Ras Activates the Epithelial Na+ Channel through Phosphoinositide 3-OH Kinase Signaling*

Aldosterone induces expression and activation of the GTP-dependent signaling switch K-Ras. This small monomeric G protein is both necessary and sufficient for activation of the epithelial Na+ channel (ENaC). The mechanism by which K-Ras enhances ENaC activity, however, is uncertain. We demonstrate here that K-Ras activates human ENaC reconstituted in Chinese hamster ovary cells in a GTP-dependent manner. K-Ras influences ENaC activity most likely by affecting open probability. Inhibition of phosphoinositide 3-OH kinase (PI3K) abolished K-Ras actions on ENaC. In contrast, inhibition of other K-Ras effector cascades, including the MAPK and Ral/Rac/Rho cascades, did not affect K-Ras actions on ENaC. Activation of ENaC by K-Ras, moreover, was sensitive to co-expression of dominant negative p85PI3K. The G12:C40 effector-specific double mutant of Ras, which preferentially activates PI3K, enhanced ENaC activity in a manner sensitive to inhibition of PI3K. Other effector-specific mutants preferentially activating MAPK and RalGDS signaling had no effect. Constitutively active PI3K activated ENaC independent of K-Ras with the effects of PI3K and K-Ras on ENaC not being additive. We conclude that K-Ras activates ENaC via the PI3K cascade.

Small monomeric G proteins, including Ras, are GTP-dependent signaling switches that control cell growth, proliferation, and differentiation, as well as playing important signaling roles in differentiated cells (1)(2)(3). There are four homologous Ras proteins, K-RasA, K-RasB, H-Ras, and N-Ras, capable of activating similar downstream effectors with Ras having three primary first effectors, Raf, RalGDS, and PI3K, 1 initiating the MAPK 1/2, Ral/Rac/Rho, and PI3K signaling cascades, respectively. Signaling mediated by Ras and other Ras-like small G proteins can impinge upon the activity of membrane-resident ion channels. For instance, Ras signaling leads to phosphorylation of K ir causing decreased channel activity (4); small G proteins within the RGK family directly interact with the ␤-subunit of L-type Ca 2ϩ channels promoting retrieval from the plasma membrane and decreased channel activity (5-7); Rac and Rho have opposing actions on ether-a-go-go-related gene K ϩ channels with the prior rapidly activating the channel and the latter quickly decreasing channel activity (8); similarly, Rap and Ras have opposing actions on muscarinic K ϩ channels (9); and Ras via MAPK signaling decreases the activity of IRK1 by promoting retrieval of the channel from the plasma membrane (10) but increases the activity of T-type Ca 2ϩ channels (11). Thus, a wide variety of ion channels are final effectors for Ras and other small G proteins.
Activity of the epithelial Na ϩ channel (ENaC) is rate-limiting for electrogenic Na ϩ transport across electrically tight epithelia, such as that lining the distal colon and renal nephron (12)(13)(14). Systemic Na ϩ levels are maintained in balance through a classic negative feedback pathway involving activation of ENaC by the steroid hormone aldosterone. Similar to other steroids, aldosterone affects gene expression to influence the activity of its target proteins; however, aldosterone-dependent increases in ENaC activity precede effects on channel expression leading to the proposal that aldosterone influences ENaC activity by promoting the expression of mobile signaling molecules capable of transducing information from the nucleus to existing channels. K-RasA is regulated by aldosterone at the level of transcription (15) with aldosterone promoting MAPK signaling in epithelia via induction of this small G protein (16,17). Moreover, K-RasA is necessary and sufficient for ENaC activation in some models (18,19), which makes this small G protein an attractive candidate to transduce aldosterone actions from the nucleus to the channel. However, the direct role of Ras in regulation of ENaC is unclear with its mechanism of action and the cell signal transduction pathways involving Ras actions on ENaC being uncertain.
Insulin, in addition to aldosterone, increases Na ϩ reabsorption with PI3K being necessary to both aldosterone and insulin actions on Na ϩ transport (20 -24). Moreover, aldosterone increases the levels of the phospholipid products of PI3K in epithelial cells, and these phospholipids have recently been shown to directly enhance ENaC activity (20,25,26). Because PI3K is a first effector of Ras, we tested whether activation of ENaC in response to K-Ras was mediated by signaling through this lipid kinase. The current results are consistent with K-Ras activating ENaC via PI3K.

EXPERIMENTAL PROCEDURES
Materials-All chemicals were of reagent grade and were purchased from Calbiochem, BioMol (Plymouth Meeting, PA), or Sigma, unless noted otherwise. The mammalian expression vectors encoding ENaC subunit cDNAs have been described previously (24,26,27). Expression vectors encoding hemagglutinin-tagged wild type, constitutively active (G12V), and dominant negative (S17N) K-Ras were from the Guthrie Research Institute. Expression vectors encoding Ras effector-specific double mutants (G12:C40, G12:E38, G12G37) as well as dominant negative PI3K (p85 ⌬iSH2-C) were from J. Downward and have been described previously (3). The expression vector encoding constitutively active PI3K (pUSE-amp-p110␣) was from Upstate Biotech (Lake Placid, NY). This plasmid encodes p110␣ fused to an amino-terminal Src myristoylation sequence, which localizes PI3K to the plasma membrane, where it is then constitutively active (28). The plasmid encoding the CFP-tagged membrane marker (CFP-M) was from Clontech (Palo Alto, CA). All materials used in Western blot analysis were from Bio-Rad. Sulfo-NHS-LC-biotin and streptavidin-agarose were from Pierce. The mouse monoclonal anti-Myc antibody was from Clontech, and the rabbit anti-Fra-2 antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-mouse and rabbit horseradish peroxidase-conjugated secondary antibodies were from Kirkegaard-Perry Laboratories (Gaithersburg, MD). ECL reagents were from PerkinElmer Life Sciences. CHO cells were maintained with standard culture conditions (Dulbecco's modified Eagle's medium ϩ 10% fetal bovine serum, 37°C, 5% CO 2 ) and transfected using the Polyfect reagent (Qiagen, Valencia, CA) as described previously (24,26,27).
Electrophysiology-Whole cell macroscopic current recordings of ENaC reconstituted in CHO cells were made under voltage clamp conditions using standard methods (24,26,27). Current through ENaC was the inward, amiloride-sensitive Na ϩ current with a bath solution of (in mM) 160 NaCl, 1 CaCl 2 , 2 MgCl 2 , and 10 HEPES (pH 7.4) and a pipette solution of (in mM) 120 CsCl, 5 NaCl, 5 EGTA, 2 MgCl 2 , 2 ATP, 0.1 GTP, 10 HEPES (pH 7.4). Current recordings were acquired with an Axopatch 200B (Axon Instruments, Union City, CA) interfaced via a Digidata 1322A (Axon Instruments) to a personal computer running the pClamp 9 suite of software (Axon Instruments). All currents were filtered at 1 kHz. Both a family of test pulses (500 ms each) stepping by 20 mV increments from a holding potential of 30 mV to 60 mV to Ϫ100 mV and voltage ramps (300 ms) from 60 mV to Ϫ100 mV were used to generate current-voltage (I-V) relations and to measure ENaC activity at Ϫ80 mV. Whole cell capacitance was routinely compensated and was ϳ9pF for CHO cells. Series resistances, on average 2-5 megaohms were also compensated.
Membrane Labeling Experiments-Western blot analysis was performed using standard procedures described previously (16,18,27,29,30). In brief, cells were lysed in gentle lysis buffer (1.0% Nonidet P-40), cleared, normalized for total protein concentration, suspended in Laemmli sample buffer and 20 mM dithiothreitol, heated at 85°C for 10 min, run on 7.5% polyacrylamide gels (100 g of total protein/well) in the presence of SDS, transferred to nitrocellulose, and probed with antibody in Tris-buffered saline supplemented with 5% dried milk (Nestle, Solon, OH) and 0.1% Tween 20.
Membrane labeling experiments closely followed those described previously (27). In brief, CHO cells were transfected with Myc-tagged ENaC in the presence and absence of K-Ras. Forty-eight h after transfection cells were washed three times with ice-cold phosphate-buffered saline (pH 8.0) and subsequently incubated with 1 mM sulfo-NHS-LCbiotin (in phosphate-buffered saline, pH 8.0) for 30 min at 4°C in the dark. Cells were washed three times with ice-cold phosphate-buffered saline and extracted in gentle lysis buffer (1.0% Nonidet P-40). Preequilibrated streptavidin-agarose beads were agitated overnight at 4°C with 50 g of total protein. Agarose beads were then washed six times with gentle lysis buffer and subsequently resuspended in Laemmli sample buffer and 20 mM dithiothreitol, heated at 85°C for 10 min, run on 7.5% polyacrylamide gels in the presence of SDS, transferred to nitrocellulose, and probed with antibody.
Evanescent Field (EF) Fluorescence Microscopy-To determine whether K-Ras co-expression affected ENaC membrane levels, we selectively illuminated the plasma membrane with EF microscopy as described previously (26,27). In brief, EF microscopy was performed using an inverted TE2000 microscope with through-the-lens fluorescence imaging. Evanescent field illumination was generated by total internal reflection after the light beam struck the interface between the glass coverslip and cellular plasma membrane at a glancing angle. Samples were viewed through a Plan Apo total internal reflection 60ϫ oil immersion, high resolution (1.45 numerical aperture) objective (Nikon). Total internal reflection generates an EF that declines exponentially with increasing distance from the interface between the cover glass and plasma membrane illuminating only a small optical slice of the cell (ϳ200 nm) including the plasma membrane. Thus, with total internal reflection only fluorophores in the plasma membrane and its immediate vicinity contribute to emission, whereas those deeper in the cell do not (26). An argon laser with excitation filters of 458 (Ϯ5) and 488 (Ϯ5) nm was used to excite a membrane marker (CFP-M) and NH 2 -YFP-tagged ␣ENaC (expressed with untagged ␤and ␥ENaC), respectively. Tagged ENaC constructs were functional. 2 Emissions from the CFP and YFP fluorophores subsequently passed through 480 nm (Ϯ15; 465 nm dichroic) and 535 nm (Ϯ25; 505 nm dichroic) single pass filters, respectively. Images were collected and processed with a chargecoupled device camera interfaced to a personal computer running Metamorph software. Integrated intensity levels for membrane YFP-tagged ENaC were normalized to internal CFP-M controls.
Statistics-All patch clamp data are presented as mean Ϯ S.E. Paired and unpaired data were compared using appropriate t tests, with p Յ 0.05 considered significant.

K-Ras Increases ENaC Activity in a GTP-dependent Man-
ner-To investigate the actions of K-Ras on ENaC, we reconstituted the channel in CHO cells in the absence and presence of co-expressed wild type, constitutively active (G12V) and dominant negative (S17N) K-Ras. Human ENaC was reconstituted by co-expressing ␣, ␤ and ␥ channel subunits together. Fig. 1A shows typical currents from a voltage clamp experiment performed on a CHO cell containing human ENaC before and after treatment with amiloride. Amiloride is an open channel blocker of ENaC (14). Currents were elicited by applying test pulses from 60 to Ϫ100 mV with 20-mV steps. Overexpression of ENaC resulted in robust amiloride-sensitive inward Na ϩ currents that were not present in untransfected cells (not shown, see Refs. 24, 26, and 27). Fig. 1B shows ENaC currents before and after treatment with amiloride in a cell expressing the channel alone (left) and in a cell expressing both the channel and constitutively active K-Ras (right). Currents were elicited by voltage ramping from 60 mV down to Ϫ100 mV (holding potential ϭ 30 mV). As summarized in Fig. 1C, co-expression of wild type and constitutively active but not dominant negative K-Ras significantly increased ENaC activity. Overexpression of K-Ras had no effect on cell capacitance as shown in Fig. 1D. We interpret these results as showing that K-Ras enhances ENaC activity but that this channel does have some basal activity in CHO cells that is independent of K-Ras signaling.
We next tested whether K-Ras actions on ENaC were dependent on GTP activation of this small G protein. Fig. 2A shows current through ENaC in the presence of constitutively active K-Ras elicited by a train of voltage ramps applied every 5 s over the course of several min. For this experiment, 2.0 mM GDP␤S replaced the GTP in typical pipette solutions. As the cytosol was dialyzed with GDP␤S, ENaC activity decreased over time indicating that GTP was necessary for K-Ras to enhance channel activity. Run-down of ENaC activity in whole cell voltage clamp experiments was never observed (not shown, see Refs. 26 and 27), and thus the decrease in activity observed in Fig. 2A resulted from GDP␤S-competing GTP. Fig. 2B shows an overlay of ENaC currents before and after dialyzing GDP␤S and the subsequent application of amiloride. As demonstrated by the summary graph in Fig. 2C, GTP is necessary for K-Ras activation of ENaC. Interestingly but not further pursued in the current study, we saw a modest effect of GDP␤S on the K-Ras-independent basal activity of ENaC suggesting that other endogenous GTP-dependent proteins also impinge upon activity of this channel.
K-Ras Activation of ENaC Is Mediated by PI3K Signaling-We next asked whether K-Ras activated ENaC through signaling via one of its downstream effector cascades, the MAPK, PI3K, and Ral/Rac/Rho cascades. For this experiment pipette GTP was replaced with GDP␤S, and amiloride was applied to the extracellular bath solution at the end of the experiment. B, an overlay of three macroscopic currents from the experiment described in A before and after dialyzing GDP␤S and then applying amiloride to the extracellular bath solution. C, summary graph of the amiloride-sensitive current density for ENaC in the absence and presence of wild type, constitutively active, and dominant negative K-Ras before and after dialyzing intracellular GTP with GDP␤S. Also shown are the current densities for ENaC in the absence and presence of constitutively active K-Ras with GTP in the pipette. *, p Ͻ 0.05 versus beginning current densities before dialysis of GDP␤S. hENaC, human ENaC. , and PI3K (wortmannin, 200 nM) signaling on K-Ras activation of ENaC. Coexpression of K-Ras with ENaC significantly increased channel current density from 157 Ϯ 10 to 487 Ϯ 34 pA/pF. Although pretreatment with the PI3K inhibitor wortmannin significantly decreased K-Ras-dependent ENaC activity to 46 Ϯ 20 pA/pF, the inhibitors of other K-Ras effector cascades had no effect. We conclude from these results that PI3K is necessary to K-Ras-dependent activation of ENaC suggesting that this small G protein activates the channel via PI3K signaling. Because PI3K is a known first effector of K-Ras (3) and we have reported previously that K-Ras interacts with PI3K in an aldosterone-sensitive manner (24), we position this lipid kinase downstream of K-Ras in the aldosterone to ENaC transduction cascade.
Results shown in Fig. 4 further supported the idea that PI3K is integral to K-Ras activation of ENaC and that the lipid kinase is positioned between the small G protein and the channel. Shown in Fig. 4A are representative current traces elicited by voltage ramps from cells expressing ENaC plus effectorspecific mutants of Ras that are capable of interacting with and thus initiating signaling through only a single first effector. The Ras effector-specific mutant G12:E38 activates only c-Raf kinase and thus, MAPK signaling; the G12:G37 mutant only activates RalGDS and thus Ral/Rac/Rho signaling; and the G12:C40 mutant only activates PI3K and its dependent signal transduction cascade (3). Only this latter mutant enhanced ENaC activity. Also as shown by a representative current trace in Fig. 4A, enhancement of activity by the C40 mutant was sensitive, as expected, to inhibition of PI3K with wortmannin. The summary graph in Fig. 4B demonstrates that ENaC activity in the presence of the constitutively active Ras effectorspecific G12:C40 mutant was not different from that observed with constitutively active Ras, with both being greater than ENaC alone and ENaC plus the other first effector-specific Ras mutants. The summary graph in Fig. 4C demonstrates that wortmannin significantly decreased ENaC activity in the presence of Ras G12:C40 to a level similar to that observed for expression of the channel alone.
Results shown in Fig. 5 are also consistent with PI3K being necessary for K-Ras activation of ENaC. Fig. 5A shows a summary graph of ENaC activity when channel subunits are expressed alone and with K-Ras in the absence and presence of co-expression of dominant negative PI3K (p85 ⌬iSH2-C) (3). Inclusion of dominant negative PI3K with K-Ras significantly decreased ENaC activity from 445 Ϯ 12 to 317 Ϯ 38 pA/pF. Fig.  5, B-D, shows population histograms describing ENaC activity in the absence of K-Ras (Fig. 5B), and presence of co-expressed K-Ras (Fig. 5C) and co-expressed K-Ras plus dominant negative PI3K (Fig. 5D). The populations describing both ENaC alone and ENaC in the presence of K-Ras were normally distributed around a single mean (ENaC alone ϭ 156 Ϯ 11 pA/pF). In contrast, two ENaC populations were observed when the channel was co-expressed with K-Ras plus dominant negative PI3K. These populations were normally distributed around means of 155 Ϯ 14 (n ϭ 13) and 479 Ϯ 36 (n ϭ 13) pA/pF. Two mechanisms may explain the appearance of the population with the lower activity when ENaC is co-expressed with both K-Ras and dominant negative PI3K. It may result from a titration effect or from the dominant negative PI3K uncoupling K-Ras signaling to the channel. Results shown in Figs. 3 and 4 support the latter possibility. Moreover, ENaC activity in the presence of co-expressed K-Ras is normally distributed, indicating that in this population there was no titration effect. Finally, when ENaC was co-expressed with K-Ras plus another construct that did not impact PI3K signaling, we saw a single distribution with a mean current density of 433 Ϯ 53 pA/pF (n ϭ 15, not shown), which was similar to that observed with K-Ras alone, excluding a titration effect. Thus, we interpret the results in Fig. 5D as demonstrating that dominant negative PI3K uncouples K-Ras to ENaC signaling. If this interpretation is correct, then there must be a threshold effect of PI3K on the channel with activated PI3K resulting in channels with high activity and inactivated PI3K leading to channels even in the presence of K-Ras with basal activity. This possibility is consistent with our previous findings that the phospholipid products of PI3K directly affect ENaC by stabilizing channel gating transitions (26).
We predicted that if K-Ras activated ENaC via PI3K, then expression of constitutively active PI3K alone with ENaC should activate the channel, and the effects of co-expressing PI3K with K-Ras on ENaC should not be additive. Indeed, co-expression of constitutively active PI3K (myr-PI 3-K p110␣), which has a c-Src myristoylation sequence localizing it to the plasma membrane (28), with ENaC significantly increased channel activity from 139 Ϯ 17 to 464 Ϯ 51 pA/pF (n ϭ 16; data not shown). Channel activity, in addition, in the presence of co-expressed constitutively active PI3K plus K-Ras G12V was 570 Ϯ 93 pF/pA (n ϭ 10), which was not different compared with K-Ras G12V (571 Ϯ 139, n ϭ 5) or constitutively active PI3K alone. Both of these results are consistent with the idea that K-Ras and PI3K signal to ENaC through the same transduction cascade.
K-Ras Enhances ENaC Activity Independent of Effects on Membrane Levels of the Channel-We next tested whether co-expression of K-Ras with ENaC activated the channel by increasing the number of channels in the plasma membrane. The rationale for testing this hypothesis was that PI3K signaling ultimately through Sgk promotes retention of ENaC in the plasma membrane (23,(32)(33)(34), phospholipids produced by PI3K directly increase ENaC open probability (26) and counter channel run-down in excised patches (25), co-expression of constitutively active K-Ras with ENaC in Xenopus laevis oocytes most likely increases channel open probability (19), and K-Ras is necessary for aldosterone to increase ENaC open probability in A6 epithelial cells (18). Shown in Fig. 6A are representative Western blots containing plasma membrane-resident (streptavidin precipitant from cells treated with sulfo-NHS-LC-biotin, top) and total cellular (middle) ENaC when the channel is expressed in the absence and presence of wild type, constitu-tively active, and dominant negative K-Ras. Blots were probed with anti-Myc antibody. For these experiments, all three ENaC subunits were Myc-tagged with ␣and ␤ENaC migrating as overlying ϳ90-kDa bands and ␥ENaC as an ϳ85-kDa band, which is similar to that observed previously for these constructs (27). A loading control (Fig. 6A, bottom, total cellular protein probed with anti-MAPK antibody) is also shown. Shown in Fig. 6B is a similar experiment with the top blot containing whole cell lysate (100 g/well) from control cells (GFP) and cells transfected with ENaC in the absence and presence of co-expression of K-Ras and constitutively active and dominant negative K-Ras. This blot was probed for Myctagged ENaC. Prior to extracting, cell membranes were labeled with biotin. The middle blot contains both the streptavidin precipitant (M, from 100 g of total protein) and the supernatant (S, half of the total supernatant) from each respective lysate. This blot was also probed with anti-Myc antibody. The bottom blot is the middle blot reprobed against a cytosolic control protein (Fra-2) to demonstrate good separation of the membrane from cytosolic fractions. The results shown in Fig. 6 demonstrate that there is no correlation between K-Ras effects on membrane ENaC levels and channel stimulation. Indeed, the relative amounts of ENaC in the membrane normalized to total cellular ENaC pools in the absence and presence of K-Ras and constitutively active and dominant negative K-Ras were 0.20 Ϯ 0.01, 0.23 Ϯ 0.02, 0.23 Ϯ 0.04, and 0.24 Ϯ 0.04, respectively (n ϭ 4).
Shown in Fig. 7A are fluorescence images of cells expressing a membrane marker (CFP-M; pseudo-colored red) alone (top row) and with YFP-tagged ENaC (pseudo-colored green) in the absence (middle row) and presence (bottom row) of constitutively active K-Ras G12V . The summary graph in Fig. 7B reports the effects of constitutively active and dominant negative K-Ras on ENaC levels in the plasma membrane normalized to an internal control (CFP-M). Membrane levels of ENaC in control cells (expressing the channel alone) and in cells cotransfected with K-Ras G12V and K-Ras S17N were 1.40 Ϯ 0.13, 1.50 Ϯ 0.10, and 1.20 Ϯ 0.10, respectively, with membrane ENaC levels not FIG. 4. The PI3K Ras effector-specific mutant activates ENaC. A, overlays of representative macroscopic currents elicited by voltage ramps before and after amiloride (amil) for ENaC plus Ras effector-specific mutants: G12:E38 (c-Raf kinase, MAPK signaling), G12:G37 (RalGDS, Ral/Rac/Rho signaling), and G12:C40 (PI3K). The effects of pretreating with wortmannin were also tested on ENaC activity when the channel was co-expressed with the C40 mutant. Shown here is a representative current overlay before and after amiloride. B, summary graph of current density for ENaC in the absence and presence of constitutively active Ras and the constitutively active effector-specific mutants of Ras. *, p Ͻ 0.05 versus ENaC alone. C, summary graph of the amiloride-sensitive current density for ENaC in the absence and presence of constitutively active PI3K effector-specific C40 Ras mutant with and without pretreatment with wortmannin. *, p Ͻ 0.05 versus ENaC; **, versus ENaC plus Ras G12:C40 . hENaC, human ENaC. different in the presence of constitutively active K-Ras, although activity is increased (see Fig. 1). The results in Figs. 6 and 7, considering those in Figs. 1-5 and our past publications (18,26), are most consistent with K-Ras via PI3K impinging upon ENaC open probability. DISCUSSION The results of the current study demonstrate that K-Ras activates ENaC in a GTP-dependent manner with PI3K being necessary for this activation. Moreover, that ENaC is activated by the PI3K-specific effector mutant of Ras, Ras G12:C40 , in a wortmannin-sensitive manner is consistent with PI3K being functionally positioned between K-Ras and the channel. Such a transduction pathway fits well with PI3K being a known first effector of K-Ras (3) and with observations showing that both K-RasA and PI3K are necessary for aldosterone actions on ENaC and Na ϩ transport, respectively, in native epithelia (18,20,21,35).
Overexpression of constitutively active K-Ras with ENaC in X. laevis oocytes was previously shown by Verrey and colleagues (19) to result in little change in channel activity in the face of decreasing membrane surface area and surface expression of ENaC suggesting that K-Ras activated ENaC by increasing open probability. In the current study, we detected K-Ras-dependent increases in ENaC activity without changes in membrane area (capacitance, Fig 1). That ENaC activity increased with no change in membrane area simplifies the interpretation of these results leading to a conclusion similar to Verrey's group (19) that K-Ras activates ENaC. This conclusion is also consistent with our previous findings demonstrating that K-RasA and proper posttranslational modification of this small G protein are necessary for aldosterone to increase ENaC open probability in native A6 epithelial cells (18, 36 -38). The finding reported here that GTP is required for K-Ras actions on ENaC (Fig. 2) is novel and suggests that K-Ras acting in a classic manner regulates ENaC as a GTP-dependent signaling switch.
Similar to K-RasA, PI3K is also necessary for aldosterone to increase Na ϩ reabsorption in renal epithelia (20,21,23,24,35). Moreover, the two aldosterone-induced proteins, K-RasA and Sgk, are part of a converging signaling cascade with PI3K positioned between them impinging upon Na ϩ reabsorption (24). Aldosterone via Sgk is believed to increase the half-life of ENaC resident to the plasma membrane by retarding channel retrieval mediated by the ubiquitin ligase Nedd4-2 (23,(32)(33)(34)  ity of ENaC and the number of channels in the plasma membrane.
We, similar to other investigators interested in aldosterone signaling to ENaC, accept the likelihood that aldosterone affects ENaC through multiple transduction cascades; however, the current results in consideration of our previous findings and those published by others, position us to propose a completed linear transduction cascade whereby aldosterone impacts ENaC activity. This cascade would be initiated by increased transcription of K-RasA in response to trans-activation alone and membrane marker plus YFPtagged ENaC (2nd row, YFP-␣ENaC, cold ␤and ␥ENaC) in the absence and presence of constitutively active K-Ras (3rd row). Data were collected with EF microscopy to isolate fluorophores within the plasma membrane. The micrographs in the first column were collected with filter sets to isolate CFP (pseudo-colored red); those in the middle column were with filter sets to isolate YFP (pseudo-colored green); and the last column contains merged images. Gain, exposure time, and scaling were held constant under each condition. B, summary graph of normalized intensity (to CFP-M) levels of membrane YFP-ENaC in the absence and presence of constitutively active and dominant negative K-Ras. by the aldosterone-mineralocorticoid receptor complex binding to a cis-acting steroid response element in the K-ras gene with subsequent increases in the GTP-complexed active levels of K-Ras via mass action, activation of PI3K by GTP-complexed K-RasA, production of phospholipids by this activated lipid kinase, and dependent activation of ENaC by these phospholipid products increasing channel open probability. This linear cascade likely is just one branch of a larger non-linear branching cascade containing several distinct signaling intermediates and points of divergence and convergence with ENaC activity ultimately reflecting the combined sum of all signaling inputs to include the linear K-RasA-PI3K-ENaC cascade.
Thus, ENaC, similar to other ion channels such as IRK1 and ether-a-go-go-related gene K ϩ channels and L-and T-type Ca 2ϩ channels (4 -11), is a final effector of a signaling cascade initiated by a small G protein. We demonstrate here that K-Ras enhances ENaC activity via PI3K signaling. This is the first report (of which we are aware) that shows that Ras modulates activity of an ion channel via PI3K.