SLC26A3, a Cl−/HCO3− exchanger, is highly expressed in intestinal epithelial cells, and its mutations cause congenital chloride diarrhea. This suggests that SLC26A3 plays a key role in NaCl absorption in the intestine. Electroneutral NaCl absorption in the intestine is mediated by functional coupling of the Na+/H+ exchanger and Cl−/HCO3− exchanger. It is proposed that the coupling of these exchangers may occur as a result of indirect linkage by changes of intracellular pH (pHi). We therefore investigated whether SLC26A3 is regulated by pHi. We generated a hemagglutinin epitope-tagged human SLC26A3 construct and expressed it in Chinese hamster ovary cells. Transport activities were measured with a fluorescent chloride-sensitive dye dihydro-6-methoxy-N-ethylquinolinium iodide (diH-MEQ). pHi was clamped at a range of values from 6.0 to 7.4. We monitored the transport activity of SLC26A3 by reverse mode of Cl−/HCO3− and Cl−/NO3− exchange. None of these exchange modes induced membrane potential changes. At constant external pH 7.4, Cl−/HCO3− exchange was steeply inhibited with pHi decrease between 7.3 and 6.8 as opposed to thermodynamic prediction. In contrast, however, Cl−/NO3− exchange was essentially insensitive to pHi within physiological ranges. We also characterized the pHi dependency of COOH-terminal truncation mutants. Removal of the entire COOH-terminal resulted in decrease of the transport activity but did not noticeably affect pHi sensitivity. These results suggest that Cl−/HCO3− exchange mode of human SLC26A3 is controlled by a pH-sensitive intracellular modifier site, which is likely in the transmembrane domain. These observations raise the possibility that SLC26A3 activity may be regulated via Na+/H+ exchanger 3 (NHE3) through the alteration of pHi under physiological conditions.
- electroneutral NaCl absorption
- downregulated in adenoma
- dihydro-6-methoxy-N-ethylquinolinium iodide
chloride ions play many physiological roles, including regulation of cell volume, fluid secretion, and acid-base balance (3, 10, 32). An efficient absorption of Cl− in the intestine is important to maintain the optimal levels of Cl− in the body. Three chloride absorptive pathways have been proposed (15, 25): 1) a paracellular pathway, which is dependent on potential difference; 2) an electroneutral pathway involving parallel functioning of Na+/H+ exchange and Cl−/HCO3− exchange; and 3) an HCO3−-dependent Cl− absorptive pathway, which is not coupled to a parallel Na+/H+ exchange. Among these chloride-absorptive mechanisms, electroneutral NaCl absorption is thought to be a predominant pathway.
NHE3 (SLC9A3) is a major Na+/H+ exchanger contributing to NaCl absorption, and its regulation was investigated extensively (21, 46). In contrast, the molecular identity of the Cl−/HCO3− exchanger involved in NaCl absorption still remains incompletely understood. At least four Cl−/HCO3− (OH−) exchangers (SLC4A1, SLC4A2, SLC26A3, and SLC26A6 ) have been found in intestinal epithelial cells (9, 35, 37, 43). Binder and colleagues (2, 33, 34) have shown that rat SLC26A3 mediates 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS)-insensitive Cl−/OH− exchange, and SLC4A1 (AE-1) mediates DIDS-sensitive Cl−/HCO3− exchange in the rat colon. Genetic analysis studies have shown that mutations in human SLC26A3 result in congenital chloride-losing diarrhea (CLD), a disease-manifested metabolic alkalosis with diarrhea containing high chloride concentrations (22). The phenotype of SLC26A3-deficient mice was similar to CLD in humans (38). Both human and mouse SLC26A3 have been shown to function as a Cl−/HCO3− exchanger in the heterologous expression systems (11, 29). These studies suggested that SLC26A3 serves an important role in intestinal Cl−/HCO3− exchange and Cl− absorption. Unfortunately, a specific inhibitor for SLC26A3 is currently not available, although it has been inhibited by niflumic acid more potently than DIDS in the heterologus expression systems (11).
The Na+/H+ exchangers and Cl−/HCO3− exchangers are coupled since they are activated/inactivated under same conditions (1, 13, 17, 31, 42). It is believed that indirect linkage through changes in intracellular pH (pHi) is involved: changes in one exchanger could generate a H+/HCO3− gradient locally that might affect the other exchanger's activity (23). This mode of coupling would have been facilited by close residency to each other since both NHE3 and SLC26A3 seem to bind to PDZ domain containing proteins (27, 36, 44). There are two possible mechanisms for the intracellular H+/HCO3− generated by one exchanger to affect the other exchangers. First, H+/HCO3− can act as a substrate by binding to the transport site. Second, H+/HCO3− might act as a regulation signal by binding to a modifier site that is independent of H+/HCO3− transport. NHE3 (and NHE1) is well known to be activated by H+ binding to a modifier site that is independent of H+ transport site. On the other hand, it has been studied less extensively whether SLC26A3 has modifier sites that are sensitive to pHi (or intracellular HCO3−) (11).
As a first step toward understanding the details of coupling between the Na+/H+ exchangers and the Cl−/HCO3− exchangers in the intestine, we conducted experiments to define the pHi sensitivity of SLC26A3 using a heterologous expression systems in the Chinese hamster ovary (CHO) cells. To measure the activity of SLC26A3, we used a chloride-sensitive fluorescent dye dihydro-6-methoxy-N-ethylquinolinium iodide (MEQ). We clamped the pHi at various levels with nigericin in media containing varying K+ concentrations. We found that Cl−/HCO3− exchange mediated by human SLC26A3 was steeply inhibited by acidic pHi. In addition, we characterized the pHi dependency of human SLC26A3 mutant in which the entire COOH terminal was deleted. Removal of the COOH terminal decreased transport activity but did not noticeably affect the pHi sensitivity.
MATERIALS AND METHODS
Materials and solutions.
Nigericin, 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF)-AM, MEQ, Alexa 546-conjugated goat anti-mouse antibody, Alexa 488-conjugated wheat germ agglutinin (WGA), Blasticidin, and Zeocin were obtained from Invitrogen (Carlsbad, CA). Mouse anti-hemagglutinin (HA) antibody was from Covance (Berkeley, CA). Fluorometric imaging plate reader (FLIPR) membrane potential kit (red) was from Molecular Devices (Sunnyvale, CA). Tunicamycin was from Sigma (St. Louis, MO).
Isotonic HEPES-buffered medium contained (in mM): 140 KCl, 1 MgCl2, 1 CaCl2, 10 glucose, and 10 HEPES (pH adjusted to 7.4 with Tris at 37°C). The HCO3−-buffered solutions contained (in mM) 115 KCl, 25 KHCO3, 1 MgCl2, 1 CaCl2, and 10 glucose, bubbled with 5% CO2-95%O2. The Cl−-free solution was prepared by equimolar substitution with gluconate or nitrate, except that 1 mM CaCl2 was replaced with 5 mM Ca-gluconate.
Establishment of cell lines harboring wild-type or mutant SLC26A3.
It has been shown that stable transfectant of SLC26A3 is hard to establish, since SLC26A3 induces growth suppression (8). To overcome this problem, we established SLC26A3-expressing cells using inducible gene expression systems (Flp-In T-Rex core kit, Invitrogen). Full-length coding SLC26A3 transcripts were amplified from human colon Marathon-Ready cDNA (Clontech, Mountain View, CA) using the Advantage HF-2 PCR kit (Clontech) with primers designed from mRNA of human SLC26A3 (NCBI accession number NM_000111). A PCR was performed with the following primers: 5′-AATGATTGAACCCTTTGGGAATCAGTA-3′ and 5′-CATAGTCAGATGAAGATCCTTCTGAATCAT-3′. The PCR conditions were as follows: 94°C for 15 s, followed by 30 cycles at 94°C for 10 s, 60°C for 30 s, and 68°C for 4 min. PCR products were cloned into the TA vector pGEM T Easy (Promega, Madison, WI) according to the manufacturer's instructions. The full-length of nucleotide sequence of SLC26A3 was confirmed. To facilitate immunological detection, we constructed a double hemagglutinin (HA)-tagged vector. An HA sequence containing linkers were cloned into HindIII-NotI sites of pcDNA5/FRT/TO (Invitrogen). The NotI fragment containing full-length of SLC26A3 was then excised from TA vector and inserted into the NotI-site of HA double tag containing pcDNA5/FRT/TO, yielding plasmid pcDNA/FRT/TO/HA-Full-SLC26A3 that was used to incorporate the full-length SLC26A3 gene into Flp-InCHO cells (Invitrogen). Flp-InCHO cells were maintained in Dulbecco's modified Eagle's medium (DMEM) /F-12 containing 10 μg/ml blasticidin and 100 μg/ml Zeocin. Stable cells harboring HA-tagged SLC26A3 protein were generated by cotransfection of pcDNA/FRT/TO/ HA-Full-SLC26A3 and the pOG44 plasmid (Invitrogen) into Flp-InCHO cells by lipofectamine 2000 (Invitrogen) and were selected by limiting dilution in the presence of hygromycin B (200 μg/ml) and blasticidin (10 μg/ml). Expression of HA-tagged SLC26A3 was followed by immunofluorescence. Protein expression in stable cells was induced by adding tetracycline or doxycycline to culture media at a final concentration of 1 μg/ml for 24 h at 37°C before experiments.
For transient transfection, NotI fragments containing full-length of SLC26A3 were excised and inserted into mammalian expression vector d-HA-pRc/CMV (Riken, Japan), which contains a double (HA) epitope tag, located at the NH2 terminus of the protein of interest. This resultant plasmid D-HA-pRc/CMV-Full-SLC26A3 was used to generate COOH-terminal deletion mutants of SLC26A3. The COOH-terminal deletion mutants were truncated at positions 524, 643, and 705 (constructs Δ524, Δ643, and Δ705, respectively) using KOD-Plus-Mutagenesis kit (TOYOBO).
Measurement of intracellular chloride.
Intracellular Cl− concentration ([Cl−]i) was determined by the Cl−-sensitive fluoroprobe dihydro-MEQ (diH-MEQ). DiH-MEQ, the membrane-permeant form of MEQ, was synthesized from MEQ according to the manufacture's instructions. The synthesized diH-MEQ was dissolved at a concentration of 300 mM in dimethyl sulfoxide and stored at −80°C for up to 1 mo. The cells were seeded onto 25-mm round glass coverslips and mounted in a custom-made perfusion chamber that allowed continuous superfusion of the cells. The cells were then loaded with diH-MEQ by incubation with 300 μM diH-MEQ for 10 min at 37°C in a Ca2+- and Mg2+-supplemented phosphate-buffered saline (PBS). Microfluorometry was performed as previously described (19) with modifications. Briefly, the chamber was placed on the stage of an inverted microscope (TE200-U, Nikon, Tokyo, Japan) equipped with a microscopic dual-wavelength fluorometer system (CAM-230, Japan Spectroscopic, Tokyo, Japan), perfused at 6 ml/min with the gas-equilibrated solution and maintained at 35–37°C. The temperature of the solutions was maintained at 37°C using a water-jacketed stainless tubing inlet. Solutions were delivered by gravity to the chamber through CO2-less impermeable tubing. Clusters of the cells were excited at 350 nm for 50 ms every 2 s, and the fluorescence was measured at 400 nm through a bandpass filter. All these procedures were controlled by a computer (Macintosh LC), which was equipped with a data acquisition and analysis system (Lab View 2, National Instruments, Houston, TX). In situ calibration was performed at the end of each experiment by measuring the fluorescence intensity F0 obtained when the cells were exposed to the Cl−-free solution containing 10 μM amphoterin B. At the end of the measurement the cells were exposed to 140 mM KSCN in the absence of Cl− to obtain the background fluorescence. Changes in F0/F in each experiment were then converted to changes in [Cl−]i using the Stern-Volmer constant obtained from the pooled data.
Measurement of pHi and membrane potential.
Measurement of the fluorescence of intracellular BCECF or FLIPR red dye was performed essentially as described for the measurements using diH-MEQ. To measure cytosolic pH, the cells were loaded with 0.5 μM BCECF-AM for 10 min at 37°C in Ca2+- and Mg2+-supplemented PBS. For membrane potential measurement, the cells were loaded with FLIPR red dye for 30 min at 37°C according to the manufacturer's instructions. Fluorescence intensity was measured at emission wavelength 520 nm and excitation at 440 and 490 nm for BCECF or at excitation wavelength 530 nm and emission 565 nm for the FLIPR red dye.
Manipulation of pHi.
To manipulate pHi, the cells were treated with K+/H+-exchanging ionophore nigericin (10 μM) and varying extracellular concentrations of K+, as previously described (18) with modifications. Since at equilibrium the ratios of intracellular and extracellular K+ and H+ concentrations are equal ([K+]i/[K+]o = [H+]i/[H+]o), the desired pHi can be calculated from the imposed [K+] gradient and pHo, assuming a [K+]i of 140 mM. Based on these considerations, the solutions (in mM) used contained the following: for pHi 7.4, 140 K+; for pHi 7.0, 56 K+; for pHi 6.8, 35 K+; for pHi 6.4, 14 K+; and for pHi 6.0, 5.6 K+. N-methyl-d-glucamine was used to balance the osmolarity.
The cells were plated onto glass coverslips, rinsed with PBS, and fixed using 4% paraformaldehyde (PFA) in PBS for 20 min. After fixation, the cells were incubated with 100 mM glycine in PBS for 20 min. The cells were then incubated with Alexa 488-conjugated WGA (100 μg/ml) for 20 min before permeabilization. The cells were next preblocked with 5% skimmed milk and permeabilized in 0.1% Triton in PBS for 30 min and then incubated with anti-HA monoclonal antibody (1:1,000 dilution) for 1 h. After being washed three times to remove unbound antibody, the cells were incubated with Alexa 546-conjugated donkey anti-mouse antibody (1:1,000 dilution) for 1 h. After final washes, the samples were mounted onto glass slides using Dako medium (Dako, Carpinteria, CA). Images were acquired using a laser-confocal microscope (Zeiss LSM510).
The adherent cells were washed with PBS and scraped off with a rubber policeman into 1× Laemmli sample buffer, and samples were subjected to SDS-PAGE and transferred onto nitrocellulose filters (Hybond, Amersham Bioscience). Blots were blocked with 5% skimmed milk in PBS containing 0.1% Triton X-100 and exposed to anti-HA monoclonal antibody (1:5,000 dilution), followed by horseradish peroxidase-conjugated secondary antibody (1:5,000 dilution, Pierce, Rockford, IL). Immunoreactive proteins were visualized by enhanced chemiluminescence (Super Signal, Pierce) and exposed to Kodak film.
Experimental values are given as the means ± SE of the indicated number of the determinations. Comparisons between two groups were made by either unpaired or paired Students's t-test, as appropriate.
Expression and functional characterization of HA-SLC26A3.
We first studied the expression of HA-SLC26A3 by immnoblotting. As shown on Fig. 1A, two immunoreactive bands were detected in the HA-SLC26A3-expressing CHO cells using anti-HA antibody: a lower band (85 kDa) and upper band (∼120 kDa). The upper band migrated as a wide band, likely attributable to glycosylation. This was confirmed by treatment of the cells with an N-glycosylation inhibitor tunicamycin. Treatment with tunicamycin resulted in the virtual disappearance of the two bands and the appearance of a single 78-kDa band, in agreement with a previous study (14). This indicates that similar to the wild-type parental protein, HA-SLC26A3 is modified posttranslationally by N-linked glycosylation (7). This also verifies that the expressed protein undergoes normal maturation in CHO cells. We next determined the subcellular distribution of HA-SLC26A3 with confocal microscopy (Fig. 1, B–D). To confirm the cell surface expression, the plasma membrane was stained with WGA before permeabilization of the cells. As shown in Fig. 1D, the epitope-tagged protein was colocalized with WGA, suggesting that it is expressed on the cell surface. Transverse reconstructions confirmed the surface staining (Fig. 1, B'–D'). Some fraction of HA-SLC26A3 appeared to be intracellular, possibly reflecting the immature protein that can be detected by SDS-PAGE as a lower molecular weight band (85 kDa).
We next verified whether the protein tagged with HA-epitope retained its functional properties by fluorometry in diH-MEQ-loaded cells. Figure 2A shows a typical calibration of diH-MEQ by exposing the cells to the amphotericin B and by varying the perfusate Cl− concentration. To calibrate [Cl−]i, we used amphoterin B to vary [Cl−]i. Since these antibiotics form Cl−-permeable pores, it is a convenient method to use for altering [Cl−]i in the cells (4, 16). Figure 2B is a Stern-Volmer plot of the pooled data. The Stern-Volmer constant was 82.3 M−1. We first assessed the effect of expression of SLC26A3 on resting pHi and [Cl−]i (Table 1). Expression of HA-SLC26A3 did not significantly change the resting level of [Cl−]i compared with the wild-type CHO cells (P = 0.19). These values are within the range of 29–47 mM, which was measured by other methods in the CHO cells (6, 40), demonstrating the validity of our measurements. Upon expression of HA-SLC26A3, basal resting pHi was slightly but not significantly decreased (P = 0.25) (Table 1). Antiport activity was assessed by the removal of extracellular Cl− from the perfusate. Since a small DIDS-insensitive Cl−/HCO3− exchange activity was occasionally seen before induction of the exchanger (data not shown), probably due to contamination of doxycycline or leakiness of the inducible expression systems, we used the wild-type CHO cells as a control for functional experiments. In wild-type CHO cells, upon removal of extracellular Cl− from the perfusate only a small reduction of [Cl−]i was observed (1.6 ± 1.7 mM/min, n = 6, Fig. 2C). As shown in Fig. 2D, HA-SLC26A3-expressing cells responded to removal of Cl− with a vigorous reduction of Cl− (61.7 ± 11.7 mM/min, final [Cl]i = 19.8 ± 1.4 mM, n = 10). The initial rate of [Cl−]i decrease was almost completely inhibited by 50 μM niflumic acid (by 89%, n = 3, Fig. 2, D and F). In contrast, H2-DIDS did not significantly inhibit the rate of [Cl−]i decrease induced by Cl− removal, the rate being 41.8 ± 11.6 and 40.4 ± 6.7 mM/min in the cells incubated with or without 250 μM H2-DIDS, respectively (Fig. 2, E and F). HA-SLC26A3-expressing cells responded to removal of Cl− with a large intracellular alkalinization (ΔpH = 0.69 ± 0.05, n = 5, Fig. 2G). Furthermore, pHi increase induced by removal of extracellular Cl− was virtually not observed in nominally HCO3−-free HEPES-buffered solution (0.11 ± 0.03, n = 3, Fig. 2H). Jointly, these observations indicated that HA-tagged SLC26A3 retained Cl−/HCO3− exchanger in agreements with previous studies (11).
Effect of Cl−/HCO3− exchange activity on membrane potential.
Shcheynikov et al. (39) found earlier that mouse SLC26A3 is an electrogenic transporter with a 2Cl−/1HCO3− exchange stoichiometory in the oocytes. In contrast, human SLC26A3 has not been shown to be electrogenic in the oocytes and HEK293 cells (11, 26). We therefore assessed the electrogenicity of SLC26A3 using a membrane potential-sensitive dye. Since fluorescence response is larger at depolarized membrane potentials with this dye, we performed experiments in media that were containing 100 mM K+. As illustrated in Fig. 3B (left of the trace), the cells expressing SLC26A3 depolarized membrane potential upon removal of extracellular Cl− in the HCO3−-buffered solutions. However, the magnitude of this depolarization did not differ significantly from that in the wild-type CHO cells (Fig. 3, A and C). These membrane potential changes could be due to endogenous Cl− conductive pathway(s) in the wild-type CHO cells. The FLIPR membrane potential assay kit contains not only an oxonol-type voltage-sensitive dye but also a fluorescent quencher (see review, 30). It is conceivable that Cl−/HCO3− exchange activity may be inhibited by the quencher. To rule out a potential artifact, we measured the rate of Cl−/HCO3− exchange with BCECF after the cells were loaded with FLIPR kit. The magnitude of alkalinization induced by the Cl−-free solution was virtually not changed after FLIPR dye loading (ΔpHi = 0.7, n = 2) (not shown). These observations suggest that Cl−/HCO3− exchange mode of SLC26A3 is electroneutral. Furthermore, the rates of Cl−/HCO3− exchange activity, as measured by diH-MEQ, were not affected by changing the perfusate K+ concentration (Fig. 3D).
Cl−/HCO3− exchange mode of SLC26A3 is sensitive to pHi.
Having established that SLC26A3 expressed in the CHO cells is fully functional, we next wanted to investigate the sensitivity of the exchanger to pHi. Therefore, we monitored the rate of [Cl−]i decrease, reflecting chloride exchange for HCO3−, at varying pHi in the CO2/HCO3−-buffered solution. We first established the conditions for clamping pHi. All measurements were performed in media that were Na+ free, to minimize the contribution of Na+-dependent pHi regulators. By adding an electroneutral K+/H+-exchanging ionophore nigericin at sufficiently high concentrations and by setting the transmembrane K+ gradient and extracellular pH to predetermined levels, pHi can be clamped to any desired value. As shown in Fig. 4, upon changing the extracellular K+ concentration from 145 to 35 mM, pHi was changed from 7.2 to 6.8. This result suggested that nigericin is present in high enough concentrations for pHi clamping in the CO2/HCO3−-buffered solution. However, it took >5 min to equilibrate pH by setting the transmembrane K+ gradient alone in the CO2/HCO3−-buffered solution. Therefore, we clamped pHi by using HEPES- or 4-morpholino-ethanesulfonic acid (MES)-buffered-solutions before each measurement (Fig. 5A). Under this condition, pHi reached equilibrium within 2 min. Switching the perfusate from a pHi-clamping solution to the experimental one containing CO2/HCO3− caused only a small shift of pHi (Fig. 5A). When the Cl−/HCO3− exchange was initiated by removal of extracellular Cl−, pHi was increased and reached its nadir, suggesting that the Cl−/HCO3− exchange activity is greater than the nigericin exchanging activity. Recovery of pHi was observed upon readdition of extracellular Cl− returned to the original level (Fig. 5B, open squares), indicating that pHi can be clamped stably at the desired pHi level. As summarized in Fig. 5B, pHi was clamped nearly at the desired levels at acid pHi in both the wild-type (open squares) and the SLC26A3-expressing CHO cells (closed circles). We next monitored the Cl−/HCO3− exchange activity by using Cl−-sensitive dye under the same experimental condition (Fig. 5C). Three measurements were performed consecutively on the same cells. The mean value of the initial rate of [Cl−]i decreases from the first and third measurements was taken as the control value, which were clamped at pHi 7.4 and compared with the rate in the acidified cells. When the cells were clamped at clamping pH of 6.4, the resting [Cl−]i was slightly but significantly increased compare with the levels at clamping pH of 7.4 (31 ± 1 and 35 ± 1 mM, n = 7, P = 0.02). In addition, the initial rate of [Cl−]i decrease upon removal of extracellular Cl− was inhibited by 50% (Fig. 5C). The open squares in Fig. 8A summarizes the results of experiments performed with clamping pHi at various levels, the initial rate being steeply inhibited with pHi decreased between 6.8 and 7.3. It should be noted here that under the present experimental condition acid-loaded cells would be expected to show a greater rate of Cl−/HCO3− exchange activity due to the increased inward-directed HCO3− gradient. Nevertheless, we found that the Cl−/HCO3− exchange activity was decreased in the acid-loaded cells.
Cl−/NO3− exchange mode of SLC26A3.
In the preceding experiments, we could not clamp the pHi alone, because intracellular HCO3− was also changed inevitably under the conditions used. To overcome this problem, we took advantage of the ability of SLC26A3 to carry NO3−. Therefore, we next monitored the changes of [Cl−]i as chloride exchanges for nitrate at varying pHi in nominally HCO3−-free HEPES-buffered solution. We first verified that NO3− was transported via HA-SLC26A3 in diH-MEQ-loaded cells. Fluorescence of MEQ is quenched in the presence of Cl− but not NO3− (5). As shown in Fig. 6A, when perfusate Cl− was totally replaced with NO3− in wild-type CHO cells, a small [Cl−]i decrease was observed and this was totally abolished by H2-DIDS (4.4 ± 0.7 vs. 0.4 ± 0.9 mM/min in the absence and presence of 125 μM H2-DIDS, respectively, n = 3), suggesting that this Cl−/NO3− exchange is via endogenous H2-DIDS-sensitive anion exchange. In contrast, in SLC26A3-expressing cells, replacement of extracellular Cl− with NO3− led to a rapid and reversible decrease of [Cl−]i (39.8 ± 6.3 mM/min, final [Cl]i = 24.3 ± 1.8 mM, n = 9). This [Cl−]i decrease was attenuated by nifulumic acid (Fig. 6, C and D) but insensitive to H2-DIDS (Fig. 6, B and D). These results suggested that SLC26A3 can mediate Cl−/NO3− exchange. It was reported that mouse SLC26A3 can also mediate Cl−/OH− exchange (24). Therefore to evaluate a possible Cl−/OH− exchange, we measured pHi using pH-sensitive dye BCECF using the same conditions as for the [Cl−]i measurements (Fig. 6, E and F). When the perfusate Cl− was totally replaced with NO3−, a small decrease of pHi was observed in the wild-type CHO cells (0.22 ± 0.02, n = 3), this magnitude of acidification was not significantly different from that in the SLC26A3-expressing cells (0.23 ± 0.06, n = 3, P = 0.91), suggesting that Cl−/OH− exchange mode is not operating. Although we did not study further this NO3−-induced acidification mechanism, it might be due to proton nitrate cotransporter (12). It was reported previously that mouse SLC26A3 operates as uncoupled anion transport when NO3− was used as substrates (39). We therefore measured membrane potential when Cl− was replaced with NO3− in the HEPES-buffered solution. As illustrated in Fig. 3B, the cells expressing SLC26A3 were hyperpolarized when Cl− was removed from the perfusate (0.15 ± 0.03 arbitrary unit, n = 3). However, this magnitude of hyperpolarization did not differ significantly from that in the wild-type CHO cells (0.11 ± 0.01 arbitrary unit, n = 3, P = 0.26, Fig. 3, A and C). These results excluded that Cl− exit in the SLC26A3-expressing cells occurred via conductive pathway(s). These membrane potential changes could be due to the entry of NO3− via endogenous conductive pathway(s) in the wild-type CHO cells. Furthermore, varying transmembrane K+ ratios did not affect the rate of intracellular Cl− decrease (data not shown). Together, these results suggest that SLC26A3 can transport NO3− in exchange for Cl− exit via an electroneutral process.
We next examined pHi dependency of SLC26A3 in the Cl−/NO3− exchange mode in nominal HCO3−-free HEPES-buffered solution. In this buffer solution, pHi was clamped at a desired level in SLC26A3-expressing cells (Fig. 7A). As shown in Fig. 7B, the rate of [Cl−]i decrease was not inhibited when the cells were acidified at clamping pH of 6.4. The open squares in Fig. 8B summarizes the results of experiments at various clamping pH. The rate of Cl−/NO3− exchange activity was not significantly inhibited at pHi values between 7.4 and 6.4. When the cells were challenged by a clamping pH of 6 (observed pHi, 6.2), the initial rate was inhibited only by 25% (P = 0.01, n = 5). In contrast to Cl−/HCO3− exchange activity, there was no apparent inhibition by acidic pHi for Cl−/NO3− exchange activity. This could be due to the absence of extracellular/intracellular CO2/HCO3−-buffered condition. However, the rates of Cl− exit when CO2/HCO3−-buffered solution containing 25 mM Cl− was switched to the one containing no Cl− and 25 mM NO3− did not differ between control and acid-loaded cells (P = 0.35, n = 3, see Fig. 8B, closed square).
COOH terminus of SLC26A3 does not contribute to pH regulation.
It has been shown that removal of the entire COOH terminal cytoplasmic domain of SLC26A3 abolished the sensitivity of pHi in the oocytes (11). It is thought that histidine residue is a candidate for a pH sensor, since their imidazole side chain have a pKa value of near 7. Comparison of amino acid sequences of the COOH terminal of SLC26A3 reveals that three histidine residues (H647, H714, and H719) are highly conserved among several species. To get insights into the pHi sensor of SLC26A3, we made various COOH terminally truncated mutants. Immunofluorescence confocal microscopy was used to define the subcellular distribution of the truncated mutants. As shown in Figs. 9, B and B ′, Δ524: the mutant lacking the entire COOH-terminal cytoplasmic domain appeared to be as punctate spots at the plasma membrane and intracellular compartments. The Δ524 mutant was detected by immunobloting as two distinct bands (53, ∼90 kDa) (Fig. 9J). This suggests that the upper band was fully glycosylated. To confirm this, the cells were treated with tunicamycin, resulting in the disappearance of the two bands and their converge to a single 49-kDa band. In contrast, as shown in Figs. 9, E and H, Δ643 and Δ709 were diffusely distributed throughout the cytoplasm and did not localize to the cell surface membrane. Immunobloting revealed that these mutants migrate as only one band, observed at the predicted molecular size of 71 and 81 kDa, respectively. This suggests that proper trafficking of these mutants to the plasma membrane is prevented.
We next established cells harboring truncated mutants of SLC26A3 using inducible gene expression systems. Using these cells we measured the Cl−/HCO3− exchange activity by diH-MEQ. As shown in Fig. 9K, the removal of the entire COOH-terminal domain resulted in a decrease in the Cl−/HCO3− exchange activity in the Δ524 mutant (12 ± 3 mM/min, n = 4) compared with wild-type SLC26A3. However, the pHi sensitivity was still observed when the pHi was clamped at clamping pH of 6.4 (3 ± 1 mM/min, n = 4 Fig. 8A, closed circle). In contrast, we could not detect any Cl−/HCO3− exchange activity in the Δ643 and Δ709 mutants, in agreement with their inability to localize to the surface as shown by the immunofluorescence images described above.
The mechanism of coupling between the Na+/H+ exchangers and the Cl−/HCO3− exchangers in the intestine remains only partially understood. It is thought that change of exchanger activities by pHi (intracellular [H+]) through affecting modifier site is one possible mechanism. We therefore developed a method to change pHi in the presence of CO2/HCO3−. We conducted experiments to define the pHi sensitivity of human SLC26A3 using heterologous expression in CHO cells.
Our results suggest that Cl−/HCO3− exchange mode of SLC26A3 is controlled by a pH-sensitive intracellular modifier site. Inhibition of Cl−/HCO3− exchange activity by intracellular acidification does not result from general toxicity of acidic pHi because this inhibition was reversible. In addition, we could observe robust Cl−/HCO3− exchange activity after acidification (Fig. 5). We assessed the effect of acidification on Cl−/HCO3− exchange activity in Cl− outward and HCO3− inward mode. In this mode, acid-loaded cells would be expected to show a greater rate of Cl− efflux as a result of the larger inwardly directed HCO3− gradient. Therefore, inhibition of Cl− exit/ HCO3− entry by acid pHi indicates that SLC26A3 possesses a H+ modifier site. Membrane potential changes could regulate SLC26A3 exchange activity, since SLC26A3 was shown to be a Cl−/HCO3− exchanger with a 2:1 stoichiometry (39) and varying transmembrane [K+] ratios altered pHi. In contrast, our results do not support the electrogenic nature of SLC26A3. First, although large [Cl−]i changes (>20 mM) occurred following removal and readdition of extracellular Cl− in SLC26A3 expressing cells, there was no discernable membrane potential change compared with that in the wild-type CHO cells (Fig. 3). Second, extracellular K+ concentrations did not affect Cl−/HCO3− or Cl−/NO3− exchange activity.
Our immunofluorescence data show that SLC26A3 was present both on the surface membrane and intracellular compartments (Fig. 1C). This implies that SLC26A3 activity may be regulated by altering the number of available molecules at the plasma membrane. However, this explanation was rendered unlikely by the observation that Cl−/NO3− exchange activity was not changed in acid-loaded cells (Fig. 7), suggesting that the number of transporters on the surface is unchanged.
Other investigators have demonstrated that removal of the entire COOH-terminal cytoplasmic domain of SLC26A3 abolished its sensitivity to pHi in oocytes (11), suggesting that a pH sensor might reside within the cytoplasmic domain. However, our results are in contrast to those findings, suggesting that pH sensor of SLC26A3 resides in the transmembrane domain. This discrepancy may be explained by the use of different expression system and different exchange mode of SLC26A3.
Interestingly, intracellular Cl− decrease under acid pHi was stopped approximately halfway to two-thirds of the way toward the values in more neutral condition, although pHi was already increased to ∼7.4 during Cl−/HCO3− exchange (compare the middle part of the traces with the left and right parts in Fig. 5, A and C). These results imply that the release from the inhibition of Cl−/HCO3− exchange by pHi is slow. These slow responses to pHi changes for activity of SLC26A3 are reminiscent of that of pH-dependent activation for NHE3 (20).
Our results failed to demonstrate pHi sensitivity of SLC26A3 activity when anion exchange activity was monitored by Cl−/NO3− exchange mode (Figs. 7 and 8). However, we could observe the modest inhibition when the cells were clamped at pHi 6.2 (Fig. 8). It is conceivable, therefore, that the pHi-sensitivity curve is shifted to the acidic side when anion exchange activity was monitored by Cl−/NO3− exchange mode. We measured Cl−/NO3− exchange activity where pHi was clamped at 5. Both extracellular and intracellular pH were simultaneously changed. However, this condition was toxic, and Cl−/NO3− exchange activity was totally diminished after acidification (data not shown). Another possibility is that the cells are exposed to large gradients of transported ions (extracellular NO3− = 144 mM) in these experiments, and this may mask pHi sensitivity of SLC26A3. To exclude this, we reduced extracellular NO3− concentration to 25 mM. Even under this condition, Cl−/NO3− exchange activity was not inhibited when pHi was clamped at 6.4 (data not shown). Since CHO wild-type cells have endogenous Cl−/NO3− exchange activity (Fig. 6A), we analyzed pHi dependency of this anion exchange. The endogenous Cl−/NO3− exchange activity was completely inhibited at pHi 6.4 (data not shown). Jointly, we conclude that the Cl−/NO3− exchange mode of SLC26A3 was essentially insensitive to physiological pHi ranges. It is noteworthy that the activity of AE3 was stimulated at alkaline pHi as measured in Cl−/HCO3− exchange mode (28) but not as measured in Cl−/NO3− exchange mode (41).
Our data indicated that the pHi dependence of SLC26A3 is steepest in the range of pH between 6.8 and 7.3, which is well within the physiological pHi range in the cells. Actually, this value is in good agreement with the values for the resting pHi in the native enterocytes (19). What are the physiological significances of this pHi dependency of SLC26A3? One of the functions of SLC26A3 would be to stabilize pHi of epithelial cells. The increase in pHi would accelerate SLC26A3 to extrude HCO3− by affecting the modifier site, a property suggested in the present study. In addition, the increase in intracellular HCO3− would energetically stimulate the activity of the exchanger. On the other hand, when pHi decreases, SLC26A3 would possibly mediate HCO3− uptake that would in turn decrease pHi. Another and probably most important function of SLC26A3 is to mediate Cl− absorption in the intestinal epithelia. Intestinal Cl− absorption is, in most cases, coupled to Na+ absorption that is mediated by the apical Na+/H+ exchanger. The Na+/H+ exchanger and SLC26A3 transports H+ or HCO3−, respectively, are both derived from hydration of CO2. Importantly, NHE3, which is suggested to be a major apica1 Na+/H+ exchanger, was demonstrated to be activated by intracellular H+ via a modifier site, with the half-maximal value of 0.11–0.13 μM (pH = 6.89–6.96) for H+. These properties are likely to promote coupling between apica1 NHE3 and SLC26A3 since the increase in the apical SLC26A3 activity (Cl− uptake) would cause pHi decrease, which then increases apical NHE3 activity (Na+ uptake). By the same mechanism, the decrease of SLC26A3 activity would induce a reduction of NHE3 activity. The present finding that SLC26A3 activity is regulated by pHi through modifier sites would suggest the existence of a similar but reciprocal coupling mechanism. The increase or decrease of apical NHE3 activity would increase or decrease pHi, which then activates or inhibits SLC26A3 and promotes the coupled Cl− and Na+ absorption. The existence of a coupling mechanism is also suggested by the well-known finding that increases in intracellular cAMP in the intestinal epithelium inhibit both Na+ and Cl− absorption simultaneously (1, 31). However, SLC26A3 does not possess consensus site for PKA. In accordance with previous findings (11) treatment of the cells with 8-bromo-cAMP did not affect the Cl−/HCO3− exchange activity of SLC26A3 in the present expression system (unpublished observation). It is well known that NHE3 activity is acutely inhibited by activation of cAMP-dependent protein kinase (45). Therefore, upon an increase in intracellular cAMP in the enterocytes, intracellular acidification resulting from inhibition of NHE3 could inhibit the Cl−/HCO3− exchange activity via the pHi-sensitive modifier sites. Whereas regulation through pHi could serve as a potent coupling mechanism, our observations do not rule out the possibility that an additional cofactor (e.g., sodium/hydrogen exchanger regulatory factor) is required in cAMP-mediated inhibition of SLC26A3 activity.
We have examined the subcellular distribution of various SLC26A3 mutant proteins. Our findings show that although the Δ524-truncated mutant, which lacks the entire COOH-terminal part, was expressed at the plasma membrane, its level at the plasma membrane was lower than that of the wild-type SLC26A3. In addition, Δ643 and Δ709 mutants showed the lack of maturation and proper trafficking to the plasma membrane (Fig. 9). These results suggest that the COOH-terminal cytoplasmic domain is needed for proper expression at the plasma membrane. This interpretation is in accordance with previous findings showing that disruption of the COOH-terminal STAS (sulfate transporters and anti-sigma-factor) domain affects steps that are involved in the folding and/or trafficking pathway (14).
In summary, we have described the pHi dependency of human SLC26A3 activity, which was regulated within physiological pH ranges. Our findings contribute to the understanding of functional coupling between the Na+/H+ exchangers and the Cl−/HCO3− exchangers in the intestine. These observations raise the possibility that, under physiological circumstances, SLC26A3 activity may be regulated by NHE3 via the changes in pHi.
This study was supported in part by grants from the Salt Science Research Foundation, No. 0340, and the Houansha Foundation (to H. Hayashi).
The authors thank Drs. Y. Suzuki (University of Shizuoka) and K. Szaszi (St. Michael's Hospital, Toronto, Canada) for helpful comments and discussions.
Present address of K. Suruga: Division of Nutritional Science, University of Nagasaki, Siebold 1-1-1 Manabino, Nagayo-cho Nishisonogi-gun, Nagasaki-ken, Japan.
- Copyright © 2009 the American Physiological Society