| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute of Pharmacology and Toxicology, University of Lausanne, 1005 Lausanne, Switzerland; and
Institut National de la Santé et de la Recherche Médicale (INSERM), Unité 478, Institut Fédératif de Recherche 02, Faculté de Médecine Xavier Bichat, 75870 Paris Cédex 18, France
1Correspondence: Institute of Pharmacology and Toxicology, University of Lausanne, Rue du Bugnon 27, CH-1005 Lausanne, Switzerland. E-mail: olivier.staub{at}ipharm.unil.ch
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: Liddle’s syndrome • hypertension • Na+ transport • ubiquitination • HECT domain • ENaC • protein–protein interaction • WW domain • isoform
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
ENaC, which facilitates the entry of Na+ from the
lumen into the cells, is composed of three homologous subunits
(
ß
) (2
, 3)
that contain two transmembrane domains,
a large extracellular loop and short intracellular amino and carboxyl
termini (4
5
6)
. All three subunits have conserved
proline-rich regions conforming to the consensus of PY motifs (xPPxY;
P=proline, Y=tyrosine, x=any amino acid) in their carboxyl-terminal
tail (7)
. In the cortical collecting duct, ENaC is under
tight hormonal control, namely, by aldosterone and vasopressin, but the
molecular mechanisms of this regulation are not known. A functional
ENaC channel is essential as evidenced by the genetic linkage with two
diseases affecting blood pressure: pseudohypoaldosteronism type I
(8
, 9)
and Liddle’s syndrome (10)
. The
latter is an autosomal, dominant form of human hypertension
(11)
characterized by early onset of severe hypertension,
hypokalemia, metabolic alkalosis, and low plasma levels of renin and
aldosterone. The disease is caused by mutations in the genes encoding
either ß- (12)
or ENaC (13)
. To date,
all identified mutations are located in the regions encoding the
carboxyl termini of these subunits and they invariably delete or modify
the PY motifs (14
15
16
17)
, known to interact with WW domains
(protein–protein interaction modules). Using the two-hybrid technique,
we have identified the rat ubiquitin-protein ligase Nedd4 (rNedd4) as a
binding partner of the rat ENaC (rENaC) PY motifs (18)
.
rNedd4, which is composed of one C2 (calcium-dependent lipid binding)
domain, three WW domains, and a HECT (ubiquitin-protein ligase) domain,
binds rENaC via its WW domains. The identification of a
ubiquitin-protein ligase as an interacting partner of ENaC pointed at
ubiquitination (the modification of target proteins with ubiquitin
moieties) as a possible mechanism of ENaC regulation. Consistent with
the well-established function of ubiquitination as a signal for the
rapid internalization/degradation of membrane proteins (reviewed in
refs 19
20
21
), we demonstrated that rENaC is a protein with
a short half-life (T1/2 
1 h) that is
regulated by ubiquitination (22)
. Moreover, we and others
have provided evidence that Nedd4 is able to regulate ENaC activity
when coexpressed in oocytes by controlling the number of channels at
the cell surface (23
24
25)
. A relationship between an
epithelial Na+ channel and Nedd4 was also
demonstrated in salivary duct cells, although the nature of the
observed channel is unclear (26)
. However, though we found
that in Xenopus oocytes Xenopus Nedd4 (xNedd4)
does regulate rENaC, we were not able to show rENaC regulation and
in vivo binding by rat Nedd4 (rNedd4). This prompted us to
look for other mammalian, physiologically more relevant Nedd4 proteins. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
total RNA and
primers 5'-CCAGATGCTGCCACTCATTTGCCGC-3' (sense, nt 916–940 GenBank no.
D85414) and 5'-GCCTAAGACTCACTACACACTG-3' (antisense, nt 3509–3530) for
PCR. For the 5'-end we performed 5'-RACE-PCR (28)
, reverse
transcribing total RNA from mpkCCDc14 with random
primers, and PCR with nested primers 5'-ATCTCTGGAGTAGGGCACTGCTGGT-3'
(antisense, nt 1734–1710) and 5'-CAGCCTGACCACTATACTCGG-3' (antisense,
nt 1228–1208). Several cDNAs were cloned and sequenced and found to
correspond to the sequence reported by Kumar et al. (GenBank accession
no. D85414). One 5'-RACE clone was cloned into pGEMT vector (Promega,
Madison, Wis.), sequenced, found to range from nucleotide 169 to 1230
(GenBank accession no. D85414), and used for subsequent reconstitution
of a full-length mNedd4–1 cDNA. This was achieved by ligating an
EcoRI/XbaI 5' fragment (nt 169-1024) and a
XbaI/EcoRI 3' fragment (nt 1025–3530) into the
pSDeasy plasmid (29)
. A catalytically inactive mNedd4–1
mutant was generated by mutating the essential cysteine 850 to serine
using PCR (30)
. A full-length clone encoding mouse
Nedd4–2 was obtained as follows: Based on the sequence information
present in several expressed sequence tags (ESTs) (AI527754, AI315046,
AA118878), we amplified from mpkCCDc14 total RNA
3 overlapping PCR fragments by RT-PCR using the following primers:
1) 5'-GCTGTTGTTAAGGAAAACAGACC-3' (nt 1–23) and
5'-TCCATGGTTGGATCTTCTGTCGG-3' (319–341; antisense); 2)
5'-CCGACAGAAGATCCAACCATGGA-3' (319–341; sense) and
5'-GCTCTATTGTAAACAGCTGAGGACC-3' (2737–2761); and
3) 5'-AGCTCATACATGCTTTAATCGCCTT-3' (2790–2814)
and 5'-AAGCAGTGGTAACAACGCAGAGTTTTTTTTT-3'. The fragments were first
subcloned into pGEMT plasmid and sequenced, then assembled by triple
ligation of fragments EcoRI/NcoI (nt 1–335),
NcoI/SalI (nt 335-2548) and
SalI/NotI (nt 2548–3325) into the pSDeasy
plasmid linearized by EcoRI and NotI. A
catalytically inactive mutant mNedd4–2, generated by mutating cysteine
822 to serine by PCR (30)
and an amino-terminally
truncated mutant (mNedd4–2-
N; amino acids 1 to 288 deleted) were
cloned into the pSDeasy plasmid. The following residues in the WW
domains of Xenopus Nedd4 (31)
were mutated by
PCR: WW domain 1: tryptophan 218 to phenylalanine and proline 221 to
alanine; WW domain 2: tryptophan 409 to phenylalanine and proline 412
to alanine; WW domain 3: tryptophan 521 to alanine and proline 524 to
alanine; WW domain 4: tryptophan 572 to 575; the mutants were cloned
into pSDeasy. The rENaC constructs have been described previously by
Schild et al. (32)
and the constructs containing a FLAG
epitope by Firsov et al. (33)
.
Expression and function of rENaC channels in
Xenopus oocytes
For functional expression studies, rENaC and mNedd4 constructs
were transcribed using SP6-RNA polymerase and 10 ng cRNA encoding rENaC
(3.3 ng of each subunit) with or without 25 ng cRNA encoding mNedd4–1,
mNedd4–2, or xNedd4 were coinjected into Xenopus oocytes
(as described previously ref 23
). Electrophysiological
measurements were performed 12 to 24 h after cRNA injection using
the two-electrode voltage-clamp technique, as described before
(32)
, by measuring the current sensitive to 10 µM
amiloride in a 120 mM Na+ frog Ringer solution at
a holding potential of -100 mV. All values were normalized to the
control value (oocytes only injected with rENaC) in one given batch of
oocytes. Data are presented as mean ± SE.
The statistical significance of the differences between the means was
estimated using bilateral Student’s t test for unpaired
data.
Culture of mpkCCDc14 cells
Cells were grown in a hormonally defined DMEM:Ham’s F12 (1:1)
medium supplemented with 2% of fetal calf serum in an atmosphere of
humidified air/5% CO2 at 37°C on
collagen-coated Transwell permeable filters (Costar, Cambridge, Mass.)
as described by Bens et al. (27)
.
Biochemical analyses
After the electrophysiological measurements, oocytes were kept
at 4°C, pooled, and lysed in (20 µl/oocyte) Triton X-100
homogenization buffer (20 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1% Triton
X-100, 1 mM PMSF, 10 µg/ml leupeptin, 10 µg/ml peptstatin A, 10
µg/ml aprotinin) at 4°C. After centrifugation at 4°C for 10 min
at 14,000 rpm, the supernatant was recovered and stored at -80°C.
Lysate from mpkCCDc14 cells growing on filter
were prepared using lysis buffer (50 mM HEPES, pH 7.4, 150 mM NaCl, 1.5
mM MgCl2, 1 mM EGTA, 1% Triton X-100, 10%
glycerol, 1 mM PMSF, 10 µg/ml leupeptin, 10 µg/ml pepstatin A, 10
µg/ml aprotinin); insoluble material was removed by centrifugation.
For the coimmunoprecipitations, oocytes coinjected with cRNA encoding
rENaC (with or without a FLAG epitope on the extracellular loop; ref
33
) and mNedd4–1 or mNedd4–2 were incubated overnight in
0.1 mCi/ml [35S]methionine and homogenized in
50 mM HEPES, pH 7.4, 83 mM NaCl, 1 mM MgCl2, 1 mM
PMSF, 10 µg/ml leupeptin, 10 µg/ml pepstatin A, 10 µg/ml
aprotinin. An aliquot was removed for analysis of the total homogenate.
Membranes were prepared by centrifugation at 4°C for 10 min at 3000
rpm in a microfuge, followed by centrifugation of the supernatant at
4°C for 20 min at 14,000 rpm. The recovered membranes were
solubilized in 50 µl/oocyte of lysis buffer. Coimmunoprecipitations
were performed with anti-FLAG antibody (Kodak) and protein G Sepharose
(Sigma). The homogenate and the immunoprecipitated material was then
analyzed either by Western blotting or by autoradiography. For Western
blot analysis, material was separated either on 5–13% sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gradient gel or
on 7% SDS-PAGE minigels, transferred onto nitrocellulose, and then
immunoblotted as described before, using an anti-rNedd4-WW2 antibody at
a dilution of 1:1000 (18)
. Two-dimensional (2D) gel
electrophoresis using the mini-gel system of Bio-Rad Laboratories was
carried out as described (34)
.
Northern blot analysis
A multiple tissue Northern blot (MTN) containing 2 µg of
purified poly(A)+ RNA per lane from eight
different mouse tissues was obtained from Clontech (Palo Alto, Calif.).
The blot was prehybridized for 1 h at 65°C in ExpressHyb
(Clontech) according to the directions of the manufacturer. A
NdeI/XhoI fragment (nt 2830–3330) situated at
the 3' noncoding portion of mNedd4–2 cDNA, a mNedd4–1
HindIII/HindIII fragment (nt 1807–3497), or
ß-actin cDNA (provided by Clontech) was used as a random primed
32P-labeled probe. Hybridization was for 1 h
at 65°C. The blot was washed quickly four times with 2x SSC, 0.05%
SDS, three times for 10 min at room temperature with 2x SSC, 0.05%
SDS, and twice for 20 min at 65°C with 0.1x SSC and autoradiography
was performed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) acts as a negative regulator of the rat epithelial
Na+ channel (rENaC) when coexpressed in
Xenopus oocytes (23)
. Both xNedd4 and human
Nedd4 (hNedd4) contain four WW domains (31)
, but there are
only three in mouse (mNedd4) or rat Nedd4 (rNedd4) (18
, 35)
. Moreover, there is another human protein with homology to
xNedd4 (KIAA0439; GenBank accession no. AB007899). To determine whether
several Nedd4 isoforms exist, we performed a BLAST search
(36)
in the mouse EST GenBank database for mouse
homologues of Nedd4. Indeed, we found two populations of mNedd4 EST
sequences, one being identical to the original mNedd4 (37)
and the other one homologous to xNedd4 (31)
and human
KIAA0439. This finding suggested the existence of a novel Nedd4 isoform
in mice. Based on the sequence information available from the mouse
ESTs, we cloned the two mNedd4 isoforms from cultured mouse
mpkCCDc14 cells. These cells are derived from the
cortical collecting duct (CCD), a model epithelium for the study of
amiloride-sensitive Na+ transport
(27)
.
The first isoform is identical to the Nedd4 protein identified by Kumar
and collaborators (35
, 37)
(accession no. D85414), which
we propose to name mNedd4–1. It contains one C2, three WW, and one
HECT domain and is homologous to rNedd4 (rNedd4–1) and the human
KIAA0093 gene product (hNedd4–1; Fig. 1
). In contrast, the other mNedd4 isoform, referred to
here as mNedd4–2, is with 89% similarity to the hypothetical gene
product of human KIAA0439 gene (which we propose to name hNedd4–2) and
86% to xNedd4 yet clearly different from Nedd4–1 (52 to 58%
similarity, depending on the species). The sequence of the cloned
mNedd4–2 cDNA (3325 base pairs) predicts an open reading frame of 855
amino acids, starting at position 337 (GenBank accession no. AF277232)
and containing a translation initiation site, which fits the Kozak
consensus sequence well (38)
. Analyzing the deduced
protein sequence with the SMART tool program (39)
, four WW
and one HECT domain are predicted. Three of these WW domains (1, 2, and
4) are highly conserved among Nedd4 isoforms (Fig. 1)
. In contrast, the
third, additional WW domain is identical to the ones in hNedd4–2 and
xNedd4 and partially similar to the inserted WW domain in hNedd4–1
(Fig. 1)
. The predicted protein of mNedd4–2 does not contain the
amino-terminal C2 domain, as we consistently found in several clones
originating from different sources (mpkCCDc14 or
total mouse kidney) a stop codon upstream of the initiating methionine.
This stop codon is also present in all EST sequences covering this
region, three of which we have confirmed by sequencing.
|
Northern blot analysis on a multiple tissue blot, using a 3'-noncoding
region cDNA probe, reveals that a 3.8 kb mNedd4–2 mRNA is highly
expressed in liver and kidney, to a lesser extent in heart, brain, and
lung and is barely detectable in spleen, skeletal muscle, and testis
(Fig. 2
, top panel). The size and tissue distribution of mNedd4–2 mRNA are
different from mNedd4–1, which is ubiquitously expressed and has a
size of 6.0 kb (middle panel; see also ref 35
). From
the estimated size of the mNedd4–2 mRNA we cannot exclude the
possibility that our cDNA clones are not full length. Indeed, there are
at least two EST sequences in GenBank that contain additional
5'-sequences, but attempts to clone these regions failed.
|
mpkCCDc14 cells express both the mNedd4–1 and
mNedd4–2 protein
To detect mNedd4–1 and mNedd4–2 in
mpkCCDc14 cells, we performed immunoblot analysis
on mpkCCDc14 lysates using an antibody
directed against the WW domain 2 of rNedd4–1 (18)
(Fig. 3A
, lane 4). The antibody recognized two proteins with
apparent molecular masses of 115 kDa and 105 kDa, respectively. As
shown on the same blot, the 115 kDa band migrates at the same size as
mNedd4–1 when expressed in oocytes (Fig. 3A
, compare lanes
1 and 4) and the 105 kDa band is equal in size to mNedd4–2 (lanes 2 or
3 and 4). In addition, both proteins are recognized by three additional
anti Nedd4 antibodies (anti-rNedd4-HECT, anti-rNedd4-Nt, and anti
xNedd4; data not shown). To further corroborate that these bands
represent mNedd4–1 and mNedd4–2, we analyzed a
mpkCCDc14 cell lysate by 2D gel electrophoresis
and Western blotting (Fig. 3B
). mNedd4–1 has a calculated
molecular mass of 103 kDa and a pI of 5.12, as compared to mNedd4–2
with 97 kDa and a pI of 5.8. Several spots are seen at an apparent
molecular mass of 115 kDa that correspond to the size of Nedd4–1
and migrate in a pH range between 5 and 5.2. These spots probably
represent differentially phosphorylated forms of mNedd4–1, as we have
found that mNedd4–1 is a phosphoprotein (E. Kamynina and O. Staub,
unpublished results). At 105 kDa we observe a protein that migrates
at a more basic pH of 5.8, the predicted pI of mNedd4–2. The same
spots are also recognized by an antibody directed against rNedd4–1
HECT (data not shown). Hence, both mNedd4–1 and mNedd4–2 proteins are
expressed in mpkCCDc14 cells.
|
mNedd4–2, but not mNedd4–1, regulates rENaC in
Xenopus oocytes
To determine whether mNedd4–2 has the potential to regulate
rENaC, we expressed either wild-type mNedd4–2 or a catalytically
inactive mutant in Xenopus oocytes. In the mutant, the
cysteine 822 in the HECT domain was replaced by serine (mNedd4–2 CS).
The correct expression of the wild-type and mutant Nedd4 was confirmed
by Western blot analysis (Fig. 3A
). We then measured
amiloride-sensitive Na+ currents in these oocytes
(Fig. 4
). We found that wild-type mNedd4–2 was able to abrogate rENaC activity
when compared to control oocytes (rENaC/H2O)
whereas the inactive mNedd4–2 mutant increased this activity by more
than twofold, likely by competing against endogenous xNedd4
(23)
. An amino-terminally truncated mNedd4–2 protein
(mNedd4–2-N) was also able to down-regulate rENaC (see below). On
the other hand, when wild-type mNedd4–1 was coexpressed with rENaC,
channel activity was not altered (Fig. 4A
, rENaC/mNedd4–1),
whereas the corresponding CS mutant had a minor stimulatory effect
(rENaC/mNedd4–1 CS). Both mNedd4–1 proteins were correctly expressed
(not shown). Similar data were obtained with rNedd4–1 or hNedd4–1 (E.
Kamynina and O. Staub, unpublished results), reinforcing the finding
that Nedd4–1 is not primarily involved in rENaC regulation.
|
mNedd4–1 does not bind to rENaC when coexpressed in
Xenopus oocytes
To understand why Nedd4–1 does not regulate rENaC, we
investigated whether differences in rENaC binding between the two
isoforms exist. We performed coimmunoprecipitation experiments, by
coexpressing rENaC, that either contained (Fig. 5A
, lanes 4–6) or not (lanes 1–3) a FLAG epitope within its
extracellular loop (33)
, with mNedd4–1 or mNedd4–2.
Oocytes were metabolically labeled, a membrane fraction was solubilized
and immunoprecipitated with an anti-FLAG antibody. The precipitated
proteins were then analyzed by SDS-PAGE and autoradiography and/or
Western blotting. All three rENaC subunits were precipitated when the
FLAG epitope was present (Fig. 5A
, top panel, lanes 4–6).
Moreover, when mNedd4–2 was coexpressed, an additional band at 105
kDa was seen (top panel, lane 6), which we identified as mNedd4–2 by
Western blotting using an anti-Nedd4 antibody (middle panel, lane 6).
No protein coimmunoprecipitated when mNedd4–1 cRNA was injected (top
and middle panel, lane 5), despite equal expression levels with
mNedd4–2 (bottom panel, compare lanes 5 and 6). This suggests that the
slight stimulatory effect of the mNedd4–1 CS mutant on ENaC activity,
as observed in Fig. 4
, is indirect, likely by affecting cellular
mechanisms such as endocytosis. We conclude that mNedd4–1 does not
have the potential to directly regulate rENaC, because it does not bind
to the channel complex when coexpressed in oocytes. When we performed
the same experiment with either wild-type or mutant Nedd4–2, we did
not observe a difference in either the quantity of the total pool of
rENaC subunits or the amount of Nedd4–2 bound to rENaC (Fig. 5B
). Because rENaC cell surface expression is down-regulated
by xNedd4 (the Xenopus homologue of mNedd4–2)
(23)
, it is likely that the primary interaction between
Nedd4–2 and rENaC observed in these experiments occurs at
intracellular locations.
|
Each rENaC PY-motif is involved in Nedd4–2 binding
We wanted to determine how the different PY motifs on the three
rENaC subunits contribute to the binding to Nedd4–2. We therefore
systematically mutated the tyrosine residues in each PY motif,
mutations expected to abolish the interaction with WW domains, and
analyzed the effect of Nedd4–2 (wild-type or CS mutant) on these
mutant channels (Fig. 6
). As previously reported (32)
, the sequential removal of
PY motifs leads to an increase of ENaC activity (Fig. 6A
,
columns 1–8), as revealed by a comparison of wild-type channels (white
column 1) to channels with one PY motifs deleted (gray, columns 2–4),
two PY motifs deleted (striped, columns 5–7), or three PY motifs
deleted (black, column 8). When wild-type mNedd4–2 is coexpressed
(columns 9–16), the negative effect of mNedd4–2 is progressively lost
as the number of mutated PY motifs increases (compare white column 9 to
gray columns 10–12, striped columns 13–15 or black column 16).
Consistently, the same PY motif-dependent pattern can be seen with the
mNedd4–2 CS mutant (columns 17–24). We also note that the presence of
a single PY motif is sufficient to yield an effect on rENaC currents
(Fig. 6A
, compare striped columns 5–7 to 13–15 or 21–23)
and interaction with Nedd4–2 (Fig. 6B
, lanes 3–5). It also
appears that the PY motif on ß rENaC is more important for rENaC
function than those on and (Fig. 6A
, compare gray
columns in each condition to each other or the striped columns in each
condition). When all the PY motifs are deleted, there is not more
interaction with mNedd4–2 (Fig. 6B
, lane 2) and no
significant difference between the control condition
H2O/Y673AßY618H628A
(Fig. 6A
, column 8) and
mNedd4–2/Y673AßY618H628A
(Fig. 6A
, column 16) or between control and
mNedd4–2-CS/Y673AßY618H628A
(Fig. 6A
, compare columns 8 and 24). Thus, these data (Fig. 6)
demonstrate that there is an absolute requirement for at least one
PY motif on rENaC in order to observe an effect of Nedd4–2 on rENaC
regulation. We note that patients with Liddle’s syndrome, possessing a
mutation in a single PY motif, are clearly affected. It is therefore
possible that through overexpression we may reach the limits of the
oocyte system.
|
The WW domains 3 and 4 mediate binding to rENaC
Having established that all PY motifs are involved in
rENaC/Nedd4–2 interaction we asked which WW domains play a role in
rENaC binding. Such experiments were done with the Xenopus
homologue (xNedd4) of mNedd4–2, which is highly homologous and
contains nearly identical WW domains. We individually mutated each WW
domain of xNedd4, the second conserved tryptophan to phenylalanine and
the conserved proline into alanine. These mutations are expected to
abolish binding to PY motifs (40
, 41)
. These mutant xNedd4
proteins were then coexpressed in Xenopus oocytes and their
effect on rENaC activity was analyzed (Fig. 7A
). We found that mutation of the first WW domain in xNedd4
did not effect xNedd4-dependent rENaC regulation, whereas mutation of
the second WW domain had a minor effect and changing the third or
fourth WW domain had more significant consequences for the xNedd4/rENaC
interaction (Fig. 7A
). We then made double mutants in which
we mutated the first and second WW domains together or, alternatively,
the third and fourth together (Fig. 7B
). We expressed these
constructs in a dose-dependent manner and observed that mutation of the
first two WW domains did not impair the capability of Nedd4–2 to
down-regulate rENaC. On the other hand, mutation of the third or fourth
WW domain completely abolished the potency of Nedd4–2 to control rENaC
activity, suggesting that WW domains 3 and 4 are principally involved
in rENaC binding. This was further validated with a truncated form of
mouse Nedd4–2, which had the first two WW domains deleted (Fig. 4
,
rENaC/mNedd4–2N); this mutant was as effective as wild-type
mNedd4–2 in regulating rENaC, corroborating the observation that
essentially the third and fourth WW domains are important for rENaC
binding.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 37)
, contains one C2, three WW, and one HECT domain,
mNedd4–2 has no C2, but has four WW and one HECT domain. Both isoforms
have homologues in humans and rats. In Xenopus laevis, only
a Nedd4–2 homologue has been identified (see Fig. 1
).
Intriguingly, mNedd4–2 does not contain the C2 domain. In all cDNA
clones analyzed, we consistently found a stop codon upstream of the
initiating methionine, which fits well the consensus sequence for
translation initiation (38)
. This stop codon is present in
all mouse EST sequences covering this region, making it unlikely that
there is a sequencing or cloning artifact. In support of this,
expression of mNedd4–2 lacking the C2 domain in Xenopus
oocytes yields a protein of similar size to the one detected by several
anti-Nedd4 antibodies in mpkCCDc14 (Fig. 3A
). When a mpkCCDc14 cell lysate is
separated on 2D gels, a protein migrates at the predicted molecular
weight and pI and is recognized by the anti Nedd4 antibodies (Fig. 3B
). Hence, mNedd4–2 expressed in
mpkCCDc14 does not contain a C2 domain at its
NH2 terminus. Moreover, our data show that
mNedd4–2 is fully functional in regulating rENaC when coexpressed in
Xenopus oocytes, indicating that, at least in these cells,
the C2 domain is not required for rENaC control. The lack of a C2
domain suggests that localization of mNedd4–2 is not regulated by
intracellular calcium levels as has been shown for Nedd4–1 in MDCK
cells (42)
and that regulation of rENaC by intracellular
calcium (43)
does involve factors other than mNedd4–2. It
is noteworthy that the presence of a C2 domain is also uncertain in
hNedd4–2, as the first methionine in this sequence corresponds to the
one in mNedd4–2; however, the sequence in GenBank may only be partial.
We demonstrate for the first time functional and biochemical
specificity among close members of the Nedd4 family of
ubiquitin-protein ligases: the observation that only mNedd4–2, but not
Nedd4–1 (from either mouse, rat, or human; data not shown), binds and
regulates rENaC illustrates that the presence of WW domains is not
sufficient for a protein to interact with any given PY motif-containing
proteins. Other factors, such as the nature of the WW domains or of the
PY motif, sterical hindrance, or cofactors like E2 enzymes, may play a
role. Our results also demonstrate the importance of in vivo
data as compared with in vitro binding analysis. We and
others have shown by various approaches, including the two-hybrid
system, binding/competition assays, and coimmunoprecipitation from
cells overexpressing rENaC, that the individual WW domains of
Nedd4–1 can bind to the PY motifs of rENaC (18
, 25
, 41
, 44)
. These results must now be reevaluated. It is important to
determine, for example, whether there is a difference in the binding
affinities to rENaC between the WW domains of Nedd4–1 and Nedd4–2.
Moreover, it will be necessary to develop methods that determine the
affinity between Nedd4 isoforms and the entire assembled rENaC complex.
It also follows that the two isoforms, which are expressed within the
same cells (mpkCCDc14), are not functionally
redundant and that the physiological substrate(s) for Nedd4–1 remains
to be identified. This is further supported by the observation that the
two isoforms exhibit different patterns of expression. Whereas
mNedd4–1 is expressed in all tissues analyzed (Fig. 2
, middle panel),
mNedd4–2 mRNA is more restricted, with highest expression in the liver
and kidney and less abundant in the other tissues examined. The
observation that mNedd4–2 expression in the lung is very weak
correlates well with the fact that no lung phenotype is observed in
Liddle’s patients.
Our results showing that Nedd4–1 does not regulate rENaC activity
contrast with the data published by Goulet et al. (24)
,
who reported that rNedd4 suppresses human ENaC activity by 80% when
coexpressed in Xenopus oocytes and that the catalytically
inactive rNedd4 mutant (rNedd4-C854S) does not influence rENaC
activity. We cannot explain this discrepancy, as Goulet et al.
(24)
provide very little information on the forms of Nedd4
they used. Nevertheless, one would expect that an enzymatically
inactive mutant protein, which preserves the binding sites for its
substrate, would enter into competition with endogenous xNedd4 (present
in the oocyte) (23)
and should therefore increase rENaC
activity.
The data in this paper provide fascinating insights into the
mechanism on how Nedd4–2 binds to and regulates rENaC. An interesting
point is revealed by the finding that the same amounts of rENaC are
immunoprecipitated from oocytes coinjected with rENaC and either
wild-type or mutant Nedd4–2 (Fig. 5B
). This is remarkable
since expression of Nedd4–2 leads to the complete inhibition of rENaC
activity, which is likely due to the loss of rENaC channels at the cell
surface (23)
. Two explanations are plausible: either the
amount of cell surface rENaC, which becomes ubiquitinated and degraded,
is negligible compared to the total cellular pool or rENaC is recycled
after internalization and deubiquitination. The finding that comparable
amounts of wild-type or mutant Nedd4 coimmunoprecipitate with rENaC
suggests that Nedd4–2 binds to intracellular ENaC channels and that
this binding does not lead to the immediate ubiquitination/degradation
of the rENaC complex. We speculate that inactive mNedd4–2 may be bound
to intracellular rENaC and would be activated by some unknown mechanism
on translocation to the plasma membrane. Such activation will then lead
to ubiquitination of rENaC, dissociation of mNedd4–2, and
internalization of the channel complex.
The experiments done with either an amino-terminally truncated
form of mNedd4–2 (Fig. 4)
or xNedd4-containing mutations in the
individual WW domains (Fig. 7)
suggest that mainly the third and fourth
WW domains, but not the first and second, are involved in rENaC
binding. On the other hand, the three PY motifs (or even four, if one
considers the channel stoichiometry of 21ß1; ref
45
) seem to play a role for Nedd4 binding. It appears that
the ß PY motif is functionally different from those on or
rENaC, as it causes a lower INa when present
alone as compared to the other PY motifs (compare columns 5–7, or
13–15, or 21–23) or a higher current when mutated alone (compare
columns 2–4, or 10–12, or 18–20). These effects seem to be
independent of mNedd4–2 and may be due to changed open probability as
has been demonstrated for Liddle’s channels (33)
. Taken
together, our data suggest that there are several functionally relevant
binding modes for the interaction with Nedd4, which also include the PY
motif of rENaC, but the nature of these interactions remain to be
determined experimentally.
Our finding that all three PY motifs are involved in interaction with
Nedd4 have important implications for understanding Liddle’s syndrome.
To date, all of the identified mutations causing Liddle’s syndrome
either delete or mutate one PY motif on ß or ENaC, thereby
destroying one binding site for Nedd4. To our knowledge, no mutations
on ENaC or Nedd4 that would cause Liddle’s syndrome have been
reported, although we predict that such mutations exist. In the case of
Nedd4–2, this may be simply because it is an essential protein (as is
the case in yeast; see ref 46
), and hence null mutations
may be lethal. It is possible, however, that investigations have
focused so far on the wrong genetic locus (encoding the classical
hNedd4–1), that localized on chromosome 15. Searching for mutations in
the gene encoding for hNedd4–2, situated on chromosome 18, will likely
be more promising.
In summary, we have shown that a novel mouse Nedd4 isoform (mNedd4–2), and not the originally identified mNedd4–1, does regulate rENaC when coexpressed in Xenopus oocytes and that the lack of regulatory competence of mNedd4–1 is mainly due to its inability to bind to the rENaC complex. Moreover, we showed that all the PY motifs on the rENaC complex and WW domains 3 and 4 on mNedd4–2 are involved in the interaction. Hence, our data suggest that Nedd4–2 is a novel suppressor protein of rENaC and therefore a susceptibility gene for arterial hypertension.
We thank Drs. Bernard Rossier, Laurent Schild, Jean-Daniel Horisberger, Dmitri Firsov, Lukas Müller, Phil Shaw, and Marc Andrew Thomas for critically reading the manuscript, Laurent Schild, Ivan Gautschi, and Estelle Schneeberger for rENaC mutants, Carole Münster for excellent technical assistance, and Dr. Silvia Barcelo for help with the 2D gel analysis. This work was supported by the Swiss National Science Foundation (31–52178.97), the Leenards Foundation, the Fondation Emma Muschamp, the Novartis Foundation, and the Fondation de 450e anniversaire de l’Université de Lausanne.
Received for publication April 17, 2000.
Revision received June 19, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and subunits of the Xenopus epithelial Na+ channel provide information about the amiloride binding site and extracellular sodium sensing. Proc. Natl. Acad. Sci. USA 94,5949-59541This article has been cited by other articles:
|
|
 |
|
  R. Kabra, K. K. Knight, R. Zhou, and P. M. Snyder Nedd4-2 Induces Endocytosis and Degradation of Proteolytically Cleaved Epithelial Na+ Channels J. Biol. Chem., March 7, 2008; 283(10): 6033 - 6039. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Mueller, O. B. Kashlan, J. B. Bruns, A. B. Maarouf, M. Aridor, T. R. Kleyman, and R. P. Hughey Epithelial Sodium Channel Exit from the Endoplasmic Reticulum Is Regulated by a Signal within the Carboxyl Cytoplasmic Domain of the {alpha} Subunit J. Biol. Chem., November 16, 2007; 282(46): 33475 - 33483. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-B. Peng and D. G. Warnock WNK4-mediated regulation of renal ion transport proteins Am J Physiol Renal Physiol, October 1, 2007; 293(4): F961 - F973. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Zhou, S. V. Patel, and P. M. Snyder Nedd4-2 Catalyzes Ubiquitination and Degradation of Cell Surface ENaC J. Biol. Chem., July 13, 2007; 282(28): 20207 - 20212. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Pao, J. A. McCormick, H. Li, J. Siu, C. Govaerts, V. Bhalla, R. Soundararajan, and D. Pearce NH2 terminus of serum and glucocorticoid-regulated kinase 1 binds to phosphoinositides and is essential for isoform-specific physiological functions Am J Physiol Renal Physiol, June 1, 2007; 292(6): F1741 - F1750. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Jespersen, M. Membrez, C. S. Nicolas, B. Pitard, O. Staub, S.-P. Olesen, I. Baro, and H. Abriel The KCNQ1 potassium channel is down-regulated by ubiquitylating enzymes of the Nedd4/Nedd4-like family Cardiovasc Res, April 1, 2007; 74(1): 64 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Ring, S. X. Cheng, Q. Leng, K. T. Kahle, J. Rinehart, M. D. Lalioti, H. M. Volkman, F. H. Wilson, S. C. Hebert, and R. P. Lifton WNK4 regulates activity of the epithelial Na+ channel in vitro and in vivo PNAS, March 6, 2007; 104(10): 4020 - 4024. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Bhalla, N. M. Oyster, A. C. Fitch, M. A. Wijngaarden, D. Neumann, U. Schlattner, D. Pearce, and K. R. Hallows AMP-activated Kinase Inhibits the Epithelial Na+ Channel through Functional Regulation of the Ubiquitin Ligase Nedd4-2 J. Biol. Chem., September 8, 2006; 281(36): 26159 - 26169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Sorkina, M. Miranda, K. R. Dionne, B. R. Hoover, N. R. Zahniser, and A. Sorkin RNA Interference Screen Reveals an Essential Role of Nedd4-2 in Dopamine Transporter Ubiquitination and Endocytosis J. Neurosci., August 2, 2006; 26(31): 8195 - 8205. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Liang, K. W. Peters, M. B. Butterworth, and R. A. Frizzell |
