Muscarinic Agonists Induce Phosphorylation-independent Activation of the NHE-1 Isoform of the Na+/H+ Antiporter in Salivary Acinar Cells*

(Received for publication, April 10, 1996, and in revised form, October 11, 1996)

Marli A. Robertson Dagger §, Michael Woodside Dagger , J. Kevin Foskett Dagger , John Orlowski par and Sergio Grinstein Dagger **

From the Dagger  Division of Cell Biology, Hospital for Sick Children, Toronto, M5G 1X8, Canada and the par  Department of Physiology, McGill University, Montreal, H3G 1Y6 Canada

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

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.


INTRODUCTION

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.


EXPERIMENTAL PROCEDURES

Materials

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.

Solutions

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+.

Preparation of Acinar Cells

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 Measurements

Approximately 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).

Calculation of Net Proton Flux

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*e-k2t) 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).


Fig. 2. pH dependence of the antiport in carbachol-activated cells. A, effect of carbachol (10 µM) on the rate of pHi recovery from an acid load. Cells suspended in Na+-rich medium (solution A) were pulsed with 40 mM NH4+ (solution C) and then perfused with Na+-free medium (solution D) with (open circles) or without (solid circles) carbachol. pHi recovery was next induced by reintroducing extracellular Na+. Traces are data from single cells representative of at least cells 30 cells from six different animals. B, pHi dependence of the Na+-induced net H+ (equivalent) flux (JH+). This was calculated as the sum of the Na+-dependent rate of pHi recovery and the rate of acid loading observed when Na+ was removed, multiplied by the buffering power, which was determined independently throughout the pH range of interest as described under "Experimental Procedures." No spontaneous acid loading was detected in control cells when Na+ was removed (n = 25). Straight lines were fitted by least squares using Cricket Graph 3.1.2. and data were analyzed using Statworks. Control cells: closed circles (fitted by the equation y = 1072-146×; r = 0.76, p < 0.001, n = 36). Carbachol-stimulated cells: open circles (fitted by y = 2125-283x; r = 0.88, p < 0.001, n = 28).
[View Larger Version of this Image (24K GIF file)]


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.

Immunoblotting and Immunoprecipitation

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 Mapping

Samples 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.

Immunofluorescence

Cells 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 Methods

Protein 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.


RESULTS

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).


Fig. 1. Carbachol stimulates H+ extrusion and Na+ uptake. A, effect of carbachol (10 µM) on pHi of a parotid acinar cell incubated in Na+-rich medium in the presence of HCO3- (solution B). Data are from a single cell representative of 38 determinations. B, effect of carbachol on pHi in the nominal absence of HCO3- in the presence (solution A) and absence of extracellular Na+ (solution D; Na+ was replaced with N-methyl-D-glucammonium). C, effect of carbachol on cytosolic free [Na+], measured by ratio imaging of intracellular SBFI. The cells were preincubated briefly with ouabain (1 mM) and then challenged with carbachol (10 µM). The samples were additionally treated with (solid squares) or without (open squares) 20 µM DMA (added 3 min prior to carbachol stimulation). Each trace was obtained from a single cell which is representative of at least 15 determinations.
[View Larger Version of this Image (17K GIF file)]


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 (approx 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 Dependence

The 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 Cells

To 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 approx 500 bp (Fig. 3, lane 2), while a smaller yield of the expected product (approx 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.


Fig. 3. Parotid glands express NHE-1 and NHE-2 transcripts. mRNA was extracted from partially purified parotid acinar cells and used as a template for RT-PCR with isoform-specific primers. M, molecular weight markers. P, template was linearized pCMV plasmid containing the full sequence of rat NHE-1 (lane 1), NHE-2 (lane 4), NHE-3 (lane 7), or NHE-4 (lane 10), hybridized with the corresponding primers. AC+, template was parotid acinar mRNA, which was reverse transcribed and hybridized with primers specific for NHE-1 (lane 2), NHE-2 (lane 5), NHE-3 (lane 8), or NHE-4 (lane 11). AC-, controls using similar template and primers as AC+, but reverse transcriptase was omitted. Representative of three separate experiments.
[View Larger Version of this Image (49K GIF file)]


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.


Fig. 4. NHE-1 protein is expressed in parotid glands. Microsomal fractions were separated by gel electrophoresis and immunoblotted with a polyclonal antibody specific for the C-terminal domain of the human NHE-1 isoform. AC mem, acinar cell membranes; AC mem+, acinar membranes after extraction of extrinsic proteins in alkaline carbonate solution; NHE-1+, membranes from antiport-deficient CHO cells stably transfected with the NHE-1 isoform; NHE-1-, untransfected antiport-deficient CHO cells. Representative of three separate experiments.
[View Larger Version of this Image (40K GIF file)]


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.


Fig. 5. Localization of NHE-1 in parotid glands. Frozen sections obtained from parotid glands (A-C) and isolated acinar cells (D) were fixed, permeabilized, and stained with polyclonal anti-NHE-1 antibody, followed by fluorescently labeled secondary (donkey anti-rabbit) antibody, as detailed under "Experimental Procedures." Samples were analyzed using a Bio-Rad 600 laser scanning confocal imaging system mounted on a Leitz Metallux-3 microscope using × 63 (1.3 NA) and × 100 (1.32 NA) oil-immersion objectives (Leitz). The arrow in B points to the apical membrane.
[View Larger Version of this Image (114K GIF file)]


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 approx 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.


Fig. 6. Inhibition of Na+-induced H+ extrusion by HOE694. A, cells suspended in Na+-rich medium (solution A) were pulsed with 40 mM NH4+ (solution C) and then exposed to Na+-free medium (solution D) with carbachol (10 µM). Finally, the cells were perfused with carbachol-containing Na+-rich medium (solution A) with (open circles) or without (solid circles) 3 µM HOE694. B, concentration dependence of the inhibitory effect of HOE694 in cells that were stimulated with carbachol and acid-loaded as in A. Na+/H+ exchange was induced by suspending the cells in a medium containing 35 mM Na+. The pHi recovery rate was measured and normalized to the rate recorded in the absence of the inhibitor. A reduced concentration of Na+ was used to slow down the pHi recovery, thereby increasing the accuracy of the determinations. Data are means ± S.E. of three experiments.
[View Larger Version of this Image (16K GIF file)]


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.


Fig. 7. Phosphorylation of NHE-1 in acinar cells. Parotid acinar cells were preloaded with [32P]orthophosphate for 3 h at 37 °C, washed, and treated with (Car) or without (Con) 10 µM carbachol for 2 min, as indicated. The cells were next extracted and immunoprecipitated using anti-NHE-1 antiserum. The immunoprecipitates were subjected to electrophoresis and blotting onto nitrocellulose. The blots were probed with the NHE-1 antibodies (alpha -NHE-1 Ab) using ECL (top right panel). After removal of the luminescent mixture, the blots were rinsed, dried, and used for autoradiography of 32P-labeled proteins (top left panel). The position of molecular mass markers (in kDa) is indicated to the left. The location of NHE-1 and the heavy chain of the immunoglobulin (IgG) used for precipitation are indicated. The blot is representative of four experiments. Bottom panel, quantitation of radioactivity incorporated into NHE-1, obtained by PhosphorImaging. Data are means ± S.E. of four experiments.
[View Larger Version of this Image (31K GIF file)]


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.


Fig. 8. Phosphopeptide composition of NHE-1 isolated from control and carbachol-treated cells. Untreated (Con) and carbachol-stimulated (Car) cells were extracted and immunoprecipitated with NHE-1 antibodies. The radiolabeled immunoprecipitates were subjected to proteolytic degradation by chymotrypsin, as described under "Experimental Procedures." The resulting phosphopeptides were resolved by SDS-PAGE using 15% acrylamide. The position of major phosphopeptides is indicated by arrows. The location of molecular mass markers is also noted. Representative of four similar experiments.
[View Larger Version of this Image (53K GIF file)]



DISCUSSION

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.


FOOTNOTES

*   This work was supported in part by the Canadian Cystic Fibrosis Foundation and the Medical Research Council of Canada. 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.
§   Current address: Dept. of Pediatrics, The University of Calgary, Calgary, Alberta, Canada.
   Current address: Dept. of Physiology, University of Pennsylvania, Philadelphia, PA.
**   International Scholar of the Howard Hughes Medical Institute. To whom correspondence should be addressed: Div. of Cell Biology, Hospital for Sick Children, 555 University Ave., Toronto, M5G 1X8 Canada. Tel.: 416-813-5727; Fax: 416-813-5028; E-mail: sga{at}sickkids.on.ca.
1    The abbreviations used are: DMA, dimethylamiloride; BCECF, 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein; [Ca2+]i, cytosolic free calcium concentration; HOE694, 3-(methylsulfonyl-4-piperidino-benzoyl)-guanidine methanesulfonate; NHE, Na+/H+ exchanger; pHi, intracellular pH; PBS, phosphate-buffered saline; SBFI, sodium-binding benzofuran isophthalate; RT-PCR, reverse transcriptase-polymerase chain reaction; bp, base pair(s); PAGE, polyacrylamide gel electrophoresis.
2    The short film exposure times required for development of enhanced chemiluminescence enabled us to perform sequential immunoblotting and autoradiography of the same samples.

REFERENCES

  1. Cook, D. I., Van Lennep, E. W., Roberts, M. L., and Young, J. A. (1994) in Physiology of the Gastrointestinal Tract (Johnson, L. R., ed), Third edition, pp. 1061-1117, Raven Press, New York
  2. Manganel, M., and Turner, R. J. (1990) J. Biol. Chem. 265, 4284-4289 [Abstract/Free Full Text]
  3. Pirani, D., Evans, L. A. R., Cook, D. I., and Young, J. A. (1987) Pflugers Arch. 408, 178-184 [Medline] [Order article via Infotrieve]
  4. Robertson, M. A., and Foskett, J. K. (1994) Am. J. Physiol. 267, C146-C156 [Abstract/Free Full Text]
  5. Dissing, S., and Nauntofte, B. (1990) Am. J. Physiol. 259, G1044-G1055 [Abstract/Free Full Text]
  6. Lau, K. R., Elliot, A. C., and Brown, P. D. (1989) Am. J. Physiol. 256, C288-C295 [Abstract/Free Full Text]
  7. Melvin, J. E., Moran, A., and Turner, R. J. (1988) J. Biol. Chem. 263, 19564-19569 [Abstract/Free Full Text]
  8. Soltoff, S. P., McMillian, M. K., Cantley, L. C., Cragoe, E. J., Jr., and Talamo, B. R. (1989) J. Gen. Physiol. 93, 285-319 [Abstract]
  9. Manganel, M., and Turner, R. J. (1989) J. Membr. Biol. 111, 191-198 [Medline] [Order article via Infotrieve]
  10. Turner, R. J., Paulais, M., Manganel, M., Lee, S., Moran, A., and Melvin, J. (1993) Crit. Rev. Oral Biol. Med. 4, 385-391 [Medline] [Order article via Infotrieve]
  11. Yun, C. H. C., Chung-Ming, T., Samir, K. N., Levine, S. A., Brant, S. R., and Donowitz, M. (1995) Am. J. Physiol. 269, G1-G11 [Abstract/Free Full Text]
  12. Noel, J., and Pouyssegur, J. (1995) Am. J. Physiol. 268, C283-C296 [Abstract/Free Full Text]
  13. Chung-Ming, T., Levine, S. A., Yun, C. H., Brant, S. R., Pouyssegur, J., Montrose, M., and Donowitz, M. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 9110-9114 [Abstract]
  14. Orlowski, J., Kandasamy, R. A., and Shull, G. E. (1992) J. Biol. Chem. 267, 9331-9339 [Abstract/Free Full Text]
  15. Klanke, C. A., Su, Y. R., Callen, D. F., Kandasamy, R. A., Orlowski, J., Shull, G. E., and Menon, A. G. (1995) Genomics 25, 615-622 [CrossRef][Medline] [Order article via Infotrieve]
  16. Foskett, J. K., Roifman, C. M., and Wong, D. (1991) J. Biol. Chem. 266, 2778-2782 [Abstract/Free Full Text]
  17. Foskett, J. K., and Melvin, J. E. (1989) Science 244, 1582-1585 [Medline] [Order article via Infotrieve]
  18. Foskett, J. K. (1988) Am. J. Physiol. 255, C566-C571 [Abstract/Free Full Text]
  19. Roos, A., and Boron, W. F. (1981) Physiol. Rev. 61, 296-397 [Free Full Text]
  20. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156-165 [CrossRef][Medline] [Order article via Infotrieve]
  21. Grinstein, S., Woodside, M., Sardet, C., Pouyssegur, J., and Rotin, D. (1992) J. Biol. Chem. 267, 23823-23828 [Abstract/Free Full Text]
  22. Lee, S. I., and Turner, R. J. (1991) Am. J. Physiol. 261, G111-G118 [Abstract/Free Full Text]
  23. Nauntofte, B. (1992) Am. J. Physiol. 263, G823-G837 [Abstract/Free Full Text]
  24. Grinstein, S., Rothstein, A., and Cohen, S. (1985) J. Gen. Physiol. 85, 765-788 [Abstract]
  25. Grinstein, S., Rotin, D., and Mason, M. J. (1989) Biochim. Biophys. Acta 988, 73-97 [Medline] [Order article via Infotrieve]
  26. Counillon, L., Scholz, W., Lang, H. J., and Pouyssegur, J. (1993) Mol. Pharmacol. 44, 1041-1045 [Abstract]
  27. Wang, Z., Orlowski, J., and Shull, G. E. (1993) J. Biol. Chem. 268, 11925-11928 [Abstract/Free Full Text]
  28. Collins, J. F., Honda, T., Knobel, S., Bulus, N. M., DuBois, R., and Ghishan, F. K. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 3938-3941 [Abstract]
  29. Burns, K. D., Homma, T., and Harris, R. C. (1991) Am. J. Physiol. 261, F607-F616 [Abstract/Free Full Text]
  30. Cohen, M. E., Reinlib, L., Watson, A. J. M., Gorelick, F., Rys-Sikora, K., Tse, M., Sharp, G. W. G., and Donowitz, M. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 8990-8994 [Abstract]
  31. Sardet, C., Counillon, L., Franchi, A., and Pouyssegur, J. (1990) Science 247, 723-726 [Medline] [Order article via Infotrieve]
  32. Bianchini, L., Woodside, M., Sardet, C., Pouyssegur, J., Takai, A., and Grinstein, S. (1991) J. Biol. Chem. 266, 15406-15413 [Abstract/Free Full Text]
  33. Bertrand, B., Wakabayashi, S., Ikeda, T., Pouyssegur, J., and Shigekawa, M. (1994) J. Biol. Chem. 269, 13703-13709 [Abstract/Free Full Text]
  34. Wakabayashi, S., Bertrand, B, Ikeda, T., Pouyssegur, J., and Shigekawa, M. (1994) J. Biol. Chem. 269, 13710-13715 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.