INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

THE VERTEBRATE Na⁺/H⁺ exchangers, or NHEs, are a family of transmembrane proteins that mediate electroneutral exchange of 1 Na⁺:1 H⁺, are inhibited by the diuretic amiloride and its analogs (8), and are important in cellular Na⁺ and pH homeostasis. At least six vertebrate exchangers have been cloned thus far: NHE1-5 from mammalian cells (21, 22, 35), -NHE from the trout erythrocyte (5), and NHE6, which has been identified but not fully characterized (22). The mammalian isoforms vary in size, tissue distribution, regulation, membrane localization, and pharmacological characteristics. They perform diverse roles as in the regulation of intracellular pH (pH_i), cell volume, growth, differentiation, and epithelial transport (8, 15, 16, 22, 35). However, they all share a similar topology: the amino-terminal region with 500 amino acids is the hydrophobic transmembrane transport domain, whereas the carboxyl-terminal region with 300 largely hydrophilic amino acids and protein kinase consensus sequences is the cytoplasmic, regulatory domain (21, 22, 26, 34, 35). Growth factors, hormones, and neurotransmitters evoking tyrosine kinases, protein kinase A (PKA), protein kinase C (PKC), or Ca²⁺-CaM kinases (6, 20-22, 24, 25, 27, 30, 31, 33, 35), as well as physical factors such as cell shrinkage, regulate exchanger activities, albeit in an isoform-specific manner (4, 9, 10, 15). NHE1, the ubiquitous isoform, is present in almost all cell types and is located on the basolateral membrane of polarized epithelial cells. In contrast, NHE2, NHE3, and NHE4 are largely epithelial in distribution. Pharmacological and immunodetection studies indicate that NHE3 and NHE2 are located on the apical membrane of epithelial cells and are involved in vectorial transport (reviewed in Refs. 21, 22, 32, and 35).

Sodium conservation is extremely important in most terrestrial vertebrates, and land birds such as the chicken have developed highly efficient Na⁺-absorptive mechanisms in their small and large intestines (7, 9). In a very recent study, NHE2 and NHE3 protein and activity in chicken small intestine and colon were reported (9). The small intestinal cells or enterocytes have a well-developed brush-border membrane for enhanced absorption. Na⁺ entry into these cells is via a prominent apical Na⁺/H⁺ exchanger (6, 19, 27). Previous studies in our laboratories characterized the activity and regulation of this exchanger. A prominent feature of the chicken intestinal brush-border exchanger is that activation of any signaling pathway involving cAMP, cGMP, Ca²⁺, or PKC inhibits its activity (6, 13, 20, 27, 28, 30). cAMP and cGMP appear to act via an increase in intracellular Ca²⁺ (27, 28). Thus it is very likely that the brush-border membrane exchanger(s) is constitutively active for efficient vectorial transport and plays a major role in Na⁺ conservation. Some features of this exchanger, i.e., its sensitivity to amiloride, inhibition by Ca²⁺ and PKC, and antigenicity, resemble those of mammalian NHE3, which is the apical exchanger involved in vectorial transport (1, 3, 6, 8, 13, 16, 27, 28, 35).

In this study we determined if the chicken also possessed a nonepithelial Na⁺/H⁺ antiporter, analogous to mammalian NHE1, that performed "housekeeping" roles such as pH_i regulation, growth, and differentiation (15, 16; reviewed in Refs. 21, 22, 32, 35). Characterization of such an isoform would indicate any unique features of another vertebrate nonmammalian exchanger, such as the trout exchanger (5), and imply the importance of adaptation and divergence during the course of evolution. Chicken embryonic fibroblasts, SL-29, were used for this purpose. They can be grown in culture and, because they are primary cells, may likely represent the in vivo situation. We report here a detailed characterization of the SL-29 Na⁺/H⁺ exchanger that shares some properties with mammalian NHE1 and yet is distinct from NHE1. It is also distinct in its properties from the chicken intestinal brush-border exchanger characterized previously (6). Our findings suggest that avian species possess a unique family of NHE isoforms.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Materials. Primary chicken embryonic fibroblasts, SL-29, were obtained from the American Type Culture Collection (Rockville, MD). Cell culture supplies were obtained from Costar (Cambridge, MA), media and serum were from GIBCO BRL (Grand Island, NY), enhanced chemiluminescence kit was from Amersham (Arlington Heights, IL), ²²Na was from NEN-DuPont (Boston, MA), Budget Solve and Bio-Safe II scintillation cocktails and vials were from Research Products International (Mt. Prospect, IL), Affi-gel 10 was from BioRad (Richmond, CA), restriction endonucleases were from GIBCO BRL and Promega (Madison, WI), oligo(dT)-cellulose columns were from Collaborative Biomedical Products (Becton Dickinson Labware, Bedford, MA), DECAprime II^TM DNA labeling kit was from Ambion (Austin, TX), and all other chemicals and supplies were from Sigma (St. Louis, MO) and Fisher Scientific (Itasca, IL).

Cell culture. All cells were grown in a humidified 95% air-5% CO₂ atmosphere at 37°C. The cells were subcultured using 0.1% (wt/vol) trypsin-1 mM EDTA in PBS, pH 7.4. Chicken embryonic fibroblasts were grown in MEM with nonessential amino acids and Earle's balanced salt solutions, supplemented with 1 mM sodium pyruvate, 5% tryptose phosphate broth, 5% fetal bovine serum (FBS), 10,000 U/l penicillin, and 10 mg/l streptomycin. Because these are cells in primary culture, they can be passaged only a few times, and different batches needed to be procured for the complete study. Cells were routinely received in passages 13-14, and most experiments were performed on cells in passages 14-18. The PS127A cell line, which was developed and kindly provided by Dr. J. Pouyssegur (26), was grown in DMEM supplemented with 5 U/ml penicillin, 5 µg/ml streptomycin, and 10% FBS. Tests to rule out mycoplasma contamination were not performed, because no glaring discrepancies in cell growth pattern were noticed.

Preparation of microsomal membranes. The cells were washed twice with PBS, scraped off, and resuspended in a small volume of PBS. The suspension was centrifuged at 2,000 g for 15 min, and the resulting pellet was homogenized in buffer (HB) containing (in mM) 10 HEPES-Tris, pH 7.4, 3 EGTA, 1 EDTA, 10 mannitol, 0.1 phenylmethylsulfonyl fluoride, 2 dithiothreitol (DTT), and 0.01 mg/ml L-1-tosylamide-2-phenylethylchloromethyl ketone, and 0.001 mg/ml each of leupeptin, pancreatic, and soybean trypsin inhibitors. The homogenate was centrifuged at 2,000 g for 15 min, and the resulting postnuclear supernatant was centrifuged at 100,000 g for 1 h to obtain the microsomal membrane pellet. The pellet was resuspended in a minimal volume of HB. Protein content was assayed by the Lowry procedure as described previously (30).

Isolation of antibody against NHE1. The generation and characterization of this antibody has been described previously (18). Briefly, a fusion protein comprising the sequence of amino acids 631-746 of human NHE1 (26) and glutathione-S-transferase (GST) was injected into New Zealand White rabbits (18). The polyclonal antibodies were affinity purified using antigen linked to Affi-gel 10. To remove antibodies to GST, the antiserum was first run on a GST column, and the resulting flow-through was run on an NHE1 antigen affinity column. The antibody was eluted with 0.2 M glycine-HCl, pH 2.2.

Immunoblotting. Proteins were resolved by SDS-PAGE on 6% gels by the Laemmli method and electrotransferred to nitrocellulose by the Towbin method using the modifications described previously (29). The membrane was blocked in Blotto buffer containing 5% (wt/vol) fat-free dry milk powder, 150 mM NaCl, 10 mM sodium phosphate, pH 7.4, 2 mM EDTA, and 0.2% (vol/vol) Nonidet-40 for 1 h. It was then incubated in primary antibody (1:200) in Blotto overnight at 4°C. It was washed (2 × 5 min and 2 × 10 min in Blotto) and incubated in horseradish peroxidase-conjugated goat anti-rabbit IgG (1:1,000) in Blotto containing 1% nonfat milk for 1 h at room temperature. The membranes were washed as follows: 3 × 5 min, 3 × 10 min, and 1 × 15 min in Blotto, and reaction products were visualized by chemiluminescence.

Northern blotting. The SL-29 fibroblasts were washed in PBS and resuspended in GTC lysis buffer containing 4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7.0, 0.5% (wt/vol) sodium lauryl sarcosinate, and 1% (vol/vol) -mercaptoethanol. Total RNA was isolated by centrifugation in a 5.7 M CsCl cushion followed by extraction with Tris-saturated (pH 7.4) phenol and chloroform-isoamyl alcohol (24:1) as described (23). Polyadenylated mRNA was isolated by affinity chromatography on an oligo(dT) cellulose column. The mRNA was separated on a 1% agarose, 6% formaldehyde gel, with RNA size markers run in parallel and transferred by capillary action to nylon membranes (Hybond N, Amersham). The following isoform-specific probes were used: full-length rat heart NHE1 cDNA probe consisting of bases 614-3452, a rat NHE2, Pst I cDNA fragment corresponding to bases 260-3598, and a full-length rat NHE3 cDNA consisting of bases 1-3036 (from a PvuI I-Nsi I fragment) (3, 15). Note that the NHE2 probe is almost full length with 789 of the 813 translated codons and is lacking the 3'-end noncoding sequence. The cDNA clones were kindly provided by Dr. Gary Shull, The University of Cincinnati, Cincinnati, OH. The vectors were incubated with the appropriate restriction enzymes, and inserts were then purified on a 1.2% agarose gel. The purified inserts were labeled by random oligonucleotide primers (Ambion DECAprime II^TM). Hybridization and washes are as described in the figure legend.

²²Na⁺ uptake studies. Cells grown in 24-well clusters were acid preloaded in a buffer containing (in mM) 50 NH₄Cl, 70 choline chloride, 5 KCl, 1 MgCl₂, 2 CaCl₂, 5 glucose, and 15 MOPS, pH 7.0 at 37°C for 1 h. They were transferred to room temperature, and the preload buffer was replaced by two changes of Na⁺-free wash buffer (containing 120 mM choline chloride and 15 mM HEPES, pH 7.4). The cells were then exposed to uptake buffer containing (in mM) 5 or 20 NaCl, 115 or 100 chloride, 1 ouabain, 1 MgCl₂, 2 CaCl₂, 20 HEPES, pH 7.4, and 1 µCi/ml of carrier-free ²²Na⁺ for 5 min unless otherwise indicated (34). External Na⁺ concentration was reduced from 20 mM in early experiments to 5 mM in later experiments to increase ²²Na⁺ specific activity. The reaction was terminated by three washes with ice-cold PBS. Samples were hydrolyzed in 0.1 N HNO₃ and counted by liquid scintillation spectrometry.

Data analysis. Data are presented as means ± SE of 3-5 experiments, unless otherwise mentioned. In each experiment the assay was performed in triplicate. Data were analyzed and compared using Student's t-test for paired data.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Time course and kinetics of Na⁺/H⁺ exchange. To determine whether SL-29 cells possess Na⁺/H⁺ exchange activity and to examine the linearity of ²²Na⁺ uptake, cells were initially acid-loaded by exposure to NH₄Cl to maximize the driving force for the exchanger. The acid-loaded cells were exposed to uptake buffer containing ²²Na⁺ for 1, 3, 5, and 10 min in the presence and absence of 10⁴ M of the amiloride analog dimethylamiloride (DMA). NHE activity was calculated as the difference between total ²²Na⁺ uptake and uptake in the presence of DMA. As shown in Fig. 1, DMA-sensitive uptake was linear up to 10 min, whereas DMA-insensitive uptake remained relatively constant, indicating the presence of an Na⁺/H⁺ exchanger in the SL-29 cells. In subsequent experiments, uptake was measured at 5 min. In the absence of an acid gradient, DMA-sensitive uptake was only 7.5-21% of that in its presence (2 experiments in quadruplicate, data not shown). In acid-loaded cells, ²²Na⁺ uptake was neither inhibited by furosemide nor by the stilbene derivative SITS (data not shown), suggesting that NHE was the major Na⁺ transport mechanism. It must be emphasized that SL-29 cells are primary cells, and, although we restricted the number of times the cells were passaged, variability in basal levels was observed from batch to batch. As it is useful for future investigations to have such variability documented in literature, we have expressed the data, as far as possible, in absolute values. However when basal values varied greatly, results are expressed as a percentage of control. Although the absolute values may differ, the trend (stimulation or inhibition) was generally the same between experiments.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Time course of Na+ uptake in SL-29 cells. Acid-preloaded SL-29 cells were washed twice in Na+-free buffer and exposed to uptake buffer containing 1 µCi/ml of 22Na+ in 20.4 mM Na+, as described in Materials and methods. 22Na+ uptake was assayed in the absence and presence of 0.1 mM dimethylamiloride (DMA). All buffers used were HCO3 free. Values are expressed as means ± SE of triplicate measurements from 1 experiment. In all further experiments, uptake was determined at 5 min.

Figure 2 shows exchanger activity in response to increasing concentrations of external Na⁺. After NH₄Cl-induced acidification, cells were exposed to uptake buffer containing a concentration of Na⁺ ([Na⁺]) in the range of 1.875 to 120 mM in the presence and absence of 10⁴ M ethylisopropylamiloride (EIPA). EIPA was used in most subsequent experiments because it was the most potent inhibitor (Fig. 3). Osmolarity was maintained constant among the buffers by adding choline chloride. A plot of EIPA-sensitive Na⁺ uptake as a function of external [Na⁺] is a rectangular hyperbola, demonstrating that the exchanger obeys Michaelis-Menten kinetics with a single external, saturable Na⁺ binding site. A linear Hanes-Woolf transformation of the data (Fig. 2, inset) revealed a K_m of 18.75 mM and a maximal velocity of 156.3 nmol · mg protein¹ · min¹.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effect of external Na+ concentration on exchanger activity. Effect of increasing Na+ concentrations on the initial rate of Na+ uptake, calculated on the basis of the ethylisopropylamiloride (EIPA; 0.1 mM)-sensitive component of total Na+ uptake. Values are expressed as means ± SE of 4 experiments, each done in triplicate. Inset: linear transformation of the data using a Hanes-Woolf plot based on the equation [S]/v = {[S] + Km}/Vmax, where Km is the Michaelis constant, Vmax is maximal velocity, [S] is Na concentration, and v is rate of uptake.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of amiloride and its analogs on Na+ uptake. Inhibition profiles of Na+ (20.4 mM) uptake into SL-29 cells exposed to varying concentrations of amiloride, EIPA, or DMA are shown. y-Axis depicts rate of uptake expressed as %control, i.e., in absence of inhibitor. Data are expressed as means ± SE of 3 experiments each in the case of amiloride and DMA and 5 experiments in the case of EIPA.

The process is electroneutral because agents known to alter membrane potential do not affect EIPA-sensitive ²²Na⁺ uptake in SL-29 cells. Thus addition of the K⁺ ionophore valinomycin (10 µM), an agent routinely used to clamp membrane potential, had little effect on exchanger activity (the EIPA-sensitive uptake values were 9.798 vs. 8.395 nmol · mg protein¹ · min¹ in control and valinomycin-treated cells, respectively). Similarly, when cells were exposed to an uptake buffer containing 120 mM KCl, there was little effect of valinomycin on EIPA-sensitive uptake. Additionally, the electrogenic protonophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP) (10 µM) had no effect in combination with valinomycin (9.798 and 9.704 nmol · mg protein¹ · min¹ in control and CCCP-treated cells, respectively) or by itself (9.798 and 8.028 nmol · mg protein¹ · min¹ in control and CCCP-treated cells, respectively). These findings suggest that the ²²Na⁺ uptake being measured occurs via an electroneutral exchanger.

Inhibition by amiloride and its analogs. A useful way to distinguish the various mammalian NHE isoforms is by their relative sensitivities to inhibition by amiloride and its analogs. SL-29 cells were acidified and exposed to increasing concentrations of amiloride and its analogs, DMA and EIPA, in the uptake buffer containing 20.4 mM Na⁺. The IC₅₀ values were calculated as 2 × 10⁶ M for amiloride, 3.5 × 10⁷ M for DMA, and 2.5 × 10⁸ M for EIPA (Fig. 3). This potency series of EIPA > DMA > amiloride for inhibition of exchange activity is found in a number of cell types (8, 19). Because EIPA was the most potent analog, NHE activity was measured as the EIPA-sensitive uptake in subsequent experiments.

An IC₅₀ value of 1-2.5 × 10⁶ M for amiloride is shared by mammalian NHE1 and NHE2; the amiloride-insensitive NHE3 has an IC₅₀ value one to two orders of magnitude higher (8, 21, 32, 35). Hence the exchanger in the SL-29 cells was likely to be related either to NHE1 or NHE2. However, NHE2 has a lower sensitivity to EIPA than NHE1; the reported IC₅₀ values of NHE1 and NHE2 range from 1.5 to 2.5 × 10⁸ M and 0.08 to 1 × 10⁶ M, respectively (21, 35). The similarity in IC₅₀ values between the SL-29 exchanger and NHE1 suggests that the chicken fibroblast exchanger is an NHE1-like molecule.

Effects of regulatory agents. Because the pharmacological studies imply the presence of an NHE1-type activity in SL-29 cells, we examined the effects of well-known regulators of NHE1, such as serum, transforming growth factor (TGF)-, okadaic acid, and hypertonicity, on its activity. Serum stimulates NHE1 activity with a concomitant increase in NHE1 phosphorylation in mammalian fibroblasts (24). To obviate any change in osmolality, dialyzed FBS (10%) was used. Short-term (5 min) serum exposure caused a 43% stimulation (Table 1). As mentioned above, the variability in the basal levels represents batch-to-batch differences, as these are primary cells.

The stimulation seen with serum indicates that the exchanger is activatable by growth factors. Epidermal growth factor (EGF) and TGF- can be used interchangeably because they bind to the same receptor as exemplified in chicken granulosa cells wherein EGF-induced NHE stimulation was mimicked by TGF- but not by TGF- (16). In highly confluent SL-29 cells, TGF- (10 pg/µl or 1.67 nM) caused a marked 230% stimulation in EIPA-sensitive Na⁺ uptake (Table 1).

Mammalian NHE1 is also stimulated by pathways that involve Ca²⁺ and PKC (16, 22, 34). The effect of PKC stimulation on the SL-29 exchanger was examined by exposing the cells to increasing concentrations of the phorbol ester, phorbol-12,13-dibutyrate (PDB). In contrast to other NHE1 systems, which exhibit a monophasic response, SL-29 cells exhibited a biphasic response to PDB as shown in Fig. 4. To determine if the SL-29 exchanger was modulated by altering intracellular Ca²⁺ levels, the cells were treated with the Ca²⁺ ionophore A-23187 (1 µM). A total of nine experiments, with values being determined in triplicate in each experiment, were performed. As shown in Fig. 5, the data could be separated into two groups. In one group comprising n = 5, there was a 30% inhibition and in the other (n = 4), a 22% stimulation. This variability could not be attributed to any obvious differences in passage number, state of confluence, batch of cells, or absolute activity values between the two groups. Because other regulators do not have such contrasting opposite effects, it is unlikely that a general phenomenon like clonal variability could account for these differences. The basal activity was 10 ± 1.73 nmol/mg in group I and 7.61 ± 1.08 nmol/mg in group II and therefore not statistically different.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effect of protein kinase C activation on Na+ uptake. A representative curve (of 5 experiments) is shown, where cells were exposed to concentrations of phorbol-12,13-dibutyrate (PDB) increasing from 1010 to 106 M. Data are means ± SE of triplicate measurements.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Effect of an increase in intracellular Ca2+ concentration on Na+ uptake. Figure depicts Na+/H+ exchanger (NHE) activity in control or Ca2+ ionophore (A-23187, 1 µM)-containing uptake buffer. Data divided into group 1 (n = 5, top) and group 2 (n = 4, bottom) on the basis of inhibitory or stimulatory response, respectively (see Results for discussion). Each n value represents a separate experiment and is average of triplicate measurements. ** Significance at P < 0.025 and **** P < 0.0005 using paired Student's t-tests.

cAMP generally does not affect mammalian NHE1, whereas it inhibits mammalian NHE3 and stimulates trout -NHE (5, 16, 33). The effect of cAMP on the SL-29 NHE was examined by exposing cells to the cAMP analog, chlorophenylthio-cAMP (cpt-cAMP). Addition of 10⁶ and 10⁴ M cpt-cAMP caused significant decreases of 36 and 65% in activity, respectively (n = 3). The activity with 10⁴ M was significantly less than that with 10⁶ M cpt-cAMP (Fig. 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of cAMP on NHE activity. Graph depicts effects of chlorophenylthio (cpt)-cAMP at 106 and 104 M (n = 3). y-Axis represents NHE activity as %corresponding control, i.e., without cpt-cAMP. x, Effect at 104 was different than that at 106. Significance of inhibition at *** P < 0.01 and ** P < 0.025 using paired Student's t-tests.

The pathways discussed thus far imply the involvement of kinases in the regulation of the SL-29 Na⁺/H⁺ exchanger. To determine the role of phosphatases in this regulation, we next examined the effects of the phosphatase inhibitors okadaic acid and calyculin. Okadaic acid, a potent inhibitor of phosphatase IIA, stimulates NHE1 activity in mammalian fibroblasts with a concomitant increase in its phosphorylation (2). In highly confluent SL-29 cells, 1 µM okadaic acid caused a twofold stimulation (Table 1). In contrast, calyculin (1 µM), which inhibits phosphatase I and IIA, had a small but significant (12%) inhibition of ²²Na⁺ uptake (data not shown).

Physical factors such as cell shrinkage are also known to affect NHE activity (4, 10). On increasing the osmolarity of the medium from 270 to 400 or 550 mosM, the activity of the SL-29 exchanger increased 1.4-fold. There was no difference in the response to 400 compared with 550 mosM (Fig. 7).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of hypertonicity on Na+ uptake. Effect of hypertonicity was examined by exposing SL-29 cells to uptake buffers of 270 (isotonic, control), 400, or 550 mosM. Osmolarity was increased by addition of choline chloride. Activity is expressed as %control values. * Significant increases (P < 0.05, paired Student's t-test; n = 6).

Molecular characterization of the exchanger. On the basis of the functional similarities between mammalian NHE1 and the SL-29 NHE, we attempted to identify the SL-29 NHE using mammalian probes. SL-29 membrane proteins were probed with a polyclonal antibody directed against a unique mammalian NHE1 sequence (amino acids 631-746 of human fibroblast NHE1) (26). This antibody was shown previously to recognize NHE1 in nontransfected human foreskin fibroblasts, HSWP (18), rat intestinal membranes (3), and more recently in turtle gastrointestinal tissues (12). Although a 105-kDa band was detected with membranes from the NHE1-transfected fibroblasts PS127A, no product was detected in SL-29 membranes (Fig. 8) or in the NHE-deficient PS120 cells.

View larger version (41K): [in this window] [in a new window]   Fig. 8.   Immunoblotting of SL-29 membranes with NHE1 fusion protein antibody. In this representative immunoblot, 10 µg of PS127A and 100 µg of SL-29 membrane proteins were probed with a 1:200 dilution of NHE1 fusion protein antibody followed by a 1:1,000 dilution of goat anti-rabbit peroxidase-labeled secondary antibody. Reaction products (arrow) were visualized by enhanced chemiluminescence. Mr, relative molecular weight.

To determine whether the SL-29 cells have a transcript related to any of the mammalian NHE isoforms, mRNA from these cells was run in parallel with mRNA from rat jejunum (as a positive control) and probed with full-length cDNAs corresponding to NHE1 or 3 and an almost full-length cDNA corresponding to NHE2. A positive reaction was seen only with the NHE1 probe in SL-29 cells at high stringency. However, the size of this transcript was 3.9 kb (Fig. 9) in comparison with the 4.8-kb NHE1 mRNA of rat jejunum (Fig. 9). Although the NHE2 and NHE3 probes recognized transcripts of 4.4 and 5.2 kb, respectively, in rat enterocytes, they did not recognize any product in SL-29 cells, even under lower stringency conditions (see Fig. 9 legend). It is unlikely that failure to detect an NHE2 is due to a partial-length NHE2 probe, because this probe accounts for all but 24 codons of the 813 codons of NHE2.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Northern blotting of SL-29 mRNA with NHE isoform-specific cDNA probes. A total of 4 µg each of rat jejunal and chicken intestinal poly(A+) RNA was subjected to Northern analysis (see MATERIALS AND METHODS). NHE1 cDNA: prehybridization at 65°C for 2-3 h, followed by overnight hybridization at 50°C. The blots were washed as follows: 2 × 15 min at 60°C in 2 × SSC (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate, pH 7.0) and 1 × 30-min at 60°C in 2× SSC + 0.1% SDS. Three additional washes of 30 min in 2× SSC, the last one with 2% SDS were included to lower background. Blot was exposed to Kodak XAR film for 4 days with intensifying screen. Heavy arrow indicates 3.9 kb and light arrow indicates 4.8 kb. Similar results were obtained with longer exposures. NHE2 cDNA: prehybridization at 50°C overnight, followed by hybridization at 50°C for 2 h. Blots were washed as follows: 1 × 15 min at 60°C in 2× SSC and 2 × 30 min at 60°C in 2× SSC, the last one with 2% SDS. Blot was exposed for 2 days with intensifying screen. Arrow depicts NHE2. Similar results were obtained with longer exposures. NHE3 cDNA: prehybridization at 65°C for 2-3 h was followed by overnight hybridization at 65°C. Washes included 1 × 20 min in 2× SSC at room temperature and 1 × 15 min in 0.1× SSC + 0.1% SDS at 65°C. Blot was exposed for 6 days with intensifying screen. Arrow depicts 5.2-kb NHE3. Similar results were obtained with lower stringency conditions of washing [2 × 15 min in 2× SSC and 1 × 30 min in 2× SSC, 0.1% (wt/vol) SDS].

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

These studies demonstrate the presence of an Na⁺/H⁺ exchanger in the chicken embryonic fibroblasts SL-29 using ²²Na⁺ uptake measurements. NHE activity, calculated as the amiloride-sensitive component of total ²²Na⁺ uptake, increased linearly with time (Fig. 1). Our data indicate that ²²Na⁺ uptake in acid-loaded cells is largely due to an Na⁺/H⁺ exchanger and not other Na⁺-dependent transporters for several reasons. First, Na⁺ uptake is dependent on H⁺ extrusion because there is <22% uptake in the absence of a pH gradient. Second, uptake was studied in HCO₃-free conditions, thereby severely attenuating contributions of HCO₃-dependent Na⁺ transport. Furthermore, SITS did not inhibit transport, suggesting that any metabolically produced CO₂ had no effect on Na⁺ uptake. Third, the Na⁺-K⁺-2Cl cotransport inhibitor furosemide had no significant effect on Na⁺ uptake (data not shown). The studies with ²²Na⁺ uptake were corroborated by pH_i measurements using the pH-sensitive fluorescent probe 2',7'-bis(2-carboxyethyl)-5 (and 6)-carboxyfluorescein as an alternate measure of NHE activity (Bhartur, Chang, and Rao, unpublished observations). The pH_i measurements confirmed the ²²Na⁺ uptake studies on effect of external Na⁺ and amiloride (Bhartur, Chang, and Rao, unpublished observations). However, because determining pH_i is at best an indirect measure of NHE activity, studies on regulation were performed using the more direct measure of amiloride-sensitive ²²Na⁺ uptake. Exchanger activity responded in a hyperbolic fashion to increasing extracellular [Na⁺] (Fig. 2) and was not altered by valinomycin. This indicates that the exchanger has a single external Na⁺ binding site and operates in an electroneutral fashion, with a probable stoichiometry of 1:1. This is similar to the finding in other vertebrate exchangers (8).

The low IC₅₀ values for inhibition of SL-29 Na⁺/H⁺ exchanger by amiloride and its analogs are more characteristic of NHE1 than NHE2 or 3. A strict comparison of the inhibition constant values is not possible due to the competitive nature of Na⁺ and amiloride interaction with the exchanger (8, 35). In contrast to the SL-29 exchanger, the chicken intestinal brush-border exchanger, described in an earlier paper (1, 9), is far less sensitive to inhibition by amiloride (IC₅₀ 28 µM) and its analogs (IC₅₀ DMA 6 µM; EIPA 1.3 µM). This intestinal exchanger resembles mammalian NHE3 in its amiloride sensitivity and regulation (8, 16, 19, 21, 35).

Regulation of SL-29 NHE exchanger by various agents has some distinctive features. The stimulation by serum TGF-, okadaic acid (Table 1), and hypertonicity (Fig. 7) are features shared with NHE1. The stimulation by serum and TGF- suggests the importance of growth factors in regulating the chicken exchanger and implies a role for tyrosine kinase pathways. The stimulation with okadaic acid suggests that basal phosphorylation is important in regulating exchange activity and that the SL-29 exchanger and/or its modulator protein has a phosphatase IIA-sensitive site. Interestingly, calyculin A, which inhibits phosphatase IIA and I with equal efficacy, has a marginal inhibitory effect on the SL-29 exchanger. We have previously reported differential effects of these two phosphatase inhibitors for the Na⁺-K⁺-2Cl cotransporter (29). Complex regulation of the basal phosphorylation state of NHEs involving multiple phosphorylation sites has been suggested (34). The stimulation by hypertonicity makes the SL-29 exchanger similar to NHE1 and NHE2 but distinct from mammalian NHE3, which is inhibited by cell shrinkage (15). Whether changes in the cytoskeleton and/or the G protein cascade and/or a phosphorylation-independent pathway (22) are involved in the effect of hypertonicity remains to be determined.

In contrast to the above agents, the responses to Ca²⁺, the PKC-activator PDB, and cAMP set the SL-29 exchanger apart from mammalian NHE1. In mammalian systems, regulation of NHE1 by Ca²⁺ is thought to involve calmodulin (31). Amino acids 636-656 and 657-700 of mammalian NHE1 are calmodulin binding sites. Increase in intracellular [Ca²⁺] results in binding of calmodulin to the 636-656 region and activates NHE1 by releasing the autoinhibitory effect of this region. Thus an antibody, RPI-c28, directed against amino acids 658-815 of NHE1 blocks the effect of [Ca²⁺]-dependent thrombin in activating NHE1 (34). On the basis of the variable responses of the SL-29 exchanger to Ca²⁺ (Fig. 5), it is tempting to postulate that the Ca²⁺-responsive sites are significantly altered in this molecule. This is substantiated by the finding that an antibody against amino acids 631-746 failed to recognize any protein in SL-29 cells (Fig. 8). The protonophore effects of A-23187 are unlikely to have caused an effect on the exchange process, because CCCP, another protonophore, was without effect (data not shown). In marked contrast to its variable effects on the fibroblast exchanger, Ca²⁺ is a predominant inhibitor of the chicken intestinal apical exchanger (20, 27, 28).

The SL-29 exchanger shows a biphasic response to increasing PDB concentrations (Fig. 4). Maximal stimulation is seen at 10⁸ M PDB, a concentration that correlates well with the affinity of PDB for purified PKC. This is in contrast to the chicken intestinal apical exchanger, where 10⁸ M PDB causes maximal inhibition and is monophasic (6). Although a biphasic response to PDB has been observed with other processes such as chloride secretion (7), studies in a variety of systems reveal a monophasic response of NHE1 activity to phorbol esters, making the SL-29 exchanger unique (reviewed in Refs. 22, 32). Whether the PKC and/or growth factor stimulation of SL-29 NHE involve a distal common step, such as the mitogen-activated protein kinase pathway suggested for mammalian NHE1 regulation (25), remains to be determined.

The cAMP analog cpt-cAMP, which acts via PKA, caused a prominent dose-dependent inhibition of the SL-29 exchanger (Fig. 6). In mammalian cells, cAMP has been reported either to stimulate (14) or not have an effect (16) on NHE1. The trout -NHE is stimulated by cAMP (21). In fact, -NHE, but not NHE1, has PKA consensus sites (21, 26). In contrast, cAMP consistently inhibits the renal NHE3 (33) and the chicken intestinal NHE3-like apical exchangers (8, 16, 27, 35). In the latter case, cAMP is known to act via an elevation in intracellular [Ca²⁺] (27) and in the former appears to involve accessory proteins (reviewed in Ref. 22).

Although the kinetic, pharmacological, and growth factor response data suggest that the SL-29 NHE is similar to mammalian NHE1, the immunoblotting, Northern blotting, and Ca²⁺ response data suggest that it is a distinct, albeit related, entity. The SL-29 exchanger is also distinct from the chicken intestinal brush-border exchanger in terms of amiloride sensitivity, internal proton binding site, and responses to Ca²⁺ and PKC. Furthermore, NHE3 and NHE2 probes fail to recognize a transcript in SL-29 cells. Due to the high degree of homology between NHE2 and 4 (21, 22, 32, 35), the results also suggest that SL-29 cells do not have an NHE4-related transcript. Because a full-length mammalian NHE1 cDNA probe (covering most of the membrane domain and all of the cytoplasmic domain), but not an antibody directed against amino acids 631-746 of mammalian NHE1, recognized a product in the SL-29 cells and in chicken enterocytes (1), the homology may be restricted to the amino-terminal membranous domain. Recently, Gupta et al. (11) detected a protein in avian osteoclasts using a monoclonal antibody raised against the entire COOH-terminal domain of porcine NHE1. The recent findings that our antibody can recognize reptilian forms of NHE1 but fails to recognize chicken NHE1 reinforces the notion that the chicken NHE is distinct, at least in one regulatory region, from its better characterized mammalian counterparts. Similar to our findings in the SL-29 fibroblasts, NHE1 probes reveal a 3.9-kb transcript in chicken osteoclasts (11). An NHE1 homolog is thus present in the chicken and, like mammalian NHE1, is present in multiple tissues. This NHE1-like protein probably performs housekeeping functions such as pH regulation and growth. Phosphorylation is important in SL-29 NHE regulation, but whether it is the sole regulatory mechanism remains to be determined. Our studies suggest the existence of an NHE family in aves, which, although sharing some features with the mammalian NHEs and the trout -NHE, are distinct entities.

Perspectives