The Amiloride-sensitive Epithelial Sodium Channel α-Subunit Is Transcriptionally Down-regulated in Rat Parotid Cells by the Extracellular Signal-regulated Protein Kinase Pathway*

Previous studies have shown that an inducible Raf-1 kinase protein, ΔRaf-1:ER, activates the mitogen-activated protein kinase/extracellular signal-regulated protein kinase (ERK)-signaling pathway, which is required for the transformation of the rat salivary epithelial cell line, Pa-4. Differential display polymerase chain reaction was employed to search for mRNAs repressed by ΔRaf-1:ER activation. Through this approach, the gene encoding the α-subunit of the amiloride-sensitive epithelial sodium channel (α-ENaC) was identified as a target of activated Raf-1 kinases. α-ENaC down-regulation could also be seen in cells treated with 12-O-tetradecanoyl-1-phorbol-13-acetate (TPA), indicating that the repression of steady-state α-ENaC mRNA level was dependent upon the activity of protein kinase C, the target of TPA, as well. Pretreatment of cells with a specific inhibitor of the ERK kinase pathway, PD 98059, markedly abolished the down-regulation of α-ENaC expression, consistent with the hypothesis that the ERK kinase-signaling pathway is involved in TPA-mediated repression. Moreover, through the use of transient transfection assays with α-ENaC-reporter and activated Raf expression construct(s), we provide the first evidence that activation of the ERK pathway down-regulates α-ENaC expression at the transcriptional level. Elucidating the molecular programming that modulates the expression of the α-subunit may provide new insights into the modulation of sodium reabsorption across epithelia.

The expression of the amiloride-sensitive epithelial sodium channel (ENaC), 1 a member of the DEG/ENaC ion channel superfamily (1), on the apical surface of select epithelia regulates the rate of sodium reabsorption. Its sodium channel activity has been identified in distal segments of kidney tubule, airway epithelium, skin, bladder, colon, sweat, and salivary glands. By using functional expression assays in a Xenopus oocyte system, three ENaC subunits, denoted ␣-, ␤-, and ␥-ENaC, have recently been identified, cloned, and characterized (2)(3)(4). These three subunits share considerable amino acid identity (34 -37%) and a similar topological structure (5). Although the ␣-subunit is sufficient for channel activity, expression of the ␤or ␥-subunit(s) of ENaC alone or concurrently does not reconstitute an amiloride-sensitive conductance in Xenopus oocytes, but either subunit augments the channel activity of ␣-ENaC by three-to five-fold. When both ␤and ␥-ENaC are co-injected with ␣-ENaC, greater than 100-fold potentiation of the amiloride-sensitive current is observed (3). Thus, optimal activity of the amiloride-sensitive epithelial sodium channel requires a heteromultimeric complex consisting of ␣-, ␤-, and ␥-subunits (6,7). ENaC plays a major role in the control of fluid and electrolyte homeostasis, as demonstrated by the discovery that mutations in ENaC subunits are associated with hypertensive or hypotensive genetic diseases such as Liddle's syndrome (8) or pseudohypoaldosteronism type 1 (9). Though the exact subunit composition of ENaC is still being resolved, the expression of a functionally active sodium channel is dependent on the presence of ␣-ENaC. Indeed, mouse ␣-ENaC(Ϫ/Ϫ) neonates, generated by homologous gene targeting, develop respiratory distress and die within 40 h from an inability to clear the lung liquid (10).
The expression of ENaC is up-regulated by glucocorticoids in the human fetal lung, where mRNAs of all three subunits increase in response to dexamethasone. In contrast, the induction of ENaC genes by aldosterone appears to be subunitspecific, since ␤and ␥-ENaC, but not ␣-ENaC mRNAs, are increased in rat distal colon (11,12), whereas in the kidney, vasopressin is reported to up-regulate all three subunits (13). However, the molecular components that constitute the complete signaling pathway(s) of these events are far from clear.
In most eukaryotic cells, gene expression is regulated by several distinct signaling modalities. Mitogen-activated protein (MAP) kinases, characterized as proline-directed serine/ threonine kinases (14), are important cellular-signaling components that transduce signals from cell membrane to nucleus in response to a wide variety of stimuli (15)(16)(17). Several MAP kinases have been identified to date, including the extracellular signal-regulated protein kinase (ERK) (18,19), the C-Jun NH 2 -terminal/stress-activated protein kinase (20,21), and p38 (also known as Mkp2/CBSP) (22)(23)(24). Each MAP kinase is activated by a distinct kinase module, which consists of MAP kinase kinase kinase, MAP kinase kinase, and MAP kinase. These kinase modules may provide a structural basis for different signaling cascades and thereby relay extracellular stimuli to specific effectors. The MAP kinase kinases in the ERKsignaling module are MAPK/ERK kinase (MEK) 1 and 2 (25), although the MAP kinase kinase kinases includes Raf-1 and MEKK1 (26,27). Biochemical and genetic studies have revealed that the Raf/MEK/ERK signaling pathway is a pivotal cascade that culminates in phosphorylation of downstream cytosolic and nuclear substrates, ultimately leading to phenotypic cellular changes.
The mechanisms involved in the activation events for the ERK kinase cascade have been studied in detail. Raf-1 kinase is activated through interaction with membrane-associated Ras-GTP and in part by phosphorylation (28,29). In addition, there is ample evidence for the protein kinase C (PKC)-dependent pathway of ERK kinase activation via Raf-1 kinase (30). Since PKC is the major effector for phorbol esters, it has been implicated in the activation of the Raf/ERK pathway (31,32). For instance, treatment with phorbol esters (e.g. TPA) leads to a rapid activation of Ras and/or Raf in most cell types (33)(34)(35). More direct evidence for the involvement of PKC in regulating this pathway has come from transfection or microinjection experiments, which report that TPA treatment or expression of PKC-␣ and PKC-can induce Raf/MEK/ERK pathway activation (34,36,37).
In our efforts to elucidate the molecular mechanisms of Raf-1-mediated cell differentiation, proliferation, and transformation, we employed the polymerase chain reaction-based differential display method (38) to identify genes that are regulated in response to Raf-1 activation. In this report, we demonstrate that the ␣-ENaC mRNA level is selectively down-regulated. Because the ␣-subunit of ENaC is critical for channel activity, we have investigated the signaling pathway that modulates its expression. Recently, the ␣-ENaC gene has been shown to be organized into 13 exons and spans at least 17 kb of the human genome (39). We have characterized the first 5 kb of the ␣-ENaC promoter and engineered several ␣-ENaC/CAT reporter constructs to study the molecular regulatory mechanism(s) for ENaC gene expression. Although much attention has focused on the up-regulation of ENaC, to date very little is known about the molecular mechanisms by which cells downregulate ENaC expression. This paper proposes one such mechanism in salivary epithelial cells, where MAP kinase/ERK activation leads to down-regulation of ␣-ENaC by repressing the transcription of the ␣-ENaC gene.
mRNA Differential Display-Total RNA (200 ng) from either parental or ⌬Raf-1:ER stably transfected Pa-4 cells was reverse-transcribed using primer pair sets provided by GenHunter's RNAimage TM kit 1. Products were labeled by incorporation of 35 S-dATP during DNA synthesis via polymerase chain reaction. Reaction products from both parental and stably transfected cells utilizing identical primer pairs were compared in parallel by size fractionation on a 6% polyacrylamide gel under denaturing conditions. Changes in gene expression between parental and transfected cells were visually indicated by asymmetric banding patterns on autoradiographic X-Omat TM film. Bands of inter-est were excised from the gel, eluted, and reamplified before being ligated into pCR TM II TA Cloning vector (Invitrogen). All clones were sequenced using a Sequenase TM version 2.0 kit (Amersham Pharmacia Biotech). DNA fragments were further analyzed using the Wisconsin Sequence Analysis Package TM to search nucleic acid data bases at the National Center for Biotechnology Information (NCBI) for sequence comparison.
Northern Blot Analysis-Total RNA was extracted from parental or ⌬Raf-1:ER-transfected Pa-4 cells with Trizol® Reagents (Molecular Research Center, Inc.) as instructed. The quality and quantity of RNA were analyzed by fractionating the same amount of RNA from each sample on a 1.5% agarose gel in the presence of 2.2 mM formaldehyde, staining with ethidium bromide, and comparing 28 S and 18 S ribosomal RNAs. Equal amounts of RNA (18 g/sample) from parental, transfected, or treated cells were electrophoresed through a denaturing agarose gel, transferred to nylon membranes (ICN Biomedicals, Inc.), and UV-cross-linked. All blots were prehybridized for 1 h with QuickHyb (Stratagene) and hybridized according to manufacturer's recommendations. To screen various blots, [ 32 P]dCTP-labeled probes were synthesized using isolated ␣-ENaC cDNA fragment (bases 1-905) as a template. A rat ␤-actin probe was also included to ensure that the quality and quantity of RNA between lanes were comparable. All blots were washed in a 0.1ϫ SSC (0.15 M NaCl and 0.015 M sodium citrate), 0.5% SDS solution at 60°C. To improve autoradiographic detection, exposure was carried out overnight at Ϫ80°C between intensifying screens. The membranes were also subjected to electronic autoradiography with the Instantimager 228 (Packard Instrument Co.) to quantitate radioactivity in each band for comparative analysis.
Western Blot Analysis-Parental and ⌬Raf-1:ER stably transfected Pa-4 cells were washed and pelleted at 4°C in Earle's balanced salt solution. Equal amounts of total protein from cell lysates were resolved on 10% SDS-polyacrylamide gel electrophoresis, electroblotted on Immobilon-P Transfer Membrane (Millipore), and incubated with either anti-␣-ENaC antibody (courtesy of Dr. Peter Smith, Allegheny University) or anti-eIF-2␣ antibody (courtesy of Dr. Roger Duncan, University of Southern California). ERK1/2 blots were probed with either Anti-Active TM MAP kinase (Promega, WI) or anti-ERK1 C-16-G (Santa Cruz Biotechnology, Inc., CA) antibodies. Horseradish peroxidase-labeled secondary antibody (Amersham) and the enhanced chemiluminescence (ECL) detection system (Amersham) were used to visualize proteins of interest.
Identification and Characterization of Genomic Clone Encoding Rat ␣-ENaC Gene 5Ј-Flanking Region-To study the rat ␣-ENaC gene promoter/enhancer, a male Sprague-Dawley rat genomic library (in DASH II, Stratagene, La Jolla, CA) was screened with the 32 P-labeled 5Ј-most ␣-ENaC cDNA probe as described previously (43). Four positive plaques were identified from an initial screening of four rat genomic equivalents. Corresponding phage plaques were selected, diluted, replated, and rescreened to obtain single phage clones. One purified positive clone, named 7, showed a strong hybridization signal. 7 was further characterized by restriction enzyme mapping and sequencing analyses. A 5.5-kb EcoRI fragment of 7 (clone 38) that contains the 5Ј-flanking region of the ␣-ENaC gene was used to make two reporter constructs: ␣-ENaC/CATl and ␣-ENaC/CATs. DNA Constructs and Transient Transfection Assays-The ␣-ENaC/ CATl and ␣-ENaC/CATs constructs used for transient transfection assays were constructed from a 4.9-or 1.4-kb DNA fragment of clone 38, respectively; the shorter representing a truncation of the longer DNA fragment (see Fig. 4), where 3.5 kb of the 5Ј-end was deleted. Both DNAs were ligated upstream of the CAT reporter gene in Promega's pCAT basic vector. These two DNA fragments contain an identical 3Ј end, which corresponds to 156 bp upstream of the putative translation start site in the first exon of rat ␣-ENaC gene.(2) Plasmids were transiently transfected into Pa-4 cells by the lipofectAMINE™-mediated method as described previously (40). In all experiments, ⌬56FosdEluciferase plasmid, whose expression level is unaffected by Raf/ERK activation, was included as an indicator plasmid to normalize transfection efficiency. The plasmid molar ratio and total amount of transfected DNA (2 g) were kept constant by supplementing with the pCAT basic vector and/or pkRSPA. (44) Twenty-four h after transfection, the medium was changed, and cells were treated (when appropriate) and harvested 24 h thereafter.

RESULTS AND DISCUSSION
We have been using the rodent salivary gland as a model system to study the genomic response(s) leading to cellular hypertrophy and hyperplasia. Recently, we have shown that changes in the morphology of the parotid salivary gland correlate with the activation of Raf-1 kinase (42). To study the phenotypic consequences of Raf-1 activation in the rat parotid salivary gland, a parotid epithelial cell line (Pa-4) was stably transfected with a gene encoding a human Raf-1 kinase/estrogen receptor fusion protein (⌬Raf-1:ER). ⌬Raf-1:ER responds to estradiol in a dose-and time-dependent manner (41,45), allowing Raf-1 kinase activity to be selectively and conditionally stimulated. Furthermore, activation of ⌬Raf-1:ER is known to be independent of the endogenous estrogen receptor (ER) since either ER agonists or antagonists activate the ⌬Raf-1:ER protein (41,45,46).
Activation of ⌬Raf-1:ER resulted in a number of morphological changes in Pa-4 cells (42) reflecting the changes in mRNA expression. We compared these changes in mRNA expression between the parental (Pa-4) and ⌬Raf-1:ER-transfected cell line (Pa-4⌬Raf-1:ER) using the polymerase chain reactionbased method, mRNA differential display. Reverse transcription-polymerase chain reaction products from Pa-4 and Pa-4⌬Raf-1:ER cells revealed a ϳ125-base pair fragment (B4) that was markedly decreased in Pa-4⌬Raf-1:ER cells (data not shown). Sequence analysis indicated that B4 was identical to the 3Ј terminus of the rat ␣-ENaC cDNA. To confirm the decrease of the ␣-ENaC message in Pa-4⌬Raf-1:ER cells, a Northern analysis was performed on total RNA prepared from Pa-4 and Pa-4⌬Raf-1:ER cells using the B4 fragment as a probe (Fig. 1A). In confirmation of the results from the mRNA differential display, a single band corresponding to the reported size of the ␣-ENaC message (ϳ3700 bases) (2) was detected in RNA from Pa-4 and Pa-4⌬Raf-1:ER cells (Fig. 1A), whereas a dra-matic decrease in signal was observed in ⌬Raf-1:ER-transfected cells (Fig. 1A, 3rd and 4th lanes).
Estradiol alone did not decrease the expression of ␣-ENaC in Pa-4 cells (Fig. 1A, 1st and 2nd lanes), indicating that the down-regulation of ␣-ENaC mRNA is a consequence of ⌬Raf-1:ER activity and not an ER-mediated event via the estrogen response element. Moreover, the basal ␣-ENaC expression in unstimulated Pa-4⌬Raf-1:ER cells was already repressed in comparison to the ␣-ENaC level in parental Pa-4 cells. This was not surprising in that the Pa-4⌬Raf-1:ER cell line has already been shown to exhibit a low level of ⌬Raf-1:ER kinase activity (42), probably due to an incomplete inhibition of the ⌬Raf-1:ER protein in these stably transfected cells (see below). Thus it is possible that this remaining ⌬Raf-1:ER activity is high enough to repress ␣-ENaC expression.
To determine the relative amount of ␣-ENaC protein, a Western analysis using antibody against ␣-ENaC was performed. Consistent with the Northern analysis (Fig. 1A), a significant decrease in the amount of ␣-ENaC protein was detected in Pa-4⌬Raf-1:ER-cells (Fig. 1B, upper panel). The amount of the translation factor eIF-2␣ in Pa-4 and Pa-4⌬Raf-1:ER cell lysates was also determined (Fig. 1B, lower panel) to demonstrate that the decrease in ␣-ENaC was not due to nonspecific down-regulation or protein degradation. The half-life of ␣-ENaC protein is reported to be relatively short at about 1 h (47). Although we do not know if ⌬Raf-1:ER activation directly affects ENaC function, it is likely that the reduced availability of both ␣-ENaC message and protein would lead to diminished sodium absorption across the parotid epithelium.
The ⌬Raf-1:ER protein has been reported to mimic the action of endogenous Raf-1 kinase (41, 42). However, the low level of ⌬Raf-1:ER kinase activity in Pa-4⌬Raf-1:ER cells made it necessary for us to investigate whether activation of the Raf/MEK/ ERK-signaling cascade by a pharmacological means would down-regulate ␣-ENaC in the parental cell. Since PKC activates Raf-1 kinase, we treated cells with a known PKC activator, TPA. A Northern analysis was performed on Pa-4 cells that were treated with TPA at intervals up to 12 h. A marked decrease in ␣-ENaC mRNA was observed within 6 h posttreatment ( Fig. 2A), implicating that the Raf-1 cascade downregulates the ␣-ENaC gene. Although PKC activation with TPA induces ERK kinase activity in many cell types (48), we investigated whether TPA could activate ERK1/ERK2 in Pa-4 cells. A Western analysis of total protein from TPA-treated Pa-4 cells was performed with antibody that recognizes activated ERK2 and cross-reacts with activated ERK1. The kinetics of TPA-induced ERK activation were transient, as indicated by an increase in ERK phosphorylation within 5 min that declined to basal levels after 6 h. There was no change in the level of either ERK1 or ERK2 after TPA treatment, as indicated by an immunoblot probed with antibody that recognizes both ERK1 and ERK2. These data correlate well with the Northern analysis (Fig. 2B) in which no further decrease in ␣-ENaC mRNA is observed between 6 to 12 h. These results suggest that TPA transiently activates the Raf/MEK/ERK-signaling module by demonstrating TPA-mediated transient activation of ERK1/2 in Pa-4 cells.
We also compared the level of ERK activation between untreated ⌬Raf-1:ER cells and TPA-treated Pa-4 cells to determine whether the sustained basal elevation of ⌬Raf-1:ER kinase was sufficient to down-regulate ␣-ENaC. Fig. 2C indicates that the basal level of activated ERK was elevated in ⌬Raf-1:ER cells to a level comparable with that in TPA-stimulated Pa-4 cells. Although the level of activated ERK is slightly higher in TPA-treated Pa-4 cells, activation is transient. Therefore, it is conceivable that in ⌬Raf-1:ER cells, the sustained

FIG. 1. A representative Northern analysis comparing ␣-ENaC mRNA expression in Pa-4 and Pa-4⌬Raf-1:ER cells treated with
␤-estradiol. A, total RNA was extracted from Pa-4 and Pa-4⌬Raf-1:ER cells treated with 1 M estradiol for the indicated time periods. Eighteen micrograms of total RNA from cells at 0 and 24 h after estradiol addition were fractionated, transferred to membrane, and hybridized with 32 P-labeled DNA probe of ␣-ENaC. A rat ␤-actin DNA probe was used as a control for the quality and quantity of RNA loading. B, immunoblots depicting ␣-ENaC expression in Pa-4 and Pa-4⌬Raf-1:ER parotid cells. Cell lysates were prepared from parental Pa-4 and Pa-4⌬Raf-1:ER cells as described under "Materials and Methods." Equal amounts (60 g) of lysates were loaded in each lane and resolved by SDS-polyacrylamide gel electrophoresis. ␣-ENaC protein level was probed with anti-␣-ENaC, an antibody against the amino acid residues 44 -57 of rat ␣-ENaC (courtesy of Dr. Peter Smith). The level of eIF-2␣ was also determined as an internal control. The antibody-bound bands were detected by ECL and exposed to x-ray film for autoradiography. elevation of activated ERK is sufficient to down-regulate ␣-ENaC mRNA expression.
To further support this model, the MEK inhibitor, PD 98059, was employed to confirm the pathway by which TPA treatment led to a decrease in ␣-ENaC mRNA. PD 98059 specifically blocks the activation of ERK kinase by inhibiting the kinase activity of the ERK upstream kinase (MEK), thereby preventing the phosphorylation and activation of ERK (49,50). Its specificity has been tested against at least 18 kinases, including PKC and the highly homologous C-Jun NH 2 -terminal kinase, where PD 98059 appears to have no effect on PKC or C-Jun NH 2 -terminal kinase activity (51). Thus, Pa-4 cells were exposed to 20 M PD 98059 15 min before TPA treatment. A Northern analysis on ␣-ENaC mRNA levels in Pa-4 cells showed that the presence of PD 98059 inhibited TPA-induced down-regulation of ␣-ENaC expression (Fig. 2C, lanes 2 and 4). Therefore, we propose that the activation of the ERK signaling is sufficient to down-regulate ␣-ENaC gene expression.
To investigate whether ERK activation regulates ␣-ENaC gene expression at the post-transcriptional level, Pa-4 cells were treated with the transcription inhibitor, actinomycin D, in the presence or absence of TPA. If TPA-mediated ERK activation increases ␣-ENaC mRNA turnover by a post-transcriptional mechanism, the addition of TPA to actinomycin D-treated cells should augment a decrease in the ␣-ENaC message. The concentration of actinomycin D (5 g/ml) used in these experiments has been shown to completely inhibit the incorporation of [ 3 H]uridine into the acid-soluble fraction (52). Total RNA was prepared from Pa-4 cells that were treated with or without TPA (100 ng/ml) in the presence of actinomycin D (5 g/ml) to block transcription for 4, 8, and 12 h. Equal amounts of RNA were used in the Northern analysis to estimate the half-life for ␣-ENaC mRNA. Based on our results (Fig. 3), the half-life for ␣-ENaC mRNA in parotid cells was calculated to be approximately 8 h. If the transcription of ␣-ENaC gene is completely inhibited by TPA treatment, there should be about 35% ␣-ENaC mRNA in cells after 12 h co-treated with actinomycin D and phorbol ester, as shown in Fig. 3B. Taken together, these results indicate that the addition of TPA does not increase the turnover rate of ␣-ENaC mRNA.
To determine whether the down-regulation of ␣-ENaC is mediated, at least in part, at the transcriptional level by the ERK kinase pathway, we subcloned a 4.9-kb and a 1.4-kb DNA 5Ј-flanking fragment of the rat ␣-ENaC gene into a CAT reporter plasmid (␣-ENaC/CATl and ␣-ENaC/CATs, respectively, Fig. 4A). The promoter/enhancer activities of these constructs were examined by transiently transfecting Pa-4 cells and measuring the CAT activity in cell lysates (Fig. 4). The truncation of the ␣-ENaC promoter/enhancer from Ϫ4.9 kb to Ϫ1.4 kb had a modest effect on basal CAT activity (Fig. 4B, lanes 1 and 5). Furthermore, the CAT activities from both ␣-ENaC/CATland ␣-ENaC/CATs-transfected cells were down-regulated by TPA treatment approximately four-to five-fold (Fig. 4B, lanes 2 and  6). Consistent with ERK-mediated down-regulation of ␣-ENaC, PD 98059 inhibited the TPA-dependent decrease in ␣-ENaC/ CAT activity. Moreover, cotransfection of Pa-4 cells with ␣-ENaC/CATs and a constitutively active form of Raf-1 kinase (RafBXB) (44) showed a dose-dependent decrease in CAT activity, whereas triple transfection with ERK2 (53) had a cooperative effect (Fig. 4C). Together, these results suggest that the major regulatory region(s) that mediates the basal and/or regulated ␣-ENaC expression is located within the 1.4 kb of the proximal 5Ј-flanking region. In summary, our results provide the first demonstration that activation of the MAP kinase/ERK cascade can specifically down-regulate ␣-ENaC promoter/enhancer activities. Eighteen g of total RNA from individual samples were electrophoresed, blotted, and probed with 32 P-labeled probes of rat ␣-ENaC cDNA fragment and ␤-actin, as indicated (upper panel). Below are the kinetics of TPA-induced ␣-ENaC mRNA down-regulation in Pa-4 cells. The percent inhibition shown is the normalized mean ϮS.E. from at least three independent experiments and was calculated relative to the level measured in the control. B, a Western analysis of total protein taken from Pa-4 cells, which were treated with TPA (100 ng/ml) over a 6-h period. Cell lysates were prepared at intervals, where 12 g of total protein was fractionated by size, transferred, and probed with either antibody against active ERK2 (upper panel) or ERK1/ERK2 (lower panel). ERK1 and ERK2 have 90% sequence identity, with either being indicative of functional activation (19). The anti-active ERK2 antibody also showed some cross-reaction with ERK1 in this study. Uniform loading of the gel was demonstrated by probing the blot with ERK1/2 antibody. C, a Western analysis of cell lysate from either Pa-4 or Pa-4⌬Raf-1:ER cells treated with either vehicle or TPA for 5 min. Blots were prepared and probed as in panel B. D, a representative Northern analysis comparing ␣-ENaC mRNA levels in TPA-treated Pa-4 cells in the presence or absence of PD 98059. Where indicated, cells were incubated for 15 min with PD 98059 prior to the 6-h TPA treatment. The Northern blot was performed and analyzed as described in panel A.
To further explore the nature of the down-regulation of ␣-ENaC expression mediated by ERK signaling, we considered whether concurrent protein synthesis is necessary to downregulate the ␣-ENaC message. Protein synthesis was blocked in Pa-4 cells prior to TPA treatment using the protein synthesis inhibitor emetine (10 g/ml). A Northern analysis indicated that ␣-ENaC expression was unaffected by emetine treatment (6 h) alone and that the TPA-mediated repression (Fig. 5, 2nd lane) was markedly inhibited (60%) by co-treatment with emetine (Fig. 5, 4th lane). The same inhibitory effect on TPAmediated down-regulation was also observed using another protein synthesis inhibitor, cycloheximide (10 g/ml, data not shown). Therefore, our results indicate that concurrent protein synthesis is required for ERK-dependent down-regulation of ␣-ENaC gene expression. These results taken together suggest that de novo protein synthesis is required for transcriptional repression of ␣-ENaC through the ERK kinase pathway. Whether or not repression is mediated through a modification of protein-DNA or protein-protein interaction by this newly synthesized factor is unknown at this moment. Detailed deletion mapping of the ␣-ENaC promoter could help identify both the cis elements and trans factors that regulate the ␣-ENaC gene.
Previous studies have provided several examples in which activated ERK targets transcription factor phosphorylation, presumably through regulating transcription via various response elements (54 -56). One of the best-studied mechanisms through which the Raf/MEK/ERK pathway modulates gene expression involves the phosphorylation of the Ets family of transcription factors, such as Elk-1 and Ets-2 (45, 57). ERK2mediated phosphorylation of Elk-1 in its carboxyl-terminal transactivation domain or of threonine 72 in the pointed do-main of Ets-2 is believed to be essential for their transactivation switch. However, little work has been done to identify and characterize the transcription factor(s) mediating negative regulation in response to activation of the ERK kinase-signaling module. One can envision at least two models to explain the differential transcriptional control of the ␣-ENaC gene expression in response to ERK activation. The first model assumes that in parotid cells, ␣-ENaC gene transcription is down-regulated by an aforementioned trans-acting factor(s), such as Elk-1. The second model postulates that the ERK kinase pathway acts as a positive regulator by inducing a novel transacting factor, which negatively regulates ␣-ENaC gene expression. The latter model is supported by the use of protein synthesis inhibitors, emetine or cycloheximide, both of which inhibit the down-regulation of ␣-ENaC mRNA mediated through the ERK kinase cascade. Current efforts are directed toward elucidating the cis-acting elements involved in down-  5) and (3 and 7) represent cells treated with vehicle alone or PD 98059, respectively. CAT assays were performed after normalizing against luciferase activity acquired from the expression of a cotransfected luciferase reporter gene. A representative CAT assay is shown in the lower panel. The average CAT activity Ϯ S.E. (n ϭ 3-4) relative to the corresponding control lane is shown in the lower panel. CAT activity in control lanes is arbitrarily represented as 100. C, Pa-4 cells were transiently transfected with 1.3 g of ␣-ENaC/CATs alone or 0.25 or 0.5 g of RafBXB (lanes 1, 2, and 3, respectively). In lanes 5 and 6, Pa-4 cells were transiently transfected with 1.3 g of ␣-ENaC/CAT and either 0.25 g of ERK2 or 0.25 g of ERK2 and 0.25 g of RafBXB, respectively. The appropriate parental vector was added to keep the total amount of DNA constant. The CAT assays were performed as described in B.
regulating ␣-ENaC steady-state level by the activation of the ERK kinase pathway.
Data in this report from transient transfection assays of two ␣-ENaC/CAT reporter genes show that the ERK kinase-mediated transcriptional repression of the ␣-ENaC promoter/enhancer activity occurs via the cis regulatory element(s) located within the 1.4-kb 5Ј-flanking region of the ␣-ENaC gene. The inhibitory effect of PD 98059 on TPA-mediated down-regulation of reporter activity is also consistent with the ability of TPA to activate ERK. Moreover, the cooperative effect between RafBXB and ERK-mediated suppression of CAT (Fig. 4C) underscores this point. Surprisingly, PD 98059 has a negligible effect on the basal level of ␣-ENaC message (Fig. 2C). This could be explained by results from recent studies, which support the notion that different magnitudes of ERK activation control its intracellular distribution (58 -60). It has been proposed that both duration and intensity of ERK activation explain how the same signaling cascade is used by different growth factor receptors or pharmacological agents, where a variety of signals use the same cascade to elicit distinct changes in gene expression, cell proliferation, and differentiation (61). Recently, it has been shown that phosphorylated (active) ERK can form a homodimer with either phosphorylation state (60). Many substrates of ERK are dimers as well, including basic helix-loop-helix and leucine zipper proteins; ERK dimers may activate substrate dimers through a single interaction. Thus, ERK dimers may be composed of phosphorylated and unphosphorylated forms. As a mixture of fully active and hemiactive dimers, ERK may impact the time course, the activity threshold, and the efficacy involved in regulating downstream events (62). Taken together, this information supports our observation that the treatment of PD 98059 alone has a negligible effect on basal ␣-ENaC expression. However, our results do not exclude that an alternative signaling pathway could exist that also down-regulates ␣-ENaC expression. The data presented here provides experimental evidence and a rationale for the analysis of other modulators of ERK activation with implications for normal and abnormal cell physiology. This raises the possibility that the negative regulation of ␣-ENaC gene expression mediated by ERK activation in parotid cells can be viewed in the larger framework of tissuespecific regulation of ␣-ENaC gene expression. Are putative negative trans-acting factor(s) constitutively expressed in tissue lacking ␣-ENaC expression? Are there cell-type-specific pathways to regulate ␣-ENaC gene expression? Each of these questions will require further studies. The biological advantage of the ERK-mediated ␣-ENaC down-regulation is that it provides a more flexible and precise way to regulate gene expression in response to extracellular stimuli or pharmacological agents. The physiological implications for ERK-mediated ␣-ENaC down-regulation may be extensive and underlie many important aspects of homeostatic regulation for epithelial sodium transport.
The hyperactivity of the epithelial sodium channel has been recently implicated in a number of human diseases. For example, mutations that increase ENaC activity are associated with hypertension. In cystic fibrosis, misregulation of ENaC by defective cystic fibrosis transmembrane conductance regulator has been proposed to cause hyperabsorption of sodium in airway epithelia (63). Under these conditions, down-regulating ␣-ENaC gene expression would be a way to restore normal rates of sodium absorption in epithelial tissues. The ability to repress ␣-ENaC gene expression has important implications for relieving symptoms in sodium channel-associated anomalies. Further investigation that focuses on the trans-acting factor(s) involved in regulating ENaC expression may also identify novel therapeutic strategies for medical conditions such as hypertension and perhaps cystic fibrosis as well. Moreover, other members of the DEG/ENaC superfamily, which includes proteins involved in mechanotransduction, propioception, neurotransmission, as well as fluid and electrolyte homeostasis (64 -66) may be regulated by an ERK-dependent pathway. Thus, it is possible that ERK-dependent regulatory pathways may play an important role in modulating various aspects of cellular function controlled by DEG/ENaC superfamily members.
FIG. 5. Down-regulation by TPA requires de novo protein synthesis. A Northern analysis of ␣-ENaC mRNA in control and TPAstimulated Pa-4 cells cultured in the absence (Ϫ) or presence (ϩ) of the protein synthesis inhibitor emetine (10 g/ml) for 6 h is shown. The Northern blot was performed as described in Fig. 2, and the experiments have been repeated for three times; the percent of inhibition on TPA-mediated repression was about 60%.