C with a solution containing 5 µM of the acetoxymethyl ester of fura 2 (fura 2-AM; Molecular Probes, Eugene, OR). The fura 2-AM, a precursor of the dye fura 2, was added from a 1 mM stock solution in DMSO. Adequate cell loading was obtained by incubating the cells for ~60 min. Cells were then washed twice with control solution and then transferred to a special chamber where the coverslips formed the bottom of the chamber. The Ringer solution was HCO₃ free (HEPES buffered) with a pH of 7.4 at 37°C, and the cells were continuously superfused with this solution. Using a PTI system (PTI, Princeton, NJ), we determined [Ca²⁺]_i values by alternating the excitation wavelength between 340 and 380 nm and measuring the emission signal at 510 nm. Change in fluorescence ratio (R_340/380) obtained in this manner is proportional to changes in [Ca²⁺]_i. Data points were recorded at 12 points/s. Approximately 15-20 cells were in the field of measurement.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Electrical transepithelial parameters: baseline values. Our initial experiments defined the time course during which M-1 cells develop transport properties that are characteristic of the CCD. To do this, daily measurements of V_te, R_te, and I_eq were taken. Measurements were conducted either before changing media or at least 3 h thereafter. All Transwells were measured daily in the laminar flow hood using the STX-2 electrodes and the EVOM ohm/volt meter. Each (one) culture plate, with six Transwells, was removed from the incubator individually, and measurements were completed within 3-5 min. This short time interval is necessary to prevent drift in measurements due to changes in temperature and pH of the cells. Measurements over a longer period of time (>10 min) displayed gradual instability, particularly in R_te, which usually increased slowly. Concurrently, daily experiments were also conducted on at least two Transwells using the Ussing chamber, as outlined above. Measurements from both techniques could thus be compared.

Four days after seeding (1 day after serum removal), cells developed apical-negative V_te ranging from 5 to 25 mV and R_te ranging from 200 to 800  · cm². The presence of Dex proved to be crucial, because the transepithelial readings consistently collapsed in its absence. Subsequent daily readings showed that adequate measurements are maintained for the next 4-6 days with optimal values usually obtained 1-2 days after removal of serum. In 45 cultures, the average peak values were as follows: V_te = 24.5 ± 1.6 mV, R_te = 2,795 ± 163  · cm², and I_eq = 8.7 ± 0.5 µA/cm². In cultures in which serum was not removed, the measurements were not significantly different. High V_te and R_te are characteristic properties of the mammalian CCD.

When measured in the Ussing chamber, the time course for the development of V_te and I_eq was essentially the same as that obtained from EVOM readings. However, two important observations were apparent. The yield of successful experiments in the Ussing chamber was very low since stable readings could often not be obtained. Usually, this was caused by a continuous downward drift in R_te that eventually led to a progressive rise in I_eq to an open reading; this occurred after a short (~10 min) lag time during which I_eq remained stable. On the other hand, when stable measurements were possible, these readings were usually a fraction (30-50%) of the measurements obtained using the EVOM. Although one would expect that the environment of cells in the Ussing chamber would be better controlled with a more stable pH and temperature, EVOM experiments were apparently less detrimental to the cells. These observations suggested that these cells are relatively fragile and very sensitive to physical perturbations, as would occur when solutions are circulating in the Ussing chamber. In support of this, transepithelial measurements (using EVOM) immediately after changing the media yielded much lower values of V_te and I_eq compared with immediately before media change. In other experiments, using EVOM, we found out that even careful removal of the bathing media, followed by returning it back to the respective compartment, substantially diminished V_te and I_eq. These effects were transient, and cells usually fully recovered if left undisturbed for 2-3 h. These effects may indicate a loss of monolayer integrity (presumably by affecting tight junctions) and diminished transport in response to even minimal perturbations. These results rendered the use of the Ussing chamber for studying transepithelial transport in these cells extremely difficult. Consequently, we performed most experiments using the EVOM on Transwells while still in the culture plates, in a converted cell culture incubator maintained at 37°C and 5% CO₂. Measurements conducted in this fashion were consistent and reliable.

Effect of amiloride. Na⁺ transport in M-1 cells, like the CCD, presumably involves luminal Na⁺ uptake through Na⁺ channels. To verify that apical Na⁺ channels are responsible for the development of I_eq and V_te in M-1 cells, amiloride was added to the apical compartment. In 22 experiments, addition of 10 µM amiloride rapidly depolarized V_te from 20.4 ± 2.2 to 4.7 ± 0.8 mV and increased R_te from 2,812 ± 131 to 3,977 ± 271  · cm², and the I_eq decreased from 8.0 ± 1 to 1.2 ± 0.2 µA/cm² (P < 0.001).

Effects of Aldo and/or Dex on transepithelial measurements. The next series of experiments was performed to determine how M-1 cells respond to treatment with glucocorticoids and/or mineralocorticoids. Three factors were determined: 1) the degree of stimulation of I_eq and V_te by Aldo and Dex, 2) the time course of steroid-dependent stimulation of Na⁺ transport, and 3) whether the effects of these hormones were additive.

Upon removal of serum from the culture medium on the third day after seeding, M-1 cells were divided into three groups. One group had 100 nM Dex continuously present. In another group, Aldo (1 µM) but not Dex was added. The third group had no steroids ("None" group) in the culture medium but did contain vehicle (ethanol). Transwells in each group were matched to have comparable initial readings of V_te and R_te. I_eq was stimulated in both the Dex and the Aldo groups. Maximal stimulation appeared at ~24 h, after which a gradual decrease in I_eq was observed; therefore all subsequent data refer to the 24 h time points. The vehicle control group (i.e., was not treated with any steroids) did not show any stimulation of I_eq and had the lowest readings. Although both glucocorticoids and mineralocorticoids significantly stimulated I_eq and V_te, stimulation by Dex was always greater than that by Aldo. Figure 1 shows the stimulatory effects of steroids on I_eq after 24 h of treatment with the respective hormones. In Dex-treated cells, I_eq (9.2 ± 0.9 µA/cm², n = 11) and V_te (22.4 ± 2.5 mV) were respectively higher than those in Aldo (6.0 ± 0.4 µA/cm² and 17.7 ± 1.5 mV) or vehicle control (None, no steroids) (4.1 ± 0.3 µA/cm² and 13.0 ± 1.4 mV). The differences in I_eq were statistically significant among all groups. There was no statistical difference in V_te between the Aldo- and the Dex-treated M-1 cells. The data summarized in Fig. 1 indicates that both Aldo and Dex stimulate Na⁺ transport in M-1 cells.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Stimulation of equivalent short-circuit current (Ieq, left) and transepithelial voltage (Vte, right) by dexamethasone (Dex) or aldosterone (Aldo). Aldo or Dex significantly increased Ieq and Vte in comparison to nontreated cells ("None"). Stimulation by Dex was highest, and the differences among all groups were statistically significant (P < 0.05, n = 11). * Significantly different from None group.

The next series of experiments was performed to determine whether the stimulatory effect of Aldo is additive to that of Dex. A nonadditive effect would argue against a separate regulatory influence of Aldo. As shown in Fig. 2, exposure of cells to Dex (100 nM) and Aldo (1 µM) increased I_eq from 7.5 ± 0.6 (Dex only) to 10.2 ± 0.5 µA/cm² (n = 9, P < 0.05). V_te in cells treated with both hormones (36.4 ± 3.4 mV) was also higher than in cells treated with Dex only (30.5 ± 4.5 mV); however, the difference was not statistically significant.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of Aldo on Ieq and Vte in presence of Dex. Stimulation of Ieq (left) in presence of both Aldo and Dex was higher than stimulation by Dex alone (P < 0.05, n = 9). Increase in Vte (right) was not statistically significant. * Significantly different from Dex group.

We also checked the effect of Aldo when stimulation by Dex was less than maximal by lowering the concentration of Dex to 10 nM (1/10th usual Dex). As shown in Fig. 3, I_eq in cells treated with 1 µM Aldo plus 10 nM Dex (5.7 ± 0.3 µA/cm²) was higher (n = 15, P < 0.01) than that in 10 nM Dex only (4.6 ± 0.3 µA/cm²). The difference in V_te was not statistically significant (15.2 ± 1.1 vs. 13.2 ± 0.8 mV). These results are qualitatively similar to the previous experiments with 100 nM Dex and still showed an additive effect of Aldo to that of Dex. However, I_eq was significantly higher (9.7 ± 0.8 µA/cm²) in the group with Dex at 100 nM concentration. These data clearly demonstrate an additional effect of Aldo to that of Dex.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Effect of Aldo on Ieq (left) and Vte (right) in presence of 10 nM Dex. Stimulation of Ieq was greater in presence of Aldo and 10 nM Dex than with 10 nM Dex alone (P < 0.01, n = 15). * Significantly different from Dex group.

To further determine whether the stimulatory effect of Dex was purely through binding to the glucocorticoid receptor, we examined the effect of spironolactone, a mineralocorticoid antagonist, on I_eq and V_te in Dex-treated M-1 cells. As shown in Fig. 4, spironolactone (10 µM) in the presence of 100 nM Dex significantly decreased I_eq from 8.3 ± 0.6 (Dex) to 5.8 ± 0.6 µA/cm² (n = 12, P < 0.01). V_te also decreased from 22.6 ± 1.7 to 16.8 ± 2.3 mV (P < 0.05). These results indicate that at least a component of the stimulatory effect of Dex is probably mediated through activation of mineralocorticoid receptors.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Effect of mineralocorticoid antagonist (spironolactone) on glucocorticoid stimulation of Ieq (left) and Vte (right). Spironolactone (Spironolac, 10 µM) significantly inhibited both Ieq and Vte (* P < 0.05, n = 12) in Dex-treated cells.

In some studies, lower concentrations of mineralocorticoids (10-100 nM) have been used to stimulate Na⁺ transport in cultured cells of the collecting duct. In the next series of experiments, we lowered the concentration of Aldo to 100 nM and examined whether stimulation of the I_eq was maintained. As can be seen in Fig. 5, Aldo, at 100 nM, still caused a significant increase in I_eq and V_te compared with cells which were not treated with steroids. On the other hand, Aldo stimulation of I_eq and V_te was less than that caused by Dex, similar to the results of Fig. 1.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Stimulation of Ieq (left) and Vte (right) by lower concentration of Aldo (100 nM) or Dex. Lower concentration of Aldo (100 nM) or Dex (100 nM) significantly increased Ieq and Vte in comparison to nontreated cells (P < 0.05, n = 8). Stimulation by Dex was highest among all groups. * Significantly different from None group.

Response of M-1 cells to acute treatment with hormones. In contrast to adrenal steroids, whose effects occur predominantly over hours to days, several other hormones acutely influence Na⁺ transport in CCD. Among these we investigated the effects of phorbol esters as activators of protein kinase C (PKC), EGF, elevations in [Ca²⁺]_i, and AVP.

PMA at 10⁹ M has been shown to acutely inhibit Na⁺ reabsorption and K⁺ secretion in the rabbit CCD (22) and to inhibit Na⁺ channels in rat CCD (17), although Rouch et al. (40) did not find an effect in the perfused rat CCD. In our experiments, addition of PMA did not affect either V_te or I_eq (Fig. 6) in M-1 cells treated with either Dex (n = 10) or Aldo (n = 6). The results suggest that activation of PKC as expected from treatment with PMA does not significantly inhibit basal Na⁺ transport in these cells.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Acute effect of phorbol myristate acetate (PMA) on Ieq (top) and Vte (bottom). PMA (108 M) did not significantly change Ieq or Vte in M-1 cells treated with Dex () or Aldo (); , effect of adding vehicle only.

EGF, administered acutely in the rabbit CCD, was shown to drastically decrease Na⁺ reabsorption with a corresponding fall in V_te (50). Further studies showed that the effect of EGF is peritubular and depends on the influx of Ca²⁺ across the basolateral membrane (51). This suggested that the effect of EGF is mediated through an increase in [Ca²⁺]_i. In our studies, as shown in Fig. 7, addition of EGF to the bath did not affect either V_te or I_eq in M-1 cells treated with either Dex (n = 8) or Aldo (n = 5). Figure 7 also shows the inhibition of transepithelial V_te and I_eq by amiloride. Since an EGF effect in intact CCD is known to be mediated by a change in activity of [Ca²⁺]_i, we checked whether EGF did increase [Ca²⁺]_i in M-1 cells. Alternatively, we also examined whether an induced increase in [Ca²⁺]_i would have any influence on I_eq and V_te in these cells. As shown in Fig. 8, top, using fura 2 to measure [Ca²⁺]_i, addition of EGF elicited a small increase in [Ca²⁺]_i. Subsequent addition of thapsigargin (1 µM) caused a substantial increase in [Ca²⁺]_i that partially recovered. Six similar experiments were conducted. In another set of four parallel experiments, thapsigargin was added to Dex-treated M-1 cells on Transwells; no change in I_eq or V_te was seen (Fig. 8, bottom).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Acute effect of epidermal growth factor (EGF) on Ieq (top) and Vte (bottom). EGF (10 ng/ml) did not significantly influence Ieq or Vte in M-1 cells treated with Dex () or Aldo (); , effect of adding vehicle only. Last segment of the tracing shows that amiloride (Amil, 10 µM) completely blocked Ieq and Vte in all groups.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Response of Ieq and Vte to changes in intracellular Ca2+ ([Ca2+]i). Top: an actual tracing showing that EGF moderately increases [Ca2+]i measured as a change in fluorescence ratio (R340/380) of fura 2 trapped intracellularly. Thapsigargin (Thapsig, 1 µM) substantially but transiently increased [Ca2+]i. Bottom: thapsigargin, which increases [Ca2+]i did not significantly affect Ieq in Dex-treated M-1 cells (); , effect of adding vehicle only.

Unlike PMA or EGF, AVP was shown to activate Na⁺ reabsorption in the rabbit CCD in an early phase (5, 8, 12), which was soon followed by an inhibitory effect (12, 23). In contrast to the rabbit CCD, addition of AVP (1 µM) to rat CCD produced a sustained increase in Na⁺ transport that was even greater when rats were treated with deoxycorticosteroid hormones (12, 37, 48). In nine experiments on M-1 cells, addition of AVP (10⁶ M) caused a transient large increase in I_eq and V_te (Fig. 9) and a much smaller sustained increase in I_eq. To investigate whether the stimulation of I_eq was due to activation of luminal Na⁺ conductance, we conducted a series of seven experiments in which AVP was added after inhibiting the luminal Na⁺ channels with amiloride (10 µM). As shown in Fig. 10, addition of AVP in the absence of amiloride (top tracing) caused the usual transient increase in I_eq similar to Fig. 9 above. In paired experiments, addition of 10 µM amiloride first (Fig. 10, bottom tracing) caused rapid and almost complete inhibition of I_eq as observed earlier (see Fig. 7). In the continued presence of amiloride, addition of AVP caused a substantial transient increase in I_eq similar to that observed in the absence of amiloride. These results indicate that AVP stimulation of I_eq in M-1 cells is not secondary to activation of amiloride-sensitive Na⁺ channels.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Response of M-1 cells to arginine vasopressin (AVP). Addition of 1 µM AVP to bath elicited a transient and large increase in Ieq (top) and Vte (bottom) of Dex-treated M-1 cells (); , no effect of adding vehicles.

View larger version (11K): [in this window] [in a new window]   Fig. 10.   Response of M-1 cells to AVP after pretreatment of cells with amiloride. Addition of 1 µM AVP to control M-1 cells () caused transient increase in Ieq as shown in Fig. 9 above. In paired experiments, addition of 10 µM amiloride () inhibited Ieq almost completely. In the continued presence of amiloride, addition of AVP (1 µM) still caused a transient and large increase in Ieq. Seven similar experiments were conducted.

Response to protease inhibition. Recently, Vallet et al. (49) reported a novel autocrine mechanism for activation of Na⁺ transport in A6 cells. They reported the identification of a novel serine protease, CAP1, which activated Na⁺ channels. To determine whether mammalian CCD cells express a similar mechanism of activation of Na⁺ transport, protocols similar to those reported in A6 cells were followed. Aprotinin (28 µg/ml), a protease inhibitor, was added to the apical solution for ~12 h and V_te, R_te, and I_eq were measured. I_eq was reduced by 49 ± 9% (to a mean of 1.6 ± 0.3 µA/cm², n = 12) in cells exposed to aprotinin. Trypsin (200 µg/ml), was then added to the luminal solution of both aprotinin-exposed and control cells. As can be seen in Fig. 11, I_eq increased significantly (by 103 ± 27%, P < 0.05) over 5-10 min in cells previously treated with aprotinin. Application of trypsin to cells that had not been exposed to aprotinin had no effect on V_te or R_te. Amiloride abolished I_eq in all cells. These data are consistent with an endogenous protease, which normally activates Na⁺ transport in mammalian CCD cells, as it does in A6 cells. Inhibition of this protease activity with aprotinin decreases this activation, and application of trypsin can lead to reactivation.

View larger version (9K): [in this window] [in a new window]   Fig. 11.   Trypsin stimulation of Ieq in M-1 cells pretreated with aprotinin. In 12 experiments, addition of trypsin (200 µg/ml) to the luminal solution significantly increased Ieq (P < 0.05) in M-1 cells that were pretreated with aprotinin (, top tracing). Control cells showed no change (, bottom tracing).

RT-PCR. No PCR products were obtained in control reactions in which the RT was omitted from the cDNA synthesis. All three primer sets yielded the expected products of ~913, 785, and 541 bp as shown in Fig. 12. Data are from cells treated with vehicle only or Aldo. Appropriately sized PCR products were consistently obtained with Aldo-treated cells. Although Fig. 12 suggests that Aldo directly increases the mRNA levels of all three subunits in M-1 cells, these results were not quantitated for the present series of experiments.

View larger version (36K): [in this window] [in a new window]   Fig. 12.   Ethidium bromide-stained bands of PCR products as labeled and as described in text. Band between  and  lanes is a ladder showing 1636, 1018, and 506 nucleotide bands. For the -, -, and -subunits in absence () of Aldo, bands were clearly present in other experiments.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The present studies demonstrate that M-1 cells represent a useful model for studying the regulation of Na⁺ transport by CCD cells. The cells respond to either glucocorticoids or mineralocorticoids; this response is mediated at least in part by mineralocorticoid receptors (spironolactone sensitive). Acute regulation in these cells is distinctive in that AVP has an effect independent of Na⁺ channels, but PKC activation and increases in intracellular calcium are without effects, in contrast to the rabbit CCD.

Most studies aimed at investigating the effects of hormones on Na⁺ regulation in the CCD, especially chronic exposure, have relied on preparations derived from whole animals or amphibian cultured cells. The main drawback of the whole animal studies is that hormone treatments induce unavoidable secondary changes that themselves affect transport. For example, administration of steroids in vivo can affect renal hemodynamics, distal fluid delivery, and electrolyte and hormone status. Although the A6 amphibian culture model has been extraordinarily valuable in delineating the mechanisms and regulation of Na⁺ transport, possible differences with mammalian tissues need to be examined. These considerations underlie the use of a mammalian cell culture model in the present studies to examine CCD Na⁺ transport.

Although M-1 cells are being increasingly utilized as a mammalian CCD model with which to study Na⁺ transport, the regulation of transepithelial Na⁺ transport in these cells has not been fully characterized. This characterization is crucial if these cells are to be used to understand the behavior of the intact CCD. A principal aim of our study was therefore to characterize the response of M-1 cells to hormones known to alter Na⁺ transport in the intact mammalian CCD. Two aspects of hormonal responses were addressed: chronic (hours/days) regulation by Aldo and Dex, and acute regulation (minutes) by EGF, AVP, PKC activation by PMA, and intracellular calcium.

The I_eq measured in these studies is clearly mediated predominantly by epithelial Na⁺ channels (ENaC). Both I_eq and V_te were immediately and almost completely inhibited by low concentrations of amiloride. In addition, the mRNA of all three subunits of ENaC were detected by RT-PCR; this confirms the findings (Northern blots) of Letz et al. (28) on the presence of mRNA for all three subunits of the epithelial Na⁺ channel in M-1 cells. Other investigators have characterized whole cell currents and the single channel properties of the Na⁺ channels in these cells (10, 11, 25, 28). The conductances are amiloride sensitive and have selectivity of Li⁺ > Na⁺ K⁺ (10, 11, 28); single channels have relatively low conductances (~5-8 pS). Using patch clamp techniques, Chalfant et al. (10) have further characterized the Na⁺ channel in M-1 cells as having complex kinetics involving more than two open and two closed states. In addition to these typical amiloride-sensitive channels, M-1 cells also possess amiloride-insensitive, nonselective cation channels (1, 26); but these are probably not involved in the amiloride-sensitive transepithelial properties reported here.

M-1 cells in the present studies responded to both Dex and Aldo. Although the degree of stimulation with Aldo or Dex is relatively small, this stimulation was consistent and significant. It should be noted that in other in vitro studies, where higher absolute currents were obtained (11, 25, 47), measurements were obtained in smaller wells (diameter 12 mm) and/or in undefined media, such as PC-1 media, which contained serum factors. In our studies, it was important to use defined media to avoid nonspecific effects that may be induced by serum or other undefined factors. Part of this response was probably mediated via mineralocorticoid receptors, since the response was sensitive to spironolactone and since Aldo augmented transport even when Dex was present at maximally effective concentrations. In other studies of cultured collecting duct cells, both Aldo and Dex were also found to stimulate electrogenic Na⁺ transport (24, 34). Response to both glucocorticoids and mineralocorticoids contrasts with the usual concept that mineralocorticoids and not glucocorticoids regulate distal nephron ion transport. However, several previous studies using a variety of models (including whole animals and primary cultures) have demonstrated effects of glucocorticoids, particularly synthetic agents (4, 13, 27, 34, 42, 43). Glucocorticoids are potent ligands for mineralocorticoid receptors, but mineralocorticoid target tissues are usually "protected" by metabolism of endogenous glucocorticoids by 11--hydroxysteroid dehydrogenase (20). Glucocorticoid receptors may also modulate Na⁺ transport in collecting duct cells directly (27, 34). However, endogenous glucocorticoids probably have little regulatory role in CCD Na⁺ transport (via either glucocorticoid or mineralocorticoid receptors) because of the action of 11--hydroxysteroid dehydrogenase in principal cells. Withdrawal of both hormones in M-1 cells resulted in loss of V_te and I_eq. This finding is reasonably consistent with findings in other preparations. For example, Reif et al. (37) showed that V_te and Na⁺ transport were minimal in CCD derived from rats that have not been treated with excess mineralocorticoids. In sum, M-1 cells can be used to study steroid stimulation of Na⁺ transport, but the relative response is small compared with some models, rendering distinction between glucocorticoid and mineralocorticoid responses difficult.

In terms of acute regulation, M-1 cells did not respond to either EGF, PMA to stimulate PKC, or increases in intracellular calcium with thapsigargin. This clearly differs from the response of the microperfused rabbit CCD, which responds to each of these stimuli with an abrupt decrease in Na⁺ transport (19, 22, 50). The lack of response of the M-1 cells does, however, parallel the lack of response of the rat CCD to similar stimuli (40). Whether these differences represent species differences or differences based on culture conditions cannot be clarified by the present data. Others have examined the response of M-1 cells to some agents; Stoos and colleagues (45, 46) have described that M-1 cell transport is inhibited by 1) endothelium-derived relaxing factor (coculture with endothelial cells and stimulation with bradykinin or acetylcholine) and 2) the combination of atrial natriuretic factor and bradykinin. However, both inhibitory responses were complex in that single agents or cGMP analogs alone had no effect (45, 46).

Antidiuretic hormone (or arginine vasopressin, AVP) transiently stimulated I_eq in the present studies (Fig. 9). The response of the intact CCD to AVP differs with species. In the rabbit, AVP inhibits Na⁺ transport after an initial brief stimulation (5, 23); in the rat CCD, AVP stimulates Na⁺ transport, particularly in tubules from deoxycorticosterone-treated animals (12, 37, 48). In cultured CCD cells from rabbit, AVP stimulates Na⁺ transport (8). Other studies have reported that a component of AVP stimulation of I_eq in cultured cells is mediated through a mechanism other than activation of Na⁺ luminal channels. For example, Canessa and Schafer (8) observed an amiloride-insensitive I_eq that was inhibited by ouabain. Using primary cultures of rabbit CCD, Nagy et al. (33) reported that AVP activated a DIDS-insensitive Cl conductance through a cAMP pathway. In fact, Letz and Korbmacher (29) reported recently that cAMP stimulated cystic fibrosis transmembrane conductance regulator-like channels in M-1 cells. In our study, AVP transiently stimulated I_eq even after inhibiting luminal Na⁺ channels with amiloride. The AVP-induced increase in I_eq in M-1 cells is not secondary to activation of amiloride-sensitive Na⁺ channels and may represent activation of Cl channels similar to other studies (29, 32, 33).

The inhibitory response to some hormones has been linked to an increase in [Ca²⁺]_i (5). This increase in [Ca²⁺]_i has been suggested to be secondary to an initial increase of apical Na⁺ entry, which in turn leads to an increase in [Ca²⁺]_i via basolateral Na⁺/Ca²⁺ exchange (6, 19). In our experiments increasing [Ca²⁺]_i by thapsigargin did not affect I_eq or V_te. Rouch et al. (40) reported similar lack of inhibition of salt and water transport by thapsigargin and ionomycin in rat CCD, although both increased [Ca²⁺]_i. In this context, one possibility is that stimulation by mineralocorticoids may decrease the effectiveness of some inhibitory hormones, particularly those linked to an increase in [Ca²⁺]_i. In agreement with this hypothesis, Frindt et al. (18, 19) have postulated that mineralocorticoid-induced increases in apical Na⁺ conductance are sufficiently high that the apical membrane is not rate limiting any more. In this context, the second messenger pathway of the response to hormones in these cells deserves further investigation.

An endogenous protease activity does appear to activate transport in M-1 cells, as has been recently reported in A-6 cells by Vallet et al. (49). Our studies on M-1 cells indicate that the responses to both aprotinin and trypsin are consistent with an endogenous protease activation. Aprotinin, a protease inhibitor, decreases transport after several hours but not acutely. Trypsin can then activate transport in these aprotinin-treated cells. However, in cells with no aprotinin in which transport has presumably been activated by endogenous protease activity, trypsin has no effect. Although the present studies extend protease activation of transport to mammalian cells, the specific mechanisms and regulation of this phenomenon remain to be elucidated.

In summary, the present study demonstrates that M-1 cells exhibit many properties characteristic of mammalian CCD and therefore serve as a suitable (but limited) in vitro mammalian model for studying CCD. In regard to hormonal regulation, either Dex or Aldo is able to enhance Na⁺ transport in these cells. Stimulation by Dex is increased in the presence of Aldo, but a component of the effect of Dex appears to be mediated through activation of mineralocorticoid receptors. Acute treatments with EGF and PMA do not affect Na⁺ transport, but AVP transiently activates I_eq via an amiloride-insensitive pathway, likely Cl channels. Also, transport is stimulated by an endogenous protease activity that can be inhibited by aprotinin and reactivated by apical trypsin.