INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE CORTICAL COLLECTING duct (CCD) of the kidney contains at least two types of intercalated cells that transport HCO₃ in opposite directions. HCO₃ reabsorption takes place in the -type intercalated cells, which express a Cl/HCO₃ exchanger on the basolateral membrane (for review, see Ref. 20). This exchanger is the product of the anion exchanger 1 (AE1) or band 3 gene. There is strong functional evidence indicating that HCO₃ secretion occurs in the -intercalated cells, via a Cl/HCO₃ exchanger located in the apical membrane (22), but the molecular identity of this exchanger remains to be established. Both the basolateral and the apical Cl/HCO₃ exchangers are regulated by acid/base balance. The expression of AE1 mRNA and protein is upregulated in acidosis (8, 13, 18), and functional studies indicate that the activity of the apical exchanger in -intercalated cell is downregulated by acidosis (22).

Recently, it was reported that AE2 mRNA and protein is present along the entire collecting duct system (4); however, its distribution among the different cell types of the collecting duct and its role in acid/base homeostasis are unknown. Several lines of indirect evidence raise the possibility that the apical exchanger of -intercalated cells might be a product of the AE2 gene. First, previous studies suggest that the inhibitor sensitivity and the kinetic properties of the apical Cl/HCO₃ exchanger are somewhat similar to those of AE2 (1, 15). In addition, AE2 is present in the apical membrane in another HCO₃-secreting epithelium, i.e., the rabbit ileum (7). If indeed AE2 would function as the apical Cl/HCO₃ exchanger of -intercalated cells, then one would expect AE2 expression to be high in these cells and to be upregulated in alkalosis to facilitate stimulated HCO₃ secretion. To test these hypotheses, we examined the effects of in vivo acid load or alkali load on AE2 mRNA expression in CCD cells and the distribution of AE2 mRNA among the three cell types of the collecting duct. We also tested the direct effects of changes in medium pH on AE2 expression in cultured CCD cells. Our results show that AE2 mRNA expression in CCD cells is significantly lower following acid load than alkali load both in vivo and in vitro, and this phenomenon can be observed in all three CCD cell subtypes.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Male New Zealand White rabbits, weighing 1.5-2.0 kg, were used. The animals were kept on standard diet and had access to water ad libitum. Alkali load was performed by intravenous infusion of 15 mmol/kg of NaHCO₃, 16-20 h before death. Metabolic acidosis was achieved by an intragastric load of 15 mmol/kg of NH₄Cl. To keep the amount of Na⁺ load in the two groups constant, this latter group also received 15 mmol/kg NaCl iv. For the last 12 h before the experiments, the rabbits were on restricted food intake (3 oz). Urine samples were taken from the bladder immediately before death for the determination of urinary pH. Plasma pH, PCO₂, and HCO₃ levels were measured using an automated blood-gas analyzer.

Cell isolation. CCD cells were isolated from the renal cortex by solid-phase immunoadsorption, using a monoclonal antibody against an ectoantigen on these cells, as described previously (9, 17). In some experiments, the immunodissected CCD cells were further fractionated into the three collecting duct cell types (i.e., - and -intercalated cells and principal cells) by fluorescence-activated cell sorting using cell-specific markers as described (10-12). Principal cells were identified with a FITC-conjugated antibody that reacts specifically with these cells (DT.17; Ref. 10); -intercalated cells were identified with peanut lectin agglutinin (PNA) coupled to phycoerythrin, as described in detail elsewhere (10-12), whereas -intercalated cells were operationally defined as the DT.17- and PNA-negative population. To aid in the discrimination between live and dead cells, CCD cell preparations were also stained with 4',6-diamidino-2-phenylindole (0.1 µg/ml), which is excluded from viable cells. The purity of the sorted cells was determined by immunocytochemistry, using other cell-specific markers as described (10-12) and was ~96% for -cells, ~94% for principal cells, and ~82% for -intercalated cells.

CCD cell culture. CCD cells were seeded on porous bottom dishes with a surface area of 0.6 cm² (Millicell HA; Millipore, Bedford, MA) at a saturating density of 6-8 × 10⁵ cells/cm². The filter cups were placed into the central wells of organ culture dishes and incubated with 0.4 ml of medium at the apical (inner) and 0.8 ml at the basolateral (outer) side. The cultures were grown in PC1 medium (Ventrex) supplemented with 5% decomplemented fetal bovine serum (Hyclone) and antibiotics (10) in an incubator containing 5% CO₂. After reaching confluence, the cultures were incubated in regular PC1 medium for 24 h. At this stage, ~40-45% of cells express markers specific for principal cells, ~60% is positive for PNA (-cell marker), ~25% for band 3, and 85% for an antibody reacting with both CCD and connecting tubule cells (ST.12; Ref. 9). The medium was changed at both sides of the monolayers to media with different pH values ranging from 6.1 to 8.1. Media of different pH were generated by the addition of NaOH or HCl, and appropriate amounts of NaCl were added to the control medium (pH 7.4) to keep osmolality constant. Cultures were maintained in media with different pH for another 24 h before RNA isolation.

RNA isolation and AE2 RT-PCR. Total RNA was isolated using TRI Reagent (Molecular Research Center), and RNA concentrations were calculated from the optical densities at 260 nm. cDNA was synthesized using 0.5-2 µg of total RNA as described (13). Sense and antisense PCR primers were designed based on the sequence of rabbit ileal AE2 (7). The sequences for primers were as follows: primer U1 (sense), 5' GGCGTGGAGCGGTTTGAAGA 3'; primer L1 (antisense), 5' TTGGTGGGGCAGCAGTGTAG 3'; primer U2 (sense), 5' GGAGCCACCCCCACCATTGA 3'; primer L2 (anti-sense), 5' CAGGAGACTGCGGAACGACA 3'. Primers U1 and L1 anneal to nucleotides 216-235 and 610-629, respectively, on the rabbit ileal AE2 and bracket a 414-bp sequence; primers U2 and L2 anneal at nucleotides 435-454 and 1180-1199, respectively. Reactions were performed in a 20-µl total volume containing 10 mM Tris · HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl₂, 75 µM dNTP, 200 nM of each primer, 0.1 U Taq polymerase (Perkin-Elmer), and 1 µCi [³²P]dCTP (NEN, 3,000 Ci/mmol) with four different amounts (1-30 ng) of template cDNA. Each sample was overlaid with 20 µl of Chill-Out (MJ Research) to prevent evaporation. After an initial 2-min denaturation at 96°C, PCR was carried out for 25 cycles with denaturation at 95°C for 1 min, annealing at 61°C for 1 min, and extension at 72°C for 1 min, then a final extension was carried out at 72°C for 8 min. The relative abundance of -actin mRNA in each CCD cell sample was determined using primers and conditions as described (13). cDNA samples derived from pairs of rabbits (one acidotic, the other alkalotic) were always amplified simultaneously.

To determine the copy number of AE2 mRNA, in several experiments known amounts of an internal standard cDNA were included in the PCR reaction. This internal standard was generated as follows. First, the 414-bp PCR product, amplified using primers U1 and L1, was digested with Ban I at 37°C for 3 h, which generated the expected 91-, 128-, and 196-bp fragments. These DNA fragments were separated on 4% NuSieve GTG agarose by electrophoresis, and fragments of 128 bp and 196 bp were religated at 16°C for 18 h. Amplification of the resulting DNA with primers U1 and L1 yielded a 324-bp PCR product. This amplicon was used as an internal standard using four different amounts (80, 20, 5, and 1.25 fg).

After amplification, 4 µl loading buffer was added to each sample, and 20 µl was run on a 6% polyacrylamide gel. Gels were dried, and the amount of radioactivity in the PCR products determined using a model 425 PhosphorImager (Molecular Dynamics).

To determine the relative abundance of AE2 mRNA, AE2 PCR product levels were compared with the levels of -actin PCR product obtained from the same sample. AE2 and -actin cDNA was amplified from 1-30 ng cDNA and 0.03-1 ng cDNA, respectively. Products were quantitated using a PhosphorImager, and slopes derived by linear regressions were compared.

Northern analysis. Northern blotting was carried out using standard protocols (3). In brief, 5-10 µg total RNA originating from CCD cells was fractionated on a 1.2% agarose gel containing 1.1% formaldehyde. RNA was transferred to a nylon membrane (0.45 µm; MSI, Westborough, MA) and probed with a gel-purified PCR fragment generated with primers U1 and L1 and labeled with ³²P using a random priming kit (DECAprime II; Ambion, Austin, TX). Prehybridization was performed at 42°C for 1 h in 5× SSC, 5× Denhardt's solution, 50% formamide, 100 µg/ml salmon sperm DNA, and 0.5% SDS. Hybridization was done using the same conditions as for prehybridization for 12 h. Two washes were carried out at room temperature for 15 min each, with 1× SSC and 0.1% SDS, followed by two washes with 0.25× SSC and 0.1% SDS. After a final wash at 45°C for 15 min with 0.1× SSC and 0.1% SDS, the blot was visualized using a PhosphorImager.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of metabolic acidosis and alkali load on expression of AE2 mRNA in rabbit CCD cells. First we established that AE2 mRNA is expressed in rabbit CCD cells, using RT-PCR and primers based on the published sequence of the rabbit ileal AE2 (7). Using primers U1 and L1, a product of the expected size, i.e., 414 bp, was amplified (Fig. 1A, lane 2). The identity of this PCR product as AE2 was verified by nested PCR, and by digestion with Ban I. In both cases products of the expected sizes were obtained (Fig. 1A, lanes 1 and 3). To examine the relative abundance of AE2 mRNA expression in CCD cells, PCR amplifications were carried out in the presence of a competitive control template, which is a DNA that is 94 bp shorter than the AE2 PCR product. These experiments revealed that AE2 mRNA is an abundant message in CCD cells with ~500 copies per cell.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   A: representative RT-PCR of AE2. RT-PCR was performed using primers U2 and L1 (lane 1) or U1 and L1 (lane 2) with 10 ng of cDNA derived from rabbit cortical collecting duct (CCD) cells. The expected size of the PCR product using these primers is 194 bp (U2 and L1) and 414 bp (U1 and L1). Position of molecular weight standards is shown on left. Lane 3: digestion of the 414-bp PCR product with Ban I. B: positions of the primers and the expected restriction fragments.

To examine whether changes in acid/base balance result in changes in the steady-state levels of AE2 mRNA, pairs of rabbits were acid or alkali loaded. Urinary pH averaged 5.4 ± 0.4 in acidotic vs. 8.3 ± 0.2 in alkali-loaded rabbits (P < 0.001). Plasma pH was 7.2 ± 0.1 in acidotic vs. 7.48 ± 0.02 in alkali-loaded rabbits (P < 0.05). Plasma HCO₃ concentration was 18.1 ± 2.7 and 32.1 ± 1.7 meq/l (P < 0.05), and PCO₂ was 31.1 ± 1.5 and 37.9 ± 1. 5 mmHg (P < 0.005), in acidotic and alkali-loaded rabbits, respectively.

AE2 mRNA levels were determined by quantitative RT-PCR using different amounts of cDNA derived from CCD cells of acid- or alkali-loaded rabbits. The amount of the AE2 PCR product increased linearly with the amount of starting CCD cDNA (Fig. 2, left). Similarly, PCR amplification was in the linear range for the internal standard, as shown in Fig. 2, right, where a standard amount (10 ng) of CCD cDNA and varying amounts (0-80 fg) of the AE2 internal standard were amplified simultaneously. AE2 mRNA levels were normalized for -actin mRNA levels, which, as found previously, are not affected by changes in acid/base balance (13). As illustrated in Fig. 3, the relative abundance of AE2 mRNA, calculated from the ratio of [³²P]dCTP incorporated into the 414-bp AE2 PCR product and into the 350-bp -actin PCR product, was significantly higher in alkali-loaded than in acidotic animals. The average ratio of AE2 mRNA levels in alkali-loaded vs. acid-loaded CCD cells, calculated from the paired treatments, was 3.7 ± 1.07-fold (Fig. 4).

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Relationship between [32P]dCTP incorporation into the AE2 PCR product and the amount of starting cDNA (A and C) or internal standard (B and D). Results of two typical AE2 PCR amplifications are shown. A and C: AE2 PCR was performed as described in METHODS, using 30-0.3 ng CCD cDNA as template, assuming 100% reverse transcription efficiency. Gel was subjected to autoradiography, and the corresponding bands were cut out from the gel and counted in a liquid scintillation counter. B and D: AE2 PCR was performed using 10 ng of CCD cDNA in each tube (top bands) and varying amounts (0-80 fg) of the internal standard (bottom bands). Amount of radioactivity in the experiment shown in B and D was calculated using a PhosphorImager.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Expression of AE2 mRNA in CCD cells isolated from acidotic and alkalotic rabbits. RT-PCR was performed with 4 different amounts (1-30 ng) of template cDNA originating from rabbits subjected to metabolic acidosis or alkalosis for 16-20 h as described in METHODS. Values shown are relative amounts that were corrected for the levels of -actin; n = 5 rabbits in each group. Data are means ± SE. * P = 0.005 using Wilcoxon's test (one-tailed).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Ratio of AE2 mRNA levels in acid- vs. base-loaded and acid-loaded vs. normal rabbits. Two series of experiments (acid vs. base load and acid load vs. control) were performed, CCD cells were isolated, and quantitative AE2 RT-PCR was carried out as described in METHODS; n = 5 for each experimental group. Values are ratios of actin-normalized AE2 mRNA levels in base-loaded vs. acid-loaded group (left) and in normal vs. acid-loaded group (right). * P < 0.05, with regard to a difference in ratio from 1.0 (Student's one-tailed t-test).

To determine whether the observed differences between acidotic and alkali-loaded animals were due a reduction in AE2 expression caused by acidosis or due to an increase caused by alkalosis, in a separate group of experiments, mRNA levels of AE2 were determined in CCD cells isolated from control (nontreated) rabbits as well as from rabbits with metabolic acidosis achieved by the same NH₄Cl loading as used in the previous set of experiments (except that these animals did not receive extra NaCl). The results are summarized on Fig. 4. These data demonstrate that AE2 mRNA levels are significantly decreased in metabolic acidosis compared with those of CCD cells from control rabbits. The average ratio of AE2 mRNA levels (normalized for -actin) in normal vs. acidotic CCDs was 1.93 ± 0.7 (n = 5). This is a smaller reduction (P < 0.05) than the difference (3.7-fold) observed in acidotic vs. alkali-loaded rabbits.

Effect of medium pH on the expression of AE2 mRNA in primary cultures of CCD cells. These experiments were aimed to determine whether changes in extracellular pH directly alter expression of AE2 mRNA. Primary cultures of rabbit CCD cells were incubated in media with different pH values both in the basolateral and the apical compartments, for 24 h, then RNA was isolated, and AE2 mRNA levels were determined by RT-PCR. As illustrated on Fig. 5, the levels of AE2 mRNA were much higher in alkaline medium (pH 7.9) than in normal or acidic medium (pH 6.6). In a separate group of cultures, the effect of extreme pH changes was studied. The increase in AE2 mRNA levels (normalized for -actin mRNA levels) was particularly high at pH 8.1 (4,632 ± 2,356% of the value obtained at pH 7.4; n = 3), whereas incubation in a medium with very low pH (pH 6.1) markedly reduced the level of AE2 mRNA compared with cultures grown at pH 7.4 (33.5 ± 13% of values obtained at pH 7.4; n = 3). These results suggest that the higher levels of AE2 mRNA seen in CCD cells from alkali-loaded vs. acidotic rabbits were probably brought about directly by changes in extracellular pH and not by secondary changes in hormonal levels or imbalances of other electrolytes.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effect of medium pH on AE2 mRNA expression in cultured CCD cells. CCD cells were isolated from untreated rabbits by immunoselection and grown on permeable membranes as described. For a 24-h period, the medium was changed at both the basolateral and apical sides of the monolayers to media with different pH values ranging from 6.6 to 7.9. After 24 h, RNA was isolated from each culture, and AE2 RT-PCR was performed. Data are expressed as percent of the value obtained for cultures maintained at pH 7.4; n = 3 for each group. * P < 0.05 and *** P < 0.001 (compared with pH 7.4). Statistical analysis was performed using Student's two-tailed t-test.

Northern analysis. The levels of expression of AE2 mRNA were also determined by Northern analysis using total RNA from CCD cells maintained on acidic, normal, and alkaline media for 24 h. As shown in Fig. 6, the amount of AE2 mRNA was much higher in CCD cells maintained in a medium with alkaline pH (7.8 and 8.1) than those at control pH (7.4) or at acidic pH (6.1 and 6.6).

View larger version (94K): [in this window] [in a new window]   Fig. 6.   Northern analysis of AE2 mRNA levels in CCD cells maintained at different pH. Primary cultures of CCD cells were incubated for 24 h in media with different pH. Total RNA was isolated, and Northern blotting was performed as described in METHODS. Results of two experiments are shown.

Distribution of AE2 mRNA in principal and intercalated cells. The above results were compatible with the idea that AE2 plays a role in bicarbonate secretion by the CCD. If AE2 were the apical Cl/HCO₃ exchanger, then one would expect its mRNA preferentially expressed in -intercalated cells. To examine this possibility, we isolated principal, -, and -intercalated cells by fluorescence-activated cell sorting and determined AE2 mRNA levels by RT-PCR. The results of these experiments, however, did not reveal a preferential expression of AE2 in -intercalated cells. As shown in Fig. 7, AE2 mRNA levels are comparable in all three subtypes of CCD cells. In addition, in all three cell types, AE2 mRNA levels were higher in alkali-loaded than in acidotic rabbits, although the differences did not reach statistical significance. It should be noted that both the comparable levels of AE2 mRNA among CCD cell subtypes, and the trend of AE2 mRNA levels to increase in alkali-loaded animals are in strong contrast with changes occurring in AE1 mRNA levels. AE1 is expressed almost exclusively in -intercalated cells, and its levels are markedly increased in acidosis (13).

View larger version (38K): [in this window] [in a new window]   Fig. 7.   Levels of AE2 mRNA in CCD cell types isolated from acid-loaded and base-loaded rabbits. RT-PCR was performed in triplicates with 10 ng of cDNA derived from principal cells (PC), -intercalated cells (-ICC), and -intercalated cells (-ICC), isolated by fluorescence-activated cell sorting. Values are means ± SE; n = 3 for PC, n = 5 for -intercalated cell, and n = 4 for -intercalated cell. Statistical analysis was performed using Student's two-tailed t-test.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Despite strong functional evidence for the existence of an apical Cl/HCO₃ exchanger in -intercalated cells (22), the molecular identity of this exchanger is still unknown. Although the possibility that the apical and basolateral exchangers are both encoded by the AE1 (band 3) gene was raised (23), the bulk of the evidence indicates that the two exchangers are separate proteins. First, no apical staining can be demonstrated in the CCD with AE1 antibodies that recognize the basolateral exchanger (1, 21). Second, there is little AE1 mRNA in sorted -intercalated cells compared with -intercalated cells (13). Third, the inhibitor sensitivity and the kinetic properties of the apical exchanger do not match those of AE1 (1, 19). On the other hand, these pharmacological data raised the possibility that another member of the AE family, AE2, could be a candidate for the apical exchanger of -intercalated cells. The observation that AE2 is present in the apical membrane in another HCO₃-secreting epithelium, i.e., the rabbit ileum (7), also supports this possibility. In addition, AE2 was shown to be expressed in rat CCD cells (4).

If AE2 were to function as the apical Cl/HCO₃ exchanger, then its expression is expected to be elevated following alkali load to facilitate bicarbonate secretion by the CCD. Certain findings of the present study seem to be compatible with this hypothesis. First, AE2 is a relatively abundant message in -intercalated cells, with ~500 copies/cell, and such high values are likely to be biologically relevant. Second, AE2 mRNA expression was found to be higher in CCD cells originating from alkali-loaded rabbits than in CCD cells from acidotic animals, and similar results were obtained in vitro in cultured CCD cells, suggesting that H⁺ (or bicarbonate) concentrations could directly regulate expression of AE2 mRNA in CCD cells. The direction of these changes in AE2 mRNA expression, both in vivo and in culture, is opposite to those observed in the expression of AE1. As we recently reported, AE1 mRNA and protein levels were significantly higher in CCD cells of acidotic than alkalotic rabbits (13).

If AE2 plays a role in bicarbonate secretion, then it is expected to be preferentially expressed in -intercalated cells. However, our results obtained with sorted cells do not support this idea. We found that AE2 mRNA levels are comparable in the three collecting duct cell types. Furthermore, our data also indicate that regulation of AE2 expression by acidosis and alkali load is not cell-type-specific within the CCD, as AE2 levels tended to be higher in all three cell types of alkali-loaded vs. acidotic animals (Fig. 7). These results suggest that AE2 might have a more general role in renal cells than functioning as the apical exchanger of -intercalated cells. For instance, AE2 could participate in the regulation of intracellular pH. The wide distribution of AE2 along the nephron (4) is also more compatible with a function in cellular homeostasis than in a specific transcellular transport event. Nevertheless, the observations that AE2 can occur with either basolateral (6, 14, 16) or apical (7) polarization in other tissues, coupled with the fact that some membrane components occur with opposite polarities in - vs. -intercalated cells (5), still leave the possibility open that polarization of AE2 is cell-type dependent. Immunohistochemical determination of the membrane localization of AE2 in -intercalated cells should yield definitive answer to this question.¹

In conclusion, the present data demonstrate that AE2 is a relatively abundant message in CCD cells, and its levels are significantly higher following both in vivo and in vitro alkali load than acid load. However, the findings that AE2 mRNA is not expressed in a cell-type-specific manner and that changes in acid/base balance have similar effects on each CCD cell subtype suggest that AE2 might serve a housekeeping function rather than being the apical anion exchanger of -intercalated cells. Thus the identity of the apical exchanger remains unknown, although preliminary data indicating that different isoforms of AE3 could be localized either basolaterally or apically in intercalated cells (2) raise the possibility that the apical anion exchanger may be an isoform of AE3.