The Na+-driven Cl−/HCO3 −Exchanger

The Na+-driven Cl−/HCO3 −exchanger is an important regulator of intracellular pH in various cells, but its molecular basis has not been determined. We show here the primary structure, tissue distribution, and functional characterization of Na+-drivenchloride/bicarbonate exchanger (designated NCBE ) cloned from the insulin-secreting cell line MIN6 cDNA library. The NCBE protein consists of 1088 amino acids having 74, 72, and 55% amino acid identity to the human skeletal muscle, rat smooth muscle, and human kidney sodium bicarbonate cotransporter, respectively. The protein has 10 putative membrane-spanning regions. NCBE mRNA is expressed at high levels in the brain and the mouse insulinoma cell line MIN6 and at low levels in the pituitary, testis, kidney, and ileum. Functional analyses of the NCBE protein expressed inXenopus laevis oocytes and HEK293 cells demonstrate that it transports extracellular Na+ and HCO3 − into cells in exchange for intracellular Cl− and H+, thus raising the intracellular pH. Thus, we conclude that NCBE is a Na+-driven Cl−/HC O3 −exchanger that regulates intracellular pH in native cells.

In pancreatic islet ␤-cells, glucose is physiologically the most important regulator of insulin secretion. It has been shown that glucose metabolism induces an increase in pH i in pancreatic ␤-cells (17)(18)(19)(20)(21) and that this glucose-induced rise in pH i is evoked primarily by the action of the Na ϩ -driven Cl Ϫ /HCO 3 Ϫ exchanger (16). To determine the structure and functional roles of the Na ϩ -driven Cl Ϫ /HCO 3 Ϫ exchanger, we attempted to clone the Na ϩ -driven Cl Ϫ /HCO 3 Ϫ exchanger from the mouse insulinsecreting cell line, MIN6. We describe here the primary structure, tissue distribution, and functional properties of a Na ϩdriven Cl Ϫ /HCO 3 Ϫ exchanger, designated NCBE.

EXPERIMENTAL PROCEDURES
cDNA Cloning and RNA Blot Analysis-A partial cDNA fragment of human kidney NBC cDNA (7) was amplified by polymerase chain reaction (PCR) using a human kidney cDNA as a template. The sense and antisense primers used were 5Ј-TTTGGAGAAAACCCCTGGT-3Ј (nt 2232-2250) and 5Ј-TGACATCATCCAGGAAGCTG-3Ј (nt 2912-2931). PCR was performed for 40 cycles under the following conditions: denaturation at 94°C for 15 s, annealing at 60°C for 30 s, and extension at 72°C for 45 s in a thermal cycle GeneAmp PCR system 9600 (Applied Biosystems, Foster City, CA). The 700-base pair PCR product was subjected to screening of a MIN6 cDNA library (22) as a probe under the low stringency conditions previously described (23). Positive clones were subcloned in pGEM-3Z vectors (Promega, Madison, WI) and sequenced in both directions using an ABI PRISM TM 377 DNA sequencer (Applied Biosystems).
RNA blot analysis was performed using 10 g of total RNA from the various tissues and cells. The RNAs were denatured with formaldehyde, electrophoresed on a 1% agarose gel, and transferred to a nylon membrane. The blot was probed with NCBE cDNA (nt 1-840) under the standard conditions described previously (24). The blots were washed in 0.1ϫ SSC and 0.1% SDS at room temperature for 1 h and then at 50°C for 1 h before autoradiography.
Reverse Transcription (RT)-PCR-Total RNA was prepared with TRIZOL reagent (Life Technologies, Inc.) from isolated mouse pancreatic islets. First-strand cDNA (10 ng) was generated using Superscript II reverse transcriptase (Life Technologies) with random primers. PCR was performed with the Expand high fidelity PCR system (Roche Diagnostics, Mannheim, Germany) using about 1 ng of template DNA in a 20-l reaction volume under standard conditions. The sense and antisense primers used were 5Ј-GTCATGTTAGACCAACAG-GT-3Ј (nt 4283-4302) and 5Ј-GTTGTAATAGCGACACTC-3Ј (nt 4911-4928). The product was resolved on a 1% agarose gel and confirmed by DNA * This work was supported by Grant-in-aid 10NP0201 for Creative Basic Research from the Ministry of Education, Science, Sports and Culture and by grants from the Ministry of Health and Welfare, Japan, Novo Nordisk Pharma Ltd., Yamanouchi Foundation for Research on Metabolic Disorders, and Suzuken Memorial Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains Fig. 1S  ¶ Supported by a JSPS research fellowship for young scientists. 1 The abbreviations used are: pH i , intracellular pH; NCBE, Na ϩdriven Cl Ϫ /HCO 3 Ϫ exchanger; NBC, sodium bicarbonate cotransporter; AE, anion exchanger; PCR, polymerase chain reaction; RT-PCR, reverse transcription-PCR; DIDS, 4,4Ј-diisothiocyanostilbene-2,2Ј-disulfonic acid; EIPA, 5-(N-ethyl-N-isopropyl)-amiloride; GFP, green fluorescent protein; nt, nucleotide(s).

sequencing.
Experimental Solutions-The composition of the experimental solutions (solutions A to F for oocyte experiments and solutions G to O for HEK293 cell experiments) is listed in Table I.
Functional Analysis of NCBE in Xenopus laevis Oocytes-The coding sequence of NCBE in pSD5 was linearized by digestion with FspI and in vitro transcribed with SP6 RNA polymerase as described previously (24). Defolliculated oocytes were injected with NCBE cRNA (50 nl, 0.5 g/l) or water and incubated in 1ϫ MBS medium (88 mM NaCl, 1 mM KCl, 0.8 mM MgCl 2 , 0.4 mM CaCl 2 , 0.3 mM Ca (NO 3 ) 2 , 2.4 mM NaHCO 3 , and 7.5 mM Tris, pH 7.4) for 3-5 days at 18°C before the studies. For the study of 22 Na ϩ or 36 Cl Ϫ uptake, the oocytes were preincubated for 1 h at 18°C in solution A (ml/oocyte) and then incubated in 1.4 ml of solution A, B, or C bubbled with 1.5% CO 2 or solution D or E without CO 2 containing 0.074 MBq of 22 NaCl or H 36 Cl (PerkinElmer Life Science Products). A 10-l aliquot was removed from the incubation solution for later determination of 22 Na ϩ -or 36 Cl Ϫ -specific activity. 22 Na ϩ or 36 Cl Ϫ uptake was terminated after 15 min by three washes with the respective ice-cold solutions, with the exception of solutions D and E, which were replaced with ice-cold solution F. The oocytes were then dissolved in 0.5 ml of 5% SDS, and 4.5 ml of aqueous counting scintillant (Amersham Pharmacia Biotech) was added. 22 Na ϩ uptake for 15 min in solution A was examined next in the presence of 300 M 4,4Јdiisothiocyanostilbene-2,2Ј-disulfonic acid (DIDS, Sigma), an inhibitor of anion transporters. For the study of 36 Cl Ϫ efflux, oocytes were preincubated in solution A with 36 Cl Ϫ for 30 min for loading. The oocytes were washed three times with solution A, B, C, D, or E at 18°C and then placed in 1.5 ml of each solution, respectively, at 18°C for 15 min to determine 36 Cl Ϫ efflux. The 36 Cl Ϫ activities of the oocytes and each solution were measured. The 36 Cl Ϫ efflux is presented as the percent relative to incorporated 36 Cl Ϫ . The 36 Cl Ϫ efflux in solution A for 15 min was also examined in the presence of 300 M DIDS. The 22 Na ϩ and 36 Cl Ϫ activities were measured with a beta scintillation counter (Aloka, Tokyo).
Functional Analysis of NCBE in HEK293 Cells-HEK293 cells were plated at a density of 3 ϫ 10 5 cells/3.5-cm diameter dish containing CELLocate coverslips (Eppendorf) and were cultured in Dulbecco's modified Eagle's medium (high glucose) supplemented with 10% fetal bovine serum, streptomycin (100 g/ml), and penicillin (60.5 g/ml) at 37°C under a humidified condition of 95% air and 5% CO 2 . Cells were cotransfected with 1 g/well of the full-length NCBE cDNA in the pcDNA3.1 vector (Invitrogen, Groningen, The Netherlands) and 0.05 g/well of enhanced green fluorescent protein (GFP) vector, pEGFP-N1 (CLONTECH, Palo Alto, California), for transfection marker, using LipofectAMINE, LipofectAMINE Plus, and Opti-MEM I reagents (Life Technologies, Inc.) according to the manufacturer's instructions. The GFP-expressing cells from more than 20 monolayers were studied 48 -72 h after transfection. HEK293 cells were loaded with 1 M 2Ј,7Јbis-(2-carboxyethyl)-5-(6)-carboxyfluorescein, acetoxymethyl ester (BCECF-AM, Molecular Probes, Eugene, OR) for 1 h and monitored for changes in pH i (7) by a dual-excitation wavelength method with a computerized image processor (490 nm/450 nm excitation, 520 -560 nm emission) (Argus-50, Hamamatsu Photonics, Hamamatsu, Japan) (7,21). ⌬pH i was estimated as the difference between pH i before and 10 min after switching to the test solution. The pH i calibration was generated using the KCl/nigericin technique (25). In all experiments, the cells were first acidified by NH 4 ϩ prepulse with 40 mM NH 4 Cl-containing solutions (solutions H, K, and N) for 5 min before switching to the Na ϩ -containing respective test solutions (7). To examine Na ϩ dependence of the pH i recovery (⌬pH i ) from intracellular acidification by solution H, solutions G and I were used, and to test for HCO 3 Ϫ dependence, solutions J, K, and L were used. To determine Cl Ϫ dependence, solutions M, N, and O (in this case, cells transfected with NCBE were preincubated in solution M for more than 1 h to decrease intracellular Cl Ϫ ) were used, and the results were compared with control and NCBEtransfected cells. All solutions were bubbled with 95% O 2 and 5% CO 2 , adjusted to pH 7.4. The osmolarity of each solution was adjusted with sucrose. The assays were carried out at 37°C.
Statistics-The results are expressed as means Ϯ S.E. The statistical significance between each experiment was determined by Student's t test.

RESULTS AND DISCUSSION
Cloning of the NCBE cDNA and Predicted Protein-Using a partial human kidney NBC (7) cDNA as a probe, we cloned a cDNA encoding NCBE by screening a MIN6 cDNA library. The composite nucleotide sequence of 5385 base pairs contains an open reading frame following an in-frame termination signal upstream of the ATG, which encodes a protein of 1088 amino acids with a predicted molecular mass of 122 kDa (GenBank TM / EBI/DDBJ data bank with accession no. AB033759, available on-line as Fig. 1S). A hydropathy analysis suggests that NCBE has 10 putative membrane-spanning segments (26). There are three potential N-linked glycosylation sites in the extracellular loops between the third (TM3) and fourth (TM4) membranespanning region (Asn-647, Asn-657, and Asn-667). Comparison of the amino acid sequences between NCBE and other bicarbonate transporters shows that NCBE has 74, 72, 55, 49, 38, and 34% amino acid identity to human skeletal muscle NBC Cloning and Characterization of a Na ϩ -driven Cl Ϫ /HCO 3 Ϫ Exchanger (9), rat smooth muscle NBC (8), human kidney NBC (7), Drosophila Na ϩ -driven anion exchanger (NDAE1) (30), mouse brain AE3 (31), and mouse erythrocyte AE1 (32), respectively, indicating that NCBE represents a novel bicarbonate transporter. NCBE is also homologous to several members of a bicarbonate transporter superfamily, the functional properties of which have not yet been characterized. NCBE has 76, 74, 52, and 49% amino acid identity to an NBC-related clone from human brain (SLC4A8, GenBank™ accession no. NM004858), putative human retinal NBC (33), Drosophila gene product alt 1 (GenBank TM accession no. AAF52496), and Drosophila gene product alt 2 (GenBank accession no. AAF52497), respectively. The amino acid sequence in the putative transmembrane regions is well conserved between NCBE and members of the bicarbonate transporter superfamily, whereas the intracellular amino-and carboxyl-terminal regions and a large extracellular loop between the third and fourth membrane-spanning region are rather divergent. Tissue Expression of NCBE-RNA blot analysis revealed a 5.5-kilobase NCBE mRNA expressed at high levels in brain and insulin-secreting cell line MIN6 cells and expressed at low levels in the pituitary, testis, kidney, and ileum (Fig. 1a). RT-PCR analysis shows that NCBE mRNA is also expressed in pancreatic islets (Fig. 1b).
Functional Expression of NCBE in Xenopus Oocytes and HEK293 Cells-We first examined the functional properties of NCBE using a Xenopus oocyte system. 22 Na ϩ and 36 Cl Ϫ uptake or efflux were measured 3-5 days after injection of the cRNAs or water as control. By bubbling with 1.5% CO 2 or using 16 mM butyric acid without CO 2 to acidify the oocytes (6), we examined the time course of 22 Na ϩ uptake. Because oocytes injected with water showed almost no increase, but NCBE-expressing oocytes showed a linear increase in 22 Na ϩ uptake from 5 to 45 min (data not shown), we measured the uptake at 15 min to calculate 22 Na ϩ uptake (nmol/oocyte/h). As shown in Fig. 2a, in standard solution (solution A) 22 Na ϩ uptake in NCBE-expressing oocytes was 31.4 Ϯ 2.1 nmol/oocyte/h (n ϭ 21), whereas uptake in water-injected oocytes was 1.6 Ϯ 0.3 nmol/oocyte/h (n ϭ 22), indicating that the 22 Na ϩ uptake in NCBE-expressing oocytes was 20-fold greater than in control oocytes (p Ͻ 0.05).
In contrast, the 22 Na ϩ uptake in NCBE-expressing oocytes was almost zero (n ϭ 10) in Na ϩ -free solution (solution B), but it increased in an extracellular Na ϩ concentration-dependent manner (data not shown), indicating that the activity of NCBE depends on extracellular Na ϩ . When oocytes were acidified with butyric acid rather than CO 2 (8), the 22 Na ϩ uptake in NCBE-expressing oocytes in HCO 3 Ϫ -containing solution (solution D) was 19 Ϯ 2.8 nmol/oocyte/h (n ϭ 8). Thus, acidification with butyric acid is also effective on uptake of 22 Na ϩ . To determine the effect of HCO 3 Ϫ , we used HCO 3 Ϫ -free solution (solution E) with butyric acid without CO 2 . Under this condition, 22 Na ϩ uptake was significantly decreased (3.1 Ϯ 0.5 nmol/ oocyte/h, n ϭ 8, p Ͻ 0.05), indicating that extracellular HCO 3 Ϫ is required to transport Na ϩ into cells. We then examined the effect of Cl Ϫ on 22 Na ϩ uptake. 22 Na ϩ uptake in NCBE-expressing oocytes in solution C was 18.6 Ϯ 1.6 nmol/oocyte/h (n ϭ 16), decreased to 50% of that in standard solution, indicating that extracellular Cl Ϫ accelerates NCBE activity and that there is 22 Na ϩ uptake even under extracellular Cl Ϫ -free conditions (also see on-line supplemental information, Table IIS). To determine whether extracellular Cl Ϫ accelerates NCBE activity and whether intracellular Cl Ϫ is exported by NCBE, we measured the uptake and efflux of 36 Cl Ϫ . 36 Cl Ϫ uptake was measured in NCBE-expressing oocytes or control oocytes injected with water in solution A, B, C, D, or E. Although control oocytes showed no increase in 36 Cl Ϫ uptake, NCBE-expressing oocytes showed a significant increase in 36 Cl Ϫ uptake in all solutions Ϫ -free. Thin arrows indicate decreased activity. Note that solutions A, B, and C were bubbled with 1.5% CO 2 , but solutions D and E were without CO 2 . Results were obtained from 2 to 3 independent experiments. The values represent the mean Ϯ S.E. of 7-22 oocytes from each experiment.
Cloning and Characterization of a Na ϩ -driven Cl Ϫ /HCO 3 Ϫ Exchanger except solution C (Fig. 2b). The values were 24 Ϯ 2.8 (n ϭ 9), 32 Ϯ 0.8 (n ϭ 10), 0.44 Ϯ 0.1 (12), 13 Ϯ 0.1 (n ϭ 11), and 48 Ϯ 1.2 nmol/oocyte/h (n ϭ 7) in solutions A, B, C, D, and E, respectively. The increase in 36 Cl Ϫ uptake of NCBE-expressing oocytes in these solutions indicates that NCBE transports extracellular Cl Ϫ into the cell and that the importing activity of Cl Ϫ is significantly increased in the absence of extracellular Na ϩ (solution A versus solution B, p Ͻ 0.05) or HCO 3 Ϫ (solution D versus solution E, p Ͻ 0.05). We then measured the Cl Ϫ efflux of the NCBE-expressing oocytes (Fig. 2c). The rate (%) of 36 Cl Ϫ efflux was 74.7 Ϯ 2.8 (n ϭ 8), 43.0 Ϯ 2.0 (n ϭ 9), 42.3 Ϯ 1.3 (n ϭ 17), 48.0 Ϯ 4.0 (n ϭ 8), and 17.0 Ϯ 4.0% (n ϭ 8) in solutions A, B, C, D, and E, respectively (solution A versus other solutions, p Ͻ 0.05). These results indicate that: 1) 75% of intracellular 36 Cl Ϫ is exported out of the cell in standard solution, 2) extracellular Na ϩ is necessary for exporting intracellular Cl Ϫ , 3) extracellular HCO 3 Ϫ is essential for exporting intracellular Cl Ϫ , and 4) extracellular Cl Ϫ accelerates Na ϩ uptake and Cl Ϫ efflux. Because Cl Ϫ can be transported into and out of the cell when NCBE-expressing oocytes are acidified, the transporting activity appears to be bidirectional. The functional properties of NCBE under different conditions in Xenopus oocytes are summarized in Fig. 2d. The direction of ion movement through NCBE under the physiological condition ( Fig. 2d-1) or the extracellular Cl Ϫ -free condition ( Fig. 2d-2) during the period of decrease in pH i is forward, whereas the ion movement through NCBE under the extracellular Na ϩ -free (Fig. 2d-3) or HCO 3 Ϫfree ( Fig. 2d-4) condition is reverse. Taken together, these findings suggest that NCBE exchanges extracellular Na ϩ and HCO 3 Ϫ with intracellular Cl Ϫ under the physiological condition, that is, in the presence of extracellular Na ϩ , HCO 3 Ϫ , and Cl Ϫ , when the cells are acidified. We also examined the effect of DIDS on 22 Na ϩ uptake and 36 Cl Ϫ efflux. 22 Na ϩ uptake in NCBE-expressing oocytes in standard solution was 6.0 Ϯ 0.7 nmol/oocyte/h (n ϭ 14) in the presence of 300 M DIDS, indicating that DIDS decreased 22 Na ϩ uptake to 20% of that in solution A (Fig. 2a). The 36 Cl Ϫ efflux in NCBE-expressing oocytes in solution A was 41 Ϯ 2% (n ϭ 9) in the presence of 300 M DIDS. Although DIDS decreased the 36 Cl Ϫ efflux to 55%, the value of 36 Cl Ϫ efflux was at almost the same level as in solution C (Fig. 2c).
To clarify the role of NCBE in regulating pH i , pH i changes under the various conditions were measured in HEK293 cells transiently transfected with NCBE. All experiments were carried out in acidified pH i conditions with NH 4 prepulse (with solutions H, K, or N). To determine whether the change in pH i is dependent on extracellular Na ϩ , the environment of the cells was changed from Na ϩ -free solution (solution G) to Na ϩ -containing solution (solution I). In the presence of 1 mM 5-(N-ethyl-N-isopropyl-amiloride (EIPA), a specific inhibitor of the Na ϩ /H ϩ exchanger, rapid pH i recovery (⌬pH i ) was observed only in the NCBE-transfected cells (⌬pH i was 0.239 Ϯ 0.028 in NCBE-transfected cells (n ϭ 97) and 0.003 Ϯ 0.015 in control (n ϭ 70), p Ͻ 0.05) (Fig. 3a). This pH i recovery was partially inhibited by 300 M DIDS (⌬pH i was 0.023 Ϯ 0.042 (n ϭ 89), p Ͻ 0.05) (Fig. 3a). To determine whether the change in pH i is bicarbonate-dependent, the environment of NCBE-transfected cells was changed from HCO 3 Ϫ -free and Na ϩ -free solution (solution J) to HCO 3 Ϫ -free but Na ϩ -containing solution (solution L) in the presence of 1 mM EIPA. No pH i recovery was detected (⌬pH i was 0.002 Ϯ 0.014 (n ϭ 71)) (Fig. 3b). We also examined Cl Ϫ dependence. NCBE-transfected cells were kept in Cl Ϫ -free solution (under the intracellular Cl Ϫ -depleted condition) throughout the experiments. Under this condition, the environment of the cells was changed from Na ϩ -free (solution M) to Na ϩ -containing solution (solution O). In the presence of 1 mM EIPA, pH i recovery was not detected (⌬pH i was 0.067 Ϯ 0.012 (n ϭ 95)) (Fig. 3c). These results show that pH i recovery from intracellular acidification is detected only in the presence of extracellular Na ϩ , HCO 3 Ϫ , and intracellular Cl Ϫ . Functional studies of NCBE heterologously expressed in Xenopus oocytes and HEK293 cells show that NCBE causes pH i recovery from acute intracellular acidification by transporting extracellular Na ϩ and HCO 3 Ϫ in exchange for intracellular Cl Ϫ in the presence of extracellular Na ϩ , HCO 3 Ϫ , and Cl Ϫ . NCBE is functionally distinct from the anion exchangers (3) and the Na ϩ -HCO 3 Ϫ cotransporters (9, 10), because NCBE-expressing oocytes show an increase in 22 Na ϩ uptake that is dependent on Cl Ϫ and HCO 3 Ϫ , and NCBE-expressing HEK293 cells show an increase in pH i that is dependent on extracellular Na ϩ , HCO 3 Ϫ , and Cl Ϫ . These properties are similar to those of the Na ϩdriven Cl Ϫ /HCO 3 Ϫ exchanger described in native cells (14,16,(27)(28)(29), indicating that the cloned NCBE is a Na ϩ -driven Cl Ϫ / HCO 3 Ϫ exchanger. The functional properties of NCBE are different from those of the recently identified Drosophila NDAE1, which does not require HCO 3 Ϫ for transport activity (30). Expression of NCBE mRNA in insulin-secreting cell line MIN6 and pancreatic islets implies its physiological relevance. , and Cl Ϫ dependence in pH i recovery. a, representative trace of control (nontransfected) cells and NCBE-transfected cells with or without 300 M DIDS as indicated. The environment of the cells was changed from Na ϩ -free solution (solution G) to Na ϩ -containing solution (solution I). Each trace is from a different cell; three separate traces are superimposed. b, the environment was changed from Na ϩ -free (solution J) to Na ϩ -containing solution (solution L) under HCO 3 Ϫ -free conditions. c, the environment was changed from Na ϩ -free (solution M) to Na ϩ -containing solution (solution O) under Cl Ϫ -free conditions. It has been shown that glucose-induced insulin secretion is accompanied by a rise in pH i in pancreatic islet ␤-cells (17)(18)(19)(20)(21). Although several pH i regulators have been suggested to be present in pancreatic ␤-cells (16,21), the molecular basis of these regulators is not known. NCBE is the first pH i -regulating exchanger of which the primary structure and functional properties have been determined. NCBE most likely contributes to the pH i recovery process in pancreatic ␤-cells that have been acidified by glucose metabolism. NCBE mRNA also is present in the testis, although its level is low. It has been shown that pH i regulates many sperm functions including sperm capacitation (1,2,14). Because sperm capacitation results in pH i increases that require a functional Na ϩ , Cl Ϫ , and HCO 3 Ϫ -dependent acid efflux pathway (14), NCBE could participate in the process. NCBE mRNA is also expressed in the brain at high levels. Based on physiological studies (12,27), NCBE may be present in hippocampal neurons and astrocytes, but its physiological significance in such cells is not known at present. Further investigation is necessary to clarify the structure and function relationships of NCBE and its physiological roles in various tissues.