HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 37, 34480-34488, September 13, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/37/34480    most recent M204615200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hendron, E. Articles by Stockand, J. D. Search for Related Content PubMed PubMed Citation Articles by Hendron, E. Articles by Stockand, J. D. Social Bookmarking                      What's this?

Identification of Cytoplasmic Domains within the Epithelial Na⁺ Channel Reactive at the Plasma Membrane^*

Eunan Hendron, Pravina Patel, Melinda Hausenfluke, Nikita Gamper, Mark S. Shapiro, Rachell E. Booth, and James D. Stockand^§

From the Department of Physiology, University of Texas Health Science Center, San Antonio, Texas 78229-2900

Received for publication, May 10, 2002, and in revised form, June 6, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The activity of membrane proteins is controlled, in part, by protein-protein interactions localized to the plasma membrane. In the current study, domains within the epithelial Na⁺ channel (ENaC) reactive at the plasma membrane were identified using a novel yeast one-hybrid screen. The cytosolic N terminus of ENaC and the cytosolic C termini of -, -, and ENaC contained domains reactive at the plasma membrane. Fluorescent micrographs of epithelial cells overexpressing fusion proteins of enhanced green fluorescent protein and mENaC cytosolic domains were consistent with those in yeast. A novel membrane reactive domain within the cytosolic C terminus of -mENaC was localized to the 17 amino acids between residues Thr⁵⁸⁴-Pro⁶⁰⁰. Two overlapping internalization signals within the C terminus of -mENaC, a WW-binding domain (PY motif) and a tyrosine-based endocytic signal, were additive with respect to decreasing complementation and expression levels of hybrid proteins. Decreases in expression levels of hybrid proteins containing the PY and endocytic motif were reversed with latrunculin A, an inhibitor of endosomal movement. Decreases in complementation and expression levels of hybrid proteins mediated by the combined PY and overlapping endocytic motif proceeded in the absence of established ubiquitination sites within ENaC. In addition, the endocytic motif was active in the absence of the PY motif, demonstrating that these two domains, while possibly interacting, also have discrete functions. The novel domains within the cytosolic N terminus of ENaC and the C termini of -, -, and ENaC identified here are likely to be involved in protein-protein and/or protein-lipid interactions localized to the plasma membrane. We hypothesize that these newly identified domains play a role in modulating ENaC activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein-protein interactions play a central role in many biological processes. The activities of intrinsic membrane proteins, for instance, are often regulated by the formation of protein-protein interactions with regulatory and accessory proteins. This is particularly true for ion channels, such as the amiloride-sensitive epithelial Na⁺ channel (ENaC)¹ (1, 2).

ENaC is located in the luminal membrane of many epithelial cell types, including those in the distal colon and renal nephron (3-5). The activity of ENaC is rate-limiting for Na⁺ (re)absorption across these epithelia. Thus, ENaC plays a pivotal role in fluid homeostasis and blood pressure control (6-8). Gain of function mutations in ENaC result in rare forms of Mendelian hypertension associated with salt sensitivity and improper salt conservation (9). Conversely, loss of function mutations in ENaC result in hypotension associated with inappropriate salt wasting (10). ENaC is also resident to the luminal membrane of lung alveolar type 2 epithelial cells where it plays a critical role in dehydration of the neonate lung and fluid clearance from the alveolar space during pulmonary edema (11). Ion channels in the ENaC/Degenerin superfamily have, in addition, been identified in neurons and sensory receptors where they play central roles in neural transmission, mechanotransduction, and sensory perception (1, 12).

Cannessa and colleagues (3, 13) and Barbry and colleagues (14, 15) were the first to determine the molecular structure of ENaC. It is accepted that ENaC is a heteromeric channel consisting of at least three homologous but distinct subunits: , , and . The subunit stoichiometry remains controversial (16-19); however, it is believed that ENaC forms the channel pore with - and ENaC serving accessory and regulatory functions (20-22). Each subunit contains two membrane-spanning domains (M1 and M2), one large extracellular domain, and two cytoplasmic tails (13-15, 23). Control of ENaC activity involves both direct post-translation modifications of the channel as well as protein-protein interactions involving the cytoplasmic tails (reviewed in Ref. 24). Whereas the central role of ENaC in fluid homeostasis is well established, less is known about the structure-function relation of this ion channel and the domains critical to protein-protein interactions involved in channel modulation.

The C-terminal cytosolic domains of all three ENaC subunits contain two well conserved proline-rich sequences (P1 and P2) as follows: a minimal consensus SH3-binding domain (P1; PXXP) and a PY motif (P2; PPPXY) that interacts with WW domains in ubiquitin-protein isopeptide ligase, such as Nedd4 (25-29). Interestingly, P1 in - but not - or ENaC interacts with -spectrin localizing this subunit to the luminal plasma membrane of polarized epithelia (27). The salt-sensitive hypertension associated with inheritable Liddle's syndrome in humans results from deletion/disruption of the PY domains in - and ENaC. Such mutations abrogate Nedd4 binding. This then leads to inappropriate overexpression of active ENaC at the apical membrane because of decreases in normal channel retrieval and degradation in the proteosomal compartment (28, 30, 31). The C-terminal cytosolic regions of - and ENaC, in addition, contain well conserved clathrin-coated pit-mediated endocytosis tags (YXX, where  is a hydrophobic amino acid) that partially overlap the PY motif (32). Several other studies implicate the cytosolic C termini of ENaC subunits as being important for channel down-regulation and have identified conserved tyrosine and Ser/Thr residues within the combined PY and endocytic motif (PPPXYX(T/S)L) that affect channel activity when differentially phosphorylated (32-38). The latter portions of the cytosolic C termini of ENaC are also believed to bind actin and interact with the cystic fibrosis transmembrane conductance regulator Cl channel both of which then affect ENaC activity (39, 40). Comparatively less is known about functional domains within the N-terminal cytosolic regions of ENaC subunits. The target for ubiquitination mediated by Nedd4 binding to C-terminal PY motifs is located in the N terminus of both - and ENaC (41). In addition, a 10-amino acid tract (Thr⁹²-Cys¹⁰¹) containing a critical glycine (Gly⁹⁵) in the N-terminal cytosolic region of ENaC has been shown to be important to proper channel gating and is mutated in some forms of the inheritable hypotensive salt-wasting disease pseudohypoaldosteronism type I, which results from loss of function mutations in ENaC (42). It also has been shown that residues 2-67 in -rENaC contain a retrieval tag, and that a distinct domain within the N-terminal tail of this ENaC subunit is required for channel activity (43).

Gupta and Canessa (44) have demonstrated previously that heterologous expression of ENaC in Saccharomyces cerevisiae results in functional channels. In the present work, we sought to further use this simple system to identify novel domains within the cytosolic regions of ENaC subunits that are reactive at the plasma membrane and to investigate further the role of C-terminal domains involved in channel internalization and ultimately degradation. Such domains are likely involved in protein-protein and protein-lipid interactions important for channel modulation, localization, and recycling.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials

All chemicals and enzymes were of reagent grade and purchased from Sigma, Invitrogen, New England Biolabs, and Tocris unless noted otherwise. The BCA Protein Assay was from Pierce. All chemicals and materials used in Western blot analysis were from Bio-Rad. Anti-SOS antibody was from BD Transduction Laboratories. Anti-mouse horseradish peroxidase conjugate secondary antibody was from Kirkegaard % Perry Laboratories. ECL reagents were from PerkinElmer Life Sciences. All DNA sequencing was performed by the molecular biology core facility at the University of Texas Health Science Center, San Antonio.

The immortalized amphibian distal tubule A6 and Chinese hamster ovary (CHO) cell lines were from American Type Culture Collection. The conditionally immortalized mouse collecting duct principal cell line mCT1 was a kind gift from Dr. C.C. Cotton. Refer to Ref. 45 for more information regarding isolation and characterization of this cell line. The S. cerevisiae cdc25H yeast strain was from Stratagene (cdc25H: MAT ura3-52 his3-200 ade2-101 lys2-801 trp1-90 leu2-3,112 cdc25-2 (ts) Gal⁺ (46)). The pEGFP-C2, pEGFP-C3, pEGFP-F, pDsRed2-Mito, and pDsRed2-Nuc plasmids were from CLONTECH. The pSOS plasmid and control pMyr plasmids were from Stratagene. The plasmids encoding mouse ENaC subunit cDNAs have been described previously (47) and were a kind gift from Dr. T. R. Kleyman.

Cell and Yeast Culture

A6 cells were maintained in culture using standard procedures described previously (48). For fluorescence microscopy, A6 cells were seeded on number 0 Micro cover glass (Electron Microscopy Sciences) and transfected after becoming 50-80% confluent (see below). Transfected cells were used for study up to 2 days after transfection. The mCT1 cell line was maintained in culture using standard techniques described previously (45). In brief, cells were maintained in a humidified incubator at 33 °C and 5% CO₂ in 1:1 DMEM/Ham's F-12 supplemented with 1.3 µg/liter sodium selenite, 1.3 µg/liter triiodothyronine, 5 mg/liter insulin, 5 mg/liter transferrin, 25 µg/liter prostaglandin E₁, 2.5 mM glucose, 5 µM dexamethasone, 5% fetal bovine serum, 10 units/ml -interferon, and antibiotics. For fluorescence microscopy, mCT1 cells were treated in a similar manner as A6 cells with the exception that plasmid DNA was introduced with a distinct transfection protocol (see below). Chinese hamster ovary cells were maintained in culture using standard techniques. For fluorescence microscopy, CHO cells were treated identically to mCT1.

Yeast cells were maintained in culture using standard techniques described previously (46, 49). In brief, yeast strains were cultured with constant agitation at 24 °C in YPAD broth from single colonies isolated from YPAD agar source plates (freshly streaked from frozen stocks) to prepare cells competent for transformation (see below). For biochemistry, isolated yeast colonies (from SD/glucose (Leu) agar plates; see below) were maintained with constant agitation at 24 in SD/glucose (Leu) broth (see below).

Molecular Biology

Chimeric hSOS-ENaC Fusion Proteins-- pSOS encodes truncated (amino acids 1-1067; catalytic domain) human Son-of-Sevenless (hSOS), which is the human homologue of yeast cdc25. This truncated hSOS is incapable of independently localizing to the plasma membrane in yeast (50-52). Expression of hSOS is constitutive under the control of the yeast ADH1 promoter (p_ADH1). Both cdc25 and hSOS are GTP-exchange factors capable of activating Ras-signaling and thus stimulate yeast growth when recruited to the plasma membrane. The cytosolic domains (N and C termini) of -, -, and -mENaC were amplified with a standard PCR using mouse ENaC subunit cDNAs originally in pBluescript SK (Stratagene) (47). PCR primers (see Table I) engineered restriction sites, which were subsequently used after gel purification of PCR products (GeneClean, QBIOGENE) to subclone the respective fragments in-frame behind hSOS in pSOS to create pSOS-N, pSOS-C, pSOS-N, pSOS-C pSOS-N, and pSOS-C. These constructs encode a hybrid protein containing hSOS proximal to the respective portions of ENaC. Every pSOS-ENaC construct was sequenced to ensure proper orientation, reading frame, and sequence identity.

View this table: [in this window] [in a new window]   Table I PCR primers used to amplify the NH2 and COOH termini of -, -, and -mENaC for cloning into pSOS

Mutagenesis of ENaC-- All mutagenesis was completed using QuikChange (Stratagene) site-directed mutagenesis per the manufacturer's instructions. The primers and constructs used to create the deletion and truncation mutations of pSOS-C and its derivatives are listed in Table II. Every pSOS-C mutant was sequenced to ensure proper mutagenesis and to confirm orientation, reading frame, and sequence identity.

View this table: [in this window] [in a new window]   Table II Primers used in PCR based mutagenesis of pSOS-C

EGFP-ENaC Fusion Proteins-- EGFP-ENaC fusion proteins were created by subcloning appropriate gel-purified cDNA fragments from pSOS-ENaC plasmids into pEGFP-C2 in-frame behind EGFP. The BamHI and SacII fragments (into BglII and SacII) from pSOS-C, -C, -N, and -C were used to create pEGFP-C, pEGFP-C, pEGFP-N, and pEGFP-C, respectively. The PstI and SacII fragment from pSOS-N was used to generate pEGFP-N (from pEGFP-C3). The BamHI and EcoRI fragment (into BglII and EcoRI) from pSOS-N was used to generate pEGFP-N. Every pEGFP-ENaC construct was sequenced to ensure proper orientation, reading frame, and sequence identity.

Yeast One-hybrid Screen-- A novel yeast one-hybrid screen developed in our laboratory was used to identify the cytosolic domains within ENaC subunits that are reactive at the plasma membrane. This screen is based upon the SOS recruitment system developed by Aronheim and colleagues (51, 52). In the current study, pSOS-N, -C, -N, -C, -N, and -C were overexpressed in the S. cerevisiae yeast strain cdc25H. A two-stepped screening process involving auxotrophic selection with a leucine marker contained within the pSOS plasmid (leu2) and complementation by hybrid hSOS-ENaC constructs of the temperature-sensitive growth defect inherent to the cdc25H yeast strain was used to identify plasma membrane-reactive ENaC domains. The rationale for this one-hybrid screen is that in order for hSOS-ENaC fusion proteins to complement temperature-sensitive growth, ENaC domains must recruit truncated hSOS (catalytic domain) (50) to the plasma membrane, where it then is able to activate Ras signaling and promote growth at restrictive temperature (37 °C). Only domains within ENaC that interact with integral or peripheral membrane proteins or membrane lipids will recruit hSOS-ENaC to the membrane and complement. Yeast transformed with the various pSOS-ENaC constructs were initially plated on SD/glucose (Leu) agar at 24 °C. Colonies were allowed to form over a 4-day period. This source plate then was either replicated with the replica plate maintained at restrictive temperatures (37 °C) or colonies were patched from the source plate in duplicate onto SD/glucose (Leu) plates to create two identical arrays (see Fig. 1). Arrays were grown in parallel at permissive (24 °C) and non-permissive (37 °C) temperatures. Colony formation (growth) was quantified after 2 days. Digital images of yeast plates were captured with a DC4800 Zoom Digital Camera (Eastman Kodak Co.) interfaced with a PC running Kodak IFSCore software.

Transformation/Transfection-- Competent yeast cells were prepared using standard techniques. In brief, cells were maintained in liquid culture (YPAD broth) at 24 °C until reaching an A₆₀₀ = 0.7. After rinsing, yeast were treated with LiSORB (100 mM LiOAc, 10 mM Tris-HCl, 1 mM EDTA, 1 M sorbitol, pH 8.0) for 30 min at room temperature, then pelleted, and resuspended in PEG/LiOAc (10 mM Tris-HCl, 1 mM EDTA, 100 mM LiOAc, 40% PEG 3350, pH 8.0) containing 10% Me₂SO and ~0.8 mg/ml sheared and boiled salmon sperm (CLONTECH) as a carrier. Competent yeast (100 µl; titer = 2 × 10⁷ colony-forming units/ml) was transformed with 0.5 µg of plasmid DNA using heat shock (42 °C for 20 min) in the presence of 2 µl of 1.4 M -mercaptoethanol. Transformed cells were then pelleted and resuspended in 0.5 ml of 1 M sorbitol and plated on SD/glucose (Leu) agar plates using 4-mm glass beads.

Amphibian cells were transfected using methods described previously (48). In brief, cells seeded on number 0 Micro cover glass were treated for 6-8 h with 1 ml of complete amphibian media (48) supplemented with 85 µl of LipofectAMINE spiked with 8 µl of PLUS reagent and 2 µl (~4 µg) of plasmid DNA. Cells then were maintained in complete media and used between 24 and 48 h.

CHO and mCT1 cells were transfected using Polyfect Transfection Reagents (Qiagen) per the manufacturer's instructions. In brief, cells seeded on number 0 Micro cover glass were treated overnight with 1 ml of DMEM (supplemented with 10% fetal bovine serum) and 150 µl of transfection media containing DMEM plus 2.5 µg of plasmid previously mixed with 15 µl of Polyfect Transfection Reagent for 10 min. Cells were maintained using standard methods following transfection and were used for experimentation over the next 24-48 h.

Protein Chemistry

For Western blot analysis of hSOS-ENaC hybrids, inoculates of single yeast colonies were developed at 24 °C in 10 ml of SD/glucose (Leu) broth for 2-3 days (A₆₀₀ >1.0). Yeast whole cell extract was prepared by lysing washed and pelleted yeast in cell lysis buffer (CLB; 5 µl/µg pellet weight) containing (in mM) 140 NaCl, 2.7 KCl, 10 Na₂HPO₄, 1.8 KH₂PO₄, and 1% Triton X-100 supplemented with 1 mM phenylmethylsulfonyl fluoride and 1 µl of protease inhibitor mixture (for fungal and yeast cells; Sigma) per 100 µl of CLB. Pellets treated with CLB were vortexed for 5 min at 4 °C in the presence of an equal volume of acid-washed glass beads (0.5 mm diameter). Lysates were then normalized for total protein concentration and sample buffer and 20 mM dithiothreitol added. All Western blot analyses were performed using standard procedures described previously (48). In brief, proteins were separated by SDS-PAGE (on 7% gels) and subsequently electrophoretically transferred to nitrocellulose (0.45 µm). Nonspecific interactions were blocked with Tris-buffered saline (TBS) supplemented with 5% dried milk (Carnation) and 0.1% Tween 20. Primary anti-SOS antibody was applied overnight at 4 °C in the presence of TBS/Tween and milk. Blots next were probed with anti-mouse horseradish peroxidase-conjugated antibody at room temperature for 1 h in TBS/Tween and milk. Blots were developed with enhanced chemiluminescent reagents and Kodak BioMax Light-1 film. Digital images of Western blots were generated using a ScanJet4200C (Hewlett-Packard).

Fluorescence Microscopy

Fluorescence microscopy was performed on an inverted Nikon Eclipse TE300 microscope in DIC configuration with an oil-immersion 40×/1.30 NA objective (Nikon). A USHIO 150-watt xenon lamp in combination with Polychrome IV monochromator (T.I.L.L. Photonics GMBH) was used as a light source. Images were acquired with an Imago 12-bit VGA CCD camera and analyzed with TILLvisION 3.3 software (T.I.L.L. Photonics GMBH). The fluorescein isothiocyanate HQ 96170M filter cube (Chroma) with excitation at 570 nm and R HQ 96171M filter cube (Chroma) with excitation at 540 nm were used for EGFP and DsRed2 imaging, respectively.

Confocal Microscopy

Confocal microscopy was performed at the University of Texas Health Science Center, San Antonio, Optical Imaging Core Facility. For these experiments, mCT1 cells overexpressing EGFP constructs were plated on number 1 circular coverslips and fixed in 4% paraformaldehyde, washed three times with phosphate-buffered saline, twice with 100 mM NaPO₄, pH 7.4, and then twice with water. Air-dried slides containing fixed mCT1 cells were then mounted on a drop of Vectashield (Vector Laboratories) and sealed with nail polish. Cells were viewed with an Olympus FV-500 confocal microscope using the laser and excitation/emission filters appropriate for EGFP. Single images were collected 3-5 µM above the surface of the coverslip.

Statistics

Complementation frequency was compared using a z test on proportions with the Yates correction where necessary. A p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The N Terminus of -mENaC and the C Termini of -, -, and -mENaC Contain Domains Reactive at the Plasma Membrane-- A novel yeast one-hybrid screen was used to test whether the cytosolic domains of each ENaC subunit contain motifs reactive at the plasma membrane. Positive domains resulted in complementation of growth at restrictive (37 °C) temperatures due to localization of hSOS-ENaC hybrids at or near the plasma membrane. Fig. 1A shows the ENaC domains (colored black) used to create hSOS-ENaC hybrids. Shown in Fig. 1B are typical original source plates (top row) grown at 24 °C for 4 days containing cdc25H yeast transformed with pSOS-N (left) and pSOS-C (right), which encode hybrids containing the complete cytosolic N and C termini, respectively, of the -mENaC subunit. Duplicate arrays, in this case with each containing 23 distinct patched colonies, were generated from the respective source plate and developed for 2 days in parallel at 24 (middle row) and 37 °C (bottom row). As expected, plaques consistently formed for every patched colony in arrays developed at permissive temperatures. Complementation was positive as seen for C if plaques formed in arrays developed at restrictive temperatures. In contrast, complementation was negative as seen for N if plaques failed to form in arrays developed at 37 °C.

View larger version (53K): [in this window] [in a new window]   Fig. 1.   A one-hybrid screen for identifying putative membrane reactive domains. A, schematic representation of ENaC subunits. The cytosolic regions (black) were subcloned behind hSOS to create hSOS-ENaC hybrids used in the one-hybrid screen. B, top row shows source plates containing cdc25H yeast transformed with pSOS-N (left) and C (right) developed at permissive temperatures. The middle and bottom rows show duplicate arrays patched from the respective source plate developed for 2 days at permissive (middle) and restrictive (bottom) temperatures, respectively.

The experimentally quantified spontaneous revertant frequency for the cdc25H yeast strain used in these studies was below 1 in 10,000. (A few such revertants are seen in the pSOS-N plate at 37 °C.) Revertants were easily identified for they failed to develop full plaques.

Fig. 2A shows duplicate arrays developed in parallel at permissive (left) and restrictive (right) temperatures containing cdc25H yeast transformed with one of the six possible pSOS-ENaC hybrids, as well as the pSOS negative control (top lanes). (Each group is represented by 6 patched colonies.) As shown by this representative experiment, hybrids containing the N terminus of - but not - and -mENaC as well as those containing the C terminus from any of the three subunits complemented growth at restrictive temperatures. Summarized data in Fig. 2B show that the complementation frequency for yeast transformed with pSOS-N, -C, -C, and -C was significantly greater than pSOS alone, whereas that for pSOS-N and -N was not. Although pSOS-N complemented, yeast overexpressing this hybrid grew slower consistently forming abundant but punctated plaques after 2 days compared with the C termini of -, -, and -mENaC, which formed more solid plaques at the same time point (see Fig. 2A, right panel). Yeast transformed with pSOS-N, pSOS-N, and pSOS failed to form plaques of any sort, excluding spontaneous revertants, up to the longest time point measured (6 days after patching). Experiments where original source plates were replicated instead of patched and then developed at restrictive temperatures resulted in identical observations (data not shown).

View larger version (42K): [in this window] [in a new window]   Fig. 2.   The cytosolic, N terminus of -mENaC and C termini of -, -, and -mENaC contain domains reactive at or near the plasma membrane. A, parallel arrays containing cdc25H yeast transformed with pSOS, pSOS-C, -C, -C, -N, -N, and -N developed for 2 days at permissive (left) and restrictive (right) temperatures. B, summary graph showing complementation frequency for each of the constructs. C, a typical Western blot containing lysate extracted from one of the colonies developed at permissive temperatures for each group in A. Blots were probed with anti-SOS antibody.

The positive control of co-overexpression of two interacting proteins (myristoylated-MAFB and hSOS-MAFB) and the negative controls of transformation with empty pSOS, as well as co-overexpression of two non-interacting proteins (myristoylated-collagenase and hSOS-MAFB or myristoylated MAFB and hSOS-lamin C; data not shown for these last two negative controls) showed that our one-hybrid screen faithfully reported membrane reactivity for pSOS fusion proteins.

The results in Figs. 1 and 2 demonstrate that the N terminus of -mENaC and the C termini of -, -, and -mENaC localized truncated hSOS to the plasma membrane in yeast allowing growth at restrictive temperatures. Because complementation in the one-hybrid screen is predicted to be sensitive to hSOS-ENaC expression levels, we tested if expression of pSOS-N and -N reached sufficiently high levels to complement growth. The typical Western blot in Fig. 2C contained lysate extracted from cdc25H yeast overexpressing hSOS-ENaC hybrids. As shown in this figure, the C-terminal fusion proteins of hSOS-ENaC expressed at consistently high levels, whereas the N-terminal fusion proteins expressed at significant but lower levels. This may explain the slower plaque formation observed with pSOS-N compared with C-terminal constructs (see Fig. 2A, right panel), but because all N-terminal constructs expressed equally well, this argues against the notion that pSOS-N and -N failed to complement due to an unusually low expression level. The results described next also argue against this notion.

Fig. 3 shows the cellular expression patterns in the immortalized principal mCT1 cell line of EGFP-C (A), -N (B), and -N (C). Images in the top row were collected under epi-fluorescence with EGFP conditions. In addition to the EGFP-ENaC constructs, cells were also transfected with a red nuclear or mitochondrial marker (2nd row and merged images shown in the 3rd row). Shown in the bottom row are transmitted light images for each group. EGFP-C, -N, and -C had expression patterns identical to EGFP-C (data not shown). Whereas membrane localization of EGFP-ENaC hybrids cannot be definitively determined from these images, the findings for EGFP-N, -C, -C, and -C are consistent with a portion of the hybrid proteins localizing to both the cytosol and at or near the plasma membrane (see also Fig. 6). More importantly, the results in these mammalian cells are entirely consistent with our findings in the yeast complementation screen. The two hybrids that failed to complement in yeast, -N, and -N were localized in mCT1 cells exclusively to a region distinct from that of the other hybrids. For -N, this region was clearly the nucleus. The -N hybrid localized in a punctated fashion to a more ambiguous compartment. Identical findings were observed with fluorescence microscopy when EGFP-ENaC hybrids were transiently overexpressed in A6 and CHO cells (data not shown).

View larger version (47K): [in this window] [in a new window]   Fig. 3.   The C termini for all three mENaC subunits and the N terminus of - but not - and -mENaC localize to the plasma membrane and cytosol in the immortalized mCT1 mouse collecting duct principal cell line. A-C, fluorescence microscopy of mCT1 cells transfected with pEGFP-C, -N, and -N fusion proteins, respectively. Top and top middle rows collected with EGFP and DsRed conditions, respectively. The bottom middle rows show merged images. The bottom row shows transmitted light images.

It is not surprising that ENaC domains, which complemented in yeast, have a dispersed expression pattern in mammalian cells considering that these domains are involved in dynamic protein-protein and/or protein-lipid interactions, predicting that only a small portion of the hybrid protein localized at or near the plasma membrane at any one time. We argue that this in fact supports the use of the yeast one-hybrid screen, which only focuses on membrane reactivity, to identify putative domains involved in protein-protein and protein-lipid interactions at the plasma membrane.

-mENaC Contains a Membrane-reactive Domain within Residues Thr⁵⁸⁴-P⁶⁰⁰-- Experiments were performed to localize further the membrane-reactive domain in the cytosolic region of the C terminus of -mENaC. Deletions and truncations removing proximal and distal portions of the cytosolic region of the C terminus of ENaC were created (shown in Fig. 4A) and tested for complementation in the one-hybrid screen. As shown in Fig. 4B and summarized in Fig. 4C, deletion of the proximal portion of C abolished complementation. The complementation frequencies for the three different proximal deletion mutants, N₁, N₂, and N₃, were 0.0 (0/80), 0.05 (3/60), and 0.0 (0/110), respectively. In contrast, constructs that had the distal portion deleted (e.g. C-C₁) maintained full complementation (complementation frequency = 1.0; 137/137) indicating that the membrane-reactive domain localizes to the first 28 residues following the 2nd membrane-spanning region. The reactive domain was further localized to the 17 amino acids between residues 584-600 (TPPSTETPSSQQGQDNP) using the M₂ C-terminal -mENaC mutant. The M₂ mutant complemented fully (complementation frequency = 0.97; 35/36), although it resulted in formation of punctated plaques, possibly due to a slightly decreased complementation efficiency. To exclude the possibility that expression levels unduly influenced these results, Western blot analysis was performed for each mutant construct. As indicated by the typical Western blots shown in Fig. 4D, all mutants expressed equally well. We conclude that the membrane-reactive domain of the C terminus of -mENaC localizes to residues 584-600.

View larger version (39K): [in this window] [in a new window]   Fig. 4.   The membrane-reactive domain of -mENaC is within residues Thr584-P600. A, schematic representation of the pSOS-C deletion/truncation mutants used for these experiments. B, parallel arrays of cdc25H yeast transformed with pSOS-C deletion/truncation mutants developed at permissive (top row) and restrictive (bottom row) temperatures for 2 days. C, summary graph showing complementation frequencies for the pSOS-C deletion/truncation mutants. D, a typical Western blot probed with anti-SOS antibody of lysates extracted from one colony developed at permissive temperatures for the respective groups in B.

Experiments next were performed to test whether the proximal region of the cytosolic portion of the C terminus of ENaC was also reactive at the plasma membrane in mammalian collecting duct principal cells (Fig. 5). Fig. 5A shows a schematic representation of the deletion construct used in these experiments. The C-C₂ EGFP fusion protein contained the 40 residues directly following the 2nd transmembrane domain of ENaC. Shown in Fig. 5B are transmitted light (top) and epi-fluorescence images (bottom) of mCT1 cells transiently transfected with either a membrane marker EGFP-F (left, EGFP modified to contain a C-terminal isoprenylation and methylation motif similar to that in Ras) or C-C₂ (right). Both constructs had discernible membrane localization. These results are consistent with those in yeast and together argue that the proximal region of the cytosolic, C terminus of ENaC contains a motif that reacts at or near the plasma membrane.

View larger version (102K): [in this window] [in a new window]   Fig. 5.   The first 40 amino acids of the cytosolic, C terminus of -mENaC localize as an EGFP fusion protein to the plasma membrane of mCT1 cells. A, schematic representation of the truncation mutant used in these experiments. B, transmitted light (top) and epi-fluorescence (bottom) images of mCT1 cells overexpressing a plasma membrane maker (EGFP-F; left) and an EGFP fusion protein containing the first 40 amino acids directly following the 2nd transmembrane domain of -mENaC (EGFP-C-C2; right). Arrows note plasma membrane localization.

Confocal microscopy was performed to better determine the cellular locale of C and its derivatives. Shown in Fig. 6 are confocal images of mCT1 cells overexpressing the EGFP-F membrane marker, EGFP, EGFP-C, EGFP-N, and EGFP-C-C₂. The EGFP-F, EGFP-C, and EGFP-C-C₂ fusion proteins had clear membrane localization; however, a portion of these proteins also localized to the cytosol. In comparison, EGFP and EGFP-N showed no membrane localization with EGFP primarily localizing to the intracellular space in a diffuse manner and EGFP-N localizing in a more punctated manner. These findings are similar to those observed with fluorescence microscopy (Figs. 3 and 5) and are also consistent with findings in yeast demonstrating that C and its C₂ derivative but not N complement. These results provide good support for the proximal portion of C localizing to the plasma membrane in mammalian epithelial cells.

View larger version (36K): [in this window] [in a new window]   Fig. 6.   Comparison of the cellular locale of N, C, and its C2 derivative. Confocal images were generated from mCT1 cells overexpressing the EGFP-F membrane marker, EGFP, EGFP-C, EGFP-N, and EGFP-C-C2. Control (CON) was untransfected cells.

The C Terminus of -mENaC Contains Two Motifs Additive with Respect to Decreasing Complementation and Hybrid Expression Levels-- It is recognized that the PY and overlapping endocytic motif in ENaC (the full domain in -mENaC is PPPRYNTL) play a role in normal channel modulation by regulating channel internalization and subsequent degradation (24). In the current study, we asked using the yeast one-hybrid screen whether these overlapping motifs combine to decrease complementation, and whether the endocytic motif was functional in the complete absence of the PY motif and in the absence of ubiquitination sites previously localized to the cytosolic N-terminal region (41). Fig. 7A shows a schematic representation of the various mutants used in these experiments. The C-C₃ and M₁ mutants contain both the PY and YXXL motifs and only the YXXL motif, respectively. Generation of both constructs deleted surrounding residues and thus as a consequence moved these retrieval signals from their normal sequence context. Fig. 7B shows duplicate arrays containing yeast transformed with the various mutants shown in Fig. 7A developed for 4 days at permissive (top) and restrictive (bottom) temperatures. Deletion of the residues directly following the combined PY endocytic motif (C-C₃) abrogated complementation. Similarly, deletion of the 20 residues directly preceding the endocytic YXXL motif (C-M₁) decreased complementation. Fig. 7C shows a summary graph of such experiments. The complementation frequencies of 0.08 (9/110) and 0.54 (63/117) for the C-C₃ and C-M₁ mutants, respectively, were significantly lower than for the full-length cytosolic, C-terminal hybrid and a hybrid containing the first 40 residues succeeding the second transmembrane domain (C-C₂; 51/51). To determine whether expression levels were involved in these results, Western blot analysis was performed. Fig. 7D shows results from typical Western blots containing lysate from colonies expressing C, -C₂, -C₃, -M₁, and -M₁ hybrids developed at permissive temperatures. The -C₃ and -M₁ lysates were from colonies that failed to complement. Fig. 7E shows an overexposed Western blot containing lysate (harvested from colonies grown at 24 °C) from yeast transformed with the various hybrids. This typical blot contains lysate from two different C-M₁ colonies: one that complemented (noted by the asterisk) and one that failed to complement. Colonies containing C-M₁ that complemented consistently had higher expression levels compared with those that failed to complement (n = 3). We interpret these results as showing that the PY and overlapping endocytic motifs became activated once they were removed from their normal context with the consequences of this activation being decreased complementation and ultimately decreased levels of expression.

View larger version (43K): [in this window] [in a new window]   Fig. 7.   The endocytic YXXL motif can function independently of the overlapping PY motif. A, schematic representation of the pSOS-C mutants used in these experiments. B, parallel arrays of cdc25H yeast transformed with pSOS-C mutants developed at permissive (top) and restrictive (bottom) temperatures for 4 days. C, summary graph showing complementation frequencies. D, a typical Western blot probed with anti-SOS antibody of lysates extracted from one colony developed at permissive temperatures for the respective groups in Fig. 6B. For this blot, non-complementing C3 and M1 colonies were chosen. E, a typical Western blot probed with anti-SOS antibody of lysates extracted from cdc25H yeast transformed with pSOS-C-N3, -M1, and -C3. All lysates, except that for -M1*, were from non-complementing colonies developed at permissive temperatures. Lysate for -M1* was from a complementing colony (developed at permissive temperatures).

An alternative interpretation of the results in Fig. 7 are that both C-M₁ and C-C₃ mutants simply fail to produce high levels of protein due to problems associated with transcription and/or translation of these plasmids. Experiments in Fig. 8 were performed to determine whether transcription and translation of the C-M₁ and C-C₃ mutants was normal. Shown in this figure are representative experiments (n = 3) where the top Western blot contains lysate from cdc25H yeast transformed with C, C-M₁, and C-C₃, and the bottom blot contains lysate from the same colonies grown in parallel in the presence of 0.5 µg/ml latrunculin A, an inhibitor of actin polymerization and thus endosomal movement (49) for 6 days. This latrunculin A concentration was experimentally established to be just below the lethal dose for cdc25H yeast, although it did a substantially slow rate of growth. In the yeast grown in latrunculin A, all three constructs clearly expressed equally well. Thus, it is likely that transcription and translation of all mutants are similar and that the M₁ and C₃ constructs are preferentially degraded. It follows then that the results in Fig. 7 are consistent with the full motif (PPPRYNTL) having a higher activity compared with the endocytic motif alone, and that the endocytic motif does function in the complete absence of the PY motif. Because these constructs only contained the cytosolic, C-terminal domain of -mENaC, the putative site for ubiquitination was also missing. This further argues that decreased complementation and expression levels must proceed in the absence of ubiquitination. We conclude that the endocytic motif is, in part, functionally distinct from the PY domain, although they may impinge upon each other's activity and are additive with respect to decreasing complementation. Because the decrease in complementation was associated with decreased expression levels that were reversed by an inhibitor of endosomal movement, we speculate that the PY domain and overlapping endocytic motif are functionally independent, in part with respect to promoting channel internalization and subsequent degradation.

View larger version (57K): [in this window] [in a new window]   Fig. 8.   The C-M1 and C3 mutants express equally well with C in the absence of lysosomal/proteosomal degradation. Typical Western blot probed with anti-SOS antibody of lysate extracted from cdc25H yeast transformed with the noted constructs developed in the absence (top) and presence (bottom) of latrunculin A.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The current study identified novel domains within the cytosolic N terminus of -mENaC and C termini of -, -, and -mENaC that interacted with lipids and/or proteins at or near the plasma membrane. The domain in the C terminus of -mENaC was determined to reside within the first 28 amino acids directly following the 2nd transmembrane-spanning region, most likely between residues Thr⁵⁸⁴ and Pro⁶⁰⁰. We believe the domains identified in the current study are novel for several reasons. First, it has not been reported previously that the N terminus of ENaC is directly involved in protein-protein and/or protein-lipid interactions. The only work that speaks to this issue identified residues in the N-terminal, cytosolic region of ENaC as being important for channel gating (42) and activity (43) and that the N-terminal, cytosolic regions of - and ENaC are targets for ubiquitination (41). It is unlikely that the reactive domain identified in the current study in the N terminus of ENaC is involved in ubiquitination, because this N-terminal tail construct was assayed in the absence of the Nedd4-binding domain contained within the C-terminal tail. Future study will determine whether this domain is involved in modulating channel gating. Second, most studies identify the C-terminal cytosolic regions of ENaC as playing a role in suppression of channel activity. As discussed further below, we also investigated this function. One study showed that the P1 domain of - but not - and ENaC interacts with -spectrin localizing ENaC to the plasma membrane (27). The C-terminal domains identified in the current study, at least definitively for -mENaC, are proximal to P1 and P2. Future investigation will further localize the novel domains identified here and determine the functional role these domains play in controlling ENaC activity. That the truncated cytosolic, C terminus of -mENaC localized to the plasma membrane in mouse collecting duct principal cells argues that they play a functional role in modulating ENaC. We speculate that these novel domains are involved in channel regulation by mediating protein-protein and/or protein-lipid interactions likely through a function distinct from promoting channel retrieval, because retrieval signals clearly functioned normally in the present study (see below).

Sequence alignment of the reactive domain in the C-terminal, cytoplasmic tail of -mENaC with ENaC from other species reveals a high degree of sequence conservation. Indeed, the sequence QGQ(D/E)XP is absolutely conserved in human, rabbit, rat, and mouse ENaC. In addition, the C-terminal cytosolic region of Xenopus laevis ENaC contains a similar motif. Interestingly, the QGQ(D/E)XP sequence is also very well conserved across species in several zinc finger proteins and in the -2 and -3 subunits of the Na⁺/K⁺-ATPase. Similar sequences, in addition, are found twice in gelsolin from halocynthia to man. Related proteins, such as scinderin, villin, and flightless 1 homologues also contain this sequence. The brush border enzyme meprin , Na⁺/K⁺-ATPase inhibitors in the trappin family, and PABP4, in addition, contain similar GQ type sequences. We speculate that this sequence may be a novel motif involved in protein-protein and/or protein-lipid interactions. Future investigation will reveal the functional role of this motif and identify proteins and/or lipids that interact with ENaC at this site.

One important outcome of the current study is development of a novel screen for identifying putative domains reactive at the plasma membrane. We believe that further use of this screen will continue to yield meaningful results about domains reactive at the plasma membrane involved in protein-protein and/or protein-lipid interactions.

In addition to identifying novel domains within ENaC reactive at or near the plasma membrane, we also further investigated the role of domains believed to be involved in channel internalization and ultimately targeted degradation. Several different laboratories have previously identified these tags as playing important roles in suppressing ENaC activity (reviewed in Ref. 24). Our results are consistent with these studies. However, it is not clear if the PY motif and the overlapping endocytic motif are functionally independent and how these domains influence the other's activity. Interestingly, a recent study reporting the solution structure of a Nedd4 WW domain-ENaC peptide complex suggests that the tyrosine and leucine within the endocytic YXXL motif are involved in coordinating Nedd4 binding to the PY motif in this channel subunit (53). The converse that residues within the PY motif influence the activity of the endocytic motif has not been investigated. Our results demonstrate for the first time that the endocytic motif is functional in the complete absence of the PY motif and in the absence of N-terminal ubiquitination sites. Moreover, the PY domain and endocytic tag were additive with respect to loss of complementation. We suggest that these results are consistent with the hypothesis, proposed first by others (32), that ENaC can be retrieved from the plasma membrane via one of two complementary routes: the lysosomal and proteosomal pathways. (In yeast, the vacuole is the terminal compartment for plasma membrane retrieval and endosomal transport-mediated degradation (reviewed in Refs.54, 55).) We argue further that although initial events of either pathway may promote the other and the end result of these pathways may be additive, ultimately they are, in part, functionally distinct and probably independently controlled with respect to modulation of ENaC turnover. Importantly, most mutations resulting in Liddle's syndrome disrupt/delete both the PY and endocytic motif, and thus the pathologically high Na⁺ reabsorption associated with this disease may result from decreased channel degradation in lysosomal as well as proteosomal compartments.

	ACKNOWLEDGEMENTS

We thank Drs. T. R. Kleyman and C. C. Cotton for sharing the mENaC cDNAs and mCT1 cell line, respectively.

	FOOTNOTES

^* This work was supported in part by NIDDK Grant R01-DK59594 from the National Institutes of Health (to J. D. S.), American Heart Association Grant SDG-0130008N, American Society of Nephrology Carl W. Gottschalk Research Scholar grant, American Physiological Society Lazaro J. Mandel Young Investigator award, and University of Texas Health Science Center, San Antonio, Howard Hughes Medical Institutes Cancer Research Enhancement Fund and Institutional Research grant.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by Frontiers in Physiology Science Teachers Summer Research Program.

^§ To whom correspondence should be addressed: Dept. of Physiology, 7756, University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX. 78229-2900. Tel.: 210-567-4332; Fax: 210-567-4410; E-mail: stockand@uthscsa.edu.

Published, JBC Papers in Press, July 1, 2002, DOI 10.1074/jbc.M204615200

	ABBREVIATIONS

The abbreviations used are: ENaC, epithelial Na⁺ channel; CHO, Chinese hamster ovary; DMEM, Dulbecco's modified Eagle's medium; EGFP, enhanced green fluorescent protein; TBS, Tris-buffered saline; hSOS, human Son-of-Sevenless.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES