Cloning and Functional Expression of a Human Kidney Na+:HCO3 −Cotransporter*

Several modes of HCO3 −transport occur in the kidney, including Na+-independent Cl/HCO3 − exchange (mediated by the AE family of Cl−/HCO3 − exchangers), sodium-dependent Cl−/HCO3 −exchange, and Na+:HCO3 −cotransport. The functional similarities between the Na+-coupled HCO3 − transporters and the AE isoforms (i.e. transport of HCO3 − and sensitivity to inhibition by 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid) suggested a strategy for cloning the other transporters based on structural similarity with the AE family. An expressed sequence tag encoding part of a protein that is related to the known anion exchangers was identified in the GenBank™ expressed sequence tag data base and used to design an oligonucleotide probe. This probe was used to screen a human kidney cDNA library. Several clones were identified, isolated, and sequenced. Two overlapping cDNA clones were spliced together to form a 7.6-kilobase cDNA that contained the entire coding region of a novel protein. Based on the deduced amino acid sequence, the cDNA encodes a protein with aM r of 116,040. The protein has 29% identity with human brain AE3. Northern blot analysis reveals that the 7.6-kilobase mRNA is highly expressed in kidney and pancreas, with detectable levels in brain. Functional studies in transiently transfected HEK-293 cells demonstrate that the cloned transporter mediates Na+:HCO3 − cotransport.


EXPERIMENTAL PROCEDURES
Cloning and Sequencing-The GenBank™ nonredundant EST data base was queried against the three known anion exchangers AE1, AE2, and AE3. A sequence with a score of 172, from a human pancreatic islet cell line (GenBank™ accession number W39298) was a close, but not an identical, match to rat AE3.
A sense-stranded oligonucleotide, 5Ј-AGG GAG CAA AGA GTC ACT GGA ACC-3Ј, from W39298 was synthesized, biotinylated and used in the GeneTrapper™ cDNA positive selection system (Life Technologies, Inc.) to screen a SuperScript™ human (38-year-old Caucasian male) kidney cDNA library (Life Technologies, Inc.) directionally cloned in pCMV⅐SPORT1. The GeneTrapper™ system uses streptavidin-linked magnetic beads to enrich for sequences complementary to the biotinylated oligonucleotide. After plating, the library was screened by hybridization with 32 P-end-labeled oligonucleotide, and 21 positive clones were selected.
The three largest unique clones were chosen for sequencing. Two of the three clones (one of approximately 5 kb and another of approximately 1.6 kb) contained the EST sequence (W39298) and also contained sequences of perfect homology to each other. An open reading frame analysis suggested that the entire coding region was not contained between the two clones. Therefore, a second oligonucleotide was synthesized (which was 5Ј to the EST sequence) for a second round of GeneTrapper™ cDNA selection. The sequence of the second oligonucleotide was: 5Ј-CAA GCC AAC AAG TCC AAA CCG AGG-3Ј. A polymerase chain reaction analysis was performed using the T7 and SP6 primers to reveal the size of the cDNA inserts of 48 randomly chosen clones. The largest (with a 3.5-kb insert) overlapped the 5-kb clone by 828 base pairs, included the EST clone W39298, and contained the remainder of the open reading frame (as well as 149 base pairs of the 5Ј-noncoding region). Two Sse8387I restriction sites, one in the region of overlap and another in the polylinker, allowed the construction of the full-length Na ϩ :HCO 3 Ϫ cotransporter clone. Transient Transfection with the Cloned cDNA-HEK-293 cells, grown for 24 h on fibronectin-coated glass coverslips (in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum) were transfected with 8 g of the full-length cDNA construct (in the cloning/expression vector pCMV⅐SPORT1) by calcium phosphate-DNA coprecipitation (13), and cells were studied 44 -52 h after transfection.
Intracellular pH Measurement-Changes in intracellular pH (pH i ) were monitored in cells using BCECF (14,15). HEK-293 cells were * This work was supported by National Institutes of Health Grants DK 46789 (to M. S.) and DK 50594 (to G. E. S.). 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF007216.
ʈ To whom correspondence and Reprint requests should be addressed: Division of Nephrology & Hypertension, University of Cincinnati College of Medicine, 231 Bethesda Ave., Cincinnati, OH 45267-0585. Tel.: 513-558-5471; Fax: 513-558-4309. loaded with 5 M BCECF and monitored for changes in pH i using a Delta Scan dual excitation spectrofluorometer (14,15). The pH i calibration was generated using the KCl/nigericin technique. To test for the presence of Na ϩ :HCO 3 Ϫ cotransport, cells were acid loaded by NH 4 ϩprepulse (16) using solution B and subsequently, solution A ( Table I). The initial rate of pH i recovery was monitored in the presence of 1 mM amiloride in a solution containing both HCO 3 Ϫ and Na ϩ (solution C). To test for the presence of Cl Ϫ /HCO 3 Ϫ exchange, cells were switched from a chloride-containing solution (solution C) to a chloride-free solution (solution D) and examined for cell alkalinization (17). All solutions were gassed with 95% O 2 , 5% CO 2 . Fig. 1 shows the nucleotide sequence and conceptual translation of the open reading frame of the 7.6-kb Na ϩ :HCO 3

RESULTS
Ϫ cotransporter cDNA. The nucleotide sequence encodes a protein of 116 kDa. Three potential N-linked glycosylation sites are found at amino acid numbers 592, 597, and 617. A single cAMP-dependent protein kinase phosphorylation site (Lys-Lys-Gly-Ser) occurs at amino acid number 979. Protein kinase C and casein kinase II phosphorylation sites are indicated in Fig.  1. 149 nucleotides precede the coding region on the 5Ј-end and are also indicated in Fig. 1. The 4.3-kb 3Ј-untranslated region is not shown, but is included in the sequence as submitted to GenBank™.
An amino acid comparison between the Na ϩ :HCO 3 Ϫ cotransporter and AE3 illustrates significant similarity (29% identity). A hydropathy plot of the Na ϩ :HCO 3 Ϫ cotransporter indicates at least nine transmembrane domains (Fig. 2, lower panel). The carboxyl terminus contains a highly hydrophilic stretch of 15 amino acids (12 of which are lysine) corresponding to the extreme hydrophilic region in the plot.
Multiple Tissue Northern Blots-A human multiple tissue Northern blot was purchased from Clontech and probed with a 32 P-labeled polymerase chain reaction product containing nucleotides 2737-2973 (the homologous region of W39298, which was common to both cDNA clones used in the construct). Fig. 3 (upper panel) shows a 7.6-kb mRNA in human kidney and pancreas hybridized strongly with the probe, indicating that these tissues express relatively high amounts of the Na ϩ : HCO 3 Ϫ cotransporter under basal conditions. A faint band can also be detected in human brain. The lower panel in Fig. 3 demonstrates the expression of GADPH in all lanes of the multiple tissue Northern blot, indicating that intact mRNA was present in all lanes.
Functional Expression of the Cloned cDNA-To determine the functional identity of the protein encoded by the cDNA, transiently transfected cells were assayed for the presence of Na ϩ :HCO 3 Ϫ cotransport, Na ϩ -dependent Cl Ϫ /HCO 3 Ϫ exchange, or Na ϩ -independent Cl Ϫ /HCO 3 Ϫ exchange. Accordingly, cells were acidified using an NH-pulse and allowed to recover in the presence of Na ϩ and 1 mM amiloride (to block Na ϩ /H ϩ exchange). In the absence of HCO 3 Ϫ , transfected cells showed negligible recovery from intracellular acidosis (data not shown). In the presence of both HCO 3 Ϫ and sodium, control cells showed no pH i recovery from the intracellular acidification (Fig. 4A). However, in transfected cells, switching from the Na ϩ -free solution (solution A, Table I) to the Na ϩ -containing solution (solution C) in the presence of HCO 3 Ϫ resulted in a rapid recovery from acidic pH i (Fig. 4B), with the recovery of 0.225 pH (⌬pH i of 0.225 Ϯ 0.035 in transfected cells versus almost 0 in nontransfected cells) (n ϭ 5). The recovery from cell acidification was observed only in the presence Na ϩ and HCO 3 Ϫ and was completely inhibited by 300 M DIDS (Fig. 4B), consistent with the presence of Na ϩ :HCO 3 Ϫ cotransport. Depleting the intracellular Cl Ϫ (18) by incubating the cells in Cl Ϫ -free media (only Cl Ϫ -free solutions were used for the duration of the experiment) did not reduce the rate of Na ϩ -dependent HCO 3 Ϫ movement into acid-loaded cells (Fig. 4C), indicating that the cloned transporter is not the Na ϩ -dependent Cl Ϫ /HCO 3 Ϫ exchanger (⌬pH i was 0.225 Ϯ 0.035 in chlorine-containing cells (n ϭ 5) and 0.28 Ϯ 0.018 in chlorine-depleted cells (n ϭ 4) Cloning and Functional Expression of a Human Kidney Na ϩ :HCO 3 Ϫ Cotransporter (p Ͼ 0.05). Transfected cells, switched from a Cl Ϫ -containing solution (solution C) to Cl Ϫ -free media (solution D), demonstrated little cell alkalinization (Fig. 4D), indicating that under physiological conditions, the cloned transporter does not function in Cl Ϫ /HCO 3 Ϫ exchange mode. To determine whether the cloned transporter can mediate HCO 3 -dependent 22 Na influx, HEK 293 cells were grown in 24-well plates, acidified with NH 4 pulse in a manner similar to Fig. 4A and assayed for 22 Na influx in the presence of HCO 3 Ϫ . The results showed that transfected cells mediated significant acid-stimulated DIDS-sensitive 22 Na influx in the presence of HCO 3 Ϫ , whereas nontransfected cells had no DIDS-sensitive 22 Na influx (7.15 Ϯ 0.5 nmol/mg of protein/4 min in transfected versus 0.25 Ϯ 0.1 nmol/mg of protein/4 min in nontransfected cells, p Ͻ 0.001, n ϭ 4). The Na ϩ :HCO 3 Ϫ cotransporter can mediate the movement of HCO 3 Ϫ out of or into the cell, depending on the ionic composition of experimental solutions (3)(4)(5). In the present study, switching the Na ϩ -containing solution to Na ϩ -free media did not result in significant cell acidification, as would have been the case if the transporter were functioning in efflux mode. Rather, the transporter functioned only in an uptake mode. Whether lack of efflux mode was caused by low intracellular Na ϩ concentration, decreased membrane potential, or other mechanisms is not clear at present.

DISCUSSION
The cDNA clone was identified by virtue of its significant homology with, but divergence from, the anion exchanger family. Both the homology with, and the divergence from AE3, are apparent in Fig. 2. Fig. 1 shows the nucleotide sequence and conceptual translation of the open reading frame of the Na ϩ : HCO 3 Ϫ cotransporter. The close relationship between the size of the cDNA clone (7586 base pairs, not including the poly(A) tail) and the mRNA found in human kidney indicate that the full sequence has been obtained.
Tissue distribution studies show high expression levels in kidney and pancreas, with lower levels of expression in the brain. Expression in the kidney (3)(4)(5), pancreas (19), and brain is consistent with functional studies (20). Heart and lung do not appear to express this transporter under basal conditions despite the fact that functional studies demonstrate the presence of Na ϩ :HCO 3 Ϫ cotransport in these tissues (18,21). This finding raises the possibility that the Na ϩ : cotransporter in the heart and lung is another isoform from this family. Functional studies upon transient expression in HEK293 cells showed that this transporter causes recovery from acute intracellular acidosis only in the presence of sodium and bicarbonate. The amiloride-insensitive pH i recovery in the presence of bicarbonate was detectable only in the presence of Na ϩ , indicating the presence of a Na ϩ and HCO 3 Ϫ -dependent cotransport process. The Cl Ϫ independence of this cotransporter and sensitivity to inhibition by DIDS indicates that this transporter is distinct from the Na ϩ -dependent Cl Ϫ /HCO 3 Ϫ exchange and thus demonstrates the presence of Na ϩ :HCO 3 Ϫ cotransport (Figs. 4, B and C).
Functional studies have shown inhibition of Na ϩ :HCO 3 Ϫ cotransport by protein kinase C and by cAMP-dependent protein kinase (22). Hence, it is not surprising to find consensus sites for phosphorylation by these two enzymes. A closer look at regulation of Na ϩ :HCO 3 Ϫ cotransport by casein kinase II may be warranted, since a large number of potential phosphorylation sites for this enzyme were found.
In certain epithelia, such as kidney and colon, this transporter mediates the exit of HCO 3 Ϫ from the cell to the blood (3)(4)(5), whereas in other epithelial tissues, as well as in nonepithelial tissues, sodium bicarbonate is transported from blood to the cell (18 -21). Whether the difference in the direction of the Na ϩ :HCO 3 Ϫ cotransporter movement in kidney and other tissues is due to differences in the membrane potential, cellular ionic composition in these tissues, or whether it suggests the presence of other isoforms of this transporter remains to be determined.
In conclusion, a cDNA encoding a Na ϩ :HCO 3 Ϫ cotransporter was cloned based on similarity of the anion exchangers to an expressed sequence tag. The Na ϩ :HCO 3 Ϫ cotransporter cDNA encodes an mRNA of 7.6 kb and a protein molecular mass of 116 kDa. The Na ϩ :HCO 3 Ϫ cotransporter mRNA is expressed in several tissues, including kidney, pancreas, and brain, indicating its pivotal role in cell pH regulation in mammalian tissues. FIG. 4. Functional expression of the cloned DNA. HEK293 cells were transfected with the cloned DNA and assayed for HCO 3 Ϫ -dependent transporter activity as described. A represents nontransfected cells, B and C indicates transfected cells. C represents assayed in Cl Ϫ -free solution. In D, cells were switched from Cl Ϫ -containing to Cl Ϫ -free.