(Received for publication, April 10, 1996, and in revised form, October 11, 1996)
From the Division of Cell Biology, Hospital for Sick
Children, Toronto, M5G 1X8, Canada and the
Department of
Physiology, McGill University, Montreal, H3G 1Y6 Canada
Cholinergic agonists stimulate isotonic fluid
secretion in the parotid gland. This process is driven by the apical
exit of Cl, which enters the cells partly via
Cl
/HCO3
exchange across the
basolateral membrane. Acidification of the cytosol by the extrusion of
HCO3
is prevented by the concomitant activation
of the Na+/H+ exchanger (NHE), which is
directly activated by cholinergic stimulation. Multiple isoforms of the
NHE have been described in mammalian cells, but the particular
isoform(s) present in salivary glands and their mechanism of activation
have not been defined. Reverse transcriptase-polymerase chain reaction
with isoform-specific primers was used to establish that NHE-1 and
NHE-2, but not NHE-3 or NHE-4, are expressed in parotid glands. The
presence of NHE-1 was confirmed by immunoblotting and
immunofluorescence, which additionally demonstrated that this isoform
is abundant in the basolateral membrane of acinar cells. The
predominant role of NHE-1 in carbachol-induced
Na+/H+ exchange was established
pharmacologically using HOE694, an inhibitor with differential potency
toward the individual isoforms. Because muscarinic agonists induce
stimulation of protein kinases in acinar cells, we assessed the role of
phosphorylation in the activation of the antiport. Immunoprecipitation
experiments revealed that, although NHE-1 was phosphorylated in the
resting state, no further phosphorylation occurred upon treatment with
carbachol. Similar phosphopeptide patterns were observed in control and
carbachol-treated samples. Together, these findings indicate that
NHE-1, the predominant isoform of the antiporter in the basolateral
membrane of acinar cells, is activated during muscarinic stimulation by
a phosphorylation-independent event. Other processes, such as
association of Ca2+-calmodulin complexes to the cytosolic
domain of the antiporter, may be responsible for the activation of
Na+/H+ exchange.
Stimulation of salivary acinar cells induces rapid and abundant
secretion of isotonic fluid, a process that is driven primarily by
efflux of Cl and HCO3
across
the apical membrane (1). The stimuli, such as cholinergic agonists,
increase secretion by elevating the anion permeability of the apical
membrane, while promoting accumulation of Cl
in the
cytosol. The latter is accomplished in part by activation of a loop
diuretic-sensitive Na+-K+-2Cl
co-transporter, but also by parallel operation of the
Cl
/HCO3
and
Na+/H+ antiporters in the basolateral membrane
(1, 2). Accordingly, fluid secretion in perfused salivary glands was
found to be sensitive to inhibitors of the
Na+/H+ antiporter (3). Moreover, in isolated
rat parotid acinar cells carbachol markedly activated Na+
influx, and the initial rate of influx was inhibited by 75% in the
presence of dimethylamiloride (DMA),1 a
relatively specific inhibitor of Na+/H+
exchange (4, 5).
Stimulation of salivary cells is accompanied by a tendency of the
cytosol to become acidic. This is attributable in part to generation of
acid equivalents by the metabolic pathways supplying energy to the
secretory process, but mainly to the electrodiffusional exit of
HCO3 across the apical membrane (6, 7).
Because the changes are caused by HCO3
itself, the cells are unable to buffer the cytosolic pH (pHi) using HCO3
/CO2 and must
resort to other regulatory mechanisms. Na+/H+
exchange fulfills this role as well, extruding the excess cytosolic acid across the basolateral membrane (5, 8, 9). The alkalinization that
accompanies activation of the antiporter not only regulates pHi, but also promotes the intracellular accumulation of
HCO3
, facilitating secretion of this anion
(10). Jointly, these observations indicate that activation of the
basolateral Na+/H+ antiporter plays an
essential role in salivary fluid secretion. Despite its importance,
however, neither the identity nor the molecular mechanism of activation
of the antiporter have been elucidated.
Five distinct isoforms of the Na+/H+ exchanger (NHE) which differ in their kinetic and pharmacological properties have been identified in mammalian cells (11, 12). They are differentially expressed in various tissues, suggesting distinct functions for the individual isoforms. NHE-1 is ubiquitously expressed and is involved in the regulation of pHi and cell volume in both epithelial and non-polarized cells (12). NHE-2 and NHE-3 are prominent in intestinal and renal tissues where they ostensibly participate in transepithelial NaCl transport. NHE-3 is located on the apical membrane, while the specific location of NHE-2 is still controversial (11, 12, 13, 14). The two other isoforms are poorly characterized. NHE-4 is abundant in the stomach (14), but its precise cellular location and function remain obscure. Similarly, the distribution and function of NHE-5 (15) are still unknown, and even its full sequence remains to be defined. Little is known about the distribution of these isoforms in the salivary gland.
The purpose of the experiments described in this article was to identify the isoform(s) of the Na+/H+ exchanger present in acinar cells of the rat parotid gland, to explore their individual contribution to the uptake of Na+, and the regulation of pHi, and to define their mechanisms of activation during cholinergic stimulation.
DMA was a generous gift from Dr. T. Kleyman
(Department of Medicine, University of Pennsylvania).
3-(Methylsulfonyl-4-piperidino-benzoyl)guanidine methanesulfonate
(HOE694) was kindly provided by Dr. A. Durckheimer, Hoechst AG,
Frankfurt, Germany. The acetoxymethyl esters of sodium-binding benzofuran isophthalate (SBFI) and of
2,7
-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) were
obtained from Molecular Probes Inc. (Eugene, OR). Nigericin was from
Calbiochem-Novabiochem Corp. (La Jolla, CA). All other chemicals used
were from Sigma or other standard commercial sources.
Unless otherwise indicated, experiments were
conducted in nominally HCO3-free medium
(solution A) consisting of (in mM) 135 Na+,
144.6 Cl
, 5.4 K+, 0.73 PO42
, 0.8 SO42
,
0.8 Mg2+, 1.8 Ca2+, 20 HEPES, 2 glutamine, and
10 glucose, pH 7.4, at 37 °C. Where specified, 25 mM
HCO3
replaced Cl
and the
solution was gassed with 95% O2, 5% CO2
(solution B). Acid loading was accomplished by pre-pulsing the cells
for the indicated time in a medium where 40 mM
NH4+ replaced Na+ (solution C).
In the Na+-free medium (solution D) all Na+ was
iso-osmotically replaced with
N-methyl-D-glucammonium+.
Parotid acinar cells from male Wistar rats were isolated by sequential treatment of the glands with trypsin (Life Technologies, Inc.) and purified collagenase (Worthington, type CLSPA), as described previously (16). The fraction used for optical studies, which consisted of single cells, doublets, triplets, and "strings" of cells, was kept at room temperature with periodic top gassing with 100% O2.
Microscopy and Fluorescence MeasurementsApproximately 200 µl of the cell suspension was layered onto a
poly-L-lysine (0.8 mg/ml) coated coverslip. Cells adhering within 2 min were covered with solution A and loaded with the dye by
incubation with either 1 µM BCECF-acetoxymethyl ester for 5 min at 37 °C or with 7 µM SBFI-acetoxymethyl
ester for 60 min at room temperature under 100% O2
gassing. After loading, cells were allowed to recover for 30-60 min in
solution A at 37 °C under 100% O2 gassing, to minimize
possible toxic effects of dye loading (4). The coverslip was next
mounted in a chamber and perfused continuously with solution A on the
stage of an inverted microscope (Zeiss Axiovert). BCECF was excited
sequentially at 440 and 490 nm (10 nm band pass) and emission was
detected at 530 nm (10-nm band pass). Fluorescence was quantified by
averaging pixel intensities throughout the cell and pHi was
determined by in situ calibration of the excitation ratio
using the K+/nigericin technique. SBFI was excited at 340 and 380 nm (10-nm band pass) and emission was measured at 500 nm (40-nm
band pass). Additional details of the optical setup were described
previously (4, 16, 17, 18). To convert the ratio of SBFI fluorescence to
[Na+]i, cells were exposed to various
extracellular concentrations of Na+ (substitution for
K+) in the presence of 10 µM gramicidin.
Ionophore-induced cell swelling was prevented by replacing 60 mM Cl with gluconate
. The data
fitted the equation [Na+]i = KDB (R
R0)/(Rmax
R), where R0 and
Rmax were the ratios measured in the absence and
presence of saturating (150 mM)
[Na+]i, respectively. KD was
20.8 ± 1.4 mM (n = 6). During all
experiments, cells were viewed simultaneously by differential
interference contrast while measuring low light-level fluorescence,
using red illumination and the dichronic mirrors and filter sets
described earlier (18). This enabled us to simultaneously estimate cell
volume, as described (16, 17).
Net proton flux,
JH+ (in mM/min), was
calculated as the product of the rate of pHi recovery
(dpH/dt) and the intrinsic buffering capacity of the cells.
The latter was measured using weak electrolyte pulses, as described
(19), while dpH/dt over a discrete pHi
interval was determined by fitting a straight line to 3 or more
consecutive data points. Lines were fitted by least squares using
Cricket Graph 1.3.2 and consistently yielded r2 > 0.95. The slope of this line was considered to be the rate of
pHi change at the mean pHi of the interval analyzed. An
alternative method of calculating the slope involved fitting all the
data of a pHi recovery curve to an exponential function
(pHi = k0 + K1*ek2t)
using the IGOR curve-fitting software. Similar results were obtained
with both approaches. Where indicated (e.g. Fig.
2B) the rate of acid accumulation induced by the stimulus
was added to the rate of extrusion, to calculate total
Na+-dependent H+ efflux. Acid
accumulation was calculated by measuring the pH changes upon removal of
Na+ in stimulated cells. No spontaneous acid loading was
detected in control cells when Na+ was removed
(n = 25).
Isolation of RNA, Reverse Transcription, and Polymerase Chain Reaction (RT-PCR)
Total RNA was isolated from partially purified
acinar cells by guanidinium thiocyanate-phenol-chloroform extraction
(Trizol; Life Technologies, Inc.), based on the method of Chomczynski
and Sacchi (20). Poly(A+) RNA was purified by affinity
chromatography with an oligo(dT)-cellulose column (Pharmacia). Parotid
mRNA was then reverse-transcribed and the complementary DNA
amplified by the polymerase chain reaction, using the GeneAmp RNA PCR
kit (Perkin-Elmer) and a Perkin-Elmer DNA thermal cycler Model 480. After completion of the PCR reaction (35 cycles), a 10-µl sample of
the PCR tube was analyzed by electrophoresis on a 0.8% agarose gel
pre-stained with 0.5 µg/ml ethidium bromide and the gel was
photographed under UV illumination. Four isoform-specific sets of
primers were used, which hybridized to unique regions of the rat NHE-1,
NHE-2, NHE-3, and NHE-4. Primers were as follows: NHE-1, 5 primer: CCT
ACG TGG AGG CCA AC, 3
primer: CAG CCA ACA GGT CTA CC, size of the PCR
product: 429 base pairs (bp); NHE-2, 5
primer: GCT GTC TCT GCA GGT GG,
3
primer: CGT TGA GCA GAG ACT CG, size of PCR product: 680 bp; NHE-3,
5
primer: CTT CTT CTA CCT GCT GC, 3
primer: CAA GGA CAG CAT CTC GG,
size of PCR product: 574 bp; NHE-4, 5
primer: CTG AGC TCT GTG GCT TC,
3
primer: C GAG GAA ATG CAG CAG C, size of PCR product: 381 bp. All
four sets of primers yielded the expected PCR products when pCMV
plasmids containing the full-length clone of the corresponding isoform
were used as a template, but did not yield discernible products when
any of the other isoforms was used as template.
The preparation and purification of anti-NHE-1 antibodies and the method used for immunoblotting of membranes have been described in detail elsewhere (21). For immunoprecipitation, acinar cells were labeled for 2 h at 37 °C in nominally phosphate-free medium containing [32P]orthophosphate (500 µCi/ml). Cells were then treated with or without carbachol in medium A for 2 min at 37 °C. The reaction was stopped by sedimentation, followed by resuspension in immunoprecipitation buffer. The samples were extracted for 30 min at 4 °C and sedimented for 30 min at 100,000 × g at 0 °C. Immunoprecipitation proceeded as described previously (21).
Peptide MappingSamples were immunoprecipitated as above and eluted from the beads by boiling for 5 min in 125 mM Tris-HCl, pH 6.8, 0.5% SDS, 10% glycerol, and 0.0001% bromphenol blue. After cooling to room temperature, 25 µg/ml chymotrypsin was added to the eluate. Digestion was stopped at the indicated times by addition of mercaptoethanol and SDS to final concentrations of 10 and 2%, respectively, and boiling for 5 min. The peptides were resolved by SDS-PAGE using 15% acrylamide, the gels were dried and autoradiograms obtained using Kodak X-Omat AR film.
ImmunofluorescenceCells attached to coverslips or frozen tissue sections were rinsed twice with PBS and then fixed by incubation with 3% paraformaldehyde for 10 min. Fixation was terminated by rinsing and incubating with 100 mM glycine in PBS, pH 7.4, for 15 min. Cells were then permeabilized by incubation with a solution of 0.1% Triton-X and 0.1% (w/v) bovine serum albumin in PBS (TA-PBS solution) for 15 min, followed by three washes with the same solution. Blocking was then performed by incubation for 20 min in TA-PBS containing 5% goat serum. The coverslips were then washed three more times with TA-PBS. All the preceding steps were at room temperature. The cells were next incubated overnight with a 1:100 dilution of anti-NHE-1 antibodies in TA-PBS at 4 °C. Where indicated, the primary antibody was omitted to control for specificity of staining. After three more washes in TA-PBS, the samples were incubated with a 1:200 dilution of fluorescently labeled donkey anti-rabbit antibody in TA-PBS for 50 min at room temperature. The cells were finally washed three times with PBS and mounted in 50% glycerol containing 1% n-propyl gallate.
Other MethodsProtein was determined using the Pierce BCA Assay Reagent. All experiments were performed at least three times. Representative radiograms or confocal images are illustrated. Quantitative data are presented as mean ± S.E. of the number of experiments (n) in parentheses.
Effect of Carbachol on pHi and on Na+/H+ Exchange
Fig. 1, A and B, illustrate
measurements of pHi in isolated acinar cells using quantitative
imaging of the fluorescence of intracellular BCECF. In agreement with
earlier observations (4, 5), we found that muscarinic stimulation of
parotid acinar cells in the presence of
HCO3 induces a rapid and transient
cytosolic acidification (Fig. 1A). Exposure of cells to 10 µM carbachol in
HCO3
-containing medium (solution B),
resulted in an drop in pHi averaging 0.20 ± 0.02 pH units
(n = 40). The transient acidification was superseded by
a secondary alkalinization, which exceeded the base-line pHi by
an average of 0.21 ± 0.01 (n = 38). As shown in
Fig. 1B, the initial acidification was absent when the cells
were stimulated in nominally HCO3
-free
medium (solution A). Instead, the cells immediately became alkaline,
reaching a final pHi of 7.48 ± 0.08 (n = 66). The occurrence of a transient acidification in the presence, but not in the absence of HCO3
, is consistent
with the notion that this pH change reflects loss of intracellular
HCO3
through Ca2+-activated
apical anion channels (22, 23).
The alkalinization that follows the transient acidosis in solution A
(Fig. 1A), as well as that elicited immediately by carbachol in solution B (Fig. 1B) are attributable to activation of
Na+/H+ exchange. This view is supported by the
following observations. First, omission of Na+ following
stimulation with carbachol induced a pronounced cytosolic acidification
(Fig. 1B), suggestive of accumulation of metabolic acid
and/or reversal of the antiport. This acidification was rapidly reversed upon reintroduction of Na+ (Fig. 1B).
Second, the alkalinization induced by the muscarinic agonist was not
observed in the presence of DMA, a relatively specific inhibitor of
Na+/H+ exchange (data not shown). Third, the
net extrusion of H+ (equivalents) after exposure to
carbachol was accompanied by Na+ influx, readily detectable
as an increase in [Na+] in cells treated with ouabain to
preclude extrusion by the Na+/K+ pump. As shown
in Fig. 1C, where intracellular [Na+] was
measured using SBFI, addition of the glycoside alone increased [Na+] at a marginal rate prior to addition of carbachol,
and DMA had little effect. Upon muscarinic stimulation, however,
intracellular [Na+] increased drastically, at a rate of
approximately 40 mM/min. Importantly, the influx was
greatly reduced in the presence of DMA, implying that at least 75% of
the Na+ enters the cell via the antiporter in
HCO3-containing medium (
55% in
nominally HCO3
-free solution A). Assuming a
1:1 stoichiometry, the amount of Na+ that enters the cell
through the DMA-sensitive pathway (22 mM/min) suffices to
account for the net H+ extrusion induced by carbachol (24 mM/min), calculated from the rate of change of pHi
and considering the buffering power of the cytosol (approximately 8 mM/pH in the pH 7.3-7.5 range). Together, these results
indicate that Na+/H+ exchange is stimulated
markedly by treatment of acinar cells with carbachol. Of note, the
failure of ouabain-treated cells to gain Na+ prior to
muscarinic activation (in the presence and absence of DMA) implies that
the antiporter is virtually quiescent in resting (unstimulated)
cells.
Effect of Carbachol on Na+/H+ Exchange
pHi DependenceThe preceding results suggest that treatment with carbachol converts the antiporter from a quiescent to an active mode. Further insight into this transition was gained by analyzing the properties of Na+/H+ exchange in resting and stimulated cells. Because the antiporter is not detectable in untreated cells at normal pH, its activity was unmasked by acid-loading the cytosol, using an NH4+ pre-pulse. A representative experiment is shown in Fig. 2A. Acinar cells were pulsed with the weak base, which was then removed while simultaneously replacing Na+ with N-methyl-D-glucammonium+ (solution D). Under these conditions, the cells underwent rapid acidification and failed to recover within the period studied, due to the absence of Na+. Upon readdition of Na+, however, a rapid alkalinization ensued. In otherwise untreated cells, pHi recovered to near the original basal level. By contrast, if the cells were treated with carbachol prior to Na+ readdition (open circles in Fig. 2A), the recovery surpassed the resting pHi level, resulting in a net cytosolic alkalinization, reminiscent of that recorded in Fig. 1. Calculation of the rates of Na+-induced H+ (equivalent) extrusion in cells with or without muscarinic stimulation are summarized in Fig. 2B. Two features are noteworthy: first, that following acid loading the rates of recovery are very large (upwards of 100 mM/min), comparing very favorably with other cell types where absolute antiport rates have been reported (e.g. Ref. 24). This likely reflects the specialized function of these secretory cells. Second, it is apparent that the pHi sensitivity of the antiporter is increased following exposure to carbachol. Although the rates of both control and carbachol-treated cells were similar at more acidic pH, exchange is clearly noticeable in the stimulated cells at H+ concentrations where the basal antiporter is essentially inactive (i.e. between 7.3 and 7.45). Similar shifts in the activation threshold or "set point" of the antiport have been reported in other systems (see Ref. 25 for review). That the antiport is truly quiescent in unstimulated cells is suggested by several observations: (i) the absence of a DMA-sensitive Na+ gain in cells treated with ouabain (Fig. 1C); (ii) the failure of DMA and other amiloride analogs to alter baseline pHi (not shown), and (iii) the absence of a cytosolic acidification upon removal of external Na+ in unstimulated cells, a finding that contrasts sharply with the large pHi drop noted when Na+ is removed after stimulation (Fig. 1B).
Isoforms of NHE in Acinar CellsTo better understand the
mechanism underlying NHE activation by muscarinic agonists, it was
important to establish which isoform(s) of the antiport operate in
acinar cells. To this end, we extracted mRNA from parotid glands
and assessed the expression of the four well known isoforms of the
antiporter (NHE-1 to 4) by RT-PCR (Fig. 3).
Isoform-specific primers which hybridized to unique regions of the rat
NHE-1, NHE-2, NHE-3, and NHE-4 were used. All four sets of primers
yielded the expected PCR products when linearized pCMV plasmids
containing the full-length cDNA clone of the corresponding isoform
were used as template (Fig. 3, lanes 1, 4, 7, and
10). No discernible products were detected when a specific
primer set was used with any of the non-corresponding isoforms as
template (not shown). When cDNA obtained by reverse transcription
of rat parotid mRNA was used as a template, the NHE-1 primers
yielded a product of 500 bp (Fig. 3, lane 2), while a
smaller yield of the expected product (
700 bp) was also observed for
NHE-2. The NHE-3 and NHE-4 primers did not yield discernible products
in repeated trials (e.g. Fig. 3, lanes 8 and
11). Omission of reverse transcriptase prevented appearance
of the 500- and 700-bp products, ruling out contamination with genomic
DNA. Thus, the predominant isoforms expressed in parotid glands are
NHE-1 and NHE-2, with no detectable NHE-3 and NHE-4.
The presence of NHE-1 was further documented immunochemically. Acinar
cell membranes were probed with an antibody raised against the
C-terminal 157 amino acids of NHE-1. The specificity of the antibody
was first ascertained comparing Chinese hamster ovary cells transfected
with NHE-1 with their untransfected, antiport-deficient counterparts
(Fig. 4). The antibody recognized a major band of 110-115 kDa, the expected size of mature NHE-1, in the transfectants but not in the deficient precursor cells. A smaller and sharper band
also present in the transfectants but missing in the controls is in all
likelihood the incompletely (core) glycosylated form of NHE-1, a
biosynthetic precursor. A third polypeptide, present in both samples,
is likely nonspecific. As shown in the leftmost lanes of Fig. 4, one
major and one minor polypeptide were also recognized by the antibody in
acinar cell membranes. Both polypeptides remained associated with the
membranes following alkaline extraction of extrinsic components,
suggesting that they are transmembrane proteins. The predominant
immunoreactive band of acinar cells likely represents the mature form
of NHE-1, which is known to be heterogenously glycosylated (12),
accounting for its diffuse mobility on SDS-PAGE. The smaller, sharper
band may be the core-glycosylated biosynthetic precursor.
Parotid glands are composed of acini and ducts. Because these were not
separated for preparation of RNA or for membrane isolation, it cannot
be definitively stated that NHE-1 is present in the acinar cells. To
verify this point, the distribution of NHE-1 in parotid slices was
assessed immunocytochemically, using the polyclonal antibody described
above. Representative confocal fluorescence images are shown in Fig.
5. The low power image of panel A
demonstrates that NHE-1 is present in both ductal and acinar cells. In
both instances, the staining is observed predominantly on the
basolateral membrane, which can be more clearly discerned in
panels B and C. Plasmalemmal immunoreactivity was
also evident in isolated acinar cells (Fig. 5D). Such
staining likely reflects the presence of NHE-1 on the basolateral
membrane, inasmuch as this occupies by far the largest fraction of
surface area in acinar cells (1). Whether NHE-1 is present also in the
apical membranes cannot be defined unambiguously, but in several
instances staining appeared to be minimal in the region of the membrane
facing the lumen of the acinus (e.g. arrow in Fig.
5B), suggesting that the apical membranes are largely devoid
of NHE-1. Discontinuities in the staining of the luminal membrane are
also apparent in ductal cells, suggesting preferential distribution of
NHE-1 on the basolateral side.
The preceding results indicate that NHE-1 is present in acinar cells,
but do not clarify the contribution of this isoform to the
Na+/H+ exchange activity across the basolateral
membrane. The fraction of the exchange mediated by NHE-1 was assessed
pharmacologically in isolated acinar cells using HOE694. This compound
inhibits NHE-1, NHE-2, and NHE-3 at widely differing concentrations
(26), thus providing a means of discerning between the isoforms. As shown in Fig. 6A, the antiport activity of
acinar cells, measured as the Na+-dependent
recovery of pHi from an acid load, could be effectively
inhibited by low doses of HOE694. The concentration required for
half-maximal inhibition was 0.06 µM (Fig.
6B), similar to that reported to inhibit NHE-1 (26), and
much lower than that needed to inhibit either NHE-2 or NHE-3
(K0.5 of 5 and 650 µM,
respectively). Together, the biochemical and functional findings indicate that NHE-1 is the primary Na+/H+
antiporter of rat acinar cells. Of note, low doses of HOE694 inhibited
the pHi recovery effectively both before and after treatment
with carbachol. In carbachol-stimulated cells the recovery was
inhibited by >95% by 3 µM HOE694, from 4.6 ± 0.53 to 0.2 ± 0.02 pH/min (n = 10; measured at
pHi 6.8). These data imply that NHE-1 is the isoform mediating
muscarinic activation of the antiport.
Mechanism of Activation of Na+/H+ Exchange
Having established that NHE-1 is the isoform activated by
muscarinic agonists in acinar cells, we proceeded to explore the mechanism(s) underlying this form of regulation. Phosphorylation of
serine residues within the cytoplasmic (C-terminal) domain has been
postulated to mediate receptor-mediated activation of NHE-1 in cultured
cells and in platelets (see Refs. 12 and 25, for reviews). Because
muscarinic stimulation triggers protein kinase activity in parotid
cells, we compared the phosphorylation of NHE-1 before and after
challenge with carbachol (Fig. 7A). Freshly
isolated cells were labeled with [32P]orthophosphate for
2 h and, after washing, they were incubated for 2 min with or
without 10 µM carbachol at 37 °C. Finally, the cells
were solubilized and NHE-1 was immunoprecipitated and analyzed by
SDS-PAGE and radiography. Immunoblotting with anti-NHE-1 antibody was
used to ensure that comparable amounts of the protein were precipitated
from control and treated cells.2 One of
four similar experiments is illustrated in Fig. 7 (top panel). As reported for other cell types (11, 12, 21) NHE-1 was
found to be phosphorylated in resting acinar cells. The absence of
radiolabel in samples treated with preimmune serum confirmed the
specificity of the immunoprecipitation protocol (not shown). Importantly, the extent of phosphorylation was indistinguishable before
and after muscarinic stimulation. Densitometric integration of
radiograms from four independent experiments indicated that phosphate
incorporation in carbachol-treated cells was 102 ± 19% (mean ± S.E.) of the control level (Fig. 7, botton
panel). Multiple exposures were performed to ensure that
differences in intensity were not obscured by film saturation.
Moreover, direct quantification by PhosphorImaging similarly showed
no significant difference between untreated and stimulated cells.
The above experiments indicate that net phosphorylation of NHE-1 does
not change when cells are acutely stimulated by muscarinic agonists.
While the overall phosphate content of NHE-1 appears to remain
constant, it is conceivable that phosphorylation of one site occurs and
is accompanied by dephosphorylation of a different site. Alternatively,
multiple phosphorylation sites may exist in the resting state. In this
case phosphorylation or dephosphorylation of a single residue may
affect the total 32P content only moderately and could go
undetected when the total radioactivity is compared. To test these
possibilities, we carried out phosphopeptide mapping of radiolabeled
immunoprecipitates from untreated and stimulated cells. The
precipitates were eluted from the beads and denatured with SDS, then
hydrolyzed using chymotrypsin. The protease yielded 4 phosphopeptides
of molecular mass between 4 and 6 kDa. This pattern was essentially
identical whether proteolysis was carried out for 1 or 5 min,
suggesting that the reaction had reached completion. Importantly, the
phosphopeptide composition and relative intensity of the bands were
identical in control and carbachol-treated samples (Fig.
8). These findings indicate that dephosphorylation of
one site with concomitant phosphorylation of another is unlikely. In
addition, no evidence was found for preferential dephosphorylation of
any one of the phosphopeptides resolved by chymotryptic cleavage of
NHE-1.
Primary fluid secretion by rat salivary glands is driven
osmotically by transepithelial salt gradients. As in most secretory epithelia, these gradients are generated by Na+-coupled
anion uptake across the basolateral membrane. In salivary glands, such
secondary active uptake of anions is mediated not only by
Na+-K+-2Cl cotransport, but also
by Cl
/HCO3
exchange coupled
to Na+/H+ exchange via the intracellular pH. By
alkalinizing the cytosol, NHE is important also in driving
HCO3
secretion and in guarding the cells
against metabolic acidosis. Despite recognition of the important role
of the antiporter in fluid secretion in salivary glands, the identity
of the isoform activated by muscarinic stimulation and the molecular
details of this activation had not been elucidated.
The present studies provide evidence that NHE-1 is the predominant
isoform in the parotid gland and that it is preferentially localized to
the basolateral membrane of acinar cells. This site is in agreement
with the reported location of NHE-1 in other epithelia and is
consistent with a role for the antiporter in promoting secondary
active, pHi-coupled Cl entry into the cell. That
this isoform is the major contributor to Na+ uptake in
carbachol-stimulated cells was shown using HOE694. At low micromolar or
submicromolar concentrations, this benzoylguanidine derivative greatly
inhibits NHE-1, while leaving the other isoforms unaffected. In our
studies, low doses of HOE694 markedly inhibited Na+-induced
H+ extrusion, implicating NHE-1 in the process.
Transcripts encoding NHE-2 have been detected in kidney medulla and cortex, jejunum, ileum, duodenum, stomach, and adrenal gland (11, 14, 27, 28), although the subcellular distribution of this isoform remains controversial. NHE-2 was also detected in the parotid, using RT-PCR. Its location and regulation were not pursued here partly because of unavailability of effective antibodies, but mainly because it seems to contribute little to the muscarinic response of the acinar cells, as judged by the effects of HOE694. NHE-2 may be located on the apical surface of the ductal cells, which failed to stain for NHE-1.
Activation of parotid cells by carbachol resulted in an apparent alkaline shift of the pH dependence of NHE, suggesting increased affinity for intracellular H+. A similar mode of activation was reported earlier in submandibular glands and in a variety of systems where NHE-1 mediates transport. In contrast, there is evidence that activation of NHE-2 entails primarily an increase in maximal velocity, at constant affinity (11). Albeit indirect, this evidence further supports the notion that NHE-1 is the isoform activated in acinar basolateral membranes.
Despite earlier controversy, it is now generally accepted that NHE-1 can be activated by elevation of intracellular [Ca2+]. In the parotid acinar cells, increased [Ca2+]i was sufficient to mimic the stimulation effected by muscarinic agonists: treatment with thapsigargin induced a Na+-dependent cytosolic alkalinization (not illustrated). Interestingly, the apical isoforms of epithelia (NHE-2 and/or NHE-3) have been suggested to become inhibited when [Ca2+] increases (29, 30). This would further argue for a predominant role of NHE-1 in parotid cells.
NHE-1 has been shown to be constitutively phosphorylated in other cells and additional phosphate groups are acquired upon stimulation by growth factors, phorbol esters, or okadaic acid (31, 32). It has therefore been postulated that phosphorylation mediates at least part of the biological responses of NHE-1. It was conceivable that the muscarinic activation, and that elicited by thapsigargin, were similarly mediated by activation of Ca2+-dependent kinases and/or inhibition of phosphatases. However, no change in phosphorylation of NHE-1 was detectable when the cells were stimulated. This observation is not without precedent, since the osmotic activation of the antiporter had been demonstrated earlier to occur in the absence of phosphorylation (21) and ionomycin-induced elevation of [Ca2+] similarly activated the exchanger without altering the phosphorylation state of NHE-1 (33, 34).
The simplest hypothesis available to explain our observations is that muscarinic stimulation of the antiporter results, at least in part, from the elevation of [Ca2+]i and formation of Ca2+-calmodulin complexes that bind and activate the antiporter directly. Wakabayashi and colleagues (33, 34) reported the existence of two calmodulin-binding domains in NHE-1, which upon binding Ca2+-calmodulin are believed to induce a conformational change in the protein that displaces an autoinhibitory domain, thereby stimulating the exchanger. Indeed, deletion of the putative autoinhibitory segment resulted in constitutively activated exchangers. We attempted to demonstrate the applicability of this model to acinar cells. At the concentrations required to block calmodulin, compound W7 completely inhibited the alkalinization of carbachol-stimulated cells. However, recovery of unstimulated cells from an acid load was also impaired. This may reflect a constitutive effect of calmodulin on NHE-1 of unstimulated acinar cells, and/or a direct (nonspecific?) effect of W7 on the antiport, which complicates the assessment of the role of calmodulin. Another calmodulin inhibitor, calmidazolium, induced a variable acidification of the cells, particularly after stimulation with carbachol. The confounding nature of this acidification again precluded evaluation of the role of calmodulin in muscarinic stimulation. Other approaches to evaluate the role of calmodulin are currently being considered.
In summary, carbachol stimulation of salivary acinar cells results in marked activation of NHE-1, which resides predominantly, if not exclusively, in the basolateral membrane. Such activation is independent of phosphorylation and can be at least partially mimicked by simply raising the concentration of [Ca2+]i to levels similar to those observed upon muscarinic stimulation. Although direct proof is as yet unavailable, we speculate that interaction of the antiporter with Ca2+-calmodulin is largely responsible for the stimulation.