ROS production as a common mechanism of ENaC regulation by EGF, insulin, and IGF-1

Cell culture and lentiviral transduction.

Immortalized mouse cortical collecting duct (mpkCCD_c14) cells were kindly provided by Dr. Alain Vandewalle (INSERM, Paris, France) and described previously (5). M-1 cells (58) were obtained from ATCC (Manassas, VA) and maintained under standard culture conditions (DMEM, 10% FBS, 1× Pen/Strep, 37°C, and 5% CO₂-95% air atmosphere). Growth medium for mpkCCD_c14 principal cells was composed of equal volumes of DMEM and Ham's F-12, 50 nM dexamethasone, 1 nM triiodothyronine, 2% FBS, and 1%/vol Pen/Strep. The mpkCCD_c14 cells were also grown in a 5% CO₂-95% air atmosphere incubator at 37°C. For electrophysiological experiments, M-1 and mpkCCD_c14 cells were grown in defined medium on permeable supports (Costar Transwells; 0.4-μm pore, 24-mm diameter) as described previously (5, 27, 47, 57). The cells were seeded onto permeable supports at a density of 0.2–0.3 × 10⁶ cells/filter and kept on filter supports for at least 7 days. The medium was changed every other day allowing cells to polarize and form monolayer with high resistance and avid Na⁺ absorption. Eighteen hours before use in any experiment, the medium of the cells incubated on filter supports was replaced with minimal medium that contained only DMEM, Ham's F-12 and antibiotics.

Rac1 short hairpin (sh)RNA lentiviral particles (sc-36352-V; Santa Cruz Biotechnology, Santa Cruz, CA) were used for the inhibition of Rac1 expression in M-1 cells (27). Control shRNA lentiviral particles (sc-108080; Santa Cruz Biotechnology) were used to confirm the selectivity of the lentiviruses. After transduction with Santa Cruz transfection reagent (sc-108061), a stable cell line expressing the shRNA was isolated via puromycin selection (27).

Electrophysiological studies.

Equivalent short circuit current (I_eq) across the mpkCCD_c14 cell monolayer was calculated using Ohm's law as the quotient of transepithelial voltage (V_T) to transepithelial electrical resistance (R_T) under open circuit conditions using a Millicel Electrical Resistance System with dual Ag/AgCl pellet electrodes (Millipore) to measure V_T and R_T. To determine the amiloride-sensitive current, 10 μM amiloride were added to the apical cell surface at the end of experiments. EGF (E4127), insulin (I6634), and apocynin (W508454) were obtained from Sigma-Aldrich (St. Louis, MO). IGF-1 (50-990-805) was received from Fisher Scientific. ML171 (492002) was obtained from EMD Millipore.

Western blot analysis.

mpkCCD_c14 were washed twice in PBS and lysed in a buffer as described previously (46). Equal amounts of proteins were separated by using 7.5% SDS-PAGE and were electrophoretically transferred onto nitrocellulose membrane (Millipore, MA), immunoblotted with the appropriate antibody, and visualized by ECL (Amersham Biosciences). Antibodies for the NADPH oxidase subunits were from Sigma-Aldrich (NOX1; cat. no. SAB4200097), BD Biosciences (NOX2; cat. no. 611414), Novus Biologicals (NOX4; NB110-58851), Santa Cruz (p22-phox; sc-271262), and Millipore (p67-phox and Rac1; 07-502 and 05-389, respectively).

Total ROS detection.

mpkCCD_c14 cells were maintained with standard culture conditions. The day before the experiment, the cells were seeded onto glass slides to ensure 50–70% confluence on the day of the experiment. Cells were incubated with apocynin (500 μM) overnight. For the measurements of H₂O₂ production, cells were loaded with 4 μg/ml of CM-H₂DCF-DA (2′,7′-dichlorofluorescein acetate, DCF; Enzo Life Sciences) dissolved in PBS and incubated for 1 h in the dark at room temperature. Incubation times and concentrations of EGF/IGF-1/insulin were carefully selected from transepithelial current measurements in response to EGF/IGF-1/insulin and published literature (8, 31, 37, 38, 57). We have selected approximately half maximum concentrations and times required to acutely activate ENaC-mediated transport. Vehicle or experimental agents were added to the cells 1 h (EGF, 50 ng/ml), 20 min (IGF, 100 ng/ml), or 10 min (insulin, 100 nM) before the end of incubation. A ROS inducer (200 μM pyocyanin; ENZ-53001-C001; Enzo Life Sciences) and inhibitor (5 mM N-acetyl-l-cysteine; ALX-105-005-G001; Enzo Life Sciences) were added 30 min before the end of the incubation to demonstrate that the cells provided the predicted response to these drugs. After incubation with the dye and the reagent, the cells were washed with 1× PBS twice, and then the glass slides were immediately overlaid with a coverslip and observed under the fluorescent microscope. For the staining with DCF, the images of the mpkCCD_c14 cells were taken with the same exposure time with a fluorescence microscope Eclipse E-600 (Nikon Instruments, Tokyo, Japan) using a ×20 objective. The signal was excited at 490 nm, and emission was collected at 525 nm. Fluorescent images were processed with the Metamorph 7.5 software (Molecular Devices) and open source software ImageJ 1.42q (http://rsb.info.nih.gov/ij). Twenty or more regions were chosen on each coverslip; at least two coverslips were treated with an agent or vehicle. Three sets of similar experiments were conducted for each agent tested, and background levels were subtracted. Smaller regions on the coverslips were chosen rather than the whole coverslip to avoid the cells with low fluorescence levels and to record the CM-H₂DCF-DA signal from the cells fluorescing above the defined threshold that would amount to 50–55% of the highest fluorescence level detected on the slide (signal levels were detected with ImageJ software; no postprocessing of the images intensity/brightness/contrast was performed).

Staining with rhodamin-phalloidin.

M-1 cells were seeded on 12-mm round coverslips coated with poly-d-lysine (BD Biosciences) 24 h before experiments and grown to confluence of ∼70%. Before staining, the cells were washed with PBS and fixed with 1.2% paraformaldehyde (Sigma-Aldrich). Then, cells were treated with 0.1% Tryton X-100 for 5 min, washed with PBS and incubated with 2 μM of rhodamine-phalloidin (Life Technologies) for 15 min at 37°C. Stained cells were mounted on glass slides with Vectashield mounting medium (Vector Laboratories). M-1 cells stained with rhodamine-phalloidin were visualized with an Eclipse E-600 microscope, (Nikon Instruments, Tokyo, Japan), using a ×60 oil objective equipped with a CCD camera (Princeton Instruments). The signal was excited at 546 nm, and emission was collected at 590 nm.

Electron microscopy.

For the electron microscopy experiments, wild-type and Rac1 knockdown M-1 cells were fixed in 2% glutaraldehyde buffered in 0.1 M cacodylate (pH 7.4) for 1 h at room temperature and washed three times for 5 min in 0.1 M cacodylate buffer. After that, the cells were postfixed in 1% osmium tetroxide reduced with 1.25% potassium ferricyanide on ice for 2 h. Then, the specimen was washed three times for 5 min in distilled water, dehydrated through graded methanol, and embedded in EMBed 812 epoxy resin. After embedding, ultrathin sections were cut on a RMC Powertome and stained with uranyl acetate and Reynolds lead citrate. Sections were viewed in a Hitachi HS600 transmission electron microscope (Hitachi High Technologies), and images were recorded via an AMT 1K digital imaging camera.

Statistics.

All summarized data are reported as means ± SE. Data are compared using either the Student's (two-tailed) t-test or a one-way ANOVA, and P < 0.05 is considered significant.

NADPH oxidase subunits are expressed in the CCD principal cells.

NADPH oxidase is comprised of the membrane-bound proteins NOX2 (gp91phox) and p22^phox, as well as several regulatory proteins termed p40^phox, p47 ^phox, and p67 ^phox, and of the small GTPase Rac (30). In addition to NOX2, other NADPH oxidase subunits, such as NOX1, NOX3, NOX4, NOX5, and DUOX1 and DUOX2 have been identified (17, 30). Mouse polarized CCD mpkCCD_c14 cells used in this study were tested for the expression of the NADPH oxidase complex subunits. As shown by Western blotting experiments (Fig. 1), mpkCCD_c14 cells express all tested proteins. Particularly, NOX1, NOX2, and NOX4 (band observed at ∼31 kDa may represent a reported splice isoform; Ref. 23), as well as p22phox, p67phox, and Rac1 were identified (Fig. 1).

EGF, insulin, and IGF-1 stimulate ROS production in mpkCCD_c14 cells.

Fluorescent microscopy measurements were employed to study changes in the production of ROS in the mpkCCD_c14 cells in response to acute application of EGF, insulin, and IGF-1. The CM-H₂DCF-DA reagent (also referred to as DCF; Ref. 16), a cell-permeable probe that fluoresces upon oxidation, was used to measure the relative production of ROS in vehicle and EGF-, insulin-, and IGF-1-treated cells. Figures 2, 3, and 4 demonstrate that treatment with 50 ng/ml of EGF, 100 nM of insulin, and 100 ng/ml of IGF-1 for 1 h, 10 min, and 20 min, respectively, results in a significant increase of the DCF fluorescence compared with vehicle. Ten minutes for insulin treatment were selected based on the published data (38), which revealed that the maximum signal for hydrogen peroxide production in A6 cells monolayers stimulated by insulin was observed after 5 min of treatment and sustained for at least 30 min. For IGF-1 and EGF stimulation, preliminary experiments were carried out to estimate the time point that would show the initial increase in DCF fluorescence (data not shown).

Treatment with EGF, insulin, and IGF-1 resulted in 1.27 ± 0.04, 1.23 ± 0.04, and 1.25 ± 0.05-fold increases in fluorescence levels for corresponding ligands, respectively, compared with 1 in vehicle-treated cells (Figs. 2–4, summary graphs). For these experiments, all data were normalized to the fluorescence levels detected for cells treated with vehicle and determined consecutively in each group. N-acetyl-l-cysteine and pyocyanin were used as negative (not shown) and positive controls of changes in ROS production and demonstrated almost no fluorescent signal and elevation of DCF fluorescence, respectively, in all studied groups.

To ensure that NADPH oxidase is involved in the mechanism of EGF-, insulin-, and IGF-1-mediated ROS production, we used the NADPH oxidase inhibitor apocynin (16, 17). Apocynin inhibits the NADPH oxidase complex by preventing the transfer of the soluble cytosolic subunits to the membrane complex. Overnight treatment with 500 μM apocynin resulted in impaired relative fluorescence signal that was not significantly increased by treatment with EGF, insulin, or IGF-1 (Figs. 2–4). Thus these data demonstrate that the NADPH oxidase complex is required for EGF, insulin, and IGF-1 to cause acute intracellular ROS production in mpkCCD_c14 cells.

Effects of EGF, insulin, and IGF on transepithelial currents are abolished by the inhibition of the NADPH oxidase complex.

To further test whether an increase in ROS production and upregulation of ENaC activity in response to EGF-, insulin-, and IGF-1 intercross, we performed equivalent transepithelial current (I_eq) measurements across the mpkCCD_c14 cell monolayer. Amiloride (10 μM) was added to the apical membrane at the end of experiments to reveal amiloride-sensitive current, which most likely represents ENaC-mediated Na⁺ absorption through the monolayer. The experimental design included treatment with vehicle and incubation with 10 ng/ml of EGF, 20 nM of insulin, and 100 ng/ml of IGF-1 either alone or against the background of overnight apocynin treatment. Overnight pretreatment with 500 μM apocynin was only used as a control treatment that was repeated independently in every experiment. EGF, insulin, and IGF-1 (all applied basolaterally) significantly increased the amiloride-sensitive transepithelial current through the mpkCCD_c14 cell monolayer (Fig. 5). However, pretreatment with apocynin significantly blunted the response of the amiloride-sensitive currents to the stimulation with EGF, insulin, or IGF-1 (Fig. 5, A–C).

Inhibition of the NADPH oxidases with ML171 decreases transepithelial currents evoked by stimulation with EGF.

Apocynin should be used with caution since it was described predominantly as an antioxidant in endothelial cells and vascular smooth muscle cells (25). Therefore, considering that apocynin may be acting predominantly as a scavenger, thereby circumventing NADPH oxidase activity, we used another structurally distinct pharmacological inhibitor of NADPH oxidases, phenothiazine compound ML171 (17, 19, 64). ML171 does not scavenge ROS, this being an important advantage of this drug compared with apocynin used in Figs. 2–5. ML171 (20 μM) significantly blunted the transepithelial currents through mpkCCD_c14 cell monolayers stimulated by 10 ng/ml EGF (Fig. 6A). ML171 alone did not have any significant effect on the current. Similarly, 20 μM ML171 diminished the IGF-1- and insulin-mediated increase in transepithelial current in the mpkCCD_c14 cells (data not shown). ML-171 has the IC₅₀ of 0.25 μM for NOX1 and the IC₅₀ >3 μM for other NADPH oxidases (17, 19). Concentration of ML171 that was reported to be selectively inhibiting for the NOX1 showed slight inhibition of the Na⁺ transport that was not statistically significant as illustrated by Fig. 6B, which shows a summary of experiments with mpkCCD_c14 cells treated with 0.5 μM of ML171 against the background of Na⁺ transport stimulated with 20 nM of insulin.

Stimulation of the ROS production by xanthine-xanthine oxidase reaction disrupts actin cytoskeleton and increases equivalent short-circuit currents in the mpkCCD_c14 cells.

We addressed here a question whether elevated production of the ROS affects the transepithelial current in these cells. I_eq measurements were performed on the monolayers of the mpkCCD_c14 cells seeded on permeable supports in the presence or absence of xanthine and/or xanthine oxidase. Xanthine oxidase catalyzes xanthine to uric acid and in an adjacent reaction reduces oxygen to superoxide ion that further spontaneously dismutes to H₂O₂. Figure 7A illustrates significant increase in equivalent short circuit current across the mpkCCD_c14 monolayers in presence of xanthine (250 μM) and xanthine oxidase (15 mU/ml). Neither xanthine nor xanthine oxidase alone had any effect on the transepithelial current. In the adjacent experiments, we tested the effects of the xanthine-xanthine oxidase reaction on the structure of the actin microfilaments. F-actin was visualized by staining with rhodamine-phalloidin. As shown in the representative images in Fig. 7B, microfilaments appear significantly destructed after 2 h of treatment with xanthine-xanthine oxidase.

Knockdown of Rac1 decreases ROS production and precludes EGF-mediated increase of transepithelial current in M-1 cells.

We have recently demonstrated an essential role for small GTPase Rac1, which represents an essential subunit of the NADPH oxidase complex, in the regulation of ENaC activity (27). To test a role of the NADPH complex in the regulation of transepithelial current by EGF, we have used a previously generated M-1 stable cell line with knockdown of Rac1, in which Rac1 expression was up to 75% decreased as previously shown by Western blotting and densitometry assays (27). Efficient downregulation of Rac1 expression in M-1 cells stably expressing anti-Rac1 shRNA was verified by Western blot analysis (Fig. 8A). As we have previously shown with transepithelial current measurements and patch clamp analysis, Rac1 knockdown significantly decreased basal transepithelial current and ENaC activity in the M-1 cells (27). Many of the Rac1 functions are mediated through regulation of actin organization (12). To test whether knockdown of Rac1 destroyed the actin cytoskeleton integrity, we performed additional experiments. Staining with rhodamine-phalloidin was performed to trace any changes in the spatial organization of the actin cytoskeleton that could be due to the knockdown of this small GTPase. However, as shown on Fig. 8B, knockdown of Rac1 did not alter the distribution of the actin cytoskeleton within the cells. To further confirm that knockdown of Rac1 did not affect the morphology of the cells and their polarity, electron microscopy experiments were performed. As demonstrated on the representative electron micrographs on Fig. 8C, Rac1 knockdown did not elicit any changes in the M-1 cell phenotype compared with control. For instance, close-up images demonstrate that tight junctions as well as monolayer integrity remain unchanged.

As we previously demonstrated, Rac1 knockdown cells exhibited diminished equivalent short circuit currents compared with control (scrambled shRNA) cells (27). We further tested whether Rac1 knockdown cells are stimulated by EGF. As shown by equivalent transepithelial current measurements, Rac1 depletion prevents an elevation in amiloride-sensitive currents in response to 10 ng/ml of EGF that is observed in wild-type cells (Fig. 9A). Therefore, Rac1 is required for acute stimulation of transepithelial currents by EGF. Corresponding transepithelial resistances (R_T) for the represented I_eq experiment are reported in Fig. 9B. Interestingly, treatment with EGF did not decrease R_T, which might suggest that EGF is affecting other components of the ion transport process in these cells. For instance, future studies are required to test whether EGF affects paracellular transport. While the distal nephron, including CCD, is considered a tight epithelium with high transepithelial resistance and low passive ion permeability (56), the paracellular pathway in the collecting duct might also play a role in transepithelial absorption.