INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CHANGES IN INTRACELLULAR pH (pH_i) may affect cardiac contractility (5, 37) and may also be associated with malignant arrhythmias during metabolic derangements (8). Protection from intracellular acidification is provided by the Na⁺/H⁺ antiport and a Na⁺/HCO symporter (9, 11, 28), whereas intracellular alkalosis is countered by the sarcolemmal Cl/HCO exchanger (49, 52) and a more recently described cardiac Cl/OH exchanger (31). In addition to pH_i homeostasis, anion exchange mechanisms are involved in regulation of intracellular Cl concentration ([Cl]_i) and cell volume (2). The former may play a significant role in cellular automaticity as Cl movements through Cl channels can modulate the action potential with potential proarrhythmic results (1, 16).

Whereas the molecular identity of the cardiac Cl/OH exchanger is not known, cardiac Cl/HCO exchange activity is likely mediated by one or more members of the anion-exchanger (AE) gene family (2). The AE1 gene encodes the erythroid band 3 (eAE1) and the type A intercalated cell kAE1 polypeptides. The gene is also expressed in spleen, heart, lungs, and liver (26). The AE2 gene is expressed most abundantly in the stomach and choroid plexus and is widely distributed in other tissues as well (26, 51). AE3 transcripts are present in greatest abundance in the heart and brain as well as in the gut (23, 26, 55). Human cardiac membranes contain AE3 polypeptides (55) and isolated adult rat cardiomyocytes contain AE3 and AE1 polypeptides in membrane fractions (39). Cardiac anion exchange can be activated by ATP (38) and by ANG II (7). Puceat and co-workers (40) have suggested that the component of cardiocyte anion exchange activated by ATP in adult rat cardiomyocytes is mediated by cardiac AE1 polypeptide and not by AE3 polypeptide.

Cardiac anion-exchange activity has been studied in adult mammalian myocardial fibers (49), in single adult rat and guinea pig cardiocytes (31, 38, 52), and in neonatal rat cardiocytes (25, 40). Both adult and neonatal cardiocytes in culture pose limitations to some types of molecular biological manipulations. Adult mammalian cardiomyocytes do not proliferate, cannot reenter the cell cycle (43), and are difficult to maintain in culture. Embryonic or neonatal cardiomyocytes may proliferate in primary cultures but cannot be passaged.

AT-1 cells are derived from atrial tumors of transgenic mice expressing the simian virus 40 (SV40) large T antigen cDNA linked to the atrial natriuretic factor (ANF) promoter (13). These cells exhibit spontaneous synchronous contractions when confluent and maintain a highly differentiated adult phenotype (30). They express -myosin heavy chain (-MHC), -cardiac actin, ANF, and connexin43; form well-organized myofibrils and gap junctions; and demonstrate the ultrastructure of normal atrial myocytes (12, 30, 46). AT-1 cells can be propagated as transplantable subcutaneous tumors in syngeneic mice and thus can be maintained via serial passage for years; they can therefore serve in the molecular study of cardiac ionic homeostasis, excitation-contraction coupling, impulse formation and propagation, anisotropic conduction, and arrhythmogenesis. Moreover, AT-1 cells can be used in studies addressing specific questions of atrial physiology.

Little is known regarding ionic homeostasis in AT-1 cells. Given the importance of the AT-1 cell line as an investigational tool in mammalian cardiac physiology, we undertook the present study to characterize the anion-exchange activity in these cells and define the expressed AE genes and their isoforms. We found that cultured AT-1 cells exhibit basal Na⁺-independent Cl/HCO activity which can be stimulated by extracellular ATP and ANG II. AT-1 cells express many mRNA products of all three AE genes (but lack kAE1 mRNA) and contain AE1, AE2, and AE3 polypeptides.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AT-1 cell preparation. AT-1 cells, the gifts of S. Izumo (Beth Israel Deaconess Medical Center) and R. Kline (Columbia University), were injected subcutaneously in syngeneic female mice 4-6 wk old (B6D2/F1, Jackson ImmunoResearch Laboratories). Subcutaneous tumors 2-3 cm in diameter grew within 8-12 wk. Tumors were removed under sterile conditions from anesthetized mice and AT-1 cells were isolated as previously described (20). Briefly, tumors were minced and incubated overnight at 4°C in a solution containing a 1:1 ratio of PC-1 medium (BioWhittaker) and 0.125% trypsin-EDTA. The tissue pieces were then incubated at 37°C sequentially for periods of 60, 30, and 30 min in PC-1 medium containing 0.1% collagenase (Sigma). Cells obtained from the second and third incubations were plated in PC-1 medium containing 10% fetal bovine serum, 10 nM dexamethasone, 100 U/ml penicillin, and 100 µg/ml streptomycin at a density of 0.5-1 × 10⁶ cells per milliliter in 25-cm² tissue-culture flasks (Corning). Within 7-10 days these cells exhibited spontaneous synchronous contractions. Beating cells were trypsinized and transferred onto sterile 25-mm Cell-Tak-coated glass coverslips. Subconfluent cells of passages 2-5 were studied within 2-3 days after plating onto coverslips, at which time they exhibited spontaneous beating.

pH_i measurements. AT-1 cells plated onto glass coverslips were incubated with 2 µM 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein-AM (BCECF-AM) for 25 min at room temperature and then rinsed. pH_i measurements and Cl/HCO exchange assays were performed as previously described (17). Briefly, each coverslip of BCECF-loaded cells was mounted in a Leiden closed perfusion chamber on the stage of an Olympus IMT-2 inverted microscope equipped with a Dage MTI CCD-72 series video camera, a Genisys image intensifier, a Pinnacle REO-650 optical disk drive, a color video monitor, and a printer. BCECF fluorescence was collected at 530 nm after alternating excitation at 495 and 440 nm at 6-s intervals. Fluorescence images were acquired and recorded with an Image-1 digital ratio-imaging system (Universal Imaging, West Chester, PA). Fluorescence ratios of at least 15 individual AT-1 cells on each coverslip were recorded as averages of 8 background-subtracted images obtained with a ×20 oil objective lens. Results were not different when fluorescence ratios were acquired from a region of interest encompassing at least 15 individual cells. Anion-exchange activity in AT-1 cells was expressed as the rate of change in pH_i (dpH_i/dt) immediately after transition from Cl-containing superfusate (in mM: 126 NaCl, 24 NaHCO₃, 5 KCl, 2 MgSO₄, 1 CaCl₂, and 10 glucose; and 5% CO₂ at pH 7.40) to Cl-free superfusate (in mM: 126 sodium gluconate, 24 NaHCO₃, 5 potassium gluconate, 2 MgSO₄, 3 calcium gluconate, and 10 glucose; and 5% CO₂ at pH 7.40). In some experiments equimolar N-methyl-D-glucamine (NMDG) chloride substituted for NaCl, and choline bicarbonate substituted for NaHCO₃.

RT-PCR. Total RNA was prepared from confluent cultures of spontaneously beating AT-1 cells or freshly resected mouse (CD-1 strain) kidney, spleen, stomach, brain, and heart using the Qiagen RNeasy kit. One microgram of total RNA was reverse transcribed from an oligo(dT) primer with the use of a First Strand synthesis kit (Ambion). RNA integrity was confirmed by PCR amplification of -actin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNAs (not shown). PCR was performed by the "hot start" procedure using Taq polymerase (Qiagen) in the supplier's recommended buffer. We used 5% of the RT reaction volume in a total PCR reaction volume of 50 µl. PCR mixes lacking only primers were preincubated at 82°C for 1 min, and appropriate primers were injected into the mix through mineral oil. The complete reaction mixes were then denatured for 2 min at 95°C and subjected to the following cycle conditions: denaturation for 45 s at 94°C, annealing for 2 min at 60°C, and elongation for 3 min at 72°C. Final extension of 15 min at 72°C was terminated by rapid cooling to 4°C after 36 cycles for AE PCR products or 25 cycles for -actin and GAPDH. Forward (F) and reverse (R) amplification primers used are listed in Table 1 and were coupled as follows: for kAE1, kAE1.F1 versus AE1.E7R, and kAE1.F1 versus AE1.E9R; and for eAE1, eAE1.E3F versus AE1.E7R, and eAE1.E3F versus AE1.E9R. For AE2 forward primers, AE2a.E2F, AE2b.F1, and AE2c.F1 were used versus the common reverse primer AE2.E8R. For AE3 forward primers, cAE3.F1 and bAE3.F1 were used versus the common reverse primer AE3.E11R.

Immunoblots. Fresh confluent AT-1 cell cultures were scrape-harvested and solubilized in 50 mM Tris · HCl at pH 8, 150 mM NaCl, and NP40 buffer containing protease inhibitor (BMB Complete, Boehringer). The lysate was microcentrifuged at 14,000 rpm at 4°C for 30 min. The supernatant was used for protein measurement by the bis-cinchoninic acid assay and was subsequently mixed to a 1:1.2 ratio with 2× Laemmli buffer containing 10% -mercaptoethanol. The mix was kept frozen at 80°C. Frozen samples were thawed, electrophoresed on 10% SDS-polyacrylamide gel, and transferred to nitrocellulose as previously described (3).

Immunocytochemistry. Coverslips of confluent AT-1 cells were washed in PBS (37°C) and then immediately fixed with 3% paraformaldehyde for 30 min, quenched with 50 mM glycine 3 × 5 min, and then washed in PBS 2 × 5 min. They were stored at 4°C in PBS with 0.02% sodium azide before being immunostained. AE proteins were visualized with affinity-purified rabbit antisera raised against mouse AE1 (aa 917-929), mouse AE2 (aa 1,224-1,237), mouse AE3 (aa 1,216-1,227), human bAE3 (aa 115-128), and human cAE3 (aa 42-53). Fixed coverslips were briefly preincubated in PBS at room temperature. In addition, coverslips to be incubated with AE2 (aa 1,224-1,237) antibodies were preincubated 10-15 min with 1% SDS in PBS and then 3 × 5 min in PBS, a process that elicits increased staining with this antibody. All coverslips were blocked with 1% BSA and 0.05% saponin in PBS for 15-20 min and then incubated with the diluted primary antibody for 1-2 h in the presence of 24 µg/ml irrelevant peptide or peptide antigen. Coverslips were then washed 3× for 5 min in PBS, incubated for 0.75-1 h with secondary antibodies conjugated to the fluorophore Cy3 (Jackson ImmunoResearch Laboratories), and washed 3× for 5 min in PBS. The coverslips were sealed onto slides and examined with an Olympus BH-2 photomicroscope equipped for epifluorescence and were photographed by using Kodak TMAX 400 film push-processed to 1,600 ASA.

Statistical analysis. Numerical data are expressed as means ± SD. Comparisons were made using Student's t-test for paired observations with the Bonferroni correction applied for multiple comparisons. The null hypothesis was rejected at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AT-1 cells express Cl/HCO exchange activity. Anion-exchange activity in superfused BCECF-loaded AT-1 cells was measured by monitoring the rate of intracellular alkalinization and reacidification in response to removal and subsequent readdition of extracellular Cl. Removal of extracellular Cl alkalinized AT-1 cells (maximal pH = 0.31 ± 0.07; dpH_i/dt = 0.042 ± 0.009 per minute; n = 42 cells in 8 coverslips) in a 100 µM DIDS-sensitive manner [maximal pH = 0.1 ± 0.03; dpH_i/dt = 0.012 ± 0.005 per minute; n = 58 cells in 11 coverslips; P < 0.001 (Fig. 1, A and B)].

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Cl/HCO exchange (AE) activity in AT-1 cells. Intracellular pH (pHi) was recorded during removal and restoration of extracellular Cl in the indicated conditions. Each panel represents an average pHi trace from at least five cells simultaneously and individually recorded within one visual field of a single coverslip. A: baseline AE activity in AT-1 cells in the presence of CO2-HCO buffer. B: inhibition of AE activity in the presence of 100 µM DIDS. C: AE activity in the presence of 50 µM dibenzamidostilbene disulfonic acid (DBDS). D: inhibition of AE activity in the nominal absence of CO2-HCO buffer.

AT-1 cell Cl/HCO exchange activity is regulated by paracrine factors. ATP and ANG II have been described as potent enhancers of Cl/HCO exchange in rat and cat cardiac tissue, respectively (7, 38, 40). We tested the effects of these agents on Cl/HCO exchange in AT-1 cells in the presence of 24 mM extracellular Na⁺. We found that 50 µM ATP increased bidirectional Cl/HCO exchange activity nearly threefold without any significant change in resting pH_i. The incremental Cl/HCO exchange activity was DIDS sensitive (Fig. 2; Table 2). Similarly, 500 nM ANG II activated bidirectional DIDS-sensitive Cl/HCO exchange without appreciable change in resting pH_i (Fig. 3; Table 3). In the complete absence of extracellular Na⁺, with initial pH_i = 7.11 ± 0.02 (n = 4), dpH_i/dt increased 3.5-fold from 0.054 ± 0.004 (n = 4) to 0.19 ± 0.01 pH units per minute in the presence of 50 µM ATP (n = 2; P < 0.02), and 3.3-fold to 0.18 ± 0.01 pH units per minute in the presence of 500 nM ANG II (n = 2; P < 0.02) again without change in initial pH_i. Stimulation of Cl/HCO exchange by ATP and by ANG II was indistinguishable also in the presence of 150 mM Na⁺ (n = 3, not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Regulation of Cl/HCO exchange activity in AT-1 cells by paracrine factors. pHi was recorded during extracellular Cl removal and restoration. Traces represent average pHi of at least 5 cells recorded simultaneously on single coverslips. A: AE activity was enhanced by addition of 50 µM extracellular ATP. B: AE activity was increased by addition of 500 nM ANG II. These studies were performed in the presence of 24 mM extracellular Na+.

View this table: [in this window] [in a new window]   Table 2.   ATP stimulation of Cl/HCO exchange in AT-1 cells

View larger version (44K): [in this window] [in a new window]   Fig. 3.   AE anion exchanger transcripts of AT-1 cells. A, left: RT-PCR phenotype of AE1 transcripts in cultured AT-1 cells (lanes 1, 3, 5, and 7), in mouse kidney (lanes 2 and 4), and in mouse spleen (lanes 6 and 8). The kidney AE1 transcript is not detectable in AT-1 cells (lanes 1 and 3). PCR fragment lengths are lane 2 (kAE1) 512 bp; lane 4 (kAE1) 764 bp; lanes 5 and 6 (eAE1) 533 bp; and lanes 7 and 8 (eAE1) 785 bp. A, right: Southern blot of AE1 PCR products transferred to nylon from gel of left panel and hybridized with a 32P-labeled internal AE1 oligonucleotide probe (AE1.E5F) common to all AE1 transcripts. B, left: RT-PCR phenotype of AE2 transcripts in cultured AT-1 cells (lanes 1, 3, and 5) and in mouse stomach (lanes 2, 4, and 6). PCR fragment lengths are lanes 1 and 2 (AE2a) 1,022 bp; lanes 3 and 4 (AE2b) 974 bp; lanes 5 and 6 (AE2c1) 482 bp and (AE2c2) 768 bp. B, right: Southern blot of AE2 PCR products transferred to nylon from gel of left panel and hybridized with a 32P-labeled internal AE2 oligonucleotide probe (AE2.E6R) common to all AE2 transcripts. A moderately abundant band in lane 5 (left) is not AE2 related and did not hybridize with the 32P-labeled probe. C, left: RT-PCR phenotype of AE3 transcripts in cultured AT-1 cells (lanes 1 and 3), in mouse brain (lane 2), and in mouse heart (lane 4). PCR fragment lengths are: lanes 1 and 2 (bAE3) 1,192 bp; and lanes 3 and 4 (cAE3) 790 bp. C, right: Southern blot of AE3 PCR products transferred to nylon from gel of left panel and hybridized with a 32P-labeled internal AE3 oligonucleotide probe (AE3.E10F) common to all AE3 transcripts.

View this table: [in this window] [in a new window]   Table 3.   ANG II stimulation of Cl/HCO exchange in AT-1 cells

AT-1 cells express AE transcripts. To determine which AE gene products are expressed in AT-1 cells and might mediate the above-described Cl/HCO exchange activity, AE transcripts were characterized by RT-PCR. PCR product identities were confirmed by Southern blotting and hybridization with a ³²P-labeled gene-specific internal oligonucleotide.

AT-1 cells express AE polypeptides. In Fig. 4A, antibody to the AE2 COOH-terminus peptide (aa 1,224-1,237) detected two polypeptides (lane 1): the upper is AE2a (and perhaps also AE2b); the lower may be AE2c forms. Recognition was specifically blocked by the peptide antigen (lane 2) but not by nonspecific peptide. Despite the presence of eAE1 mRNA in AT-1 cells (Fig. 3A), AE1 polypeptide was not detected by this anti-AE2 antibody known to be cross reactive with AE1 (47). However, the anti-AE2 antibody detected AE1 polypeptide in mouse heart membranes (Fig. 5, lane 1). Antibody to the AE3 COOH-terminus peptide (aa 1,216-1,227) detected two polypeptides likely corresponding to bAE3 and cAE3 (lane 3, top band and bottom band, respectively) in AT-1 cells, and a single polypeptide, likely cAE3, in mouse heart membranes (Fig. 5, lane 3).

View larger version (46K): [in this window] [in a new window]   Fig. 4.   Immunoblot detection of AE polypeptides in AT-1 cells. Triton X-100 lysates of whole AT-1 cells were fractionated by SDS-PAGE, transferred to nitrocellulose, and incubated with affinity-purified antibodies in the presence of 2 µg/ml excess peptide antigen (+) or excess irrelevant peptide (). A: lanes 1 and 2 were incubated with anti-mouse AE2 COOH-terminus aa 1,224-1,237, and lanes 3 and 4 were incubated with anti-human AE3 COOH-terminus aa 1,216-1,227, using 44 µg of protein per lane. B: each lane was incubated with bAE3-specific anti-human bAE3 aa 115-128, using 44 µg of protein per lane. C: lanes 1 and 2 were incubated with anti-human AE3 COOH-terminus aa 1,216-1,227, and lanes 3 and 4 were incubated with cAE3-specific anti-human cAE3 aa 42-53, using 100 µg of protein per lane.

View larger version (75K): [in this window] [in a new window]   Fig. 5.   Immunoblot detection of AE polypeptides in mouse heart postnuclear membranes. Mouse heart postnuclear membranes solubilized in SDS were fractionated by SDS-PAGE, transferred to nitrocellulose, and incubated with affinity-purified antibodies in the presence of 2 µg/ml excess peptide antigen (+) or irrelevant peptide (); 100 µg of protein were used per lane. Lanes 1 and 2 were incubated with anti-mouse AE2 COOH-terminus aa 1,224-1,237 (which cross reacts with the corresponding COOH-terminus residues of AE1). Lanes 3 and 4 were incubated with anti-human AE3 COOH-terminus aa 1,216-1,227. Lanes 5 and 6 were incubated with cAE3-specific anti-human cAE3 aa 42-53.

Immunocytochemical detection of AE polypeptides in AT-1 cells. All three antibodies to AE3 exhibit appropriate specificity of immunostaining in bAE3- and cAE3-transfected CHOP cells (55) and 293 cells, in sections of Epon-embedded rat heart, and (for cAE3-specific antibody) in isolated rat ventricular myocytes (results not shown). Figure 6 demonstrates the presence in AT-1 cells of both bAE3 and cAE3. Specific AE3 immunostaining is inhibited in the presence of respective peptide antigens (Fig. 6, B, D, and F) but not by irrelevant peptides (Fig. 6, A, C, and E). In addition, cAE3 was detected in cryosections of AT-1 tumor (Fig. 6, G and H).

View larger version (89K): [in this window] [in a new window]   Fig. 6.   Immunocytochemical detection of AE3 polypeptides in cultured AT-1 cells and AT-1 tumor. Paraformaldehyde-fixed AT-1 cells on coverslips (a-f) or 1-µm sections of periodate-lysine-paraformaldehyde-fixed LX-112-embedded AT-1 tumor (g, h) were incubated with affinity-purified anti-mouse AE3 COOH-terminus aa 1,216-1,227 (a, b); anti-human bAE3 aa 115-128 (c, d); or anti-human cAE3 aa 42-53 (e-h) in the presence of 10 µg/ml irrelevant peptide (a, c, e, g) or peptide antigen (b, d, f, h). Bar for a-f = 30 µm; bar for g and h = 20 µm.

View larger version (134K): [in this window] [in a new window]   Fig. 7.   Immunocytochemical detection of AE2 (A-D) and AE1 (E and F) polypeptides in AT-1 cells. Paraformaldehyde-fixed AT-1 cells on coverslips were incubated with the indicated affinity-purified antibodies in the presence of 10 µg/ml irrelevant peptide (A and E), AE1 COOH-terminus peptide (C and F), AE2 COOH-terminus peptide (B), or AE3 COOH-terminus peptide (D). Bar = 30 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study establishes that AT-1 cells in culture exhibit Cl/HCO exchange activity sensitive to DIDS and insensitive to DBDS. This Cl/HCO exchange activity is Na⁺ independent and can be activated by the paracrine factors ATP and ANG II. AT-1 cells in culture expressed mRNAs encoding bAE3, cAE3, AE2a, AE2b, AE2c1, AE2c2, and eAE1, but kAE1 mRNA was not detectable. These cells also express AE2, cAE3, and bAE3 polypeptides detectable by immunostaining and by immunoblot as well as an AE1 epitope detectable only by immunostaining.

AT-1 cells remain the only mammalian cardiocyte cell line that can be passaged in cell culture or in syngeneic host mice while maintaining a highly differentiated, spontaneously beating, contractile phenotype. Cultured AT-1 cells express adult cardiac-specific proteins such as -MHC, -cardiac actin, and connexin43 and exhibit the ultrastructural features of adult cardiocytes (12, 30, 46). They respond to muscarinic stimuli with a pertussis toxin-sensitive hyperpolarization (20), phosphoinositide hydrolysis, and adenyl cyclase inhibition (35). They express a rapidly activating delayed rectifier current (54) as well as multiple cardiac K⁺-channel subunit mRNAs including minK (53, 54). They increase intracellular free calcium and mitogen activated protein kinase activity in response to endothelin via protein kinase C and G proteins (18). In addition to maintaining a proliferative phenotype in cell culture, genetically engineered AT-1 cells can be grafted into syngeneic host myocardium (21). Cultured and engrafted AT-1 cells could aid the study of basic arrhythmogenic mechanisms, especially of atrial fibrillation because AT-1 cells are of atrial origin.

The understanding of pH_i and [Cl]_i regulation in AT-1 cells is important before their use in the study of arrhythmogenesis. Intracellular acidification can lead to intracellular calcium increase (36, 50) and with more extreme acidification to decreased gap-junction conductance (37). Both effects may provoke proarrhythmic consequences. The role of [Cl]_i dysregulation in arrhythmogenesis is less well understood. In isolated perfused hearts, substitution of nitrate for chloride in the perfusate reduced the incidence of arrhythmias after ischemia and reperfusion (42), and the chloride-transport AE blocker SITS had a similar antiarrhythmic effect (10). More recently, it has been shown that guinea pig papillary muscle exhibits a marked SITS-sensitive increase in [Cl]_i during ischemia and that ischemia-induced depolarization is also inhibited by SITS (29). ATP, a potent stimulator of Cl/HCO exchange in cardiac tissue (38), is abundantly released from cardiac cells during hypoxia (14). Direct application of ATP onto isolated rat cardiomyocytes evokes DIDS-suppressible delayed afterdepolarizations (45).

Our finding of DIDS-inhibitable Na⁺-independent Cl/HCO exchange activation in AT-1 cells treated with 50 µM ATP (Fig. 2A, Table 2) resembles similar results in adult rat cardiocytes (40). Our finding of DIDS-inhibitable Na⁺-independent Cl/HCO exchange activation in cells treated with 500 nM ANG II reproduces that reported in intact cat papillary muscle (7) (Fig. 2B, Table 3). ANG II has also been shown to stimulate two types of myocardial acid extruders: Na⁺/H⁺ antiport (34) and Na⁺/HCO symport (15, 22).

AE polypeptides have been detected previously in human and rat cardiac tissues. Anti-cAE3-specific antibody detected cAE3 protein in postnuclear membranes from human hearts with advanced cardiac failure, whereas an antibody against the common AE3 COOH terminus failed to detect bAE3 protein in the same preparations (55). Expression of AE1 and cAE3 proteins was detected by immunoblot in adult and neonatal rat cardiocyte preparations (25, 39). The current study documents the presence in mouse AT-1 cells of AE2a polypeptide, of another AE2 polypeptide of relative molecular weight consistent with AE2c, of cAE3 polypeptide, and, at lower abundance, bAE3 polypeptide. AE1 polypeptide was detected immunocytochemically but not by immunoblot. In contrast, immunoblots of mouse heart postnuclear membranes documented the presence of AE1 polypeptide as well as cAE3 polypeptide (Fig. 5).

We do not yet know whether the AE gene products described in this study constitute the full cadre of AT-1 cell Cl/HCO exchangers. Puceat and co-workers (40) introduced AE3 antisense oligonucleotides into neonatal rat cardiocytes by incubation, resulting in respective 70% and 60% decreases in cAE3 and bAE3 polypeptides but without change in ATP-stimulated Cl/HCO exchange activity. However, this latter transport activity was inhibited after microinjection of select anti-AE1 polyclonal antibodies. The authors concluded that a cardiac AE1 polypeptide was responsible for the purinergic receptor-stimulated Cl/HCO exchange in neonatal rat cardiocytes. Kudrycki and colleagues (26) detected two AE1 mRNA species by Northern analysis of whole adult rat heart mRNA, but the molecular identity of the shorter of the two hybridizing bands remains unknown. Unlike our current findings in AT-1 cells and whole adult heart of mouse and rat (not shown), Richards and co-workers (41) detected no eAE1 mRNA by RT-PCR in isolated rat cardiomyocytes. These authors proposed the presence of a novel AE1 transcript of yet-undefined 5'-exons which lacks exons 1-5.

Cultured AT-1 cells will provide a useful system in which to identify mediators of regulated Cl/HCO exchange activity within the AE gene family. This work will facilitate future experiments in which AT-1 cells are genetically modified and grafted into intact neonatal hearts.