(Received for publication, January 16, 1997, and in revised form, March 31, 1997)
From the Departments of Medicine and Physiology, The epithelial brush border
Na+/H+ exchanger isoform 3 (NHE3) is
regulated by growth factors and protein kinases. When stably expressed
in PS120 fibroblasts, NHE3 is stimulated by serum and fibroblast growth
factor (FGF) and inhibited by phorbol esters. To examine the role of
phosphorylation of NHE3 in growth factor/protein kinase regulation,
NHE3 was C-terminally tagged with an 11-amino acid epitope of the
vesicular stomatitis virus glycoprotein (VSVG) and stably expressed in
Na+/H+ exchanger null PS120 fibroblasts
(PS120/NHE3V). NHE3V was regulated by serum, FGF, and phorbol ester in
a manner identical to wild type non-VSVG-tagged NHE3. Phosphorylation
of NHE3V was evaluated via immunoprecipitation with anti-VSVG antibody
after in vivo labeling of PS120/NHE3V cells with
[32P]orthophosphate. NHE3V was phosphorylated under basal
conditions. However, FGF and PMA, under conditions in which these
agonists regulate NHE3V, altered neither the amount of phosphorylation of NHE3V as analyzed by one-dimensional SDS-polyacrylamide gel electrophoresis and autoradiography nor two-dimensional phosphopeptide maps of tryptic digests of NHE3V. In contrast, while changes in NHE3V
phosphorylation were not observed with serum exposure by one-dimensional SDS-polyacrylamide gel electrophoresis, two-dimensional studies showed increases in two phosphopeptides. Under all these conditions, phosphoamino acid analysis showed that NHE3V was
phosphorylated only on serine residues. By cell surface protein
biotinylation studies under basal conditions, at least 27% of the
NHE3V was expressed on the cell surface. To further analyze the
phosphorylation status of the surface and intracellular forms of NHE3V
under basal conditions and determine whether the amount of
phosphorylation of the surface form changes upon serum, FGF, and PMA
regulation, the surface form of NHE3V was separated from intracellular
form by biotinylation/avidin-agarose precipitation. Under basal
conditions, both intracellular and surface forms of NHE3V were
phosphorylated. However, the amount of phosphorylation of the surface
form of NHE3V did not change upon stimulation by serum and FGF and
inhibition by PMA based on one-dimensional SDS-polyacrylamide gel
electrophoresis and autoradiography. Thus, we conclude that when
expressed in PS120 cells, while NHE3 is a phosphoprotein under basal
conditions, its regulation by FGF and PMA is not by changes in the
phosphorylation of NHE3, while regulation by serum may involve changes
in its phosphorylation. Regulation of NHE3 probably involves
intermediate associated regulatory proteins. The function of basal
phosphorylation of NHE3 is not known.
The mammalian Na+/H+ exchanger
(NHE)1 gene family is made up of plasma
membrane transport proteins that are involved in the regulation of
intracellular pH (pHi), cell growth, cell volume, and
transcellular Na+ absorption (1, 2). Five mammalian
Na+/H+ exchanger isoforms have been identified
with different tissue distributions and functional properties (2, 3),
and three of these have been expressed and characterized in detail
(NHE1, NHE2, NHE3). By Northern blot analysis and immunocytochemistry, NHE1 is present in nearly all cell types and is localized to the basolateral surfaces of renal and intestinal polarized epithelial cells
(4), whereas NHE2 and NHE3 are predominantly epithelial and are
co-localized to the brush border of these cells (5).
NHE3 is stimulated by FGF, serum, and okadaic acid but inhibited by
phorbol esters and by calmodulin, the latter under basal [Ca2+]i conditions (6-8). Unlike NHE1, the
regulation of NHE3 is through changes in the maximum exchange rate
(Vmax) rather than an alteration in
K [32P]Orthophosphoric acid was from
NEN Life Science Products. Fetal bovine serum was from Hyclone Corp.
(Logan, UT), basic FGF was from Boehringer Mannheim, and the rest of
the reagents were purchased from Sigma. The polyclonal (rabbit) and
monoclonal (mouse) anti-VSVG antibodies were generous gifts from Drs.
D. Louvard and T. Kreis. The polyclonal anti-phosphotyrosine antibody
was obtained from Zymed. NHS-SS-biotin and avidin-agarose were from Pierce. Protein A-Sepharose 6M beads were from Pharmacia Biotech Inc.
Monoclonal anti-actin antibody (clone JLA20) was from Calbiochem.
PS120 cells, Chinese hamster lung fibroblasts
deficient in endogenous Na+/H+ exchange-
(originally donated by J. Pouyssegur) and NHE3V- (NHE3 epitope tagged
with an 11-amino acid VSVG epitope; see below) transfected PS120 cells
were grown in Dulbecco's modified Eagle's medium supplemented with 25 mM Hepes, 50 IU/ml penicillin, 50 µg/ml streptomycin, and
10% fetal bovine serum in the presence of 5% CO2. Acid
killing and G418 (400 µg/ml) selections were applied to
NHE3V-transfected PS120 cells as described previously to maintain high
Na+/H+ exchange activity (6, 8).
To construct the VSVG epitope tag on the C terminus of
NHE3 (NHE3V), we made use of an NHE3 truncation mutant, E3/585, which had been VSVG-tagged on the C terminus. E3/585 is a cDNA of NHE3 truncated from the C terminus at amino acid 585. Briefly, E3/585V cDNA was obtained by cleaving the untagged E3/585 cDNA from the pMAMneo vector with BamHI and XhoI. This
BamHI/XhoI cassette was subcloned into the 5 PS120 cells were transfected with the NHE3V/pMAMneo construct using the
calcium phosphate precipitation technique and then subjected to acid
selection 3 days after the transfection. This procedure was repeated
every 3-4 days over a period of 3 weeks. Geneticin (800 µg/ml) was
present on initial splitting in the cell culture medium to further
enhance selection pressure (5, 10).
Cellular Na+/H+ exchange
activity in PS120/NHE3V cells was determined fluorometrically using an
intracellular pH-sensitive dye BCECF with a computerized fluorimeter
and a perfusion system with cells grown on glass coverslips and
serum-starved overnight, as described previously (6-8). The effect of
FGF (10 ng/ml) and PMA (1 µM) on NHE3 was studied by
initial rates. The effect of 10% dialyzed FBS on NHE3V was studied by
the initial rates and at steady-state pHi, as described
previously (6-8). In these studies, the rate of
Na+-dependent alkalinization was obtained by
calculating the first order derivative of the
Na+-dependent pH recovery curve. The hydrogen
ion efflux rate (µM H+/s), equivalent to the
rate of Na+/H+ exchange, was then determined by
multiplying the rate of change in pHi by the cellular buffering
capacity at the corresponding pHi values. Scatter plots of
H+ efflux rate versus intracellular
[H+] were constructed (6). Na+/H+
exchange rate data were analyzed using a nonlinear regression data
analysis program, ENZFITTER (Biosoft Corp.). In each experiment, control cells were studied at the same time in parallel with treated cells (FGF, 10 ng/ml, or PMA, 1 µM) to control for the
variability in the basal exchange rate among cells from different cell
passages and acid selection.
Confluent PS120 or PS120/NHE3V
cells grown in 10-cm dishes were washed twice with phosphate-free
Dulbecco's modified Eagle's medium. Cells were then labeled in
vivo for 4 h with 3 ml of phosphate-free Dulbecco's modified
Eagle's medium containing 2.5 mCi of
[32P]orthophosphate. At the end of the incubation, cells
were treated with 10% dialyzed FBS, PMA (1 µM), or FGF
(10 ng/ml) for 5 min. It has previously been shown that under such
conditions, these growth factors affect Na+/H+
exchange by PS120/NHE3 cells (6-8). In the present study, we also
tested whether these growth factors regulate PS120/NHE3V cells under
the conditions used for phosphorylation. Cells were then washed with
ice-cold phosphate-buffered saline at the end of the incubation period
and processed immediately for immunoprecipitation.
All procedures of immunoprecipitation
were performed at 4 °C. The confluent layer of cells labeled with
[32P]orthophosphate as described above were scraped and
resuspended in 500 µl of 50 mM Hepes/Tris, pH 7.4, 150 mM NaCl, 3 mM KCl, 25 mM sodium
pyrophosphate, 5 mM EDTA, 2 mM EGTA, 1 mM Na3VO4, and protease inhibitors
(0.1 mM phenylmethylsufonyl fluoride, 1 mM
phenanthroline, and 1 mM iodoacetamide) (IP buffer)
(14-16). Cells were then collected by centrifugation for 10 min at
12,000 × g in an Eppendorf centrifuge and were
resuspended in IP buffer containing 1% Triton X-100 (IPT buffer),
sonicated for 20 s, and agitated on a rotating rocker at 4 °C
for 30 min, followed by centrifugation at 12,000 × g
for 30 min to remove insoluble cellular debris. The supernatants were
first precleared with Protein A-Sepharose 6M beads by rocking for
1 h. The beads were spun down, and the supernatants were incubated
overnight with 5 µl of the anti-VSVG polyclonal antibody. Protein
A-Sepharose beads previously treated with PS120 cell extract
solubilized by Triton X-100 (1%) were then added and allowed to rock
for a further 2 h. The beads were then washed eight times with IPT
buffer. Immunoprecipitated proteins were eluted by boiling in 70 µl
of Laemmli sample buffer. For each sample, an aliquot of 50 µl was
analyzed by electrophoresis on a 10% SDS-polyacrylamide gel, and
autoradiography on a dried gel was performed.
The phosphoprotein corresponding to NHE3 was further characterized by
two-dimensional tryptic phosphopeptide mapping and phosphoamino acid
analysis.
The band
corresponding to NHE3 was identified by autoradiography and prestained
high molecular weight markers (Bio-Rad), excised, and washed (10%
methanol, 5% glacial acetic acid and then washed with 50% methanol).
The gel pieces were incubated with 0.3 mg/ml L-1-tosylamide-2-phenylethylchloromethyl ketone
(TPCK)-treated trypsin in 0.4% NH4HCO3 at
37 °C overnight. The digested proteins were lyophilized and
resuspended in 10 µl of H2O and spotted at the origin on
thin layer cellulose plates together with 0.5 µl each of the marker
dyes, phenol red and basic fuschin. Phosphopeptides were separated by
electrophoresis in acetic acid/pyridine/H2O (10:1:89) (pH
3.5, 500 V) in the first dimension until the marker dyes had travelled
~6 cm from the origin in opposite directions, and then by ascending
chromatography in pyridine/butanol/acetic acid/H2O
(15:10:3:2) in the second dimension until the marker dyes had ascended
to the edge of the thin layer plate. Analysis of the amount of
phosphorylation of each phosphopeptide was by use of a PhosphorImager
and ImageQuant software, with a comparison made by normalization to
spots felt not to change in phosphorylation in response to serum, PMA,
and FGF, based on preliminary studies. Comparison was by paired
t tests.
To determine the identity of the
phosphoamino acids of NHE3, 32P-labeled protein was excised
from acrylamide gels, treated with trypsin, and lyophilized as
described above. The lyophilized protein was further hydrolyzed in 6 M HCl under nitrogen at 105 °C for 1 h. The
acid-hydrolyzed peptides were relyophilized, resuspended in 10 µl of
H2O, and spotted on thin layer cellulose plates with 100 µg each of phosphoserine, phosphothreonine, and phosphotyrosine used
as internal standards and 0.5 µl of phenol red marker also added at
the origin. The plate was subjected to the first phase of
electrophoresis in formic acid/acetic acid/H2O, (10:1:89)
(pH 1.9, 500 V) until the phenol red had moved ~5 cm from the origin, and then electrophoresis was continued in the same direction until the
marker had travelled another ~8 cm in another tank containing acetic
acid/pyridine/H2O (10:1:89) (pH 3.5, 500 V). The cellulose plates were dried, developed in 1% ninhydrin in acetone to identify the internal phosphoamino acid standards, and then subjected to autoradiography to reveal the 32P-labeled phosphoamino
acids.
Cell surface biotinylation was
performed at 4 °C. PS120/NHE3V cells were grown to confluence in
10-cm Petri dishes. Cells were washed twice in phosphate-buffered
saline (150 mM NaCl, 20 mM
Na2HPO4, pH 7.4) and once in borate buffer (154 mM NaCl, 10 mM boric acid, 7.2 mM
KCl, 1.8 mM CaCl2, pH 9) (17, 18). The surface
plasma membrane proteins were then biotinylated by gently shaking the
cells for 20 min with 3 ml of borate buffer containing 1.5 mg of
NHS-SS-biotin (biotinylation solution). An additional 3 ml of the same
biotinylation solution was then added, and the cells were rocked for an
additional 20 min. The biotinylation solution was then discarded, and
the cells were washed extensively with the quenching buffer (20 mM Tris and 120 mM NaCl, pH 7.4) to scavenge
the unreacted biotin, and then the cells were washed twice with
phosphate-buffered saline. Cells were then scraped and solubilized with
500 µl of IPT buffer and were then sonicated for 20 s and
agitated on a rotating rocker at 4 °C for 30 min, followed by
centrifugation at 12,000 × g for 30 min to remove insoluble cellular debris. The supernatant was then incubated with
avidin-agarose to separate the biotinylated proteins from nonbiotinylated proteins by binding the former to avidin-agarose. Separation and analysis of the surface form from the intracellular form
of NHE3V was performed as detailed in the figure legends.
Regulation of NHE3V by Serum, PMA, and FGF Regulation of NHE3V by serum, FGF, and PMA.
A, NHE3V/PS120 cells were acidified by NH4Cl
prepulse and were allowed to recover in the presence of sodium medium
to a steady-state pHi. The addition of 10% dialyzed FBS (added
at the time indicated by
Since NHE3V is regulated by serum,
PMA, and FGF in the same way these growth factors/protein kinases
regulate NHE3 and the anti-VSVG antibody allowed quantitative
immunoprecipitation of NHE3V, we determined whether NHE3V is a
phosphoprotein under basal conditions and whether serum, FGF, and PMA
change the amount of its phosphorylation. PS120/NHE3V cells were
labeled in vivo with [32P]orthophosphate, and
NHE3V was immunoprecipitated with a polyclonal anti-VSVG antibody. The
immunoprecipitated NHE3V was then analyzed by SDS-PAGE and
autoradiography. As shown in Fig. 2A, under
basal conditions, the rabbit polyclonal anti-VSVG antibody
immunoprecipitated a single phosphoprotein of 85 kDa, which was
recognized by the mouse monoclonal anti-VSVG antibody by Western
blotting (Fig. 2B). This is the same size of the protein
recognized on Western analysis by anti-NHE3 antibody (5). There was no
detectable phosphoprotein immunoprecipitated by anti-VSVG antibody from
untransfected PS120 cells (Fig. 2A). These results indicate
that anti-VSVG antibodies specifically immunoprecipitate NHE3V, and
NHE3V is an 85-kDa phosphoprotein under basal conditions.
In contrast to our original expectation, as shown in Fig.
2A, there was no change in the degree of total
phosphorylation of NHE3V in response to a 5-min exposure to dialyzed
FBS (10%), PMA (1 µM), or FGF (10 ng/ml). These findings
were obtained in a series of five experiments, in which these
regulators caused no consistent change in the amount of phosphorylation
of NHE3. For instance, while serum apparently caused a small increase
in phosphorylation of NHE3 in Fig. 2A, in other studies such
as in Fig. 2B, this did not occur. Also, direct Cerenkov
counting of the lyophilized immunoprecipitated digests gave similar
counts of radioactivity among the control and samples treated with FBS,
FGF, and PMA (data not shown).
To eliminate the possibility that any potential differences in
phosphorylation could have been masked by unequal amounts of NHE3V
being immunoprecipitated under various conditions, an aliquot of the
immunoprecipitated NHE3V was separated by SDS-PAGE and transferred onto
nitrocellulose. Then, sequentially on the same blots, autoradiography
followed by Western analysis was performed to analyze the amount of
NHE3V phosphorylation and protein, respectively, in the same specimen
(Fig. 2B). It was shown that the amount of NHE3V
phosphorylation per amount of NHE3V protein was constant under basal,
FBS-, FGF-, and PMA-regulated conditions. Thus, stimulation of NHE3
transport activity by FBS and FGF and its inhibition by PMA are not
associated with detectable changes in total phosphorylation of the
exchanger protein.
While there is no change in the total amount of
phosphorylation of NHE3V in response to FBS, FGF, and PMA, it was
possible that there was an increase in phosphorylation of one site of
NHE3V that was too small to change total NHE3 phosphorylation and/or a
simultaneous decrease of another site, leading to no change in the
total amount of NHE3V phosphorylation. It was also possible that there
was a site that was dominant in the amount of phosphorylation and
masked any changes in the amount of phosphorylation of other sites in
response to growth factors/protein kinases. Therefore, we performed
two-dimensional phosphopeptide mapping on immunoprecipitated NHE3V.
As shown in Fig. 3A, two-dimensional tryptic
phosphopeptide mapping reproducibly revealed 14 phosphopeptides of NHE3
under basal conditions. However, the patterns of maps obtained in the basal state were indistinguishable from those obtained upon treatment with PMA (1 µM) and FGF (10 ng/ml) in terms of the number
of phosphopeptides identified and the relative intensity or migration
positions within each map and between different maps (Fig.
3B). In contrast, 10% FBS exposure increased
phosphorylation in two phosphopeptides, 9 and 12. When normalized to
phosphopeptide 3, which did not appear to change in phosphorylation
with FBS exposure, FBS significantly increased the phosphorylation of
those spots (phosphorylation of phosphopeptide 9 as a percentage of
phosphopeptide 3 phosphorylation as quantified by densitometry using
ImageQuant software was as follows: control, 19 ± 8%; FBS,
51 ± 13%; n = 3, p < 0.05;
phosphorylation of phosphopeptide 12 as a percentage of phosphopeptide
3 phosphorylation was as follows: control, 45 ± 8%; FBS,
135 ± 35%, n = 3, p < 0.05).
To determine the types of
phosphoamino acids in NHE3V, the phosphorylated NHE3V immunoprecipitate
was subjected to phosphoamino acid analysis as described under
"Experimental Procedures." Under basal conditions, the phosphoamino
acids detected were entirely phosphoserine, with no phosphothreonine or
phosphotyrosine identified (Fig. 4). Identical findings
were obtained under FGF (10 ng/ml), PMA (1 µM) -exposed
conditions. FGF stimulates and PMA inhibits NHE3V (Fig. 1 and Refs. 6
and 7). Simultaneous exposure of FGF (10 ng/ml) and PMA (1 µM) to PS120/NHE3V cells also revealed only phosphoserine
but no phosphotyrosine and phosphothreonine in NHE3V immunoprecipitates
(Fig. 4). The absence of tyrosine phosphorylation on NHE3V
immunoprecipitate was also confirmed with a polyclonal
anti-phosphotyrosine antibody by Western blotting on basal, FGF-, PMA-,
and FGF plus PMA-exposed conditions (data not shown).
It is possible that in an
overexpressing system, such as the expression of NHE3 in PS120 cells
(7), not all expressed proteins are targeted to the plasma membranes.
It is also not known whether the intracellular pools of NHE3 are
phosphoproteins, although these pools of NHE3 are believed not to be
involved in the exchange function of NHE3, at least under basal
conditions. Therefore, we estimated the fraction of total NHE3 protein
that is on the cell surface and determined whether the phosphorylation
of cell surface NHE3 changed in response to growth factor/protein
kinase regulation.
The surface and intracellular forms of NHE3V were separated by
biotinylation of membrane surface proteins and subsequent affinity binding of biotinylated membrane surface proteins to avidin-agarose. Biotinylation was performed at 4 °C to restrict the biotin labeling to the cell surface proteins by minimizing internalization of biotin
(17). After biotinylation, the cells were solubilized with IPT buffer,
and the solubilized crude extract was incubated with avidin-agarose to
which biotinylated surface proteins bound (avidin fraction), and the
surface NHE3V was identified by Western blotting with anti-VSVG
antibodies (Fig. 5, lane 2). Repeated incubation of the solubilized crude extract with avidin-agarose did not
increase the recovery of the biotinylated NHE3V (data not shown). The
crude extract after clearing with avidin-agarose represents the
intracellular form of NHE3V and was analyzed by Western blotting (Fig.
5, lane 1) and compared with total NHE3 determined by
Western blot analysis (Fig. 5, lane 3). Although the
efficiency of biotinylation of the NHE3V in the present studies is not
known, there are 11 lysine residues located on the putative extracellular surfaces of NHE3, and these lysine residues are theoretically able to react with biotin. The result shown in Fig. 5
suggested that at least 27% of NHE3V is expressed on the plasma membranes in PS120/NHE3V cells under basal conditions. To eliminate the
possibility of the labeling of intracellular proteins with biotin, both
the avidin-precipitated fraction and intracellular fraction (left after
avidin precipitation) were probed with an anti-actin antibody. The
42-kDa actin protein was identified in both the total and the
intracellular form fractions but not in the avidin fraction (data not
shown), suggesting that there was unlikely to be any labeling of
intracellular proteins with biotin.
We next determined whether both the surface and intracellular forms of
NHE3V are phosphoproteins. PS120/NHE3V cells were labeled in
vivo with [32P]orthophosphate, and the surface
plasma membrane proteins were biotinylated at 4 °C with
NHS-SS-biotin. The cells were then solubilized with 1% Triton X-100.
To obtain the intracellular form, the solubilized crude cell extract
was first incubated with 100 µl of avidin-agarose to remove
biotinylated proteins. The remaining nonbiotinylated proteins were
immunoprecipitated with anti-VSVG antibody to recover the
nonbiotinylated [32P]NHE3V (intracellular form). In
parallel experiments to obtain the surface form, total
[32P]NHE3V was immunoprecipitated from the solubilized
crude cell extract with anti-VSVG antibody. The
[32P]NHE3V was then eluted from the antigen-antibody
complex with 10 mM Tris, 1% SDS (18). The biotinylated
form of [32P]NHE3V was recovered from the eluted
[32P]NHE3V by incubating with avidin-agarose (surface
form). Both forms of NHE3V were analyzed with SDS-PAGE and transferred
onto nitrocellulose membrane, which was exposed to obtain an
autoradiograph. After autoradiography, the same nitrocellulose membrane
was used to analyze the amount of NHE3V protein by Western blotting
using the anti-VSVG polyclonal antibody, and the blot was detected by enhanced chemiluminscence. As shown in Fig. 6, both the
surface and intracellular forms of NHE3 are phosphoproteins, and both forms of NHE3 are phosphorylated to a similar extent.
Since both the surface and intracellular forms of NHE3V are
phosphoproteins, the latter might mask the changes in phosphorylation of the former in response to growth factor/protein kinase regulation. Therefore, we studied the effect of serum, FGF, and PMA on the phosphorylation of the surface form of NHE3V. PS120/NHE3V cells were
labeled in vivo with [32P]orthophosphate. At
the end of incubation with isotope, cells were incubated with serum
(10%), FGF (10 ng/ml), PMA (1 µM) for 10 min. The
surface plasma membrane proteins were then biotinylated. Total
[32P]NHE3V was immunoprecipitated, followed by elution of
the immunoprecipitated protein from the Protein A-Sepharose beads and
reaffinity purification of the surface [32P]NHE3V by
incubating with avidin-agarose. As shown in Fig. 7, there was no change in the amount of total phosphorylation of surface
NHE3V in response to stimulation by serum and FGF and inhibition by PMA
(upper panel) when normalized to the amount of surface NHE3V
as indicated below by Western blotting (lower panel).
In the present studies, we showed that the epithelial isoform
Na+/H+ exchanger, NHE3, exists as a
phosphoprotein under basal conditions. Regulation of the transport rate
of NHE3 by FGF, FBS, and phorbol esters is not, however, accompanied by
a simultaneous change in the degree of total phosphorylation of NHE3,
compared with the basal state. Further, two-dimensional tryptic
phosphopeptide maps of NHE3V obtained in the presence or absence of FGF
and phorbol ester were identical. Thus, we conclude changes in NHE3
phosphorylation are not involved in regulation by FGF and phorbol
esters. In contrast, FBS increased phosphorylation of two
phosphopeptides based on two-dimensional tryptic phosphopeptide maps.
The magnitude of the change in phosphorylation of NHE3 phosphopeptides
9 and 12 in response to FBS (2-3-fold) is similar to that reported in
regulation of other transport proteins (19, 20). These results suggest that, as for NHE1, NHE3 is regulated by mechanisms independent of
phosphorylation of the exchanger (FGF, phorbol ester) as well as
mechanisms that may be dependent on changes in phosphorylation of NHE3
(FBS). However, we are aware that we have not established that the
changes in phosphorylation of NHE3 are involved in its regulation by
FBS. This is especially important to note, since FBS causes the largest
magnitude change in NHE3 stimulation among agents studied, although it
is not known which growth factor is responsible.
There are alternate interpretations for failure to observe changes in
phosphorylation of NHE3 in parallel with FGF and phorbol ester
regulation of NHE3 transport activity. For instance, the C-terminal
VSVG epitope tag could have interfered with FGF/phorbol ester
regulation of NHE3 and therefore of phosphorylation. This was shown not
to be the case based on the FBS/phosphorylation studies as well as on
the fact that PS120/NHE3V responded similarly to non-epitope-tagged
NHE3 with stimulation by FGF and inhibition by PMA (Fig. 1). These
results do exclude the possibility that the lack of change in total
phosphorylation of NHE3 by FGF and phorbol esters is due to the
selective dephosphorylation of the sites phosphorylated under basal
conditions with concomitant phosphorylation of other sites.
Furthermore, as a parallel positive control experiment using identical
experimental procedures, we confirmed that NHE1 is a phosphoprotein and
that its amount of phosphorylation increases with response to serum
stimulation of NHE1 (data not shown), as reported (10, 11).
Similar to the results shown here with NHE3, regulation of NHE1 occurs
by phosphorylation-dependent and -independent mechanisms. NHE1 is not only a phosphorylated protein under basal conditions, but
also growth factors/protein kinases activated NHE1 and increased its
phosphorylation in parallel (14-16). While there are two major signaling pathways of NHE1 expressed in fibroblasts (thrombin/PI turnover; epidermal growth factor/tyrosine phosphorylation), both pathways increased phosphorylation of the same set of NHE1
phosphopeptides, indicating action via a common kinase intermediate
(16). It was recently shown that the location of the NHE1
phosphorylation sites was C-terminal of amino acid 635, in the
cytoplasmic tail of NHE1, which contains 8 serine residues (16).
However, growth factors/kinases were still able to cause 50%
activation of an NHE1 truncation mutant that lacks the entire
phosphorylation domain. This suggests that 50% of the growth
factor/kinase regulation of NHE1 is phosphorylation-independent and may
involve associated regulatory proteins (16).
The concept of regulation by associated regulatory proteins was
further supported by the observation that osmotic regulation of NHE1
occurs without changes in NHE1 phosphorylation (13). In fact, three
NHE1-associated regulatory proteins have been identified. Wakabayashi
showed that calmodulin (Cam) binds to two separate domains with
different affinities on the cytoplasmic C terminus of NHE1. The high
affinity binding site normally inhibits NHE1 activity, and in the
presence of Ca2+, binding of calmodulin to the high
affinity site leads to stimulation of NHE1 activity (22-24). Grinstein
and co-workers (25) identified an NHE1-associated protein, p24, which
co-immunoprecipitates with NHE1. This protein is not constitutively
phosphorylated, nor could phosphorylation be induced by serum or
phorbol ester. Binding of this protein to NHE1 is
Ca2+-independent. However, the identity of this regulatory
associated protein is not yet known. Independently, Lin and Barber (26) used an interaction cloning approach to identify NHE1 regulatory proteins by using a glutathione S-transferase fusion protein
of the NHE1 C-terminal cytoplasmic domain as a hybridization probe to
screen an expression library. They identified a calcineurin B- and
calmodulin-homologous protein, which they called calcineurin-homologous protein (CHP). CHP binds to the "maintenance domain" of NHE1 and is
a phosphoprotein. Interestingly, serum stimulates NHE1 activity and
increases NHE1 phosphorylation. Expression of CHP inhibits serum-stimulated NHE1 activity, while the amount of phosphorylation of
CHP decreases in response to serum. Therefore, it is suggested that the
phosphorylation state of the CHP is important for regulation of NHE1
and that it has an inhibitory effect.
A strong case can be made that some regulation of NHE3 is via
associated regulatory proteins, with the associated proteins probably
regulated directly or indirectly by phosphorylation. The clearest
example is calmodulin (CaM). CaM binds and regulates NHE1 and also
regulates and binds NHE3 (8, 24). Inhibiting CaM stimulates NHE3, with
the CaM acting by both CaM kinase II-dependent and
-independent mechanisms (11). CaM binds to the C terminus of NHE3 and
does so in the same area that is required for its effect on
Na+/H+ exchange (24).2
Further evidence to support a role for associated regulatory proteins
in regulating NHE activity comes from studies of the cell specificity
of NHE regulation. Regulation by protein kinases of NHE3 stably
expressed in PS120 cells and in the human colon cancer cell line,
Caco-2 cells, is remarkably similar and also, with one exception, is
the same as the regulation of NHE3 expressed endogenously on the apical
membrane of the rabbit small intestine and rat colon. The major
difference is that cAMP inhibits intestinal brush border NaCl
absorption and brush border Na+/H+ exchange, at
least in some studies, while cAMP has no effect on NHE3 in PS120 cells
and on the apical surface of Caco-2 cells (2, 6). In contrast, Moe
et al. showed that when rat NHE3 is stably expressed in AP-1
cells, cAMP inhibits NHE3 along with causing an increase in NHE3
phosphorylation as determined by immunoprecipitation of NHE3 after
in vivo phosphorylation (11). These results show that
protein kinase A inhibits NHE3 stably expressed in AP-1 cells along
with increasing NHE3 phosphorylation. To date, as with the FBS-induced increase in NHE3 phosphorylation reported here, there has
been no demonstration that stimulation by cAMP of NHE3 phosphorylation in AP-1 cells is what inhibits NHE3 activity.
The simplest explanation for the differences in cell-specific NHE
regulation is that there are cell-specific differences in associated
regulatory proteins, since, for instance, cAMP in Caco-2 cells, causes
other effects, including chloride secretion. Two such related
associated regulatory proteins called Na+/H+
exchanger regulatory factor and NHE3 kinase A regulatory protein were
recently cloned by Weinman et al. (27) and Yun et
al. (28), respectively. These proteins are suggested as mediating
cAMP inhibition of NHE3 in brush border membranes of kidney and
intestine. Based on Western blotting using antibodies to these
proteins, rabbit intestinal brush border membranes and AP1 cells were
found to have this protein, while PS120 cells lack it. This is
consistent with the observation that NHE3 is regulated by cAMP in
intestinal brush border membranes and AP1 cells but not in PS120 cells.
Recently, it was shown that co-expression of
Na+/H+ exchanger regulatory factor or NHE3
kinase A regulatory protein with NHE3 in PS120 cells reconstitutes cAMP
inhibition of NHE3 (28).
The function of basal NHE3 serine phosphorylation is not known.
Multiple phosphorylation sites of NHE3 exist, since at least 14 phosphopeptides were generated by trypsin digestion when studied by
two-dimensional phosphopeptide mapping (Fig. 3A).
Possibilities for the function of basal or FBS-stimulated
phosphorylation in NHE3 include plasma membrane targeting or removal
from the plasma membrane. Recently, based on preliminary studies, we
suggested that vesicle membrane shuttling is involved in regulation of
NHE3 in rat kidney (29). Also unknown is whether basal NHE3
phosphorylation is required for associated protein regulation of
NHE3.
In conclusion, NHE3 stably transfected in PS120 fibroblasts is a
phosphoprotein under basal and growth factor/kinase-regulated conditions. Moreover, similar to results with NHE1, growth
factor/kinase regulation of NHE3 involves both NHE3
phosphorylation-independent (FGF, phorbol ester) and potentially
phosphorylation-dependent (FBS) mechanisms when NHE3 is
studied in PS120 cells, the same cell in which NHE1 was studied. The
difference between NHE1 and NHE3 regulation is that, for NHE3, certain
regulators act independently of changes in phosphorylation (FGF,
phorbol esters), while these same regulators of NHE1 act by both
phosphorylation-dependent and phosphorylation-independent
components. Our results suggest that NHE3 stably transfected in PS120
fibroblasts is regulated at least by FGF and phorbol esters via
associated regulatory proteins. Unknown is whether NHE3 is regulated by
both phosphorylation-dependent and -independent mechanisms
in epithelial cells of intestine and kidney, where NHE3 is
localized to the brush border membranes (2, 5).
We thank Helen McCann for her assistance in
manuscript preparation and editorial assistance, Drs. T. Kreis and D. Louvard for providing anti-VSVG antibodies, and Dr. J. Pouyssegur for the PS120 cells.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
(Hi+) (9). At a
protein level, NHE1 and NHE3 share 50-60% identity in amino acid
sequences in the N-terminal transmembrane domain (9, 10). The
cytoplasmic tail, which is the main area for regulatory function and
contains multiple putative protein kinase consensus sites, however, is
very different, with only 20-30% amino acid identity (9, 10). Like
NHE1, in the absence of ATP, basal Na+/H+
exchange activity of NHE3 is greatly decreased and growth
factor/protein kinase regulation ceases (6). NHE3 is, either directly
or indirectly, regulated by phosphorylation. Okadaic acid, a
phosphatase inhibitor, increases the Vmax of
NHE3, which implies a basal phosphorylated state of at least some
proteins involved in NHE3 regulation. Conversely, genistein, a tyrosine
kinase blocker, inhibits basal Na+/H+ exchange
activity, which suggests that tyrosine phosphorylation may be involved
in stimulation of basal Na+/H+ exchange (8). In
addition, by performing C-terminal truncations of NHE3 and expressing
the truncated cDNAs in PS120 cells, several separate subdomains of
the cytoplasmic C terminus of NHE3 were identified, including areas
responsible for regulation by calmodulin, phorbol ester, okadaic acid,
FGF, and serum (8). NHE3 is phosphorylated under basal conditions, and
this has been shown for NHE3 stably transfected in PS120 cells and in
another fibroblast cell line, AP-1 cells (11). Both the PS120 cells and
AP-1 cells are deficient in endogenous NHEs (12, 13). The amount of
phosphorylation of NHE3 increases with cAMP inhibition of NHE3 in AP-1
cells (11). In the present study, we determine whether stimulation of
NHE3 by FGF and inhibition of this protein by phorbol ester are
associated with or are due to changes in phosphorylation of NHE3 stably
expressed in PS120 cells. These regulators were compared with serum
with its mixed growth factors. This study was also designed to separate the cell surface and intracellular forms of NHE3 and to determine whether changes in the rate of Na+/H+ exchange
by NHE3 as caused by protein kinases/growth factors are associated with
changes in cell surface NHE3 phosphorylation.
Materials
-end
of the VSVG tag sequence in pBluescript (kindly provided by Dr. D. Louvard). The XhoI site was retained at the boundary of the
VSVG tag sequence and the E3/585 cDNA. The tag sequence, when
translated, contained an 8-amino acid spacer arm sequence (RGEGPPGP)
followed by the 11-amino acid VSVG epitope (YTDIEMNRLGK). To obtain the
E3V/pBluescript construct, a 0.9-kilobase pair NHE3 C terminus cDNA
fragment (nucleotides 1536-2496, corresponding to amino acids
522-832) was obtained by polymerase chain reaction using the paired
primers (5
primer, GTCGGCCAGAAGTCTCGG; 3
primer, TTTCCTCGAGATGTGTGTGGACTCGGGG) where the 3
primer contained the NHE3
stop codon mutated into an XhoI site. This polymerase chain reaction fragment was then digested with StuI and
XhoI (corresponding to amino acids 581-832) to replace the
StuI-XhoI fragment in E3/585V/pBluescript, thus
generating NHE3V/pBluescript. The NHE3V cDNA was then subcloned into pMAMneo, generating NHE3V/pMAMneo.
We have previously
shown that FBS and FGF stimulate, whereas PMA inhibits, NHE3 (6-8).
Mechanistically, all growth factors regulate NHE3 by a Vmax change (6). To confirm that the VSVG
epitope on NHE3V does not interfere with growth factor regulation of
NHE3V, we studied the effect of FBS, FGF, and PMA on NHE3V stably
expressed in PS120 cells. As shown in Fig.
1A, when PS120/NHE3V cells, which were
exposed to sodium medium (containing 130 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgSO4, 1 mM NaPO4,
25 mM glucose, 20 mM HEPES, pH 7.4) and had
come to a steady-state pHi, were then exposed to 10% FBS, an
increase in pHi of 0.1 unit occurred. This intracellular
alkalinization induced by serum could be blocked by 1 mM
amiloride, confirming the activation of Na+/H+
exchange (data not shown). Similarly, like wild-type NHE3, initial rate
analysis revealed that FGF (10 ng/ml) stimulated NHE3V by a
Vmax change (from 1671 ± 211 µM/s to 2205 ± 171 µM/s, a 32% increase) (Fig. 1B), and PMA inhibited the
Vmax of Na+/H+ exchange
by NHE3V (decreased from 1796 ± 150 µM/s to
1086 ± 88 µM/s, a 40% decrease). In both cases,
there is no change in
K
(Hi+) or the
Hill coefficient. Similarly, 10% FBS stimulated the initial rate of
Na+/H+ exchange by NHE3V by 45% (data not
shown). These results show that NHE3V is regulated by protein
kinases/growth factors similarly to wild-type NHE3.
Fig. 1.
) stimulated NHE3V and increased the
pHi by 0.1 unit. B, initial rate analysis of
NHE3V/PS120 cells in response to FGF (10 ng/ml) (
, control cells;
, FGF-treated cells). FGF stimulated the
Na+/H+ exchange of NHE3V by a
Vmax change from 1671 ± 211 (control) to
2205 ± 171 µM H+/s (treated) with no
change in the Hill coefficient (1.9 ± 0.2 and 1.8 ± 0.1 for
control and treated cells, respectively) and in
K
(Hi+) (0.18 ± 0.08 µM and 0.18 ± 0.05 µM for
control and treated cells, respectively). C, inhibition of NHE3V by PMA. PMA (1 µM) inhibited the Vmax of the
Na+/H+ exchange activity of NHE3V by 40% (from
1796 ± 150 (control,
) to 1086 ± 88 µM/sec
(PMA-treated cells,
)). There was no change in the Hill coefficient
(1.8 ± 0.1 and 1.8 ± 0.2 for control and treated cells,
respectively) and in
K
(Hi+) (0.22 ± 0.07 µM and 0.18 ± 0.06 µM for
control and treated cells, respectively).
[View Larger Version of this Image (11K GIF file)]
Fig. 2.
Phosphorylation of NHE3V. A, the
autoradiogram showed that NHE3 is a phosphoprotein under basal
conditions and in PMA (1 µM)-inhibited and FGF (10 ng/ml)
and serum (10%)-stimulated conditions. PS120/NHE3V cells were labeled
in vivo with [32P]orthophosphate, transport
was altered under growth factor-regulated conditions, and then NHE3V
was immunoprecipitated with the rabbit polyclonal anti-VSVG antibodies.
The immunoprecipitate was then analyzed by SDS-PAGE and
autoradiography, which showed that NHE3 is an 85-kDa phosphoprotein.
PMA, FGF, and serum, which regulate NHE3V, did not change the amount of
the phosphorylation on NHE3V compared with the control. PS120 cells
were used as an negative control. B, to ensure that equal
amounts of immunoprecipitate obtained from each condition were used for
SDS-PAGE and autoradiography, PS120/NHE3V cells were labeled in
vivo under basal and FGF-, PMA-, and serum-stimulated conditions
as described in A, and the immunoprecipitated NHE3V was
separated by SDS-PAGE and transferred onto nitrocellulose membrane,
which was then exposed to obtain an autoradiograph (upper panel). After autoradiography, the same nitrocellulose membrane was used to analyze the amount of NHE3V protein by Western blotting using the mouse monoclonal anti-VSVG antibody, and the blot was detected by ECL (lower panel). This confirms that serum,
FGF, and PMA do not change the amount of NHE3V phosphorylation and that
similar amounts of immunoprecipitate were obtained under all
conditions. This is a representative of five similar experiments. Molecular size markers are indicated on the left.
[View Larger Version of this Image (34K GIF file)]
Fig. 3.
Two-dimensional phosphopeptide mapping of
NHE3V. PS120/NHE3V cells were labeled in vivo and then
treated under basal, FGF-, PMA-, and serum-exposed conditions as
described in the legend to Fig. 2. NHE3V was then immunoprecipitated
with the rabbit polyclonal anti-VSVG antibody. The immunoprecipitated
protein was separated by SDS-PAGE. The polyacrylamide gel was dried and
exposed for autoradiography. The phosphoprotein band (NHE3V
immunoprecipitate) was sliced out and digested with trypsin (0.3 mg/ml,
TPCK-treated). The tryptic phosphopeptides were analyzed in the first
dimension by thin layer electrophoresis and in the second dimension by
thin layer ascending chromatography. The two-dimensional phosphopeptide map was analyzed by autoradiography. A, two-dimensional
tryptic phosphopeptide mapping reproducibly revealed 14 phosphopeptides in the control NHE3V phosphopeptide map. B, identical
two-dimensional phosphopeptide maps of NHE3V were obtained under the
control basal condition and FGF- and PMA-regulated conditions. Under
FBS conditions, a relative increase in size of phosphopeptides 9 and 12 was reproducibly seen when compared with the control. Shown is a
representative of five similar studies.
[View Larger Version of this Image (33K GIF file)]
Fig. 4.
Phosphoamino acid analysis of NHE3V.
Phosphoamino acid analysis was performed on 32P-labeled
NHE3V immunoprecipitate obtained under basal and FGF (10 ng/ml)-, PMA
(1 µM)-, and FGF (10 ng/ml) plus PMA (1 µM)-regulated conditions as described under
"Experimental Procedures." The results showed that the phosphoamino
acids detected were entirely phosphoserine with no phosphothreonine or
phosphotyrosine under basal, FGF-, PMA-, and FGF plus PMA-treated
conditions. This is a representative of two similar experiments.
[View Larger Version of this Image (29K GIF file)]
Fig. 5.
Separation of cell surface and intracellular
forms of NHE3. PS120/NHE3V cells were biotinylated with
NHS-SS-biotin at 4 °C. After biotinylation, the cells were
solubilized with IPT buffer, and a of the total mixture (by
volume) of the crude solubilized extract was analyzed by Western
blotting using anti-VSVG antibody (lane 3, total NHE3V). The
biotinylated surface NHE3V in the crude solubilized extract was then
adsorbed onto avidin-agarose, and a
of the total mixture
(by volume) of this fraction (lane 2, surface) was analyzed
by Western blotting, to keep the amount of samples loaded into each
lane the same for quantitation. The crude extract after clearing with
avidin-agarose represents the intracellular form and a
of
the total mixture (by volume) of this fraction was analyzed by Western
blotting (lane 1) for quantitative comparison with total
NHE3V (lane 3) and the surface form of NHE3V (lane
2). Quantitation by densitometry (ImageQuant software) revealed
that ~27% of the 85-kDa NHE3V protein is expressed on the cell
surface (compare lanes 2 and 3). Shown is a
representative result of three experiments with similar results, with
control surface NHE3V being 27 ± 8% of total NHE3V.
[View Larger Version of this Image (36K GIF file)]
Fig. 6.
Both the surface form and the intracellular
form of NHE3V in PS120/NHE3V cells are phosphoproteins under basal
conditions. PS120/NHE3V cells were labeled in vivo with
[32P]orthophosphate, and the surface plasma membrane
proteins were biotinylated with NHS-SS-biotin as described under
"Experimental Procedures." The cells were then solubilized with IPT
buffer. To obtain the intracellular form of NHE3V, the solubilized
crude cell extract was first incubated with 100 µl of avidin-agarose to remove biotinylated proteins. Remaining nonbiotinylated proteins were then immunoprecipitated with anti-VSVG antibody to recover the
nonbiotinylated [32P]NHE3V. In parallel experiments to
obtain the surface form of NHE3V, total [32P]NHE3V was
immunoprecipitated from the solubilized crude cell extract with
anti-VSVG antibody. The [32P]NHE3V was then eluted from
the antigen-antibody complex with 100 µl of 10 mM Tris,
1% SDS. The eluted [32P]NHE3V was then diluted with 1 ml
of IPT buffer. The biotinylated surface form of
[32P]NHE3V was then recovered by incubating with
avidin-agarose. Both forms of NHE3V were analyzed by SDS-PAGE and
transferred onto nitrocellulose membranes, which were exposed to obtain
an autoradiograph (upper panel). After autoradiography, the
same nitrocellulose membrane was used to analyze the amount of NHE3V protein by Western blotting (lower panel) using the
anti-VSVG polyclonal antibody. The blot was detected by enhanced
chemiluminscence. This result suggests that both the surface form and
intracellular form of NHE3 are phosphoproteins and that both forms of
NHE3 are phosphorylated to a similar extent.
[View Larger Version of this Image (36K GIF file)]
Fig. 7.
Phosphorylation of the surface form of
NHE3V. PS120/NHE3V cells were labeled in vivo with
[32P]orthophosphate. At the end of incubation, cells were
exposed to serum, FGF, and PMA as described in the legend to Fig. 2 and under "Experimental Procedures." The surface form of NHE3 was then
obtained under the control and growth factor-regulated conditions as
described in Fig. 6. The upper panel is an autoradiogram
showing that serum, FGF, and PMA, which regulate NHE3, do not change
the amount of the phosphorylation of surface NHE3V compared with the control. The lower panel is Western blotting confirming that
similar amounts of the surface form of NHE3V were obtained under all
conditions.
[View Larger Version of this Image (27K GIF file)]
*
This work was supported by NIDDK, National Institutes of
Health, Grants R01DK26523, P01DK44484, and R29DK43778 and the Meyerhoff Digestive Diseases Center.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 a postdoctoral fellowship from the Medical Research
Council, United Kingdom. Present address: Dept. of Medicine, California
Pacific Medical Center, San Francisco, CA.
To whom correspondence should be addressed: GI Division, Dept.
of Medicine, The Johns Hopkins University School of Medicine, 925 Ross
Research Bldg., 720 Rutland Ave., Baltimore, MD 21205-2195. Tel.:
410-955-9685; Fax: 410-955-9677.
1
The abbreviations used are: NHE,
Na+/H+ exchanger; PAGE, polyacrylamide gel
electrophoresis; FGF, fibroblast growth factor; FBS, fetal bovine
serum; PMA, phorbol 12-myristate 13-acetate; VSVG, vesicular stomatitis
virus glycoprotein; pHi, intracellular pH; TPCK,
L-1-tosylamide-2-phenylethylchloromethyl ketone; CHP, calcineurin-homologous protein.
2
C.-M. Tse, R., Kambadur, M. Zizak, S. Nath, and
M. Donowitz, unpublished observation.
©1997 by The American Society for Biochemistry and Molecular Biology, Inc.