Molecular Cloning, Chromosomal Localization, Tissue Distribution, and Functional Expression of the Human Pancreatic Sodium Bicarbonate Cotransporter*

We report the cloning, sequence analysis, tissue distribution, functional expression, and chromosomal localization of the human pancreatic sodium bicarbonate cotransport protein (pancreatic NBC (pNBC)). The transporter was identified by searching the human expressed sequence tag data base. An I.M.A.G.E. clone W39298 was identified, and a polymerase chain reaction probe was generated to screen a human pancreas cDNA library. pNBC encodes a 1079-residue polypeptide that differs at the N terminus from the recently cloned human sodium bicarbonate cotransporter isolated from kidney (kNBC) (Burnham, C. E., Amlal, H., Wang, Z., Shull, G. E., and Soleimani, M. (1997) J. Biol. Chem. 272, 19111–19114). Northern blot analysis using a probe specific for the N terminus of pNBC revealed an ∼7.7-kilobase transcript expressed predominantly in pancreas, with less expression in kidney, brain, liver, prostate, colon, stomach, thyroid, and spinal chord. In contrast, a probe to the unique 5′ region of kNBC detected an ∼7.6-kilobase transcript only in the kidney. In situhybridization studies in pancreas revealed expression in the acini and ductal cells. The gene was mapped to chromosome 4q21 using fluorescentin situ hybridization. Expression of pNBC in Xenopus laevis oocytes induced sodium bicarbonate cotransport. These data demonstrate that pNBC encodes the sodium bicarbonate cotransporter in the mammalian pancreas. pNBC is also expressed at a lower level in several other organs, whereas kNBC is expressed uniquely in kidney.

Sodium bicarbonate cotransport mediates the coupled movement of Na ϩ and HCO 3 Ϫ ions across the plasma membrane of many cells (1). This transport process is involved in bicarbonate secretion/absorption and intracellular pH (pH i ) 1 regulation.
Functional Na(HCO 3 ) n cotransport was first identified in the salamander Ambystoma tigritum kidney (2) and has since been documented functionally in numerous other cell types including pancreas (3)(4)(5)(6)(7)(8), colon (9), liver (10 -12), heart (13,14), retinal Mü eller cells (15,16), glial cells (17,18), parietal cells (19),and type II alveolar cells (20). Depending on the cell type, the stoichiometry of Na ϩ to HCO 3 Ϫ flux is 3:1, 2:1, or 1:1. As characterized in many cell types, several features distinguish the Na(HCO 3 ) n cotransporter from other bicarbonate-dependant transporters.1) Na(HCO 3 ) n cotransport is not dependent on the presence of Cl Ϫ , 2) transport is inhibited by stilbenes, and 3) transport is stimulated in the presence of HCO 3 Ϫ . In the kidney, Na(HCO 3 ) n cotransport was initially localized by functional studies to the basolateral membrane of the proximal tubule where it plays an important role in mediating electrogenic basolateral bicarbonate efflux (2,21,22). Although Na ϩ -dependent and -independent Cl Ϫ /base exchangers also contribute to basolateral bicarbonate transport in the proximal tubule (23)(24)(25)(26)(27), current evidence suggests that electrogenic Na(HCO 3 ) n cotransport mediates the majority of bicarbonate efflux in this nephron segment (1). Romero et al. have recently cloned a renal electrogenic sodium bicarbonate cotransporter (NBC) from rat (28) and salamander kidney (29). Burnham et al. have reported the cloning of a sodium bicarbonate cotransporter from human kidney (30). The NBC clone isolated from salamander kidney encoded a 4.2-kb mRNA transcript that was expressed predominantly in kidney, with less expression in small intestine, large intestine, brain, eye, and bladder (29). Human NBC isolated from kidney encoded a ϳ7.6-kb mRNA and was reportedly also expressed in pancreas and brain by Northern analysis using a probe to the 3Ј region of kNBC (30). NBC expression in the kidney has recently been shown to be highest in the S1 proximal tubule, with less expression in the proximal straight tubule (31). The high level of expression in S1 proximal tubules is in keeping with the high rate of transepithelial bicarbonate transport in this segment (32,33).
The pancreas secretes digestive enzymes dissolved in a HCO 3 Ϫ -rich fluid (34). Pancreatic bicarbonate secretion is mediated by principal cells lining the pancreatic ducts (34,35). Previous functional studies have led to a cell model that can account for transcellular bicarbonate secretion. Apical bicarbonate secretion is thought to be mediated by an apical Cl Ϫ / base exchanger acting in parallel with a small conductance cystic fibrosis transmembrane conductance regulator Cl Ϫ chan-* This work was supported by National Institutes of Health Grant DK46976, the Iris and B. Gerald Cantor Foundation, the Max Factor Family Foundation, the Verna Harrah Foundation, the Richard and Hinda Rosenthal Foundation, and the Fredericka Taubitz 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM  nel on the apical membrane (5,36). Influx of H ϩ equivalents during the process of apical bicarbonate secretion requires the efflux of H ϩ or the influx of bicarbonate in the steady state. Studies of pig, rat, and guinea pig pancreatic ducts have demonstrated the presence of a basolateral Na ϩ /H ϩ antiporter, which serves an important housekeeping role (5,8,37,38). A basolateral vacuolar-type H ϩ /ATPase and Na(HCO 3 ) n cotransporter are thought to play an important role in agonist-mediated bicarbonate secretion (5)(6)(7)(8)39). The relative contribution of these transporters to basolateral H ϩ /base transport and their respective stimulation by secretogues in different species is controversial. The importance of basolateral Na(HCO 3 ) n cotransport has recently been documented in studies of isolated rat and guinea pig pancreatic ducts (5)(6)(7)(8)39). Ishiguro et al. (7) reported that basolateral Na(HCO 3 ) n cotransport contributed up to 75% of basolateral bicarbonate uptake during stimulation of transepithelial bicarbonate transport by secretin. Furthermore, in isolated pancreatic acini, Na(HCO 3 ) n cotransport has been shown to participate in the regulation of pH i after acid loads (3). On the basis of HCO 3 Ϫ flux measurements and thermodynamic considerations, it was concluded that this transporter contributes to HCO 3 Ϫ efflux under unstimulated conditions (3, 7) with a stoichiometry of 3:1 (3), although direct measurements of the stoichiometry have thus far not been performed. After depolarization of the basolateral membrane by secretin (40), the electrochemical driving forces would favor basolateral bicarbonate influx (7).
Although the functional characteristics of pancreatic Na(HCO 3 ) n cotransport have begun to be investigated, the protein responsible for this function has not been identified. We report here the cloning the human pancreatic Na(HCO 3 ) n cotransporter (pNBC). The predicted pNBC polypeptide is 1079 amino acids in length, whereas the NBC variant expressed in kidney (kNBC) consists of 1035 amino acids. The C-terminal 994 amino acids of pNBC and kNBC are identical. pNBC has a unique N terminus of 85 amino acids that replaces the initial 41 amino acids in kNBC. Expression of the cDNA encoding pNBC in Xenopus oocytes results in sodium-dependent and chloride-independent HCO 3 Ϫ transport, which is inhibited by DIDS.

EXPERIMENTAL PROCEDURES
Cloning and Sequencing of pNBC-A 159-bp PCR product (2795-2954 bp in human pNBC) was generated using the human pancreas NBC EST clone W39298 (I.M.A.G.E clone) as a template, random primer-labeled with 32 P and used to screen a human pancreas gt10 cDNA library (CLONTECH, Palo Alto, CA). A similar approach was utilized by Burnham et al. (30) while the present studies were in progress to obtain an NBC clone from human pancreas. Standard hybridization conditions were employed (42°C, 50% formamide, 5ϫ standard saline phosphate EDTA (SSPE), 5ϫ Denhardt's solution, 0.5% SDS, 0.2 mg/ml prehybridization herring sperm DNA). The filters were washed three times with 1ϫ SSC/0.1%SDS (42°C) and once with 0.1ϫ SSC/0.1%SDS (25°C) (1ϫSSC ϭ 0.15 M NaCl and 0.015 M sodium citrate). Positive clones were verified by sequencing. Two overlapping clones (7.1) and (9.2.1) were obtained that contained the entire coding region. To obtain a full-length clone containing the complete open reading frame, these two clones were spliced together using a common SpeI restriction site and subcloned into pPCR-Script SK(ϩ) (Stratagene, La Jolla, CA). To increase the stability of the capped RNA transcribed from this clone, a poly(A) (70-mer) oligonucleotide was added to the 3Ј end of the clone between a SalI and KpnI restriction site. The 5Ј end of the coding sequence for pNBC was confirmed by 5Ј rapid amplification of cDNA ends PCR amplification and primer extension analysis. Furthermore, to confirm that the pNBC amino acid sequence was derived from a bona fide transcript, we amplified the entire open reading frame of human pNBC by reverse transcription-PCR using Marathon Ready cDNA prepared from human pancreas (CLONTECH, Palo Alto, CA) as a template. Nucleotide sequences were determined bidirectionally by automated sequencing (ABI 310 Perkin-Elmer) using Taq polymerase (Ampli-Taq FS, Perkin-Elmer). Sequence assembly and analysis was carried out using Geneworks software (Oxford, UK).
Northern Analysis and Tissue Distribution-Northern blots with various human tissues were obtained from CLONTECH. The various probes were random prime-labeled with [ 32 P]dCTP to a specific activity of about 1.5 ϫ 10 9 dpm/g. The filters were prehybridized at 42°C for 2 h using 50% formamide, 6ϫ SSPE, 0.5% SDS, Denhardt's solution, and 0.1 mg/ml of sheared herring testes denatured DNA. After the prehybridization, the filters were incubated with the 32 P probe using 25 ml of hybridization buffer. The probes were denatured and added to the hybridization solution at 10 7 dpm/ml. The filters were probed at 42°C for 18 h and washed in 1ϫ SSC, 0.1% SDS at 45°C for 60 min (3 changes, 350 ml/wash); after exposure for 9.5 h, the filters were rewashed in 1ϫ SSC, 0.1%SDS at 65°C for 30 min and 0.1ϫ SSC, 0.1%SDS at 65°C for 1 h. The glyceraldehyde-3-phosphate dehydrogenase DNA was T4 polynucleotide kinase (New England Biolabs, Beverly, MA) labeled with [ 32 P]␥ATP to a specific activity of 2.5 Ci/pmol. The filters were prehybridized at 42°C for 2 h using 50% formamide, 6x SSPE, 0.5% SDS, Denhardt's solution, and 0.1 mg/ml sheared herring testes denatured DNA. After the prehybridization, the filters were probed with the 32 P probe using 25 ml of hybridization buffer. The probe was denatured and added to the hybridization solution at 0.5 pmol/ml. The filters were probed at 42°C for 18 h and then washed 3 times in 1x SSC, 0.1%SDS at 45°C for 30 min. The following probes were used in the Northern blot studies: 1) a 159-bp PCR product with a sequence common to both pNBC and kNBC (nucleotides 2795-2954 pNBC sequence and 2695 to 2854 in the kNBC sequence); 2) synthetic oligonucleotide specific for pNBC (nucleotides 118 to 212 in the pNBC sequence); 3) synthetic oligonucleotide specific for kNBC (nucleotides 175 to 268 in the kNBC sequence).
Preparation of Riboprobes for in Situ Hybridization-To prepare the riboprobes, the insert (9.2.1) was subcloned into pPCR-script SK(ϩ) (Stratagene) Riboprobes were synthesized by in vitro transcription and labeled with 35 S-CTP. For generation of the antisense riboprobe, the plasmid was linearized with SstI and transcribed by T7 RNA polymerase. For generation of the sense riboprobe, the plasmid was linearized with KpnI and transcribed with T3 RNA polymerase. The RNA transcripts were purified by phenol-chloroform extractions and Sephadex G-50 spin columns (Sigma). The final products were suspended in Tris-EDTA buffer with 0.1 M dithiothreitol. The RNA transcripts were then sheared by alkaline hydrolysis at 68°C for 5 min. After the shearing, the reaction was neutralized by adding 3 M sodium acetate, pH 5, to make a final acetate concentration of 0.3 M. Slices of mouse pancreas (1 mm) were fixed in 4% formalin, and 5-m sections were attached to glass slides (Fisher). The slides were prewashed and digested for 15 min at 37°C with proteinase K. To reduce nonspecific background staining, the slides were succinylated with succinic anhydride and acetylated with acetic anhydride. The riboprobes were hybridized for 18 h at 45°C. The slides were then washed for 15 min in 2ϫ SSC at room temperature, followed by a wash (15 min) in 1 ϫ SSC/50% formamide at 45°C, then three washes in 2 ϫ SSC/0.1% Triton X-100 at 60°C for 15 min each, followed by two washes in 0.1 SSC at 60°C for 15 min each. The slides were then digested by RNase A (25 g/ml; Sigma) and RNase T1(25 units/ml; Sigma) for 40 min at 37°C. The slides were washed twice in 2ϫ SSC at 60°C for 15 min each and then dehydrated in 0.3 M ammonium acetate, 70% ethanol for 5 min followed by a further 5 min of dehydration in 0.3 M ammonium acetate, 95% ethanol. The slides were dipped into NTB2 emulsion solution (Eastman Kodak Co.) and exposed for 3 days at 4°C followed by hematoxylin/eosin staining. The sections were imaged using a Zeiss Axiophot microscope (Max Erb, Los Angeles, CA) and digitized using a Sony 3CCD color video camera (model DXC-960MD, Compix Imaging Systems, Tuscon, AZ) with C Imaging software (Compix Imaging Systems, Tuscon, AZ).
Fluorescent In Situ Hybridization-The PCR probe used to screen the pancreatic cDNA library was also used to screen an arrayed PAC human genomic library (Genome Systems, St. Louis, MO). DNA from clone F335 was identified by sequencing and was then labeled with digoxigenin dUTP by nick translation. Labeled probe was combined with sheared human DNA and hybridized to normal metaphase chromosomes derived from phytohemagglutinin-stimulated peripheral blood lymphocytes in a solution containing 50% formamide, 10% dextran sulfate, and 2ϫ SSC. Specific hybridization signals were detected by incubating the hybridized slides in fluoresceinated antidigoxigenin antibodies followed by counterstaining with 4Ј,6-diamidino-2-phenylindole, dihydrochloride for one color experiments. Probe detection for two color experiments was accomplished by incubating the slides in fluoresceinated antidigoxigenin antibodies and Texas red avidin followed by counterstaining with 4Ј,6-diamidino-2-phenylindole, dihydrochloride.
The sequences submitted to GenBank TM have been scanned against the data base, and the following related human sequences have been identified (AF007216, P02730, P48751, U62531).

Isolation of cDNA Clones and Characterization of Multiple
Human NBC Transcripts-The cloning of pNBC was based on the identification of human pancreatic cDNA clone in the Gen-Bank TM data base. To obtain full-length pNBC, we screened a human pancreas gt10 cDNA library using a probe generated by PCR amplification of the human EST sequence. Two overlapping clones were obtained that were fused at a shared SpeI restriction site to generate a full-length clone containing the entire open reading frame. To confirm that the pNBC sequence was derived from a bona fide transcript, reverse transcription-PCR was used to generate a full-length PCR product containing the complete open reading frame. Analysis of the full-length clone revealed a 1079-amino acid open reading frame beginning with the initial methionine as well as 117 bp of 5Ј-untranslated region. Additional overlapping 3Ј-untranslated rapid amplification of cDNA ends sequences were almost identical to the sequence recently published by Burnham et al. (30) except for minor changes likely due to polymorphisms. The nucleotide sequence of human pNBC has been submitted to the GenBank TM (accession number AF011390).
Structure of pNBC-The overall structure of pNBC is similar to kNBC (30) and other members of the anion exchange gene family (43). Specifically, pNBC has 12 predicted transmembrane regions and hydrophilic intracellular N-and C-terminal regions. As shown in Fig. 1, the sequences of pNBC and kNBC   FIG. 1. Comparison of the N terminus of pNBC and kNBC. a, ATG translation start sites are underlined. The nucleotide sequence common to pNBC and kNBC is highlighted in gray. b, predicted pNBC and kNBC polypeptide sequences. The sequences of pNBC and kNBC differ before the Ser residue at position 42 of kNBC and position 86 of the pancreatic sequence. In pNBC, several consensus phosphorylation sites are depicted: 1) protein kinase A beginning at Lys 46 (dotted line; 2) protein kinase C beginning at Ser 38 and Ser 65 (thick line); and 3) casein kinase II at Ser 68 (thin line). Amino acids 1-41 of kNBC lack consensus phosphorylation sites. The amino acid sequence common to pNBC and kNBC is highlighted in gray.
are identical after the Ser residue at position 42 of kNBC and position 86 of the pancreatic sequence. Unlike the kidney sequence, the N terminus of pNBC before the region common to both polypeptides contains blocks of charged amino acids. Further distinctive features of the N terminus of pNBC are 1) the consensus phosphorylation site for protein kinase A beginning at Lys 46, 2) the consensus phosphorylation sites for protein kinase C beginning at Ser 38 and Ser 65 , and 3) the casein kinase II phosphorylation site beginning at Ser 68 . In contrast, amino acids 1-41 of kNBC lack consensus phosphorylation sites. pNBC and kNBC polypeptides are predicted to have identical transmembrane and C-terminal regions.
Tissue Expression of pNBC and kNBC-The expression of pNBC and kNBC was examined in various human tissues by Northern blot analysis (Fig. 2). Specific probes were prepared that contained the unique N-terminal region of each isoform. A probe to the common 3Ј-coding region of pNBC and kNBC recognized transcripts in RNA samples from the pancreas, with less expression in kidney, brain, liver, prostate, colon, stomach, thyroid, and spinal chord. Burnham et al. (30), using a probe to nucleotides 2737 to 2973 of kNBC (2837 to 3073 in pNBC), detected transcripts in kidney, pancreas and brain. Impor-tantly, the results of the present study demonstrate that the 5Ј-coding region of pNBC differs from the 5Ј terminus of the human NBC sequence reported by Burnham et al. (30). The C-terminal 994 amino acids of pNBC and kNBC are identical. pNBC has a unique N terminus of 85 amino acids that replaces the first 41 amino acids in kNBC. This observation is consistent with the failure of a probe derived from the 5Ј terminus of kNBC to detect the ϳ7.7-kb pNBC transcript in the Northern blot experiments. As shown in Fig. 2, the expression of the ϳ7.6-kb kNBC transcript was restricted exclusively to kidney. The kNBC probe failed to detect a transcript in any tissue other than kidney (despite longer exposure times, and lower stringency), suggesting that the kNBC N terminus is unique. The mechanism responsible for the tissue-specific expression of kNBC and pNBC is unknown but may involve activation of a downstream promoter and/or differential splicing of the primary transcripts in kidney.
In separate experiments, a Northern blot from a variety of human tissues was screened with a probe derived from the unique 5Ј portion of the pancreas cDNA to confirm that the sequence is represented in the ϳ7.7-kb mRNA transcript. As shown in Fig. 2, this probe detected a ϳ7.7-kb mRNA abundant in pancreas and present at lower levels in kidney, brain, liver, prostate, colon, stomach, thyroid, and spinal chord. The size of the transcripts in most tissues was identical to that of the pancreatic NBC transcript identified with a probe common to the 3Ј-coding region of pNBC and kNBC. However, in thyroid, a transcript of higher mobility was detected by both the specific pNBC probe and the C-terminal probe common to kNBC and pNBC. This transcript may represent a variant of pNBC.
In Situ Hybridization of pNBC-To determine in greater detail the distribution of pNBC within the pancreas, we used in situ hybridization with a pNBC 35 S-labeled riboprobe. Mouse pancreas was used because of tissue availability. Mouse pNBC was cloned using a PCR-based strategy. The sequence of mouse pNBC was found to be 93% identical to the human sequence and has been deposited in GenBank TM (accession number AF020195). Microautoradigraph analysis of frozen pancreas sections hybridized with the pNBC riboprobe showed strong expression in the pancreatic ducts and pancreatic acini (Fig. 3). A signal was not detected in the islets.
Chromosomal Localization of Human NBC-The initial experiment resulted in the specific labeling of the long arm of a group B chromosome, which was believed to be chromosome 4 on the basis of size, morphology, and banding pattern. A second experiment was conducted in which a biotin-labeled probe that is specific for the centromere of chromosome 4 (D4ZI) was cohybridized with clone F335. This experiment resulted in specific labeling of the centromere of chromosome 4. Measurements of 10 specifically labeled chromosomes 4 demonstrated that clone F335 maps to a position that is 19% that of the distance from the centromere to the telomere of chromosome arm 4q, an area that corresponds to band 4q21. A total of 80 metaphase cells were examined, with 71 exhibiting specific labeling (Fig. 4).
Functional Expression of pNBC in Xenopus Oocytes-We examined the functional properties of pNBC using measurements of pH i , 22 Na, and 36 Cl Ϫ uptake after injecting the corresponding polyadenylated cRNA into Xenopus oocytes. Polyadenylated cRNA prepared as described above was injected into oocytes and allowed to express for 72 h. Augmented 22 Na uptake was observed in oocytes injected with pNBC cRNA (Fig. 5). The uptake was 16-fold greater than control oocytes, p Ͻ 0.001. Uptake was significantly inhibited in the presence of DIDS (0.3 mM). Cl Ϫ uptake was not significantly affected by pNBC cRNA injection: 0.16 Ϯ 0.05 nmol/h/oocyte in controls (n ϭ 8) and Each lane was loaded with ϳ2 g of poly(A) ϩ human RNA. The multiple tissue Northern blots were from CLONTECH. The following 32 P-labeled NBC probes were used: a, a 159-bp PCR product with sequence common to both pNBC and kNBC (nucleotides 2795-2954 pNBC sequence and 2695 to 2854 in the kNBC sequence); b, synthetic oligonucleotide specific for pNBC (nucleotides 118 to 212 in the pNBC sequence); c, synthetic oligonucleotide specific for kNBC (nucleotides 175 to 268 in the kNBC sequence). The blots were also probed for glyceraldehyde-3phosphate as shown in d.
Further studies were done in which pH i transients were measured in control oocytes and those injected with pNBC cRNA (Fig. 6). Resting pH i was similar in both groups of oocytes, ϳ7.1 After extracellular Na ϩ removal, in control oocytes, pH i increased by ϩ0.008 Ϯ 0.001 pH/min (n ϭ 7). In contrast, in the cRNA-injected oocytes, pH i decreased at a rate of Ϫ0.008 Ϯ 0.001 pH/min (n ϭ 5), p Ͻ 0.001. Na ϩ removal caused a similar decrease in pH i in cRNA-injected oocytes in the absence of chloride with 10 M EIPA; Ϫ0.010 Ϯ 0.001 pH/min (n ϭ 3), p ϭ NS. DIDS (0.3 mM), significantly decreased the Na ϩ -induced pH i transient in cRNA-injected oocytes to Ϫ0.0025 Ϯ 0.0003 pH/min (n ϭ 3), p Ͻ 0.05.
The Physiological Role of pNBC-mediated Na(HCO 3 ) n Cotransport-The highest level of pNBC expression was found in the pancreas, with lower levels of expression in kidney, brain, liver, prostate, colon, stomach, thyroid, and spinal chord. The results of the present study are compatable with previous functional studies that have demonstrated Na(HCO 3 ) n cotransport in pancreatic ductal cells and acini (3,(5)(6)(7)(8)39). It has been hypothesized that basolateral Na(HCO 3 ) n cotransport in pancreatic ductal cells plays a modulatory role in ductal fluid secretion (5,7). Under resting conditions, the basolateral cotransporter would mediate cellular bicarbonate efflux when the basolateral membrane potential is ϳϪ70 mV (40). After stimulation of bicarbonate secretion by secretin, the basolateral membrane voltage of rat duct cells depolarizes to ϳϪ40 to Ϫ20 mV (40). Under these conditions, the cotransporter would mediate bicarbonate influx (7). After stimulation by secretin, basolateral bicarbonate uptake by guinea pig pancreatic ductal cells is mediated in part by the basolateral Na(HCO 3 ) n cotransporter (5)(6)(7)(8)39), although a basolateral H ϩ -ATPase may also play a role (5,6,8). Two important physiological roles for bicarbonate secretion by pancreatic centroacinar cells and ductal cells have been proposed (44) 22 Na uptake (expressed as nmol/h/oocyte) in oocytes injected with water or with pNBC cRNA is shown. Expression was assessed in the presence and absence of 0.3 mM DIDS. Each bar represents the mean Ϯ S.E. of 10 to 14 oocytes. 22 Na uptake was significantly increased by pNBC cRNA versus H 2 O-injected oocytes and inhibited in the presence of DIDS (0.3 mM); * indicates p Ͻ 0.001 cRNA versus H 2 O-injected and cRNA versus cRNA plus DIDS. 36 Cl Ϫ uptake (expressed as nmol/h/oocyte) was also assessed in oocytes injected with water or with pNBC cRNA. Cl Ϫ uptake was not significantly affected by pNBC cRNA injection. the acinar lumen and 2) neutralization of the acidic chyme delivered into the upper intestine from the stomach. In the absence of secretogogues, cellular bicarbonate efflux via by the basolateral Na(HCO 3 ) n cotransporter coupled to apical Na ϩ /H ϩ exchange may mediate transepithelial H ϩ secretion in the main and common pancreatic ducts (5).
The results of the present study indicate that the pNBC is also expressed at lower levels in kidney, brain, liver, prostate, colon, stomach, thyroid, and spinal chord. The N terminus of pNBC has a unique consensus phosphorylation site for protein kinase A beginning at Lys 46 , consensus phosphorylation sites for protein kinase C beginning at Ser 38 and Ser 65 , and a casein kinase II phosphorylation site beginning at Ser 68 , which kNBC lacks. Of interest, cAMP stimulates transepithelial bicarbonate secretion and basolateral Na(HCO 3 ) n cotransport in pancreatic ducts (7,45), whereas in the renal proximal tubule, cAMP inhibits basolateral Na(HCO 3 ) n cotransport (46). The unique N terminus of pNBC not shared by kNBC could have an important regulatory role in functioning as a target for phosphorylation by protein kinase A. Na(HCO 3 ) n cotransport has been functionally demonstrated in pancreas (3)(4)(5)(6)(7)(8), kidney (1,2,21,22,(25)(26)(27)31), leech glial cells (17,18), retinal Mü ller cells (15,16), type II alveolar cells (20), hepatocytes (10 -12), colon (9), gastric parietal cells (19), cardiac Purkinje fibers (13) and ventricular myocytes (14), although the stoichiometry of the transported species appears to be tissue-dependent. The lack of detectable transcripts in heart and lung with any of the three probes used in this study suggests the possibility that cardiac and lung Na(HCO 3 ) n cotransport is mediated by an alternative protein(s). The finding that prostate and thyroid have transcripts that are labeled by the pNBC probe is of interest, given that these tissues have not been previously investigated for the presence of functional Na(HCO 3 ) n cotransport. Furthermore the higher mobility of the thyroid transcript suggests that this tissue expresses a variant of pNBC not present in other organs.